基于向量加权平均算法优化最小二乘支持向量机的电价短期预测  

Short-term electricity price forecasting based on least squares support vector machine optimized by weighted mean of vectors algorithm

作　　者：陈晓华 吴杰康[2] 杨国荣 CHEN Xiaohua;WU Jiekang;YANG Guorong(Zhanjiang Power Supply Bureau of Guangdong Power Grid Co.,Ltd.,Zhanjiang 524005,China;School of Automation,Guangdong University of Technology,Guangzhou 510006,China)

机构地区：[1]广东电网有限责任公司湛江供电局,广东湛江524005 [2]广东工业大学自动化学院,广州510006

出　　处：《黑龙江电力》2025年第1期1-7,共7页Heilongjiang Electric Power

基　　金：国家自然科学基金项目(项目编号:50767001)。

摘　　要：针对电价短期预测精度低等问题,提出一种基于向量加权平均算法优化最小二乘支持向量机的电价短期预测模型。将电价的历史数据归一化后作为输入变量;利用INFO优化LSSVM的惩罚因子和核函数参数,从而利用最优的参数值构建INFO-LSSVM预测模型;选取某地区2010年1月1日-15日的电力价格数据进行分析。仿真结果表明:与核极限学习机、长短期记忆神经网络、LSSVM预测模型相比,INFO-LSSVM预测模型的预测效果更好;利用果蝇优化算法优化LSSVM的惩罚因子和核函数参数构建FOA-LSSVM预测模型的预测效果不及INFO-LSSVM预测模型,并且INFO的收敛速度比FOA快。通过与对照预测模型对比表明,INFO-LSSVM预测模型具有更好的预测性能。Aiming at the problem of low accuracy of short-term electricity price forecasting,proposes a short-term electricity price forecasting model based on least squares support vector machine optimized by weighted mean of vectors algorithm.The historical data of the electricity price is normalized as an input variable.INFO is used to optimize the penalty factor and kernel function parameters of LSSVM,and the INFO-LSSVM prediction model is constructed by using the optimal parameter values.The electricity price data of a certain area in Australia from January 1 to 15,2010 is selected for analysis.The simulation results show that compared with the kernel extreme learning machine,long-term and short-term memory neural network and LSSVM prediction model,the prediction effect of INFO-LSSVM prediction model is better,and the prediction effect of FOA-LSSVM prediction model constructed by using fruit fly optimization algorithm optimize the penalty factor and kernel function parameters of LSSVM is not as good as that of INFO-LSSVM prediction model,and the convergence speed of INFO is faster than that of FOA.Compared with the control prediction model,the INFO-LSSVM prediction model has better prediction performance.

关 键 词：向量加权平均算法 最小二乘支持向量机 电价预测 短期预测 INFO-LSSVM预测模型 

分 类 号：TM744[电气工程—电力系统及自动化]