含分布式电源点的区域配电网三相线损测试研究

Research on Three-Phase Line Loss Test of the Regional Distribution Network with Distributed Power Points

当前区域配电网三相线损测试难度大,为解决该问题,提出了基于神经网络的含分布式电源点的区域配电网三相线损测试方法。根据神经网络结构,分析目标不同层次特征,在特征响应支持下,分析三相四线制连接方式下的线损。利用神经网络的监视控制层实时采集线损数据,通过粒子群算法对神经网络进行优化,确定用于全局搜索的神经网络阈值,设定粒子群算法收敛条件,重新赋予线损测试值,动态调整学习因子,评估种群适应度,由此设计三相线损测试流程,实现含分布式电源点的区域配电网三相线损测试。实验结果表明,该方法测试干扰幅值较低,测试精准度高,有助于降低线损。

At present,it is difficult to test the three-phase line loss of a regional distribution network. To solve this problem,a method based on the neural network is proposed to test the threephase line loss of the regional distribution network with distributed power points. According to the structure of the neural network,the characteristics of different levels of the target are analyzed. Under the support of the characteristic response,the line loss under the three-phase four-wire connection mode is analyzed. The monitoring control layer of the neural network is used to collect the line loss data in real time;the neural network is optimized using particle swarm algorithm;the threshold value of the neural network used for global search is determined;the convergence condition of particle swarm algorithm is set;and the line loss test values are reassigned,learning factors adjusted dynamically and the population fitness evaluated to design the test procedure of the three-phase line loss and thus realize the three-phase line loss test of the regional distribution network with distributed power points. The experimental results show that this method has low interference amplitude and high accuracy,which is helpful to reduce the line loss.