System Identification Performance Research Articles

System Identification and Seismic Performance Assessment of a Stone Arch Bridge

ABSTRACT This paper presents system identification and seismic performance assessment of a masonry arch bridge located on the railway route which is on the northeastern part of Turkey. Forty-one masonry arch bridges were registered as historical and needed to be preserved on the route. On the other hand, the railway line passes through North Anatolian Fault, resulting in high seismic demand on bridges. Therefore, seismic assessment of the bridges was carried out by finite element analysis; however, masonry structures such as stone arch bridges have significant uncertainties in terms of material properties, boundary conditions, and modeling assumptions. As a result, it becomes almost unavoidable to perform dynamic identification tests to validate finite element models (FEM). Modal properties of 12 bridges such as modal frequencies, mode shapes, and modal damping ratios were identified through vibration measurements collected under ambient conditions, impact loading, and train passage. Based on identified modal parameters, FEM of the bridges were updated to obtain actual values of Young’s modulus of masonry and soil. FEM updating procedure was performed by minimizing the difference between experimental and analytical modal properties. Afterwards, seismic performance assessment of a representative bridge was carried out using ANSYS software. Two different nonlinear approaches were considered, namely, pushover and nonlinear time history (NLTH) analyses. In the analyses, macro-modeling approach was followed to develop homogenized behavior of stone and mortar. It was found that mass proportional pushover analysis gave similar results to NLTH in terms of seismic performance. The other significant point was that modeling soil medium reflects the actual behavior of the bridge in the sense that artificial stress concentrations do not occur at the pier base. It was also observed that tensile strength capacity was reached on spandrel walls which may result in a probable failure.

Generalized Correntropy Induced Metric Memory-Improved Proportionate Affine Projection Sign Algorithm and Its Combination

In this brief, an efficient memory-improved proportionate affine projection sign algorithm with the generalized correntropy induced metric (GCI-M-IP-APSA) is proposed to modify the filtering performance for sparse system identification. In addition, a simplified implementation of GCI-M-IP-APSA is derived, and called as SGCI-M-IP-APSA, which has lower computational complexity and realizes comparable filtering performances with that of GCI-M-IP-APSA. However, the major limitation of proposed APSA-type algorithm is a poor convergence behavior in non-stationary situations. To overcome such issue, a convex combination method is applied to SGCI-M-IP-APSA to enhance the tracking performance. The good performances of proposed algorithms are verified in sparse system identifications with impulsive noise and abrupt changes.

A bias‐compensated proportionate NLMS algorithm with noisy input signals

SummaryThis paper proposes a novel proportionate normalized least‐mean‐squares (PNLMS) algorithm that is robust to input noises. Through compensating for biases due to input noise added at the filter input, the proposed PNLMS algorithm avoids performance deterioration owing to the noisy input signals. Moreover, since the proposed PNLMS algorithm uses a new gain‐distribution matrix, it has a fast convergence rate compared with the existing PNLMS algorithms, even when there is no input noise. The experimental results verify that the proposed PNLMS algorithm enhances the filter performance for sparse system identification in the presence of input noises.

Robust Variable Step-Size Reweighted Zero-Attracting Least Mean M-Estimate Algorithm for Sparse System Identification

The reweighted zero-attracting least mean square (RZA-LMS) algorithm has good performance in sparse system identification. However, the convergence performance of the RZA-LMS algorithm degrades in the impulsive noise environment. In this brief, a reweighted zero-attracting least mean M-estimate (RZA-LMM) algorithm is proposed to address the issue. The RZA-LMM algorithm employs the M-estimate cost function with a sparsity-aware penalty term and is derived by using the unconstrained minimization method. In order to further enhance the convergence speed and reduce the steady-state error of the RZA-LMM algorithm, a variable step-size reweighted zero-attracting least mean M-estimate (VSSRZA-LMM) algorithm is proposed. The variable step-size scheme, which adopts the M-estimate cost function with a sparsity-aware penalty term and a constrained term of the step-size, is obtained by using the constrained minimization method. Besides, the mean-square stability is analyzed to obtain the range of the step-size, which guarantees the stabilization of the proposed algorithms. The computer simulation results show that the proposed RZA-LMM algorithm achieves better convergence performance than the RZA-LMS algorithm, and the proposed VSSRZA-LMM algorithm exhibits better convergence performance in terms of convergence speed and steady-state normalized mean square deviation (NMSD) than the RZA-LMM algorithm for sparse system identification in the impulsive noise environment.

Experimental Verification of a Novel Quad-RLS Technique for Improving Real-time System Identification Performance: A Practical Approach to CMG

This paper presents a Quad RLS(Q-RLS) technique with four RLSs under a fixed forgetting factor condition to improve the identification performance of a dynamical system in a real-time fashion. Although an adaptive RLS method with a variable forgetting factor and a higher order model may provide the modeling accuracy, their implementations are not easy because of leakage effects and relatively complex modeling. In practice, the fixed forgetting factor is still used for the RLS-based system identification. Therefore, a novel Q-RLS scheme with concurrent four RLSs is proposed as an alternative way for the better estimation performance in an on-line fashion. In the Q-RLS scheme, the first pair of the forward and inverse RLSs is to identify the forward and inverse models independently. The second pair of the forward and inverse RLSs is to improve the identification of the previously identified model. The proposed approach has several advantages: 1) The RLS with a fixed forgetting factor can avoid the leakage problem. 2) Both forward and inverse models are separately identified to improve the accuracy. 3) Q-RLS can have the 4th order filter structure, but provide the better identification performance. Three schemes such as a second-order RLS, a fourth-order RLS, and the Q-RLS are experimentally tested and their performances are compared for the state observation accuracy of the control moment gyroscope(CMG) system.

Extended-Kalman-filter-based dynamic mode decomposition for simultaneous system identification and denoising

A new dynamic mode decomposition (DMD) method is introduced for simultaneous system identification and denoising in conjunction with the adoption of an extended Kalman filter algorithm. The present paper explains the extended-Kalman-filter-based DMD (EKFDMD) algorithm which is an online algorithm for dataset for a small number of degree of freedom (DoF). It also illustrates that EKFDMD requires significant numerical resources for many-degree-of-freedom (many-DoF) problems and that the combination with truncated proper orthogonal decomposition (trPOD) helps us to apply the EKFDMD algorithm to many-DoF problems, though it prevents the algorithm from being fully online. The numerical experiments of a noisy dataset with a small number of DoFs illustrate that EKFDMD can estimate eigenvalues better than or as well as the existing algorithms, whereas EKFDMD can also denoise the original dataset online. In particular, EKFDMD performs better than existing algorithms for the case in which system noise is present. The EKFDMD with trPOD, which unfortunately is not fully online, can be successfully applied to many-DoF problems, including a fluid-problem example, and the results reveal the superior performance of system identification and denoising.

Modelling of Horizontal Flexible Plate Structure Using Artificial Bee Colony Algorithm

This paper presents the performance of system identification for modeling the horizontal flexible plate system using artificial bee colony and recursive least square algorithms. Initially, the experimental rig of flexible plate was designed and fabricated with all edges clamped boundary condition at the horizontal position. Then, the instrumentation and data acquisition systems were integrated into the rig for acquiring the input-output vibration experimentally. The collected data in the experiment will be used later for modeling the dynamic system of horizontal flexible plate system using system identification. The effectiveness of the developed model will be validated using mean squared error, one step ahead prediction, correlation tests and pole zero diagram stability. The estimated of the developed models were found are acceptable and possible to be used as a platform of controller development for vibration suppression of the undesirable vibration in the flexible plate structure. It was found that the artificial bee colony algorithm has performed better in this study by achieving the lowest mean squared error, good correlation test and high stability in the pole zero diagram.

A Novel Approach for Modeling Neural Responses to Joint Perturbations Using the NARMAX Method and a Hierarchical Neural Network.

The human nervous system is an ensemble of connected neuronal networks. Modeling and system identification of the human nervous system helps us understand how the brain processes sensory input and controls responses at the systems level. This study aims to propose an advanced approach based on a hierarchical neural network and non-linear system identification method to model neural activity in the nervous system in response to an external somatosensory input. The proposed approach incorporates basic concepts of Non-linear AutoRegressive Moving Average Model with eXogenous input (NARMAX) and neural network to acknowledge non-linear closed-loop neural interactions. Different from the commonly used polynomial NARMAX method, the proposed approach replaced the polynomial non-linear terms with a hierarchical neural network. The hierarchical neural network is built based on known neuroanatomical connections and corresponding transmission delays in neural pathways. The proposed method is applied to an experimental dataset, where cortical activities from ten young able-bodied individuals are extracted from electroencephalographic signals while applying mechanical perturbations to their wrist joint. The results yielded by the proposed method were compared with those obtained by the polynomial NARMAX and Volterra methods, evaluated by the variance accounted for (VAF). Both the proposed and polynomial NARMAX methods yielded much better modeling results than the Volterra model. Furthermore, the proposed method modeled cortical responded with a mean VAF of 69.35% for a three-step ahead prediction, which is significantly better than the VAF from a polynomial NARMAX model (mean VAF 47.09%). This study provides a novel approach for precise modeling of cortical responses to sensory input. The results indicate that the incorporation of knowledge of neuroanatomical connections in building a realistic model greatly improves the performance of system identification of the human nervous system.

Linear phase sparse system identification in the presence of impulsive noise

ABSTRACTIn this paper, we incorporate an ℓp-norm sparsity constraint into linear phase constrained affine projection-sign algorithm (LPCAP-SA). The proposed algorithm combines the effect of linear phase constrained affine projection sign algorithm and ℓp-norm sparsity constraint .The proposed algorithm has better performance for linear phase sparse system identification in the presence of vigorous impulsive interference. The proposed ℓp-norm LPCAP-SA is based on the ℓ1-norm minimisation criterion on posterior error and ℓp-Norm on weight coefficients and a linear phase constraint. Simulations in Matlab software show that the proposed ℓp-norm LPCAP-SA has robustness against impulsive interferences and sparse environment for linear phase system.

Modifying the ERA and fast ERA to improve operational performance for structural system identification

Abstract The eigensystem realization algorithm (ERA) and fast ERA are two popular methods of structural-system identification; however, these two powerful methods have deficiencies when employed in real-life identification projects. The main deficiencies of the ERA and fast ERA methods are that, when the order of the structural system is not known or is infinite, the sample time is not appropriate and/or initial condition is not zero, the identification results using these algorithms are not correct. This paper introduces a theoretical foundation called the modified ERA (mERA) and modified fast ERA (mfERA) to tackle these deficiencies. The simulation and experimental results are presented to demonstrate the validity of these analytical developments.

Identification of Nonlinear State-Space Systems From Heterogeneous Datasets

This paper proposes a new method to identify nonlinear state-space systems from heterogeneous datasets. The method is described in the context of identifying biochemical/gene networks (i.e., identifying both reaction dynamics and kinetic parameters) from experimental data. Simultaneous integration of various datasets has the potential to yield better performance for system identification. Data collected experimentally typically vary depending on the specific experimental setup and conditions. Typically, heterogeneous data are obtained experimentally through 1) replicate measurements from the same biological system or 2) application of different experimental conditions such as changes/perturbations in biological inductions, temperature, gene knock-out, gene over-expression, etc. We formulate here the identification problem using a Bayesian learning framework that makes use of “sparse group” priors to allow inference of the sparsest model that can explain the whole set of observed heterogeneous data. To enable scale up to large number of features, the resulting nonconvex optimization problem is relaxed to a reweighted Group Lasso problem using a convex–concave procedure. As an illustrative example of the effectiveness of our method, we use it to identify a genetic oscillator (generalized eight species repressilator). Through this example we show that our algorithm outperforms Group Lasso when the number of experiments is increased, even when each single time-series dataset is short. We additionally assess the robustness of our algorithm against noise by varying the intensity of process noise and measurement noise.

Modified memory‐improved proportionate affine projection sign algorithm based on correntropy induced metric for sparse system identification

A novel memory-improved proportionate affine projection sign algorithm with correntropy induced metric (CIM-MIP-APSA) is proposed to enhance the filtering performance for sparse system identification. Furthermore, a simplified implementation of the CIM-MIP-APSA (SCIM-MIP-APSA) with low computational complexity is proposed while improving the convergence performance and filtering accuracy. Simulation results demonstrate that the CIM-MIP-APSA and SCIM-MIP-APSA algorithms outperform the existing works for sparse system identification.

Performance of Vector Fitting Algorithm Applied to Bandpass and Baseband Systems

This article presents the performance evaluation of Vector Fitting Algorithm (VFA) from a system identification perspective. In this paper, VFA has been first applied to known baseband and bandpass systems such as Butterworth lowpass and bandpass filters to analyze the algorithm’s pole-residue extraction ability for band-limited noisy data. The poles identified by the algorithm for different bandwidths and noise powers are compared with the actual system poles of the baseband and bandpass systems. It is concluded that the algorithm is capable of identifying the actual system poles even if the capture bandwidth is less than the 3 dB bandwidth, which is a significant observation of this paper. It is also seen that the system identification performance with noisy data is better for baseband systems when compared to bandpass systems. Further, a practical investigation has been done to evaluate VFA performance for modeling a microstrip coupled line filter in the presence of noise.

Robust Proportionate Adaptive Filtering Algorithms Against Impulsive Interference

The Proportionate Normalized Least Mean Square (PNLMS) algorithm, as a popular tool of signal processing, achieves excellent performance for sparse system identification. However, in previous studies, most of the cost functions used in proportionate-type sparse adaptive algorithms are based on the Mean Square Error (MSE) criterion, which is optimal only when the measurement noise is Gaussian. However, when the noise is impulsive noise, the performances of the existing proportionate-type sparse adaptive algorithms deteriorate severely. In this paper, two novel proportionate adaptive algorithms, namely, Proportionate Maximum Correntropy Criterion (PMCC) algorithm based on maximum correntropy criterion algorithm and proportionate normalized least mean square algorithm based on arctangent cost function (P-Arc-NLMS) are proposed to address this problem. The corresponding computational complexity is analyzed. The simulations are done to confirm that the proposed algorithms have advantages, compared with the existing other algorithms proposed against impulsive interference in sparse system identification.

An iterated cubature unscented Kalman filter for large-DoF systems identification with noisy data

Structural and mechanical system identification under dynamic loading has been an important research topic over the last three or four decades. Many Kalman-filtering-based approaches have been developed for linear and nonlinear systems. For example, to predict nonlinear systems, an unscented Kalman filter was applied. However, from extensive literature reviews, the unscented Kalman filter still showed weak performance on systems with large degrees of freedom. In this research, a modified unscented Kalman filter is proposed by integration of a cubature Kalman filter to improve the system identification performance of systems with large degrees of freedom. The novelty of this work lies on conjugating the unscented transform with the cubature integration concept to find a more accurate output from the transformation of the state vector and its related covariance matrix. To evaluate the proposed method, three different numerical models (i.e., the single degree-of-freedom Bouc–Wen model, the linear 3-degrees-of-freedom system, and the 10-degrees-of-freedom system) are investigated. To evaluate the robustness of the proposed method, high levels of noise in the measured response data are considered. The results show that the proposed method is significantly superior to the traditional UKF for noisy measured data in systems with large degrees of freedom.

Closed-loop identification for plants under model predictive control

Model predictive controllers incorporate step response models for pairings of independent and dependent variables. Motivated by the fact that it may be time-consuming to conduct open-loop experiments to identify the step response models, the paper assesses the performance of closed-loop system identification on MPC-equipped plants, using both simulated and actual plant data. Pure feedback closed-loop system identification is shown to be effective for an identifiable simulated system and an industrial hydrogen production plant. The use of closed-loop system identification as a mechanism for monitoring model quality in MPC implementations may enhance the long-term sustainability of the implementation.

A Novel Networked Online Recursive Identification Method for Multivariable Systems With Incomplete Measurement Information

Network-based identification of multivariable systems plays a key role in future smart manufacturing systems in achieving the goals of industry 4.0. The incomplete information caused by network traffic congestion or cyber attacks in the networked environment will inevitably deteriorate the performance of system identification, and in the extreme cases, it will cause nonconvergence of the identifier. Unlike the traditional recursive least-squares algorithms based on the complete data, this paper investigates a novel networked online recursive identification method for multivariable systems with incomplete information. In this new algorithm, the characteristics of data packet dropouts are first formulated as a Bernoulli process, and the lost data are compensated by an auxiliary model. A new information set including networked parameters is then constructed, and the corresponding networked online identification algorithm for multivariable systems is proposed. The proposed algorithm can overcome the negative effect of data packet losses on the identification performance and can be updated recursively. Furthermore, using the Lyapunov and martingale methods, the convergence rate of the proposed algorithm as well as its computational complexity is analyzed in detail. Simulation examples confirm the feasibility and efficiency of the proposed method.

FFT formulations of adaptive Fourier decomposition

Adaptive Fourier decomposition (AFD) has been found to be among the most effective greedy algorithms. AFD shows an outstanding performance in signal analysis and system identification. As compensation of effectiveness, the computation complexity is great, that is especially due to maximal selections of the parameters. In this paper, we explore the discretization of the 1-D AFD integration via with discrete Fourier transform (DFT), incorporating fast Fourier transform (FFT). We show that the new algorithm, called FFT-AFD, reduces the computational complexity from O ( M N 2 ) to O ( M N log N ) , the latter being the same as FFT. Through experiments, we verify the effectiveness, accuracy, and robustness of the proposed algorithm. The proposed FFT-based algorithm for AFD lays a foundation for its practical applications.

Estimating the unknown time delay in chemical processes

Although time delay is an important element in both system identification and control performance assessment, its computation remains elusive. This paper proposes the application of a least squares support vector machines driven approach to the problem of determining constant time delay for a chemical process. The approach consists of two steps, where in the first step the state of the system and its derivative are approximated based on the LS-SVM model. The second step consists of modeling the delay term and estimating the unknown model parameters as well as the time delay of the system. Therefore the proposed approach avoids integrating the given differential equation that can be computationally expensive. This time delay estimation method is applied to both simulation and experimental data obtained from a continuous, stirred, heated tank. The results show that the proposed method can provide accurate estimates even if significant noise or unmeasured additive disturbances are present.

System with Distributed Lag: Adaptive Identification and Prediction

Adaptive algorithms of parametric identifica- tion of discrete systems with lag variables are proposed. Adaptive algorithms (AA) in the presence of lag input variables are developed. The convergence of the AA and the boundedness of the trajectories the adaptive system is proved. Convergence domain АА depends on operating disturbance. Models with multiplicative parameters (MPM) for the decrease of a number estimated parame- ters are offered. The process for selection of the vector of base parameters MPM was developed. The performance of adaptive system identification for this case is proved. It is shown that parameters of system estimation at the ap- plication of multiplicative identification must be chosen from a condition of minimization of the criterion of the prediction error. Transformation of interdependence be- tween the lagged variables is offered, allowing eliminat- ing their effect on system work. In the second part of work, the method of synthesis АА identification of the systems containing lagged output variables is offered. We consider a case of linear correlation between an output of the system and operating disturbance. For a solution of a problem, we suggest fulfilling an estimation of operating disturbance. Corresponding procedures are described and proved their efficiency. Simulation results are presented that confirm the efficiency of the adaptive methods. Index Terms—Adaptive algorithm, identification, dis- tributed lag, convergence, model, decision making.