Robust optimization for multimodal route planning integrating shared taxi and bus under uncertain traveling time

Multimodal transportation that integrates multiple transport modes, such as buses, taxis, and subways, plays a crucial role in alleviating environmental pollution and improving mobility. However, unpredictable traffic conditions may disrupt the travel time. This work studies a robust multimodal route planning problem (RMRP), focusing on the integration of shared taxis and buses. It devises an innovative route planning approach for shared taxis to enable passengers to seamlessly transition between the two modes, while reducing the impact of uncertainty and allowing passengers to arrive on time. It establishes a multiobjective optimization model that considers travel time uncertainty. The objectives are minimizing the total travel distance traversed by shared taxis and maximizing the passenger satisfaction. A novel nondominated sorting genetic algorithm with uncertainty repair (NSGA-UR) is proposed to solve the problem. It incorporates innovative encoding and decoding methods, evolution strategies, and an uncertainty repair strategy. NSGA-UR demonstrates higher robustness compared to several widely used multiobjective optimization algorithms, including NSGA-II, MOPSO, and MOGWO. In addition, experimental results show the superiority of the algorithm in solving RMRP. This work can contribute to the advancement of intelligent public transportation services.