项目名称: 基于机器视觉和惯性测量的轮式滑动转向移动机器人定位导航与遥感知
项目编号: No.60875056
项目类型: 面上项目
立项/批准年度: 2009
项目学科: 金属学与金属工艺
项目作者: 刘景泰
作者单位: 南开大学
项目金额: 30万元
中文摘要: 轮式滑动转向移动机器人(Skid-Steered Mobile Robot)具有结构简单、高度灵活的特点,在户外作业、军事、勘探等领域中得到了广泛的应用。这种机器人的转向是通过车轮打滑来实现的,其运动学模型和动力学模型非常复杂,在户外复杂环境下的定位和导航都相当困难。本项目中我们首先针对四轮独立驱动移动机器人建立运动学和动力学模型,应用视传感器(Vision Odemetry)、惯性测量装置(IMU)和光电码盘来对移动机器人进行定位和导航控制。利用异类多传感器的信息融合技术,基于扩展Kalman滤波方法来对移动机器人的位置和运动速度进行估计和预测,并研究新的机器人运动打滑模型和传感器的测量误差模型。设计了一套基于高精度网络摄像机的全局轨迹实时视觉测量系统,设定和记录机器人运动的基准轨迹。通过比较实际轨迹与基准轨迹之间的误差,可以衡量出定位与导航系统的性能。本项目已经能够对移动机器人高速移动时进行较为精准角度的快速运动控制。
中文关键词: 轮式移动机器人;滑动转向模型;多传感器信息融合;定位与导航;高速控制
英文摘要: The Skid-Steered Mobile Robot has the characteristics of simple structure and highly flexible,it has been widely used in the outdoor work,military,exploration and other fields.The robot's steering is achieved by wheel slip,its kinematic and dynamic model is very complex,the positioning and navigation is very complex in outdoor environment.In this project,we first estblished the kinematic and dynamic model of the four-wheel independent drive mobile robot,and uses the Vision Odemetry,inertial measurement unit(IMU) and the optical encoder to make the mobile robot's positioning and navigation control.Use the heterogrnrous multi-sensor information fusion technology,bases on extended Kalman filter approach to estimates and predicts the position and velocity of the mobile robot.We design a overall trajecory of real time visual measurement system which based on a high-precision network camera to set and record the robot's trajectoties.By comparing the error between the actual trajectory and the reference one,we can measure the performance of the position and navigation system.This project has been able to make a more accurate control for the mobile robot when it moves in a high speed.
英文关键词: Wheeled Mobile Robot;Sliding steering model;Multi-sensor information fusion;Positionging and navigation;High-speed control