A novel augmented Kalman filter with mixed norm constraint for impact force identification

Current deep learning approaches to linear prediction coefficient (LPC) estimation for the augmented Kalman filter (AKF) produce bias estimates, due to the use of a whitening filter. This severely degrades the perceived quality and intelligibility of enhanced speech produced by the AKF. In this paper, we propose a deep learning framework that produces clean speech and noise LPC estimates with significantly less bias than previous methods, by avoiding the use of a whitening filter. The proposed framework, called DeepLPC, jointly estimates the clean speech and noise LPC power spectra. The estimated clean speech and noise LPC power spectra are passed through the inverse Fourier transform to form autocorrelation matrices, which are then solved by the Levinson-Durbin recursion to form the LPCs and prediction error variances of the speech and noise for the AKF. The performance of DeepLPC is evaluated on the NOIZEUS and DEMAND Voice Bank datasets using subjective AB listening tests, as well as seven different objective measures (CSIG, CBAK, COVL, PESQ, STOI, SegSNR, and SI-SDR). DeepLPC is compared to six existing deep learning-based methods. Compared to other deep learning approaches to clean speech LPC estimation, DeepLPC produces a lower spectral distortion (SD) level than existing methods, confirming that it exhibits less bias. DeepLPC also produced higher objective scores than any of the competing methods (with an improvement of 0.11 for CSIG, 0.15 for CBAK, 0.14 for COVL, 0.13 for PESQ, 2.66% for STOI, 1.11 dB for SegSNR, and 1.05 dB for SI-SDR over the next best method). The enhanced speech produced by DeepLPC was also the most preferred by 10 listeners. By producing less biased clean speech and noise LPC estimates, DeepLPC enables the AKF to produce enhanced speech at a higher quality and intelligibility.

The performance of the Kalman Filter (KF) technique is most appropriate when both the process and observation are considered as white Gaussian noise. But in real time applications, this assumption is not always true. Two sorts of Kalman Filtering techniques i.e Augmented Kalman Filtering and Second Moment Information bases Kalman Filtering with colored system noise i.e (process noise and observation noise) are examined in this paper. Augmented Kalman Filtering (AKF) with colored measurement noise is shortly analyzed. Modified Kalman Filtering method (SMIKF) is proposed for colored process and observation noise. In order to check the performance of the system a real-time Mass-Spring Damper system having colored system noises is considered. The computational burdens of the proposed scheme is compared with AKF. The simulation results endorse the Performance of the suggested algorithm.

Multiview image registration is widely used in computer vision applications. In this paper, an approach for multiview 3D image registration based on augmented Kalman filter is proposed by taking account of various uncertainties. The position and orientation of viewpoint are considered as system state. System augmentation model and system observation model are constructed. The position and orientation of each viewpoint is augmented and updated recursively. The global transformation parameters of image are computed with the state estimation of corresponding viewpoint. The proposed multiview image registration method can handle the uncertainty efficiently and the registration result is accurate and globally consistent. Some experimental results are provided to validate the performance of the proposed method.

Knowledge of the structural excitations induced in bridges due to traffic loading is an important component of structural health monitoring. However, direct measurement of such traffic-induced structural excitations in real-world structures is extremely challenging. While model-based methods provide indirect ways to estimate structural excitation, uncertainties associated with modelling have not been effectively considered in currently available methods for the identification of traffic-induced excitations. An optimised state estimation method that can accurately identify such traffic-induced structural excitations under uncertainties is proposed. The method uses an augmented Kalman filter (AKF) and a genetic algorithm (GA). However, the selection of error covariance values on the model, measurement and excitations is a critical challenge when using the AKF, especially for cases when excitations are spatially distributed over a large number of locations. The proposed method addresses such issues using GA-based optimisation with the objective of minimising the estimation error between measured and estimated excitations. Furthermore, heterogeneous structural measurements are used in the identification of excitations to improve the accuracy and stability of the process.

Traffic loads on the highway bridges have been increasing in the past years. Actual truck axle configurations and weights may deviate significantly from standard legal design trucks depending on States, freight corridors, and seasons. Changes in legal truckloads are often made by state legislatures based on the pressure by the trucking industry. It is important to understand how trucks impact bridges and whether they cause deterioration and degradation so that policy decisions on legal truckloads can be based on scientific facts. The information of the actual passing vehicles that would be required to address such concerns includes actual truck axle loads and configurations, speeds, and dynamic effects. The critical questions are: How can we measure truck axle configurations and weights over the long term? How can we identify and measure the corresponding critical bridge responses over the long term, and how can we relate these to changes in the bridge life cycle? Finding objective answers to these questions requires complex research. This study is a first step through exploring bridge weigh-in-motion (B-WIM) algorithms, which help to capture the truck's information for instrumented bridges in operations. Several B-WIM algorithms were inspected first and classified based on their assumptions in incorporating dynamic bridge vehicle interaction. For instance, B-WIM algorithms can be classified as static and dynamic. For the static case, the dynamic effect caused by the interaction between vehicle and bridge is mostly ignored, which results in significant inaccuracies when the dynamic component of the coupled systems is not insignificant or challenging to filter out. Therefore, this study mainly focuses on dynamic algorithms. Various dynamic algorithms, such as dynamic programming and augmented Kalman filters (AKF), have been investigated for their challenges and opportunities they offer into B-WIM applications. Various techniques were employed to address the potential deficiency of dynamic algorithms. These techniques include modal superposition technique to deal with complex structures, mode truncation, singular value decomposition to mitigate the singularity of the formulation matrices, and dummy measurements to improve the identification process's observability. The influence of different parameters on the performance of AKF based B-WIM algorithms were investigated via a parametric study of a simply supported beam subject to moving forces. The results show that those parameter has been carefully chosen for the success of the force prediction. A B-WIM framework was developed. The framework is composed of all the components required to estimate the vehicular weights all the way up from conceptualizations. Critical ones include structural identification to prepare the digital twin, eigenvalue reduction to reduce the dimension of the problem, the numerically verified AKF algorithm, and a parameter tuning method based on optimization technique. A scale model inspired by a typical highway bridge span was designed and constructed in the laboratory following the Similitude Theory. It aims to demonstrate the capability of the proposed B-WIM framework. Multiple tests with different vehicular speeds, weights, and moving paths were conducted and analyzed. Complete force-time histories were predicted and discussed. For B-WIM, the vehicular weights were calculated by averaging the force-time history segment where the vehicle was entirely on the bridge, and results show good agreement between the predicted forces and the static weights.

Improvement of condition monitoring (CM) systems for wind turbines (WTs) and reduction of the cost of wind energy are possible if knowledge about the condition of different WT components is available. CM based on the WT drive train shaft torque signal can give a better understanding of the gearbox failure mechanisms as well as provide a method for detecting mass imbalance and aerodynamic asymmetry. The major obstacle preventing the industrial application of CM based on the shaft torque signal is the costly measurement equipment which is impractical for long‐term use on operating WTs. This paper suggests a novel approach for low‐cost, indirect monitoring of the shaft torque from standard WT measurements. The shaft torque is estimated recursively from measurements of generator torque, high speed shaft and low speed shaft angular speeds using the well‐known Kalman filter theory. The performance of the augmented Kalman filter with fading memory (AKFF) is compared with the augmented Kalman filter (AKF) using simulated data of the WT for different load conditions, measurement noise levels and WT fault scenarios. A multiple‐model algorithm, based on a set of different Kalman filters, is designed for practical implementation of the shaft torque estimator. Its performance is validated for a scenario where there are frequent changes of operating points. The proposed cost‐effective shaft torque estimator overcomes all major problems, which prevent the industrial application of CM systems based on shaft torque measurements. Future work will be focused on validating the method using experimental data and developing suitable signal processing algorithms for fault detection. Copyright © 2013 John Wiley & Sons, Ltd.

The performance of Kalman filters (KF) in speech enhancement is highly dependent on the estimation accuracy of linear predictive coefficients (LPCs). To address the significant estimation bias of LPCs in complex noisy environments, this paper proposes a time-frequency dual-domain Conformer-based Kalman filter speech enhancement method—CAKFNet. The model adopts an encoder-decoder architecture: during the encoding phase, a Dilated DenseNet module is introduced to exponentially expand the receptive field and enhance local feature continuity, thereby achieving multi-scale feature aggregation. Subsequently, dual Conformer modules in both time and frequency domains jointly model the long-term dependencies of speech signals and local convolutional features, fully capturing spatio-temporal correlations. The decoding stage employs a decoupled dual-branch architecture to separately regress the LPC power spectrum (LPC-PS) of clean speech and noise. Building upon this, an autocorrelation matrix is generated via inverse Fourier transform. The Levinson-Durbin recursive algorithm is then employed to compute the clean speech and noise LPC coefficients required for the Augmented Kalman Filter (AKF), along with their prediction error variance, ultimately achieving speech enhancement. The proposed method undergoes systematic experimental evaluation on the publicly available VoiceBank + DEMAND dataset. Results demonstrate that CAKFNet effectively reduces spectral distortion in LPC estimation. Objective metrics, including PESQ and STOI, show significant improvements over baseline methods, while subjective listening tests confirm substantial enhancements in speech quality and intelligibility.

This paper addresses the issue of the correction of registration errors due to the presence of biases in the measurements when multiple sensors are used for tracking targets. The problem is formulated as one of recursively estimating biases, and is addressed by an application of the technique, using an augmented Kalman filter, of simultaneously estimating the sensor biases and the state vector (of track parameters), the latter assuming that no biases are present. Simulation results are presented comparing the performance of the recursive algorithm with that of a generalized least squares method, which is applicable when targets with known positions are available for measurement by the sensors. Results are also presented for a two-sensor case when the extended Kalman filter is used as the tracking filter.

Real-time results of a receding horizon state-space controller, applied to a laboratory plant with hybrid characteristics, are reported in This work. The control strategy is based on a multi-model scheme and the corresponding switching logic. The control law involves a state-space receding horizon tracking system, a constant-disturbances compensator and a state and constant-disturbances estimator based on an augmented Kalman filter. The pilot plant consisting of three tanks, a pump and several valves exhibits nonlinear and hybrid behavior. Results are satisfactory and better than the results obtained for a unique controller.

Multivariable generalized state-space receding horizon control in a real-time environment

A data-cleaning augmented Kalman filter for robust estimation of state space models

The inverse dynamic analysis of vehicle-bridge (VB) systems provides a superior scheme for track irregularity identification using the onboard measurement data of vehicle dynamic responses. However, available algorithms face a great challenge when applied to high-speed railway bridges because of the increase in the number of unknown inputs, resulting in a degradation of identification accuracy. Here, we propose a condensation procedure to reduce the unknown inputs of the VB system, and then propose a novel algorithm for track irregularity identification using the augmented Kalman filter. A numerical simulation of a real continuous three-span simply supported railway bridge was conducted to validate the estimation accuracy of the proposed algorithm. Numerical results illustrate that the proposed algorithm can accurately estimate the track irregularities by using the vertical acceleration responses of the vehicle body only, thanks to the unknown input condensation. A comparative study demonstrates that the proposed algorithm outperforms the conventional track irregularity identification methods in terms of estimation accuracy. To evaluate the actual performance of the proposed algorithm, a set of onboard measurement data recorded by a typical high-speed train when running on the Hangzhou–Changsha high-speed railway in China was employed to identify the track irregularities of the railway line, and the results suggest that the proposed algorithm is highly efficient and low in cost. This study may help develop highly efficient real-time track irregularity monitoring in the years ahead.

This letter presents a new method for identification of fast-fading mobile channels (for which combinations of vehicle speed and carrier frequency give rise to significant fading). Our new algorithm estimates both the channel statistics and the time-varying channel impulse response on-line. Simulation studies demonstrate the performance of the new estimator which couples an augmented Kalman filter with a recursive least squares algorithm.

Various applications in structural dynamics may require the real-time estimation of unknown input. A recently developed joint input-state estimator for linear systems treats the unknown input as white Gaussian noise with finite covariance. The performance of this finite input covariance (FIC) estimator is validated using simulated data from a finite element model of the Air Force Research Laboratory’s experimental testbed called the DROPBEAR (Dynamic Reproduction of Projectiles in Ballistic Environments for Advanced Research). The estimator performance is compared with a few well-known estimators, including the augmented Kalman filter (AKF) and the weighted least squares (WLS) estimators. The results show that the FIC estimator is capable of accurately estimating an impact load applied to the beam when acceleration is measured at a small number of locations. Additionally, the results show that the FIC estimator eliminates the low-frequency drift error that other well-known estimators are susceptible to.KeywordsInput estimationDynamic testingLinear stochastic systemImpact load identification

An interacting Fuzzy-Fading-Memory-based Augmented Kalman Filtering method for maneuvering target tracking