基于阵列波束形成技术的压缩机噪声测量与控制

发布时间：2018-08-06 10:12

【摘要】：在非接触式机械故障诊断及室内噪声控制领域，空间声场成像意义重大。而基于麦克风阵列的波束形成方法是当前较为稳定可靠的噪声源识别成像技术，其主要依据阵列面上不同位置的麦克风接收到的声波信号相位差，人为加以特定扫描补偿，以达到对被测目标声源空间方位以及能量强度进行重建识别的目的。依照生成的空间场强云图进而对机械设备特性加以分析。文章以分布式阵源测量技术引出波束形成声源成像方法，对波束在平面波与球面波两种声场背景下的聚焦形成原理进行论述，通过数学推演与仿真对比，对制约成像分辨率以及截止频率的空间采样率，阵元间隔，阵元数目，阵列张角以及测量距离等参数进行研究，得出了优良识别效果所应具备的优化阵元参数。在一维阵列的基础上设计了八组平面麦克风阵列，应用不同的评价指标，证明了非规则旋转车轮阵的成像优势所在。为解决频带受限以及分辨率不足所引起的虚像问题，又分析设计出三维立体空间阵，经对比验证，该阵列性能表现稳定可靠，在同等麦克风数目下，进一步改善了近场声源成像效果。为更好的应用于实际噪声控制工程中，对基于波束形成互功率谱改进算法进行了研究，实验验证，该算法不但能够聚焦主瓣，削减旁瓣，还有突出的滤除工业背景噪声干扰的能力。为满足声学诊断的要求，采用波束频域变化算法，在复杂多声源情况下，对不同频率，不同幅值的声源又进行了分离提取。依托中石油吉林石化公司乙二醇压缩机房的噪声治理项目，应用波束形成方法，在以往相位补偿的基础上又加入幅值补偿校准处理，，经与手持式声级计对比测试，证明了该校准处理的真实可行性。通过对车间噪声源进行分析识别，采用对压缩机房各个位置安装隔声罩，阻尼隔声板，声屏障，空间吸声体的方案，达到降噪治理的预期目标。
[Abstract]:In the field of non-contact mechanical fault diagnosis and indoor noise control, spatial acoustic field imaging is of great significance. However, the beamforming method based on microphone array is a stable and reliable method for identifying and imaging noise sources. It is mainly based on the phase difference of acoustic signals received by microphone at different positions on the array surface and is compensated by special scanning. In order to achieve the goal of reconstruction and identification of the target sound source spatial orientation and energy intensity. According to the generated spatial field intensity cloud map, the characteristics of mechanical equipment are analyzed. In this paper, the beamforming sound source imaging method is introduced by using distributed array source measurement technology. The focusing principle of beam-forming under the background of plane wave and spherical wave is discussed, and the mathematical deduction and simulation are compared. The parameters such as spatial sampling rate, array spacing, array number, array angle and measurement distance, which restrict the imaging resolution and cut-off frequency, are studied, and the optimized array element parameters are obtained for the excellent recognition effect. Eight sets of planar microphone arrays are designed on the basis of one-dimensional array. Different evaluation indexes are used to prove the advantage of irregular rotating wheel array. In order to solve the problem of virtual image caused by limited frequency band and insufficient resolution, the three-dimensional spatial array is also analyzed and designed. The performance of the array is proved to be stable and reliable under the same number of microphones. The near field sound source imaging effect is further improved. In order to be better applied to the actual noise control engineering, the improved cross-power spectrum algorithm based on beamforming is studied. The experimental results show that the algorithm can not only focus on the main lobe, but also reduce the sidelobe. There is also a prominent ability to filter industrial background noise interference. In order to meet the requirements of acoustic diagnosis, the beam-beam frequency-domain variation algorithm is used to separate and extract the sound sources with different frequencies and amplitudes in the case of complex multi-sound sources. Relying on the noise control project of ethylene glycol compressor room of PetroChina Jilin Petrochemical Company, applying the beamforming method and adding amplitude compensation calibration processing on the basis of previous phase compensation, the noise control project has been compared with the hand-held sound level meter. The feasibility of the calibration process is proved. Through the analysis and identification of the noise sources in the workshop, the scheme of installing the sound isolating cover, damping the sound insulation board, the sound barrier and the space sound absorber to each position of the compressor room is adopted to achieve the expected goal of noise reduction and control.
【学位授予单位】：沈阳工业大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TB535

【参考文献】