刮板机异常监测系统设计  

Design of Scraper Anomaly Monitoring System

作　　者：齐健 包国强 尉维洁 刘峰 高磊 陈廷官 冯化一 吴昊 冯俊 QI Jian;BAO Guoqiang;WEI Weijie;LIU Feng;GAO Lei;CHEN Tingguan;FENG Huayi;WU Hao;FENG Jun(National Energy Group Xinjiang Energy Co.,Ltd.,Urmqi 830000,China;Tianjin Meiteng Technology Co.,Ltd.,Tianjin 300000,China)

机构地区：[1]国家能源集团新疆能源有限责任公司,新疆乌鲁木齐830000 [2]天津美腾科技股份有限公司,天津300000

出　　处：《自动化仪表》2024年第8期58-63,共6页Process Automation Instrumentation

摘　　要：为了实时识别刮板机上的异常小目标,确保刮板机的正常、安全运行,设计了基于机器视觉的刮板机异常监测系统。数据采集层的工业摄像机采集单元基于机器视觉原理获取刮板机实时监测图像,经通用串行总线(USB)接口传输图像给数据处理层。对采集的刮板机图像作降噪、增强处理后,通过数据传输层的基于现场可编程门阵列(FPGA)的以太网通信模块完成图像的上传。数据监测层的异常状态监测模块依据接收到的图像,创新性地调用改进的掩蔽区域卷积神经网络(Mask R-CNN)模型,由异常报警模块发送报警信息,并通过数据显示层呈现异常监测结果及报警提示信息,以实现刮板机异常监测。试验结果表明:该系统处理后的刮板机图像峰值信噪比显著提升、均方根误差显著降低;增强后的刮板机图像异常识别损失更低。该系统可识别刮板机不同类型的异常,并标记异常目标。To real-time identify anomaly small targets on the scraper,ensure the normal and safe operation of the scraper,the anomaly monitoring system of scraper based on machine vision is designed.The industrial camera acquisition unit in the data acquisition layer acquires real-time monitoring images of the scraper based on the principle of machine vision and transmits images to the data processing layer through the universal serial bus(USB)interface.After noise reduction and enhancement of the collected scraper image,the image is uploaded through the Ethernet communication module based on field programmable gate array(FPGA)in the data transmission layer.The anomaly state monitoring module in the data monitoring layer innovatively calls the improved mask region-convolutional neural network(Mask R-CNN)model based on the received images,sends alarm information by the anomaly alarm module,and realizes the anomaly monitoring of the scraper by presenting the anomaly monitoring results and alarm prompt messages through the data display layer.The test results show that:the peak signal-to-noise ratio of the scraper image processed by the system is significantly improved,the root mean square error is significantly reduced;the anomaly recognition loss of the enhanced scraper image is lower.The system can realize the recognition of different types of anomalies in the scraper and mark the abnormal targets.

关 键 词：机器视觉 刮板机 异常监测 图像异常 现场可编程门阵列 掩蔽区域卷积神经网络模型 

分 类 号：TH-39[机械工程]