Construction of Electricity Revenue Prediction Model Based on PSO-BP Neural Network and Coordinated Filtering Algorithm

Electricity revenue forecasting is an important part of power enterprise management. However, traditional forecasting models have problems such as high model complexity, high computational cost, and insufficient dynamics and adaptability when facing complex market environments and changing user behaviors. This paper studies an electricity revenue forecasting model based on PSO-BP neural network and coordinated filtering algorithm, aiming to improve the forecasting accuracy and adaptability of the model. First, historical electricity revenue data, electricity sales data, electricity price data and related external factor data are collected and preprocessed. Then, the particle swarm optimization (PSO) algorithm is used to optimize the initial weights and biases of the BP neural network, and the coordinated filtering algorithm is introduced to calculate the similarity between users and perform weighted averaging of electricity revenue, further improving the accuracy and stability of the forecast. The results show that the forecasting model based on PSO-BP neural network and coordinated filtering algorithm performs best in both mean square error (MSE) and mean absolute error (MAE), with MSE of 0.019 and MAE of 0.037. In terms of dynamics and adaptability, the adaptability score can reach 8.78 points.