抽油机音频故障分析研究与应用  

Research and Application of Audio Fault Analysis of Pumping Unit

作　　者：李兴 朱苏青 刘松林 Li Xing;Zhu Suqing;Liu Songlin(Sinopec Jiangsu Oilfield Company,Yangzhou,225009,China)

机构地区：[1]中国石化股份有限公司江苏油田分公司,江苏扬州225009

出　　处：《石油化工自动化》2022年第5期82-86,103,共6页Automation in Petro-chemical Industry

摘　　要：在目前油田生产信息化系统条件下,中心控制室对抽油机设备的故障监控、发现与预警的手段和方法较少,仅能通过视频的方式进行巡回检查。针对当前现状,开展抽油机故障音频及预警技术研究与应用,利用物联网、机器学习、大数据分析等技术,实现连续性的设备监控,及时、精准地发现和诊断抽油机机械故障并预警,避免机械事故的发生,减轻员工的劳动强度,提高设备信息化管理水平。Under the conditions of current oilfield production information system, the central control room has few means and methods to monitor, discover and early warn the fault of pumping unit, and the patrol inspection can only be conducted by video. In response to the current status, the research and application of pumping unit fault audio and early warning technology is carried out. The Internet of Things, machine learning, big data analysis and other technologies are utilized to realize continuous equipment monitoring, to timely and accurately detect and diagnose mechanical faults of pumping units, and to early warn. Labor intensity of employees is reduced and equipment information management is improved.

关 键 词：梅尔频率倒谱系数 语谱图 深度残差网络 抽油机音频 故障识别 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置]