Smoothing Technique Research Articles

A Smoothing EKF-UI-WDF Method for Simultaneous Identification of Structural Systems and Unknown Seismic Inputs without Direct Feedthrough

It is of great significance to identify structural state-parameters and the unknown seismic inputs using partial measurements of structural acceleration responses for the rapid evaluation of structures after unknown seismic excitations. However, unknown seismic inputs do not directly appear in the observation equations of measured absolute floor accelerations of building structures, i.e., there is no direct feedthrough of unknown seismic inputs in the observation equations. Current methods for the identification of joint structural systems and unknown inputs are either inapplicable or greatly influenced by measurement noises. In this paper, a method so-called smoothing extended Kalman filter with unknown input without direct feedthrough (smoothing EKF-UI-WDF) is proposed. The identification algorithm is derived in the framework of minimum-variance unbiased estimation (MVUE), and the smoothing technique is adopted to introduce subsequent observation steps in the current identification step. Then, structural states, parameters, and unknown seismic excitations without direct feedthrough are simultaneously identified recursively with only a few steps delay, and the identification results are tolerant to measurement noises. The proposed method is verified by a numerical simulation model and a practical engineering case study. Both identification results validate the effectiveness of the proposed method for the simultaneous identification of structural systems and seismic inputs without direct feedthrough.

Simultaneous estimation of time-varying boundary heat flux and temperature-dependent thermophysical and optical parameters of grey translucent medium

A hybrid optimization method, which combines unscented Kalman smoothing (UKS) technique and sequential quadratic programming (SQP) algorithm, is proposed to estimate the time-dependent boundary heat flux and temperature-dependent optical and thermophysical parameters (such as absorption coefficient, refractive index and thermal conductivity) of grey translucent medium simultaneously. The transient coupled radiative and conductive problem in grey translucent medium is solved by finite volume method (FVM) in forward model. In inverse problem, the measured signals consist of the temperature and radiative intensity on the boundaries of the medium. The effects of model parameters (such as the number of future time step and process noise covariance) and different incident heat fluxes on the reconstructed results are studied by the UKS-SQP technique. The results indicate that the proposed hybrid optimization technique is accurate and effective for determining the time-varying boundary heat flux, temperature-dependent optical and thermophysical parameters of grey translucent medium even with the measurement error.

A probabilistic method for dynamic force identification of uncertain structures

To address the dynamic force identification problem of uncertain structures with random parameters, this study proposes a novel approach to describe the probabilistic properties of the identified forces. First, the integral equation for the force probability density function (PDF) is derived from the mathematical mapping relationship between the identified force and random structural parameters. Next, the sample space of the random parameters is partitioned into non-overlapping subdomains using a point selection technique based on the generalized F-discrepancy. Meanwhile, the function fitting technique and Tikhonov regularization method are applied to identify the force samples under the corresponding partitioned subdomains. Thus, the integral equation of the force PDF is transformed into a summation of the force probabilities for all the subdomains. Subsequently, a smoothing technique is introduced to obtain the expression of the identified force PDF. Finally, the confidence interval over the entire time domain is constructed from the force cumulative distribution function, which effectively quantifies the influence of random structural parameters on the identified results. The efficiency and superiority of the proposed method are verified using an example of a plane truss structure with multiple random structural parameters, and force identification for a simplified launch vehicle model with connection uncertainties is discussed.

An inhomogeneous stabilized node-based smoothed radial point interpolation method for the multi-physics coupling responses of functionally graded magneto-electro-elastic structures

In this paper, an inhomogeneous stabilized node-based smoothed radial point interpolation method (ISNS-RPIM) is developed for analyzing functionally graded magneto-electro-elastic (FGMEE) structures in hygrothermal environment. The change of material gradient follows the exponential function distribution. Combining with the gradient smoothing technique (GST) and the coupling constitutive relationship, and considering the hygrothermo-magneto-electro-elastic (HMEE) coupling effects of FGMEE material, the equations for ISNS-RPIM are deduced. The generalized displacement of the FGMEE structures is obtained, and the error in total energy norm, CPU time and efficiency of ISNS-RPIM, NS-RPIM and FEM are researched. Numerical examples are provided to explore the influence of the following parameters in FGMEE structures including exponential factor, hygrothermal loading and empirical constants, and demonstrate the correctness, high effectivity and insensitivity to mesh distortion of ISNS-RPIM. The present method has shown great potential in solving practical complex problems.

Cohort-based smoothing methods for age-specific contact rates.

The use of social contact rates is widespread in infectious disease modeling since it has been shown that they are key driving forces of important epidemiological parameters. Quantification of contact patterns is crucial to parameterize dynamic transmission models and to provide insights on the (basic) reproduction number. Information on social interactions can be obtained from population-based contact surveys, such as the European Commission project POLYMOD. Estimation of age-specific contact rates from these studies is often done using a piecewise constant approach or bivariate smoothing techniques. For the latter, typically, smoothness is introduced in the dimensions of the respondent's and contact's age (i.e., the rows and columns of the social contact matrix). We propose a smoothing constrained approach-taking into account the reciprocal nature of contacts-introducing smoothness over the diagonal (including all subdiagonals) of the social contact matrix. This modeling approach is justified assuming that when people age their contact behavior changes smoothly. We call this smoothing from a cohort perspective. Two approaches that allow for smoothing over social contact matrix diagonals are proposed, namely (i) reordering of the diagonal components of the contact matrix and (ii) reordering of the penalty matrix ensuring smoothness over the contact matrix diagonals. Parameter estimation is done in the likelihood framework by using constrained penalized iterative reweighted least squares. A simulation study underlines the benefits of cohort-based smoothing. Finally, the proposed methods are illustrated on the Belgian POLYMOD data of 2006. Code to reproduce the results of the article can be downloaded on this GitHub repository https://github.com/oswaldogressani/Cohort_smoothing.

Near real-time prediction of thermophysical parameters and time-dependent boundary heat flux of participating medium

The scattering coefficient (κs), absorption coefficient (κa), thermal conductivity (λ) and time-varying boundary heat flux (q(t)) of 1D participating medium are simultaneously determined in near real time. The forward problem of coupled conductive and radiative heat transfer is solved by finite volume method (FVM). The unscented Kalman filter (UKF) and corresponding smoothing technique (UKS) are employed to estimate the q(t), λ, κa and κs. To achieve non-intrusive detection, the measured signals on boundaries of the medium (such as radiative intensity on right boundary and temperatures on boundaries) are employed as inputs in the inverse problem. The effects of measurement error covariance, process error covariance, future measured signals, sampling interval and thermophysical properties on estimation results are thoroughly investigated. The UKS algorithm, when compared to the UKF algorithm, can improve the stability and accuracy of estimated q(t), λ, κa and κs. It is found that the UKS algorithm is efficient and accurate to solve the near real-time prediction of thermophysical parameters and boundary heat flux of participating medium even with measurement errors.

Electromyogram (EMG) Signal Classification Based on Light-Weight Neural Network with FPGAs for Wearable Application

Recently, the application of bio-signals in the fields of health management, human–computer interaction (HCI), and user authentication has increased. This is because of the development of artificial intelligence technology, which can analyze bio-signals in numerous fields. In the case of the analysis of bio-signals, the results tend to vary depending on the analyst, owing to a large amount of noise. However, when a neural network is used, feature extraction is possible, enabling a more accurate analysis. However, if the bio-signal time series is analyzed as is, the total neural network increases in size. In this study, to accomplish a light-weight neural network, a maximal overlap discrete wavelet transform (MODWT) and a smoothing technique are used for better feature extraction. Moreover, the learning efficiency is increased using an augmentation technique. In designing the neural network, a one-dimensional convolution layer is used to ensure that the neural network is simple and light-weight. Consequently, the light-weight attribute can be achieved, and neural networks can be implemented in edge devices such as the field programmable gate array (FPGA), yielding low power consumption, high security, fast response times, and high user convenience for wearable applications. The electromyogram (EMG) signal represents a typical bio-signal in this study.

The influence of smoothing techniques on the accuracy of the reference finite helical axis when applied to 2D-3D registrations.

Highspeed Biplanar Videoradiography (HSBV) permits recording of 3D bone movements with sub-millimeter precision. 2D-3D registrations are performed to quantify bone movements, providing a series of affine transformation matrices (ATMs). These registrations may result in alignment errors that produce inaccurate kinematics. Smoothing techniques can be applied to the ATMs to reduce these inaccuracies. Which techniques are best for this application remain unknown. The purpose of this study was to investigate the performance of six smoothing techniques on ATMs obtained from HSBV. Performance was assessed by measuring the accuracy of three reference finite helical axis (rFHA) measures during a turntable rotation: orientation, dispersion, and rotation speed difference (RSD = rFHA RS-turntable RS). A 3D printed femur and tibia were mounted to the turntable and rotations recorded with HSBV. The rFHA was calculated for the bones using each smoothing technique and ranked using a Friedman test. The relative percent change to the unsmoothed data was reported. A spline filter with outlier removal (SPOUT) was ranked the best technique, producing the most accurate RSDs for the femur (-79.64%) and tibia (-70.59%). SPOUT was the top performing smoothing technique. Further investigations using SPOUT are required for in-vivo human movements.

Adaptive deep learning-based remaining useful life prediction framework for systems with multiple failure patterns

Recent advances in multivariate data fusion technology have promoted the applications of neural network-based models for remaining useful life (RUL) prediction. However, the interpretability of these models is usually poor since they are developed in a black-box manner. It is difficult to use them in engineering systems with multiple failure modes (FMs) under various operation conditions (OCs). This work proposes an adaptive deep learning-based RUL prediction framework with FM recognition. First, a FM recognizer fusing physics-informed FM classifier with deep convolutional neural networks (DCNN) is developed, which improves the interpretability and the accuracy of the recognition model. Then, a framework which can adaptively train models and select them for RUL prediction according to FM recognition results is presented. An OC-based smoothing technique is proposed to improve the RUL prediction accuracy and robustness. Extensive experiments based on turbofan datasets are conducted to validate the effectiveness of the proposed framework. The results show that the RUL prediction accuracy is improved by 7% under the proposed framework when compared with other methods. It proves the performance gains of the proposed framework by incorporating prior FM recognition with RUL prediction. It also provides insights for RUL prognostics subject to distinct FMs and OCs.

Multiscale feature cross‐layer fusion remote sensing target detection method

AbstractTarget detection based on remotely sensed images, which has recently attracted much attention, is a fundamental but challenging task. In remote sensing images, the problem of difficult recognition of small targets or targets with a large aspect ratio arises because the targets have the characteristics of small proportion, dense distribution, and multidirectionality. To address the above problems, this article proposes an improved multiscale feature cross‐layer fusion remote sensing target detector based on YOLOv5. First, this method introduces the circular smooth label technique, using YOLOv5 as a rotation detector to solve the angular boundary condition and angle prediction problem for large aspect ratio targets. Second, the explicit visual centre module is introduced to solve the problem of missed detection in target‐dense distribution tasks. Finally, a multiscale feature cross‐layer fusion structure (S‐160) is proposed based on YOLOv5, which improves the detection accuracy of each scale target by fusing shallow and deep feature information and introduces new large‐scale features for small target detection to solve the problem that ultrasmall targets in remote sensing images cannot be recognised. Our experiments were conducted on three public remote sensing datasets, DOTA, DIOR‐R, and HRSC2016, and the average accuracy (mAP) on the datasets was 76.50%, 70.34%, and 97.68%, respectively, demonstrating the substantial detection performance of the proposed method.

Forecasting Stock Prices on the Basis of Technical Analysis in the Industrial Sectors of the UK Stock Market

This study aims to evaluate, critically and rigorously the weak-form market efficiency and forecasting power of technical analysis in different industries in the London Stock Exchange. Weekly data were collected from the FTSE-all share index, FTSE-350 general industrial index and twenty companies of four different industries, for the period between 1997 and 2017. Bai-Perron’s multiple breaks test was applied to diagnose plain data period for the purpose of forecasting. The statistical inference was made from the application of the runs test, variance ratio tests, Ljung-Box’s test and ADF-unit root test that the market is not weak-form efficient and stock prices are predictable. This study extends the current literature by considering the existence of weak-form inefficiency in different industrial sectors. The findings do not support for weak-form efficiency over the periods tested from the application of the ARIMA and GARCH (1, 1) models and double and triple exponential smoothing techniques

Noise Reduction of Low-Count STEM-EDX Data by Low-Rank Regularized Spectral Smoothing.

Statistically weighted principal component analysis (wPCA) is widely used to reduce the noise of scanning transmission electron microscopy-energy-dispersive X-ray (STEM-EDX) spectroscopic data. It is beneficial to retain the spatial resolution of observation in each step of the analysis, but the direct application of wPCA without preprocessing, such as spatial averaging, often fails against low-count STEM-EDX data. To enhance the applicability of wPCA while retaining spatial resolution, a step-by-step noise reduction method is considered in this study. Specifically, a numerical optimization is developed to simultaneously characterize the smoothness of EDX spectra and the low rankness of the data. In the presented approach, low-count data are first spectrally smoothed by solving this optimization problem, and then further denoised by using wPCA to project onto a subspace rigorously spanned by a small number of components. A challenging example is provided, and the improved noise reduction performance is demonstrated and compared to existing spectral smoothing techniques.

Forecasting daily admissions to an emergency department considering single and multiple seasonal patterns

When dealing with several years of daily data, such as the number of daily admissions to a hospital’s emergency department (ED), how complex does it get to forecast into the future? With that in mind, this study has two main goals: to explore the differences between several methodologies, considering both single and multiple-seasonal patterns; and to select the most suitable model for the administration of a Portuguese hospital to use while managing their ED. To that end, we first considered the data as a time series with a single weekly seasonal pattern. We then modelled the data using time series regression, linear regression with autoregressive integrated moving average (ARIMA) errors, seasonal ARIMA and exponential smoothing techniques. Second, the data was set to be a time series with weekly and annual seasonal patterns. Then, using Fourier terms, we applied time series regression, linear regression with ARIMA errors and trigonometric exponential smoothing state space models with Box–Cox transformation, ARMA errors, Trend and Seasonal components (TBATS) for the analysis. After selecting the best-fitting models using the Akaike Information Criteria (AIC) values, we forecasted into the future and compared the results using both training and test datasets’ root mean square error (RMSE), mean absolute error (MAE) and mean absolute percentage error (MAPE) values. The time series regression model based on seasonal variables and a weekly seasonal pattern gives the best results. However, we decided to use linear regression with ARIMA errors, seasonal variables, and both weekly and annual seasonal patterns. This produces similar results but allows for the annual seasonality to be considered, which is useful when more data is added.

Multi-scored sleep databases: how to exploit the multiple-labels in automated sleep scoring.

Inter-scorer variability in scoring polysomnograms is a well-known problem. Most of the existing automated sleep scoring systems are trained using labels annotated by a single-scorer, whose subjective evaluation is transferred to the model. When annotations from two or more scorers are available, the scoring models are usually trained on the scorer consensus. The averaged scorer's subjectivity is transferred into the model, losing information about the internal variability among different scorers. In this study, we aim to insert the multiple-knowledge of the different physicians into the training procedure. The goal is to optimize a model training, exploiting the full information that can be extracted from the consensus of a group of scorers. We train two lightweight deep learning-based models on three different multi-scored databases. We exploit the label smoothing technique together with a soft-consensus (LSSC) distribution to insert the multiple-knowledge in the training procedure of the model. We introduce the averaged cosine similarity metric (ACS) to quantify the similarity between the hypnodensity-graph generated by the models with-LSSC and the hypnodensity-graph generated by the scorer consensus. The performance of the models improves on all the databases when we train the models with our LSSC. We found an increase in ACS (up to 6.4%) between the hypnodensity-graph generated by the models trained with-LSSC and the hypnodensity-graph generated by the consensus. Our approach definitely enables a model to better adapt to the consensus of the group of scorers. Future work will focus on further investigations on different scoring architectures and hopefully large-scale-heterogeneous multi-scored datasets.

Adaptative finite element simulation of unsteady turbulent compressible flows on unstructured meshes

The simulation of high-speed turbulent compressible flows using Large Eddy Simulation (LES) combined with the Characteristic-Based Split scheme (CBS) and a mesh adaptation strategy is presented in this work. A compressible dynamic Smagorinsky model is employed for the compressible LES model and the CBS scheme is employed in a Finite Element Method (FEM) context for space and time discretization using unstructured meshes. An accurate mesh adaptation scheme is frequently applied along the simulation including mesh refinement, mesh coarsening, edge or face swapping and a vertex smoothing technique in order to control the interpolation error and the mesh size. A feature-based error estimation is employed considering multi-scale and multi-metric intersection in order to capture turbulence effects. Two and three-dimensional applications are evaluated in order to investigate the capabilities of the present approach for evaluating external turbulent and compressible aerodynamics problems. The turbulent transonic flow around a circular cylinder in 2D and 3D domains and the turbulent supersonic flow past a blunt cylinder-flare model are investigated. Aerodynamic coefficients, frequency analysis and pressure distributions are compared against experimental data and other numerical simulations.

2D Ti3C2Tx MXene-derived self-assembled 3D TiO2nanoflowers for nonvolatile memory and synaptic learning applications

Two-dimensional (2D) semiconducting materials and transition-metal oxides are promising materials for nonvolatile memory and brain-inspired neuromorphic computing applications. However, it remains challenging to obtain high-quality stacked 2D films with low energy consumptions (or drive currents) because of their high interfacial resistance. In this study, we synthesized 2D Ti3C2Tx MXene-derived three-dimensional (3D) TiO2 nanoflowers (NFs) as a feasible resistive switching (RS) material with outstanding electronic properties and synaptic learning capabilities. The electrical and optical characteristics of the synthesized material were determined through density functional theory calculations. Electrical measurements of the Al/Ti3C2Tx-TiO2 NF/Pt memory device indicated the occurrence of forming-free switching phenomena with extremely low switching voltages (0.68–0.53 V), stable ON/OFF ratio (2.3 × 103), and retention greater than 105 s. The Holt–Winters exponential smoothing technique was used for modeling and predicting the switching voltages of the RS device. The mechanism underlying the reliable RS was confirmed by observing the dense conductive filaments through conductive atomic force microscopy. Interestingly, the 2D Ti3C2Tx MXene-derived 3D TiO2 NF-based RS device mimicked the potentiation/depression and spike-time-dependent plasticity of a biological synapse. Finally, a convolutional neural network was implemented based on the observed synaptic weights of Al/Ti3C2Tx-TiO2 NF/Pt for image-edge detection.

Detection-estimation strategies for delayed systems with unavailable packet sequence

In this paper, we study the problem of remote state estimation on networks with random delays and unavailable packet sequence due to malicious attacks. Two maximum a posteriori (MAP) schemes are proposed to detect the unavailable packet sequence. The first MAP strategy detects the packet sequence using data within a finite time horizon; the second MAP strategy detects the packet sequence by a recursive structure, which effectively reduces the computation time. With the detected packet sequence, we further design a linear minimum mean-squared error (LMMSE) estimation algorithm based on smoothing techniques, rather than using the classic prediction and update structure. A wealth of information contained in the combined measurements is utilized to improve the estimation performance. Finally, the effectiveness of the proposed algorithms is demonstrated by simulation experiments.

Kinetically driven thermal roughening of semiconductor surfaces: experiment on GaAs and Monte Carlo simulation

We present the results of GaAs annealing experiments in the conditions near equilibrium, which clarify the reasons of the transition from surface smoothing to roughening at temperatures above 650 °C. The roughening is due to kinetic instabilities arising under deviation of annealing conditions towards growth or sublimation. These instabilities reveal themselves in appearing islands (for sublimation) and pits (for growth) of multilayer heights and depths, respectively. The islands and pits appear due to the motion of atomic steps through surface spots, at which sublimation and growth are suppressed. Pinning of the steps at these spots also lead to step bunching at surfaces with sufficiently small terrace widths. This explanation is consistent with Monte Carlo simulations of atomic processes on the GaAs surface. The similarity and distinctions in surface roughening under sublimation and growth, along with the role of Schwöbel barrier, are discussed. Annealing experiments in the cavities, which are formed by GaAs substrates with well-controlled atomic step densities enabled us to clarify the roughening mechanisms and to improve the efficiency of GaAs thermal smoothing technique by increasing smoothing temperature up to 775 °C.

Software development to standardize the clinical diagnosis of exercise oscillatory ventilation in heart failure.

Exercise oscillatory ventilation (EOV) is characterized by periodic oscillations of minute ventilation during cardiopulmonary exercise testing (CPET). Despite its prognostic value in chronic heart failure (HF), its diagnosis is complex due to technical limitations. An easier and more accurate way of EOV identification can contribute to a better approach and clinical diagnosis. This study aims to describe a software development to standardize the EOV diagnosis from CPET's raw data in heart failure patients and test its reliability (intra- and inter-rater). The software was developed in the "drag-and-drop" G-language using LabVIEW®. Five EOV definitions (Ben-Dov, Corrà, Kremser, Leite, and Sun definitions), two alternative approaches, one smoothing technique, and some basic statistics were incorporated into the interface to visualize four charts of the ventilatory response. EOV identification was based on a set of criteria verified from the interaction between amplitude, cycle length, and oscillation time. Two raters analyzed the datasets. In addition, repeated measurements were verified after six months using about 25% of the initial data. Cohen's kappa coefficient (κ) was used to investigate the reliability. Overall, 391 tests were analyzed in duplicate (inter-rater reliability) and 100 tests were randomized for new analysis (intra-rater reliability). High inter-rater (κ > 0.80) and intra-rater (κ > 0.80) reliability of the five EOV diagnoses were observed. The present study proposes novel semi-automated software to detect EOV in HF, with high inter and intra-rater agreements. The software project and its tutorial are freely available for download.

Simplified Spatial Smoothing for DOA Estimation of Coherent Signals

In coherent direction of arrival (DOA) estimation, subspace-based methods suffer from performance deterioration because of the rank loss of the signal covariance matrix. A variety of spatial smoothing preprocessing techniques have been proposed for decorrelation, among which the enhanced spatial smoothing preprocessing (ESS) technique shows outstanding performance by exploiting the signal subspace. However, ESS is applied by squaring the matrix whose columns span the signal subspace (called signal matrix), which involves unnecessary computational loads. Besides, the smoothed covariance matrix after ESS is a linear combination of the sub-matrices of signal matrix, where the coefficients of the combination may undermine the decorrelation performance. In this context, a simplified spatial smoothing (SSS) technique is proposed for decorrelation by averaging the sub-matrices of the signal matrix directly, and avoids redundant operations in squared signal matrix. The proposed method is tested numerically in terms of the signal-to-noise ratio (SNR), the number of snapshots, angle separation, and the execution time. Simulation results show the improvement of the decorrelation performance, efficiency, and robustness with the proposed method in coherent scenarios, compared with the other spatial smoothing preprocessing based methods.