| Traffic Prediction Based Battery State-of-Charge Planning for Connected Plug-in Hybrid Electric Vehicle |
| Dongwei Yao1,2, Fanlong Zeng1, Xinwei Lu1, Junhao Shen1, Kaiyang Huang1, Feng Wu1 |
1The Power Machinery and Vehicular Engineering Institute, College of Energy Engineering, Zhejiang University, Zhejiang, 310027, China
2The Key Laboratory of Smart Thermal Management Science & Technology for Vehicles of Zhejiang Province, Taizhou, 317200, China |
|
|
|
Received: January 22, 2025; Revised: May 5, 2025 Accepted: May 6, 2025. Published online: July 8, 2025. |
|
|
|
| ABSTRACT |
|
Battery state-of-charge (SOC) planning utilizing traffic information from the intelligent transportation system (ITS) and the Internet of Vehicles is promising to improve the performance of the connected plug-in hybrid electric vehicle (PHEV) energy management strategy. In this paper, a novel reference SOC planning method based on traffic prediction was proposed. A data-driven spatio-temporal graph convolutional neural network model with adaptive graph generation module (A-STGCN) was proposed to predict the traffic speed of road networks. Based on the predicted traffic speed, the long-term global speed profile can be predicted. A dynamic programming (DP)-oriented simplified vehicle power balance model was established and the global reference SOC trajectory was planned through online DP algorithm. Finally, the proposed SOC planning algorithm was combined with an adaptive equivalent consumption minimization strategy (A-ECMS) and verified through simulation. Simulation results using real world traffic data illustrate that A-ECMS combined with the proposed reference SOC planning method achieves 98.52% of fuel economy of the off-line global optimal solution, which outperforms other reference SOCs. |
| Key Words:
SOC planning · Graph convolutional neural network · Traffic speed prediction · Energy management strategy |
|
Corresponding Author.
Dongwei Yao , Email. 
Preview 
Full text via DOI 
Download Citation
Print
