Researching Network Development Methods
I’ve been working at the University of Maryland in College Park on methods for selecting and scheduling interrelated projects in transportation networks. These methods fill an important gap in the technical literature. Satisfactory methods have long been available for dealing with projects that are mutually exclusive, such as those where at most one project may be selected, or independent ones, where the benefits and costs of projects do not depend on whether and when other projects are implemented.
However, in transportation networks, projects are largely interdependent because traffic flows, and hence benefits depend on which other projects are selected and when those are implemented. Some interdependence may also arise from economies of constructing multiple projects. Thus, methods that can deal with interrelated projects must still be devised.
To deal with such problems my students and I have developed general methods for optimizing the selection, sequencing and scheduling of projects. These methods rely on artificial intelligence heuristics, such as genetic algorithms, simulated annealing and swarm algorithms for optimization, provided that a problem-specific method (such as a traffic assignment algorithm, a queuing network or a traffic simulation model) is available to evaluate the network at any development stage.
We initially developed such methods for inland waterways and later extended them to other kinds of transportation systems. Recently, Ms. Elham Shayanfar, who is a doctoral candidate, developed such models for rural and urban road networks. Dr. Yanshuo Sun has developed models for the phased development of airport facilities. Mr. Uros Jovanovic has recently completed an M.S. thesis in which he incorporated intersection improvement projects within urban road networks. Ms. Ya-ting Peng, a visiting Ph.D. student from China, is working on scheduling the development of rail transit networks. The methods developed by this team account for various constraints, such as budgets and other resources, and the inevitable uncertainties regarding demand, costs, construction times and funding.
Paul Schonfeld is a professor in Civil and Environmental Engineering at the University of Maryland, College Park. His expertise is in transportation engineering.