页岩气开采往复式压缩机微弱故障智能诊断方法  

Intelligent Diagnosis Method for Weak Faults of Reciprocating Compressors in Shale Gas Mining

作　　者：杨柳 唐洋[2] 杨鑫 梁谷 YANG Liu;TANG Yang;YANG Xin;LIANG Gu(Sichuan Changning Natural Gas Development Co.,Ltd.,Sichuan Chengdu 610051,China;School of Mechatronic Engineering,Southwest Petroleum University,Sichuan Chengdu 610500,China)

机构地区：[1]四川长宁天然气开发有限责任公司,四川成都610051 [2]西南石油大学机电工程学院,四川成都610500

出　　处：《机械设计与制造》2025年第4期375-379,共5页Machinery Design & Manufacture

基　　金：页岩气勘探开发设备管理体系研究(SCCN-FW-2020-0020)。

摘　　要：往复式压缩机是保证页岩气开采中增压的重要设备,其发生故障将可能会给企业造成运营中断,甚至造成重大经济损失。然而,通过往复式压缩机故障诊断系统采集的运行状态数据具有非线性特征耦合的特点,导致其故障特征不明显,系统所集成算法对于特征参数提取困难。因此,为了解决上述问题,将网络中的最小RMSE作为适应度函数,并采用GOA算法自适应调整DBN网络结构的最优参数组合,提出了针对页岩气开采往复式压缩机微弱故障智能诊断方法。结果表明:通过GOA算法实现了自适应调节DBN网络结构参数,避免了手动调节时间长,诊断性能差的问题。与其他方法在页岩气开采往复式压缩机故障诊断中应用比较,本研究方法准确率高达90.78%。本研究表明了无监督特征学习的优越性,为故障信号非平稳性、非周期性的复杂动设备实现精确故障诊断提供理论参考。Aiming at the serious coupling of nonlinear features in the massive and complex data of reciprocating shale gas compres⁃sors,resulting in weak fault characteristics and difficulty in extracting characteristic parameters,a weak fault diagnosis of recip⁃rocating shale gas compressors is proposed.This method takes the RMSE in the network as the fitness function,and uses the GOA algorithm to adaptively adjust the optimal parameter combination of the DBN network structure.The GOA algorithm realizes adap⁃tive adjustment of DBN network structure parameters,avoiding the problems of long manual adjustment time and poor diagnostic performance.This method is applied in the fault diagnosis of reciprocating shale gas compressor.Compared with other methods,the accuracy rate is as high as 90.78%.This research shows the superiority of unsupervised feature learning,and provides a theoreti⁃cal reference for accurate fault diagnosis of complex moving equipment with non-stationary and non-periodic fault signals.

关 键 词：页岩气开采 往复式压缩机 故障诊断 DBN GOA 

分 类 号：TH16[机械工程—机械制造及自动化]