CN-122024744-A - Engine pipeline acoustic mode identification method based on cross spectrum matrix

CN122024744ACN 122024744 ACN122024744 ACN 122024744ACN-122024744-A

Abstract

The invention relates to an engine pipeline acoustic mode identification method based on a cross spectrum matrix, which comprises the steps of arranging M microphones on the circumferential pipe wall of a pipeline to be tested, arranging the M microphones at intervals in the axial direction of the pipeline to be tested to form a circumferential microphone array, carrying out pipeline acoustic mode identification measurement by adopting the M microphones, taking an obtained noise sound pressure signal as a time domain signal, obtaining a measurement cross spectrum matrix of the noise signal measured by the M microphones through the conversion from the time domain signal to a frequency domain signal, removing self-spectrum elements, correspondingly establishing a simulation model cross spectrum matrix with the self-spectrum elements removed, and enabling the mean square error between the measurement cross spectrum matrix and the simulation model cross spectrum matrix to be minimum, thus obtaining an engine pipeline acoustic mode identification result for inhibiting background noise. According to the invention, when the microphone is damaged or the signal is interfered, the damaged or received microphone signal is removed, and the engine pipeline acoustic mode can be still identified.

Inventors

LIAN JIANXIN
WANG HONGMEI
ZHANG LIANGJI
QIAO WEIYANG

Assignees

西北工业大学

Dates

Publication Date: 20260512
Application Date: 20260213

Claims (7)

1. An engine pipeline acoustic mode identification method based on a cross spectrum matrix is characterized by comprising the following steps of: firstly, arranging M microphones on the circumferential pipe wall of a pipeline to be tested, and arranging the M microphones at intervals in the axial direction of the pipeline to be tested to form a circumferential sound transmission array; wherein, M is more than or equal to 2, and the interval in the axial direction is equidistant or non-equidistant, at this moment, the maximum circumferential modal decomposition order K= (M-1)/2 of M microphones is rounded downwards; Carrying out pipeline acoustic mode recognition measurement by using M microphones, taking the obtained noise acoustic pressure signals as time domain signals, and combining time differences of the acoustic signals received by the M microphones to obtain a cross-correlation function of the noise acoustic pressure signals measured by each noise frequency of the M microphones and a cross-power spectrum function after Fourier transformation of any two microphone measurement signals, so as to obtain a measurement cross-spectrum matrix of the noise signals measured by the M microphones; the fourier transform is to transform the time domain signal into a frequency domain signal; Step two, establishing simulation models of M microphones under each frequency, wherein the sound field in the circular or annular hard wall pipeline is formed by linearly superposing waves of different modes, so that a cross spectrum matrix of the simulation models of the M microphones and the different modes is sequentially established based on the maximum circumferential mode identification order K; And thirdly, according to the simulated model cross spectrum matrixes of different modes, determining circumferential mode amplitude values of the M microphones, so that the mean square error between the measured cross spectrum matrix and the simulated model cross spectrum matrix is minimum, and obtaining an engine pipeline acoustic mode identification result, and removing self-spectrum elements of the measured cross spectrum matrix and the simulated model cross spectrum matrix in the calculation process, so that the engine pipeline acoustic mode identification result of background noise can be restrained.
2. The method for identifying the acoustic modes of the engine pipeline based on the cross spectrum matrix according to claim 1, wherein in the first step, the M microphones are fixed on the circumferential outer pipe walls of the inlet and outlet pipelines to be tested of the annular blade grid or the aeroengine test bed, so as to realize the identification and measurement of the acoustic modes of the pipeline to be tested in the inlet and outlet directions.
3. The method for identifying engine pipeline acoustic modes based on cross-spectrum matrix according to claim 1, wherein, In the first step, noise sound pressure signals measured by any two microphones of the M microphones are And And sound pressure signal of noise Time domain signal recorded as noise sound pressure of mth microphone, noise sound pressure signal The time domain signal recorded as noise sound pressure of the nth microphone is a cross correlation function of noise signals measured by M microphones Expressed as: , In the formula, Is the time difference of the sound signals received by the M microphones, t is the time of the sound signals received by the M microphones, M and n respectively represent the M-th and n-th microphones, and M represents the total number of the microphones; Further, noise sound pressure signals measured by the mth and nth microphones are obtained And Is a cross-power spectral function after Fourier transformation The method comprises the following steps: , In the formula, F represents the frequency of noise; obtaining a measurement cross spectrum matrix of noise signals measured by M microphones The method comprises the following steps: , where M represents the total number of microphones.
4. The method for identifying engine pipeline acoustic modes based on cross-spectrum matrix according to claim 1, wherein in the second step, the constructed M microphone analog model signals The method comprises the following steps: , Wherein k represents a mode order; Representing the amplitude of mode k; An analog model signal for the mth microphone; represents a steering vector, m represents an mth microphone, and θ m represents a circumferential angle of the mth microphone; Furthermore, the constructed simulation model cross spectrum matrix of different modes The method comprises the following steps: , Wherein, the Represents a conjugate transpose; And M and n respectively represent the mth microphone and the nth microphone, namely any two microphones in M microphones, and K is the maximum circumferential mode identification order.
5. The method for identifying engine pipeline acoustic modes based on cross spectrum matrix as claimed in claim 1, wherein the third step is as follows: Determining the values of the k-order circumferential modes B k of the M microphones so as to measure a cross-spectrum matrix Cross spectrum matrix of simulation model The mean square error F (B) between the two is the smallest: , Wherein k represents a mode order; a power spectrum representing a k-order modality; represents a conjugate transpose; And M and n respectively represent the M number and the n number of microphones, M represents the total number of microphones, K is the maximum circumferential mode identification order; Then, the following form is developed: , Wherein, the , , , Wherein m and n represent the mth and nth microphones, respectively; , , , all represent steering vectors; represents conjugate transpose, k and u represent modal orders, M represents total number of microphones; to minimize F (B), let: , I.e. solve the minimization problem: , wherein k and u represent mode orders, and B k represents a power spectrum of a k-order mode amplitude; Then, written as a matrix equation: , wherein: , , , And solving a matrix equation to obtain the value of each order circumferential mode B k .
6. The method for identifying engine pipeline acoustic modes based on cross-spectrum matrix according to claim 5, wherein self-spectrum elements in the measurement cross-spectrum matrix and the simulation model cross-spectrum matrix are removed first in order to reduce the influence of background noise And Namely, removing m=n elements in the minimization problem U u , and then performing matrix calculation and solving to achieve the purpose of suppressing background noise, so that the mode identification result is more accurate and the imaging is clearer.
7. The method for identifying engine pipeline acoustic modes based on cross-spectrum matrix according to any one of claims 1 to 6, wherein mode identification calculation is performed at each noise frequency of 0 to 10000 hz.

Description

Engine pipeline acoustic mode identification method based on cross spectrum matrix Technical Field The invention relates to the technical field of recognition of pipeline acoustic modes of an aeroengine. Background The unique inlet/outlet duct layout of an aero gas turbine jet engine is such that its noise always propagates through the air duct in a characteristic acoustic mode when radiating outwards. As is well known, noise generated by various components of an aero-engine, except that a jet noise source is exposed to the free space of the atmosphere, propagates in space according to spherical waves, and noise of other components of the engine always propagates to an engine inlet and a jet outlet through a circular pipeline or an annular pipeline of engine air inlet/exhaust, and then radiates outwards through an engine inlet and an engine outlet. It is known from the basic theory of pipe acoustics that in a circular (annular) airflow pipe, only a portion of a specific type of sound wave in the pipe propagates along the pipe due to the influence of the boundary conditions of the pipe wall, and this specific type of sound wave structure is generally referred to as acoustic mode. In a circular (annular) pipe, the acoustic mode of the pipe is characterized by a rotational motion (circumferential mode) that rotates about the pipe axis and a radial amplitude profile (radial mode) described by a Bessel function, and the propagation characteristics of the acoustic mode of the pipe are determined by the specific geometry, size, and flow field characteristics within the pipe. Because turbulent broadband noise propagates in the engine pipeline in a mode waveform form with statistical average, and the acoustic energy is distributed on each cut-off mode, the broadband noise pipeline acoustic mode needs to be described by adopting a statistical method for the research of broadband noise. As early as 1972, harel & Perulli et al proposed the use of a cross-spectrum method (Cross Spectral Approach) for identifying and measuring the acoustic modes of wideband noise pipes, which were the first time to replace the hot wire anemometer with a microphone in the acoustic pressure measurement. Bolleter et al (1973) have developed a study of the fan broadband noise pipeline acoustic modes using the Cross SPECTRAL DENSITY Approach, which decomposes broadband noise acoustic power into modes by measuring the Cross correlation function between different microphones, but during the study they assume that the pipeline end walls are not reflective and neglect the effect of the air flow in the pipeline. Kerschen et al (1981) developed a first decoupling method, transient (Instantaneous Approach), for acoustic modes of both monophonic and wideband noise pipelines, and analyzed the principle of modal decomposition and its characteristic of suppressing flow noise. In terms of the measurement work of the broadband noise acoustic power in the pipeline, based on the research work of Chung et al (1977), michalke et al (1996) and Chun et al (2003), enghardt (2004) proposes a fan broadband noise acoustic power measurement method based on the recognition of the pipeline acoustic mode, namely a reference microphone method. In recent years, research work on turbulent broadband noise pipeline acoustic modes of aeroengines gradually becomes a research hotspot and an important point in the aerodynamic acoustic field, and pipeline acoustic mode identification methods such as instantaneous pipeline acoustic mode identification (Instantaneous approach), cross-correlation pipeline acoustic mode identification (Cross Correlation-based applied) and reference microphone pipeline acoustic mode identification (REFERENCE SENSOR APPROACH) are developed successfully in the world. In summary, although a plurality of engine turbulence broadband noise pipeline acoustic mode identification methods have been developed at present, the following defects still exist at present (1) in actual test measurement, microphone damage is often encountered, or part of microphone signals are interfered, so that the traditional pipeline acoustic mode identification method cannot be used, and (2) the traditional pipeline acoustic mode identification method needs to use a Zhou Xiangmai gram wind array which is uniformly distributed, and in the actual measurement process, the reasons such as hardware shielding and the like can be met, so that the uniform arrangement of circumferential microphones cannot be met, and the use of the pipeline acoustic mode identification method is affected. (3) The conventional pipeline acoustic mode identification method cannot overcome the influence of background noise. Disclosure of Invention The invention aims to avoid the defects of the prior art, and provides a cross spectrum matrix-based engine pipeline acoustic mode identification method which can identify the engine pipeline acoustic mode when a microphone is damaged or a signal is interfered, rejec