Multi-arm spiral subarray cooperative adaptive acoustic imaging system and method

Abstract

The invention provides a multi-arm spiral subarray cooperative adaptive acoustic imaging system and method. The system adopts a circular substrate to construct an Archimedes spiral array. The innovative dynamic processing strategy comprises low-frequency sub-array combination and aperture expansion, intermediate-frequency independent processing and load balance maintenance, and high-frequency cross-arm virtual focusing array construction. The key technology covers multi-channel dynamic sampling, frequency band conflict arbitration, phase compensation beam forming and confidence weighted fusion. In the data fusion stage, a blind area is eliminated by adopting Delaunay triangulation and radial basis interpolation, and edge features are reserved in combination with anisotropic diffusion filtering. A temperature-sound velocity correction model is integrated in the system, and delay summation, MVDR and compressed sensing hybrid beam forming algorithms are applied to real-time processing. The output unit generates a three-dimensional acoustic thermodynamic diagram fusing Gaussian smoothing and edge enhancement. According to the scheme, the spatial sampling efficiency is optimized through spiral topology, the resolution and precision in a complex scene are improved through a dynamic recombination mechanism, and the method is suitable for industrial monitoring and noise source positioning.

Description

Technical Field

[0001] The present invention relates to the field of acoustic imaging technology, specifically to a multi-arm spiral subarray collaborative adaptive acoustic imaging system and method. This system optimizes sensor layout through Archimedean spiral array topology and combines it with a dynamic frequency band arbitration mechanism to achieve multimodal processing of acoustic signals. This encompasses low-frequency subarray aperture expansion, mid-frequency conflict resolution, and high-frequency virtual focusing array construction. Background Art

[0002] As a non-contact, non-destructive detection method, acoustic imaging technology plays an important role in industrial inspection, biomedicine, geological exploration, and other fields. However, with the increasing complexity of application scenarios and the increasing demand for imaging accuracy, traditional acoustic imaging systems have gradually exposed numerous technical bottlenecks. In particular, under complex working conditions, the physical structure limitations of traditional uniform arrays, the signal attenuation problems of non-uniform arrays, and the real-time challenges of multi-subarray collaborative processing have seriously restricted the further development of acoustic imaging technology.

[0003] Internationally, as early as the 1990s, Brüel & Companies such as Head Acoustics have launched acoustic cameras based on large-scale microphone arrays and beamforming algorithms. These products enable real-time monitoring and analysis of complex sound fields in industrial and architectural settings. Head Acoustics' products also utilize digital signal processing technology to achieve real-time imaging and noise source location for automotive and industrial noise. NTi Audio's high-precision acoustic measurement instruments have also found success in environmental noise detection. While these technical solutions have improved the performance of acoustic imaging systems to a certain extent, they still face technical bottlenecks when dealing with broadband noise, dynamic environments, and interference from multiple sound sources.

[0004] Domestically, with the development of intelligent sensing technology and signal processing algorithms, universities and research institutions such as Tsinghua University and Zhejiang University have conducted in-depth research in noise source localization, sound field reconstruction, and multimodal data fusion. Some of these findings have been applied to industrial monitoring and equipment fault diagnosis. However, existing research has largely focused on fixed array applications, making it difficult to balance the processing requirements of different frequency bands and robustness in complex scenarios.

[0005] To address the above-mentioned issues, the present invention proposes a multi-arm spiral subarray collaborative adaptive acoustic imaging system and method. The system utilizes a circular carrier substrate, with multiple spiral arms evenly arranged via Archimedean spirals. Each spiral arm is equipped with M miniature microphone sensors to form a subarray. Specifically, the system involves dynamic sampling rate allocation, confidence-weighted data fusion, and multi-algorithm hybrid beamforming technology. By adaptively reorganizing the array structure and performing real-time sound velocity correction, it solves the problem of multiple sound source localization and trajectory tracking in complex sound field environments. The system is suitable for scenarios such as industrial equipment status monitoring, noise source localization, and acoustic feature analysis. It effectively addresses the problem of multiple sound source localization and interference in complex sound fields, providing a new approach for intelligent industrial monitoring. Summary of the Invention

[0006] The present invention proposes a multi-arm spiral subarray collaborative adaptive acoustic imaging system and method, which solves the problem of multi-source localization and imaging in complex sound field environments through innovative array topology design and dynamic processing strategies. The core of the system adopts a circular carrier substrate, on the surface of which multiple spiral arm structures are arranged in an Archimedean spiral. Each spiral arm integrates a miniature microphone sensor to form a subarray. The system dynamically adjusts the working mode through real-time frequency domain analysis: in the low-frequency band, it automatically merges adjacent subarrays to expand the aperture to enhance the near-field resolution; in the mid-frequency band, it maintains the independent operation of the subarray to achieve load balancing; in the high-frequency band, it selects sensors across the spiral arms to construct a virtual focusing array to improve the far-field positioning accuracy; and when broadband noise interference is detected, it dynamically reorganizes into compact or fan-shaped subarrays according to the distance to the sound source.

[0007] In the present invention, each part of the system is described as follows:

[0008] 1. The array structure utilizes a multi-layer composite design, with a special alloy substrate as the core support platform. Optimized surface treatment processes ensure structural stability and signal integrity. The spiral layout utilizes precise mathematical modeling to create sub-array modules with robust interference resistance. Sensing units achieve high-density integration based on micro-electromechanical systems (MEMS) technology and are arranged in three dimensions based on the propagation characteristics of acoustic waves, forming a broadband, multi-channel distributed measurement network.

[0009] 2. The signal processing unit integrates a multi-channel dynamic sampling mechanism, which can adaptively allocate the sampling rate according to the working status of the sub-array, and extract the dominant frequency band in real time through fast Fourier transform. To address the frequency band competition problem of multiple sub-arrays, the system adopts a conflict resolution strategy based on signal-to-noise ratio evaluation to prioritize key acoustic features. The construction of the virtual focusing array combines the core ring array and the peripheral supplementary array, corrects the spatial position difference through the phase compensation algorithm, and applies a differentiated window function to optimize the focusing performance. A confidence weighting mechanism is introduced in the data fusion stage, and the weight coefficient is generated by comprehensively considering multiple factors such as the sub-array signal-to-noise ratio, frequency band matching and historical consistency. The Delaunay triangulation and radial basis interpolation are combined to eliminate spatial blind spots, and anisotropic diffusion filtering is applied to retain the edge characteristics of the sound source.

[0010] 3. The system integrates a real-time temperature-speed correction model. During the beamforming phase, it integrates a delayed-sum algorithm, a minimum-variance distortion-free response algorithm, and a compressed sensing algorithm to compensate for the Doppler effect of moving sound sources. The output unit generates a three-dimensional acoustic heat map that incorporates Gaussian smoothing and edge enhancement, and simultaneously outputs an analysis report that includes spectral characteristics, energy distribution, and position reliability. During implementation, a sound speed mapping table and noise baseline are established through initial environmental calibration. During the real-time processing phase, signal acquisition, dynamic reconstruction, and hybrid algorithm calculations are performed in parallel. Ultimately, a trajectory prediction algorithm is used to continuously track moving sound sources.

[0011] Key features of this invention: The acoustic array imaging system utilizes a multi-layer spiral topology, integrating intelligent frequency division processing and dynamic calibration mechanisms. Through a multimodal fusion architecture, adaptive switching of array configurations and algorithms enables high-precision, wideband sound field reconstruction. An innovative arbitration strategy optimizes resource allocation, while hybrid beamforming technology enhances dynamic target tracking capabilities, significantly enhancing imaging stability and resolution in complex environments.

[0012] The benefits and application prospects of this invention: It can be applied to industrial noise source monitoring, mechanical failure abnormal noise detection, and realize multi-modal fusion output. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 The multi-arm spiral subarray cooperative adaptive acoustic imaging system proposed by the present invention

[0014] Figure 2 Schematic diagram of array topology

[0015] Figure 3 Schematic diagram of differentiated processing mode

[0016] Figure 4 Signal processing flow chart

[0017] Figure 5The multi-arm spiral subarray collaborative adaptive acoustic imaging method proposed by the present invention DETAILED DESCRIPTION

[0018] like Figure 1 As shown in the figure, a multi-arm spiral subarray cooperative adaptive acoustic imaging system is constructed. The array topology is as follows: Figure 2 shown.

[0019] The multi-arm spiral subarray collaborative adaptive acoustic imaging system proposed in the present invention achieves high-precision positioning and imaging in complex sound field environments through a multi-level technical architecture. Figure 1 In the system, a 7075 aluminum alloy circular substrate 1 with a diameter of Φ=420mm is used, and its surface is anodized with a thickness of 50μm to improve wear resistance and signal stability. Eight spiral arms 2 are evenly distributed on the surface of the substrate according to the Archimedean spiral equation r(θ=25+0.8θ) (θ is in degrees, r is in millimeters), and the spacing between adjacent spiral arms is designed to be 12mm to ensure that there is no blind spot in the sound wave coverage. 16 MEMS microphone sensors 3 are integrated on each spiral arm. The sensor spacing is distributed according to a logarithmic law, with a proximal spacing of 6mm and a distal spacing extended to 18mm, forming a distributed array with a total of 128 channels. The sensor sensitivity is uniformly calibrated to -38dB±1dB, and the frequency response range is 20Hz-20kHz.

[0020] In the signal processing unit 4 stage, the system activates the sub-array control module 7 through the multi-channel dynamic sampling module 5 and the frequency domain analysis module 6. Figure 3 The differentiated processing modes shown achieve intelligent resource allocation. In low-frequency mode 8, the system automatically merges three adjacent subarrays to form an extended aperture, with a sampling rate set to 48kHz to match low-frequency wavelength characteristics; medium-frequency mode 9 adopts an independent subarray processing strategy, with a sampling rate increased to 96kHz to capture transient characteristics; high-frequency mode 10 selects sensors across the spiral arms to construct a virtual focused array, with a sampling rate of 192kHz to meet the Nyquist criterion. The frequency domain analysis module performs a fast Fourier transform with a period of 500ms. When it detects that the energy proportion of a certain frequency band exceeds 25%, it triggers the dominant frequency identification and prioritizes the allocation of computing resources. For multiple subarrays competing for frequency bands, the system uses a signal-to-noise ratio comparison arbitration mechanism to prioritize signal sources with a signal-to-noise ratio of more than 3dB, ensuring the effective capture of key acoustic features.

[0021] The high-frequency virtual focusing array is constructed using a layered principle: 32 sensors (spacing Δ = λ / 4 @ 15kHz) from each spiral arm are selected to form the core annular array. A Hanning window function is applied to control the main lobe width to 12°. The remaining 96 sensors serve as the peripheral supplementary array and are uniformly weighted. A phase compensation algorithm is used to correct for differences in the spatial position of the sensors, with a compensation accuracy of 0.1°.

[0022] The data fusion unit 11 stage adopts a three-level weighting strategy: the signal-to-noise ratio weight is ω SNR =1-e -SNR / 10 The calculation shows that when the frequency band matching degree exceeds 70%, the weight is doubled, and the historical consistency weight of the sub-array with a positioning deviation of less than 5° for three consecutive times is increased by 30%. The spatial interpolation algorithm first constructs the initial sound intensity distribution surface based on Delaunay triangulation, and uses thin plate spline radial basis function for the sub-array gap area. Interpolation is performed, and the smoothing coefficient λ = 0.01 balances detail preservation and noise suppression. The interpolation result is iterated by 5 times of anisotropic diffusion filtering, and the transmission coefficient Effectively maintain the sharpness of the sound source edge.

[0023] The beamforming algorithm switches dynamically according to the characteristics of the frequency band: the low-frequency mode uses a delayed sum algorithm combined with a Hamming window to achieve -42dB sidelobe suppression; the medium-frequency mode uses the MVDR algorithm with a diagonal loading set to 10% of the noise power to improve interference suppression capabilities; the high-frequency mode uses the orthogonal matching pursuit (OMP) compressed sensing algorithm, sets the sparsity K = 10 and the residual threshold to 0.1, and can complete the sound field reconstruction within 20ms. For moving sound sources, the system superimposes a Doppler effect compensation module and adjusts the delay estimate in real time according to the radial velocity v of the sound source. The temperature-sound speed correction model updates the environmental parameters every 30 seconds to ensure that the positioning error rate is less than 0.3%. The results are finally sent to the output unit 12 to generate an analysis report containing spectral characteristics, energy distribution and positioning confidence.

Claims

1. A multi-arm spiral subarray collaborative adaptive acoustic imaging system, characterized by: The multi-arm spiral subarray collaborative adaptive acoustic imaging system includes a circular carrier substrate, spiral arms, microphone sensors, a signal processing unit, a data fusion unit, and an output unit; the subsystems work in coordination: N spiral arms arranged in an Archimedean spiral are arranged on the surface of the circular carrier substrate; M miniature microphone sensors are configured on each spiral arm to form a subarray; the signal processing unit includes a multi-channel acquisition module, a frequency domain analysis module and a subarray control module, which is responsible for processing the signals received by the sensor, the multi-channel acquisition module is responsible for dynamically allocating the sampling rate according to the subarray working mode, and the frequency domain analysis module is responsible for real-time detection. The frequency band energy distribution of each subarray signal is measured, and the subarray control module is responsible for activating differentiated processing modes. In low-frequency mode, P adjacent subarrays are merged to form an extended aperture array to enhance spatial resolution. In medium-frequency mode, each subarray signal is processed independently to achieve load balancing. In high-frequency mode, sensors are selected across the spiral arms to construct a virtual focusing array to improve positioning accuracy. The data fusion unit maps the local coordinate system of each subarray to the global coordinate system through spatial alignment, and uses a confidence-weighted method to assign weight coefficients based on the signal-to-noise ratio of each subarray. Finally, an image synthesis method is used to fuse multi-source data to generate a three-dimensional acoustic thermal map. The output unit displays the enhanced heat map with superimposed sound source markers in real time, and simultaneously generates an analysis report including spectral characteristics, energy distribution and positioning reliability.

2. The multi-arm spiral subarray cooperative adaptive acoustic imaging system according to claim 1, characterized in that: The frequency domain analysis module includes the following functional units: a fast Fourier transform unit, which is used to update the spectral distribution of each subarray signal in real time at a preset period; a dominant frequency identification unit, which is used to automatically extract frequency bands whose energy proportion exceeds the threshold defined by the present invention as the priority for subarray processing; and a frequency band conflict resolution unit, which prioritizes frequency bands with higher signal-to-noise ratios based on signal quality assessment when multiple subarrays detect competing frequencies, thereby ensuring the effective capture of key acoustic features.

3. The multi-arm spiral subarray cooperative adaptive acoustic imaging system according to claim 1, characterized in that: The subarray control module's dynamic strategy includes the following: when broadband noise interference is detected, if the distance between the sound source and the center of the substrate is less than a preset value, the sensors are automatically reorganized into a compact subarray to perform near-field sound source localization; if the distance between the sound source and the center of the substrate is greater than a preset value, the sensors are divided into Q sector-shaped scanning subarrays based on the signal arrival time difference to achieve directional monitoring of the far-field area; in high-frequency mode, some sensors are selected from each spiral arm at intervals to construct a concentric annular virtual array to optimize the annular spatial focusing capability of the high-frequency signal; at the same time, when the subarray signal strength continues to be lower than the predetermined background noise threshold for a predetermined period of time, the current subarray configuration is automatically released to release computing resources.

4. The multi-arm spiral subarray cooperative adaptive acoustic imaging system according to claim 1, characterized in that: The method for constructing the virtual focusing array includes: selecting K sensors from the middle area of each spiral arm to form a core ring array; distributing the remaining sensors according to the density of the spiral lines to form a peripheral supplementary array; applying a preset window function weight to the core array, while using uniform weighting for the peripheral array; and correcting the spatial position differences between the sensors using a phase compensation algorithm.

5. The multi-arm spiral subarray cooperative adaptive acoustic imaging system according to claim 1, characterized in that: The calculation method of the confidence weighted method includes: defining a weight for the subarray signal-to-noise ratio; doubling the frequency matching weight when the overlap between a detected frequency band and the subarray's optimal operating frequency band exceeds a matching threshold defined in the present invention; and increasing the historical consistency weight priority of subarrays whose positioning results have a deviation less than an angle threshold defined in the present invention for one consecutive time.

6. The multi-arm spiral subarray cooperative adaptive acoustic imaging system according to claim 1, characterized in that: The image synthesis method implements the following steps: performing Gaussian smoothing on the low-frequency subarray results to expand the sound source coverage area; applying an edge enhancement algorithm to the high-frequency subarray results to improve detail resolution; using a Poisson equation fusion algorithm to eliminate interference fringes between subarrays; and superimposing the maximum value of multiple subarray detection results in the sound source boundary area to achieve a comprehensive display.

7. A method for cooperative adaptive acoustic imaging using a multi-arm spiral subarray, characterized by: The multi-arm spiral subarray cooperative adaptive acoustic imaging system according to claims 1 to 6 is adopted, The following stages are included: a) Initialization phase: Perform environmental calibration and collect background noise to establish a frequency energy baseline; Pre-calculate the spatial response matrix of each sub-array; b) Real-time processing: Acquire the time domain signals of each subarray in parallel, perform bandpass filtering and downsampling on them; calculate the signal delay between subarrays using a cross-correlation algorithm; dynamically select the optimal subarray combination for beamforming calculations; and initiate directional scanning of the mid- and high-frequency subarrays in the low-frequency prediction area. c) Data fusion stage: Spatial interpolation compensation is performed on the positioning results to detect blind spots; a filtering algorithm is used to track the movement trajectory of the sound source; and acoustic imaging results with confidence intervals are output.

8. The multi-arm spiral subarray collaborative adaptive acoustic imaging method according to claim 7, characterized in that: The environmental calibration includes: establishing a temperature-sound velocity correction model; measuring the background noise threshold of each subarray and setting a dynamic detection threshold; and triggering a calibration warning when the sensor position error exceeds a predetermined ratio.

9. The multi-arm spiral subarray collaborative adaptive acoustic imaging method according to claim 7, characterized in that: The beamforming calculation in the real-time processing stage includes: using a delay-sum algorithm in the low-frequency mode to improve anti-interference capability; applying a minimum variance distortion-free response algorithm in the medium-frequency mode; reconstructing the sound field distribution using a compressed sensing algorithm in the high-frequency mode; and superimposing Doppler effect compensation on moving sound sources.

10. The multi-arm spiral subarray collaborative adaptive acoustic imaging method according to claim 7, characterized in that: The spatial interpolation compensation adopts: constructing a sound intensity distribution surface based on Delaunay triangulation; applying radial basis function interpolation in the gap area covered by the subarray; and performing anisotropic diffusion filtering on the interpolation result to maintain edge features.

Images