Abstract
For near-field acoustic holography, sparse array measurement for cost reduction can result in inaccuracy due to aliasing error. To attenuate it, there are data-driven methods based on artificial intelligence theories. Among these, the JTCSA-NAH method has not adopted measures for robustness enhancement despite its high accuracy in practice. In this work, the influence of measuring noise on JTCSA-NAH is analyzed followed by the principle of adding Gaussian noise for robustness improvement. Based on the relevant prior conditions, the ICCSA-NAH method, which relies on self-identity constraint data working as the Kalman gain is proposed. Subsequently, numerical example and experiment are carried out, and the results show that compared with JTCSA-NAH method, the mean errors of near-field vibration velocity reconstruction are theoretically and experimentally reduced from 15.19% and 23.64% to 6.03% and 12.45%, respectively, by the ICCSA-NAH method, which verifies the feasibility and superiority of the proposed method.
Similar content being viewed by others
Data availability
Data available on request from the authors.
References
Pézerat C et al (2009) Identification of vibration excitations from acoustic measurements using near field acoustic holography and the force analysis technique. J Sound Vib 326(3):540–556
Guozhi B et al (2017) Identification and contribution analysis of vehicle interior noise based on acoustic array technology. Adv Mech Eng 9(11):1687814017730031
Lu W et al (2012) A fault diagnosis scheme of rolling element bearing based on near-field acoustic holography and gray level co-occurrence matrix. J Sound Vib 331(15):3663–3674
Williams EG (1999) Fourier acoustics : sound radiation and nearfield acoustical holography. Academic Press, Cambridge
Bai MR (1992) Application of BEM (boundary element method)-based acoustic holography to radiation analysis of sound sources with arbitrarily shaped geometries. J Acoust Soc Am 92(1):533–549
Williams EG, Houston BH, Herdic PC (2003) Fast Fourier transform and singular value decomposition formulations for patch nearfield acoustical holography. J Acoust Soc Am 114(3):1322–1333
Jørgen Hald (2009) Basic theory and properties of statistically optimized near-field acoustical holography. J Acoust Soc Am 125(4):2105–2120. https://doi.org/10.1121/1.3079773
Pasqual AM (2017) A patch near-field acoustical holography procedure based on a generalized discrete Fourier series. Mech Syst Signal Process 90:285–297
Koopmann GH, Song L, Fahnline JB (1989) A method for computing acoustic fields based on the principle of wave superposition. J Acoust Soc Am 86(6):2433–2438
Wang Z, Wu SF (1997) Helmholtz equation–least-squares method for reconstructing acoustic pressure fields. J Acoust Soc Am 102(5):3090–3090
Donoho D (2006) Compressed sensing. IEEE Trans Inf Theory 52:1289–1306
Chardon G et al (2012) Nearfield acoustic holography using sparsity and compressive sampling principles. J Acoust Soc Am 132(3):1521–1534
Fernandez-Grande E, Xenaki A, Gerstoft P (2017) A sparse equivalent source method for near-field acoustic holography. J Acoust Soc Am 141(1):532–542
Bi C-X et al (2017) Sound field reconstruction using compressed modal equivalent point source method. J Acoust Soc Am 141(1):73–79
Fernandez-Grande E, Daudet L (2018) Compressive acoustic holography with block-sparse regularization. J Acoust Soc Am 143(6):3737–3746
He Y et al (2019) A compressed equivalent source method based on equivalent redundant dictionary for sound field reconstruction. Appl Sci 9(4):408
Hald J (2018) A comparison of iterative sparse equivalent source methods for near-field acoustical holography. J Acoust Soc Am 143(6):3758–3769
Zhao R et al (2019) Deep learning and its applications to machine health monitoring. Mech Syst Signal Process 115:213–237
Dai J et al (2019) Signal-based intelligent hydraulic fault diagnosis methods: review and prospects. Chin J Mech Eng 32(1):75
Niu S, Srivastava V (2022) Ultrasound classification of interacting flaws using finite element simulations and convolutional neural network. Eng Comput. https://doi.org/10.1007/s00366-022-01681-y
He Y et al (2022) A framework of structural damage detection for civil structures using a combined multi-scale convolutional neural network and echo state network. Eng Comput. https://doi.org/10.1007/s00366-021-01584-4
Wang Q, Zhuang X (2022) A CNN-based surrogate model of isogeometric analysis in nonlocal flexoelectric problems. Eng Comput. https://doi.org/10.1007/s00366-022-01717-3
Ping G et al (2020) Three-dimensional source localization using sparse Bayesian learning on a spherical microphone array. J Acoust Soc Am 147(6):3895–3904
Antoni J et al (2019) Sparse acoustical holography from iterated Bayesian focusing. J Sound Vib 446:289–325
Zan M et al (2020) A sound source identification algorithm based on bayesian compressive sensing and equivalent source method. Sensors. https://doi.org/10.3390/s20030865
Fan B, Das S (2021) Synthetic aperture acoustic imaging with deep generative model based source distribution prior. In: ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).
Olivieri M et al. (2021) Near field acoustic holography on arbitrary shapes using convolutional neural network. In: 2021 29th European Signal Processing Conference (EUSIPCO).
Olivieri M et al (2021) A physics-informed neural network approach for nearfield acoustic holography. Sensors 21(23):7834
Wu H et al (2020) Acoustic spatial patterns recognition based on convolutional neural network and acoustic visualization. J Acoust Soc Am 147(1):459–468
Wang J et al (2021) A 3D convolutional neural network based near-field acoustical holography method with sparse sampling rate on measuring surface. Measurement 177:109297
Wang J et al (2023) A cylindrical near-field acoustical holography method based on cylindrical translation window expansion and an autoencoder stacked with 3D-CNN layers. Sensors. https://doi.org/10.3390/s23084146
Wang J et al (2022) Research on joint training strategy for 3D convolutional neural network based near-field acoustical holography with optimized hyperparameters. Measurement 202:111790
Hald J (2014) Scaling of plane-wave functions in statistically optimized near-field acoustic holography. J Acoust Soc Am 136(5):2687–2696
Bi C-X et al (2007) Reconstruction and separation in a semi-free field by using the distributed source boundary point method-based nearfield acoustic holography. J Vib Acoust 129:323
Bishop CM (1995) Training with noise is equivalent to tikhonov regularization. Neural Comput 7(1):108–116
Acknowledgements
This work was funded by the Liuzhou Science and Technology Planning Project under (Grant nos. 2020GAAA0403), Project supported by the Education Department of Hainan Province (Project no. Hnky2023ZD-3) and Hainan University research start-up fund (Grant no. KYQD(ZR)23085). The authors gratefully acknowledge the supports.
Funding
This article is funded by Liuzhou Science and Technology Planning Project, 2020GAAA0403, Qibai Huang, Project supported by the Education Department of Hainan Province, Hnky2023ZD-3, Zhifu Zhang, Hainan University research start-up fund, KYQD(ZR)23085, Zhifu Zhang.
Author information
Authors and Affiliations
Corresponding authors
Ethics declarations
Conflict of interest
We declare that we have no financial and personal relationships with other people or organizations that can inappropriately influence our work. There is no professional or other personal interest of any nature or kind in any product, service and/or company that could be construed as influencing the position presented in, or the review of, the manuscript entitled.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Electronic supplementary material
Below is the link to the electronic supplementary material.
A. Appendix
A. Appendix
The JTCSA-NAH method is composed of two parts. The first is the SAE module from original velocity to reconstructed velocity. The other part is the combined pressure to velocity (PV)-NN composed of holography feature extraction neural network (HFENN) and the decoder in the first part, which maps from the theoretical sparse pressure to the reconstructed velocity. Then we train the network to obtain the optimal parameter sets:
When the measured pressure data includes noise n, the near-field velocity generalized by the trained network is:
Assuming the network consists of N layers, \(\hat{{v}}_{{\varvec{n}}}\) is the output of layer N. The output xn of each layer is determined by xn-1, and xN = \(\hat{{v}}_{{\varvec{n}}}\).
where On-1 is the function vector composed of activation functions of all neurons in layer n, ⊗ denotes generalized multiplication, representing the mapping of the function from its multiplied term. w* and b* are the weight vector and bias of each layer, respectively. There is no parameter in the upsampling layer, so it is not considered.
When the system is linear, the difference between generalized two different sound pressure data \(\hat{{v}}_{{\varvec{n}}}\) and \(\hat{{v}}\) is \(\Delta \hat{{v}}\):
In Eq. (25), the difference between the surface vibration velocity obtained by the generalization of the noise-free and the noisy data are substituted into the MSE loss function to obtain the final deference of loss function:
When the noise n on each pixel follows a Gaussian distribution with a mean value of 0 (\({\varvec{n}}\sim \mathcal{N}\left(0,\delta \right)\)), and there are enough samples with random additional noise, the mean of each pixel noise is 0, and the distribution of noise n and vibration velocity v is independent, and so the left term of ΔL is 0 in Eq. (29). Therefore, for linear systems, when there is measurement noise, the model trained by only clean data will magnify the reconstruction error. As for the PV-NN, although the activation function ReLU: y = (x + x·sign(x))/2 adopted in the network has a certain degree of linearity, the network framework is nonlinear globally.
For nonlinear systems, substituting the clean and noisy pressure into MSE loss function will yield:
where the second term of ΔL (the last line of Eq. (30)) is larger than 0. For the remaining cross term (the penultimate line of Eq. (30)), when the statistical energy of noise is relatively low, and random combination samples are sufficient, the second-order infinitesimal term of it can be ignored, and therefore it turns to:
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Wang, J., Huang, Y., Li, Z. et al. A novel robust approach of 3D CNN and SAE-based near-field acoustical holography relying on self-identity constraint data for Kalman gain. Engineering with Computers (2023). https://doi.org/10.1007/s00366-023-01911-x
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s00366-023-01911-x