An Approach of Highway Toll Revenue Calculation Based on the Prediction of OD Matrix

Accurate toll revenue calculation is essential for highway operation management and construction planning. Reliable forecasting requires a thorough understanding of evolving traffic flow patterns and regional origin–destination (OD) distributions. However, challenges such as incomplete data collection and difficulties in capturing transit traffic patterns often hinder accurate OD estimation. To address this, we propose a bidirectional long short‐term memory (Bi‐LSTM)–based method for OD matrix estimation in highway revenue prediction. Using real‐world data from the Hanghui Highway in China, the proposed model is evaluated against eight benchmark approaches, including traditional machine learning and deep learning models. Results show that Bi‐LSTM achieves the best performance, with a root mean squared error (RMSE) of 2.8922 and a mean absolute error (MAE) of 1.1890, outperforming all comparison methods. These findings demonstrate not only the precision and robustness of the Bi‐LSTM approach but also its novelty in bridging OD estimation with revenue forecasting, providing new insights for the intelligent highway management and operation as well as the highway revenue prediction and analysis.