基于标识线导航的自动导引车跟踪控制  

TRACKING CONTROL OF AUTOMATIC GUIDED VEHICLE BASED ON LANE MARKER NAVIGATION

作　　者：陈无畏[1] 孙海涛[1] 李碧春[1] 王启瑞[1] 李进[1] 

机构地区：[1]合肥工业大学机械与汽车工程学院,合肥230009

出　　处：《机械工程学报》2006年第8期164-170,共7页Journal of Mechanical Engineering

基　　金：安徽省自然科学基金(050420302);江苏省汽车工程重点实验室资助项目(QC200505)资助。

摘　　要：为实现自动导引车对标识线路径的识别和跟踪,提出一种原理简单的视觉图像处理方法,包括阈值分割、滤波、边缘检测、变形矫正和改进的车体相对位置参数提取等。此方法能有效降低噪声干扰、图像变形对参数提取的影响以及增加图像处理的实时性；还提出一种适于自动导引车跟踪控制的最优控制策略。为验证理论分析的正确性,先对标识线图像进行处理,然后进行自动导引车的跟踪控制试验。结果表明,采用此种图像处理方法和最优控制策略的自动导引车,具有较为准确和可靠的路径跟踪效果。In order to make an automatic guided vehicle （AGV） identify the guidance path, a method about simplified image processing, including image segmentation, filter, border detection, distortion recti fying and an improved way to get the relative position of the body is presented. The method can reduce noise disturbance and the influence of the distorted image and make AGV perform the real-time image processing. Meanwhile, an optimal control strategy for path tracking controlling of the AGV is presented. In order to testify the correctness of the analysis of theory, the image of lane marker is processed and then the experiment of the path tracking is carried out. The results show that AGV with the method and the optimal control strategy could carry out the path tracking accurately and reliably.

关 键 词：自动导引车 最优阈值 路径跟踪 图像处理 

分 类 号：TP242.6[自动化与计算机技术—检测技术与自动化装置]