Towards a more flexible demand responsive transit service with compensation mechanism considering boundedly rational passengers

Abstract

AbstractDemand responsive transit (DRT) with app‐based reservation platforms is experiencing a renaissance to bring the tremendous potential for mobility in the urban universe. Nevertheless, how to attract and retain passengers for long‐term use has become one of the most significant problems. The decision‐making psychology of passengers is often overlooked but incredibly critical in the practical applicability of DRT services. This paper proposes a more flexible DRT service with soft time windows considering boundedly rational passengers. A compensation mechanism is developed to alleviate the dissatisfaction of passengers while considerably promoting the system efficiency. A two‐stage model is designed to incorporate bounded rationality into the optimization process of mixed demand, including the static phase for reservation passengers and the dynamic phase for real‐time passengers. To enhance the computational efficiency, a hybrid heuristic algorithm combining spatiotemporal clustering and non‐dominated sorting genetic algorithm (NSGA)‐II is constructed to obtain the Pareto solutions set. An illustrative example of the Nguyen–Dupuis network is presented to demonstrate the validity of the algorithm. Subsequently, a large‐scale case study in Beijing evaluated the applicability of DRT in the real‐world network. The results reveal that dynamic DRT with compensation mechanism can substantially improve the system performance while ensuring the service quality. The response rate of passengers has been dramatically promoted to 80%. The operating profit has been enormously improved by up to 73%. Therefore, this study is radically conducive to understanding the passenger's decision‐making psychology while constructing a more cost‐efficient flexible strategy for the service provider.