基于机器视觉的矿井提升机首绳抖动监测系统设计与应用

Design and Application of First Rope Jitter Monitoring System for Mine Hoist Based on Machine Vision

在线阅读下载PDF

导出

摘要为了解决提升机首绳抖动带来的隐患问题,本系统基于改进YOLOv5算法,开发首绳抖动视频智能分析监测系统。该系统首先获取视频流,将视频流的图像取出进行双线性插值法处理之后输入到训练好的YOLOv5模型中,进行提升机首绳的异常识别;再通过PyQt控件设计图形化界面对系统进行控制。在实现提升机首绳异常监测的基础上,系统增加了视频储存、视频回放以及输出报警信号等功能,以首绳偏离10 cm抖动区分正负样本,抖动检测精度达到95%以上。 In order to solve the hidden trouble caused by the first rope jitter of the hoist,this system is based on the improved YOLO v5 algorithm to develop the video intelligent analysis and monitoring system for the first rope jitter.Firstly,the system obtains the video stream,and the image of the video stream is taken out and processed by the bilinear interpolation method.Then it is input into the trained YOLO v5 model to identify the anomaly of the first rope of the hoist.Then through the PyQt control design graphical interface to control the system.On the basis of realizing the abnormal monitoring of the first rope of the hoist,the system adds the functions of video storage,video playback and output alarm signals.The positive and negative samples are distinguished by the jitter of the first rope deviating from 10 cm,and the jitter detection accuracy reaches more than 95%.

作者郑伟卫王奕 ZHENG Weiwei;WANG Yi(Cheji Coal Mine of Henan Longyu Energy Co.,Ltd.;School of Information and Control Engineering,China University of Mining and Technology)

机构地区河南龙宇能源股份有限公司车集煤矿中国矿业大学信息与控制工程学院

出处《现代矿业》 CAS 2022年第6期202-204,208,共4页 Modern Mining

关键词钢丝绳抖动机器视觉智能识别 YOLOv5 算法 wire rope jitter machine vision intelligent recognition YOLO v5 algorithm

分类号 TD534 [矿业工程—矿山机电] TP391.41 [自动化与计算机技术—计算机应用技术]

作者简介郑伟卫(1982-),男,工程师,476600河南省永城市。