基于机器视觉的绳轮升降器滑轮总成装配线设计

发布时间：2018-06-06 02:45

  本文选题：PLC + ARM　； 参考：《浙江理工大学》2017年硕士论文

【摘要】：检测及装配是汽车玻璃升降器生产过程中非常重要的工序,尤其是随着汽车行业自动化水平的不断提高,单纯依靠人工视觉检测并进行装配很显然不能满足当今汽车行业的发展要求。在汽车玻璃升降器的生产装配过程中,需要对滑轮位置和尺寸进行检测,并根据检测结果进行装配。传统的检测方法如光电传感器只可以检测升降器滑轮及其它部件的漏装,无法检测到升降器部件的尺寸和部件间相对距离等详细情况,无法适应玻璃升降器快速和可靠制造的需求。尤其是滑轮总成的检测与装配很大程度上影响了绳轮升降器的使用性能和寿命,为此本文提出了一种基于机器视觉的绳轮升降器滑轮总成装配线,旨在克服目前绳轮升降器滑轮总成的人工检测及装配方法的低效率、高成本、精确度差等缺陷,实现滑轮总成自动化检测与装配的目的。本文首先设计了基于机器视觉嵌入式检测系统的总体方案,分析了滑轮检测图像处理算法,特别是对于滑轮表面的正反面检测方法,着重分析了边缘检测算法。为了提高检测效率,对连续检测算法进行了研究。其次,在装配线的布置方面,设计了滑轮总成装配线的机械结构,采用ABAQUS软件进行了仿真,验证了滑轮总成与导轨旋铆过程中卡紧力与铆头速度和摩擦系数的关系。最后,采用PLC与ARM设计了整套设备的控制系统。基于Qt平台上设计了用户操作界面具有视觉系统控制、检测数据显示、实时监控、模式切换等功能,实现了对生产过程进行实时监控。在工厂实际装配线上进行了长时间连续测试,测试结果表明,本文所开发的装置降低了滑轮总成误装率及错装率数据,完全可以替代现有的人工检测环节以及装配环节,降低了企业的用工成本,提高了产品可靠性(误检率0.7‰)。本检测系统也可以满足其它行业的表面质量检测需求,具有重要的理论意义和实际应用价值。
[Abstract]:Detection and assembly are very important processes in the production of automotive glass hoists, especially with the continuous improvement of automation level in the automobile industry. Simply relying on manual visual inspection and assembly is obviously not able to meet the development requirements of the automotive industry today. In the process of production and assembly of automobile glass hoist, the position and size of pulley should be detected and assembled according to the test results. Traditional detection methods such as photoelectric sensors can only detect the leakage of hoist pulleys and other components, and cannot detect details such as the size of the hoist components and the relative distance between them. Unable to meet the need for fast and reliable manufacture of glass hoists. In particular, the detection and assembly of pulley assembly greatly affect the performance and service life of rope wheel hoist. In this paper, a new assembly line of pulley hoist pulley assembly based on machine vision is proposed. The purpose of this paper is to overcome the low efficiency, high cost and poor precision of manual inspection and assembly method of pulley assembly of rope wheel hoist, and to realize the purpose of automatic inspection and assembly of pulley assembly. In this paper, the overall scheme of embedded detection system based on machine vision is designed, and the image processing algorithm of pulley detection is analyzed, especially for the front and back detection of pulley surface, and the edge detection algorithm is emphatically analyzed. In order to improve the detection efficiency, the continuous detection algorithm is studied. Secondly, the mechanical structure of the assembly line of pulley assembly is designed and simulated by ABAQUS software, which verifies the relationship between the clamping force and the speed of riveting head and friction coefficient in the process of pulley assembly and guide rail rotation riveting. Finally, the control system of the whole equipment is designed by using PLC and ARM. The user interface based on QT platform has the functions of visual system control, detection data display, real-time monitoring, mode switching and so on. The real-time monitoring of production process is realized. A long time continuous test has been carried out on the actual assembly line of the factory. The test results show that the device developed in this paper can reduce the misassembly rate and misfit rate data of the pulley assembly, and can completely replace the existing manual inspection and assembly links. The cost of employment is reduced and the product reliability is improved (the false detection rate is 0.7 鈥，

本文编号：1984698