Research on Application of Neural Network Based on Cloud Computing in Power Load Forecasting

TIn the process of short-term load forecasting, the non-linear mapping relationship between the factors causing load changes and the load is one of the reasons for the deviation between the predicted results and the actual results. Neural networks have strong nonlinear mapping ability and self-learning ability. In this paper, the distributed idea is introduced, the BP neural network is distributed, and the distributed BP neural network prediction algorithm is implemented on Spark cloud platform. Combined with its advantages in analyzing non-linear problems, we propose a load forecasting model based on neural network, analyze the existing problems of BP algorithm in power load forecasting and make improvements. Starting from the input layer, hidden layer and output layer, a load forecasting model based on neural network is designed and implemented. Taking the daily load data of a certain area as a sample, the collected electric quantity is preprocessed. Then the number of hidden layers and nodes are analyzed and designed. Finally, the short-term daily load is predicted. The comparison results show that the predicted value is in good agreement with the actual value.