Evolutionary Dual Rule-based Fuzzy Path Planner for Omnidirectional Mobile Robot

Abstract

Fuzzy navigator has been widely used in path planning for mobile robots because of its fast response. In this paper, evolutionary dual rule-based fuzzy path planner is proposed as a nobel path planner for omnidirectional mobile robots. The newly proposed fuzzy path planner utilizes a multiobjective evolutionary algorithm, which can take user's preference into account in path planning, for optimization of the path through scaling parameters used in fuzzification and defuzzification. The performance of evolutionary dual rule-based fuzzy path planner is evaulated through comparison of created paths and estimated Pareto-optimal fronts from dual multiobjective particle swarm optimization and dual multiobjective qauntum-inspired evolutionary algorithm. To show the validity of our path planner, path planning for omnidiretional mobile robot is simulated in MATLAB. Through the use of quantum-inspired evolutionary algorithm in Webots simulator, each unit movement of the omnidirectional mobile robot is optimized so that the non-holonomic motion of the mobile robot can be modified. The proposed evolutionary dual rule-based fuzzy path planner demonstrates the effectiveness through computer simulation for the omnidirectional mobile robot navigating in an environment with obstacles.