Research on a hybrid LSSVM intelligent algorithm in short term load forecasting

Abstract

For enhancing the prediction accuracy of power load forecasting, a support vector machine (SVM) prediction model based on wavelet transform and the mutant fruit fly parameter optimization intelligent algorithm (WT-MFOA-LSSVM) was presented. The load data were pretreated by wavelet transform, and the load curves were decomposed into different scales, in order to strengthen the regularity and randomness of historical data. Aiming at overcoming the problems of low convergence precision and easily relapsing into local extreme in basic fruit fly optimization algorithm (FOA), judge whether the intelligent algorithm was trapped in local extreme from the fitness variance of the population and the current optimal. Then, it was conducted by optimal individual perturbation and Gauss mutation operation and the mutant fruit flies were second times optimized, which made the accuracy of prediction model be obviously enhanced. The next few days of historical load data of a certain area of Henan Province, China, in 2015 were predicted by using WT-MFOA-LSSVM, and then the prediction results were compared with the results predicted by the SVM model and by the SVM model based on particle swarm optimization model. The results showed that WT-MFOA-LSSVM had high precision in short term load forecasting, and it had a very good practical significance.