A multiple perspective and multiple objective approach to real-time dynamic traffic assignment for advanced traffic management and information systems

Intelligent Transportation Systems aim to mitigate traffic congestion by adding computer intelligence to vehicle and roadway use. How best to efficiently route vehicles through a network with variable and interacting demands and limited supply is considered in this thesis. This problem is termed Dynamic Traffic Assignment and is applicable to Advanced Traffic Management and Information Systems.; Dynamic Traffic Assignment methods to date are computationally burdensome for real-time application. Centralized processing conceptions would require two-way communications in order to use the vehicle's destination in the solution and transmit the optimal route back to the vehicle. The extant methods ignore driver behavior except for travel time optimization as a driver or system goal.; In this work an in-vehicle approach is taken to dynamic assignment in which path choice is performed autonomously, requiring only one-way communication of real-time traffic data. Driver characteristics are treated as multidimensional where optimizing trip time and trip complexity are the goals.; Several demand-side strategies are examined: Pre-Trip Routing, in which the optimal path at the start of the trip is followed; Dynamic Routing, in which the optimal path is recalculated at each decision node; and Equilibrium Routing, a baseline case, in which drivers without information aim to use the paths of equal time at the peak volume.; To the in-vehicle path choice perspective is added a supply-side control tactic used in conjunction with the Dynamic Routing strategy, termed the Arc Metering strategy. Arc availability is determined by a Transportation Management Center using network flows only, without knowledge of destinations. The Arc Metering strategy benefits from proactive anticipation of breakdown in a network arc's performance by invoking arc opportunity costs for the transmitted arc time and complexity costs. Arc Metering aims to closely match flows to arcs that have the capacity to allow uncongested flow.; The various routing strategies are simulated over a range of trip time and trip complexity tradeoffs for a freeway network with peak period demands. The results show that substantial improvements over an array of performance measures can result, especially at high demand, by using the Arc Metering supply control strategy.