Analysis of urban network performance under stochastic incident occurrences is a new subject in traortation literature. Such incidents may be Modeling problems in group (b) above has been initiated by constructing a model to improve the performance of the networks to contain the negative impacts of the random events with long-term effects. An algorithm is proposed to solve this problem to maximize the overall reliability of the network to withstand those stochastic events under resource constraint. Application of the proposed method over several networks is shown numerically, and the results are discussed. In the analysis of the problems in group (c) above, this dissertation discusses the random nature of volume and travel time of the links in the network, particularly for the morning peak trips, which are basically comprised of those with trip purpose “work” and “school”. These trips are such that they should be done before certain time, and hence the respective trip-makers are argued to have a pessimistic or conservative attitude in making them. A new equilibrium flow problem, called pessimistic equilibrium flow problem (PEF), has been proposed to model such behavior in route choice in the network, and a procedure is devised to solve it. Next, a pessimistic network design (PND) problem has been proposed which minimizes an objective function that is a linear combination of the travelers’ social cost and time cost, and has its network flows based on PEF problem. A procedure is devised to solve this problem. Application of this procedure to solve a PND problem in a test network, and comparison of the results obtained with the corresponding results obtained by solving a conventional network design problem on the same test network, reveals several interesting points, discussed in this dissertation. This dissertation is concluded by summing up the results obtained in this research, and pointing several avenues, for future research.
