基于MCKD-HED-CNN的连杆轴承故障诊断  

Fault Diagnosis of Connecting Rod Bearing Based on MCKD-HED-CNN

作　　者：贾继德 沈杨 徐彩莲 JIA Jide;SHEN Yang;XU Cailian(The Higher Educational Key Laboratory for Flexible Manufacturing Equipment Integration of Fujian Province,Xiamen Institute of Technology,Xiamen 361021,China;State Key Laboratory for Manufacturing Systems Engineering,Xi’an Jiaotong University,Xi’an 710049,China)

机构地区：[1]厦门工学院柔性制造装备集成福建省高校重点实验室,福建厦门361021 [2]西安交通大学机械制造系统工程国家重点实验室,陕西西安710049

出　　处：《车用发动机》2024年第1期86-92,共7页Vehicle Engine

基　　金：机械制造系统工程国家重点实验室开放基金(sklms2020021)。

摘　　要：针对强背景噪声干扰下连杆轴承故障诊断难的问题,提出了基于MCKD-HED-CNN的连杆轴承故障诊断方法。首先采用最大相关峭度解卷积(MCKD)对获取的信号进行降噪处理,增强信号中因故障引起的周期性冲击,其次通过Hilbert包络解调(HED)进一步增强周期性冲击,最后将故障特征通过对称点模式(SPD)映射到极坐标图上,并将SDP图像输入CNN网络进行训练,建立连杆轴承故障诊断模型。结果表明:该方法能有效诊断连杆轴承故障,CNN训练样本和测试样本的诊断准确率均为100%。Aiming at the difficult fault diagnosis of connecting rod bearing under strong background noise,the fault diagnosis method of MCKD-HED-CNN was proposed.Firstly,the maximum correlation kurtosis deconvolution(MCKD)algorithm was used to reduce noise and enhance the periodic impact caused by fault.Secondly,the Hilbert envelope demodulation(HED)was used to further enhance the periodic impact.Finally,the fault features were mapped to the polar map by the symmetric point mode(SPD)and the SDP image was input into CNN network for training to establish the fault diagnosis model of connecting rod bearing.The results show that the method can effectively diagnose the fault of connecting rod bearing,and the diagnosis accuracy of CNN training samples and test samples is 100%.

关 键 词：内燃机 连杆轴承 故障诊断 信号处理 

分 类 号：TK407[动力工程及工程热物理—动力机械及工程] TP206[自动化与计算机技术—检测技术与自动化装置]