Nonholonomic motion planning of mobile robot with ameliorated genetic algorithm

Abstract

The optimal motion planning problem of mobile robot system with nonholonomic constraints amounts to the optimal control problem of nonlinear system. The kinematic and dynamic equations of the mobile robot can be set up according mechanical and control theory. After stepping into the problem, some shortcomings in choosing the specific genetic operator for the current used Genetic Algorithm (abbreviated as GA) have turned up. In this paper, a new method of mobile robot path planning based on the Ameliorated Genetic Algorithm (abbreviated as AGA) is presented. Compared with GA, the ameliorated one can improve system's motion precision with specific crossover operator of AGA operations. The effectiveness of the optimal motion planning problem of mobile robot system is demonstrated by the numerical simulation. And AGA has also been proven to have good convergence, higher operative speed and improved accuracy. © 2006 IEEE.