基于L型阵列的故障声源位置估计

发布时间：2018-06-10 18:16

本文选题：D-MUSIC算法 + 均匀线性阵　；参考：《电子测量与仪器学报》2017年04期

【摘要】：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。
[Abstract]:The traditional fault diagnosis method is mainly based on the contact vibration signal analysis. In some special environment, this method has great limitations, but the problem can be solved based on the sound signal during the operation of the equipment. A position estimation method based on L-type acoustic array is proposed for fault diagnosis. The commonly used DOA estimation based on 2-D music in radar is adopted, and the estimation performance of the algorithm is studied. The computer simulation results of 2D-DOA estimation under different array number, SNR and rapid-beat number are studied, and compared with the DOA estimation based on rectangular array. The results show that the estimation results of the algorithm will be very different under different parameters. Therefore, the algorithm can be applied to the gearbox fault diagnosis under the condition of selecting appropriate parameters.
【作者单位】：北京信息科技大学现代仪器与测控教育部重点实验室;
【基金】：国家自然科学基金(51275052) 现代测控技术教育部重点实验室开放课题(KF20161123202) 北京市重点实验室开放课题(KF20161123204)资助项目
【分类号】：TN911.7

【相似文献】