The 25th COTA International Conference on Transportation Professionals (CICTP2025) concluded successfully in Guangzhou. Experts and scholars from the world’s top universities and research institutions gathered to jointly discuss cutting-edge issues and innovative solutions in “Transportation-Artificial Intelligence-Energy.”
The TRANS research group performed outstandingly and achieved fruitful results at this conference, winning three awards in one go, fully demonstrating the group’s continuous exploration in modeling and AI applications in transportation energy systems:
- Hao Qi, a 2024 doctoral student at SCUT’s Future Technology College, won the Best Paper Award at CICTP2025 for “Optimizing Safety, Efficiency, and Battery Health: A Deep Reinforcement Learning Model for Car-Following in Electric Vehicles” (Advisor: Prof. Shiqi Ou). The study proposes a deep reinforcement learning-based ecological driving strategy to optimize energy consumption, ensure safety, and reduce battery degradation in car-following scenarios. Trained on a physics-integrated powertrain simulation platform, the model outperformed traditional methods in 70% of short-term tests. This work establishes a quantitative framework for analyzing the impact of driving behavior on battery aging, supporting EV sustainability. Future research will explore long-term battery health in complex scenarios.
- Xinyun Zhang, a 2023 undergraduate at SCUT’s Future Technology College, received the Best Lightning Talk Award at CICTP2025 for “Research on Electric Logistics Vehicle Charging Management Based on M/M/c and Simulation Models” (Advisor: Prof. Shiqi Ou).
The study analyzes commercial vehicle charging patterns, applying queuing theory (M/M/c for logistics vehicles, M/D/c for buses) to quantify daytime charging delays. Findings were validated using data from the Guangzhou government. Future work will investigate the nonlinear relationship between vehicle-to-charger ratios and queuing times to optimize charging infrastructure planning. - Jiayi Wang (2021 undergrad) and Hao Jing (2024 PhD candidate) from SCUT’s Future Technology College earned the Best Poster Award at CICTP2025 for “Optimization of Electric Vehicle Battery Charging Strategy Based on Reinforcement Learning” (Advisor: Prof. Shiqi Ou).
The study develops an EV powertrain energy model to improve battery state prediction accuracy under real driving conditions. Using Q-learning, it provides customized charging recommendations (charger type, timing, duration) while incorporating rental cost penalties for insufficient battery capacity. Results show that smart charging strategies can extend battery life, reduce costs, and meet travel demands, thereby enhancing the feasibility of the EV lifecycle. Future work will integrate battery discharge simulation into capacity assessments.
The TRANS research Group will continue to uphold a rigorous and realistic scientific research attitude and an innovative spirit, delve deeply into the field of transportation and energy system optimization, be committed to solving complex practical problems, produce more high-level research results with international influence, and contribute to the development of the discipline and the progress of the industry!
