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

Development of an Automatically Steered Vehicle Controlled with Multisensor

多传感器控制自动转向车辆的研制

基本信息

批准号：
13650267
负责人：
EMURA Takashi
金额：
$ 2.24万
依托单位：
Tohoku 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 2002
项目状态：
已结题

来源：
https://kaken.nii.ac.jp/grant/KAKENHI-PROJECT-13650267/
关键词：
Automatically Steered Vehicles ALV AHS ITS Sensor-fusion Sonar Vision Sensor Navigation

项目摘要

In order to decrease driver's duty and avoid traffic accidents, it becomes important to develop automatically steered vehicles. Therefore, some automatically steered vehicles have been developed in the world since 1985. The most difficult problem in this development is to make collision avoidance systems that have environmental sensors whose reliability. However, most of environmental sensors for collision avoidance have poor performance if they are used alone. Thus, some methods to fuse information obtained from CCD cameras and a scanning laser range finder were attempted. The methods to avoid collision by images obtained with CCD cameras require a very high-speed computer. In particular, the stereo view requires extremely high-speed computer to find corresponding points from two different images. Thus, the information obtained from CCD cameras has been used usually together with that obtained from the laser range finder.However, the minimum detectable range of the scanning laser range finder is more than 5m. This means that we cannot use it in the sort range detection within 5m. Therefore, the authors used ultrasonic sensors together with the scanning laser range finder and CCD cameras and fused the information obtained with one scanning laser range finder, eight ultrasonic sensors, and four CCD cameras.Experiments were carried out. In experiments, other sensors such as optical fiber gyroscope, steering angle sensor, angle sensor of tire, etc. were added to the above three kinds of sensors, and the information obtained with these many sensors was fused. Experimental results showed that an automatically steered vehicle designed by the authors has good performance and it can avoid collision with high reliability. In this paper, details of the mobile robots designed by the authors and a fusing method to increase performance and reliability for collision avoidance.

为了减轻驾驶员的工作负担，避免交通事故的发生，发展自动驾驶汽车显得尤为重要。因此，自1985年以来，世界上已经开发了一些自动转向车辆。在这种发展中最困难的问题是使防撞系统具有可靠的环境传感器。然而，大多数用于避碰的环境传感器单独使用时性能较差。因此，尝试了一些方法来融合从CCD相机和扫描激光测距仪获得的信息。利用CCD摄像机获得的图像来避免碰撞的方法需要非常高速的计算机。特别是，立体视图需要极高速的计算机从两个不同的图像中找到对应点。因此，CCD摄像机所获得的信息通常与激光测距仪所获得的信息一起使用，而激光扫描测距仪的最小探测距离大于5m。这意味着我们不能在5米范围内的分拣范围检测中使用它。为此，作者将超声波传感器与激光扫描测距仪和CCD摄像机配合使用，将1台激光扫描测距仪、8台超声波传感器和4台CCD摄像机获得的信息进行融合，并进行了实验。在实验中，在上述3种传感器的基础上，加入光纤陀螺、转向角传感器、轮胎转角传感器等传感器，并对多传感器获得的信息进行融合。试验结果表明，作者设计的自动转向小车具有良好的避碰性能和较高的可靠性。本文详细介绍了作者设计的移动的机器人，并提出了一种提高避碰性能和可靠性的融合方法。