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

Network Robot Tractors Associating with Human

与人类联动的网络机器人拖拉机

基本信息

批准号：
13660251
负责人：
NOGUCHI Noboru
金额：
$ 1.92万
依托单位：
HOKKAIDO UNIVERSITY
依托单位国家：
日本
项目类别：
Grant-in-Aid for Scientific Research (C)
财政年份：
2001
资助国家：
日本
起止时间：
2001 至 2003
项目状态：
已结题

项目摘要

The goal of the research is to make the vehicle follow the leading vehicle at a given relative distance and a given angle. The 2-D laser range finder (SICK Co.ltd.) mounted on the front of the follower can provide the outline of the leading vehicle and some noises. To make the follower recognize the leading vehicle, pattern recognition method was adopted in the algorithm. two-dimensional shape of the leading vehicle as a template was made by surveying with a total-station. The basic idea of the algorithm is to measure the relative angle and range from the follower to the leading vehicle in each control step of the follower. The 2-D laser range finder can sense the range to obstacles within 80 m and over 180 deg. The developed algorithm can provide these navigation signals at update rate of 10 Hz by matching a template with an actual range data set. Then, the control algorithm for both lateral and longitudinal motions was also developed. Lateral error is controlled by steering and longitudinal motion is controlled by transmission and engine speed change. Finally, we investigated our vehicle follower control algorithm by field tests. The leading vehicle was driven by human and the following algorithm was installed to our robot tractor. The follower could track the leading vehicle under various given offset and distance. To evaluate the following accuracy, a RTK/GPS and a Fiber Optic Gyroscope were mounted on both the leading vehicle and the follower. The overall tracking error indicated about 10 cm of lateral and 3 deg of offset angle.

研究的目的是使车辆以给定的相对距离和给定的角度跟随前车。安装在从动器前部的二维激光测距仪（SICK公司）可以提供领先车辆的轮廓和一些噪音。该算法采用模式识别方法，使跟随车辆能够识别前车。通过全站仪测量，得到了以前车为模板的二维形状。该算法的基本思想是在从动器的每个控制步骤中测量从动器与前导车辆的相对角度和距离。2d激光测距仪可以感知80米范围内180度以上障碍物的距离。开发的算法可以通过将模板与实际距离数据集相匹配，以10 Hz的更新速率提供这些导航信号。然后，提出了横向运动和纵向运动的控制算法。横向误差由转向控制，纵向运动由变速器和发动机变速控制。最后，通过现场试验对车辆跟随器控制算法进行了验证。前面的车辆由人驾驶，下面的算法安装在我们的机器人拖拉机上。在给定的不同偏移量和距离下，跟随车辆可以跟踪前车。为了评估跟踪精度，在前导车辆和跟随车辆上分别安装了RTK/GPS和光纤陀螺仪。总体跟踪误差约为10厘米的横向和3度的偏置角。