计量学报2026，Vol.47Issue(5)：693-701,9.DOI:10.3969/j.issn.1000-1158.2026.05.08
基于音频信号工业机器人关节异常运行检测方法研究
Research on Detection Method for Abnormal Joint Operation of Industrial Robots Based on Audio Signals
摘要
Abstract
An industrial robot operational status monitoring system is designed.The system adopts a joint monitoring device to acquire audio signals generated by the operation of robot joints.To address the difficulty in abnormal feature analysis of audio signals,an SVMD_IBWO_MCKD method is proposed.First,decomposes the audio signal into multiple intrinsic mode functions(IMFs)using the sequential variational mode decomposition(SVMD)method,and then screens out the optimal IMF through the Gaussian weighted kurtosis index.Secondly,the improve beluga whale optimization(IBWO)algorithm is utilized to adaptively select the parameters T,M and L of maximum correlation kurtosis deconvolution(MCKD),and perform MCKD processing on the selected optimal IMF.Finally,fault features in the robot joint audio signals are extracted through envelope spectrum analysis.The experimental results show that neither the BWO_MCKD nor the WOA_MCKD method can extract the effective octave,and the SSA_MCKD method can only extract the 3-octave component.In contrast,the SVMD_IBWO_MCKD method can effectively extract the fourfold frequency of the periodic fault frequency in the joint audio signal.关键词
力学计量/关节异常检测/工业机器人测试/音频信号/逐次变分模态分解/最大相关峭度解卷积/SVMD_IBWO_MCKD方法/改进白鲸优化算法Key words
mechanics metrology/joint abnormality detection/industrial robot testing/audio signal/SVMD/MCKD/SVMD_IBWO_MCKD method/IBWO algorithm分类
通用工业技术引用本文复制引用
奚明,乔贵方,徐思敏,乔子烁,刘娣..基于音频信号工业机器人关节异常运行检测方法研究[J].计量学报,2026,47(5):693-701,9.基金项目
国家自然科学基金(51905258) （51905258）
江苏省研究生科研与实践创新计划项目(SJCX25_1268) （SJCX25_1268）
