电网技术2018，Vol.42Issue(2)：585-590,6.DOI:10.13335/j.1000-3673.pst.2017.1403
基于MFOA-GRNN模型的年电力负荷预测
An Annual Load Forecasting Model Based on Generalized Regression Neural Network With Multi-Swarm Fruit Fly Optimization Algorithm
李冬辉 1尹海燕 1郑博文1
作者信息
- 1. 天津大学电气自动化与信息工程学院,天津市南开区300072
- 折叠
摘要
Abstract
Accurate annual power load forecasting can provide reliable guidance for power grid operation and power construction planning.Influenced by various factors,an annual load curve shows non-linear characteristics.Therefore,solution of annual load forecasting problem needs to be based on a nonlinear model.Generalized regression neural network (GRNN) is proven effective in dealing with non-linear problems.There is only one spread parameter in the network.And determination of appropriate spread parameters with GRNN for power load forecasting is a key point.This paper proposes a hybrid load forecasting model combining multi-swarm fruit fly optimization algorithm (MFOA) and GRNN to deal with this problem,where MFOA is used to choose the appropriate parameters for GRNN load forecasting model.Finally,according to analysis of experimental data based on MFOA-GRNN model,the mean absolute percentage error (MAPE) is 0.510％,and the mean square error (MSE) is 0.281.Results of MFOA-GRNN forecasting model were compared with support vector regression with differential evolution algorithm (DE-SVM),GRNN model with particle swarm optimization (PSO-GRNN) and generalized regression neural network with fruit fly optimization algorithm (FOA-GRNN).The comparison shows that the proposed hybrid model outperforms other three forecasting models in annual power load forecasting.关键词
年电力负荷预测/广义回归神经网络/参数优化/多种群/果蝇优化算法/相对误差Key words
annual power load forecasting/generalized regression neural network/parameter optimization/multi-swarm/fruit fly optimization algorithm/relative error分类
信息技术与安全科学引用本文复制引用
李冬辉,尹海燕,郑博文..基于MFOA-GRNN模型的年电力负荷预测[J].电网技术,2018,42(2):585-590,6.
