PASSENGER FLOW CONTROL IN CONGESTED TRANSIT NETWORKS

This paper presents an approach to model the impact of ramp metering in public transit assignment. In overcrowded situations the operator might choose to meter the access to platforms for safety reasons and to minimise the delays passengers incur through failing-to-board. The approach is based on introducing a fail-to-access probability at the station entrance if the demand exceeds the available capacity. This fail-to-access probability is also thought to deter passengers from choosing a congested route. In the second part of the paper the ramp metering is then optimised in the form of a bi-level optimisation problem. The control variables are the station-access capacities. An elitist non-dominated sorting Genetic Algorithm is used in order to find the pareto-optimal set of solutions. With an example it is illustrated that it might be beneficial to introduce ramp metering for the operator as well as for the passenger.