System Input-output Data Research Articles

Interval compression‐based model‐free control algorithm for reducing actuator execution frequency

AbstractThe complexity of real‐world systems poses challenges to model‐based control, sparking significant interest in model‐free control methods. By depending exclusively on the system's input–output data, the proposed method eliminates the need to construct intricate internal system models. The implementation is straightforward, can satisfy bounded control inputs, and allows for arbitrary adjustment of the actuator's execution frequency. The proposed method establishes an iterative mechanism under the constraint of bounded control inputs. It guarantees the algorithm's convergence by ensuring the continuous narrowing of the control interval. Furthermore, the update conditions within the iterative strategy can adapt to extremely low and continuously adjusting actuator execution frequencies. The bounded stability of the control method is proven using the continuity definition of functions. Its effectiveness and feasibility are validated through simulation and experimental verification.

Adaptive Dynamic Programming-Based Event-Triggered Robust Control for Multiplayer Nonzero-Sum Games With Unknown Dynamics1-4mmPlease verify and confirm the term "Multi-Player" has been changed to "Multiplayer" in the title of this article.

In this article, the event-triggered robust control of unknown multiplayer nonlinear systems with constrained inputs and uncertainties is investigated by using adaptive dynamic programming. To relax the requirement of system dynamics, a neural network-based identifier is constructed by using the system input-output data. Subsequently, by designing a nonquadratic value function, which contains the bounded functions, the system states, and the control inputs of all players, the event-triggered robust stabilization problem is converted into an event-triggered constrained optimal control problem. To obtain the approximate solution of the event-triggered Hamilton-Jacobi (HJ) equation, a critic network for each player is established with a novel weight updating law to relax the persistence of excitation condition based on the experience replay technique. Furthermore, according to the Lyapunov stability theorem, the present event-triggered robust optimal control ensures the multiplayer system to be uniformly ultimately bounded. Finally, two simulation examples are employed to show the effectiveness of the present method.

Predictive Control of Networked Nonlinear Multiagent Systems With Communication Constraints

This paper investigates the control problem of networked nonlinear multiagent systems with communication constraints and unknown dynamics, based on system input–output data. A cost function for the design of networked nonlinear multiagent control systems is presented, which considers not only the coordination performance between agents but also the performance of individual agents. Utilizing a data model to represent discrete-time nonlinear multiagent systems, a networked data-driven predictive control scheme is proposed to optimize control performance, compensate for communication constrains, and achieve both consensus and stability of networked nonlinear multiagent systems. An optimal control protocol is derived, which minimizes the cost function and makes compensation for communication constraints. The criteria of achieving both consensus and stability of networked nonlinear multiagent systems are provided. An example illustrates the performance of the proposed networked data-driven predictive control scheme.

Online optimal solutions for multi-player nonzero-sum game with completely unknown dynamics

Abstract In this paper, a data-driven approximate neural network (NN) learning scheme is developed to solve the multi-player nonzero-sum (NZS) game problem with completely unknown system dynamics. An augmented NN identifier based on a new parameter estimation algorithm is first established to approximate the completely unknown system dynamics. Then approximated dynamic programming (ADP) with neural networks is constructed to approximate the optimal solutions of the coupled Hamilton-Jacobi equations for each player. The approximated NN value functions are then used to synchronously calculate the optimal control policies for every player. The identifier and ADP NN weights are online updated with the system input-output data based on a novel adaptive law, which could achieve a faster convergence speed. Moreover, the convergence of all NN weights and the stability of the closed-loop system are proved based on the Lyapunov approach. Finally, a dual-driven servo motor system and a three-player nonlinear game system are simulated to verify the feasibility of the developed methods.

Fuzzy Sliding‐mode Strategy for Air–fuel Ratio Control of Lean‐burn Spark Ignition Engines

AbstractMinimization of emissions of carbon dioxide and harmful pollutants and maximization of fuel economy for lean‐burn spark ignition (SI) engines relies to a large extent on precise air–fuel ratio (AFR) control. However, the main challenge of AFR control is the large time‐varying delay in lean‐burn engines. Since the system is usually subject to external disturbances and uncertainties, a high level of robustness in AFR control design must be considered. We propose a fuzzy sliding‐mode control (FSMC) to track the desired AFR in the presence of periodic disturbances. The proposed method is model free and does not need any system characteristics. Based on the fuzzy system input–output data, two scaling factors are first employed to normalize the sliding surface and its derivative. According to the concept of the if‐then rule, an appropriate rule table for the logic system is designed. Then, based on Lyapunov stability criteria, the output scaling factor is determined such that the closed‐loop stability of the internal dynamics with uniformly ultimately bounded (UUB) performance is guaranteed. Finally, the feasibility and effectiveness of the proposed control scheme are evaluated under various operating conditions. The baseline controllers, namely, a PI controller with Smith predictor and sliding‐mode controller, are also used to compare with the proposed FSMC. It is shown that the proposed FSMC has superior regulation performance compared to the baseline controllers.

Development of LTV subspace system identification using basis functions approach to assessing the performance of control loops for nonlinear processes

Abstract A new procedure is proposed for the evaluation of the control loop performance assessment for nonlinear systems using the linear time-varying (LTV) subspace system identification approach. First, a unified LTV subspace system identification is developed. Using the basis function approach, the basis function can make the LTV subspace identification look mathematically the same as the linear time-invariant one. The subspace spanned by the columns of the extended observability matrix from the system input–output data is constructed. And the system matrices are determined directly by the extended observability matrix. To simplify the complicated model structure in LTV systems, two methods are proposed. One is the kernel based method that reduces the computations, and the other, multivariate orthogonal least squares that extracts the useful basis function. Then this work presents an LQG based performance assessment of the LTV system with the subspace framework, and the performance bounds at each time point are set up. They lead to simple monitoring charts, easy tracking of the progress in each operation point, and monitoring the occurrence of the observable upsets. To demonstrate the potential applications of the proposed strategies, a real industrial problem is used.

Identification of transfer functions from surge motion response of a semisubmersible platform using time-varying NARX model

Identification of transfer functions from surge motion response of a semisubmersible platform from model test data is presented in this paper. The identification is carried out by estimating the time-varying model coefficients of a time-varying NARX (TVNARX) model. The coefficients are estimated using proposed method, named artificial bee colony-based Kalman smoother (ABC-KS). System input–output data for identification process are wave height and surge motion from a scaled 1:100 model of a prototype semisubmersible. The applicability of proposed method is assessed numerically and experimentally under unidirectional long-crested random waves. The results show that the linear and quadratic frequency response functions as well as the wave and low frequency responses of a semisubmersible platform can be well identified either in time or frequency domains. The evolution of the nonlinear wave-structure interactions also can be revealed with respect to measured time duration.

Identification of slow drift motions of a truss spar platform using parametric Volterra model

Identification of the first and second-order surge motion transfer functions of a truss spar platform from model test data is presented in this paper. The identification is carried out by estimating the time-varying kernels coefficients of a second-order Volterra model. The coefficients are estimated using proposed method, named particle swarm optimization based Kalman smoother (PSO-KS). System input–output data for identification process are wave height and surge motion from a scaled 1:100 model of a prototype truss spar. The applicability of proposed method is assessed numerically and experimentally under unidirectional long-crested random waves. The results show that the linear and quadratic frequency response functions (LFRF and QFRF) as well as the wave and low frequency responses of a truss spar platform can be well identified either in time or frequency domains. The LFRF and QFRF have high resolution so that evolution of the nonlinear wave interactions can be revealed.

Fault detection method based on bounded error and dynamic threshold techniques

SummaryThis work presents a new fault detection method applicable to dynamic systems. This method is based on historic system input–output data and on the so‐called residual signal. At each discrete time instant, the detection system checks history to locally approximate the dynamics of the faultless system and thus, obtain a prediction of the system output to calculate the residual. Likewise, the maximum threshold allowed by this residual is obtained by applying bounded‐error techniques. The main contribution of this paper is the expression that allows this threshold to be obtained. Detection is based on checking whether the residual exceeds the maximum threshold determined by the method. Copyright © 2015 John Wiley & Sons, Ltd.

Large signal-to-noise ratio quantification in MLE for ARARMAX models

It has been shown that closed-loop linear system identification by indirect method can be generally transferred to open-loop ARARMAX (AutoRegressive AutoRegressive Moving Average with eXogenous input) estimation. For such models, the gradient-related optimisation with large enough signal-to-noise ratio (SNR) can avoid the potential local convergence in maximum likelihood estimation. To ease the application of this condition, the threshold SNR needs to be quantified. In this paper, we build the amplitude coefficient which is an equivalence to the SNR and prove the finiteness of the threshold amplitude coefficient within the stability region. The quantification of threshold is achieved by the minimisation of an elaborately designed multi-variable cost function which unifies all the restrictions on the amplitude coefficient. The corresponding algorithm based on two sets of physically realisable system input–output data details the minimisation and also points out how to use the gradient-related method to estimate ARARMAX parameters when local minimum is present as the SNR is small. Then, the algorithm is tested on a theoretical AutoRegressive Moving Average with eXogenous input model for the derivation of the threshold and a gas turbine engine real system for model identification, respectively. Finally, the graphical validation of threshold on a two-dimensional plot is discussed.

State Estimate-Based Robust Adaptive Control of a Linked Mass-Spring System With Nonlinear Friction

This paper aims to develop a state estimate-based friction fuzzy modeling and robust adaptive control techniques for controlling a class of multiple degrees of freedom (MDOF) mechanical systems. A fuzzy state estimator is proposed to estimate the state variables for friction modeling. Under some conditions, it is shown that such a state estimator guarantees the uniformly ultimate boundedness (UUB) of the estimate error. Based on system input–output data and our proposed state estimator, a robust adaptive fuzzy output-feedback control scheme is presented to control multiple degrees of freedom system with friction. The adaptive fuzzy output-feedback controller can guarantee the uniformly ultimate boundedness of the tracking error of the closed-loop system. A typical mass-spring system is employed in our simulation studies. The results demonstrate that our proposed techniques in this paper have good potential in controlling nonlinear systems with uncertain friction.

Recent developments of Bayesian model class selection and applications in civil engineering

Bayesian model class selection has attracted substantial interest in recent years for selecting the most plausible/suitable class of models based on system input–output data. The Bayesian approach provides a quantitative expression of a principle of model parsimony or of Ockham’s razor which in engineering applications can be stated as simpler models are to be preferred over unnecessarily complicated ones. In this paper, some recent developments are reviewed. Linear and nonlinear regression problems are considered in detail. Bayesian model class selection is particularly useful for regression problems since the regression formula order is difficult to be determined solely by physics due to its empirical nature. Applications are presented in different areas of civil engineering, including artificial neural network for damage detection and seismic attenuation empirical relationship.

Extrinsic curvature-based fault isolation for multi-input—multi-output systems with non-linear fault models

This paper presents an online fault isolation methodology for identifying faulty signals in a multiinput—multi-output dynamical system. It is hypothesized that faults in a dynamical system can be suitably represented via non-linear functions. The isolation scheme, which is implemented online, relies on adaptive non-linear estimates of these non-linear fault functions based on the system input—output data. The non-linear fault estimation is achieved using radial basis function neural network (RBFNN) architecture while the fault isolation is accomplished using extrinsic curvature of the learned RBFNN model. The proposed approach is implemented on an F-5A Freedom Fighter aircraft's lateral-directional model and the results are presented to illustrate the concept.

An improved model-free adaptive control with G function fuzzy reasoning regulation design and its applications

This paper proposes an improved model-free adaptive control with G function fuzzy reasoning regulation (MFCF) for general systems where the equations governing the system are unknown. Such an approach has potential advantages in accommodating complex systems with possibly time-varying dynamics and uncertainties. Model-free adaptive control basic algorithm (MFCB) has two features: first, it does not need any mathematical model; second, the control algorithm is simple and is feasible to be applied in actual industry. This paper considers the use of fuzzy reasoning system, which also only requires observed system input—output data. Moreover this improved control algorithm can make use of human experiences. Related to this, a convergence proof of this improved control algorithm is established. The scheme has been used in the control of welding pool dynamics of the arc welding process, and the experiment results validate the control scheme developed. At the same time, this improved controller can enhance control performance.

THE IDENTIFICATION OF A CLASS OF NONLINEAR SYSTEMS USING A CORRELATION ANALYSIS APPROACH

In this paper, a new correlation analysis based algorithm is proposed for the identification of a class of nonlinear systems which can be described by the NARX (Nonlinear AutoRegressive with eXogeneous input) model with input nonlinearities. Without any assumptions about the structure of an approximating function for the system nonlineariy, the algorithm recovers the functional values of the nonlinearity over a discrete point set associated with the levels of the applied input and estimates the model parameters from the system input output data. An optimal approximating polynomial can then be determined from the nonlinear functional values to produce an optimal estimate for the system nonlinearity. Simulation studies are included to demonstrate the effectiveness of the new method.

Robust fuzzy Gustafson–Kessel clustering for nonlinear system identification

This paper deals with Takagi–Sugeno (TS) fuzzy model identification of nonlinear systems using fuzzy clustering. In particular, an extended fuzzy Gustafson–Kessel (EGK) clustering algorithm, using robust competitive agglomeration (RCA), is developed for automatically constructing a TS fuzzy model from system input–output data. The EGK algorithm can automatically determine the ‘optimal’ number of clusters from the training data set. It is shown that the EGK approach is relatively insensitive to initialization and is less susceptible to local minima, a benefit derived from its agglomerate property. This issue is often overlooked in the current literature on nonlinear identification using conventional fuzzy clustering. Furthermore, the robust statistical concepts underlying the EGK algorithm help to alleviate the difficulty of cluster identification in the construction of a TS fuzzy model from noisy training data. A new hybrid identification strategy is then formulated, which combines the EGK algorithm with a locally-weighted, least-squares method for the estimation of local sub-model parameters. The efficacy of this new approach is demonstrated through function approximation examples and also by application to the identification of an automatic voltage regulation (AVR) loop for a simulated 3 kVA laboratory micro-machine system.

Non-Linear Unsteady Aerodynamic Response Approximation Using Multi-Layer Functionals

Non-linear functional representation of the aerodynamic response provides a convenient mathematical model for motion-induced unsteady transonic aerodynamic loads response, that accounts for both complex non-linearities and time-history effects. A recent development, based on functional approximation theory, has established a novel functional form; namely, the multi-layer functional. For a large class of non-linear dynamic systems, such multi-layer functional representations can be realised via finite impulse response (FIR) neural networks. Identification of an appropriate FIR neural network model is facilitated by means of a supervised training process in which a limited sample of system input-output data sets is presented to the temporal neural network. The present work describes a procedure for the systematic identification of parameterised neural network models of motion-induced unsteady transonic aerodynamic loads response. The training process is based on a conventional genetic algorithm to optimise the network architecture, combined with a simplified random search algorithm to update weight and bias values. Application of the scheme to representative transonic aerodynamic loads response data for a bidimensional airfoil executing finite-amplitude motion in transonic flow is used to demonstrate the feasibility of the approach. The approach is shown to furnish a satisfactory generalisation property to different motion histories over a range of Mach numbers in the transonic regime.

Neural model input selection for a MIMO chemical process

In modelling non-linear systems using neural networks (NN), a commonly used method for the selection of network inputs, or to determine system order and time-delay, is to try different combinations of the system input–output data and choose the best one, giving minimum prediction error. The method is increasingly difficult to apply to industrial systems, due to their multivariable nature and complexity. A systematic method for the selection of model order and time-delay is developed in this paper, and applied to the neural modelling of a multivariable chemical process rig. The method is much simpler compared to the structure identification of the Non-linear Auto-Regressive with eXogenous inputs model (NARX), since the latter also needs to determine the significant terms from a linear-in-parameters polynomial. The orders and delays for system input and output are determined by identifying linearised models of the system. The method can also be applied to other approximations of a MIMO non-linear system, such as fuzzy logic models, etc. The application example demonstrates the selection procedure. Finally, the process rig is modelled using NNs according to the chosen structure, and the modelling error is compared with that of models with different structures to show the effectiveness of the method.

FUZZY PREDICTIVE CONTROL OF UNCERTAIN CHAOTIC SYSTEMS USING TIME SERIES

In this paper, a simple fuzzy logic based intelligent mechanism is developed for predicting and controlling a chaotic system to a desired target, using only input–output data obtained from the unknown (or uncertain) underlying chaotic system. In the chaos prediction phase, a fuzzy system approach incorporating with Gaussian type of fuzzy membership functions is used. Only system input–output data are needed for prediction, and a recursive least-squares computational algorithm is employed for the calculation. In the controller design phase, the Lyapunov stability criterion is used, which forms the basis of the main design principle. Some simulation results on the chaotic Sin map and Hénon map are given, for both prediction and control, to illustrate the effectiveness and control performance of the proposed method.

System Identification Techniques Using the Group Method of Data Handling

A self-organizing nonlinear mapping algorithm, known as the Group Method of Data Handling (GMDH), is examined. A procedure is specifically formulated for using GMDH in the area of system identification. The GMDH algorithm organizes a system model only from information contained in the system input-output data base. The accuracy of the GMDH model is increased by selecting, from the input-output data base, the data vectors which contain the most information concerning the data structure. The selection is achieved using information filtering - a procedure which extracts data structure information by clustering the vectors in the input-output data base. A nonparametric clustering algorithm is the basic tool employed in the information filter. A nonlinear single effect vaporizer with load disturbances and set point changes is identified by the GMDH technique to demonstrate the algorithm and the effects of data selection.