一种循环平稳声场的声源识别定位方法

摘要

The radiated sound fields of rotating or reciprocating machinery are the cyclostationary sound fields ,but these sound fields are approximately regarded as stationary sound fields when power spectrum analysis is used in the traditional near‐field acoustic holography ,which ignores the periodical time‐variant property and makes it hard to reflect the radiation characteristics accurately based on the hologram results .Thus ,a method of cyclostationary signal analysis is utilized in pre‐processing holo‐graphic surface data .The cyclic spectral density (CSD) is adopted as one of the holographic recon‐struction parameters .The characteristic frequency of the signal is extracted by the gathering slices method of CSD .The simulation and experimental results show that the proposed method enables to accurately analyze the cyclic properties of cyclostationary signals and the sources of cyclostationary sound fields can be exactly identified and localized .%传统的近场声全息技术在全息面声场预处理时，采用功率谱分析将声场简化为稳定的平稳声场，而旋转或往复机械的辐射声场为循环平稳声场，采用这种简化方法得到的分析结果并不能准确反映声场的特性。鉴此，对此类声场采用循环平稳信号处理方式进行了预处理，并选择循环谱密度作为重要参数进行全息重建，然后通过循环谱密度组合切片的分析方法提取信号特征频率。仿真及实验结果表明：提出的方法可准确分析循环平稳信号的周期特性，并对循环平稳声源进行准确的识别定位。