A Knowledge-Based Temporal Planning Approach for Urban Traffic Control

Abstract

The global trends in urbanization have caused many problems, among which Urban Traffic Control (UTC) becomes a priority issue for most big cities in many countries. As traffic demand changes rapidly, appropriate control policies are required to be generated in real time in order to, e.g., minimize congestion to reduce average travel time and air pollution. A practical way to meet the challenge is to build an intelligent control mechanism of road traffic. In this context, automated planning, a powerful and effective technique, can be exploited as an aid to dynamically produce plans to alleviate the problems of UTC. In this paper, we present an approach based on automated planning, in particular temporal planning scheme that aims for producing predictable management strategies of UTC. Meanwhile, a logic style control knowledge is employed to provide useful guidance for the search process in planning. We show the preliminary evaluations on simulation benchmarks closely related to UTC. Experimental results show the feasibility and effectiveness of our approach, compared with the state-of-the-art planners that participate in recent International Planning Competitions.