基于L型阵列的故障声源位置估计

Fault sound sources position estimation based on L array method

传统的故障诊断方法主要是基于接触式的振动信号分析,在某些特殊环境下该方法具有很大的局限性,而基于设备运转过程中的声音信号就可以解决该问题。提出了基于L型声阵列的位置估计方法来进行故障诊断,采取了雷达中常用的基于二维MUSIC的DOA估计,并研究了算法的估计性能情况。分别研究了在不同的阵元个数、信噪比和快拍数下的2D-DOA估计的计算机仿真结果,并和基于矩形阵列的DOA估计进行了比较。研究结果表明,在不同的参数下该算法的估计结果也会有很大的不同。所以在选择合适的参数条件下,该算法可以应用于齿轮箱的故障诊断中。

The traditional method of fault diagnosis is mainly based on the analysis of the vibration, whichhas great limitations in some special circumstances, but the method based on the sound signal from the equipment in the process can solve the problem. A new method for fault diagnosis based on L type acoustic array is presented in this paper. The DOA estimation based on two-dimensional MUSIC is used in radar, and the performance of the algorithm is studied. And with three kinds of different parameterson 2DDOA estimation, the computation is carried out. The comparison with the result of the rectangular arraysis carried out. The resultsshowthat the corresponding algorithm performance is changed while the different parameters have different values. And the method with suitable parameters could be better for the fault diagnosis of the gear drive system.