Rail transit systems, fundamental to urban mobility, frequently encounter disruptions necessitating prompt and effective emergency responses, particularly for connecting bus services that transport passengers to affected rail lines. This research paper explores emergency dispatch methods for abnormal connecting buses in rail transit, concentrating on enhancing the responsiveness and efficiency of dispatch protocols during non-standard operational scenarios. By delineating the emergency shuttle service process and identifying key factors, a shuttle bus emergency dispatch model was developed for both single-line and multi-line emergency scenarios, considering passenger travel behavior and vehicle operation modes. The decision variables included the stopping plan, dispatch quantity, and departure frequency, with the objective of minimizing total passenger travel time. Constraints related to resources, time, demand, safety, and physical limitations were incorporated. Given the integer nature of the decision variables concerning the number of vehicles dispatched and the stopping plan, a solution process was designed using a discrete particle swarm optimization (DPSO) algorithm, and the model was subsequently solved.