Neural network based on dynamic multi-swarm particle swarm optimizer for ultra-short-term load forecasting

Abstract

Ultra-Short-Term Load Forecasting plays an important role in Power Load Forecasting. Back Propagation Neural Network(BPNN) has become one of the most commonly used methods in Power System Ultra-Short-Term Load Forecasting for its ability of computing complex samples and training largescale samples. However, traditional BPNN algorithm needs to set up a large amount of network training parameters, and it is easy to be trapped into local optima. A new algorithm which is Neural Network based on Dynamic Multi- Swarm Particle Swarm Optimizer (DMSPSO-NN) is proposed for Ultra-Short- Term Load Forecasting in this paper. DMSPSO-NN overcomes the shortage of traditional BPNN and has a good global search and higher accuracy which shows that it is suitable to be used for Ultra-Short-Term Load Forecasting.