In order to improve the accuracy of power load forecasting, a short-term load forecasting model based on lifting wavelet and improved PSO-Elman neural network is proposed. Firstly, aiming at the fluctuation and trend of load, the lifting wavelet algorithm is used to decompose the original load data and extract its main features. Then, in the improved particle swarm optimization (GPSO) algorithm of ant colony algorithm, chaos theory is used to disturb some particles with poor fitness. The CGPSO algorithm is proposed to improve meticulousness. The accuracy of the search and the ability of global search are improved. The CGPSO algorithm is used to optimize the initial parameters of Elman neural network. Finally, the load forecasting model is established. In this paper, the actual data of a certain area in northern China are used to simulate. The experimental results show that the prediction accuracy of this method is 2.362 6% higher than that of the traditional ENN method.
