Localization of acoustic source on solids: A linear predictive coding based algorithm for location template matching

In recent years some progresses have been made in acoustic source localization (ASL) in highly anisotropic plates when the plate material properties are not known. Some simplifying assumptions in many of the earlier source localization techniques that the wave propagates along a straight line from the source to the receiving sensor is not correct for many anisotropic plates. Wave energy in anisotropic solids generally propagate along curved lines and forms non-circular wave fronts. As a result, for highly anisotropic solids the acoustic source localization techniques that assume straight-line propagation of waves from the source to the receiver are bound to give erroneous results. In this paper earlier techniques as well as new developments for acoustic source localization in homogeneous isotropic and anisotropic plates are discussed. The newest source localization techniques in anisotropic plates analyze the non-circular shapes of the wave front. Following this analysis, the acoustic source has been successfully localized in some anisotropic plates without knowing the plate properties. Limitations of the current techniques that still exist and require further investigation are discussed.

The aim of the study was to inspect the acoustic properties and sleep characteristics of a preapneic snoring sound. The feasibility of forecasting upcoming respiratory events by snoring sound was also investigated. Participants with habitual snoring or a heavy breathing sound during sleep were recruited consecutively. Polysomnography was conducted, and snoring-related breathing sound was recorded simultaneously. Acoustic features and sleep features were extracted from 30-second samples, and a machine learning algorithm was used to establish 2 prediction models. A total of 74 eligible participants were included. Model 1, tested by 5-fold cross-validation, achieved an accuracy of 0.92 and an area under the curve of 0.94 for respiratory event prediction. Model 2, with acoustic features and sleep information tested by Leave-One-Out cross-validation, had an accuracy of 0.78 and an area under the curve of 0.80. Sleep position was found to be the most important among all sleep features contributing to the performance of the 2 models. Preapneic sound presented unique acoustic characteristics, and snoring-related breathing sound could be deployed as a real-time apneic event predictor. The models, combined with sleep information, serve as a promising tool for an early warning system to forecast apneic events. Wang B, Yi X, Gao J, etal. Real-time prediction of upcoming respiratory events via machine learning using snoring sound signal. J Clin Sleep Med. 2021;17(9):1777-1784.

Acoustic emission source localization is the basic function of the application of acoustic emission technology. For complex structures, mathematical analysis positioning algorithms cannot be obtained, and accurate acoustic emission source localization cannot be obtained, which has always been one of the problems in the actual applications of acoustic emission technology. To solve this problem, this paper proposes an acoustic emission source localization method based on deep learning, which can obtain high-precision acoustic emission source positioning without the need for mathematical analysis positioning algorithms. The AI deep learning acoustic emission source localization method adopts the method of meshing, dividing the grid in the measured structure, generating label data at the grid position, using the label data to conduct AI deep learning training to establish a positioning model, and using the trained model to analyze the actual acoustic emission data (non-labeled data) to locate the acoustic emission source. This paper uses a multi-layer perceptron model to train multi-dimensional features (arrival time and amplitude). This method effectively improves the accuracy of acoustic emission source localization. Experimental results show that the positioning accuracy of the designed deep learning model in the test set reaches 99.625%, which is significantly better than the traditional Time Difference of Arrival (TDOA) positioning algorithm. In addition, this paper further verifies the stability and reliability of the model in localization tasks through credibility metrics such as Score, Margin and Entropy. This article provides a new solution for locating acoustic emission sources in complex structures, and lays a theoretical and practical foundation for the future development of non-destructive testing technology.

Source localization of EEG signals identifies the loci of neural activity that generates the EEG voltage or power distribution on the scalp. This chapter provides a practical guide to source localization and its usage in EEG analysis in the frequency domain. Techniques proposed to overcome the inverse problem of source localization have progressed from equivalent dipole modeling to distributed source modeling, such as sLORETA and eLORETA. This chapter discusses the rationale behind these techniques, as well as the advantages and disadvantages of using different techniques for source localization. Recent advances in high-resolution structural MRI make it possible to accurately delineate and model the tissues inside the head. This chapter discusses the importance of using realistic head models constructed with anatomical MRIs for EEG source localization. Finally, it reviews the applications of source localization as a neuroimaging tool in EEG frequency analysis.

The source localization problem concerns the detection and localization of an emitter whose transmission is observed by geographically separated receivers. The passive multistatic radar (PMR) problem concerns the detection and localization of a target that scatters an illumination signal to geographically separated receivers. By modeling the scattering target as an emitter, the techniques of source localization can be applied to the PMR problem. Indeed, this approach has recently been introduced in the literature. However, the exact relationship between the two problems has not been made explicit. In this work, we derive a centralized generalized likelihood ratio test (GLRT) detector that performs the processing characteristic of both source localization and PMR. This detector is used to assess the relative detection benefit provided by source localization in PMR. We show that source localization techniques are of limited utility in PMR due to the SNR regimes of the target-scattered and direct-path signals typical of the PMR signal environment.

For the burst-type acoustic emission signal, time delay estimation of two sensors calculated by signal cross-correlation is not accurate, and it leads the source localization results is not also accurate. A new method is proposed to improve accuracy of source localization results based on multi-scale analysis and multi-sensors. Acoustic emission signal multi-scale analysis using wavelet transform, then the most accurate time delay is selected in sub-band of signal multi-scale analysis by multi-sensors time delay vector close rule. Finally, the simulation acoustic emission source localization experiment has high accuracy.

Linear predictive coding (LPC) parameters are widely used in various speech coding applications for representing the short-time spectral envelope information of speech [1]. For low bit rate speech coding applications, it is important to quantize these parameters using as few bits as possible. Considerable workhas been done in the past to develop both scalar and vector quantization procedures to quantize the LPC parameters [2, 3,4]. Scalar quantizers quantize each of the LPC parameters independently, while vector quantizers consider the entire set of LPC parameters as an entity and allow for direct minimization of quantization distortion. Because of this, the vector quantizers result in smaller distortion than the scalar quantizers at any given bit rate. The vector quantizers, however, have one major problem; their computational complexity is high. In our earlier paper [3], we have reported on a vector quantizer where the LPC parameter vector is split in the line spectral frequency (LSF) domain to overcome this complexity problem. We have shown that this quantizer can quantize the LPC parameters at 24 bits/frame with an average spectral distortion of 1 dB, less than 2% frames having spectral distortionl in the range 2-4 dB and no frame having spectral distortion greater than 4 dB.

Composite materials are increasingly being used in the structure of aircraft. However, this type of materials is commonly encountering damage companied with acoustic emission (AE) events. In order to ensure the safety of structures during service, it is urgent to develop structural health monitoring methods able to localize AE source in composite components as they occur. Kundu et al [1] for the first time proposed a technique using L-shaped clusters to localize AE source in anisotropic plates without knowing the pre-defined velocity. However, the direction dependent group velocity may lead to a large error if the group velocities in some anisotropic materials vary widely. In this paper, whether the characteristic of anisotropy has a great influence on the localization results and the principle of how this characteristic affects the AE source localization results are investigated. The results show that localization results of isotropic plate are obviously better than that of anisotropic plate due to the direction dependent velocities of anisotropic plate, which proves that the characteristic of anisotropy will affect AE source localization results. It is also found that the variation of localization orientation error is consistent with the local velocity gradient.

Advances in Bayesian computational technology in the last decade have enabled the development of new source separation and source localization algorithms. These algorithms are greatly improved by the encoding of prior information about a specific problem in the form of the chosen relevant model parameters, the assignment of the likelihood functions, and the assignment of the prior probabilities of the model parameter values. I refer to such source separation algorithms as informed source separation for the reason that they are endowed with specific and often vital information about the problem. Furthermore, the Bayesian methodology allows source separation to be united with source localization simply by including the model parameters that are of interest to the researcher. Here, I will discuss the union of source separation and source localization under the Bayesian methodology, the incorporation of prior information, and the construction of an informed algorithm using the new computational technologies that allow us to estimate the values of the parameters that define these high‐dimensional problems.

Chapter 15 - A novel sound source localization method using a global-best guided cuckoo search algorithm for drone-based search and rescue operations

Equivalent dipole source localization is a well-established approach to localizing the electrical activity in electroencephalogram (EEG). So far, source localization has been used primarily in localizing the epileptic source in human epileptic patients. Currently, source localization techniques have been applied to account for localizing epileptic source among the epileptic patients. Here, we present the first application of source localization in the field of sleep spindle in mouse brain. The spatial distribution of cortical potential was obtained by high density EEG and then the anterior and posterior sleep spindles were classified based on the K-mean clustering algorithm. To solve the forward problem, a realistic geometry brain model was produced based on boundary element method (BEM) using mouse MRI. Then, we applied four different source estimation algorithms (minimum norm, eLORETA, sLORETA, and LORETA) to estimate the spatial location of equivalent dipole source of sleep spindles. The estimated sources of anterior and posterior spindles were plotted in a cine-mode that revealed different topographic patterns of spindle propagation. The characterization of sleep spindles may be better be distinguished by our novel visualization method.

Functional delay and sum beamforming for three-dimensional acoustic source identification with solid spherical arrays

An elegant relationship (here called symmetry) between the linear predictive coding (LPC) analysis and the composite sinusoidal modeling (CSM) analysis is described. The objective is to provide a systematic view of the theory of voice spectral modeling such as LPC, PARCOR, CSM, and line spectral pairs (LSP). All of these methods of analysis are reduced to the problems of deriving the model parameters for lower‐order autocorrelation functions equal to the number of degrees of freedom of the model. From the point of view of the theory of orthogonal polynomials, LPC and CSM are the theory of orthogonal polynomials of a unit circle (in the case of LPC) and on the real axis (in the case of CSM). Beautiful symmetries hold in regard to the formulation, definition of various parameters, and solution algorithm. Also, a new interpretation is given here for the LSP from the point of view of orthogonal polynomials. © 2002 Scripta Technica, Electron Comm Jpn Pt 3, 85(6): 42–54, 2002

In this paper, we derived a relationship between Linear Predictive Coding (LPC) coefficients and Immittance Spectral Pairs (ISP) pseudo-cepstrum. In Adaptive Multi-Rate Wideband(AMR-WB) speech codec, ISP parameters were used for quantisation and interpolation. This paper first described the representation of ISP pseudo-cepstrum. ISP pseudo-cepstrum was defined as the inverse z-transform of half natural logarithm of immittance function. References [2] and [3] proposed two kinds of new perceptual weighting filters based on ISP pseudo-cepstrum. Spectral tilt is more pronounced in wideband signals due to the wide dynamic range between low and high frequencies. The proposed filters are helpful for solving this problem. Then we gave the conversion method between ISP pseudo-cepstrum and polynomials cepstrum coefficients. Finally we derived the recursive relations between ISP pseudo-cepstrum and LPC coefficients. There are two methods of converting LPC coefficients to ISP pseudo-cepstrum. The one is directly from LPC coefficients to ISP pseudo-cepstrum. The other is from ISP parameters.

Newton's method is one of the commonly used methods for acoustic emission partial discharge (AEPD) source localization in power transformers. The major problem in the AEPD source localization is the difficulty in accurately measuring the signal arrival time from the source to the multiple sensors. The exact instant at which the partial discharge (PD) occurs is not known. Therefore, instead of the absolute time, the time-delay in signal reception of each sensor with respect to the sensor nearest to the PD source is estimated. In the present study the effect of time-delay error on the accuracy of the PD source localization is analyzed through numerical experimentations. The solution of the Newtons's method in turn depends on the choice of the initial guess. Hence, different initial guesses are selected for Newton's method randomly. For a fixed set of coordinates for the four sensors and the PD source positions, the time-delays are theoretically calculated. The error is systematically increased in the calculated time-delay to form different groups of time-delays. The PD source is localized using each initial guess for all the groups of time-delays to analyze the effect of error in time-delay on accuracy of PD source localization. The error in time-delay greatly affects the AEPD source localization. Moreover, if Newton's method with a bad initial guess is used then the errors in AEPD source localizations are not systematic.