Method for improving sound source positioning performance for large civil aircraft acoustic wind tunnel test

Abstract

The application discloses a sound source positioning performance improvement method for large civil aircraft acoustic wind tunnel test, and belongs to the technical field of aviation wind tunnel test. After microphone array signal acquisition, block filtering and noise reduction preprocessing, a sound source dirty map is generated by using a beam forming algorithm, CLEAN-SC technology is combined to iteratively eliminate sidelobe interference and reconstruct a high-resolution clean map, and a CPU and GPU cooperative computing architecture is used, through cross-spectrum matrix optimization, mixed precision floating point operation and matrix vectorization acceleration strategy, the calculation efficiency of sound source positioning is greatly improved under the premise of ensuring sound imaging accuracy, and the problems of large consumption of calculation resources and slow execution speed of the traditional method in large-scale test are effectively solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of image generation, in particular to a sound source positioning performance improvement method for large civil aircraft acoustic wind tunnel tests. BACKGROUND

[0002] As the most direct noise testing means, the aeroacoustic wind tunnel test can intuitively reflect the aeroacoustic performance of various equipment and guide the noise reduction design direction. In the past few decades, various countries led by the United Kingdom and the United States have developed various acoustic testing technologies, such as sound source positioning and pipe modal identification, which are applied to aeroengine noise, wing noise, rotor noise and other aeroacoustic noise testing, aiming to improve the acoustic performance of various aircraft and guide the design and optimization direction of the next generation of products. Acoustic imaging is an important tool for identifying the position, amplitude and shape of the sound source, and is often used with near-field acoustic holography and far-field beamforming technology.

[0003] Beamforming is a non-contact noise source measurement technology based on microphone array sound pressure signals, which is widely used in the fields of aviation, aerospace, navigation and rail transportation, and can effectively identify the noise source of aircraft fuselage and aeroengine and other key components, providing visual reference for noise control and noise reduction. The delay-and-sum beamforming method is the most classic time-domain acoustic imaging method. Because the spatial positions of the microphones in the array are different, the propagation time of the sound waves emitted by the same sound source to different microphones is different. This causes the beam after delay and summation operation to present directivity, thereby achieving the purpose of positioning the sound source. In order to reduce the error caused by time-domain signal translation, the time-domain method is often converted to the frequency domain, and the phase difference is used to represent the delay time, and the background noise is eliminated by removing the autocorrelation spectrum on the diagonal line of the cross-spectrum matrix. In order to identify moving sound sources, time-frequency domain beamforming methods have been further derived.

[0004] Traditional beamforming methods can better identify medium and high frequency sound sources, but it is difficult to distinguish low frequency sound sources. This has given rise to deconvolution beamforming methods and many similar derivative methods. Functionally, deconvolution methods aim to extract the local sound intensity maximum value; from the principle, they aim to identify the point spread function in the acoustic imaging results of traditional beamforming methods and gradually replace it with a single point or narrow beam. The DAMAS method is based on non-negative constraints and uses the Gauss-Seidel iteration method to solve the linear equation inverse problem related to the point spread function. In practice, sound sources usually have spatial characteristics, rather than being concentrated at a point. In this regard, Pieter Sijtsma proposed the CLEAN method based on spatially coherent sound sources, referred to as CLEAN-SC. The core idea of CLEAN-SC is to abandon the point spread function and extract the sound source from the traditional beamforming results by analyzing the spatial coherence.

[0005] The CLEAN-SC method is known for its stability and intuitive results, and has been verified by many units of wind tunnel tests worldwide. It is considered the gold standard in the industry. The widespread application of this method in acoustic imaging has made it one of the preferred acoustic measurement methods in the industry.

[0006] However, during the operation of the CLEAN-SC program, it is necessary to continuously iterate to extract the local sound intensity maximum value, resulting in high cost of sweep calculation (number of frequencies > 1000), so it is urgent to accelerate the method to meet the actual engineering needs. Through algorithm acceleration, the calculation time and resources can be saved, the data analysis, processing and decision-making process can be accelerated, and the efficiency of the wind tunnel can be improved, thereby indirectly promoting the development of China's noise control industry. SUMMARY

[0007] In order to solve the technical problems of large consumption of computing resources and low execution efficiency of the sound source positioning method in traditional large aircraft acoustic wind tunnel test, the present application provides a sound source positioning performance improvement method for large civil aircraft acoustic wind tunnel test, comprising the following steps:

[0008] S1. Receive the sound pressure signal through the microphone array, and block and store the received signal;

[0009] S2. Batch processing of the blocked sound pressure signal;

[0010] S3. Fourier transform of the processed sound pressure signal, converting the processed sound wave signal from time domain to frequency domain;

[0011] S4. According to the beamforming algorithm, the Fourier transformed microphone sound pressure signal is weighted and phase controlled, the sound power and cross-spectrum matrix are calculated, and the image on the scanning plane is obtained, which is called dirty map;

[0012] S5. Suppressing the sidelobes in the acoustic imaging result and improving the image resolution, finding the sound power maximum value and its position in the scanning plane from the dirty map;

[0013] S6. Based on the sound power maximum value change the steering vector of the cross-spectrum matrix to obtain the spatial correlation of the maximum point and the remaining points;

[0014] S7. Remove the influence of the maximum point from the cross-spectrum matrix to obtain a new cross-spectrum matrix, and the corresponding dirty map and clean map;

[0015] S8. After i iterations or , the final acoustic imaging is the superposition of the clean beam and the remaining sound sources;

[0016] S9. Defining the cross-spectrum matrix, the clean and dirty map variables as GPU variables, using a graphic processor to process; while the rest of the variables and related operations are processed by a general-purpose CPU;

[0017] S10. Accelerating the processing of S1-S8 by vectorization or matrix operation, converting the loop process in S1-S8 into matrix operation;

[0018] S11. Accelerating the processing of S1-S8 by adjusting the storage precision of floating point numbers.

[0019] Further, the sound power A is obtained by:

[0020]

[0021] wherein, is a directivity vector, C is a cross-spectrum matrix, and the superscript * represents a conjugate transpose;

[0022] The directivity vector is obtained by:

[0023]

[0024] wherein, is a circular frequency, is the distance from the microphone to the discrete points on the scanning plane, and the symbol || || represents the two-norm, is a three-dimensional Green's function in free space;

[0025] The three-dimensional Green's function in free space is obtained by:

[0026]

[0027] wherein i represents a complex number, and c is the speed of sound;

[0028] Further, the cross-spectrum matrix is obtained by:

[0029]

[0030] wherein the elements of the lower triangle of the cross-spectrum matrix are equal to the complex conjugate of the corresponding matrix elements in the upper triangle;

[0031] A single element of the cross-spectrum matrix is obtained by:

[0032]

[0033] wherein m and n are microphone index numbers, S is an effective subset of microphones, is a microphone sound pressure signal after Fourier transform.

[0034] Further, considering The steering vector of the cross-spectrum matrix By:

[0035]

[0036] obtained.

[0037] Further, in S7, the new cross-spectrum matrix is obtained by:

[0038]

[0039] obtained, wherein, is a safety factor, , is the outer product of the steering vector and its conjugate transpose.

[0040] The application also provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program stored in the memory, the processor executes the method for improving the sound source positioning performance for the acoustic wind tunnel test of large civil aircraft.

[0041] The application also provides a computer readable storage medium, wherein the computer readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for improving the sound source positioning performance for the acoustic wind tunnel test of large civil aircraft is realized.

[0042] The application has the following advantages: based on the existing acoustic imaging method based on traditional beamforming, the application discards the point spread function, suppresses the sidelobes in the acoustic imaging result and improves the image resolution by analyzing the spatial coherence, maintains the algorithm accuracy, and significantly improves the execution efficiency of the acoustic imaging subject of the acoustic wind tunnel test of large civil aircraft. BRIEF DESCRIPTION OF DRAWINGS

[0043] Figure 1 It is a flowchart of the method for improving the sound source positioning performance for the acoustic wind tunnel test of large civil aircraft.

[0044] Figure 2 It is a schematic diagram of the microphone array sensor arrangement of the acoustic wind tunnel test of large civil aircraft landing gear.

[0045] Figure 3 It is a "dirty map" calculated by the beamforming algorithm.

[0046] Figure 4 It is a "clean map" calculated by the CLEAN-SC algorithm. DETAILED DESCRIPTION

[0047] In order to make the technical solutions and advantages of the embodiments of the present application clearer, the exemplary embodiments of the present application are further described in detail below with reference to the drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. It should be noted that the embodiments in the present application and the features in the embodiments can be combined with each other without conflict.

[0048] Embodiment 1, in combination Figure 1 In this embodiment, the present application provides a method for improving the sound source positioning performance for large civil aircraft acoustic wind tunnel test, including the following steps:

[0049] S1. receiving sound pressure signals through a microphone array, and blocking and storing the received signals; the wind speed of the large civil aircraft landing gear acoustic wind tunnel test is 68 m / s, the temperature is 25°, the humidity is 50%, and the atmospheric pressure is 100.43 kPa. The sensor position coordinates of the microphone array are as shown in Figure 2 , the total number of microphones is 143, the center point coordinates are (0, 0, 0) m, the sampling frequency is 81920 Hz, the sampling time is 10 s, and it is divided into 200 blocks.

[0050] S2. batch processing the blocked sound pressure signals; the batch processing includes filtering, removing background noise and other operations to improve signal quality.

[0051] S3. Fourier transforming the processed sound pressure signals to convert the processed sound wave signals from time domain to frequency domain;

[0052] S4. according to the beamforming algorithm, weighting and phase control of the Fourier transformed microphone sound pressure signals, calculating the sound power and cross-spectral matrix, obtaining the image on the scanning plane, which is called dirty map; the calculation result is as shown in Figure 3 . The scanning plane size is 4 m x 4 m, the resolution is 100 x 100, and the center coordinates are (0, 0.5, 8.24) m.

[0053] S5. suppressing the sidelobes in the sound imaging result and improving the image resolution, finding the sound power maximum value and its position in the scanning plane from the dirty map;

[0054] S6. based on the sound power maximum value changing the steering vector of the cross-spectral matrix to obtain the spatial correlation of the maximum point and the remaining points;

[0055] S7. removing the influence of the maximum point from the cross-spectral matrix to obtain a new cross-spectral matrix, and the corresponding dirty map and clean map;

[0056] S8. after i iterations or Finally, the resulting acoustic image is the superposition of the clean beams and the remaining sources; the results of S8 are shown in FIG. 8. Figure 4

[0057] S9. The cross-spectral matrix, the clean and dirty variables are defined as GPU variables, and are processed using a graphics processor; the remaining variables and related operations are processed using a general-purpose CPU; the central processing unit is suitable for processing complex logical operations and sequential tasks, and has strong general computing ability and flexibility; the graphics processor is suitable for processing large-scale data and highly parallelized tasks.

[0058] S10. The steps S1-S8 are accelerated by vectorization or matrix operations, and the loop process in S1-S8 is converted into matrix operations.

[0059] S11. The steps S1-S8 are accelerated by adjusting the storage precision of floating-point numbers. For acoustic imaging, the eighth digit after the decimal point has little effect on the final sound pressure level, and can be considered to be much less than 1%. Therefore, the default double-precision data is reduced to single-precision data to further improve performance.

[0060] Specifically, after the acceleration processing of S9-S11, the comparison between the basic algorithm and the accelerated algorithm is shown in the following table:

[0061] Table 1 Comparison of basic algorithm and accelerated algorithm

[0062] Parallel computation time consumption (CPU) Parallel computation time consumption (GPU) Base algorithm 1557.27s NAN Accelerated algorithm 308s 232s Acceleration effect 80.2% 85.2%

[0063] As can be seen from the comparison table, the accelerated algorithm significantly reduces the sound imaging calculation time from 1557.27 seconds of the basic algorithm to 308 seconds of the CPU (80.2% acceleration) and 232 seconds of the GPU (85.2% acceleration), proving that hardware adaptation and algorithm optimization significantly improve efficiency.

[0064] Further, the sound power A is obtained by:

[0065]

[0066] wherein, is a directional vector, C is a cross-spectral matrix, and the superscript * represents a conjugate transpose;

[0067] The directional vector is obtained by:

[0068]

[0069] wherein, is a circular frequency, ​is the distance from the microphone to the discrete point on the scanning plane, and || || denotes the two-norm, is the three-dimensional Green's function in free space;

[0070] The three-dimensional Green's function in free space is obtained by:

[0071]

[0072] where i denotes a complex number, and c is the sound speed;

[0073] Further, the cross-spectral matrix is obtained by:

[0074]

[0075] where the lower triangular elements of the cross-spectral matrix are equal to the complex conjugate of the corresponding matrix elements in the upper triangular;

[0076] The individual elements of the cross-spectral matrix are obtained by:

[0077]

[0078] where m and n are microphone index numbers, and S is an effective subset of microphones, is the Fourier-transformed microphone pressure signal.

[0079] Further, considering the steering vector of the cross-spectral matrix is obtained by:

[0080]

[0081]

[0082] Further, in S7, the new cross-spectral matrix is obtained by:

[0083]

[0084] wherein is a safety factor, , is the outer product of the steering vector and its conjugate transpose.

[0085] Embodiment 2: The computer device of the present application can be a device comprising a processor and a memory, such as a single-chip microcomputer comprising a central processing unit. The processor is used to execute the computer program stored in the memory to realize the steps of the method for improving the performance of sound source positioning for large civil aircraft acoustic wind tunnel tests.

[0086] Embodiment 3: Computer-readable storage medium embodiment.

[0087] ​The computer readable storage medium of the present application can be any form of storage medium readable by the processor of the computer device, including but not limited to non-volatile memory, volatile memory, ferroelectric memory, etc., and the computer readable storage medium stores a computer program, when the processor of the computer device reads and executes the computer program stored in the memory, the steps of the above-mentioned sound source positioning performance improvement method for large civil aircraft acoustic wind tunnel test can be realized.

[0088] The above merely provides one specific embodiment of the present application, but the protection scope of the present application is not limited thereto, any simple replacement or change within the technical concept disclosed by the present application and according to the technical solution of the present application should be within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for improving the positioning performance of sound sources for acoustic wind tunnel testing of large civil aircraft, characterized in that, The method comprises the following steps: S1. receiving sound pressure signals through a microphone array, and blocking and storing the received signals; S2. batch processing the blocked sound pressure signals; S3. performing Fourier transform on the processed sound pressure signals to convert the processed sound wave signals from time domain to frequency domain; S4. according to a beamforming algorithm, weighting and phase regulating the Fourier-transformed microphone sound pressure signals, calculating sound power and cross-spectrum matrix, and obtaining an image on a scanning plane, which is called a dirty map; S5. Suppression of side lobes in the acoustic imaging result, increase of image resolution, finding of the acoustic power maximum from the dirty map and its position in the scan plane; S6. Based on the sound power maximum Changing the steering vector of the cross-spectral matrix to obtain the spatial correlation of the maximum point with the remaining points; S7. removing the influence of the maximum point from the cross-spectrum matrix to obtain a new cross-spectrum matrix, and corresponding dirty and clean maps; S8. After i iterations or where represents the cross-spectral matrix, and the final acoustic image is the sum of the clean beams and the residual sources; S9. defining the variables of the cross-spectrum matrix, the clean and dirty maps as GPU variables, and processing them using a graphics processor; while the remaining variables and related operations are processed using a CPU; S10. accelerating the processing of steps S1-S8 through vectorization or matrix operation, and converting the loop process in S1-S8 into matrix operation; S11. accelerating the processing of steps S1-S8 by adjusting the storage precision of floating-point numbers, and further improving the performance by reducing the default double-precision data to single-precision data and customizing the mixed precision.

2. The method for improving the positioning performance of sound sources for large civil aircraft acoustic wind tunnel tests according to claim 1, characterized in that, The sound power A is calculated by: obtained, wherein is the directional vector, C is the cross-spectral matrix, and the superscript * denotes the conjugate transpose; The directivity vector is calculated by: get, wherein is the circular frequency, is the distance of the microphone to the discrete points on the scan plane, the sign denotes the two-norm, is the three-dimensional Green's function in free space; The three-dimensional Green's function in free space is calculated by: obtained, where i represents a complex number, and c is the sound speed.

3. The method for improving the positioning performance of sound sources for large civil aircraft acoustic wind tunnel tests according to claim 2, characterized in that, The cross-spectrum matrix is calculated by: obtained, where the lower triangular elements of the cross-spectrum matrix are equal to the complex conjugate of the corresponding matrix elements in the upper triangular matrix; A single element of the cross-spectrum matrix is calculated by: obtained, m, n are microphone index numbers, S is the effective subset of microphones, is the Fourier transformed microphone sound pressure signal.

4. The method for improving the positioning performance of sound sources for large civil aircraft acoustic wind tunnel tests according to claim 3, characterized in that, Consider the steering vector of the cross-spectral matrix by: obtained.

5. The method for improving the positioning performance of sound sources for large civil aircraft acoustic wind tunnel tests according to claim 4, characterized in that, In S7, the new cross-spectrum matrix is calculated by: obtained, wherein is a safety factor, , is the outer product of the pointing vector with its conjugate transpose.

6. Computer device comprising a memory and a processor, characterized in that The memory stores a computer program, and when the processor executes the computer program stored in the memory, the processor executes the sound source positioning performance improvement method for large civil aircraft acoustic wind tunnel test according to any one of claims 1-5.

7. A computer readable storage medium characterized by, The computer readable storage medium stores a computer program, and when the processor executes the computer program, the sound source positioning performance improvement method for large civil aircraft acoustic wind tunnel test according to any one of claims 1-5 is realized.

Images