Application Of System Identification Research Articles

Synthesizing spatiotemporally sparse smartphone sensor data for bridge modal identification

Smartphones as vibration measurement instruments form a large-scale, citizen-induced, and mobile wireless sensor network (WSN) for system identification and structural health monitoring (SHM) applications. Crowdsourcing-based SHM is possible with a decentralized system granting citizens with operational responsibility and control. Yet, citizen initiatives introduce device mobility, drastically changing SHM results due to uncertainties in the time and the space domains. This paper proposes a modal identification strategy that fuses spatiotemporally sparse SHM data collected by smartphone-based WSNs. Multichannel data sampled with the time and the space independence is used to compose the modal identification parameters such as frequencies and mode shapes. Structural response time history can be gathered by smartphone accelerometers and converted into Fourier spectra by the processor units. Timestamp, data length, energy to power conversion address temporal variation, whereas spatial uncertainties are reduced by geolocation services or determining node identity via QR code labels. Then, parameters collected from each distributed network component can be extended to global behavior to deduce modal parameters without the need of a centralized and synchronous data acquisition system. The proposed method is tested on a pedestrian bridge and compared with a conventional reference monitoring system. The results show that the spatiotemporally sparse mobile WSN data can be used to infer modal parameters despite non-overlapping sensor operation schedule.

Nonparametric Variable Step-Size LMAT Algorithm

This paper proposes a nonparametric variable step-size least mean absolute third (NVSLMAT) algorithm to improve the capability of the adaptive filtering algorithm against the impulsive noise and other types of noise. The step-size of the NVSLMAT is obtained using the instantaneous value of a current error estimate and a posterior error estimate. This approach is different from the traditional method of nonparametric variance estimate. In the NVSLMAT algorithm, fewer parameters need to be set, thereby reducing the complexity considerably. Additionally, the mean of the additive noise does not necessarily equal zero in the proposed algorithm. In addition, the mean convergence and steady-state mean-square deviation of the NVSLMAT algorithm are derived and the computational complexity of NVSLMAT is analyzed theoretically. Furthermore, the experimental results in system identification applications presented illustrate the principle and efficiency of the NVSLMAT algorithm.

Nonlinear system identification—Application for industrial hydro-static drive-line

The goal of the paper is to describe the added value and complexities of nonlinear system identification applied to a large scale industrial test setup. The additional important insights provided by the frequency domain nonlinear approach are significant and for such systems the nonlinear system identification is important, for example to estimate the noise and non-linearities levels, which can indicate mechanical and configuration issues. It is not the goal to provide a final full-scale model, but to explore what is the applicability of the nonlinear system identification theories for a complex multi-physical non-academic test-case.

A shrinkage variable step size for normalized subband adaptive filters

The conventional normalized subband adaptive filter (NSAF) using a constant step-size generally faces an inherent trade-off between the steady-state misalignment and the convergence rate. We propose herein a variable step-size NSAF algorithm by minimizing the mean-square deviation (MSD) between the optimal weight vector and the weight vector estimate with the utilization of the shrinkage denoising technique. With the estimation error involved in the step-size adaptation for each subband individually, the proposed algorithm is capable of tracking non-stationary environments. Without the explicit whitening assumption of the input signal in each subband, the proposed algorithm exhibits low steady-state MSD even when the input signal of each subband is colored. Simulation results validate the low misalignment and good tracking ability of the proposed algorithm in system identification application.

System identification application using Hammerstein model

Generally, memoryless polynomial nonlinear model for nonlinear part and finite impulse response (FIR) model or infinite impulse response model for linear part are preferred in Hammerstein models in literature. In this paper, system identification applications of Hammerstein model that is cascade of nonlinear second order volterra and linear FIR model are studied. Recursive least square algorithm is used to identify the proposed Hammerstein model parameters. Furthermore, the results are compared to identify the success of proposed Hammerstein model and different types of models.

Extending the Kalman filter for structured identification of linear and nonlinear systems

This paper considers a novel approach to system identification which allows accurate models to be created for both linear and nonlinear multi-input/output systems. In addition to conventional system identification applications, the method can also be used as a black-box tool for model order reduction. A nonlinear Kalman filter is extended to include slow-varying parameter states in a canonical model structure. Interestingly, in spite of all model parameters being unknown at the start, the filter is able to evolve parameter estimates to achieve 100% accuracy in noise-free test cases, and is also proven to be robust to noise in the measurements. The canonical structure ensures a well-conditioned model which simultaneously provides valuable dynamic information to the engineer. After extensive testing of a linear example, the model structure is extended to a generalised nonlinear form, which is shown to accurately identify the handling response of a full vehicle model.

Nonlinear Systems Identification and Control Using Uncertain Rule-based Fuzzy Neural Systems with Stable Learning Mechanism

This paper proposes an uncertain rule-based fuzzy neural system (UFNS-S) with stable learning mechanism for nonlinear systems identification and control. The proposed UFNS-S system not only preserves the ability of handling uncertain information but also performs less computational effort. The sinusoidal perturbations are adopted to combine with the fuzzy term sets of UFNS-S. For training the UFNS-S systems on system identification and control applications, the gradient descent method with adaptive learning rate is derived. This guarantees the convergence of UFNS-S by choosing adaptive learning rates which enhance the convergent speed. This provides a simple way for choosing the learning rates for training the UFNS-S which also guarantees convergence and faster learning. Finally, the effectiveness and performance of the proposed approach is illustrated by several examples, computational complexity analysis, nonlinear system identification, and tracking control of two-link robot manipulator system.

Nonlinear identification of the total baroreflex arc: chronic hypertension model.

The total baroreflex arc is the open-loop system relating carotid sinus pressure (CSP) to arterial pressure (AP). Its linear dynamic functioning has been shown to be preserved in spontaneously hypertensive rats (SHR). However, the system is known to exhibit nonlinear dynamic behaviors. The aim of this study was to establish nonlinear dynamic models of the total arc (and its subsystems) in hypertensive rats and to compare these models with previously published models for normotensive rats. Hypertensive rats were studied under anesthesia. The vagal and aortic depressor nerves were sectioned. The carotid sinus regions were isolated and attached to a servo-controlled piston pump. AP and sympathetic nerve activity were measured while CSP was controlled via the pump using Gaussian white noise stimulation. Second-order, nonlinear dynamics models were developed by application of nonparametric system identification to a portion of the measurements. The models of the total arc predicted AP 21-43% better (P < 0.005) than conventional linear dynamic models in response to a new portion of the CSP measurement. The linear and nonlinear terms of these validated models were compared with the corresponding terms of an analogous model for normotensive rats. The nonlinear gains for the hypertensive rats were significantly larger than those for the normotensive rats [-0.38 ± 0.04 (unitless) vs. -0.22 ± 0.03, P < 0.01], whereas the linear gains were similar. Hence, nonlinear dynamic functioning of the sympathetically mediated total arc may enhance baroreflex buffering of AP increases more in SHR than normotensive rats.

A geometric characterisation of persistently exciting signals generated by continuous-time autonomous systems

Abstract: The persistence of excitation of signals generated by time-invariant, continuous-time, autonomous linear and nonlinear systems is studied. The notion of persistence of excitation is characterised as a rank condition which is reminiscent of a geometric condition used to study the reachability properties of a control system. The notions and tools introduced are illustrated by means of simple examples and of an application in system identification.

System Identification of Vessel Steering With Unstructured Uncertainties by Persistent Excitation Maneuvers

System identification of vessel steering associated with unstructured uncertainties is considered in this paper. The initial model of vessel steering is derived by a modified second-order Nomoto model (i.e., nonlinear vessel steering with stochastic state-parameter conditions). However, that model introduces various difficulties in system identification, due to the presence of a large number of states and parameters and system nonlinearities. Therefore, partial feedback linearization is proposed to simplify the proposed model, where the system-model unstructured uncertainties can also be separated. Furthermore, partial feedback linearization reduces the number of states and parameters and the system nonlinearities, given the resulting reduced-order state model. Then, the system identification approach is carried out, for both models (i.e., full state model and reduced-order state model), resorting to an extended Kalman filter (EKF). As illustrated in the results, the reduced-order model was able to successfully identify the required states and parameters when compared to the full state model in vessel steering under persistent excitation maneuvers. Therefore, the proposed approach can be used in a wide range of system identification applications.

Direct Differentiation of the Particle Finite-Element Method for Fluid–Structure Interaction

AbstractSensitivity analysis of fluid–structure interaction (FSI) provides an important tool for assessing the reliability and performance of coastal infrastructure subjected to storm and tsunami hazards. As a preliminary step for gradient-based applications in reliability, optimization, system identification, and performance-based engineering of coastal infrastructure, the direct differentiation method (DDM) is applied to FSI simulations using the particle finite-element method (PFEM). The DDM computes derivatives of FSI response with respect to uncertain design and modeling parameters of the structural and fluid domains that are solved in a monolithic system via the PFEM. Geometric nonlinearity of the free surface fluid flow is considered in the governing equations of the DDM along with sensitivity of material and geometric nonlinear response in the structural domain. The analytical derivatives of elemental matrices and vectors with respect to element properties are evaluated and implemented in an open ...

A TUTORIAL ON APPLYING ARTIFICIAL NEURAL NETWORKS AND GEOMETRIC BROWNIAN MOTION TO PREDICT A STOCHASTIC TIME SERIES

Several challenges in the engineering or financial world can be resolved with a proper handle on data. Amongst other applications in engineering, system identification and parameter estimation are widely used in developing control strategies for automation. In this domain, there would be requirements to design an adaptive control system. In order to design an adaptive control system, an adaptive model needs to be estimated or identified. This is generally done by studying the data and creating a transfer function. In the process, regression, artificial neural networks (ANN), random walk theory and Markov chain estimates are used to understand a time series and create a model. While some of these processes are stationary, some are non-stationary. These methods are chosen based on the nature and availability of historical data. One of the issues that always remain is which method is appropriate for a certain application. The objective of this tutorial is to illustrate how artificial neural network and Geometric Brownian motion can be used in this regard. An attempt is made to predict the future price of a stock of a corporation. Stock prices are an example for a stochastic time series. Initially, an artificial neural network is used to predict the stock price. The network is designed as a Multi layer Back propagation type network. Profit over earnings and S&P are used as inputs. Thereafter, Geometric Brownian motion is explained and used on the same dataset to come up with its predictions. The results from both neural network and geometric Brownian motion are compared.

An application of system identification in metrology

Metrology is advancing by development of new measurement techniques and corresponding hardware. A given measurement technique, however, has fundamental speed and precision limitations. In order to overcome the hardware limitations, we develop signal processing methods based on the prior knowledge that the measurement process dynamics is linear time-invariant.Our approach is to model the measurement process as a step response of a dynamical system, where the input step level is the quantity of interest. The solution proposed is an algorithm that does real-time processing of the sensor's measurements. It is shown that when the measurement process dynamics is known, the input estimation problem is equivalent to state estimation. Otherwise, the input estimation problem can be solved as a system identification problem. The main underlying assumption is that the measured quantity is constant and the measurement process is a low-order linear time-invariant system. The methods are validated and compared on applications of temperature and weight measurement.

Dynamic Model Extraction for Speed Control of Induction Motor Driven by a Voltage Source Inverter with Using System Identification Application

Dynamic Model Extraction for Speed Control of Induction Motor Driven by a Voltage Source Inverter with Using System Identification Application

Regularization Paths for Re-Weighted Nuclear Norm Minimization

We consider a class of weighted nuclear norm optimization problems with important applications in signal processing, system identification, and model order reduction. The nuclear norm is commonly used as a convex heuristic for matrix rank constraints. Our objective is to minimize a quadratic cost subject to a nuclear norm constraint on a linear function of the decision variables, where the trade-off between the fit and the constraint is governed by a regularization parameter. The main contribution is an algorithm to determine the so-called approximate regularization path, which is the optimal solution up to a given error tolerance as a function of the regularization parameter. The advantage is that we only have to solve the optimization problem for a fixed number of values of the regularization parameter, with guaranteed error tolerance. The algorithm is exemplified on a weighted Hankel matrix model order reduction problem.

Adaptive filtering implemented over TMS320c6713 DSP platform for system identification

This paper presents the experimental development of software and hardware configuration to implement two adaptive algorithms: LMS (Least Mean Square) and RLS (Recursive Least Square), using TMS320C6713 DSP platform of Texas Instruments, for unknown systems identification. Methodology for implementation and validation analysis for the adaptive algorithms is described in detail for real-time systems identification applications, and the experimental results were evaluated in terms of performance criterions in time domain, frequency domain, computational complexity, and accuracy.

The Design and Implementation of Key Technologies of the Computer Algorithm Dynamic System

This topic mainly aimed at the dynamic simulation of some actual problems. The research is suitable for the dynamic system of modeling artificial model and learning algorithm. First it summarizes the existing network simulation model of dynamic system and network delay unit. All the feedback network and part of the feedback network model, the network topology is given and used for system simulation. The method of simulation learning algorithm and simulation instance. And then established the time-varying input and output process neural networks and network model of two kinds of discrete process neural network for its properties are analyzed and proved that the concrete learning algorithm is deduced. Finally combined with the actual problem, time-varying input and output process neural network is given in the application of system identification.

Continuous Mixed &lt;formula formulatype="inline"&gt; &lt;tex Notation="TeX"&gt;$p$&lt;/tex&gt;&lt;/formula&gt;-Norm Adaptive Algorithm for System Identification

We propose a new adaptive filtering algorithm in system identification applications which is based on a continuous mixed $p$ -norm. It enjoys the advantages of various error norms since it combines p-norms for $1 \leq p \leq 2$ . The mixture is controlled by a continuous probability density-like function of $p$ which is assumed to be uniform in our derivations in this letter. Two versions of the suggested algorithm are developed. The robustness of the proposed algorithms against impulsive noise are demonstrated in a system identification simulation.

Affine-Projection-Like Adaptive-Filtering Algorithms Using Gradient-Based Step Size

A new class of affine-projection-like (APL) adaptive-filtering algorithms is proposed. The new algorithms are obtained by eliminating the constraint of forcing the a posteriori error vector to zero in the affine-projection algorithm proposed by Ozeki and Umeda. In this way, direct or indirect inversion of the input signal matrix is not required and, consequently, the amount of computation required per iteration can be reduced. In addition, as demonstrated by extensive simulation results, the proposed algorithms offer reduced steady-state misalignment in system-identification, channel-equalization, and acoustic-echo-cancelation applications. A mean-square-error analysis of the proposed APL algorithms is also carried out and its accuracy is verified by using simulation results in a system-identification application.

Fractional-order system identification and proportional-derivative control of a solid-core magnetic bearing

This paper presents the application of fractional-order system identification (FOSI) and proportional-derivative (PDµ) control to a solid-core magnetic bearing (MB). A practical strategy for closed-loop incommensurate FOSI along with a modified error criterion is utilized to model the MB system and a corresponding, verification experiment is carried out. Based on the identified model, integer-order (IO) PD and fractional-order (FO) PDµ controllers are designed and compared with the same specifications. Besides, the relation between the two categories of controllers is discussed by their feasible control zones. Final simulation and experimental results show that the FO PDµ controller can significantly improve the transient and steady-state performance of the MB system comparing with the IO PD controller.