基于机器视觉的智能配送单元导航系统研究

发布时间：2018-03-04 15:36

本文选题：智能配送单元　切入点：机器视觉　出处：《合肥工业大学》2017年硕士论文　论文类型：学位论文

【摘要】：随着工业4.0概念的提出,智能制造成为今后制造业的发展趋势,它要求工厂、生产和物流等均实现智能化,而其中智能物流是实现智能制造的核心目标之一。本课题正是在这种发展趋势下,提出以AGV(Automated Guided Vehicle,自动导引车)为基础,开发一种智能配送单元(Intelligent Distribution Unit),用于智能工厂中的物料配送等环节中。随着近年来图像处理技术的进步和计算机视觉技术的发展,以及低功耗高集成度的嵌入式处理器研发技术的进步,基于嵌入式Linux系统的视觉导航方式具有非常广阔的研究和应用前景。基于机器视觉的智能配送单元系统设计的研究内容主要涉及两个方面:图像的采集与处理和控制器设计。图像的采集与处理方面,本文建立了一个典型的图像采集系统,并对机载摄像头的安装位置以及路径标示线的设置进行了讨论;对于采集后的图像,在经过灰度化和二值化处理后,采用形态学处理进行边缘提取,然后利用逆透视变换将导引轨迹边缘线映射为世界坐标系中的直线,最终利用霍夫变换得到边缘线的直线方程并提取出导航偏差参数。控制器设计方面,首先以车载视觉传感器所采集图像中智能配送单元的控制中心特征与导引轨迹的期望特征之间的偏差为依据,建立智能配送单元运动学模型,然后采用改进的滑膜变结构控制方法,并通过基于遗传算法和非线性规划的函数寻优算法对控制器参数进行优化。最后,通过在MATLAB/Simulink上进行两组对比仿真试验,分别是改进滑模变结构控制仿真和参数优化的滑模变结构控制仿真。经过对相关实验结果与数据进行分析、对比验证本课题中导航方式的可行性和导航算法的有效性。
[Abstract]:With the development of the concept of industry 4.0, intelligent manufacturing has become the development trend of manufacturing industry in the future. It requires factories, production and logistics to be intelligent. Among them, intelligent logistics is one of the core goals to realize intelligent manufacturing. This topic is based on AGV(Automated Guided vehicle. An intelligent Distribution unit is developed for material distribution in intelligent factories. With the progress of image processing technology and the development of computer vision technology in recent years, And the development of low power and high integration embedded processor technology, The visual navigation mode based on embedded Linux system has a very broad research and application prospect. The design of intelligent distribution unit system based on machine vision mainly involves two aspects: image acquisition and processing; Controller design. Image acquisition and processing, In this paper, a typical image acquisition system is established, and the installation position of the airborne camera and the setting of the path marking line are discussed. The edge of the guided trajectory is mapped to a straight line in the world coordinate system by using the inverse perspective transformation. Finally, the linear equation of the edge line is obtained by using the Hough transform and the navigation deviation parameters are extracted. Firstly, based on the deviation between the control center feature of intelligent distribution unit and the expected feature of guidance track, the kinematics model of intelligent distribution unit is established. Then the improved synovial variable structure control method is adopted, and the controller parameters are optimized by the function optimization algorithm based on genetic algorithm and nonlinear programming. Finally, two sets of comparative simulation experiments are carried out on MATLAB/Simulink. The simulation of improved sliding mode variable structure control and parameter optimization of sliding mode variable structure control is presented respectively. The feasibility of navigation mode and the effectiveness of navigation algorithm are compared and verified by analyzing the relevant experimental results and data.
【学位授予单位】：合肥工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP391.41;F252.1

【参考文献】