基于MOPSO-CNN模型的压缩机气阀故障诊断技术  

Compressor valve fault diagnosis technology based on a MOPSO-CNN model

作　　者：张平 孙霖 史建超 李亚民 ZHANG Ping;SUN Lin;SHI JianChao;LI YaMin(Western Branch,State Pipe Network Group United Pipe Co.,Ltd.,Urumqi 830011;CNOOC Energy Development Equipment Technology Co.,Ltd.,Tianjin 300450;Tangshan Xingshi Technology Co.,Ltd.,Tangshan 063000,China)

机构地区：[1]国家管网集团联合管道有限责任公司西部分公司,乌鲁木齐830011 [2]中海油能源发展装备技术有限公司,天津300450 [3]唐山行世科技有限公司,唐山063000

出　　处：《北京化工大学学报（自然科学版）》2024年第3期107-113,共7页Journal of Beijing University of Chemical Technology(Natural Science Edition)

摘　　要：针对传统方法难以提取有效的气阀故障信号,无法建立气阀状态与信号间复杂映射关系的问题,将气阀振动信号转为频域信号输入卷积神经网络(CNN)进行气阀状态诊断,采用多目标粒子群算法(MOPSO)对CNN的超参数进行优化,构建自适应CNN模型,并针对分类结果进行可视化分析,探讨了不同训练测试比对分类准确率的影响。结果表明:MOPSO-CNN模型可完成数据降噪、特征提取和故障分类的一贯式处理,实现端到端的故障诊断,其分类准确率和训练时间均优于传统方法;通过t-分布随机邻域嵌入(t-distributed stochastic neighbor embedding,t-SNE)可视化分析,证明了CNN模型在逐层特征提取和特征分离上的优越性;所建立模型在不同训练测试比的条件下表现良好,对训练数据的需求量不大。研究结果可为往复式压缩机气阀故障诊断提供实际参考。It is difficult to extract effective valve fault signals by traditional methods,and the complex mapping relationship between valve states and signals cannot be established.In response to this problem,valve vibration signals have been converted into frequency domain signals and input into convolutional neural networks(CNN).Multi-objective particle swarm optimization(MOPSO)was used to optimize the hyperparameters of CNN and an a daptive CNN model was constructed in order to analyze the classification results visually.The influence of different training tests on the classification accuracy is discussed.The results show that MOPSO-CNN model can achieve consistent processing of data denoising,feature extraction and fault classification,and achieve end-to-end fault diagnosis.The classification accuracy and training time of the MOPSO-CNN model are better than traditional methods.The superiority of the CNN model in terms of feature extraction and feature separation was demonstrated using t-distributed stochastic neighbor embedding(t-SNE)visualization analysis.The model performs well under different training test ratios and requires little training data.These results provide a practical reference for reciprocating compressor valve fault diagnosis.

关 键 词：多目标粒子群算法(MOPSO) 卷积神经网络(CNN) 压缩机 气阀 故障诊断 

分 类 号：TH457[机械工程—机械制造及自动化]