基于机器视觉的自动分拣码放系统研究-包装工程.PDF

包 装 工 程 第39 卷 第 15 期 ·194 · PACKAGING ENGINEERING 2018 年8 月 基于机器视觉的自动分拣码放系统研究 杜恩明，张仁朝 （广州铁路职业技术学院，广州 510430 ） 摘要： 目的 为实现包装码垛自动生产线上不同型号工件的分类码放，有效提高生产效率，降低生产成 本。方法 在工件识别应用中，提出一种基于机器视觉、PLC 控制技术、HMI 人机界面的分拣码放系统。 在完成工业机器人相应的硬件及关键算法设计的基础上，利用以太网构建 PLC 网络控制系统，并与上 位机管理系统实现交互。结果 经实验测试，该自动分拣码放系统对工件的识别精度较高，均在 99% 以 上，处理时间较短，均在0.2 s 以下，能够满足生产要求。结论 生产实践表明，该系统提高了工件分拣 效率，减小了工人劳动强度，降低了生产成本，具有很高的推广应用价值。 关键词：机器视觉；工业机器人；PLC 网络；分拣 中图分类号：TB486 ；TP273 文献标识码：A 文章编号： 1001-3563(2018)15-0194-05 DOI ：10.19554/j.cnki.1001-3563.2018.15.030 Automatic Sorting and Stacking System Based on Machine Vision DU En-ming , ZHANG Ren-chao (Guangzhou Institute of Railway Technology, Guangzhou 510430, China) ABSTRACT: The work aims to realize the classification and stacking of different types of workpiece on automatic pro- duction line of packaging and stacking, and improve production efficiency and reduce production cost effectively. During the application of workpiece identification, a kind of sorting system based on machine vision, PLC control technology and HMI man-machine interface was proposed. Based on the design of the corresponding hardware and key algorithms of the industrial robot, the PLC network control system was built with Ethernet and the interaction with the upper management system was realized. According to the experimental test, the automatic sorting and stacking system had a high recognition accuracy of over 99% of the workpieces. The processing time was short, below 0.2 s, which could meet the production requirements. Production practice shows that the system has improved the sorting efficiency of