Joint Dynamic Bus Platooning Formation and Trajectory Planning Optimization on a Transit Corridor

Bus platooning on urban bus corridors can improve the efficiency of dedicated bus lanes and reduce energy consumption. However, due to the multiple routes operating on the bus corridor, bus platooning may induce congestion at the bus stops and disturb regular headway. To address these bus operational issues, this paper proposes a two-stage dynamic bus platooning model considering bus routes and provides safe and efficient trajectory planning for buses in each platoon. To solve the proposed model, a Particle Swarm Optimization embedded Dynamic Programming algorithm is customized to find the time of leaving and joining the platoon, and Broyden-Fletcher-Goldfarb-Shanno Sequential Quadratic Programming method is used to control the trajectories of bus platooning in real time. Using the SUMO simulator, a real-world case study is conducted to show the effectiveness of the proposed method over traditional methods. The results show that the fuel consumption is reduced by 18.23%, operational efficiency is improved by 8.15%, and the average passenger waiting time is decreased by 5.73%. Sensitivity analysis reveals the impact of the number of routes, fleet size and stop spacing on system performance. The proposed dynamic platooning method outperforms the CACC method across different fleet sizes. Results show that the desired fleet size is greater than 60 buses, the stop spacing is greater than 500 m, especially in large cities or long-distance bus systems.