Abstract: It is necessary to adjust the train timetable and let the train recovery on time as soon as possible when the train of Urban Transit in the process of operation is late due to equipment fault, passengers congestion, etc. Taking the minimum total delay time as the objective function, this article proposed a train adjustment model based on Ant Colony Optimization (ACO) Algorithm. Shenzhen Metro Line 6 was taken as an example to verify the practicality of the model under the Visual C++ 6.0 programming environment.