Prediction of crucial material procurement prices for power grid based on the CNN-LSTM and the RIME-BP methods

The material procurement of power grid enterprises has the characteristics of large scale, diverse types, and long cycle, and its procurement price is influenced by multiple factors, making price prediction difficult. In addition, its price prediction often relies on empirical judgment, which leads to insufficient prediction accuracy. To address these issues, this paper proposes a hybrid prediction method that combines multiple models, RIME-BP and CNN-LSTM, to study the prices of critical materials in the power grid. Based on the unit price excluding tax for the procurement of cable protection pipe materials by a particular power company from 2020 to 2024. Firstly, five key features were selected using the Pearson correlation coefficient. Secondly, establish RIME-BP and CNN-LSTM prediction models to forecast the price of cable protection pipes. Finally, compare the prediction results of the two intelligent algorithms with those of the multiple linear regression method. The results indicate that the intelligent algorithm models have better predictive performance than the traditional linear regression model, and the results provide some reference for power grid companies to formulate bidding control prices.