Planning Passenger Rail Services in a Competitive and Shared-Use Environment
thesisposted on 01.07.2016, 00:00 by Ahmadreza Talebian
This research couples game theory and optimization models and puts forward a comprehensive framework aiming at efficient planning of intercity passenger rail services in competitive and shared-use environment. This framework consists of three models: air-rail competition model, train scheduling model, and capacity allocation and access charging model. In the air-rail competition module, we propose a game-theoretic model whose aim is to determine rail frequency and fare. In addition, we investigate impacts of train technology on air-rail competition. We present a three-stage game, in which the rail agency determines the train technology, as its most strategic decision, at the top stage. Then the rail agency and airline compete sequentially on frequency and fare. We adopt a backward approach to solve the three stage game. In the train scheduling module, we develop strategic-level train models for planning high performance passenger and freight train operations on shared-use corridors, given passenger rail service frequency. In particular, we develop a hypergraph-based, two-level approach to sequentially schedule train services. On the passenger side, we assume that each passenger has a preferred departure time (PDT) and tries to minimize his/her schedule delay defined as the difference between the traveler’s desired departure time and the closest scheduled departure time. The objective is to minimize total passenger schedule delay cost subject to a set of constraints guaranteeing operational feasibility of passenger train schedules. We then insert freight trains among the fixed schedule of passenger trains such that freight total cost, which consists of freight lost demand cost, departure delay cost, and en-route delay cost, is minimized subject to constraints ensuring operational feasibility of freight train schedules. The upper level of the sequential scheduling model yields the preferred passenger train schedules, given which schedules for freight train are determined subsequently. However, the host freight railroad may not be willing to implement such schedules and would like to bargain with the passenger rail agency for either improved freight service or compensation for not getting that. To model this bargaining process, we assume a set of feasible passenger train schedules is given. For each feasible passenger train schedule, we develop the corresponding freight train schedules and calculate the associated cost. The access charging and capacity allocation model receives the set of feasible passenger train schedules and corresponding freight train schedules, as well as rail fare from the air-rail competition model, and emulates the negotiation process between the passenger rail agency and the host freight railroad. We develop a bargaining model which determines rail access charges and detailed plans for allocating the shared-use line capacity.