摘要
在电气化铁道的混合有源补偿电流平衡装置中,为了减少谐波对PWM逆变装置的影响,采用感应滤波变压器代替输出变压器。感应滤波变压器的滤波绕组所接的LC滤波器补偿一定的无功,消除大部分主要次谐波,使牵引变电所的谐波,或者回灌的谐波电流对PWM逆变器的直流侧电压抬升的影响减小,有助于PWM逆变装置的稳定。为了消除牵引变电站的负序电流,提出了基于神经网络的PSO-BP算法来检测有功电流,采用对电源电流的直接控制方法来平衡两个牵引臂的有功电流,并动态补偿无功和谐波电流,使牵引变压器两臂输出的有功电流相等。本文详细介绍了该装置的拓扑结构和工作原理,仿真和实验结果表明这种装置能够有效地改善电网质量。
In order to reduce the influence of harmonic on PWM inverter device in the hybrid active compensation device for current balance applied to electrical railway system, the inductive filtering transformer is used instead of the output transformer. The filtering winding of the inductive filtering transformer accessed to LC filters can compensate most reactive power and main character harmonic currents, its feature can avoid voltage harmonic or reflowing current influence on DC side voltage rise, and help the PWM inverter device stability. In order to eliminate the negative sequence current of traction substation, the PSO-BP algorithm based on neural network is proposed to detect active current, direct control method for power supply current to balance the active current of two traction arms, and dynamic compensation reactive power and harmonic currents, the active current of traction transformer's two output arms was equal. This paper introduces the topology and principle of the device, and the simulation and experimental resuls show that this device can effectively improve the quality of power grid.
出处
《电工技术学报》
EI
CSCD
北大核心
2013年第8期268-275,共8页
Transactions of China Electrotechnical Society
基金
湖南省自然科学基金(11JJ9013)
湖南省科技厅(2012FJ4140
2012GK3099)
湖南省教育厅(11K027)资助项目
关键词
感应滤波变压器
负序电流
神经网络PSO-BP算法
混合有源补偿装置
The inductive filtering transformer, negative sequence current, neural network PSO-Bpalgorithm, hybird active compensation device