利用机器视觉远程监测煤矿带式输送机故障

Remote monitoring of faults of coal mine belt conveyor by machine vision

为及时排除煤矿带式输送机故障,避免造成连锁事故,确保设备长距离连续输送物料的工作效率,提出一种机器视觉远程故障监测方法。利用电荷耦合器件工业相机、图像处理器及照明装置、除尘装置等部件,搭建机器视觉远程监测架构,在输送机托辊、胶带等主要部件上安装多种类传感器。针对运行中的KM-SSJ型号输送机,搭建实验场景,增强、降噪、分割处理输送带撕裂图像。利用支持向量机判定图像中含有的异常故障。各阶段处理的视觉效果验证出,所提方法能够从复杂的环境背景中成功分离出监测目标,为维修人员及时发现异常故障、迅速作出相应举措提供了有效依据。经综合分析监测故障种类与用时发现,该方法对多种故障均具有良好的监测能力,且实时性较为理想。

In order to eliminate the faults of coal mine belt conveyor in time,avoid chain accidents and ensure the working efficiency of long-distance continuous conveying of materials,a machine vision remote fault monitoring method was proposed.The charge coupled device industrial camera,image processor,lighting device,dust removal device and other components were used to build the machine vision remote monitoring framework,and various sensors were installed on the main components such as conveyor idler and belt.For the running KM-SSJ conveyor,the experimental scene was set up,and the image of the torn conveyor belt was enhanced,denoised and segmented.The support vector machine was used to determine the abnormal faults contained in the image.The visual effects of each stage of processing verified that the proposed method successfully separates the monitoring targets from the complex environmental background,so as to provide an effective basis for maintenance personnel to find abnormal faults in time and take corresponding measures quickly.Through comprehensive analysis of the types and time of monitoring faults,it was found that this method has good monitoring ability for a variety of faults,and the real-time performance is ideal.