基于BP神经网络的智能电网配电系统改进算法的研究

Research on improved algorithm based on BP neural network in smart grid distribution system

提出一种基于BP神经网络的智能电网配电系统改进算法。由于BP网络是一种按误差逆传播算法训练的多层前馈网络,具有学习性,可以根据已有的配电参数样本集进行训练,从中分析出内蒙古各地区根据时间不同所配电的分配情况的内在联系,实现对以后配电系统进行自适应控制。该算法的优点就是在构造过程考虑了BP的预测精度和训练时间,采用了梯度下降法的方法,进行Matlab仿真实验,获得了较为准确的预测结果。

This paper presents an improved algorithm based on BP neural network in smart grid distribution system. BP network, with the learning ability, is the error backpropagation algorithm for training multilayer feed forward network. It can be trained by the distribution parameters of the sample set, and the internal relations of power distribution status according to different time among several counties of Inner Mongolia can be analyzed, and thus the power distribution system can achieve adaptive control. The advantage of this algorithm is that the algorithm considers BP＇s prediction accuracy and training time during the construction and adopts gradient descent method. The performance of the system is verified by Matlab simulation and finally, more accurate predicition results are obtained.