文档下载 文档下载导航
设为首页 | 加入收藏
搜索 请输入内容:  
 导航当前位置: 文档下载 > 所有分类 > Fuzzy Logic and Behavior Control Strategy for Autonomous Mobile Robot Mapping
侵权投诉

Fuzzy Logic and Behavior Control Strategy for Autonomous Mobile Robot Mapping

Abstract-- In support of research efforts to develop reactive mobile robots capable of autonomous navigation, we are concentrating on providing human reasoning capabilities as a resource for intelligence and integrating the world mapping and navigation tas

(appears as invited paper in IEEE Intl Conf on Fuzzy Systems, Orlando, FL, June 26-July 2, 1994)

Fuzzy Logic and Behavior Control Strategy for Autonomous Mobile Robot Mapping

Edward Tunstel1 and Mo Jamshidi

CAD Laboratory for Intelligent and Robotic SystemsDepartment of Electrical and Computer Engineering

University of New MexicoAlbuquerque, NM 87131 USA

Abstract -- In support of research efforts to developreactive mobile robots capable of autonomousnavigation, we are concentrating on providing humanreasoning capabilities as a resource for intelligenceand integrating the world mapping and navigationtasks. The underlying philosophy is that anautonomous mobile robot should be able to make amap of its environment as it navigates through thatenvironment. Fuzzy logic control and behavior-basednavigation are proposed as means to implement thisphilosophy. Fuzzy logic provides the approximatereasoning necessary for handling the uncertaintyinherent in mobile robot navigation. Bydecomposing the problem into simpler tasks we candevelop mobile robots that exhibit a sufficient levelof intelligence for performing navigation andmapping. In this paper, we discuss a strategy forincorporating fuzzy reasoning into the framework ofreactive behavior control systems for autonomousmapping.

time, real-time performance and relatively lowcomputation must be maintained. Traditionalapproaches to mobile robot navigation have dealtwith these requirements with the use ofcomputationally intensive planning algorithms andexplicit pre-determined world models [1]. Theresulting systems exhibited robust behavior but at thegreat expense of real-time performance. In addition,the need for significant amounts of memory andprocessing power to store and maintain world modelsprecluded the use of smaller (lightweight) computingsystems. As pointed out in [2], this is notnecessarily a drawback for fixed-base manipulators butit can be a problem for mobile robots for whichcomputational resources must be carried onboard.These issues are of particular importance for mobilerobots intended for use in space [3, 4].

We contend that the theory of fuzzy logic can beapplied to alleviate some of the difficulties alluded toabove. A fuzzy logic based system has the advantagethat it allows the intuitive nature of collision-freenavigation to be easily modeled using linguisticterminology. The computational demand of fuzzyinference systems is considerably lighter than that oftraditional methods. The use of reactive fuzzy controlsystems permits intelligent decisions to be made inreal-time, thus allowing for uninterrupted robotmotion. In intelligent systems, the reactive controlstrategy dictates faster control cycles than moreconventional systems since sensors are sampled at ahigher rate. Moreover, accurate (expensive) sensorsand detailed models of the environment are notabsolutely necessary [5]. The successful managementof uncertain and/or unmodeled data is a provenattribute of fuzzy logic based inference engines.

Furthermore, it has been suggested [6] that areactive, behavior-based control system can reduce thecomputational requirements of traditional systems andstill produce robust autonomous performance. Recentmicro-rover rough terrain navigation work [7] is anexcellent example of this. Behavior control, whichmakes reactive control possible, involvesdecomposing the motion control system into a set ofspecial-purpose routines (behaviors) that achieve

1.Introduction

An aim of intelligent robotics research is todevelop mobile robots capable of navigatingautonomously in unstructured and/or unexploredenvironments. The practicality of such a robot is afunction of its ability to utilize a map to locate itselfand its goals as it navigates throughout. In somesituations a map may be available a priori in whichcase goal locations on the map are specified and pathplanning algorithms are employed to direct the robot'smotion while avoiding known (mapped) obstacles. Amore desirable, and more challenging situation is forthe robot to be able to generate and maintain its mapautonomously.

The development of mobile robots capable ofautonomous mapping and navigation requiresintelligent control strategies that can handle theuncertainties presented by the real world. At the same

1On leave from NASA Jet Propulsion Laboratory,Pasadena, CA.

第1页

你可能喜欢

热门文档

站点地图 | 文档上传 | 侵权投诉 | 手机版
本站所有资源均来自互联网,本站只负责收集和整理,均不承担任何法律责任,如有侵权等其它行为请联系我们.
文档下载 Copyright 2013 doc.guandang.net All Rights Reserved.  -  email