反向传播算法运用于电力系统神经网络的分析研究

着重研究了神经网络模型中的反向传播算法即BP算法，并对电力系统中两个 一次结线特殊的厂站的检修批答神经网络建立了BP模型，分析了BP算法中 初始权值、学习因子、冲量因子、隐含层数和隐含层节点数对该神经网络学习 过程的影响，实验结果证明，适当地选择初始权值、学习因子、隐含层节点数 等可以大大提高神经网络的学习速度,减少迭代时间，满足电力系统神经网络在线 学习的要求。

ANALYSIS OF THE BACK PROPAGATION METHOD APPLIED TO THE NEURAL NETWORK IN POWER SYSTEM

Application of the back propagation method in the Neural Network is studied in this paper. The emphasis is put on the investigation of effects of the initial weighting values, learning factor, accelerating factor, hidden layers and the number of the nodes in the hidden layer on the learning process, including the converging speed and the converging property. Based on the results obtained from the studies, two neural networks are developed for the purpose of a practically used expert system, in which the Neural Networks are used for equipment overhauling arrangement in two specially constructed power plants.Application of the neural network in the power system shows that appropriate selection of the initial weighting values, learning factor, accelerating factors, hidden layers and the number of the nodes in the hidden layers, can greatly increase the learning speed. This helps to realize an on-line real- time neural network in power system.