摘要
针对常用BP神经网络算法进行电力系统负荷预测时容易陷入局部最优的缺点,建立一个基于GA-BP神经网络的电力系统负荷预测模型。将负荷数据、最高温度和最低温度作为输入变量,电力系统负荷数据作为输出变量,对某小区一个季度的负荷数据进行仿真分析。经实验验证,在GA-BP电力系统负荷预测模型下的电力系统负荷预测的平均误差比BP神经网络算法负荷预测的平均误差降低了3.4%,因此采用GA-BP神经网络进行电力系统负荷预测具有可行性和有效性,可以提高电力系统负荷预测的精度。
Aiming at the disadvantage that the commonly used BP neural network algorithm is apt to fall into local optimum in power system load forecasting,a load forecasting model of power system based on GA-BP neural network is established.By using load data,maximum temperature and minimum temperature as input variables,and power system load data as output variables,the load data of a quarter of a residential area are simulated and analyzed.The experimental results show that the average error of power system load forecasting under GA-BP load forecasting model is 3.4%lower than that of BP neural network algorithm.Therefore,it is feasible and effective to use GA-BP neural network for power system load forecasting,which can improve the accuracy of power system load forecasting.
作者
周杰
张矿伟
金龙奎
ZHOU Jie;ZHANG Kuang-wei;JIN Long-kui(Zhoukou Senior Technical School Training Center,Zhoukou 466000 China;School of Physics and Electronic and Engineering,Yuxi Normal University,Yuxi 653100 China)
出处
《科技创新与生产力》
2019年第10期61-63,67,共4页
Sci-tech Innovation and Productivity
基金
云南省应用基础研究项目(2018FD093)
云南省大学生创新项目(2018A03)
关键词
电力系统
负荷预测
BP神经网络
遗传算法
GA-BP
power system
load forecasting
back-propagation neural network
genetic algorithm
GA-BP
