Short-term Load Forecasting for Power System Based on IWOA-ELM

Accurate prediction of short-term power loads can guarantee the security of power system operation and improve the utilization rate of electric energy. To obtain accurate and reliable short-term load prediction results, an improved whale optimization algorithm(IWOA) is proposed to optimize the extreme learning machine(ELM) short-term power load combination prediction model. First, the population of the whale algorithm is initialized using Logistic chaos mapping, and the search ability of the whale algorithm at each stage is enhanced by a nonlinear convergence factor, to realize the optimization of the input weights and deviations of the ELM. Then the IWOA-ELM prediction model is established based on the optimized parameters to predict the dataset and output the final load prediction results. Finally, the actual load data of a region in Gansu Province is used as a dataset for experimental analysis, the experimental results show that the mean absolute percentage error(MAPE), root mean square error(RMSE) and mean absolute error(MAE) of the IWOA-ELM model are 0.87%, 1.12 MW, 1.08 MW, which are more accurate than that of a single algorithm's load forecasting model, and the model can improve the accuracy and reliability.