Short-term load forecasting of power system based on genetic algorithm improved BP neural network algorithm

Forecasting method is the key in short-term load forecasting. Genetic algorithm is used in this paper to improve the BP neural network short-term load prediction model of power system, and the BP neural network model of hidden layer 10 is optimized and improved by genetic algorithm, so as to solve the problem that the prediction effect is not ideal due to excessive internal optimization of the model. The 24 o ’clock historical load, the meteorological characteristics of the day, the maximum and minimum temperature and other information were input into the prediction model as input variables. Finally, the genetic algorithm was used to get the prediction results of power load in different time periods and took them as output variables. In view of the proportion and threshold initialization problems in the model, and I use the genetic algorithm to optimize BP network. According to the daily meteorological characteristics, this paper uses genetic algorithm to improve BP network. prediction model. In the empirical study, our prediction results also meet the expected requirements, which fully shows that genetic algorithm is used to improve the prediction accuracy of BP network in this paper.