Simulation-based multiobjective optimization for recovery planning in disrupted public transport networks

Urban public transport systems are increasingly vulnerable to operational disruptions that can compromise servicecontinuity, increase passenger delays, and reduce network resilience. This study proposes a multi-objective optimizationframework for the recovery of urban bus operations under different disruption scenarios. The framework integratesvehicle allocation, reserve-fleet activation, route continuity, service-frequency adjustment, passenger demand coverage,and recovery planning within a Mixed-Integer Linear Programming (MILP) formulation. Six conflicting objectives areconsidered: passenger delay, recovery cost, cancelled trips, recovery time, passenger service, and operational resilience.The model is implemented in Python and solved using the Gurobi Optimizer, while a weighted-sum method is employedto generate alternative efficient solutions and construct the Pareto frontier. The framework is evaluated through fivedisruption scenarios involving fleet and network disturbances. Results show that reduced fleet availability substantiallyaffects network performance: decreasing fleet availability from 100% to 70% increases delay from 120 to 1620 min andrecovery time from 12 to 88 min, while the resilience index decreases from 1.00 to 0.75. The optimized recovery strategiesachieve delay reductions of 18.4–35.4%, recovery-time reductions of 15.2–39.6%, and demand-served levels of 89.8–94.0%,with resilience indices ranging from 0.81 to 0.88. These results demonstrate that coordinated fleet allocation and recoveryplanning can mitigate disruption impacts while balancing operational efficiency, passenger service, and network resilience.The proposed framework provides a decision-support methodology for transit operators to identify effective recoverystrategies under heterogeneous disruption conditions.