基于模糊支持向量机的电力系统中期负荷预测

考虑到气温因素对电力系统负荷的影响,提出了一种模糊支持向量机(FSVM)算法,基于欧洲智能技术网络(EUNITE)竞赛数据进行了中期电力负荷预测的应用研究。首先利用隶属度函数对影响负荷的气温因素进行模糊化处理,然后结合已知数据得到支持向量机(SVM)的训练样本集,采用序列极小优化(SMO)算法实现对支持向量机(SVM)的快速训练,最终得到预测结果,并与不将气温模糊化的仿真结果进行比较,表明本文所提方法简便且预测精度较高。

Power system mid-term load forecasting based on fuzzy support vector machines

Considering the infection of temperatures to loads,the text advanced an arithmetic called Fuzzy Support Vector Machines(FSVM).This arithmetic was applied for load forecasting with the data of EUNITE competition.Firstly,fuzzied the temperatures that infecting loads,then gained the training data based on the known data.Using Sequential Minimal Optimization(SMO) arithmetic achieved the fast training to Support Vector Machines(SVM).The results show that the arithmetic advanced was simple and highly accurate.