基于粒子群算法的BP神经网络电价预测研究  

A Study of Electricity Price Forecasting by BP Neural Network Based on Particle Swarm Optimization(PSO)

作　　者：鲁娅楠 王金梅 孙帆 

机构地区：[1]宁夏大学物理与电子电气工程学院,宁夏银川750021 [2]宁夏沙漠信息智能感知自治区重点实验室,宁夏银川750021

出　　处：《科技创新与应用》2018年第28期15-17,共3页Technology Innovation and Application

基　　金：国家自然科学基金项目(编号:51167015)

摘　　要：电力负荷预测是电力系统规划和运行的主要内容,而实时电价是影响负荷预测精度的一个重要因素,文章通过分析某电网电价历史数据,结合PSO算法和BP网络优点,提出一种PSO-BP神经网络预测模型,用PSO算法优化BP神经网络的初始权值和阈值,将电力系统电价的不确定性变为可预测性。Matlab仿真结果表明,PSO-BP神经网络预测模型收敛速度快和预测精度高,可运用到未来实际电价预测当中。Power load forecasting is the main content of power system planning and operation, and the spot price is an important factor that affects the precision of load forecasting. This paper analyzes the historical data of electricity price in a power network and combines the advantages of PSO algorithm and BP network. A forecasting model of PSO-BP neural network is proposed. The PSO algorithm is used to optimize the initial weights and thresholds of the BP neural network, and the uncertainty of the electricity price in the power system can be changed into predictability. The Matlab simulation results show that the PSO-BP neural network prediction model has fast convergence speed and high forecasting accuracy, which can be applied to the actual electricity- price fore-casting in the future.

关 键 词：BP神经网络 粒子群算法 优化算法 电力负荷预测 

分 类 号：TM71[电气工程—电力系统及自动化]