基于改进遗传算法的移动机器人路径规划(初稿).pdf

基于改进遗传算法的移动机器人路径规划 郭首亮，孙海洋，陈 珍 (南京大学金陵学院信息科学与工程学院，南京 210089) 摘要：为了解决传统遗传算法在移动机器人路径规划中，在对路径长度，路径平滑度以及路径安全度进行多目 标优化时，难以找到合适的比例因子的问题。本文提出了一种改进的遗传算法，在传统的遗传算法选择，交叉，变 异的基础上，增加了基因定向变异，将多目标优化转换成单目标优化，有效地解决了比例因子的难以寻找的问题。 最后在模拟环境下利用Matlab 进行仿真，证明了这一改进遗传算法的高效性，可行性。 关键词：遗传算法；机器人路径规划；基因定向变异；Matlab 仿真 中图分类号： 文献标志码：A Path Planning Method for Mobile Robot Based on Improved Genetic Algorithm Guo Shouliang, Sun Haiyang, Chen Zhen (Information Science and Technology Academy, Nanjing University Jinling College, Nanjing 210089, China) Abstract:In order to resolve the problem conventional genetic algorithm caused in planning an appropriate path for mobile robot that it is a great challenge to find a well-matched scale factor when we try to shorten the length of path, smoothen the path and increase the security of the path, in this paper, an improved genetic algorithm is proposed, which, based on orthodox genetic algorithm in terms of algorithm selection, crossover and mutation, increases the possibility of genetic oriented-mutation and simplifies the multi-target optimization into mono-target one. Whereby the improved path planning method, the problem of finding a well-matched scale factor can be settled efficiently.And it can be proved through Matlab simulation that this improved genetic algorithm is of great efficiency and practicability. Keywords:genetic algorithm;path planning for mobile robot;genetic oriented-mutation;Matlab simulation. 0 引言 1 直角坐标法环境建模 [1] 1.1建模前提 目前，人工智能和机器人学 的发展极为迅速， 移动机器人的路径规划问题是其中重要的研究部 笔者的研究重点是在全局信息已知的前提下， 分。所谓路径规划，就是在一定的约束条件下，寻 寻找一条从起始点到目标点最短避障路径。所以， 找出一条从起始点到目标点的最佳路线。一般的约 将机器人视为一个质点，不考虑