A Novel Hybrid GA Based SVM Short Term Load Forecasting Model

The increasing importance and complexity of STLF necessitates more accurate load forecast methods. A novel genetic algorithm (GA) based support vector machine (SVM) forecasting model with determinstic annealing (DA) clustering is presented in this paper. For NN forecasting, too many training data may lead to long training time and slow convergent speed. First deterministic annealing (DA)for load data clustering technique is adopted first to solve the problem. After data clustering, GA based SVM forecasting model is established. The parameters for SVM were optimized through genetic algorithms, which were used in SVM model. The hibrid GA-SVM forecasting model is tested by using Hebei Province practical load data. The experimental results demonstrate the GA-SVM model outperforms the BP neural network model based on the root mean square error (RMSE) and the mean absolute percentage error (MAPE). And the proposed method provided a satisfactory improvement of forecasting accuracy.