Stochastic dynamic traffic assignment for intermodal transportation networks with consistent information supply strategies
MetadataShow full item record
Private car use continues to increase in most urban areas around the world, exacerbating various associated problems such as traffic congestion, environmental degradation and high fuel consumption. With the dispersion of land-use activity patterns in many urban areas, serious challenges face the design of public transportation systems to provide an effective substitute for the private car. A more plausible approach would be to design and promote an intelligent intermodal transportation system, which integrates the private car with existing or planned transit modes. Several Intelligent Transportation Systems (ITS) capabilities and user services, which fall in the areas typically referred to as Advanced Travelers Information Systems (ATIS), Advanced Traffic Management Systems (ATMS), and Advanced Public Transportation Systems (APTS), offer promising opportunities to manage traffic and improve operation in intermodal networks. Providing real-time information to users on network congestion, and availability and status of transit modes, could contribute to efficient integration among the existing modes, and hence to more efficient trips for travelers. This research develops a stochastic dynamic trip assignment model for urban intermodal networks. The model overcomes limitations of static tools used in current practice. These limitations relate to the type of alternative measures that may be evaluated, and the policy questions that planning agencies are increasingly asked to address. The model captures the interaction between mode choice and dynamic assignment under different information provision strategies and network control schemes. A simulation-based solution algorithm is presented for the problem. The model incorporates a stochastic network loading model for intermodal networks and a multi-objective shortest path algorithm. The algorithm solves for the time-dependent flows for each feasible mode-path combination in the network. This research also addresses the design of real-time consistent-normative information supply strategies for the auto travelers in the network. Providing travelers with real-time information could help them plan their trips efficiently and achieve better spatial-temporal network traffic distribution. The problem is formulated as a bilevel mathematical program where the main objective is to minimize the total travel time in the network, while the secondary objective is to minimize the difference between the provided information and experienced travel times.