Forecasting of Electricity Load Based on Particle Swarm Optimization and Support Vector Machine

The values of training parameters of support vector machine have close contact with its forecasting accuracy.Therefore, particle swarm optimization algorithm and support vector machine (PSVM) is proposed to predict electricity consumption in the study.In the model, particle swarm optimization algorithm is used to select the optimal training parameters of support vector machine.The electricity consumption data and relevant features data of JiangXi province from July to October in 2008 are used as the experimental data.The experimental results indicate that the PSVM model has higher prediction accuracy than BP neural network in the forecasting of electricity load.