文章摘要
李得民,吴在军,赵波,等.基于模型预测控制的孤岛微电网二次调节策略[J].电力系统自动化,2019,43(10):60-67. DOI: 10.7500/AEPS20180710002.
LI Demin,WU Zaijun,ZHAO Bo, et al.Secondary Regulation Strategy for Islanded Microgrid Based on Model Predictive Control[J].Automation of Electric Power Systems,2019,43(10):60-67. DOI: 10.7500/AEPS20180710002.
基于模型预测控制的孤岛微电网二次调节策略
Secondary Regulation Strategy for Islanded Microgrid Based on Model Predictive Control
DOI:10.7500/AEPS20180710002
关键词: 模型预测控制  二次调节策略  状态空间模型  滚动优化  反馈校正
KeyWords: model predictive control  secondary regulation strategy  state-space model  rolling optimization  feedback correction
上网日期:2019-01-31
基金项目:国家自然科学基金资助项目(51877039);江苏省“六大人才高峰”项目(XNY-005)
作者单位E-mail
李得民 东南大学电气工程学院, 江苏省南京市 210096  
吴在军 东南大学电气工程学院, 江苏省南京市 210096 zjwu@seu.edu.cn 
赵波 国网浙江省电力有限公司电力科学研究院, 浙江省杭州市 310014  
张雪松 国网浙江省电力有限公司电力科学研究院, 浙江省杭州市 310014  
章雷其 国网浙江省电力有限公司电力科学研究院, 浙江省杭州市 310014  
摘要:
      受线路阻抗的影响,传统的P-f和Q-V下垂控制无法实现无功功率按比例精确分配,且所有分布式电源输出电压的幅值偏离电压基准值;负荷波动时,频率无法维持在额定频率值。提出一种基于模型预测控制(MPC)的二次调节策略,通过优化下垂控制的频率和电压参考值,将频率和平均电压维持在各自的基准值,实现无功功率按额定容量比例精确分配的目标,并减小了输出电压与基准电压间的偏移量。该策略以一定的采样周期将分布式电源的状态空间模型离散化,经滚动优化和反馈校正获得频率和电压的控制变量,分别添加到P-f和Q-V初级控制中,实现优化频率与电压的二次调节,且减少了比例-积分(PI)控制参数,降低了对系统参数的敏感度。MATLAB/Simulink仿真验证了所提策略与基于PI控制的二次调节策略相比,在负荷波动时具有更快的瞬态响应特性,同时具有较强的鲁棒性。
Abstract:
      Due to the influence of line impedance, the conventional P-f and Q-V droop control cannot realize the target of accurate reactive power sharing proportional to their rated capacity and the output voltage amplitudes of all distributed generators(DGs)deviate from nominal voltage. The frequency cannot be maintained at the rated frequency value when load fluctuates. A secondary regulation strategy based on model predictive control(MPC)is proposed. By optimizing the frequency and voltage references of droop control, respectively, the frequency and average system voltage are always maintained at their respective nominal values. Meanwhile, the target of accurate reactive power sharing proportional to their rated capacity is realized and the output voltage amplitude deviation from nominal voltage is reduced. The state-space model of DG is discretized with a certain sampling period in this strategy. The control variables of frequency and voltage are obtained by rolling optimization and feedback correction and then added to P-f and Q-V primary controls, respectively, to realize the secondary regulation for frequency and voltage optimization. The number of proportional-integral(PI)control parameters decreases and the sensitivity to the system parameters is reduced. The MATLAB/Simulink simulation verifies that the proposed strategy has faster transient response characteristics and stronger robustness when load fluctuates compared with the PI based secondary regulation strategy.
查看全文(Free!)   查看附录   查看/发表评论  下载PDF阅读器