【Purpose】In order to optimize the emergency response and address the inefficiencies observed in traditional dispatch models, this study tries to develop a scheduling method, which considers a range of emergency dispatch modes.【Method】Considering the characteristics of emergency parking lot dispatching capacity, vehicle transportation capacity, and vehicle rescue time, a multi-objective combination scheduling optimization model is constructed with the goal of minimizing the evacuation cost of emergency public transportation vehicles and minimizing the average passenger delay. Based on the characteristics of the model, a fast non dominated sorting genetic algorithm (NSGA-Ⅱ) was proposed to solve the problem. In order to improve the diversity of the population and the performance of the algorithm, corresponding improvements were made to the algorithm. The Pareto distribution optimization solution set was obtained, and the optimal compromise solution was selected from it using a membership function. Finally, the Nanchang Rail Transit Line 1 was used as an example to separately evaluate the emergency bus combination dispatch plan and the single scheduling scheme are solved separately.【Result】Results show that the proposed combination dispatch plan reduced passenger delay time by 20.48% and transportation costs by 16.96% compared to traditional single dispatch plans. Additionally, the improved NSGA-Ⅱ algorithm further reduced passenger delay time and transportation costs by 6.72% and 3.59%, respectively.【Conclusion】Sensitivity analysis shows that, with clear demands of stranded passengers, fleet size negatively correlated with the average delay time of stranded passengers and positively correlated with emergency bus transportation costs.