基于传统BP神经网络和经云遗传优化的神经网络配电网故障定位研究

摘要

针对配电网故障定位问题提出一种基于人工神经网络(ANN)中结构较为简单并且可塑性强的误差反向传播(Error Back Propagation,BP)神经网络方法的定位模型.建立BP网络模型,并将训练好的BP网络模型和通过云遗传算法改进后的BP网络模型,应用于同一个简单的配电网系统中,分别对不同分支的反射信息进行特征提取与模式识别.通过对两种算法的训练曲线图和诊断精度的比较来反映优化算法的高效性和准确性,最终得以确定诊断的实际输出值,实现故障分支的判别和精确定位.%Aiming at the problem of distribution network fault location, a positioning model based on BP neural network with simple structure and strong plasticity in artificial neural network(ANN)is proposed. BP network model is established and the trained BP network model is improved and applied to the same simple distribution network system through cloud genetic algorithm. Feature extraction and pattern recognition are respectively performed on the reflection information of different branches. By comparing the two algorithms The comparison between the training curve and the diagnostic accuracy reflects the efficiency and accuracy of the optimization algorithm and finally determines the actual output value of the diagnosis to determine and pinpoint the fault branch.