【摘要】 非平稳声源识别与定位是旋转机械状态监测和故障诊断的重要环节。利用非平稳旋转机械的转速特征,将Vold-Knalman阶次追踪方法与传统波束形成方法结合,提出一种新的非平稳声场信号波束成形声源定位方法。首先对旋转机械转速进行追踪,结合时频分析方法用Vold-Kalman分解提取出非平稳声信号的特定阶次成分,然后利用分离出的特定阶次信号重建声场,从而定位声源。该方法利用对特征阶次声信号的提取,避免多个阶次声信号成像时的混叠现象,通过多重线性调频信号叠加声场声源识别仿真实验与电机升速实验进行声源识别验证,证明所述方法能够准确有效定位声源。这种方法在非平稳声源识别和噪声控制中有良好工程应用潜力。更多还原

【Abstract】 Identification and localization of non-stationary sound sources is an important issue in condition monitoring and fault diagnosis of rotary machines. In this article, a new non-stationary beamforming method is proposed by combining the Vold-Kalman order tracking with the traditional beamforming and using the speed characteristics of non-stationary rotating machinery. Firstly, the rotation speed was tracked and the Vold-Kalman method combining with time-frequency analysis was used to extract the specific order components of non-stationary sound signals. Then the extracted signals were applied to reconstruct the sound field to localize the sound source. This method uses the extraction of signals of specific orders to avoid the phenomenon of aliasing in the imaging of multiple infrasound signals. Meanwhile, it is verified that this method can localize the sound source accurately and effectively through the simulation experiment of multiple chirp signal superposition sound field and motor acceleration experiment. This method has the potential for engineering application in non-stationary source identification and noise control.更多还原