基于机器视觉的煤矿井下带式输送机堆煤故障监测系统研发

Development of Fault Monitoring System for Coal Stacking of Belt Conveyor in Underground Coal Mine Based on Machine Vision

提出了一种基于机器视觉的煤矿井下带式输送机堆煤故障监测系统。在图像预处理阶段分别采用小波去噪算法和加权平均法进行去噪和灰度化处理,随后采用Canny算子进行堆煤图像边缘检测,进而将检测结果与设定阈值对比后进行堆煤故障判断,若发生故障则控制继电器关闭井下带式输送机。试验结果显示,该系统对堆煤故障识别的准确率可达96%,可为煤矿安全生产、监测设备及装备故障诊断方面的研究人员提供一定的借鉴。

A coal stacking fault monitoring system for belt conveyor in underground coal mine based on machine vision was proposed.The wavelet denoising algorithm and weighted average method were used for denoising and grayscale processing in the image preprocessing stage,then the Canny operator was used for edge detection of coal stacking images,and then the detection results were compared with the set threshold for coal stacking fault judgment.If the fault occurs,the relay is controlled to close the underground belt conveyor.The test results show that the accuracy of this system for coal stacking fault identification can reach 96%.It can provide a certain reference for researchers in the aspects of coal mine safety production,monitoring equipment and equipment fault diagnosis.