文章摘要
苏小林,刘孝杰,阎晓霞,等.计及需求响应的主动配电网短期负荷预测[J].电力系统自动化,2018,42(10):60-66. DOI: 10.7500/AEPS20170829008.
SU Xiaolin,LIU Xiaojie,YAN Xiaoxia, et al.Short-term Load Forecasting of Active Distribution Network Based on Demand Response[J].Automation of Electric Power Systems,2018,42(10):60-66. DOI: 10.7500/AEPS20170829008.
计及需求响应的主动配电网短期负荷预测
Short-term Load Forecasting of Active Distribution Network Based on Demand Response
DOI:10.7500/AEPS20170829008
关键词: 负荷预测  负荷分解  主动配电网  需求响应  组合预测模型
KeyWords: load forecasting  load decomposition  active distribution network  demand response  combination forecast model
上网日期:2018-03-06
基金项目:
作者单位E-mail
苏小林 山西大学电力工程系, 山西省太原市 030006 xiaolin5759@sina.com 
刘孝杰 国网大同供电公司, 山西省大同市 037008  
阎晓霞 山西大学电力工程系, 山西省太原市 030006  
王穆青 山西大学电力工程系, 山西省太原市 030006  
韩学楠 山西大学电力工程系, 山西省太原市 030006  
摘要:
      随着分布式电源、电动汽车及储能等广义需求响应资源的接入,用户在电力市场各种激励影响下进行需求响应,将改变负荷特性并影响负荷预测。根据需求响应计划信号的可预知性及季节性基础负荷的独立性,利用小波分解等方法对主动配电网负荷在不同层面上进行了分解,形成季节性基础负荷和需求响应信号及各种气象因素作用的负荷部分,利用时间序列模型对季节性基础负荷进行预测,利用支持向量回归模型对需求响应信号及气象因素影响的负荷部分进行预测,形成组合预测模型,两部分预测负荷叠加得到总负荷。利用线性时变模型仿真的主动配电网负荷数据算例,进行了预测测试与分析,通过与其他方法相比较,证明了所提方法预测计及需求响应的主动配电网负荷的有效性及精确度。
Abstract:
      General demand response resources, such as distributed generators, electric vehicles and energy storage, which increasingly access to power consumer side of power distribution systems and take demand response under the incentive of the power market, will change the load characteristics and load profile and affect power load forecasting. Based on predictability of the demand response planning signals and independence of seasonal basis load, the historical load data of active distribution network is decomposed into the seasonal basis load and the other load associated with demand response signals and meteorological factors by applying the wavelet decomposition method. The time series forecasting model is applied to forecast seasonal basis load, and support vector regression model is applied to forecast the related load with demand response and meteorological factors. Then, the total load can be obtained with the combination forecasting model. The load data of active distribution network resulted from simulation of linear time-varying model is used to forecast load profile with different load forecasting methods, and the example results demonstrate that the presented load forecasting method is more effective and accurate for load forecasting of active distribution network including demand response.
查看全文(Free!)   查看附录   查看/发表评论  下载PDF阅读器