| 周中华*,刘祖斌* **.基于参数优化变分模态分解和马田系统的工业缝纫机故障诊断方法浙[J].高技术通讯(中文),2025,35(1):73~84 |
| 基于参数优化变分模态分解和马田系统的工业缝纫机故障诊断方法浙 |
| Fault diagnosis method of industrial sewing machines based on parameter optimization VMD and MTS |
| |
| DOI：10. 3772 / j. issn. 1002-0470. 2025. 01. 008 |
| 中文关键词: 工业缝纫机； 故障诊断； 变分模态分解； 马田系统； 多域特征融合 |
| 英文关键词: industrial sewing machine, fault diagnosis, variational mode decomposition, Mahalanobis Taguchi system, multi-domain feature fusion |
| 基金项目: |
| 作者 | 单位 | | 周中华* | (*浙江工业大学特种装备和先进加工技术重点实验室杭州 310023)
(**浙江工业大学台州研究院台州 318014) | | 刘祖斌* ** | |
|
| 摘要点击次数: 171 |
| 全文下载次数: 248 |
| 中文摘要: |
| 针对工业缝纫机出厂质检的人耳听音传统方式准确率不高、耗时耗力的问题，提出了一种基于参数优化变分模态分解（variational mode decomposition，VMD）和马田系统(Mahalanobis-Taguchi system，MTS)的工业缝纫机故障诊断方法。首先，通过樽海鞘群算法(salp swarm algorithm，SSA)对变分模态分解的相关参数进行迭代寻优，并利用获得最优参数的VMD对工业缝纫机声信号进行分解得到不同中心频率的固有模态函数(intrinsic mode function，IMF)；然后，分别对IMF分量进行多域特征融合，并且采用正常样本构建了MTS的基准空间，进一步利用了少量故障样本来验证和优化基准空间；最后，结合马氏距离的阈值实现了准确的故障识别分类。通过仿真信号的对比分析，证明了SSA-VMD算法分解信号的可行性和优越性；实验数据和实测数据的研究结果表明了所提出的故障诊断方法具有一定的实际应用价值。 |
| 英文摘要: |
| Aiming at the problems that low accuracy and time-consuming in the traditional way of human ear listening in the factory quality inspection of industrial sewing machine, a fault diagnosis method of industrial sewing machine based on parameter optimization variational mode decomposition (VMD) and Mahalanobis-Taguchi system (MTS) is proposed. First, the relevant parameters of the variational mode decomposition are iteratively optimized by the salp swarm algorithm (SSA), and the VMD with the optimal parameters is used to decompose the sound signal of industrial sewing machines, so as to obtain the intrinsic mode function (IMF)with different central frequency. Then, the multi domain feature fusion of IMF components is performed separately, and a reference space of MTS is constructed with normal samples, and a small number of fault samples are used to verify and optimize the reference space. Finally, combined with the threshold of Mahalanobis distance, the accurate fault identification and classification is achieved. Through the comparative analysis of simulation signals, it is proved that the SSA-VMD algorithm is feasible and superior in decomposing signals. The research results of experimental data and measured data show that the proposed fault diagnosis method has certain practical application value. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
