Optimization-based operations control for public transportation service with transfers

Abstract

This research develops a real-time simulation-based control framework that attempts to coordinate the operations of public transportation services to allow smoother transfers and to maintain service regularity. The control actions, which include holding and change speed, are set as the solution of an optimization problem with the objective to minimize total passengers’ time in the system within a prediction horizon. The prediction horizon is defined by a number of downstream stops and subsequent buses. The predictions made include the arrival and departure times of vehicles at downstream stops and the passenger demands they are expected to serve. The model is demonstrated with a case study of three lines of the BRT system in Haifa, Israel. The results show that the model outperforms headway control and no-control alternatives in terms of total passengers’ time and on-time performance. It is only slightly worst then the headway control in terms of waiting times and denied boarding at stops. The combination of holding and change speed actions provides substantially better results than each of them individually.