This proposal addresses one of the most complex technological problems confronting today's society: the minimization of congestion on our nation's roads and high ways. It is unlikely that significant improvements can be made without the application of the principles of Intelligent Control. Toward achieving this goal, we propose to formulate system models, to analyze vehicle and network modalities, and to design control, estimation, an d identification algorithms for Intelligent Vehicle/Highway Systems (IVHS). We will develop a theoretical basis for applying intelligent control to large-scale, complex systems, using IVHS as a substantive example. Our approach is rooted in optimization and control theory, as well as decisions analysis an artificial intelligence. This is a multidisciplinary problem; hence, principal investigators from Princeton's Departments of Civil Engineering and Operations Research, Electrical Engineering, and Mechanical and Aerospace Engineering will work together to solve it. Furthermore, we will collaborate with the Vehicle Systems Research Department of the General Motors Research Laboratories (providing a link to the TravTek IVHS Program currently underway in Orlando, Florida) and the TRANSCOM Interagency IVHS Program (Northern New Jersey and Metropolitan New York area) to assure that this research is not only scientifically meritorious but relevant to real needs. The proposed research separates naturally into three tasks: 1) Modeling Intelligent Vehicles and Schools of Intelligent Vehicles, 2) Intelligent Traffic Management and Information Transfer for Minimizing Congestion, 3) Graphic Simulation of Intelligent Vehicle/Highway Systems. These three tasks will be conducted concurrently and jointly by the principal investigators and their students. Anticipated outputs of the proposed research include a new discrete-event formalism for analyzing systems of rule-based controllers and physical plants, a hierarchical structure for intelligent management of large-scale systems, and a flexible graphics tool for investigating mixed-variable dynamic systems, with emphasis on IVHS scenarios.
