Elucidating Price Variability Drivers in Highway Electromechanical Equipment Using CV Predictions with PSO-XGBoost

Abstract

The exponential development of expressways has resulted in increased demand for highway electromechanical (E&M) equipment. However, the constant changes in parameters and models of electromechanical equipment make national quotas outdated and complicate price forecasting due to data collection challenges and significant price variability. To address this, the study introduces the coefficient of variation (CV) to reduce variability. Therefore, this study further introduces a novel hybrid forecasting model for the CV of E&M equipment pricing and proposes a coefficient analysis and prediction system based on particle swarm optimization (PSO) and XGBoost to analyze the driving factors behind the price fluctuations of highway E&M equipment which aims to provide a scientific and rational reference for the pricing of E&M equipment. The evaluation metrics is R² = 0.90066, MAE = 0.033412, MAPE = 11.5624%, MSE = 0.0017948, and RMSE = 0.045769, showed that the PSO–XGBoost model outperformed the other models in terms of precision and stability in prediction, demonstrating the proposed model's effectiveness in forecasting CVs for E&M equipment pricing and demonstrating the feasibility of the model to elucidate the factors influencing price changes.