Traffic-aware energy management helps fuel cell buses cut costs near bus stops

Fuel cell buses are a promising route toward cleaner public transport, but their real-world efficiency depends on more than the vehicle powertrain alone. A study in Green Energy and Intelligent Transportation presents a hierarchical predictive energy management strategy that uses traffic information around bus stops to guide how fuel cell buses plan battery state of charge and allocate power as they approach stopping events.

Bus stops create a demanding operating scenario. Vehicles slow down, queue, dwell, and accelerate again, often under traffic conditions that change from one stop to the next. For fuel cell buses, these repeated transients can make energy management difficult: the system must satisfy driving demand while avoiding unnecessary cost and keeping the state of charge close to an efficient trajectory. The authors frame this challenge as a networked-vehicle problem, asking how traffic information can be converted into practical control decisions.

To address it, the researchers built a scenario model for buses entering bus stops based on actual data. Their strategy has two connected layers. In the upper layer, dynamic programming is used to solve optimal state-of-charge trajectories under different historical traffic conditions. A bidirectional long short-term memory network then maps traffic information to the best SOC trajectory, enabling fast generation of a long-term reference that can meet real-time requirements.

The lower layer uses this SOC reference to guide real-time predictive energy management. In other words, the vehicle does not only react to immediate power demand; it follows a traffic-informed reference that reflects what is likely to happen as it approaches and enters a bus stop. This structure is designed to bring the control result closer to the dynamic-programming optimum while remaining more suitable for online use than a full global optimization performed in real time.

Simulation results showed clear benefits. Compared with a strategy without an SOC trajectory reference, the proposed approach reduced life cost by 13.8% and total cost by 3.61%. The SOC trajectory also stayed closer to the dynamic-programming optimal solution, suggesting that the traffic-aware hierarchy can improve both economic performance and control quality for stop-and-go bus operations.

The work points to a practical direction for intelligent public transport. As buses become increasingly connected to roadside and traffic information systems, energy management can shift from isolated vehicle control to scenario-aware decision-making. Because the reported results are simulation-based, future work will still need broader validation under real operating conditions. Even so, the study offers a useful framework for reducing fuel cell bus operating costs in one of urban transit's most frequent and energy-sensitive scenarios.

Reference

Author:

Mei Yan a, Hongyang Xu a, Menglin Li a, Hongwen He b, Yunfei Bai c

Title of original paper:

Hierarchical predictive energy management strategy for fuel cell buses entering bus stops scenario

Article link:

https://www.sciencedirect.com/science/article/pii/S2773153723000312

Journal:

Green Energy and Intelligent Transportation

DOI:

10.1016/j.geits.2023.100095

Affiliations:

a School of Vehicle and Energy, Yanshan University, Heibei, 066004, China

b National Engineering Laboratory for Electric Vehicles, School of Mechanical Engineering, Beijing Institute of Technology, Beijing, 100081, China

c School of Engineering, University of Warwick, Coventry, CV4 7AL, United Kingdom

Image usage restrictions:

News organizations may use or redistribute this image, with proper attribution, as part of news coverage of this paper only.

Image credit:

GREEN ENERGY AND INTELLIGENT TRANSPORTATION