Abstract:
Synchronized transfers in schedule-based public transport (PT) networks are used to reduce interroute or intermodal passenger transfer waiting time and provide a well-connected service. However, in practice, synchronized transfers do not always materialize because of some stochastic and uncertain factors, such as traffic disturbances and disruptions, fluctuations in passenger demand, and erroneous behavior of PT drivers. As a result, missed direct transfers not only frustrate existing users but also discourage potential passengers from using PT service. This work presents an optimization procedure based on a model predictive control (MPC) to increase the actual occurrence of synchronized transfers in schedule-based PT networks. The procedure aims to reduce the uncertainty of meetings between PT vehicles. MPC uses selected online operational tactics based on real-time data, such as skip stop, speed change, and holding. First a library of operational tactics was built to serve as a basis for the sequential receding horizon control process in the MPC. Then, an event activity network with dynamic moving elements was constructed to represent the logical process of the PT transfer synchronization problem. The MPC procedure for a real-time deployment of operational tactics was explicated. A detailed example was used as an expository device to illustrate the procedure developed, along with a real-life example from Auckland, New Zealand.
