System Identification Algorithm Research Articles

A stochastic gradient-based two-step sparse identification algorithm for multivariate ARX systems

A stochastic gradient-based two-step sparse identification algorithm for multivariate ARX systems

Concurrent Learning-Based Two-Stage Predefined-Time System Identification

This paper proposes a novel two-stage predefined-time system identification algorithm for uncertain nonlinear systems based on concurrent learning. The main feature of the algorithm is that the convergence time of estimation error is an exact predefined parameter, which can be known and adjusted directly by users. Historic identification data are stored in the first stage to guarantee that a finite-rank condition is satisfied. In the second stage, the estimation error converges to zero for linearly parameterized uncertain systems, or it is regulated into the neighborhood of zero for unknown systems modeled by neural networks. The identification algorithm takes effect without the restrictive requirement of the persistent excitation condition. Simulation examples verify the effectiveness of the proposed method.

Vibration suppression of smart composite beam using model predictive controller

Abstract This work presents an adaptive model predictive control (MPC) strategy to suppress the vibration in a laminated composite beam. The control method incorporates a system identification algorithm to estimate the system parameters online, which provides a precise simulation of system dynamics. A fixed-free cantilever composite beam equipped with piezoelectric actuators was used to evaluate the efficacy of the control method. The sensors and actuators are securely bonded to the upper and lower surfaces at arbitrary locations along the beam’s length. A unified mechanical displacement field is applied to all layers, while displacements are considered independently for each layer. The beam is composed of eight layers of material, each with a thickness of 0.2 mm and orientations specified as (90°/0°/90°/0°). To achieve the best performance, the parameters of the MPC were adjusted numerically. The numerical analysis revealed that placing the actuator near the clamped end at the fixed end resulted in superior control outcomes, with a settling time of approximately 1.8 s. Conversely, the longest settling time occurred when the actuator was positioned at the free end, taking around 4 s. This model could potentially be expanded to address vibration in more intricate beams exhibiting nonlinear characteristics. The deflection readings measured at the end of the beam have been utilized as feedback control signals for predicting future behavior over a predetermined control horizon. The subsequent cost function is minimized through a quadratic equation to determine the sequence of optimal yet constrained control inputs. The suggested active vibration control system is then implemented and assessed numerically to examine the effectiveness of the control method.

Application of a modified wavelet threshold denoising algorithm in system identification of WPTS

Application of a modified wavelet threshold denoising algorithm in system identification of WPTS

Highly-efficient filtered hierarchical identification algorithms for multiple-input multiple-output systems with colored noises

Multiple-input multiple-output (MIMO) systems have extensive applications in industrial processes and systems engineering. This letter applies the filtering identification idea to establish a filtered identification model and investigate a filtered auxiliary model-based recursive least squares (F-AM-RLS) algorithm for parameter identification of MIMO systems with colored noises. To improve the computational efficiency, this work further proposes a four-stage filtered auxiliary model-based recursive least squares (4S-F-AM-RLS) algorithm by means of the hierarchical identification principle. Then, by incorporating the forgetting factor, a four-stage filtered auxiliary model-based forgetting factor recursive least squares (4S-F-AM-FF-RLS) algorithm is given to improve the convergence speed and the estimation accuracy. Additionally, the computational complexity analysis of the proposed algorithms indicates that the 4S-F-AM-RLS algorithm effectively reduces the computational burden and improves computational efficiency. Finally, the effectiveness of the F-AM-RLS, 4S-F-AM-RLS and 4S-F-AM-FF-RLS algorithms is validated through a numerical example.

Grey Wolf Optimization Algorithm for Embedded Adaptive Filtering Applications

Nowadays, meta-heuristic algorithms have been emerged as a potential solution in adaptive filtering applications since they offer good convergence properties. Nonetheless, most of them fall into local minimum since their optimization is based on single-solution technique. As a consequence, these algorithms present a high misadjustment level and require a large population to find the optimal solution. Recently, the grey wolf optimization (GWO) algorithm have emerged as a potential solution since it requires a smaller population and possesses a stronger global optimization ability with lesser control parameters. From an engineering perspective, its compactness is an attractive feature. Therefore, this opens new horizons in the implementation of this algorithm in resource-constrained devices. In this paper, we present for the first time the use of the GWO algorithm for system identification and acoustic echo canceller (AEC) and its implementation in a field programmable gate array (FPGA) device to validate its effectiveness. Our results show that the use of the GWO algorithm achieves lower steady-state mean square error (MSE) and requires less computational resources when compared with one of the most used meta-heuristic algorithm.

Machine Learning Classification of Event-Related Brain Potentials during a Visual Go/NoGo Task.

Machine learning (ML) methods are increasingly being applied to analyze biological signals. For example, ML methods have been successfully applied to the human electroencephalogram (EEG) to classify neural signals as pathological or non-pathological and to predict working memory performance in healthy and psychiatric patients. ML approaches can quickly process large volumes of data to reveal patterns that may be missed by humans. This study investigated the accuracy of ML methods at classifying the brain's electrical activity to cognitive events, i.e., event-related brain potentials (ERPs). ERPs are extracted from the ongoing EEG and represent electrical potentials in response to specific events. ERPs were evoked during a visual Go/NoGo task. The Go/NoGo task requires a button press on Go trials and response withholding on NoGo trials. NoGo trials elicit neural activity associated with inhibitory control processes. We compared the accuracy of six ML algorithms at classifying the ERPs associated with each trial type. The raw electrical signals were fed to all ML algorithms to build predictive models. The same raw data were then truncated in length and fitted to multiple dynamic state space models of order nx using a continuous-time subspace-based system identification algorithm. The 4nx numerator and denominator parameters of the transfer function of the state space model were then used as substitutes for the data. Dimensionality reduction simplifies classification, reduces noise, and may ultimately improve the predictive power of ML models. Our findings revealed that all ML methods correctly classified the electrical signal associated with each trial type with a high degree of accuracy, and accuracy remained high after parameterization was applied. We discuss the models and the usefulness of the parameterization.

CONTROL AND IDENTIFICATION OF CONTROLLED AUTO-REGRESSIVE MOVING AVERAGE (CARMA) FORM OF AN INTRODUCED SINGLE-INPUT SINGLE-OUTPUT TUMOR MODEL

The article investigates the parameter estimation for controlled auto-regressive moving average models with gradient based iterative approach and two-stage gradient based iterative approach. Since deriving a new model for tumor model is substantial, introduced system identification algorithms are used in order to estimate parameters of a specific nonlinear tumor model. Besides, in order to estimate tumor model a collection of output and input data is taken from the nonlinear system. Apart from that, effectiveness of the identification algorithms such as convergence rate and estimation error is depicted through various tables and figures. Finally, it is shown that the two stage approach has higher identification efficacy.

System Identification and Fractional-Order Proportional–Integral–Derivative Control of a Distributed Piping System

The vibration of piping systems is one of the most important causes of accelerated equipment wear and reduced work efficiency and safety. In this study, an active vibration control method based on a fractional-order proportional–integral–derivative (PID) controller was proposed to suppress pipeline vibration and reduce pipeline damage. First, a mathematical model of the distributed piping system was established using the finite element analysis method, and the characteristics of the distributed piping system were studied effectively. Further, the time-frequency domain parameter identification method was used to realise the system identification of the cross-point vibration transfer function between the brake and sensor, and the particle swarm optimisation algorithm was utilised to further optimise the transfer function parameters to improve the system identification accuracy. Therefore, a fractional-order PID controller was designed using the D-decomposition method, and the optimal controller parameters were obtained. The experimental and numerical simulation results show that the improved system identification algorithm can significantly improve modelling accuracy. In addition, the designed fractional-order PID controller can effectively reduce the system’s overshoot, oscillation time, and adjustment time, thereby reducing the vibration response of piping systems.

Integrally Bladed Rotor Modal Identification Under Traveling Wave Excitation With High Density Measurement Points

Abstract Vibration testing of an integrally bladed rotor (IBR) is often completed through traveling wave excitation (TWE) bench tests composed of multiple, simultaneously excited inputs. Often, each blade has many output locations. For IBRs with many blades, as often found in the high pressure compressor, the total number of outputs can be several orders of magnitude. Formulation of a full output spectral density matrix from all measurements will then contain an exponential number of values at each frequency bin that can be detrimental to computational resources during the spectral density matrix formulation as well as down-stream system identification algorithms. An online algorithm is proposed for collecting and analyzing TWE data to reduce the large, computationally burdensome datasets into a manageable number of subsets for subsequent system identification. Furthermore, a frequency domain decomposition technique is also proposed for system identification that also attempts to reduce the data size through singular value decomposition. Identified system poles can be averaged from each subset, but mode shapes require stitching each subset together to identify the full mode shape at all output locations. The developed approaches are tested on synthetic TWE data and compared to baseline system identification results obtained using the full spectral density matrix. Results indicate the data subsets accurately compare to the baseline without much loss in accuracy.

Pattern Search Ameliorated Arithmetic Optimization Algorithm for Engineering Optimization and Infinite Impulse Response System Identification

Pattern Search Ameliorated Arithmetic Optimization Algorithm for Engineering Optimization and Infinite Impulse Response System Identification

Solar Energy Prediction Based on Intelligent Predictive Controller Algorithm

The technological advancement in all countries leads to massive energy demand. The energy trading companies struggle daily to meet their customers’ power demands. For a good quality, disturbance-free, and reliable power supply, one must balance electricity generation and consumption at the grid level. There is a profound change in distribution networks due to the intervention of renewable energy generation and grid interactions. Renewable energy sources like solar and wind depend on environmental factors and are subject to unpredictable variations. Earlier, energy distribution companies faced a significant challenge in demand forecasting since it is often unpredictable. With the prediction of the ever-varying power from renewable sources, the power generation and distribution agencies are facing a challenge in supply-side predictions. Several forecasting techniques have evolved, and machine learning techniques like the model predictive controller are suitable for arduous tasks like predicting weather-dependent power generation in advance. This paper employs a Model Predictive Controller (MPC) to predict the solar array’s power. The proposed method also includes a system identification algorithm, which helps acquire, format, validate, and identify the pattern based on the raw data obtained from a PV system. Autocorrelation and cross-correlation value between input and predicted output 0.02 and 0.15. The model predictive controller helps to recognize the future response of the corresponding PV plant over a specific prediction horizon. The error variation of the predicted values from the actual values for the proposed system is 0.8. The performance analysis of the developed model is compared with the former existing techniques, and the role and aptness of the proposed system in smart grid digitization is also discussed.

Statistical Analysis of Block Coordinate Descent Algorithms for Linear Continuous-Time System Identification

Statistical Analysis of Block Coordinate Descent Algorithms for Linear Continuous-Time System Identification

Nonlinear autoregression-based non-destructive evaluation approach for railway tracks using an ultrasonic fiber bragg grating array

The safe operation of railway systems relies on rapid and accurate non-destructive damage detection and evaluation techniques to help rule out potential threats brought by rail track damages. In this paper, an innovative fiber Bragg grating (FBG) array-based sensing system is developed for rail defect inspection by integrating the ultrasonic testing technology, and a nonlinear autoregressive neural network with exogenous inputs (NARX) is proposed for rail damage detection based on measured ultrasonic signals. The developed sensing system can realize the high-speed and multiplexed ultrasonic signals demodulation. The NARX is proposed to process the ultrasonic signals obtained by the developed sensing system, and a specified probabilistic damage sensitive feature (DSF) is tailored to indicate rail health status, derived from the probability density function (PDF) of NARX prediction errors. To verify the proposed sensing system and damage identification algorithm, experimental studies considering four rail conditions (i.e., intact, cracked, bump, and welded conditions) are conducted in this study. First, baseline NARX models are priorly trained using ultrasonic signals acquired on an intact rail segment; then, DSFs corresponding to each condition are calculated enabling effective rail damage diagnosis. To optimize the training parameters of NARX models, three different training functions are utilized to select the most appropriate parameters that are sensitive to rail damages. Last, large-scale testing is conducted, further proving the applicability and reliability of the proposed rail damage evaluation approach. Through this study, it is suggested that integrating high-performance sensing technologies with effective data processing tools would maximumly enable efficient and robust non-destructive testing, especially on important infrastructures such as rails.

Intelligent Gangue Sorting System Based on Dual-Energy X-ray and Improved YOLOv5 Algorithm

Intelligent gangue sorting with high precision is of vital importance for improving coal quality. To tackle the challenges associated with coal gangue target detection, including algorithm performance imbalance and hardware deployment difficulties, in this paper, an intelligent gangue separation system that adopts the elevated YOLO-v5 algorithm and dual-energy X-rays is proposed. Firstly, images of dual-energy X-ray transmission coal gangue mixture under the actual operation of a coal mine were collected, and datasets for training and validation were self-constructed. Then, in the YOLOv5 backbone network, the EfficientNetv2 was used to replace the original cross stage partial darknet (CSPDarknet) to achieve the lightweight of the backbone network; in the neck, a light path aggregation network (LPAN) was designed based on PAN, and a convolutional block attention module (CBAM) was integrated into the BottleneckCSP of the feature fusion block to raise the feature acquisition capability of the network and maximize the learning effect. Subsequently, to accelerate the rate of convergence, an efficient intersection over union (EIOU) was used instead of the complete intersection over union (CIOU) loss function. Finally, to address the problem of low resolution of small targets leading to missed detection, an L2 detection head was introduced to the head section to improve the multi-scale target detection performance of the algorithm. The experimental results indicate that in comparison with YOLOv5-S, the same version of the algorithm proposed in this paper increases by 19.2% and 32.4% on mAP @.5 and mAP @.5:.95, respectively. The number of parameters decline by 51.5%, and the calculation complexity declines by 14.7%. The algorithm suggested in this article offers new ideas for the design of identification algorithms for coal gangue sorting systems, which is expected to save energy and reduce consumption, reduce labor, improve efficiency, and be more friendly to the embedded platform.

The q-gradient LMS spline adaptive filtering algorithm and its variable step-size variant

In this study, an innovation q-gradient least-mean-square (LMS) spline adaptive filtering (S-AF) algorithm (SAF-qLMS) on the basis of the theory of q-derivative is proposed. The q-calculus confronts the issue of slow convergence by mitigating the over-reliance of LMS-type algorithms on the diffusion of eigenvalues in the input correlation matrix. Compared to conventional derivatives, the SAF-qLMS exploits q-calculus to compute the secant of the cost function, enabling it to take larger steps in the search direction for q>1. Furthermore, for balancing the convergence rate and steady-state error of SAF-qLMS and solving the deficiency of the fixed step-size, the SAF-VqLMS based on variable step-size is further proposed. Finally, the convergence conditions of the SAF-qLMS are discussed. Simulations in a correlated Gaussian input environment confirm the outstanding performance of the proposed algorithms for nonlinear system identification.

A hybrid rotordynamic modeling method for a rotor system with flexible foundation and nonlinear support force: Numerical and experimental investigation

The dynamic characteristics of a flexible foundation would greatly affect the dynamic behavior of the rotor system. In this paper, a hybrid dynamic model method in time domain combines the finite element (FE) method and the subspace-based identification method to model a rotor system with the consideration of a flexible foundation and nonlinear support forces. The state-space model of flexible foundation is identified by the subspace state-space system identification algorithm with measured frequency response functions (FRFs). To solve the calculation problem of the hybrid dynamic model in time domain, a general algorithm named the hybrid model fusion solution (HMFS) method is proposed, which combines the linear and nonlinear nodes separation method used in the FE model and the trapezoidal rule used in the state-space model, improving the calculation efficiency of solving transient response with nonlinear forces. Simulations for the hybrid dynamic model under constant speed and speed-up cases were carried out, resulting that the dynamic characteristics of the entire system can be represented when the nonlinear factors of rolling bearings and gravity were also taken into consideration. Additionally, owe to the consideration of tested dynamic behaviors of flexible base, experimental results in the laboratory test rig are well agreed with the numerical ones under two different test configurations. The investigation provides a complete modeling and highly efficient solution framework for the rotor system with a flexible foundation and nonlinear rolling bearings, with the abilities of further consideration of more nonlinear components such as squeeze film dampers and rubbing forces.

A Partial Approximate Least Absolute Deviation-Based Identification Algorithm for a Multivariable Closed-Loop System with Spike Noise

Considering the stability and safety of industrial production, the open-loop system cannot meet the requirements of industrial production, so more and more attention has been paid to multivariable closed-loop system identification. This study focuses on the identification of a multivariate closed-loop system with spike noise in which the model order of the feedback channel is lower than that of the forward channel. Combining principal component analysis (PCA), which is used to eliminate the correlation between the data matrix elements, and the derivable approximate least absolute criterion function, partial approximate least absolute deviation (PALAD) for multivariable closed-loop system identification is proposed. By introducing a deterministic function to replace the absolute value in partial approximate least absolute deviation, the non-differentiable problem of the least absolute deviation function can be solved, and the identifiability of the multivariable closed-loop system in the case mentioned above is theoretically verified. Simulation experiments show the validity of the PALAD algorithm. Compared with the partial least squares (PLS) method, PALAD can effectively restrain the spike noise that follows an SαS distribution and shows stronger robustness when white and spike noises exist simultaneously.

A new backward shift algorithm for system identification by a good choice of frequencies

AbstractIn this study, we propose a novel backward shift algorithm for pole estimation of linear time‐invariant (LTI) systems. By making adjustments, we address the question: How should the frequencies of input signals be chosen? We select the frequencies based on algebraic accuracy in the numerical integral and use the Gaussian–Legendre formula in this process. After poles are located, the discrete Fourier transform (DFT) and the new backward shift algorithm (new BSA) along with the least squares (LS) method are used to identify the systems. Two examples show that the proposed method for pole estimation performs better and the corresponding identification algorithm is practical.

An improved proportionate arctangent framework based LMS/F algorithm for system identification

This paper proposes an improved proportionate arctangent framework that relies on a least mean square/fourth (IPALMS/F) algorithm for the physical sparse system identification problem. The ALMS/F algorithm has significant robustness against impulsive noise, whereas the improved proportionate concept when utilized with the ALMS/F takes advantage of the sparse characteristic to increase convergence time. Finally, the IPALMS/F algorithm is implemented using a recursive adaptive sparse exponential functional link neural network (RASETFLN) nonlinear filter and is named the RASETFLN-IPALMS/F algorithm which resulted in enhanced performance compared to other existing filters in terms of robustness in an impulsive environment, convergence rate, and steady-state error for system identification and acoustic echo cancellation.