Research on Real-Time Vehicle Load Prediction Method Based on Neural Networks

Vehicle load prediction is a key research focus in the field of transportation. Existing load prediction methods primarily rely on sensor measurements or dynamic models, which face challenges in accurately obtaining real-time vehicle load remotely. This limitation hinders their widespread application in large-scale, multi-vehicle scenarios, as well as in monitoring and regulatory tasks for enterprises and governments. To address this issue, this paper proposes a commercial vehicle load prediction method based on a BP neural network optimized by the whale optimization algorithm (WOA). The method enables load prediction development using real-world road data, achieving a single-vehicle prediction model deviation of 6.66%. Furthermore, the developed Simulink model was applied to real vehicle validation, where real-road testing demonstrated that the average deviation of real-time load prediction converged to 3.47%.