节点文献
有杆泵抽油井工况远程监测与故障诊断系统研究
Research on Remote Monitoring and Fault Diagnosis of Sucker Rod Pumping System
【作者】 王昱;
【导师】 黎洪生;
【作者基本信息】 武汉理工大学 , 控制理论与控制工程, 2006, 硕士
【摘要】 在国内外原油生产中,机械采油占据了首要地位,而在机械采油中又以有杆泵抽油为主。有杆泵采油设备对整个原油生产起着至关重要的作用,而由于抽油装置特别是井下装置的故障所造成的经济损失是极其巨大的。鉴于此,本文以有杆泵抽油井工况远程监测与故障诊断系统为研究对象,针对系统中涉及到的抽油井数据远程传输、监测系统的网络共享以及深井泵故障诊断的关键技术进行了研究。主要研究内容如下: 介绍了有杆泵抽油井远程监测与故障诊断系统的体系架构。该系统通过安装于抽油井现场的远程遥测终端采集抽油井各种工况参数并对紧急故障进行实时诊断,通过GPRS将数据传至中心监测站,中心监测站对深井泵工作状况进行在线诊断,并设置远程访问服务器,使得授权用户均可通过虚拟专用网对抽油井工况进行远程实时监测。 在抽油井工作参数的远程传输方面,研究了基于GPRS的多客户通信服务程序的设计。采用异步非阻塞I/O通信方式,解决了中心监测站和多个RTU实时通信问题,减小了系统资源开销。通过抽油井数据的O/R映射方式使数据入库的过程更加便捷。 针对抽油井工况的远程实时监测,研究了提高远程监测实时性的方法。采用消息传递机制实现了远程监测用户和中心监测站的实时互动,即通过消息发布和订阅的方式,实现实时数据从服务器到客户端的“推出”,达到了远程实时监测的效果。用虚拟专用网和多级用户权限认证技术保证抽油井数据网络传输的安全性。 对于深井泵的故障诊断,研究了示功图的分类识别方法。通过多边形近似算法对示功图数据进行压缩,采用傅立叶描述子对示功图进行特征提取,用支持向量机对反映不同深井泵工作状况的示功图进行识别分类。给出了实验过程及结果,实验结果表明该方法具有较好的分类能力。
【Abstract】 Mechanical oil pumping is the main method of crude oil production in the world, and sucker rod oil pumping system is the most popular method of mechanical oil pumping. The equipments of sucker rod oil pumping system are vital important in the whole production of crude oil. And the loss, which is caused by the fault of sucker rod oil pumping equipment, especially those equipments which are in the deep well, is extremely huge. Therefore, this paper researches on the remote monitoring and fault diagnosis system of sucker rod oil well, and researches on the key techniques of the system, including the remote transmit of the oil well’s working parameters, the remote share of monitoring system and the fault diagnosis of sucker rod. Main achievements of this paper are showed in the next paragraphs.This paper introduces the architecture of remote monitoring and fault diagnosis system, which collects data from the field and diagnoses emergent fault of oil well by RTU, and transmits the data to central monitoring station by GPRS. The central monitoring station, which diagnoses the fault of sucker rod, is built-up with remote accessing server, which makes remote authorized users could monitor the real-time working condition of oil well through VPN.As regards the remote transmission of the oil wells’ working parameters, the GPRS based multi-client data communication application, which is based on asynchronous non-blocking I/O technique, and achieves the real-time communication between central monitoring station and RTU, is researched. This method reduces the cost of OS’s resource. Besides, The O/R mapping is used to make the operation on database more conveniently.As regards the remote real-time monitoring of oil wells’ working condition, this paper researches on a method which improves the real-time ability of remote monitoring. Message transmission system is employed to achieve the real-time communication between central monitoring station and remote monitoring users. "Push" the data from central monitoring station to remote monitoring users by publishing and subscribing messages. This method achieves real-time monitoring of oil well. Besides, Virtual Private Network (VPN) and authority authentication technique are used to ensure the safety of network transmission.As for the fault diagnosis of sucker rod, this paper researches on the classification method of dynamometer card. Compresses the dynamometer card data by polygon approximation algorithm, and adopts Fourier Descriptors (FD) to extract the feature of underground dynamometer card. Uses the Support Vector Machine (SVM) to classify different kinds of dynamometer cards which reflect different working conditions of sucker rod. In the end the experiment and results are given, and the experiment proves that this method has satisfying results.
- 【网络出版投稿人】 武汉理工大学 【网络出版年期】2006年 08期
- 【分类号】TP277
- 【被引频次】29
- 【下载频次】593