A comparative study of smooth path planning for a mobile robot by  evolutionary multi-objective optimization 

Abstract

This  paper  studies  the  evolutionary  planning strategies for mobile robots to move smoothly along efficient  collision-free paths in known static environments. The cost of  each  candidate  path  is  composed  of  the  path  length  and  a  weighted sum of penetration depth to vertices of polygonal  obstacles. The path is composed of a pre-specified number of  cubic spiral segments with constrained curvature. Comparison  of the path planning performance between two Pareto-optimal  schemes, the parallel genetic algorithm scheme based on the  island method (PGA) and the non-dominated sorting genetic  algorithm (NSGA-II), are conducted in terms of success rate in  separate runs and path length whenever collision-free paths are  found. Numerical simulation results are presented for three  types of obstacles: polygons, walls, and combinations of both. 

جهت دانلود به ادامه مطلب مراجعه نمایید.