Research On Electric Power System Short-Term Load Forecasting Using BP-PSO Network

Power load forecasting is one of great importance for the power department. To improve the accuracy of the load forecasting can help us to make the scientific power construction planning and to make the reasonable arrangement for the operation of the electric power consumption. With the development of the power system, the traditional electric power load forecasting method has been difficult to satisfy the demand on accuracy.The basic principle and characteristics of power load forecasting are introduced in the paper firstly. Several parameters which can affect the accuracy of the prediction algorithm have been analyzed. On the basis of the conclusions above, the procedure of the prediction are presented.Secondly, the particle swarm optimization (PSO) algorithm are described. And then an novel PSO algorithm has been proposed to improve the convergence speed, the accuracy and to aviod falling into local minima. Furthermore, based on the analysis above, the factors such as the weather, date, and the temperature etc. are also considered. Then a BP neural network combined with the improved particle swarm optimization algorithm are proposed to predict the power short-term load. Simulation results show that the method can accelarate the network learning speed, and to improve the precision of the prediction.