Research on Combination Forecasting Method of Short-Term Electric Load in Nuclear Power Industrial Park

Accurate short-term electric load forecasting results are of important reference value to the energy system scheduling optimization and economic operation in nuclear power industrial parks. However, meteorological conditions, consumption habits, work and rest time and other factors will cause short-term load fluctuations, a single forecasting model is hard to ensure the accuracy and stability of the forecast. To solve this problem, a method of constructing FA-EEMD-ELM combined model is proposed in this paper. Combining Firefly Algorithm (FA) and Ensemble Empirical Mode Decomposition (EEMD) algorithm, the parameters of the Extreme Learning Machine, ELM) model is optimized. The electric load data set of a nuclear power industrial park was used to test and verify the combined forecasting model. The results show that the combined model of FA-EEMD-ELM has an ideal forecasting effect. Compared with the three models of ELM, EEMD-ELM and FA-ELM, it has the advantages of better forecasting stability and higher accuracy.