室内环境的元胞自动机SLAM算法-机器人.PDF

第38 卷第2 期 机器人 ROBOT Vol.38, No.2 2016 年3 月 Mar., 2016 DOI ：10.13973/j.cnki.robot.2016.0169 室内环境的元胞自动机SLAM 算法 陈炜楠，刘冠峰，李俊良，苏泽荣，管贻生，张 宏 （广东工业大学机电工程学院，广东广州 510006 ） 摘 要：针对人工室内环境的同时定位与构图（SLAM ）问题，将元胞自动机（CA ）引入到传统SLAM 算法 的迭代过程，建立“SLAM －CA 生长－重定位”的闭环作用机制．利用室内规则环境的特性，以及大多数室内机 器人应用领域已知先验地图的特点，对室内SLAM 问题开展针对性的研究．通过仿真与实验证明，针对人工室内 环境，该算法使得墙壁、出入口与直角特征更加明显，“穿墙”现象得到一定程度解决，改善了构图效果，提高了 定位精度和路径规划可行性． 关键词：同时定位与构图；元胞自动机；移动机器人；自主导航 中图分类号：TP301.6 文献标识码：A 文章编号：1002-0446(2016)-02-0169-09 An Indoor SLAM Algorithm Based on Cellular Automata CHEN Weinan ，LIU Guanfeng ，LI Junliang ，SU Zerong ，GUAN Yisheng ，ZHANG Hong (School of Electro-mechanical Engineering, Guangdong University of Technology, Guangzhou 510006, China) Abstract: To improve simultaneous localization and mapping (SLAM) in indoor environments, the cellular automata (CA) is introduced into the iterative process of traditional SLAM algorithm and “SLAM-CA-relocation” closed-loop mechanism is established. Based on the characteristics of the regular indoor environment and a priori map in most mobile robot applications, the indoor SLAM problem is further studied. In the simulations and experiments in artificial indoor environments, the features of wall, exit, doors and right-angle parts become more clear, the problem of “through walls” is solved to some extent, the mapping quality is improved, the localization precision is increased, and the feasibility of path planning is also enhanced. Keywords: SLAM (simultaneous localization and mapping); cellular automata; mobile robot; autonomous navigation 1 引言（Introduction ）