• Adaptive Search Algorithm for Distribution Network Security Regions Considering Uncertainty of Wind and Photovoltaic Output

  LIU Bo, YUE Ting

  Available online:January 09, 2026  DOI: 10.7500/AEPS20250417003

  Abstract(89) HTML(18) PDF(1.58 M)( 123)

  Abstract:Existing algorithms for distribution systems security regions often suffer from low searching efficiency and redundant search paths when handling with uncertainties of wind and photovoltaic outputs. This is due to their reliance on single-distribution assumptions for scenario generation, which ignores spatiotemporal correlations. To address these issues, this paper proposes an adaptive search algorithm for distribution systems security region that accounts for the uncertainties of wind and photovoltaic output. Firstly, a joint distribution model of wind and photovoltaic outputs is constructed using kernel density estimation and the mixed Copula function. Spatiotemporally coupled scenarios are then generated through Monte Carlo sampling and reduced via K-means clustering. Secondly, an adaptive hybrid search strategy is proposed, combining adaptive interval segmentation with sensitivity matrix guidance in a two-step process of initial iteration and secondary correction. This strategy dynamically adjusts the search direction based on key constraints, significantly reducing redundant search paths and computation time while achieving robust boundary approximation. Finally, algorithm performance is optimized through a parallel computing acceleration mechanism. The effectiveness of the proposed method is validated on an IEEE 33-bus distribution system.

• Hierarchical Coordinated Optimization Method for Urban Transportation Network-Fast Charging Station-Distribution Network

  JIANG Tao, CHI Dashuo, WU Chenghao, JIN Xiaolong, LI Xue, MU Yunfei

  Available online:January 09, 2026  DOI: 10.7500/AEPS20250218002

  Abstract(43) HTML(12) PDF(2.71 M)( 48)

  Abstract:With the continuous advancement of transportation electrification, the coupling degree between urban transportation network (UTN) and distribution network (DN) is gradually increasing, which makes the calculation of traffic flow and power flow more complex. In order to comprehensively describe the impact of electric vehicles on UTN and DN, this paper proposes a dispatching strategy of hierarchical coordinated optimization for UTN-FCS-DN based on dynamic traffic flow and power flow to realize the economical and flexible operation of UTN, fast charging station (FCS) and DN. Firstly, the power flow optimization is carried out in the DN optimization layer with the goal of minimizing the operation cost, the traffic flow assignment is carried out based on the mixed user balance in the UTN optimization layer, and the electric vehicles are dispatched with the goal of maximizing the operator’s revenue in the FCS optimization layer. Then, in order to simplify the model, the traffic flow allocation problem is transformed into the constraints of the FCS revenue model and linearized, and then the alternating direction multiplier method is used to solve the above problem to protect the information privacy of different subjects. Finally, the effectiveness of the proposed model is verified by the system coupled with the IEEE 33-bus distribution network and the Nguyen-Dupuis transportation network.

• Diffusion-Policy Learning for Multi-scenario Generalized Boosting of Integrated Energy System Scheduling

  DOU Jiaming, WANG Xiaojun, LIU Zhao, SI Fangyuan, HAN Yi

  Available online:January 09, 2026  DOI: 10.7500/AEPS20250109005

  Abstract(26) HTML(8) PDF(2.05 M)( 40)

  Abstract:Reinforcement learning (RL) plays a pivotal role in advancing the digitalization and intelligence of integrated energy system (IES) dispatch due to its capabilities for high-dimensional state perception and rapid decision-making. However, conventional RL is constrained in multi-scenario IES dispatch because of its limited model expressiveness and poor cross-scenario generalization. To address this, an RL dispatch method based on denoising diffusion-policy learning is proposed. First, the developmental concepts and fundamental principles of diffusion generative artificial intelligence models are introduced. Next, a diffusion-policy dispatch learning framework is constructed, and a diffusion learning layer and an agent optimization layer are designed. Finally, the generalization performance is analyzed through extensive experiments considering key diffusion parameters, sample sizes, and sample distributions. Meanwhile, the post-hoc interpretability mining of the diffusion-policy is conducted via t-distributed stochastic neighbor embedding. The results show that, compared with classic RL methods such as deep deterministic policy gradient used in IES dispatch, the proposed method retains considerable decision generalization under drastic source-load fluctuations and changing constraints while incurring only a minor degradation in optimization performance.

• Time-of-Use Traceback of Carbon Emissions for Electricity Consumers Based on Contract Paths

  ZHANG Xian, XIE Kai, ZHANG Xiaojing, XU Chuanbo, GUO Qinglei, LIU Dunnnan

  Available online:January 07, 2026  DOI: 10.7500/AEPS20250718004

  Abstract(57) HTML(24) PDF(971.04 K)( 54)

  Abstract:For the carbon emission peak and carbon neutrality, accurate accounting of indirect carbon emissions associated with electricity consumption of electricity consumers is of significant importance. Addressing the problems in the carbon emission accounting on the electricity consumer side such as insufficient traceback accuracy and ambiguous allocation of environmental benefits, a traceback method for carbon emissions based on contract paths is proposed. First, by constructing the virtual transmission paths for electricity trading contracts, a direct correlation mechanism between the electricity consumption behaviors of electricity consumers and the carbon attributes on power generation sides is established. Second, a coordinated decomposition model for contracted electricity quantities of multiple power sources is constructed to achieve the temporal matching between power output of renewable energy and load of electricity consumers. A carbon responsibility attribution mechanism based on trading types is further established, forming a complete accounting method system for carbon emissions on the electricity consumer side. On this basis, the differences in accuracy and fairness between the proposed method and the conventional accounting method are compared and analyzed. Finally, a case study using real trading data is conducted, and the results show that the proposed method can effectively identify the power supply structure characteristics of power sources for electricity consumers, and accurately quantify the carbon emission reduction contribution of green electricity.

• Grid-forming Electricity-Hydrogen Energy Storage Planning for Renewable Energy Base Considering Frequency Regulation Parameter Optimization

  ZHANG Yumin, GAO Yehao, YE Pingfeng, JI Xingquan, YANG Ming, WANG Chengfu

  Available online:January 07, 2026  DOI: 10.7500/AEPS20250513003

  Abstract(30) HTML(18) PDF(2.26 M)( 48)

  Abstract:Regarding the frequency security issues caused by lack of conventional power support in the remote Gobi Desert and other arid areas renewable energy bases, this paper proposes an electricity-hydrogen energy storage planning model considering virtual inertia response and frequency regulation parameter optimization. It leverages the complementary advantages of electric energy storage’s fast inertia response and hydrogen energy storage’s long-term large-scale energy time-shifting. At the planning level, a grid-forming electricity-hydrogen energy storage configuration model under DC transmission scenarios is established. By analyzing the virtual inertia characteristics of hydrogen energy storage systems and electric energy storage systems, a multi-device collaborative inertia support model is formulated. At the operational level, for three safety indicators—maximum frequency deviation, rate of frequency change (RoCoF), and quasi-steady-state deviation—the virtual inertia time constant and primary frequency regulation coefficient of grid-forming control are dynamically optimized to maximize the frequency support capability of energy storage systems under different operational scenarios. Simulation studies based on a renewable energy base in Northwest China demonstrate that the proposed model improves frequency security indicators and reduces energy storage configuration costs through coordinated electricity-hydrogen energy storage frequency regulation and adaptive parameter adjustment.

• Dynamic Partitioning Method of Power Transmission System for Over-frequency Generator Tripping Control

  YANG Chenyang, WANG Haiyun, KANG Pengpeng, ZHANG Hongli, YANG Guixing

  Available online:January 06, 2026  DOI: 10.7500/AEPS20250522010

  Abstract(40) HTML(14) PDF(1.17 M)( 42)

  Abstract:The integration of high-proportion renewable energy sources leads to changes in the frequency characteristics of power systems. As a result, transient loss-of-synchronism assessment has become crucial for system security, requiring accurate calculation of transient stability margins and the implementation of over-frequency generator tripping measures to mitigate the risk of system splitting induced by generator desynchronization. However, most existing studies consider generator tripping from a system-wide perspective and lack regional partitioning for refined generator tripping. On this basis, this paper proposes a dynamic cluster partitioning strategy based on multi-dimensional indicators. By using the analytic hierarchy process, electrical and dynamic characteristics are integrated to form generator clusters. Combined with a flexible partitioning method, membership functions are constructed to enable re-partitioning of boundary nodes. Furthermore, a fast assessment model of transient stability regions is developed based on the extended equal area criterion to analyze the transient stability margins of the partitioned regions. Simulation analysis is carried out in a power transmission scenario in Northwest China. The results show that the proposed cluster partitioning method is rational and effective, demonstrating its engineering applicability in complex power transmission corridor scenarios.

• Security Region of Voltage Source Converter Parameters Considering Phase-locked Stability During Low Voltage Ride-through and Transient Overvoltage

  LUO Shujie, XIA Deming, CHEN Lei, WANG Kefei, LI Zhonghui, LIU Qingchen

  Available online:January 04, 2026  DOI: 10.7500/AEPS20250317005

  Abstract(42) HTML(18) PDF(1.46 M)( 52)

  Abstract:Short-circuit faults of the power grid cause voltage source converters (VSCs) to enter the low voltage ride-through (LVRT) state. During this process, the stability of the phase-locked loop (PLL) and the transient overvoltage after recovery are two critical challenges to security and stability. The PLL parameters and the current reference values during the LVRT process are key influencing parameters. This paper proposes a method for calculating the parameter security region that comprehensively considers PLL stability during the LVRT process and transient overvoltage after recovery. Firstly, to address the issue of deviation between the initial point and the equilibrium point affecting the effectiveness of traditional small-disturbance stability criteria, a phase-locked stability analytical analysis method and a parameter security region considering initial point deviation are proposed. Secondly, transient overvoltage constraints are specified, and the parameter security region under transient overvoltage constraints is calculated based on the analytical expression of the terminal voltage throughout the entire process. Finally, the intersection is found to obtain the parameter security region that comprehensively considers PLL stability and transient overvoltage. The effectiveness of the derived parameter security region is validated by MATLAB/Simulink simulation results.

• Load-Storage Coordinated Planning Considering Uncertainties of Distributed Photovoltaic Coupling

  SHEN Zikang, WANG Yunchu, YU Shunjiang, WANG Xuchao, YANG Li, LIN Zhenzhi

  Available online:December 31, 2025  DOI: 10.7500/AEPS20250625005

  Abstract(67) HTML(40) PDF(2.21 M)( 68)

  Abstract:With the large-scale integration of distributed photovoltaics (DPVs), distribution networks face increasing uncertainty and challenges in renewable energy accommodation. Rational planning of load-storage resources is crucial for enhancing system renewable energy accommodation capacity and flexibility. Therefore, a load-storage coordinated planning method considering uncertainty of DPV coupling is proposed. Firstly, the Copula function is employed to characterize the spatiotemporal coupling characteristics of DPV output, and a Copula-Wasserstein distance-based ambiguity set construction method is proposed to describe the coupled uncertainty among multiple DPVs. Then, considering the complementary characteristics between physical energy storage and demand response, a load-storage coordinated distributionally robust planning model considering uncertainty of DPV coupling is formulated, and a solution method combining McCormick relaxation and strong duality theory is proposed. Finally, two actual 10kV distribution networks in a certain area of China are used for cases. The results demonstrate that the ambiguity set construction method based on Copula-Wasserstein distance can effectively exclude probability distributions that violate coupling characteristics between DPVs, significantly reducing system planning and operational costs. Moreover, the coordinated load-storage planning can effectively solve problems of local DPV accommodation and regional overloading.

More+

Volume 50,2026 Issue 1

>Review·Perspective

• Development Situation and Key Technologies of Source-Grid-Load-Storage Integrated Projects

  MENG Xuyao, LIU Jiajia, WEN Fushuan, LIU Fuyan, YING Qi

  2026,50(1):1-19, DOI: 10.7500/AEPS20250121002

  Abstract(58) HTML(123) PDF(1.53 M)( 37)

  Abstract:The construction and operation of the source-grid-load-storage integrated projects (hereinafter referred to as the integrated projects) can enhance local renewable-energy consumption and clean-electricity share, increase power-market flexibility, and propel the development of a new power system that supports the goals of carbon emission peak and carbon neutrality. At present, the progress of integrated projects is unsatisfactory, with slow advancement and difficult deployment being prominent issues. This paper first summarizes the latest policy promotion and application-implementation status of integrated projects. It then reviews the current state of technology deployment and market participation. Next, it identifies the key technical challenges encountered during project rollout. Finally, it outlines future development trends and research directions.

>Key Technologies of Smart Distribution Network Based on Flexible Interconnection

• Analysis on Flexible Interconnection Operation Characteristics of Distribution Network Based on Modified Sen Transformer

  JI Yueyang, LI Peng, JI Haoran, CHEN Ran, ZENG Ping, WANG Chengshan

  2026,50(1):20-28, DOI: 10.7500/AEPS20241210002

  Abstract(26) HTML(52) PDF(806.80 K)( 12)

  Abstract:Flexible interconnection is an effective solution to enhance the operation flexibility of distribution networks, with flexible interconnection devices (FIDs) serving as the key component for enabling power sharing and flexible operation between feeders. However, the cost of fully-controlled power electronic-based FIDs is relatively high, making them suitable for deployment only at critical interconnection points within the distribution network. To achieve comprehensive flexibility across the entire distribution network and reduce the cost of flexible interconnection between feeders, this paper analyzes and describes the operation characteristics of distribution network flexible interconnection based on a modified Sen transformer (MST). First, the power regulation mechanism of the MST is elaborated, and the node voltage constraints for flexible interconnection between feeders via the MST are derived. Then, the regulation range of active power transfer and reactive power support provided by the MST is characterized, and the impact of node voltage variations at the connected terminals on the operation boundaries of the MST is analyzed. Finally, the regulation performance of the MST is validated by using parameters from an actual 10.5 kV distribution network in Tianjin of China. The results demonstrate that, under conditions satisfying flexible interconnection requirements between feeders, the MST enables a wider range of power transfer regulation with a smaller capacity of power electronic devices, thereby supporting flexible control of the distribution network.

• Economically Efficient Optimization Layout and Scheduling Strategy for Honeycomb Distribution Networks

  HAN Haiteng, XU Yiteng, ZU Guoqiang, CAO Shuyu, WU Chen, ZANG Haixiang

  2026,50(1):29-38, DOI: 10.7500/AEPS20241128002

  Abstract(34) HTML(284) PDF(819.18 K)( 19)

  Abstract:With the advancement of the national “carbon emission peak and carbon neutrality” goal, a significant number of distributed energy sources, hybrid energy storage systems are being integrated into the distribution network. This integration has enriched the composition of distribution network elements, thereby making the exploration of new network architectures and functionalities increasingly relevant. Among these, the honeycomb distribution network has emerged as a promising new architecture due to its efficient energy dispatch and the synergistic operation of multiple microgrids. This paper proposes a streamlined hexagonal honeycomb distribution network (SHHDN) architecture optimized through the strategic placement of intelligent base stations (IBSs). Firstly, a structural model of SHHDN based on the non-Cartesian coordinate system (NCCS) is developed to determine the optimal siting of IBSs and their internal energy storage capacities. Secondly, an SHHDN dispatch model incorporating multiple IBSs is constructed based on the optimized IBS system layout. Finally, the proposed model is validated through simulation on a modified IEEE 33-bus system and a modified 97-bus system. The results demonstrate that the proposed approach effectively smooths load fluctuations, reduces network losses, and facilitates renewable energy source (RES) integration with minimal IBS investment. This study provides both theoretical support and practical guidance for the flexible planning and low-carbon operation of next-generation distribution networks.

• Robust Reinforcement Learning-based Dynamic Reconfiguration Method for Distribution Networks Considering Line Electro-Thermal Coupling Characteristics

  GAO Haishu, SUN Kaining, HUANG Gang, ZHANG Feng

  2026,50(1):39-50, DOI: 10.7500/AEPS20250506002

  Abstract(35) HTML(30) PDF(1.09 M)( 18)

  Abstract:As the penetration of photovoltaics (PV) in distribution networks is increasing, the dynamic reconfiguration method based on the coordination of the soft open point (SOP) with sectionalizing and tie switches has become a key technological pathway for ensuring safe and stable operation of distribution networks. However, the electro-thermal coupling characteristics of lines are often overlooked during dynamic reconfiguration, leading to errors in resistance calculations that can skew reconfiguration results and compromise the safe and economic operation of the power grid. To address this, a robust reinforcement learning-based dynamic reconfiguration method considering the line electro-thermal coupling characteristics is proposed for SOP-equipped distribution networks. Firstly, to mitigate system modeling errors caused by the assumption of constant line resistance, a dynamic reconfiguration model is developed for SOP-equipped distribution networks considering the electro-thermal coupling of lines. Secondly, the original optimization problem is transformed into a Markov decision process, and a reward function based on first-order affine polynomials is constructed to evaluate the operation risks arising from PV and load fluctuations, thereby enhancing decision robustness. On this basis, a robust deep reinforcement learning algorithm is proposed, which realizes effective learning of robust optimization strategies through confidence-based action selection and robust updating mechanism of action network parameters. Finally, simulation tests on IEEE 34-bus and 123-bus systems show that the proposed method can better capture the dynamic variation of line resistance compared with traditional modeling approaches, improve decision reliability, and effectively reduce operation cost and risk under short-term PV generation and load fluctuations.

• Topology and Control Strategy of Soft Open Point Integrated with Photovoltaic and Energy Storage for Medium-voltage Grid Connection

  WANG Jingyue, ZHU Di, LIU Chuang, SUN Heyang

  2026,50(1):51-60, DOI: 10.7500/AEPS20241225001

  Abstract(25) HTML(35) PDF(5.17 M)( 18)

  Abstract:Aiming at the flexible interconnection among renewable energy grid-connected nodes in medium-voltage distribution networks, a soft open point (SOP) integrated with photovoltaic and energy storage for medium-voltage grid connection is proposed. This SOP has the capability to efficiently aggregate photovoltaic and energy storage resources and to improve the uneven distribution of power flow in distribution networks. Firstly, the topological structure of the photovoltaic-battery-hybrid cascaded SOP (PVBH-SOP) is proposed and the topology and modulation strategy of its single-stage isolated cascaded sub-modules are presented. Secondly, an average circuit model of the PVBH-SOP is established to analyze the control principle of its power flow. Then, three operation modes of the PVBH-SOP are described, along with their corresponding control strategies. Finally, a 10 kV/2 MW simulation platform for the PVBH-SOP system is established in MATLAB/Simulink to verify the effectiveness of the proposed PVBH-SOP topology and control strategy.

>Basic Research

• Hierarchical Early-warning for Multi-period Balance Risk in Look-ahead Dispatch of Power Grid Considering Uncertainty of Wind Power

  MU Zeyu, CHEN Siyuan, XU Peidong, SI Ruiqi, ZHANG Jun, XU Jian, HUANG He, CHEN Yiping

  2026,50(1):61-73, DOI: 10.7500/AEPS20241127005

  Abstract(31) HTML(55) PDF(1.83 M)( 17)

  Abstract:As wind power increasingly becomes the mainstay of electricity supply, intra-day prediction deviations stemming from wind power uncertainty pose severe challenges to power balance. Look-ahead dispatch serves as an effective bridge between day-ahead dispatch and intra-day automatic generation control. To this end, a hierarchical early-warning method for balance risks in look-ahead dispatch of power grid considering the uncertainty of wind power is proposed. Furthermore, a sample enhancement method based on point estimation and matrix normal distribution theory is introduced. The parametric data generation mode is used to expand minority-class samples and enhance the model prediction accuracy and scenario coverage. Finally, based on the two key indicators of net load deviation and adjustable resource capacity, an early-warning model for multi-period balance risk in look-ahead dispatch is constructed to provide hierarchical early-warning for the balance risk of the system in the next few hours. Simulation on the IEEE 118-bus test system verify that the proposed method can rapidly and accurately provide early-warning of balance risks in the next few hours, and the hierarchical warnings furnish valuable reference information for look-ahead dispatch.

• Hybrid Data-model-driven Fast Unit Commitment Solution Method for Credibility Enhancement

  WANG Wenye, FENG Chuan, GUAN Yuxiang, MA Wenhao, CHE Liang

  2026,50(1):74-85, DOI: 10.7500/AEPS20250324002

  Abstract(27) HTML(23) PDF(904.83 K)( 16)

  Abstract:With the increasing complexity of the power source structure and grid topology in new power systems, the number of system nodes and units continues to rise. The solution for the security constrained unit commitment (SCUC) model using traditional optimization methods faces problems such as the curse of dimensionality and slow calculation speed. Although the data-driven decision methods can quickly solve the SCUC model, the lack of interpretability makes the decision results unusable. To address these issues, a hybrid data-model-driven fast unit commitment solution method for the credibility enhancement is proposed. First, an SCUC solution model based on deep reinforcement learning (DRL) is constructed to achieve fast pre-solution for the unit start-up/shut-down decision results. Then, by comprehensively considering DRL behavior-level interpretability indicators and strategy-level interpretability indicators, a credibility evaluation system for the start-up/shut-down decisions is constructed to identify high-credibility unit start-up/shut-down results, and enhance the interpretability of decision results. Finally, the hybrid data-model-driven SCUC is constructed to achieve fast model solution and optimize and adjust the low-credibility decision results. The simulation verification based on a 748-bus system of a provincial power grid shows that the proposed method achieves fast SCUC solution on the premise of enhancing the interpretability of unit start-up/shut-down decision results.

• Regulation Demand Decomposition Strategy of Automatic Generation Control for Dynamically Matching Real-time Operating Conditions of Power Sources

  XIANG Mingxu, LI Weihao, YANG Zhifang, YIN Xujia, ZHOU Ce, WANG Haolin

  2026,50(1):86-96, DOI: 10.7500/AEPS20250522007

  Abstract(29) HTML(27) PDF(752.01 K)( 16)

  Abstract:Maintaining the system frequency quality is a key challenge to the construction of new power systems. The strong source-load fluctuations in the power system with high proportion of renewable energy have brought a higher requirement for the rationality of the total command allocation strategy of the automatic generation control (AGC). Existing command allocation strategies do not consider the response performance differences of regulation sources to various AGC commands under various operating conditions. The mismatch between AGC commands and power source operating conditions is prone to causing the disqualification response performance. To address this issue, an AGC regulation demand decomposition strategy for dynamically matching the real-time operating conditions of power sources is proposed. First, a data-driven qualification perception model of power source regulation performance is established. Whether the regulation performance of the power source is able to meet the assessment standard can be perceived under given operating conditions and received AGC commands. Based on this, the command range that the power source can effectively respond to is reversely inferred under the given operating conditions. Then, for the commonly used advanced allocation and real-time allocation, the AGC allocation strategies embedded with the command matching range are proposed to improve the matching degree between the AGC command allocation results and the power source operating conditions. Finally, the effectiveness of the proposed method is validated based on the actual data of gas turbines and hydropower units in a provincial power grid in China.

• Emergency Frequency Control Method for Power Systems Based on Knowledge-embedded Deep Reinforcement Learning

  LI Jiaxu, WU Junyong, SHI Fashun, ZHANG Zhenyuan, LI Lusu

  2026,50(1):97-107, DOI: 10.7500/AEPS20250611008

  Abstract(33) HTML(32) PDF(2.78 M)( 23)

  Abstract:The rapid development of new power systems has exacerbated frequency security challenges, making emergency control crucial for restoring stability during faults. This paper proposes a power system emergency frequency control method based on knowledge-embedded deep reinforcement learning. First, the emergency frequency control problem is formulated as a Markov decision process, with a simulation system serving as the reinforcement learning environment, and a deep reinforcement learning (DRL) agent is constructed based on the deep deterministic policy gradient (DDPG) algorithm. Furthermore, theoretical knowledge guides the action space optimization, integrating both over-frequency generator tripping and under-frequency load shedding scenarios. Finally, the proposed method is validated on the IEEE 39-bus system, demonstrating that the DRL agent can generate effective emergency frequency control strategies to ensure system security, and the knowledge-embedded technique enhances training stability and significantly improves policy learning efficiency and decision quality.

• Frequency Response Modeling Method for Large-scale Grid-connected Photovoltaic System Considering Low Voltage Ride-through Characteristic

  LIU Xianchao, YANG Di, LI Xue, JIANG Tao, LI Guoqing

  2026,50(1):108-117, DOI: 10.7500/AEPS20241220010

  Abstract(23) HTML(37) PDF(1.75 M)( 18)

  Abstract:Photovoltaic (PV) units equipped with low voltage ride-through (LVRT) control adopt a reactive power priority principle during significant voltage drops, resulting in variations in active power output and subsequent participation in system frequency response. To address this, a modeling method for system frequency response is proposed, taking into account the LVRT characteristics of large-scale PV. First, based on the traditional frequency response model, the impact of large-scale integration of renewable energy sources such as PV on system inertia and primary frequency regulation is analyzed. This analysis characterizes the active power-frequency response pathway triggered by control switching behavior at PV stations during significant voltage drops. Subsequently, considering the numerous and diverse characteristics of units within a PV station, three typical LVRT active power characteristics are classified, and their dividing boundaries are established. The equivalent unit parameters are rapidly obtained by using a capacity-weighted approach. A two-stage slope method is employed to approximate the nonlinear active power characteristics of clusters, thereby establishing an equivalent active time-domain expression for the entire LVRT process of PV power stations. Furthermore, the active power variation due to LVRT at PV power stations is introduced as a disturbance excitation into the traditional frequency response model. Based on the temporal characteristics of the active power variation due to LVRT, it is segmented and represented as step changes, slopes, and other forms to obtain the analytical solution of frequency response. The impact of the LVRT characteristics on indicators such as the maximum rate of frequency change, the maximum frequency deviation, and the quasi-steady-state frequency deviation is analyzed. Finally, the accuracy and effectiveness of the proposed modeling method are validated in a 5-bus test system incorporating PV power stations.

• Virtual Synchronous Generator Based Privacy Protection Strategy for Self-triggered Power Economic Allocation in AC Microgrids

  YANG Jun, FAN Yantong, QIN Jie

  2026,50(1):118-129, DOI: 10.7500/AEPS20250408001

  Abstract(23) HTML(17) PDF(2.17 M)( 14)

  Abstract:Aiming at the problem of reduced inertia in microgrids due to the high proportion of renewable energy integration, this paper focuses on the coordinated optimization of power economic allocation and data security in AC microgrids controlled by virtual synchronous generators. Firstly, an improved incremental cost consistency algorithm integrated with model predictive control is proposed to calculate the power imbalance value, breaking the limitation of traditional consistency algorithms where the leader node needs to calculate global information. Secondly, a two-dimensional coordinated self-triggered privacy protection strategy is proposed, and the differential privacy mechanism is adopted to inject noise into the interaction information to solve the data leakage problem. At the same time, the event-driven self-triggered mechanism is combined to reduce the computational burden and lower communication and encryption costs. Finally, the 4-machine 7-bus system is used for simulation verification. The results show that the proposed self-triggered privacy protection strategy significantly reduces communication and computing resources and enhances the robustness and privacy security of microgrid control.

• Risk Assessment of Cascading Failures in Electricity-Gas Interconnection System Considering Multi-states of Components

  NAN Lu, MA Yiyang, HE Chuan

  2026,50(1):130-142, DOI: 10.7500/AEPS20250407002

  Abstract(28) HTML(27) PDF(926.55 K)( 17)

  Abstract:With the deepening coupling between power and natural-gas systems, cross-infrastructure cascading failures have become a critical threat to the secure and stable operation of the interconnection system To address this challenge, this paper proposes a risk assessment method of cascading failures in electricity-gas interconnection system considering multi-states of components. Firstly, a cascading-failure simulation model is formulated with the objective of minimizing the total cost of load shedding, transmission-line outages, and gas-pipeline ruptures while accounting for component multi-state behavior. Secondly, a multi-state component failure probability model is established to calculate the accident probability. Finally, a risk assessment method of cascading failures in electricity-gas interconnection system considering the severity of interconnected system load loss is proposed. The effectiveness of the proposed model and method is demonstrated on a test system that combines the IEEE 24-bus power system with the 20-node Belgian gas system.

• Cascading Failure Blocking Method Based on Coordination of Multiple Types of Source-Grid-Load-Storage Resources

  KANG Haipeng, ZHANG Heng, ZHANG Shenxi, CHENG Haozhong, LI Changcheng

  2026,50(1):143-156, DOI: 10.7500/AEPS20241009005

  Abstract(25) HTML(53) PDF(1015.65 K)( 13)

  Abstract:Conducting research on cascading failure defense is of great significance for ensuring the safe and stable operation of power systems. This paper proposes a cascading failure blocking method based on the coordination of multiple types of source-grid-load-storage resources, aiming to address the problems of high economic cost and blackout risk caused by the one-sided consideration of factors in the fault stage selection method for implementing existing blocking strategies and the insufficient coordination of multiple types of resources. First, considering the characteristics of phased load loss during the cascading evolution of faults, an improved fault chain risk assessment index is established. Then, considering the system response capability to fault cascading propagation and the degree of risk increase at different fault stages, a fault stage selection method for implementing blocking strategies based on the time-risk integrated importance is constructed. Finally, considering the mechanism and complementary potential of resources on the source, grid, load, and storage sides in cascading failure blocking, a coordinated cascading failure blocking model is proposed that integrates the characteristics of multi-type power output regulation on the source side, the flexible adjustment capability of the transmission network topology on the grid side, the graded power supply guarantee requirements for loads of various importance on the load side, and the rapid response characteristics of energy storage equipment. Simulation results show that applying the proposed blocking method at the selected fault stages can effectively reduce the blackout risk caused by cascading failures and significantly improve the economy of the blocking strategy and the safety of the power system.

• Fault Current Suppression Strategy for High-voltage DC Based on Current Source Converter with Enhanced Fundamental Frequency Modulation

  FENG Dingteng, XIONG Xiaoling, YAO Chenhao, ZHOU Zihan, WANG Shengwei, ZHAO Chengyong

  2026,50(1):157-167, DOI: 10.7500/AEPS20250224001

  Abstract(23) HTML(22) PDF(916.98 K)( 19)

  Abstract:The current source converter with enhanced fundamental frequency modulation (EFFM-CSC), which utilizes fully controllable integrated gate-commutated thyristor (IGCT) devices, has garnered widespread attention. Its advantages include simple topology, compact size and low weight, favorable harmonic characteristics, and flexible adjustment of active/reactive power. However, high-voltage direct current (HVDC) systems based on EFFM-CSC also face the severe fault of DC line short-circuits, and there is currently a lack of systematic quantitative analysis and research. Therefore, a computational circuit model in the complex frequency domain is established based on the topology of the EFFM-CSC converter station and the switching states at the short-circuit instant. Subsequently, the equivalent RLC parameters are derived according to the energy storage equivalence principle, and a fault equivalent model of the EFFM-CSC under DC short-circuit conditions is constructed. An analytical calculation formula for the DC fault current is derived, and the effects of system parameters and fault timing on the fault current are discussed. Based on the DC short-circuit fault characteristics of EFFM-CSC, a fault identification criterion and a fault current suppression strategy are proposed, and the control parameters of the suppression strategy are tuned. Finally, the correctness of the proposed fault current calculation method and fault suppression strategy is verified through simulation analysis.

• Optimal Planning of Power Collection System for Offshore Wind Farm Based on Large-scale Mixed-integer Programming

  GAO Wenhao, SHEN Xinwei, DING Xiaochi, JI Xinzhe, LI Hongke, WANG Ke, YANG Wenbin

  2026,50(1):168-177, DOI: 10.7500/AEPS20241224007

  Abstract(27) HTML(25) PDF(2.06 M)( 18)

  Abstract:In recent years, the offshore wind power industry has experienced rapid development, with levelized grid parity gradually becoming the mainstream trend. In the construction of large-scale offshore wind farms,cost reduction and efficiency improvement have emerged as critical factors for the sustainable development of offshore wind farms. Based on this, this paper proposes an optimal planning method of power collection systems for offshore wind power based on mixed-integer programming. The mathematical optimization model considers submarine cable selection, turbine elevation, node connectivity, and other factors, with objectives of minimizing both power collection system investment and life-cycle network loss costs. To enhance computational efficiency and convergence, an acceleration solving strategy using a simplified model for hot-start initialization is introduced. Based on two actual offshore wind farm cases, a numerical analysis is conducted to compare the planning results of proposed method and widely used planning software in the industry. Results show that using the proposed method in two offshore wind farms can effectively reduce the overall investment cost of the power collection system, demonstrating good optimization characteristics and confirming the application prospects of the proposed model and algorithm in the field of power collection system planning for offshore wind farms.

• Profiling Modeling and Response Capability Assessment of Integrated Demand Response Based on Heterogeneous Data

  YANG Shensheng, XU Peng, WANG Beibei, SHANG Xiuyi, SHI Bin, WU Min

  2026,50(1):178-187, DOI: 10.7500/AEPS20250405003

  Abstract(23) HTML(31) PDF(968.30 K)( 13)

  Abstract:As a key outcome of Energy Internet development, integrated demand response (IDR) serves as an essential approach for enabling demand-side participation in interactions with the power grid. Its response methods mainly include energy substitution and energy time-shift, and it has significant advantages such as high user comfort level, strong response enthusiasm, large response potential, and relatively low uncertainty, thereby demonstrating broad development potential. First, the mechanisms of IDR are analyzed and an IDR profiling model that incorporates heterogeneous data is developed. Subsequently, for users in electricity-gas integrated energy systems, a gas substitution model is constructed based on vertical federated learning, while a power load shifting model is developed by integrating consumer psychology models to describe the characteristics of user demand response. Finally, by coupling the power load shifting model with the gas substitution model, an electricity-gas IDR profile is established to realize the precise characterization of user response potential based on price data. Case studies indicate that the proposed profiling model of IDR effectively considers the coupling and coordination between electricity and gas based on heterogeneous data, and can effectively characterize the response capabilities of users in electricity-gas integrated energy system, which provides a scientific basis for the optimal dispatching of energy systems.

• Dynamic Pricing Model and Game Analysis of Multiple Virtual Power Plants Under Electricity-Carbon-Green Certificate Trading Coupling

  ZHAO Chen, PENG Siyuan, HE Ping, WANG Shuai, WU Xiaopeng, FAN Jiale

  2026,50(1):188-198, DOI: 10.7500/AEPS20250221005

  Abstract(29) HTML(44) PDF(1.47 M)( 22)

  Abstract:As the market logic of new power systems evolves into a complex, open market with multi-layer interactions and multi-market coupling, it is of great significance to study the pricing issues and game strategies of virtual power plants (VPPs) in the multi-market environment with coupling of electricity, carbon, and green certificate trading. Therefore, this paper proposes a coupled trading framework for the electricity-carbon-green certificate markets, involving the joint trading between distribution system operators (DSO) and multiple VPPs. A Stackelberg game model is established with the DSO as the leader and multiple VPPs as followers, exploring the dynamic pricing mechanism of the DSO and the bidding strategies of the VPPs in the coupled market. To further solve this model, an artificial neural network-enhanced regional dung beetle optimization (ANN-RDBO) algorithm is proposed. This algorithm, through model prediction, reduces the information interaction between upper and lower layers and lowers the number of lower-layer model calls, significantly improving the computation speed and accuracy. Simulation results verify the rationality and effectiveness of the proposed theoretical model, indicating that the framework and model enhance the autonomous regulation capability of VPPs in the multi-market environment, reduce the total trading costs within the region, and achieve carbon emission reduction in the system.

>工程技术前沿

• Test Analysis and Application Discussion on Active Frequency Support Capability of Renewable Energy in East China Power Grid

  HUANG Hui, MIAO Yuancheng, LI Zhaowei, DING Haoyin, WU Xuelian, SUN Jinwen

  2026,50(1):199-207, DOI: 10.7500/AEPS20250208001

  Abstract(22) HTML(72) PDF(824.99 K)( 26)

  Abstract:With the installed capacity of renewable energy and the proportion of DC input increasing year by year, the frequency stability problem of East China power grid is becoming increasingly complex and severe. In response to the active frequency support capability test of renewable energy conducted by East China power grid on September 29, 2021, this paper analyzes the primary frequency regulation performance of various types of power generation resources including wind power, energy storage, wind storage, and thermal storage under actual disturbances in East China power grid. Tests demonstrate that the modified renewable energy units generally have better fast frequency regulation capability. Based on the evolution trend of frequency characteristics of East China power grid, this paper explores the effect and adaptability of the existing active frequency support capability of renewable energy on improving the future frequency stability level of East China power grid and proposes the suggestions on follow-up research, providing decision support for further targeted enhancement of the active frequency support capability of renewable energy in East China.

• Wind Turbine Fault Diagnosis Based on Sequential Attention and Spatio-Temporal Feature Fusion

  FANG Xiangqing, ZENG Xiangjun, LI Qiuxin, WANG Qiujie, YANG Ming, ZHANG Binqiao

  Available online:December 22, 2025  DOI: 10.7500/AEPS20241202001

  Abstract(94) HTML(78) PDF(1.37 M)( 48)

  Abstract:The redundant characteristics of wind turbine supervisory control and data acquisition (SCADA) data can easily obscure effective information, affecting the accuracy of fault diagnosis. Therefore, an fault diagnosis method for wind turbines combining improved sequential attention feature selection and spatio-temporal feature fusion is proposed. First, a random sparsification strategy is introduced into the sequential attention feature selection to enhance the model's focus on the global information interaction among variables. Second, a matrix sample set incorporating local spatio-temporal features is constructed to improve the mining effectiveness of spatio-temporal features. Next, an adaptive spatio-temporal information fusion model combining a gate-convolution neural network and a weighted bidirectional sequential network is proposed. This model achieves multi-scale global perception of the spatio-temporal matrix through hierarchical receptive field design and adaptively regulates information flow between layers using a convolutional forget gate, thereby enhancing the rapid response capability to abrupt electrical fault features. Simultaneously, the weighted bidirectional sequential network fully models forward and backward temporal dependencies, effectively capturing gradual mechanical fault features. On this basis, the spatio-temporal features extracted by different networks are fused to further improve the model's comprehensive representation capability for multiple fault modes. Finally, validation results based on actual wind turbine SCADA data show that the proposed method can accurately identify various typical wind turbine faults.

• Power Interface Modelling Method for Hybrid Simulation System Based on Predictive Active Disturbance Rejection Control

  LI Guoqing, YANG Mingcheng, JIN Guobin, XIN Yechun, YANG Jun, WANG Wei

  Available online:December 16, 2025  DOI: 10.7500/AEPS20250520009

  Abstract(63) HTML(26) PDF(1.48 M)( 55)

  Abstract:To address issues such as signal delay, impedance mismatch, and stability challenges at the interface of hybrid simulation systems, a power interface modeling method for the hybrid simulation system based on active disturbance rejection control (ADRC) is proposed. Firstly, by analyzing the principles and stability conditions of the voltage-source ideal transformer method, it is proposed to applying the ADRC in the real-time estimation and compensation of interactive errors and unmodeled dynamics between the digital and physical sides, thereby improving the anti-interference capability of the hybrid simulation system. Secondly, a Smith predictor is designed to obtain a delay-free system output, which is then fed back into the extended state observer to enhance the control system performance. Then, a control parameter design method is proposed based on the bandwidth-parameter method from the controller perspective. This method designs a control parameter tuning process suitable for digital-analog hybrid simulation system by setting system bandwidth and tracking performance indicators. Finally, the correctness of the theoretical analysis and the effectiveness of the proposed method are verified based on digital simulations and power hardware-in-the-loop experiments.

• Active Power Control System and Strategy for Full-voltage-level Renewable Energy Provincial-local Bidirectional Coordinated Support

  ZHANG Liudong, ZHANG Haibo, XIONG Hao, LI Yanliu, PENG Zhiqiang, LI Feng

  Available online:November 04, 2025  DOI: 10.7500/AEPS20240731007

  Abstract(114) HTML(138) PDF(2.09 M)( 78)

  Abstract:To address the diverse requirements for renewable energy control scenarios at the provincial and local levels, such as provincial-local grid peak regulation and transmission section control, as well as provincial-local main transformer overload risk control, research has been conducted on the architecture and strategies of a provincial-local coordinated automatic active power control system for renewable energy. A coordinated technical approach combining “provincial dispatch-local dispatch” and “centralized-distributed” strategies is adopted. Renewable energy automatic power control modules are deployed at both provincial and local dispatch master stations, establishing a “1+N+X” coordinated automatic active power control system for renewable energy across full voltage levels. Based on this system, two types of bidirectional coordinated support strategies for provincial-local renewable energy resources are proposed. Specifically, local dispatch can aggregate and transmit its adjustable resources to provincial dispatch to meet the latter"s regulation needs. Meanwhile, local dispatch can also generate regulation commands in response to local emergencies, such as N－1 overload of main transformers. These commands are then uniformly decided upon by provincial dispatch and issued to unified-dispatch resources for execution, thereby comprehensively enhancing the grid"s balance adjustment capabilities. The effectiveness of the proposed system and strategy is verified by the actual operation effect of a provincial and local dispatch to ensure the safe and stable operation of the power grid and the maximum consumption of renewable energy.

• Development Status and Engineering Application of Grid-forming Converter Technology

  GUO Xianshan, ZHANG Guohua, LI Jiandong, HUANG Qinxin

  Available online:November 04, 2025  DOI: 10.7500/AEPS20250324004

  Abstract(62) HTML(14) PDF(7.87 M)( 62)

  Abstract:With the ongoing advancement of new power system construction in China, the green and low-carbon transformation of energy is accelerating and upgrading. The proportion of renewable energy installed capacity has significantly increased, and a large number of power electronic devices have been integrated into the power grid. Profound changes, including reduced moment of inertia and weakened reactive power support capabilities, have occurred on the power source, load, and grid sides of the power system. In this context, the power system requires equipment to possess new characteristics, namely corresponding features of synchronous machines and the ability to actively support the power grid. This paper elaborates on the main characteristics of the new power system from three aspects: demands of the new power system, grid-forming converter technology, and typical equipment and engineering applications. In addition, the working principle and various topologies of grid-forming converters are analyzed. Subsequently, based on the existing research and practical projects, the engineering application cases of four typical grid-forming converters, namely grid-forming flexible DC, grid-forming static compensator, grid-forming static var generator, and grid-forming energy storage, are described. Finally, development suggestions and prospects are proposed based on the current research status of grid-forming technology, providing references for further engineering applications.

• Weakly-correlated Measurement Anomaly Recognition of Low-voltage Distribution Network Based on Neural Network with Variable Weights

  JIN Yangxin, XU Yongjin, WANG Jinrong

  Available online:November 04, 2025  DOI: 10.7500/AEPS20241117002

  Abstract(68) HTML(10) PDF(2.30 M)( 20)

  Abstract:As an important data source of low-voltage distribution network,the measurement anomaly control of smart energy meter concerns electricity trading fairness. A recognition model of weakly-correlated measurement anomaly of line loss is developed, whose inputs are frozen active power and voltage recorded by energy meter, as well as partial archival information. The principals of the model is the residual neural network with variable weights, which is utilized to forecast the measurement anomaly index. Aimed at the accuracy degradation owing to insufficient inputs, the weight variable algorithm of neural network is proposed, which modifies the view that each weight maintain fixed after training, whereas regards part of them as superposition of several eigenstate weights. Through the node correlation state that is not able to be input directly, the superimposing parameters of eigenstate weights are calculated via density matrices merging of sub-dimensions, which significantly improves model recognition accuracy. Finally, the validity of residual neural networks with variable weights is verified with the enhanced training and validation datasets. Furthermore, the model is applied to the control of high-loss range in a city in Zhejiang Province of China to confirm its detection effect under actual conditions.

• Fault Section Identification and Location for Transmission Lines Using Current Traveling Waves at Single Measuring Point

  ZHANG Guangbin, SHU Banggui, SHU Hongchun, SI Dajun

  Available online:September 01, 2025  DOI: 10.7500/AEPS20250204003

  Abstract(38) HTML(24) PDF(2.22 M)( 52)

  Abstract:Distributed travelling wave devices along the transmission lines need to be paired and can only perform multi-terminal fault location algorithm, which strictly relies on time clock synchronization. Single measuring point cannot independently identify fault segment or locate fault. In order to solve the drawbacks, the paper proposes a fault segment identification and fault location algorithm utilizing current travelling waves observed by single measurement along transmission lines. Firstly, by analyzing the the topological features of multiple lines connection at both ends of the transmission lines, it is found that the relative polarity determined by such topological type can be used to identify the properties of the subsequent current travelling waves at a single measuring point. Furthermore, the essential difference between the fault segment and the sound segment is found, that is, whether there is fault point reflection after the travelling wave enters the segment. Transmission line is divided into two segments of unequal length by using measuring point along the line, and the wave impedance at the measuring point is continuous. Wavefronts are identified according to the relative polarity and inherent arrival time differences of travelling waves in the segments, ultimately achieving fault segment identification and single-ended fault location. Massive data of actual measurement and simulation tests demonstrate the feasibility and effectiveness of the proposed algorithm. The method is not affected by different fault types, various fault resistances, and the deviations of travelling wave velocity and line length. The method exhibits no dead zones for the nearby faults. The installation positions of the traveling wave devices are versatile and flexible. The proposed method can serve as a backup algorithm and essential complement to existing double-ended fault location of distributed traveling wave devices. And it can be used as an independent fault segment indicator of transmission line.

• Real-time Stochastic Dispatch Method for Active Distribution Networks Using Fuzzy Neural Network Pre-decision

  CHENG Lilin, LUO Zijie, LI Qun, ZHANG Ningyu, LI Yaran, ZANG Haixiang

  Available online:August 25, 2025  DOI: 10.7500/AEPS20250213005

  Abstract(38) HTML(8) PDF(1.31 M)( 74)

  Abstract:With the large-scale integration of distributed resources into the active distribution network, internal uncertainties increase. To establish an accurate scheduling model, the uncertainty of photovoltaic output and the randomness of load are described as the uncertainty of corresponding prediction errors, and the probability distribution of uncertain variables is obtained through data-driven methods. Considering that the power flow model based on second-order cone relaxation (SOCR) may violate relaxation constraints and introduce errors, the power flow model is re-derived based on the Euler equations, thereby further establishing a stochastic optimal power dispatch model for active distribution networks that is both economical and secure. Aiming at the characteristics of the proposed model, this paper presents a real-time stochastic dispatch method for active distribution networks using fuzzy neural network (FNN) pre-decision. First, the FNN is employed to provide a fuzzy description of the probability distribution of uncertain variables, and its output is used as the initial value for solver optimization. Then, the solver is utilized to accelerate the solution process. Finally, the effectiveness of the proposed model and method is validated through a modified IEEE 33-bus system.

• Synchronous Measurement Data-driven Classification Method for Disturbance Sources in Distribution Networks

  CHEN Zhilin, LIU Hao, BI Tianshu

  Available online:August 19, 2025  DOI: 10.7500/AEPS20250110002

  Abstract(80) HTML(10) PDF(3.00 M)( 66)

  Abstract:To address the challenges brought by the integration of a large number of distributed generators, synchronous measurement technology has been introduced into the distribution network, providing more dimensional data for the safety and stability analysis of the system. To further enhance the situational awareness of the distribution network, a synchronous measurement data-driven method for classifying disturbance sources in the distribution network is proposed in this paper. Firstly, a fast identification method for disturbance sources based on Granger causality analysis theory is proposed to distinguish between the main network side and the distribution network side. Subsequently, a disturbance identification method based on kernel support vector machine algorithm is proposed to identify disturbances caused by errors in instantaneous estimation algorithms, thus achieving rapid classification and extraction of real-time distribution network disturbance data. Finally, the effectiveness of the proposed method is verified based on IEEE 123-bus system simulation data and on-site synchronous measurement data.

• Analysis and Optimization Measures of Current Breaking Mechanism for Back-to-back Conventional HVDC

  XIE Huifan, WU Jiyang, PENG Guangqiang, LIU Tao, HUANG Weihuang, CHEN Qian

  Available online:July 08, 2025  DOI: 10.7500/AEPS20240626007

  Abstract(14) HTML(8) PDF(5.31 M)( 50)

  Abstract:Regarding the incident on August 19, 2023, where the Luxi back-to-back conventional HVDC project experienced a prolonged DC current breaking due to an external AC fault, causing a bipolar lockout, the current establishment mechanism of the back-to-back conventional HVDC project is analyzed. The reasons for the unsuccessful establishment of DC current in the Luxi back-to-back conventional HVDC project are identified. The valved-controlled compensated pulse optimization strategy in low DC current scenarios is proposed. The simulation results of the real-time digital simulator (RTDS) platform fully verified its effectiveness and successfully implemented on-site application. The practical application results show that the proposed optimization strategy effectively eliminated the prolonged DC current breaking in the Luxi back-to-back conventional HVDC project caused by the external AC fault and further prevented the bipolar lockout risk, significantly improving the operational reliability and stability of the back-to-back conventional HVDC project. This optimization strategy is worthy of reference and application in similar projects.

• Active Single-ended Location Method for Grounding Electrode Lines Based on Short-time Matrix Pencil

  TIAN Xincui, CHEN Kaiwen, SHAN Jieshan, ZHANG Yining, YU Jinyun, LI Qiang

  Available online:July 03, 2025  DOI: 10.7500/AEPS20240904005

  Abstract(131) HTML(2) PDF(9.05 M)( 55)

  Abstract:The fault information in grounding electrode lines is weak and well-hidden, making detection and fault location very difficult. On this basis, a single-ended location method for grounding electrode lines is proposed based on broadband excitation injection and the short-time matrix pencil method (STMPM). First, the Gaussian signal excitation with an oscillatory decay characteristic is injected into the grounding electrode line in differential mode, ensuring that the injected excitation does not leak into the DC system side through the neutral bus and minimizing waveform distortion during the propagation of the signal along the grounding electrode line, thereby improving the detection efficiency of fault traveling waves. Secondly, sliding short-time windows are used to perform singular value decomposition (SVD) on the fault traveling waves. The eigenvalues obtained from the decomposition are used to distinguish the interference signals and fault signals, effectively amplifying the weak fault signals while suppressing the interference signals. Finally, the damping factor of the fault traveling wave within the short window is determined, a one-to-one mapping relationship between the zero-crossing moment of the damping factor and the arrival time of the fault traveling wave is established, and the fault distance is then determined. Extensive simulations show that the location method can effectively detect fault traveling waves and achieve high fault location accuracy.

• Comprehensive Identification Method for Transformer Excitation Inrush Current Based on Multi-perspective Time-frequency Characteristics

  CHEN Chun, ZHAN Luxin, CAO Bozhong, CAO Yijia, LI Yong, LIU Junle

  Available online:June 13, 2025  DOI: 10.7500/AEPS20240723005

  Abstract(216) HTML(18) PDF(2.49 M)( 156)

  Abstract:The increasing integration of power electronic devices in distribution networks has exacerbated harmonic currents, challenging the reliability of traditional transformer differential protection reliant on second harmonic restraint. Additionally, the single-characteristic identification methods are affected by the distributed generator (DG) types and closing angles, and cannot accurately distinguish fault currents and inrush currents in different scenarios. In order to improve the identification accuracy of excitation inrush current, this paper proposes a multi-perspective time-frequency analysis method that comprehensively integrates time-domain, frequency-domain, and time-frequency-domain characteristics. Bayes algorithm is used to optimize the classification parameters of eXtreme Gradient Boosting (XGBoost), improve the generalization ability of the model, and achieve accurate identification of fault current and excitation inrush current under different capacities and types of DG connected. The SHapley Additive exPlanations (SHAP) value analysis method is used to reveal the contribution of each characteristic value in the identification model. Based on simulation and on-site measurement data, the proposed excitation inrush current identification method is validated, and the accuracy of identification for sample data is close to 100%.

• A Rapid Fault Tracing Method for Large-Scale Complex Distribution Networks Based on Dynamic Graph ProjectionZHANG Yujia1^, YUAN Ye1^, ZHOU Suyang2^, ZHU Hong3^, ZHOU Aihua3^, CHEN Qingquan⁴

  ZHANG Yujia, YUAN Ye, ZHOU Suyang, ZHU Hong, ZHOU Aihua, CHEN Qingquan

  Available online:June 09, 2025  DOI: 10.7500/AEPS20240112003

  Abstract(177) HTML(2) PDF(3.48 M)( 251)

  Abstract:With the rapid growth of the distribution network and the high penetration of distributed resources, the topology of distribution networks has become increasingly complex, posing significant challenges to fault location analysis. When applying matrix algorithms and intelligent optimization algorithms to fault location, it is necessary to construct network matrices or establish optimization models based on changing topology information. This greatly increases the computational burden and complexity, leading to low efficiency in data processing and computation. Therefore, this paper first constructs a graph data model for the distribution network topology. Utilizing graph projection techniques, it extracts optimized subgraphs tailored for fault tracing scenarios from the panoramic power grid graph. On this basis, the Yen"s shortest path search algorithm is employed to find potential fault paths from the power source to the abnormal nodes. By traversing the line nodes and assessing their overcurrent information, the fault section is identified, thereby resolving the issues of accurate representation and rapid search of the power grid topology. This enables quick and precise fault localization in large-scale complex distribution networks, greatly enhancing fault search efficiency while ensuring the accuracy of fault tracing.

• Construction Scheme of Voltage-supporting Renewable Energy Station Based on Distributed Synchronous Condensers—Part Two Coordinated Control Strategy and Test Verification

  LIANG Hao, QIN Chuan, XIE Huan, LIANG Beihua, WU Tao, WANG Xuanyuan, WU Long

  Available online:May 29, 2025  DOI: 10.7500/AEPS20240807002

  Abstract(220) HTML(4) PDF(3.04 M)( 269)

  Abstract:Deploying the distributed synchronous condenser in the renewable energy station is an effective measure to improve the voltage-supporting ability of the power supply side in the whole process. However, the existing cascade strategy of “reactive power outer loop + voltage inner loop” is adopted to integrate the synchronous condenser into the automatic voltage control system (AVC) of the station, which limits the low-frequency voltage source characteristics of the synchronous condenser. First, this paper describes the existing problems of the synchronous condenser integrated into AVC and proposes the requirement for constant voltage for the synchronous condenser throughout the whole process. Based on the topology of the renewable energy station with distributed synchronous condensers, the reactive voltage conversion coefficient and the reactive power shunt influence factor are analyzed. Then, this paper proposes a new scheme of “constant voltage + reactive power shunt suppression” for integrating the synchronous condenser into AVC. The program is developed on an equipment of domestic mainstream manufacturer, and the effectiveness of the proposed scheme is verified by the hardware-in-the-loop simulation of the dual controllers of synchronous condenser excitation and AVC at the station. Finally, the engineering application has been completed in an actual renewable energy station. The field operation results show that the proposed scheme realizes the steady-state regulation of multi-type reactive power equipment of the station, effectively reduces the voltage fluctuation amplitude at the grid-connection point of the station, fully utilizes the voltage source characteristics of the synchronous condenser in the whole process, and guarantees the voltage stability margin of the system.

• Construction Scheme of Voltage-supporting Renewable Energy Station Based on Distributed Synchronous Condensers—Part One Mechanism Analysis

  LIANG Beihua, XIE Huan, ZHAO Tianqi, LIANG Hao, HAO Jing, XIN Huanhai

  Available online:May 12, 2025  DOI: 10.7500/AEPS20240807001

  Abstract(36) HTML(14) PDF(2.63 M)( 49)

  Abstract:The integration of distributed synchronous condensers into renewable energy stations can effectively improve the short-circuit ratio index and the system transmission capacity. However, the influence and quantitative evaluation of the control characteristics of synchronous condensers on the voltage-supporting capability and the power grid strength are rarely studied. First, this paper explains the requirements of the voltage-supporting function of the station throughout the entire process. Based on the frequency domain index, the difference in voltage-supporting capability with two strategies of voltage control and cascaded reactive power outer loop control is compared and analyzed. It is revealed that the introduction of the reactive power outer loop will cause the characteristic of the synchronous condenser at the low frequency to be transformed into the current source characteristic, reducing the voltage-supporting capability of the system during the dynamic adjustment process after fault clearance. Next, a short-circuit ratio index for the entire process of a renewable energy station containing synchronous condensers is proposed. Its physical meaning is voltage sensitivity. This index can effectively describe the influence of different control strategies on voltage stability at multiple time scales, which explains the adaptability of using subtransient reactance parameters of the synchronous condenser to calculate operating limits. Finally, the correctness of the analysis conclusions is verified through RTDS hardware-in-the-loop simulation.

• Preliminary Exploration of Carbon Emission Estimation Technology Based on Power Big Data

  LIU Yuliang, LI Yaowang, DU Ershun, ZHANG Ning, KANG Chongqing, DU Siyuan

  Available online:April 27, 2025  DOI: 10.7500/AEPS20240905004

  Abstract(122) HTML(124) PDF(2.75 M)( 531)

  Abstract:Energy activities are the primary source of carbon dioxide emissions. Accurate and near real-time monitoring of carbon emissions from energy activities is fundamental for governments and enterprises to grasp the total carbon emissions and formulate carbon management strategies. However, existing accounting methods are insufficient for high-frequency carbon emission calculations, whereas direct measurement methods necessitate additional equipment to be installed. Consequently, low-cost and high-frequency carbon emission estimation techniques remain unavailable. To address this issue, by using the correlation between electricity consumption and energy activities, this paper proposes an estimation technology for carbon emissions from energy activities based on power big data (abbreviated as “electricity-based carbon emission estimation” technology). Firstly, the methodological framework for “electricity-based carbon emission estimation” technology is proposed. From a data-centric perspective, a detailed investigation and analysis on related foundational data are conducted, and the potential application scope of existing data is explored. Secondly, a time-series based “electricity-based carbon emission estimation” model applicable at regional, industrial, and enterprise levels is constructed, followed by an evaluation of its effectiveness using actual data from China. The analysis results demonstrate that the proposed method can significantly improve the timeliness of carbon emission estimation. Finally, the future research and potential application directions for the “electricity-based carbon emission estimation” technology are prospected.

• New Transmission Test System Generation Method Based on the Concept of Grid Evolution

  LI Bo, WEI Guangrui, ZHONG Haiwang, LIU Hui

  Available online:April 15, 2025  DOI: 10.7500/AEPS20240409003

  Abstract(355) HTML(52) PDF(1.75 M)( 197)

  Abstract:The IEEE test case has been widely used for simulation testing in various fields such as power system planning and operation. However, due to data privacy concerns, it isn"t easy to access publicly available datasets of actual power system generation and network structures. To address the issue, based on the evolution concept of three-generation power grids, a new transmission test system generation method is proposed to build test cases that reflect actual power system network structures. Firstly, a transmission expansion planning model considering N-1 security constraints is established. Secondly, the optimization objectives and constraints are proposed according to the characteristics of different stages of network development to simulate the evolution process of the electricity network. To enhance the solving efficiency of the model, a binary representation method of transmission corridors is introduced to reduce the number of 0-1 variables. Finally, the provincial power grid is used as an example to verify the effectiveness of the proposed open-source power system dataset, according to the statistical characteristics of complex networks. The proposed dataset is also applied to optimal transmission switching for further verification.

• Anomaly Detection in Distribution Main Station Logs Based on Large Language Model

  Wang Shen, WEI Xingshen, ZHU Weiping, ZHU Daohua, GUAN Zhitao

  Available online:February 27, 2025  DOI: 10.7500/AEPS20240523001

  Abstract(341) HTML(372) PDF(1016.64 K)( 571)

  Abstract:Log anomaly detection is one of the key technologies to monitor the operation of distribution master station system and identify abnormal behavior. Existing log anomaly detection methods based on deep learning rely on a large amount of in-domain training data, and the scarcity of training data will lead to a significant decline in performance. Aiming at the above problems, based on the contextual reasoning characteristics of large language models, an adaptive hint strategy is designed and a training-free anomaly detection scheme for distribution master logs is implemented. Firstly, a demonstration example filtering algorithm is designed to dynamically select several high-quality demonstration examples from a small number of labeled local logs for different online logs. Then, combined with the task description and human experience knowledge, a text hint is automatically constructed to guide the large language model to complete the anomaly detection task of distribution master station logs. The experimental results on the general data set and the self-built distribution master station data set show that the proposed scheme has better performance than the existing methods, showing higher flexibility and generalization.

• Emergency Frequency Control Strategy of Microgrid Clusters for Extreme Survival of Critical Urban Loads

  WANG Ziyuan, XU Yin, WU Xiangyu, LI Jiaxu

  Available online:February 26, 2025  DOI: 10.7500/AEPS20241024004

  Abstract(143) HTML(106) PDF(1.97 M)( 708)

  Abstract:The proportion of electricity received outside the urban power grid is high, and extreme events leading to connectivity failures in the power transmission channels of the urban superior power grid may cause major power outages. In this extreme scenario, by flexible self-configuration operation and supply guarantee of microgrid clusters, the critical load survival can be achieved. However, the large transient frequency fluctuation of microgrids during sudden power shortages directly affects the success or failure of self-configuration operation and supply guarantee. Firstly, a framework for solving the extreme survival problem of microgrid clusters is proposed, and dynamic frequency constraints for microgrids considering multi-source collaboration are constructed based on the system frequency response model. Secondly, the frequency response model of microgrid containing nonlinear constraints such as ordinary differential equations and limiting links is differentially discretized based on the forward Euler method. Then, according to the generator tripping and load shedding during the transient frequency response of microgrid clusters, a microgrid cluster model for extreme survival is constructed considering inertia equivalent constraints, control action delay constraints and dynamic segmentation constraints. By solving a mixed-integer linear programming model, the emergency frequency control strategy and the segmentation state of microgrid clusters are obtained to coordinate multiple frequency regulation resources and minimize the load shedding volume while ensuring frequency safety. This model can be efficiently solved using mature commercial solvers. Finally, the effectiveness and superiority of the proposed emergency frequency control strategy of microgrid clusters are verified through numerical simulation analysis.

• Optimal Configuration Method for Backup Carbon Meter System for Carbon Measurement of Electricity Consumption

  GUO Kuixing, LI Yaowang, HE Xiaoyi, LIU Heng, LI Xuecheng, ZHANG Ning

  Available online:January 24, 2025  DOI: 10.7500/AEPS20240613008

  Abstract(43) HTML(88) PDF(1.82 M)( 725)

  Abstract:The electric carbon meter system is an important carrier for realizing real-time and accurate electric carbon measurement. With the advancement of the goal of carbon emission peak and carbon neutrality, the requirements for the accuracy and continuity of measurement by electric carbon meter become increasingly stringent. It will be difficult to cope with the increasing requirements relying on an independent carbon meter system, and it is urgent to explore the optimal configuration method for the backup carbon meter system. Considering the characteristics of “one hair affecting the whole body” of the electric carbon measurement and the strong coupling relationship between the electric carbon measurement and the power flow, this paper proposes an optimal configuration method of the backup carbon meter system. The method is conducted in a stepwise optimization manner. Firstly, to ensure that the carbon meter system can measure the full carbon information of the power system, this paper proposes an optimization method to determine the number of spare carbon meter devices with the lowest configuration cost. Then, with the aim of minimizing the measurement error of the backup carbon meter system, this paper proposes a method of optimizing the distribution of the backup carbon meter device. Finally, the simulation test is carried out based on the PJM 5-bus system and the actual power system data of a certain city in China. The results show that the proposed method can achieve a carbon emission measurement error of less than 0.2% with the minimum configuration cost of the backup carbon meter system, with the premise of ensuring the complete carbon emission information measurement of the power system.

• Optimal Scheduling Method for Power Grid Considering Participation of Electric Vehicles in Flexible Regulation During Cold Wave Weather

  JIAO Zhijie, XU Yin, LIU Zhao, WANG Xiaojun, HE Jinghan, SI Fangyuan

  Available online:December 30, 2024  DOI: 10.7500/AEPS20240519001

  Abstract(65) HTML(156) PDF(3.06 M)( 819)

  Abstract:The arrival of a cold wave triggers a sudden drop in weather temperature, at this time, the energy consumption of the load increases, the output of renewable energy sharply decreases, the system backup and the power supply capacity of the higher-level power grid are insufficient, resulting in a significant power shortage problem within the power grid in a short period of time. With the increasing number of electric vehicles on the load side and the improvement in the responsiveness of flexible resources, the load-side flexible resources are adjusted to compensate compensate for the power shortage caused by the cold wave. This paper first clarifies that the flexibility of resources varies with changing scenarios, explores and constructs models of flexible resources such as electric vehicles during cold wave events. Secondly, considering that electric vehicles in flexible resources need to participate in grid scheduling through aggregation, a method for aggregating electric vehicles with uncertain connection times is developed. Subsequently, based on the flexible resources during cold wave events, non-residential loads, and residential loads, a rolling optimal scheduling method for the power grid during cold wave weather is proposed with the objective of minimizing social losses. The electricity adjustment of various resources in the grid during cold wave weather is determined. Finally, through case studies, the proposed method is shown to effectively address power shortages in the grid during cold wave events.

• Identification Method for Tree-touching Grounding Fault of Medium-voltage Line Based on Long-term Variation Features of Zero-sequence Current

  WANG Pengwei, XU Bingyin, LIANG Dong, WANG Lianhui, WANG Chao, ZOU Guofeng

  Available online:October 31, 2024  DOI: 10.7500/AEPS20240315002

  Abstract(71) HTML(56) PDF(6.79 M)( 157)

  Abstract:Distinguishing whether faults in medium voltage distribution lines are caused by lines touching trees is of great significance for clarifying the causes of forest fires and preventing line faults from causing forest fires. The zero-sequence currents of various high-impedance grounding faults are obtained through prototype experiments in the paper, and the long-term variation features of the zero-sequence current waveforms of high-impedance grounding faults are analyzed. Analysis shows that there are significant differences in the fluctuation, monotonicity, and sharpness of the waveforms of the effective value of the zero-sequence currents for line touching trees grounding faults compared to other high-impedance grounding faults. A multi feature fusion parameter set including standard deviation, discrete coefficient, kurtosis, skewness of the zero-sequence current effective value curve is designed, and a ientification method for tree-touching grounding fault of medium-voltage line based on support vector machine is constructed. The results showed that the proposed method achieved a fault recognition accuracy of 98%.

• Frequency Safety Assessment of Power Systems Based on Multivariable-sample Convolution and Interaction Network

  LIU Jie, SHI Fang, SONG Xuemeng, TIAN Shuoshuo, NIE Liqiang

  Available online:October 17, 2024  DOI: 10.7500/AEPS20231101004

  Abstract(61) HTML(58) PDF(2.34 M)( 252)

  Abstract:The existing intelligent assessment methods for transient frequency in power systems do not adequately consider the temporal characteristics of input data. Therefore, a frequency safety assessment method for power systems based on intelligent prediction of transient frequency response curves is proposed in the paper. A multivariate sample convolutional interactive network is designed to fully exploit the temporal characteristics of power system measurement data, thereby improving the prediction accuracy of power system transient frequency response curves; Key indicators such as maximum frequency deviation, occurrence time of maximum frequency deviation, and the metastability frequency are calculated based on the predicted frequency response curves, and the frequency safety of the system is comprehensively assessed. Simulation tests are conducted on frequency stability standard examples, and the results showed that the proposed method effectively improved the accuracies of frequency response curve prediction and system frequency safety assessment compared to classical methods such as deep learning.

• Selectivity Enhancement Scheme for Zero-sequence Current Protection Based on Distance Relay for High-resistance Grounding Fault

  LI Yong, LI Yinhong, LIU Huanzhang, LIU Yang

  Available online:October 09, 2024  DOI: 10.7500/AEPS20240228008

  Abstract(39) HTML(26) PDF(2.76 M)( 173)

  Abstract:The last section of zero-sequence current protection of AC line adopts 300 A, which has the risk of disordered tripping.Therefore, a new principle of high-resistance grounding distance relay based on zero-sequence reactance line and non-fault phase polarization is proposed. The relay adopts the technical route of phase selection before measurement. The phase selection element combines zero-sequence reactance line and non-fault phase polarization method to form a variety of combined criteria to complete the phase selection. Due to the phase difference between the zero-sequence current at the protection installation and the zero-sequence current at the fault point, the zero-sequence reactance lines of the single-phase grounding fault phase and the advance phase of the inter-phase grounding fault have aliasing region when the fault point is near the setting point. The large variation of the operation voltage of the non-fault phase is not conducive to distinguishing the two types of faults in the aliasing region, and thus the phase selection element is divided into low-resistance module and high-resistance module. The low-resistance module adopts the zero-sequence reactance line with the downward bias, which is used to identify the near end and low-resistance short circuit. With the assistance of the low-resistance module, the high-resistance module only needs to deal with the faults near the set point, which reduces the difficulty of distinguishing the two types of faults . After phase selection, the operation voltage before fault is obtained by non-fault phase polarization method, so as to determine the operation characteristics of the relay. The ability of high-resistance distance relay to withstand the transition resistance is far beyond the requirements of the regulations, which improves the selectivity of ground backup protection to high-resistance faults.

• High-resistance Grounding Fault Detection Method for Distribution Network Considering Capacitor Switching Disturbance

  HAN Zhaoru, SHI Fang, ZHANG Hengxu, JIN Zongshuai, YUN Zhihao

  Available online:September 29, 2024  DOI: 10.7500/AEPS20240116008

  Abstract(26) HTML(98) PDF(4.03 M)( 468)

  Abstract:The accurate and reliable detection of high-impedance grounding fault (HIGF) is a challenging issue in fault handling for distribution networks, and normal capacitor switching operations can cause interference. Addressing this problem, a disturbance-resistant detection method for HIGFs based on zero-sequence Lissajous curve analysis is proposed in this paper. Firstly, the zero-sequence electrical quantities of HIGFs and capacitor switching disturbances are theoretically derived. There is no regular difference in the traditional time-frequency domain feature aspect between the two, thereby clarifying the cause of the interference. Further, the zero-sequence current and voltage waveforms are reconstructed into zero-sequence Lissajous curves. A quantitative index for the distortion complexity of the Lissajous curve trajectory shape based on the mathematical morphology theory is proposed, and an adaptive starting criterion is designed in combination with the probability distribution law of the zero-sequence Lissajous curve area. The disturbance-resistant detection algorithm for high-impedance grounding faults in the noise scenario is presented. Finally, the effectiveness and reliability of the proposed method are verified through electromagnetic transient simulation examples and real fault tests in the distribution network.

• Development and Application of Controllable Line Commutated Converter

  HE Zhiyuan, GAO Chong, YE Hongbo, YANG Jun, WANG Chenghao, SHENG Caiwang

  Available online:August 29, 2024  DOI: 10.7500/AEPS20240506007

  Abstract(16) HTML(6) PDF(2.88 M)( 171)

  Abstract:Controllable line-commutated converter(CLCC) is a new type of DC converter equipment proposed to solve the commutation failure problem of conventional DC transmission converter. In June 2023, the world"s first set of controllable phase converter valves for the ±500kV/1200MW Gezhouba to Shanghai Nanqiao high-voltage DC transmission system renovation project (hereinafter referred to as "Genan renovation project") was successfully put into operation. The technical requirements and principles of CLCC are analyzed, as well as its technical and economic benefits. Combined with the input conditions of the Genan renovation project system, the electrical parameters and structural design scheme of the controllable line-commutated converter valve were proposed, and the equipment development was completed. The type test scheme and test parameters were proposed, and the type test assessment was completed. According to the technical characteristics of the controllable line-commutated converter valve, field tests such as low-voltage test, open line test, and artificial short-circuit test were carried out to ensure the smooth operation of the project. The operation performance since the project put into operation was introduced, and the correctness of the control sequence during the AC fault was analyzed based on the field recording. Finally, the application prospects of the controllable? line-commutated converter valve in UHVDC projects, provincial AC liaison line renovation and other scenarios were prospected and analyzed, providing a reference for the further promotion and application of the controllable commutation converter valve.

• Ultra-short-term Regional Load Forecast Based on Spatio-Temporal Graph Attention Network

  ZHAO Ziyu, CHEN Yuanrui, CHEN Tingwei, LIU Junfeng, ZENG Jun

  Available online:April 28, 2024  DOI: 10.7500/AEPS20230914003

  Abstract(66) HTML(40) PDF(3.29 M)( 185)

  Abstract:A regional-level ultra-short-term load forecasting model based on a spatio-temporal graph attention network is proposed in this paper. Firstly, based on the existing regional-level load, cell partitioning is carried out to construct a graph topology that considers cell correlation. Secondly, effective features are extracted from the spatial, feature, and temporal dimensions through the graph attention network, one dimensional convolutional network and gated recurrent unit, connecting the fully connected layers to output the results. Finally, simulation validation is conducted based on real power load data from the New England region of the United States, and model attention weights are extracted to analyze spatial dependencies between cells. The results show that, compared with traditional models, the proposed model provides higher accuracy and stability with different prediction steps, effectively exploiting the spatial dependence of regional spatial load.

• Pricing Strategy of Charging Station in Power-Transportation Coupling Network Considering User Travel Cost Budget

  XIE Longtao, XIE Shiwei, CHEN Kaiyue, ZHANG Yachao, CHEN Zhidong

  Available online:January 04, 2024  DOI: 10.7500/AEPS20230628010

  Abstract(61) HTML(158) PDF(2.23 M)( 209)

  Abstract:With the large-scale development of electric vehicles, it is of great significance to study how to effectively consider the travel behavior mechanism of users and formulate rational charging prices for charging stations for the collaborative optimization and scheduling of power-transportation networks. To solve this problem, this paper proposes a pricing strategy for charging stations in the power-transportation coupling network considering the user travel cost budget. Firstly, a transportation user equilibrium model considering the travel cost budget is established, and the equilibrium state is equivalently described through variational inequalities, so as to characterize the travel demands and charging behaviors of electric vehicles. Secondly, a second-order cone optimization model for distribution networks considering power reduction is constructed. The charging station pricing problem has been transformed into an optimization problem with variational inequality constraints, and an alternating iteration algorithm combined with an extra-gradient algorithm is designed to solve the problem. Finally, the effectiveness of the proposed model and methods is verified through a case, and the results show the necessity of considering the travel cost budget for charging pricing in coupled networks.

• Non-Intrusive Real-Time Monitoring System for Electric Bicycle Based on Wavelet Detection and Feature Graph Decision

  LI Xiang, Liu Yuhang, ZHANG Qi, WU Xin

  Available online:September 12, 2023  DOI: 10.7500/AEPS20230425001

  Abstract(67) HTML(70) PDF(2.53 M)( 195)

  Abstract:The illegal charging behavior of electric bicycles (EBs) in households has temporal randomness and spatial concealment,which poses significant safety hazards and is difficult to effectively manage. A non-intrusive real-time monitoring system for EB charging based on wavelet detection and feature graph decision is proposed, utilizing the characteristics of real-time autonomous execution and easy promotion of non-invasive monitoring systems. Considering the physical structure and charging characteristics of EB loads, the typical common characteristics of EB loads are analyzed from both transient and steady-state perspectives. The EB proprietary feature map with strong distinguishability and universality is constructed in advance to realize consistent and structured expression of EB steady-state common features. In the actual monitoring process, in order to reduce the computational power demand and data transmission pressure of the system, EB specific transient phenomena with high-frequency components are accurately located based on wavelet transform to complete EB like event detection. Finally, extract event waveforms and train efficient classifiers through graphs for load identification and real-time upload. By monitoring actual users, the effectiveness of the monitoring system has been verified, which can effectively solve the problem of charging EBs in buildings and households.

• Power Consumption Model of Server in Data Center Based on Temperature Estimating by CPU Operating Statuses

  ZHU Jiebei, LIU Donghao, LI Bingsen, YU Lujie, TIAN Na, JIA Hongjie

  Available online:June 26, 2023  DOI: 10.7500/AEPS20220905004

  Abstract(23) HTML(14) PDF(14.17 M)( 65)

  Abstract:To reduce energy consumptions of data centers, the accurate establishment of data center server power consumption model is crucial. The traditional server power consumption model, which is only based on CPU utilization, ignores the delayed dynamic characteristics of server power changes influenced by temperature variables, and cannot distinguish the effect of CPU operating statuses based power consumption layering (CPL) under different CPU operating statuses, resulting in large modeling errors. To this end, a server power consumption model based on temperature estimating by CPU operating statuses (PMTC) is proposed in this paper. By taking the temperature variables into account in the model building stage, PMTC captures the delayed dynamic characteristics of server power consumption changes influenced by the delayed change of the CPU core temperature, eliminating the modeling error caused by the CPL effect. In the power calculating stage, PMTC identifies specific CPU operating statues, and then precisely estimates the CPU core temperature, avoiding extra measurement on temperature variables. A server power consumption test platform is established to compare the proposed PMTC and traditional server power consumption models, verifying that PMTC can restore the delayed dynamic characteristics of server power consumption change without increasing input date dimension and reduce the model calculation error.

• Spatial Feature Based Detection of False Data Injection Attack on Synchronous Grid Measurements

  ZHENG Yao, ZHANG Jie, YAO Wenxuan, QIU Wei, TANG Sihao

  Available online:March 16, 2023  DOI: 10.7500/AEPS20220813001

  Abstract(42) HTML(202) PDF(2.51 M)( 49)

  Abstract:As the power system gradually moves toward a new ecosystem of energy interconnection and the deep coupling of the network layer and physical layer, the threat of network attacks on the power system continues to rise. The source identity (ID) spoofing attack, as a new and complex, strong stealthy false data injection attack, can cause the grid control system to misjudge and cause system paralysis. To address this problem, a spatial feature-based method is proposed for detecting false data injection attacks on synchronized measurements of power grids. It has extracted different spatial features of the synchronized measurement devices at different locations by variational modal decomposition (VMD) and improved discrete orthonormal Stockwell transform (IDOST), so as to extract the authentication information of the measurement data without losing the spatial features of the measurements. Combined with light convolutional neural network (LCNN) to evaluate the likelihood of measurement data being attacked by source ID to enhance the speed of detection response. The effectiveness of the method is verified by the detection results of actual multi-point synchronous measurement data.

• Short-term Wind Power Interval Prediction Based on Combined Mode Decomposition and Deep Learning

  XIAO Bai, ZHANG Bo, WANG Xinwei, GAO Ningyuan

  Available online:February 20, 2023  DOI: 10.7500/AEPS20220807002

  Abstract(236) HTML(618) PDF(3.51 M)( 167)

  Abstract:Wind power prediction is very important for the economic dispatch of power systems containing wind power. Aiming at the problem that point prediction is difficult to describe the uncertainty of wind power, a short-term wind power interval prediction method based on combined mode decomposition and deep learning is proposed. Firstly, the original wind power sequence is decomposed into multiple modal components by using the improved complete ensemble empirical mode decomposition with adaptive noise, and the high-frequency strong non-stationary components are decomposed again by using the variational mode decomposition. On this basis, the sample entropy is used to calculate the complexity of each component and reconstruct them into trend components, oscillation components and random components. Then, the three components are input into the Bayesian optimization bidirectional long short-term memory neural network to establish their respective prediction models, and the point prediction values of the three components are obtained. The mixed kernel density estimation method is used to estimate the error distribution of the prediction results of oscillation components and random components, and the overall interval prediction results are obtained by combining the point prediction values. Finally, the actual examples show that this method has higher prediction accuracy than other models.

• Named Entity Recognition for Electric Power Industry Based on Enhanced Text Features

  LIU Wensong, HU Zhuqing, ZHANG Jinhui, LIU Xuejing, LIN Feng, YU Jun

  Available online:September 27, 2022  DOI: 10.7500/AEPS20210323003

  Abstract(40) HTML(42) PDF(1.43 M)( 54)

  Abstract:Considering the characteristics of small scale, nested entities and abbreviated entities for electric corpus, the named entity recognition (NER) based on enhanced vectors of text features is proposed. Firstly, by the way of the low grain word segment and the preset dictionary, the semantic information in Chinese words is properly utilized, and the transmission errors caused by word segment are decreased. Secondly, the features of inner structure of a single Chinese word is learned by the word-level bidirectional gated recurrent unit (Word BiGRU). Together with the features of the part of speech for words and word length, the enhanced word vector is built by concatenating these features vectors with word vectors. Finally, the NER model is designed with BiGRU, attention mechanism and conditional random field (CRF). The proposed method is verified using electric corpus and F1 measurement reaches 87.02%, which proves the effectiveness of NER for electric power industry.

• Identification of Abnormal Operation Condition for Environmental Protection in Enterprises Based on Power Quality Monitoring Data

  ZHANG Yi, YAO Wenxu, SHAO Zhenguo, ZHANG Liangyu

  Available online:August 18, 2022  DOI: 10.7500/AEPS20211203007

  Abstract(88) HTML(136) PDF(4.68 M)( 219)

  Abstract:Aiming at the problems of abnormal operation condition monitoring for environmental protection in polluting enterprises at present, such as difficult implementation, large identification errors and easy tampering with the results, this paper proposes an identification method of abnormal operation conditions for environmental protection based on power quality monitoring data. The multi-dimensional power quality data obtained from non-invasive load monitoring at the public power entrance of enterprise equipment are used to train the model of condition classification, to realize abnormal condition identification, which is different from the existing scheme of power consumption monitoring with a separate meter installed for each device. First, the time series change-point detection and the clustering calculation for the characteristic data representing the production conditions are carried out to realize the division of production operation conditions of enterprises. Then, combined with the operation of environmental protection equipment, the categories of environmental protection operation conditions for classification are obtained. Furthermore, the operation condition scenarios related to environmental protection are classified and learned by the Stacking learning model. Finally, the trained classification model is used to identify the abnormal operation conditions for environmental protection in the enterprise. The effectiveness of the proposed method is verified by the simulation test data and the actual enterprise data.

• Causes and Suppression Measures of Resonant Overvoltage in Offshore Wind Power AC Transmission System

  YANG Daye, SONG Ruihua, XIANG Zutao, LIU Dong, CHAO Wujie, YAN Yuesheng

  Available online:May 26, 2022  DOI: 10.7500/AEPS20220126001

  Abstract(32) HTML(18) PDF(3.69 M)( 44)

  Abstract:Since the capacitance per unit length of cable is more than 20 times that of overhead lines with the same voltage level, more and more offshore wind power is connected to the grid through AC cable, which reduces the natural resonant frequency of the transmission system and increases the resonant risk of the system. Aiming at the resonant overvoltage phenomenon in the process of grid connection of an offshore wind farm, the fault recording data are analyzed. Based on the impedance model of the transmission system, the mechanism is analyzed, and through the electromagnetic transient simulation, it is verified that the natural resonant frequency near the double frequency of the transmission system is the fundamental reason for the resonant overvoltage caused by the operation of no-load transformer and no-load line in the wind farm. Combined with the engineering practice, the joint suppression measures of changing the operation mode of the transmission system, optimizing the control and protection system of the static var generator and increasing the access load are proposed. Finally, the effectiveness of the proposed measures is verified by simulation and field tests.

• Calculation of Power Supply Reliability for Distribution Network Based on Neo4j Graph Database

  JIANG Wei, WANG Minghua, CHEN Jinming, LIU Jiangdong, PU Shi, XU Zhiqi

  Available online:May 13, 2022  DOI: 10.7500/AEPS20211031001

  Abstract(118) HTML(214) PDF(1.74 M)( 101)

  Abstract:In order to improve the efficiency of power supply reliability calculation for the complex distribution system, a reliability calculation method based on the Neo4j graph database is proposed. Firstly, the topology structure of the distribution network is stored in vertex-edge form through the graph database. Meanwhile, the feeder classification and load partitioning in the complex distribution system is completed by using the characteristics of different types of edges in the Neo4j graph database, and the distribution network diagram model based on the graph database is built. Secondly, the subgraph division of the distribution network diagram model is combined with path search, and the model simplification is completed based on each subgraph. Finally, the power supply reliability analysis of the distribution system is realized based on the minimal path reliability algorithm combined with the efficient shortest path query and other functions of the Neo4j graph database. The effectiveness of the proposed method is verified by comparing the Roy Billinton test system and an actual 10 kV distribution network in China for algorithm verification.

• Flexible Ring Network Controller Without Interface Transformer for Medium Voltage Distribution Network

  LI Zheng, CHEN Wu, HOU Kai, SHI Mingming, MOU Xiaochun, ZHU Jinsong

  Available online:April 26, 2022  DOI: 10.7500/AEPS20210806005

  Abstract(137) HTML(172) PDF(2.92 M)( 102)

  Abstract:When the flexible ring network controller cancels the interface transformer, the transmission of the zero-sequence voltage component cannot be prevented when the AC side fails, thus increasing the fault range. Therefore, this paper uses classical circuit analysis and positive and negative sequence analysis algorithms to explain the basic principles of the formation and transmission of zero-sequence voltage components. A topology of the flexible ring network controller without the interface transformer is proposed. The AC side converters are all modular multilevel converters with traditional half-bridge sub-modules. And the full-bridge sub-module valve strings are connected in series on the positive and negative polarity busbars. Utilizing the ability of the full-bridge sub-module to output positive and negative voltages, the DC side voltage fluctuations are suppressed, and the fault range is prevented from expanding. By using the MATLAB/Simulink software, the characteristics of zero-sequence voltage suppressed during the fault are simulated and analyzed. The simulation results verify the correctness of the theoretical analysis and the effectiveness of the proposed topology.

• Electricity Theft Detection Based on Multiple Different Learners Fusion by Stacking Ensemble Learning

  YOU Wenxia, LI Qingqing, YANG Nan, SHEN Kun, LI Wenwu, WU Zeli

  Available online:March 30, 2022  DOI: 10.7500/AEPS20210731001

  Abstract(238) HTML(154) PDF(1.38 M)( 116)

  Abstract:Aiming at the problems that the consumer power consumption data categories are unbalanced in electricity theft detection, and the ensemble learning method using voting as a combination strategy can not give full play to the advantages of multiple different learners, a model using Stacking ensemble learning to fuse multiple different learners is proposed and applied to electricity theft detection. Firstly, starting from the factors affecting electricity metering, six electricity theft behavior modes are simulated according to five common electricity theft methods; Secondly, synthetic minority oversampling technique (SMOTE) is used to process the unbalanced power consumption data, and k-fold cross-validation method is used to divide the balanced training sets to alleviate the overfitting caused by repeated learning; Then, the evaluation indicators and diversity metrics are employed to optimize different primary learners and meta-learners of the model, and a Stacking combination learning electricity theft detection model integrating the advantages and differences of different learners is constructed; Finally, the comparative analysis results of examples show that the proposed electricity theft detection model can effectively solve the imbalance of power consumption data categories, give full play to the advantages of different learners, and the evaluation index is good.

• Topology Optimization of Power Systems Combining Deep Reinforcement Learning and Domain Knowledge

  YAN Ziming, XU Yan

  Available online:December 01, 2021  DOI: 10.7500/AEPS20210510001

  Abstract(92) HTML(68) PDF(1.43 M)( 186)

  Abstract:While the flexibility of power systems operation can be improved by topology optimization, the dimension of system-level discrete decision variables, including the connections of lines and substation busbars, is prohibitively high. Thus, the topology optimization problem of power systems can hardly be solved by the conventional mixed-integer optimization method. Aiming at this problem, a reinforcement learning based method is proposed combining asynchronous advantage actor-critic and power system domain knowledge, which transfers the computational burden of online optimization to the offline agent training stage. The defined reward function is adopted to minimize the violations of power transmission line flow limits. Forced constraints verification is employed to reduce the searching space and improve the efficiency of the reinforcement learning. The fast computation of the topological structure optimization of power system operation is realized,and the operation security of power systems is enhanced. The effectiveness of the proposed method is validated by simulation testing results.