Acoustical inverse problems regularization: Direct definition of filter factors using Signal-to-Noise Ratio

Acoustic imaging aims at localization and characterization of sound sources using microphone arrays. In this paper a new regularization method for acoustic imaging by inverse approach is proposed. The method first relies on the singular value decomposition of the plant matrix and on the projection of the measured data on the corresponding singular vectors. In place of regularization using classical methods such as truncated singular value decomposition and Tikhonov regularization, the proposed method involves the direct definition of the filter factors on the basis of a thresholding operation, defined from the estimated measurement noise. The thresholding operation is achieved using modified filter functions. The originality of the approach is to propose the definition of a filter factor which provides more damping to the singular components dominated by noise than that given by the Tikhonov filter. This has the advantage of potentially simplifying the selection of the best regularization amount in inverse problems. Theoretical results show that this method is comparatively more accurate than Tikhonov regularization and truncated singular value decomposition.     

A subgrid computation of the curvature by a particle/level-set method. Application to a front-tracking/ghost-fluid method for incompressible flows

A particle/level-set method is developed to capture the interface between two immiscible phases. No redistance equation is used for the level-set function which is built from an accurate cubic spline interpolation of the markers describing the interface. Mass losses which usually appear with level-set functions are drastically reduced. Interface coupling with the incompressible momentum equation is taken into account thanks to the ghost-fluid method. Indeed, the particles allow an accurate computation of the local interfacial curvatures, and the capillary part of the pressure jump computed across the interface is more accurately evaluated. Results on static drop show a large decrease of spurious currents compared to other methods. Scalar advection validations and convergence studies are carried out as well as tough test-cases involving large density ratios (typically air/water). A better agreement with literature results is shown compared to classical approaches.     

An MPI parallel level-set algorithm for propagating front curvature dependent detonation shock fronts in complex geometries

We present a parallel, two-dimensional, grid-based algorithm for solving a level-set function PDE that arises in Detonation Shock Dynamics (DSD). In the DSD limit, the detonation shock propagates at a speed that is a function of the curvature of the shock surface, subject to a set of boundary conditions applied along the boundaries of the detonating explosive. Our method solves for the full level-set function field, φ(x, y, t), that locates the detonation shock with a modified level-set function PDE that continuously renormalises the level-set function to a distance function based off of the locus of the shock surface, φ(x, y, t)=0. The boundary conditions are applied with ghost nodes that are sorted according to their connectivity to the interior explosive nodes. This allows the boundary conditions to be applied via a local, direct evaluation procedure. We give an extension of this boundary condition application method to three dimensions. Our parallel algorithm is based on a domain-decomposition model which uses the Message-Passing Interface (MPI) paradigm. The computational order of the full level-set algorithm, which is O(N ⁴), where N is the number of grid points along a coordinate line, makes an MPI-based algorithm an attractive alternative. This parallel model partitions the overall explosive domain into smaller sub-domains which in turn get mapped onto processors that are topologically arranged into a two-dimensional rectangular grid. A comparison of our numerical solution with an exact solution to the problem of a detonation rate stick shows that our numerical solution converges at better than first-order accuracy as measured by an L1-norm. This represents an improvement over the convergence properties of narrow-band level-set function solvers, whose convergence is limited to a floor set by the width of the narrow band. The efficiency of the narrow-band method is recovered by using our parallel model.     

The application of regularization technique based on partial optimization in the nearfield acoustic holography

The regularization technique for stabilizing the reconstruction based on the nearfield acoustic holography(NAH) was investigated on the basis of the equivalent source method.In order to obtain higher regularization effect,a regularization method based on the idea of partial optimization was proposed,which inherits the advantages of the Tikhonov and another regularization method—truncated singular value decomposition(TSVD).Through the numerical simulation,it is proved that the proposed method is stabler than the Tikhonov,and more precise than the TSVD.Finally the validity and the feasibility of the proposed method are demonstrated by an experiment carried out in a semi-anechoic room with two speakers.     

联合波叠加法的全息理论与实验研究

当空间声场中同时存在多个相干声源时，运用常规近场声全息方法无法重建每个相干声源表面的声学信息，当然也无法预测每个声源单独产生的空间声场，相干声场的全息重建与预测已成为全息技术推广应用过程中亟待解决的问题.在提出联合波叠加法并将其应用于空间声场变换的基础上，对其进行了实验研究.通过对实际相干声场的全息重建与预测，验证了常规波叠加法在相干声场重建中的局限性、联合波叠加法在相干声场全息重建与预测过程的可行性和准确性，还研究了Tikhonov正则化方法在抑制声学逆问题的非适定性中的有效性和滤波系数的选择原则的可行性，以提高全息重建与预测的精度. 关键词： 近场声全息 联合波叠加 相干声场 Tikhonov正则化     

Numerical simulation of premixed combustion using an enriched finite element method

In this paper we present a novel discretization technique for the simulation of premixed combustion based on a locally enriched finite element method (FEM). Use is made of the G-function approach to premixed combustion in which the domain is divided into two parts, one part containing the burned and another containing the unburned gases. A level-set or G-function is used to define the flame interface separating burned from unburned gases. The eXtended finite element method (X-FEM) is employed, which allows for velocity and pressure fields that are discontinuous across the flame interface. Lagrange multipliers are used to enforce the correct essential interface conditions in the form of jump conditions across the embedded flame interface. A persisting problem with the use of Lagrange multipliers in X-FEM has been the discretization of the Lagrange multipliers. In this paper the distributed Lagrange multiplier technique is adopted. We will provide results from a spatial convergence analysis showing good convergence. However, a small modification of the interface is required to ensure a unique solution. Finally, results are presented from the application of the method to the problems of moving flame fronts, the Darrieus–Landau instability and a piloted Bunsen burner flame.     

3D cerebrovascular segmentation combining fuzzy vessel enhancement and level-sets with anisotropic energy weights

The aim of this work is to present and evaluate a level-set segmentation approach with vesselness-dependent anisotropic energy weights, which focuses on the exact segmentation of malformed as well as small vessels from time-of-flight (TOF) magnetic resonance angiography (MRA) datasets.     

Level-set based topology optimization for electromagnetic dipole antenna design

This paper presents a level-set framework for a typical electromagnetic design problem of dipole antenna. In this study, the geometrical configuration of an antenna is represented by the zero-level contour of a higher-dimensional level-set function. The governing equation for the induced current flow on a metal surface is the Electric Field Integral Equation (EFIE), which takes into account the electric component of the incident wave. The design objective is formulated in terms of the surface current and incident electric field. The normal velocity of the level-set model, which reflects the sensitivity of the objective function, is derived from the adjoint variable method and shape derivative. By optimizing the objective function, the area with the highest current density, to which the voltage feeding should be applied, can be reshaped. The advantages of adopting the level-set technique for electromagnetic design lie in its capacity for capturing sophisticated topological changes and facilitation in mathematical representation of the design configuration. The demonstrative examples of dipole antenna design show that the level-set method results in a fairly smooth optimization process, where the vacuum/metal interface gradually attains its optimal configuration. A series of design cases with self-adjoint and non-self-adjoint sensitivity analyses are studied and compared to the benchmarking problems in dipole antenna.     

Level-Set Minimization of Potential Controlled Hadwiger Valuations for Molecular Solvation

A level-set method is developed for the numerical minimization of a class of Had-wiger valuations with a potential on a set of three-dimensional bodies. Such valuations are linear combinations of the volume, surface area, and surface integral of mean curvature. The potential increases rapidly as the body shrinks beyond a critical size. The combination of the Hadwiger valuation and the potential is the mean-field free-energy functional of the solvation of non-polar molecules in the recently developed variational implicit-solvent model. This functional of surfaces is minimized by the level-set evolution in the steepest decent of the free energy. The normal velocity of this surface evolution consists of both the mean and Gaussian curvatures, and a lower-order, "forcing" term arising from the potential. The forward Euler method is used to discretize the time derivative with a dynamic time stepping that satisfies a CFL condition. The normal velocity is decomposed into two parts. The first part consists of both the mean and Gaussian curvature terms. It is of parabolic type with parameter correction, and is discretized by central differencing. The second part has all the lower-order terms. It is of hyperbolic type, and is discretized by an upwinding scheme. New techniques of local level-set method and numerical integration are developed. Numerical tests demonstrate a second-order convergence of the method. Examples of application to the modeling of molecular solvation are presented.     

Refining Spin–Spin Distance Distributions in Complex Biological Systems Using Multi-Gaussian Monte Carlo Analysis

Pulse dipolar electron paramagnetic resonance spectroscopy provides means of distance measurements in the range of ~ 1.5–10 nm between two spin labels tethered to a biological system. However, the extraction of distance distribution between spin labels is an ill-posed mathematical problem. The most common approach for obtaining distance distribution employs Tikhonov regularization method, where a regularization parameter characterizing the smoothness of distribution is introduced. However, in case of multi-modal distance distributions with peaks of different widths, the use of a single regularization parameter might lead to certain distortions of actual distribution shapes. Recently, a multi-Gaussian Monte Carlo approach was proposed for eliminating this drawback and verified for model biradicals [1]. In the present work, we for the first time test this approach on complicated biological systems exhibiting multi-modal distance distributions. We apply multi-Gaussian analysis to pulsed electron–electron double resonance data of supramolecular ribosomal complexes, where the 11-mer oligoribonucleotide (MR) bearing two nitroxide labels at its termini is used as a reporter. Calculated distance distributions reveal the same conformations of MR as those obtained by Tikhonov regularization, but feature the peaks having different widths, which leads to a better resolution in several cases. The advantages, complications, and further perspectives of application of Monte-Carlo-based multi-Gaussian approach to real biological systems are discussed.     

光偏折层析的Tikhonov正则化重建技术

从不适定问题的求解思想出发,建立了使用Tikhonov正则化方法计算流场参量分布的光偏折层析重建技术。理论上详细讨论了偏折信息的转化、投影方程组的正则化、以及由共轭梯度法实现求解等三个主要步骤。用数值模拟考察该方法对非对称温度分布的重建效果,分别计算了对应不同采样数的欠定与超定投影方程组。结果表明,10次迭代后重建分布的平均误差和峰值误差分别为2.1%和5.2%。使用正则化方法对采集的多方向叠栅偏折投影进行计算,重建出双峰温度场。结果表明,使用正则化方法的双峰值相对误差略高,迭代次数减少,可提高重建效率。     

Crack reconstruction using a level-set strategy

We present a level-set based technique to recover key characteristics of a defect or crack (e.g. location, length and shape) in a two-dimensional material from boundary electrical measurements. The key feature of this work is to extend the usual level-set technique for modeling volumetric objects to very thin objects. Two level-set functions are employed: the first one models the location and form of the crack, and the second one models its length and connectivity. An efficient gradient based method is derived in order to define evolution laws for these two level-set functions which minimize the least squares data misfit. Numerical experiments show the utility of this method even in the presence of a significant noise level in the measurements. A finite element method is used to simulate the electric field behavior in the presence of very thin objects.     

Mode shape expansion using perturbed force approach

A new approach for expanding incomplete experimental mode shapes is presented which considers the modelling errors in the analytical model and the uncertainties in the vibration modal data measurements. The proposed approach adopts the perturbed force vector that includes the effect of the discrepancy in mass and stiffness between the finite element model and the actual tested dynamic system. From the developed formulations, the perturbed force vector can be obtained from measured modal data and is then used for predicting the unmeasured components of the expanded experimental mode shapes. A special case that does not require the experimental natural frequency in the mode shape expansion process is also discussed. A regularization algorithm based on the Tikhonov solution incorporating the generalized cross-validation method is employed to filter out the influence of noise in measured modal data on the predictions of unmeasured mode components. The accuracy and robustness of the proposed approach is verified with respect to the size of measured data set, sensor location, model deficiency and measurement uncertainty. The results from two numerical examples, a plane frame structure and a thin plate structure, show that the proposed approach has the best performance compared with the commonly used existing expansion methods, and can reliably produce the predictions of mode shape expansion, even in the cases with limited modal data measurements, large modelling errors and severe measurement noise.     

An efficient method for model refinement in diffuse optical tomography

Diffuse optical tomography (DOT) is a non-linear, ill-posed, boundary value and optimization problem which necessitates regularization. Also, Bayesian methods are suitable owing to measurements data are sparse and correlated. In such problems which are solved with iterative methods, for stabilization and better convergence, the solution space must be small. These constraints subject to extensive and overdetermined system of equations which model retrieving criteria specially total least squares (TLS) must to refine model error. Using TLS is limited to linear systems which is not achievable when applying traditional Bayesian methods. This paper presents an efficient method for model refinement using regularized total least squares (RTLS) for treating on linearized DOT problem, having maximum a posteriori (MAP) estimator and Tikhonov regulator. This is done with combination Bayesian and regularization tools as preconditioner matrices, applying them to equations and then using RTLS to the resulting linear equations. The preconditioning matrixes are guided by patient specific information as well as a priori knowledge gained from the training set. Simulation results illustrate that proposed method improves the image reconstruction performance and localize the abnormally well.     

A conditioned level-set method with block-division strategy to flame front extraction based on OH-PLIF measurements

A novel approach to extract flame fronts, which is called the conditioned level-set method with block division （CLSB）, has been developed. Based on a two-phase level-set formulation, the conditioned initialization and region-lock optimiza-tion appear to be beneficial to improve the efficiency and accuracy of the flame contour identification. The original block- division strategy enables the approach to be unsupervised by calculating local self-adaptive threshold values autonomously before binarization. The CLSB approach has been applied to deal with a large set of experimental data involving swirl- stabilized premixed combustion in diluted regimes operating at atmospheric pressures. The OH-PLIF measurements have been carried out in this framework. The resulting images are, thus, featured by lower signal-to-noise ratios （SNRs） than the ideal image; relatively complex flame structures lead to significant non-uniformity in the OH signal intensity; and, the mag- nitude of the maximum OH gradient observed along the flame front can also vary depending on flow or local stoichiometry. Compared with other conventional edge detection operators, the CLSB method demonstrates a good ability to deal with the OH-PLIF images at low SNR and with the presence of a multiple scales of both OH intensity and OH gradient. The robustness to noise sensitivity and intensity inhomogeneity has been evaluated throughout a range of experimental images of diluted flames, as well as against a circle test as Ground Truth （GT）.     

Structural source identification using a generalized Tikhonov regularization

This paper addresses the problem of identifying mechanical exciting forces from vibration measurements. The proposed approach is based on a generalized Tikhonov regularization that allows taking into account prior information on the measurement noise as well as on the main characteristics of sources to identify like its sparsity or regularity. To solve such a regularization problem efficiently, a Generalized Iteratively Reweighted Least-Squares (GIRLS) algorithm is introduced. Proposed numerical and experimental validations reveal the crucial role of prior information in the quality of the source identification and the performance of the GIRLS algorithm.     

Transient nearfield acoustic holography based on an interpolated time-domain equivalent source method

Transient nearfield acoustic holography based on an interpolated time-domain equivalent source method (ESM) is proposed to reconstruct transient acoustic fields directly in the time domain. Since the equivalent source strengths solved by the traditional time-domain ESM formulation cannot be used to reconstruct the pressure on the source surface directly, an interpolation function is introduced to develop an interpolated time-domain ESM formulation which permits one to deduce an iterative reconstruction process. As the reconstruction process is ill-conditioned and especially there exists a cumulative effect of errors, the Tikhonov regularization is used to stabilize the process. Numerical examples of reconstructing transient acoustic fields from a baffled planar piston, an impulsively accelerating sphere and a cube box, respectively, demonstrate that the proposed method not only can effectively reconstruct transient acoustic fields in the time domain, but also can visualize acoustic fields in the space domain. And, in the first numerical example, the cumulative effect of errors and the validity of using the Tikhonov regularization to suppress the errors are described.     

Designing shielded radio-frequency phased-array coils for magnetic resonance imaging

In this paper, an approach to the design of shielded radio-frequency (RF) phased-array coils for magnetic resonance imaging (MRI) is proposed. The target field method is used to find current densities distributed on primary and shield coils. The stream function technique is used to discretize current densities and to obtain the winding patterns of the coils. The corresponding highly ill-conditioned integral equation is solved by the Tikhonov regularization with a penalty function related to the minimum curvature. To balance the simplicity and smoothness with the homogeneity of the magnetic field of the coil’s winding pattern, the selection of a penalty factor is discussed in detail.     

Compensation for discarded singular values in vibro-acoustic inverse methods

The paper addresses the inverse problem where source strengths are back-calculated from a sound pressure field sampled at several points. Regularization techniques, such as singular value discarding or Tikhonov regularization, are commonly used to improve estimates of source strength in such situations. However, over-regularization can result in even worse errors. A simple procedure is proposed here to compensate for errors of over-regularization. The basis is to constrain the solution such that the spatial mean of the measured and reconstructed sound pressure are equal. In other words, to set the overall sound power of the equivalent (calculated) sources equal to that of the real source. It is argued that the overall sound power is the most stable and reliable quantity on which to base source strength estimates. Examples of both singular value discarding and Tikhonov regularization are given.     

Damage identification in plates using finite element model updating in time domain

A response sensitivity-based approach is presented for identifying the local damages in isotropic plate structures from the measured structural dynamic responses. The local damage is simulated by a reduction in the elemental Young's modulus of the plate. In the forward analysis, the forced vibration responses of the plate under external force are obtained from Newmark direct integration. In the inverse analysis, a response sensitivity-based finite element model updating approach is used to identify local damages of the plate in time domain. The damage identification results are obtained iteratively with the penalty function method with Tikhonov regularization using the measured structural dynamic responses. Two numerical examples are investigated to illustrate the correctness and efficiency of the proposed method. Both single damage and multiple damages cases are studied. The effects of measurement noise and measurement point on the identification results are investigated. Studies in this paper indicate that the proposed method is efficient and robust for both single and multiple damages for plate structures. Good identified results can be obtained from the short time histories of a few number of measurement points.