Power Transformer Price Forecasting Method Based on Variational Modal Decomposition and Improved Chaotic Grey Wolf Optimization-Random Forest

Abstract

To reasonably estimate the cost of power transformers, the price trends of power transformers are analyzed based on data mining techniques. A power transformer price prediction method is proposed. This method first conducts Pearson correlation analysis on the influencing factors of power transformer prices, and extracts the main influencing factors to obtain the training data set. Second, the historical price data of power transformers are decomposed using variational modal decomposition, and the trends of each modal component are analyzed. Third, the decision tree parameters and splitting feature parameters in the random forest regression model are optimized using the improved chaotic gray wolf algorithm, and each modal component is further predicted. Finally, multilayer prediction results are accumulated to calculate the power transformer price results. The results of the computational examples show that the improved random forest can accurately predict the price changes of power transformers. Thus, it can effectively improve the level of material procurement and reduce the influence of human factors.