An integrated PSO for parameter determination and feature selection of SVR and its application in STLF

A novel support vector regression (SVR) optimized by an integrated particle swarm optimization (PSO) was proposed. The optimization mechanism combined the discrete-valued PSO with the continuous-valued PSO to optimize the input feature subset selection and the SVR kernel parameter setting. By incorporating two types of PSO, the parameters and the input features of SVR were optimized simultaneously. Based on the operational data provided by a regional power grid in north China, the method was used in short-term load forecasting (STLF). The experimental results showed the proposed approach can correctly select the discriminating input features. Compared with the simple PSO-SVR without feature selection and the traditional SVR, the average time of the proposed method in the experimental process reduced and the forecasting accuracy increased respectively. Therefore, the hybrid method is better than the other two models.