Optimal bus scheduling with dynamic stochastic demand: simulation framework and result analyses

The paper presents computer simulation approach to developing optimal bus scheduling under stochastic and deterministic network environments with static and dynamic demand. The primary objective is to minimize the total passenger waiting time. The mean transit demand at each stop and the travel time between neighboring bus stops are assumed as known information. An example route showing an application of the proposed model is presented. Comparisons between waiting times under different network scenarios have been made using the example route. The results obtained from the analyses are promising and thus the method can be implemented as a substitute to real time control strategies when implementation cost is an issue as the latter is relatively expensive. It may also be used in coordination with the real time control strategies to reduce holding time. In addition, it can assist transit users in planning their trips in advance so that misleading travel information is reduced.