Research On Short-term Load Forecasting Of Power System Based On Integrated Elman Neural Network

Power system is one of the most complex artificial intelligence systems in recent decades.With the continuous development and expansion of the system,accurate load prediction becomes more and more important in the safe,efficient and economic operation of the system.Short-term load forecasting is an indispensable work in the operation of power system,which is the basis of reasonably arranging the starting and stopping plans of units and formulating the scheduling plans.Therefore,how to improve the accuracy of short-term load forecasting has become one of the key issues that domestic power workers pay close attention to.This paper first introduces the characteristics,influencing factors and basic forecasting process of short-term load forecasting.Then briefly explains the artificial neural network which plays an important role in short-term load forecasting.Finally,the Elman neural network with dynamic recursive properties is highlighted.In view of the shortcomings of traditional Elman neural network,such as randomness and low prediction accuracy,this paper constructs a short-term load forecasting model based on Bagging algorithm and Elman neural network.This model uses Bagging algorithm to eliminate the problem of large randomness of Elman neural network prediction results and improve the prediction accuracy and stability of the prediction model.The simulation results show that compared with the traditional Elman neural network prediction model,this model can effectively improve the accuracy and stability of prediction.In the short-term load forecasting model based on Bagging algorithm and Elman neural network,the sub-training set samples of each Elman neural network are generated by the same probability sampling method,and the samples with large prediction errors cannot be given more attention,and the final prediction results are only a simple average of the prediction results of each Elman neural network,all these factors will affect the prediction accuracy.In order to eliminate the impact of these unfavorable factors,this paper proposes a new short-term load forecasting model based on Boosting regression algorithm and Elman neural network.In this model,each sub-train set of Elman neural network model is generated by probability sampling.And the establishment of the current Elman neural network depends on the training results of the previous Elman neural network,so that the Elman neural network is continuously optimized.And the final prediction result of this model is the weighted sum of the prediction results of each Elman neural network.The simulation analysis shows that the model has higher prediction accuracy than the short-term load forecasting model based on Bagging algorithm and Elman neural network.The traditional Elman neural network in the short-term load forecasting model based on Boosting regression algorithm and Elman neural network has the disadvantages of slow convergence speed and easy to fall into local minimum.In order to solve this problem,the traditional Elman neural network is optimized by PSO algorithm.Then the optimized network is combined with Boosting regression algorithm to construct a short-term load forecasting model based on Boosting regression algorithm and PSO-Elman neural network.The simulation result shows that compared with the short-term load forecasting model based on Boosting regression algorithm and Elman neural network,this model further improves the prediction accuracy.