Combination Forecasting Model for Mid-long Term Load Based on Least Squares Support Vector Machines and a Mended Particle Swarm Optimization Algorithm

Mid-long term load forecasting (MTLF) plays an important role in power system. With more factors involved, single forecasting method becomes hard to satisfy requirement. This paper proposes a new combination model for MTLF based on least squares support vector machines (LS-SVM) and particle swarm optimization (PSO) algorithm. LS-SVM is a new kind of SVM which regresses faster than standard, and a mended particle swarm optimization (MPSO) algorithm is employed to optimize the parameters of LS-SVM. With a real case test, the result shows proposed model outperforms tradition combination model.