【摘要】 针对永磁同步电机(PMSM)噪声源分离识别问题，应用一种基于变分模态分解(VMD)与鲁棒性独立分量分析(RobustICA)结合时频分析的方法。首先，采用VMD把永磁同步电机噪声信号分解为多种变分模态分量；然后，通过RobustICA提取主要信号的独立成分。最后，结合时频分析结果，对独立成分结果进行分析识别。该组合方法可以有效处理集成经验模态分解(EEMD)中存在的模态混叠问题，同时能对测试过程中传感器数量多于或等于噪声源的分离问题进行有效缓解。结果表明，提取的主要独立分量对应于PMSM产生的电磁噪声和机械噪声，采用该方法可以分离识别PMSM噪声中的电磁噪声和机械噪声。通过对PMSM不同噪声源的有效分离和准确识别，可以为降噪、运行状态监测和故障诊断提供可靠依据。更多还原

【Abstract】 Aiming at the problem of noise source separation and identification of permanent magnet synchronous motor(PMSM), a method combining variational mode decomposition(VMD)-robust independent component analysis(RobustICA) and time-frequency analysis is applied. First, the PMSM noise signal is decomposed into multiple variational modal components with VMD. Then, the independent components of the main signal are extracted according to RobustICA. Finally, the results are identified by combining the time-frequency analysis results. The combined method can effectively deal with the modal aliasing problem in the integrated empirical mode decomposition(EEMD), and effectively alleviate the separation problem when the number of sensors is more than or equal to the noise source during the test. The results show that the extracted main independent components correspond to the electromagnetic noise and mechanical noise generated by PMSM, and the method can accurately separate and identify the electromagnetic noise and mechanical noise of PMSM. Through the effective separation and accurate identification of different noise sources of PMSM, it can provide a reliable basis for noise reduction, running status monitoring and fault diagnosis.更多还原