基于视觉导引的AGV系统研制

发布时间：2018-03-23 05:26

本文选题：自动导引小车　切入点：视觉导引　出处：《中国计量大学》2016年硕士论文　论文类型：学位论文

【摘要】：制造业中的物流体系往往采用人工方式进行物料搬运、装配,但是随着劳动力成本的上升,这种方式暴露出物流效率低、可靠性低、成本高的问题。因此AGV技术迎来全新的发展机遇,但是传统的磁导引AGV获取的偏差信息单一,控制效果差,无轨导引方式价格又十分昂贵,本文引入视觉导引技术,对制造业生产线的自动化仓储物流问题进行了研究,主要研究内容如下:首先,为解决AGV传统导引方式获取偏差信息单一,控制效果差,而无轨导引方式价格非常昂贵的问题,根据模块化的设计思想,完成视觉导引AGV的总体方案设计,包括AGV机械结构设计、各功能模块的硬件部分设计以及车载控制系统软件设计。其次,阐述了基于机器视觉的AGV导引原理,提出采用AGV横向偏移量e和方向偏差角θ作为导引信息。为实现对导引信息的提取,本文基于Visual Studio 2008集成开发环境配置OpenCV 2.4.4开发包进行图像处理,包括图像滤波、阈值分割、边缘检测、色带识别以及色带中心线拟合。再次,分析了差速式视觉导引AGV的轨迹纠偏过程,并建立AGV轨迹纠偏数学模型。根据建模结果,纠偏过程中的两个控制变量存在耦合关系,因此设计了二维模糊控制器实现轨迹纠偏过程的非线性控制。最后,根据总体方案设计,搭建了视觉导引AGV实验平台;通过实验验证所设计的AGV在图像处理、负载能力、纠偏能力、停障能力以及定位精度方面的性能。
[Abstract]:The logistics system in the manufacturing industry often adopts the manual way to carry on the material handling, the assembly, but along with the labor cost rise, this kind of way exposes the logistics efficiency is low, the reliability is low, The problem of high cost. Therefore, AGV technology ushered in a new development opportunity, but the deviation information obtained by traditional magnetic guidance AGV is single, the control effect is poor, and the price of trackless guidance is very expensive. The automatic warehousing logistics problem of manufacturing line is studied. The main contents are as follows: firstly, in order to solve the problem of single deviation information and poor control effect in the traditional AGV guidance mode, the trackless guidance method is very expensive. According to the modular design idea, the overall scheme design of visual guided AGV is completed, including the mechanical structure design of AGV, the hardware design of each functional module and the software design of vehicle control system. Secondly, This paper expounds the principle of AGV guidance based on machine vision, and proposes to use AGV lateral offset e and directional deviation angle 胃 as guidance information. This paper is based on Visual Studio 2008 integrated development environment to configure OpenCV 2.4.4 development kit for image processing, including image filtering, threshold segmentation, edge detection, ribbon recognition and ribbon centerline fitting. The trajectory correction process of differential vision guided AGV is analyzed, and the mathematical model of AGV trajectory correction is established. Therefore, a two-dimensional fuzzy controller is designed to realize the nonlinear control of trajectory correction process. Finally, according to the overall scheme design, a visual guided AGV experimental platform is built, and the designed AGV is verified in image processing and load capacity through experiments. Ability to correct deviation, stop obstacle and position accuracy.
【学位授予单位】：中国计量大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP391.41;TP242

【参考文献】