基于模糊神经网络的MCR电压无功控制

针对磁控电抗器MCR恒压/恒功率因数控制模式无法兼顾电压及功率因数的不足,提出综合考虑电压及功率因数的模糊神经网络加权控制策略,详细阐述模糊神经网络控制器的结构及其参数调整方法,设计由恒压和恒功率因数模糊神经网络控制器组成的控制系统,控制MCR输出的感性无功。算例仿真结果表明,该控制系统具有良好的负载跟踪特性,能实时跟踪电网电压及功率因数的变化;与传统PI闭环控制模式相比,具有超调量更小及响应时间更短等特点。

Fuzzy-neural network based MCR voltage and reactive power control

The magnetic control reactor （MCR）control mode with constant voltage or constant power factor cannot consider both of voltage and power factor.This paper proposed a fuzzy-neural network control strategy with voltage and power factor comprehensively considered,and presented the structure of fuzzy-neural network controller and its parameter adjustment method.The designed control system was mainly composed of fuzzy-neural network based controllers with constant voltage and constant power factor,and the output perceptual reactive power of MCR was controlled.Some cases simulation results showed that the designed control system had good load tracing feature,which can track the change of power system voltage and power factor in real time.Compared with traditional PI close-loop control mode,it had a smaller overshoot volume and shorter response time.