近场子空间聚焦的碰摩故障声发射定位方法

发布时间：2018-06-24 22:48

本文选题：碰摩故障 + 近场　；参考：《声学学报》2017年06期

【摘要】：针对宽带多源声发射信号的相干、多模态和能量衰减快问题,提出一种近场多重相干信号子空间聚焦的定位算法用于碰摩故障声发射源的定位检测。首先,为滤除干扰模态波、减小频散效应,采用基于模态声发射传播特性分析的小波分解滤波方法,从碰摩初期的声发射信号中获取零阶模态波及波速用于定位计算;其次,为实现信号解相干,提出基于双边相关变换(TCT)的近场聚焦矩阵估计方法;最后,针对声发射信号的能量衰减快问题,利用近场基于特征分解的多重信号分类(N-MUSIC)的空间谱估计方法来实现声源的精确定位。理论分析和实验结果表明:该方法定位精度高、计算复杂度低、稳定性强,能有效识别多个相干碰摩声源。相比传统相干子空间算法(CSM),该方法减少了信号初值和聚焦频点的计算量,对双声源的分辨概率较现有修正近场多重信号分类算法提高了17%,是一种有效的碰摩故障源检测方法。
[Abstract]:Aiming at the problems of fast coherence, multi-mode and fast energy attenuation of wideband multi-source acoustic emission signals, a localization algorithm based on subspace focusing of near-field multi-coherent signals is proposed to detect the location of acoustic emitters from rub-impact faults. First of all, in order to filter the interference mode wave and reduce the dispersion effect, wavelet decomposition filtering method based on modal acoustic emission propagation characteristic analysis is used to obtain zero order mode wave velocity from the acoustic emission signal in the initial stage of rub-impact for location calculation. In order to realize signal decoherence, a near field focusing matrix estimation method based on bilateral correlation transform (TCT) is proposed. The spatial spectrum estimation method based on eigen-decomposition (N-MUSIC) in near field is used to realize the accurate location of sound source. The theoretical analysis and experimental results show that the proposed method has the advantages of high positioning accuracy, low computational complexity and strong stability, and it can effectively identify multiple coherent rub-impact sound sources. Compared with the traditional coherent subspace algorithm (CSM), this method reduces the computation of the initial signal and the focusing frequency, and improves the resolution probability of the dual sound source by 17 times compared with the existing modified near field multi-multiple signal classification algorithm. This method is an effective method for rub-impact fault source detection.
【作者单位】：东南大学信息科学与工程学院;火电机组振动国家工程研究中心(东南大学);中国矿业大学资源与地球科学学院;
【基金】：国家自然科学基金项目(61673108,61375028) 国家工程中心创新能力建设项目(6750100000)资助
【分类号】：TH17;TN911.7

【相似文献】