摘要
为了解决现有电价预测中BP神经网络法对初始权值敏感、易陷入局部最小值和收敛速度慢等问题,在神经网络训练中引入基于全局随机优化思想的粒子群优化(PSO)算法,先利用PSO优化BP神经网络的初始权值,然后采用神经网络完成给定精度的学习,建立了粒子群-BP神经网络模型.与传统BP神经网络、粒子群广义神经网络相比,该方法收敛速度快、所需历史数据少、预报精度高,可用于电力系统的短期电价预测.
In order to improve the problem in BP neural network of electricity price forecast at present which is sensitive with the initial weights, easy to fall into the local least value and have slow convergence speed etc. , the particle swarm optimization （PSO） algorithm based on the random global optimization is inducted into the network training; the particle swarm optimization algorithm is used for glancing study in order to confirm the initial values; then the neural network is used for given accuracy to found the PSO-BP neural network model. Comparing to the traditional BP neural network and PSO generalized regression neural network, the PSO-BP neural network model has the merits of faster convergence, needing fewer historical data, higher forecast precision etc. ; and it can be used in the short-term electricity price forecast of electric power systems.
出处
《武汉大学学报（工学版）》
CAS
CSCD
北大核心
2008年第4期102-105,共4页
Engineering Journal of Wuhan University
基金
陕西省教育厅专项科研计划项目(编号:05JK266)
关键词
电力市场
电价预测
粒子群算法
BP神经网络
electric power market
electricity price forecast
particle swarm optimization algorithm
BP neural network
