基于p-V图特征的往复式压缩机缸内易损件故障诊断技术研究--诊断模型  

Research on Fault Diagnosis Technique of Vulnerable Parts in Reciprocating Compressor Cylinder Based on p-V Diagram Characteristics-Fault Diagnosis Model

作　　者：杨毅帆 祝钟青 张戟 陈青松 肖植 吴伟烽[1] YANG Yi-fan;ZHU Zhong-qing;ZHANG Ji;CHEN Qing-song;XIAO Zhi;WU Wei-feng(School of Energy and Power Engineering,Xi′an Jiaotong University,Xi′an 710049,China;Sinopec-SK(Wuhan)Petrochemical Co.,Ltd.,Wuhan 430082,China;Chengdu Meixun Testing Equipment Co.,Ltd.,Chengdu 610041,China)

机构地区：[1]西安交通大学能源与动力工程学院,陕西西安710049 [2]中韩(武汉)石油化工有限公司,湖北武汉430082 [3]成都美讯检测有限公司,四川成都610041

出　　处：《压缩机技术》2023年第2期1-7,共7页Compressor Technology

基　　金：国家自然科学基金项目(52076166)。

摘　　要：近年来,对大型往复式压缩机越来越多地提出了长周期运行的要求,合理有效地应用故障诊断技术维持压缩机机组的可靠安全运行,对于保障企业经济效益,预防安全事故的发生具有十分重要的意义。将BP神经网络、GA-BP神经网络、POS-BP神经网络应用于往复压缩机p-V图故障诊断中,结合故障诊断实验提取出表征压缩机工作状态的关键p-V图特征参数,以膨胀过程综合指数为关键判断指标,完成了神经网络的搭建以及有效性验证,实现了p-V图特征参数、神经网络在压缩机故障诊断中的结合。In recent years,more and more requirements for long-term operation of large type of reciprocating compressors have been put forward.Reasonable and effective application of fault diagnosis technology to maintain the reliable and safe operation of compressor units is an effective mean to ensure the economic benefits of enterprises and prevent safety accidents.In this paper,BP neural network,GA-BP neural network and POS-BP neural network are applied to the fault diagnosis of reciprocating compressor with p-V diagram.Based on the fault diagnosis experiment,taking the comprehensive index of expansion process as the key judgment index,a neural network based on p-V diagram characteristic parameters for fault diagnosis of reciprocating compressors is constructed and verified.The combination of p-V diagram characteristic parameters and neural network in compressor fault diagnosis is realized.

关 键 词：往复压缩机 故障诊断 神经网络 故障模拟实验 

分 类 号：TB652[一般工业技术—制冷工程] TH457[机械工程—机械制造及自动化]