权益分类	功能权益	普通用户	{{item.name}}会员
{{category.name}}	{{benefitItem.name}}

Coordinated Control for a Non-Holonomic Modular Omnidirectional Mobile Robot

非完整模块化全向移动机器人的协调控制

基本信息

批准号：
11650268
负责人：
HASHIMOTO Masafumi
金额：
$ 2.3万
依托单位：
Hiroshima University
依托单位国家：
日本
项目类别：
Grant-in-Aid for Scientific Research (C)
财政年份：
1999
资助国家：
日本
起止时间：
1999 至 2000
项目状态：
已结题

项目摘要

This project studied a coordinated control for a non-holonomic modular omnidirectional mobile robot, in which wheel units assembled are modularized on their mechanisms and controllers, so that we can easily tailor the robot to individual transfer applications. The project is concluded as follows :1. A dead reckoning system for the robot was designed based on the extended Kalman filter. The dead reckoning system can precisely identify the robot position and the wheel radius of each wheel unit. The accurate estimation in the wheel radius succeeds in reducing the motion interactions among the wheel units.2. A motion control system was designed to coordinate wheel units and to ensure correct robot motion. In the control system, a path-tracking controller decides the desired motion of the robot based on the Ackerman geometry to track its reference path, while the coordinated controller controls wheels to produce the desired motion of the robot. To reduce motion interactions among the wheel units, we constituted two types of wheel units : active wheel units and passive wheel units. The active wheel units produce the desired robot motion, and the passive wheel units correct their own velocities to mimic caster behavior.3. The performance of the dead reckoning and motion control systems was evaluated through both simulations and experiments.4. A motion control system consisting of path-tracking and chassis-tilting controllers was designed for a modular omnidirectional vehicle equipped with chassis-tilting mechanisms. The control system can prevent the robot from tumbling by centrifugal acceleration while the robot following a predetermined path. A dead reckoning system for the robot was also designed.5. As the first step toward applying the proposed methods to real robots, a dead reckoning system for a large omnidirectional AGV in outdoor environments was designed, and the performance was evaluated through fundamental experiments.

本项目研究了一种非完整模块化全方位移动机器人的协调控制，将装配好的轮子单元在其机构和控制器上模块化，以便我们可以很容易地为机器人量身定做，以适应个性化的传输应用。1.设计了一种基于扩展卡尔曼滤波的机器人航位推算系统。航位推算系统可以精确识别机器人的位置和每个车轮单元的车轮半径。对车轮半径的精确估计成功地减少了车轮单元之间的运动干扰。设计了一种运动控制系统，用于协调车轮单元，确保机器人正确运动。在控制系统中，路径跟踪控制器根据Ackerman几何确定机器人的期望运动以跟踪其参考路径，而协调控制器控制车轮以产生机器人的期望运动。为了减少车轮单元之间的运动相互作用，我们构造了两种类型的车轮单元：主动车轮单元和被动车轮单元。主动轮单元产生所需的机器人运动，而被动轮单元校正自身的速度以模仿脚轮的行为。通过仿真和实验对航位推算和运动控制系统的性能进行了评估。针对装有底盘倾动机构的模块化全方位车辆，设计了由路径跟踪控制器和底盘倾摆控制器组成的运动控制系统。该控制系统可以防止机器人在遵循预定路径时因离心力加速而翻滚。还设计了机器人的航位推算系统。作为将提出的方法应用于真实机器人的第一步，设计了室外环境下大型全方位AGV的航位推算系统，并通过基础实验对其性能进行了评估。