MOTION PLANNING FOR DYNAMIC SYSTEMS USING FIRST-ORDER LOGIC

摘要

The problem of motion planning for a class of dynamic systems is considered in this study. A knowledge-based approach is used to determine the initial conditions that will yield a certain desired state of the dynamic system. The search space is limited by using a set of rules because reasoning about dynamic systems is basically searching an infinite space. In this study, first-order logic is used for knowledge representation and reasoning. The methodology is applied to playing a pool game. The dynamics of the motion of the balls are complicated and significant expertise is required to know how to strike the balls. Simulated results presented show how the rules help in finding the appropriate strategies for playing the game.