Short-Term Electric Load Forecasting Based on SAPSO-ANN Algorithm

For the economical, secure and stable operation of the electric power system, the short-term load forecasting plays a vital role. The paper applies the SAPSO-ANN model to forecast the short-term electric load. In order to enhance the generality of the model, as well as the capabilities of training and learning in the forecasting, the capability of searching the optimum in the overall situations in the PSO algorithm has been strengthened by using the SA algorithm which has the characteristic of global optimization, and the learning algorithm of a typical three-layer feed-forward neural network BP has been replaced by the mixed PSO algorithm. Taking actual load data of a power grid company in South China as a sample, the PSO-ANN model has been compared to the traditional model. The results show that this model has better capability of forecasting and network learning.