Global path planning of mobile robots as an evolutionary control problem

Abstract

A new method for global path planning of mobile robots is presented using an Evolutionary Strategy (ES). A two-dimensional autonomous path planning problem of a point mobile robot among polygonal obstacles is formulated as a constrained time-optimal control problem considering motions of the robot. From the visibility concept, each polygonal obstacle is enclosed by a circle so that the visible side of the obstacle will be contained inside a circle. An appropriate fitness function for the problem is constructed by applying a penalty function approach. Simulation results show that the proposed evolutionary approach is quite robust, and it is guaranteed to yield a path terminating at the goal with minimum time and distance while avoiding polygonal obstacles.