Aiming at the problems of strong randomness, poor stability and unsatisfactory forecasting accuracy of power load, a short-term power load forecasting method combining particle swarm optimization (PSO) and least squares support vector machine (LS-SVM) is proposed in this paper. The input factors of the model are load data and meteorological information. The particle swarm optimization algorithm is adopted to realize the automatic optimization of the parameter of the support vector machine, short-term load forecasting model of the least squares support vector machine based on particle swarm optimization is established. The accuracy and validity of the improved prediction model are verified by simulation, the results show that the improved prediction method brings benefits to convergence, prediction accuracy and training speed. This study provides a reference for the development of short-term load forecasting methods in China.