Capacity research is applied to describe the relationship between capacity and operation quality within a defined investigated area of a certain infrastructure network. When simulation methods are used in capacity research, the integrated dispatching modules will influence the final results of capacity research. In the worst case scenario with a large number of trains, an unsuitable dispatching algorithm perhaps cannot realize the timetabled capacity, which underlines the importance of choosing suitable dispatching algorithms for the target capacity. In order to evaluate these influences, a new state-dependent dispatching algorithm, which can be adjusted according to the system states, was designed, and a classification method of system states was developed. In the further dispatching process, the system states should be identified primarily, and the most suitable dispatching algorithm can be automatically carried out. Moreover, a multi-scale simulation model characterized by continuous scaling was built to calculate the waiting times based on timetables with stochastic deviations. With the simulation method, the influence of dispatching on the results of capacity research was systematically investigated, and convincing results were eventually produced.