文章检索
首页» 过刊浏览» 2021» Vol.6» Issue(2) 282-291     DOI : 10.3969/j.issn.2096-1693.2021.02.022
最新目录| | 过刊浏览| 高级检索
基于机器视觉的注水泵智能监控方法研究
刘珈铨,侯磊,毕新忠 ,段闯,任赆慈
1 中国石油大学(北京)机械与储运工程学院,北京 102249 2 中国石油大学(北京)油气管道输送安全国家工程实验室/石油工程教育部重点实验室,北京 102249 3 中国石化胜利油田有限公司桩西采油厂,东营 257237
Research into an intelligent monitoring method based on machine vision for a water injection pump
LIU Jiaquan1,2, HOU Lei1,2, BI Xinzhong3 , DUAN Chuang3 , REN Jinci
1 College of Mechanical and Transportation Engineering, China University of Petroleum-Beijing, Beijing 102249, China 2 National Engineering Laboratory for Pipeline Safety/MOE Key Laboratory of Petroleum Engineering, China University of Petroleum-Beijing, Beijing 102249, China 3 Zhuangxi Oil Production Company of Sinopec Shengli Oilfield, Dongying 257237, China

全文:   HTML (1 KB) 
文章导读  
摘要  随着智慧油田建设的高速推进,对油田设备的智能监控技术提出了更高要求。目前油田储存有大量监控 视频,面向人脸识别、泄漏检测等领域的视频已有研究,但面向旋转设备的视频还未被充分挖掘。针对某油田 注水泵监控视频图像噪点多及干扰目标多等问题,本文提出基于Faster Region Convolution Neural Network(Faster R-CNN)的注水泵智能监控方法。利用特征提取网络(FEN)对输入图像的柱塞区域进行特征提取;利用区域推荐 网络(RPN)基于已提取特征图生成一系列候选区域;利用目标检测网络(ODN)综合FEN提取的特征图及RPN产 生的候选区域进行柱塞区域识别和柱塞区域坐标确定,实现了变化背景中泵柱塞区域的自动检测。通过二值化 与高斯滤波对柱塞区域图像进行预处理,减少图像噪点以使柱塞运动期间的帧间差值显著增大。通过帧间差分 法判别各帧中柱塞区域的当前运动状态,并基于多个帧间差值的运动状态判定标准判别柱塞区域的整体运动状 态,实现了泵柱塞运动状态的智能监控。与传统的基于数据采集与监控系统(SCADA)中数值参数的监控方法相 比,基于机器视觉的智能监控方法更加准确直观。油田生产现场的真实视频验证表明,该方法能够快速准确地 对注水泵柱塞的运动状态进行检测,检测总准确率达到 96.75%,显著高于传统的帧间差分法及光流法,能够为 油田设备智能化管理提供技术支撑。
服务
把本文推荐给朋友
加入我的书架
加入引用管理器
关键词 : 运动目标检测;深度学习;注水泵;智能监控
Abstract
With the development of intelligent oil fields, the technology for intelligent surveillance of equipment in oil fields    
is required to reach a higher level. At present, there is a large number of surveillance videos stored in the databases of oil fields.    
There has been research on video interpretation for face recognition, leak detection and other fields, but the video for rotating    
equipment has not been fully exploited. In order to solve the problem of serious noise and various kinds of interference targets    
in the surveillance video images of water injection pumps in oil fields, an intelligent machine vision method based on the Faster    
Region Convolution Neural Network (Faster R-CNN) algorithm is introduced. Through the Feature Extraction Network (FEN),    
the image features of the plunger region of the input image are effectively extracted. Through the Region Proposal Network    
(RPN), a series of candidate regions are generated based on the extracted feature maps. Through the Object Detection Network    
(ODN), the feature maps extracted by FEN and the candidate regions generated by RPN are integrated to identify the plunger    
region and determine the coordinates of the plunger region. Therefore, automatic detection of the precise position of the plunger    
region of the pump in the changing background is realized. By using a binarization algorithm and a Gaussian filtering algorithm,    
each image of the plunger region is preprocessed to reduce image noise, which can help the frame difference during plunger    
movement become significantly larger. The current motion state of the plunger region in each frame is determined by the    
inter-frame difference method, and the overall motion state of the plunger region is determined based on the determination of the    
standard motion state through the multiple frame differences, so as to realize the intelligent surveillance of the movement state of    
the pump plunger. Compared with the traditional surveillance method based on numerical parameters from Supervisory Control    
and Data Acquisition (SCADA), this method is more accurate and intuitive. Based on the machine vision method, the character  
istics of serious noise and various kinds of interfering targets in the video image of the water injection pump can be responded    
to effectively. The surveillance video in actual oil field production sites is utilized to verify the outstanding performance of this    
intelligent surveillance technology for water injection pump. The total accuracy reached 96.75   %   , which is significantly higher    
than the traditional inter-frame difference method and the optical flow method. The proposed method can provide technical    
support for the intelligent management of oil field equipment.  


Key words: moving object detection; deep learning; water injection pump; intelligent monitoring
收稿日期: 2021-06-30     
PACS:    
基金资助:国家重点研发计划项目“油气长输管道及储运设施检验评价与安全保障技术”(2016YFC0802100) 资助
通讯作者: houleicup@126.com
引用本文:   
刘珈铨, 侯磊, 毕新忠, 段闯, 任赆慈. 基于机器视觉的注水泵智能监控方法研究. 石油科学通报, 2021, 02: 282-291 LIU Jiaquan, HOU Lei, BI Xinzhong, DUAN Chuang, REN Jinci. Research into an intelligent monitoring method based on machine vision for a water injection pump. Petroleum Science Bulletin, 2021, 02: 282-291.
链接本文:  
版权所有 2016 《石油科学通报》杂志社