摘要
往复压缩机是海洋石油平台中的A类关键设备,对海洋石油油气的安全生产起着重要作用。传统的预防性维修模式不能有效及时地消除风险,造成维修成本相对较高、时效性差,不能满足设备管理水平提升和数字化建设的要求。为此,在梳理海洋石油常见故障的基础上,建立以振动监测为主的在线监测系统,并基于推理机的机理模型和支持向量机的数学模型的算法实现对往复压缩机的智能预警诊断。通过测试能够有效实现对往复压缩机的智能预警和诊断。
Reciprocating compressor is the key class equipment on offshore oil platform,which plays an important role in the safe production of offshore oil and gas.The traditional preventive maintenance mode causes relatively high maintenance cost,poor timeliness and equipment management level to be improved,which cannot effectively and timely eliminate risks.Therefore,the common faults of offshore oil are sorted out and an on-line monitoring system,based on vibration monitoring,is established.By building the mechanism model based on inference engine and the mathematical model of support vector machine,the on-line monitoring intelligent early warning diagnosis of reciprocating compressor is realized.
作者
李进
LI Jin(China National Offshore Oil Corporation EnerTech Equipment Technology Company Limited, Tianjin 300450, China)
出处
《北京石油化工学院学报》
2021年第2期24-29,共6页
Journal of Beijing Institute of Petrochemical Technology
关键词
智能预警
支持向量机
推理机
在线监测
往复压缩机
intelligent early warning
support vector machine
inference engine
online monitoring
reciprocating compressor
