未知环境下机器人导航算法与避障算法研究  

Research on Robot Navigation Algorithm and Obstacle Avoidance Algorithm in Unknown Environment

作　　者：包震洲[1] 俞鸿飞[1] 金文德[2] 李伟达[3] BAO Zhen-zhou;YU Hong-fei;JIN Wen-de;LI Wei-da(State Grid Zhejiang Electric Power Company,Shoushuitan Water Power Plant,Zhejiang Hangzhou310000,China;State Grid Zhejiang Electric Power Company,Zhejiang Hangzhou310000,China;School of Mechanical and Electric Engineering,Soochow University,Jiangsu Suzhou215021,China)

机构地区：[1]国网浙江省电力公司紧水滩水发电厂,浙江杭州310000 [2]国网浙江省电力公司,浙江杭州310000 [3]苏州大学机电工程学院,江苏苏州215021

出　　处：《机械设计与制造》2020年第5期257-260,264,共5页Machinery Design & Manufacture

基　　金：国家自然科学基金项目(61105106、61005071);国网浙江省电力公司科技项目(5211JS160001)。

摘　　要：针对机器人在复杂动态工作环境下如何进行定位和导航的问题做了深入研究,全文阐述了机器人混合导航和地图重建的基本方法。首先,介绍了机器人机械结构与软件平台,并分析了经典双轮差动模型,并总结了机器人的应用范围;然后,通过对视觉数据的数据采集实现了对于环境特征的提取,建立了导航图像和导航机器人导航;最后,建立了基于超声波传感器的模糊避障系统,解决了由于障碍物定位不准确、检测盲区、镜子反射等缺陷导致避障不稳定的问题。经过实验证明,该机器人具有良好的避障性能和导航性能,实现了机器人视觉导航算法。The robots working in the complex dynamic environment how to carry out in-depth study of positioning and navigation,the full text expounds the new method of reconstructing robot hybrid navigation and map. Firstly,this paper introduces the mechanical structure and software platform of robot,analyzes the classic double-wheel differential model,and summarizes the application scope of the robot. Then,the data acquisition of visual data has realized the extraction of environmental characteristics,and established navigation images and navigation robot navigation. Finally,the fuzzy obstacle avoidance system based on ultrasonic sensor is established,which can solve the problem of obstacle avoidance caused by defects such as inaccuracy of obstacle positioning,blind spot detection and mirror reflection. The experimental results show that the robot has good obstacle avoidance performance and navigation performance,and realizes the robot vision navigation algorithm.

关 键 词：机器人 视觉导航 导航定位 未知环境 

分 类 号：TH16[机械工程—机械制造及自动化] TP242.3[自动化与计算机技术—检测技术与自动化装置]