Volume 58

A Comparative Study for Vehicle Power Prediction Based on Intelligent Transportation Information Zhengxian Chen, Han Hai, Huiming Jiang, Yuyang Pan, Yichong Li, Chaosheng Huang, Jun Li

Abstract

In the context of the rapid development of smart cities and intelligent transportation systems, the energy management efficiency of new energy vehicles has become a critical factor in achieving carbon peak and carbon neutrality targets. Among these, the accuracy of power prediction significantly impacts the rationality of energy allocation and overall vehicle response performance. This paper systematically investigates two representative power prediction paths: the direct method, which predicts future power directly, and the indirect method, which first predicts vehicle speed and then computes power using a dynamic model. To evaluate their applicability in real traffic environments, operating conditions are categorized into high-speed, medium-speed, and low-speed scenarios, and performance is compared under different prediction time windows (1s, 3s, and 5s), with MAE and RMSE as the primary evaluation metrics. The results indicate that under varying average speeds, each prediction path has its own advantages and disadvantages. Furthermore, this paper highlights the complementarity of the two methods. The findings can provide a reference for optimizing predictive energy management strategies in intelligent new energy vehicles, thereby promoting further improvements in vehicle operation efficiency within intelligent transportation systems.

Keywords Smart cities, Intelligent new energy vehicles, Power prediction, Energy management, Operating condition analysis