Short-term Load Forecasting Based on Multiple PSO-LSSVM under Electricity Market Environment

Abstract

Affected by the development of the electricity market and many other factors, it is very difficult to obtain high-performance short-term load forecasting methods. The day-ahead load forecasting of the power grid is deeply studied by dividing 24 hours forecast outputs into several groups and optimizing them respectively. As a result, a new short-term load forecasting method is proposed by considering the load variation trend in different periods. First, through the historical load variation trend analysis, the 24 hours load forecasting output is divided into several sets. Then, the prediction algorithm combining multi-group particle swarm optimization and least squares support vector machine is proposed. Finally, the improved algorithm and the original particle swarm optimization, and the least squares support vector machine algorithm are used to predict the short-term load data of Jiangxi Province. The experiment results show that the improved algorithm can identify the load characteristics of a different time with high prediction accuracy, which has practical application value.