Online Model Identification Research Articles

Adaptive pseudo-derivative feedback with a feed-forward gain controller for permanent magnet synchronous motor servo system based on integrated iterative learning control

This article proposes an adaptive pseudo-derivative feedback with a feed-forward gain controller based on the integrated iterative learning control for the permanent magnet synchronous motor servo system. At first, the improved just-in-time online model identification method is adopted to identify and linearize the nonlinear servo system to obtain the model information for the control strategy. Second, a model-based iterative learning control strategy is presented for the tracking control of permanent magnet synchronous motor servo system. Meanwhile, to guarantee the robust convergence of the iterative learning control system, a new tuning methodology considering the model uncertainties is proposed to select the weighting matrices of the iterative learning control. Third, to further improve control performance, an online generalized predictive control is integrated in the iterative learning control framework, referred to as integrated iterative learning control. By combining generalized predictive control and iterative learning control, the integrated iterative learning control can complement both control methods to obtain good performance, because online generalized predictive control can respond to disturbances immediately and iterative learning control can correct bias left uncorrected by the online controller. Finally, an adaptive pseudo-derivative feedback with a feed-forward gain controller is designed based on the integrated iterative learning control. Since the integrated iterative learning control can be expressed by the pseudo-derivative feedback with a feed-forward gain parameter, the design can achieve both performance improvement and simple controller structure. Experiments confirm the effectiveness of the proposed adaptive pseudo-derivative feedback with a feed-forward gain controller.

A new approach for online T-S fuzzy identification and model predictive control of nonlinear systems

This paper proposes a new unsupervised fuzzy clustering algorithm (NUFCA) to construct a novel online evolving Takagi–Sugeno (T-S) fuzzy model identification method and an adaptive predictive process control methodology. The proposed system identification approach consists of two main steps: antecedent T-S fuzzy model parameters identification and consequent parameters identification. The NUFCA combines the K-nearest neighbour and fuzzy C-means methods into a fuzzy modelling method for partitioning of the input–output data and identifying the antecedent parameters of the fuzzy system; then the recursive least squares method is exploited to obtain initialization type consequent parameters and to construct a method for on-line fuzzy model identification. The integration of the proposed adaptive identification method with the generalized predictive control results in an effective adaptive predictive fuzzy control methodology. For better demonstration of the robustness and efficiency of the proposed methodology, it is applied to the identification of a model for the estimation of the flour concentration in the effluent of a real-world wastewater treatment plant (WWTP); and to control a simulated continuous stirred tank reactor (CSTR), and a real experimental setup composed of two coupled DC motors. The results show that the developed evolving T-S fuzzy model methodology can identify nonlinear systems satisfactorily and can be successfully used for a prediction model of the process for the generalized predictive controller. It is also shown that the algorithm is robust to changes in the initial parameters, and to unexpected disturbances.

A new method for muscle fatigue assessment: Online model identification techniques

The purpose of this study was to propose a method that allows extraction of the current muscle state under electrically induced fatigue. The triceps surae muscle of 5 subjects paralyzed by spinal cord injury was fatigued by intermittent electrical stimulation (5 × 5 trains at 30 Hz). Classical fatigue indices representing muscle contractile properties [peak twitch (Pt) and half-relaxation time (HRT)] were assessed before and after each 5-train series and were used to identify 2 relevant parameters (Fm , Ur ) of a previously developed mathematical model using the Sigma-Point Kalman Filter. Pt declined significantly during the protocol, whereas HRT remained unchanged. Identification of the model parameters with experimental data yielded a model-based fatigue assessment that gave a more stable evaluation of fatigue than classical parameters. This work reinforces clinical research by providing a tool that clinicians can use to monitor fatigue development during stimulation.

Practical Research on Predictive Fuzzy-PID Control in Reactor Temperature Control

Against the characteristics of the temperature in reactor such as time-delay, time-varying and difficulty to build a precise mathematical model in the chemical industry. Through the analysis of dynamic characteristics of the controlled object, the method of fuzzy-PID control was designed based on a predictive model. According to the detected temperature signal, the output deviation of the controller and the on-line identification of prediction model, this algorithm gains the predictive value which uses a generalized predictive model and the fuzzy-PID control. Then compare the predictive value with the reference trajectory to get the deviation. Finally use this deviation and the change of the deviation to optimize the PID control parameters and attain the appropriate amount of system control. The simulation results show that the fuzzy-PID control based on prediction model has strong adaptability, good robustness, control accuracy and higher practical value.

Hammerstein–Wiener nonlinear model based predictive control for relative QoS performance and resource management of software systems

Runtime management of Quality of Service (QoS) performance and resource provisioning is a vital issue in shared resource software environments. A useful performance management technique for such software systems is the relative guarantee feedback control scheme. The existing approaches for this class of control systems are mainly based on off-line linear or on-line model identification and control techniques, which tend to have performance issues in the presence of nonlinearities induced by this scheme. Instead of using such modeling techniques, this paper proposes a new approach for QoS performance management and resource provisioning by using an off-line identification of Hammerstein and Wiener nonlinear block structural model. Using the characteristic structure of the nonlinear model, a predictive feedback controller based on a gain schedule technique is incorporated in the design to achieve the performance objectives. The proposed approach is validated using experiments based on a prototype, demonstrating superior runtime QoS performance management and resource provisioning in a complex software system.

Online model identification of lithium-ion battery for electric vehicles

In order to characterize the voltage behavior of a lithium-ion battery for on-board electric vehicle battery management and control applications, a battery model with a moderate complexity was established. The battery open circuit voltage (OCV) as a function of state of charge (SOC) was depicted by the Nernst equation. An equivalent circuit network was adopted to describe the polarization effect of the lithium-ion battery. A linear identifiable formulation of the battery model was derived by discretizing the frequent-domain description of the battery model. The recursive least square algorithm with forgetting was applied to implement the on-line parameter calibration. The validation results show that the on-line calibrated model can accurately predict the dynamic voltage behavior of the lithium-ion battery. The maximum and mean relative errors are 1.666% and 0.01%, respectively, in a hybrid pulse test, while 1.933% and 0.062%, respectively, in a transient power test. The on-line parameter calibration method thereby can ensure that the model possesses an acceptable robustness to varied battery loading profiles.

Differential constraints for bounded recursive identification with multivariate splines

The ability to perform online model identification for nonlinear systems with unknown dynamics is essential to any adaptive model-based control system. In this paper, a new differential equality constrained recursive least squares estimator for multivariate simplex splines is presented that is able to perform online model identification and bounded model extrapolation in the framework of a model-based control system. A new type of linear constraints, the differential constraints, are used as differential boundary conditions within the recursive estimator which limit polynomial divergence when extrapolating data. The differential constraints are derived with a new, one-step matrix form of the de Casteljau algorithm, which reduces their formulation into a single matrix multiplication. The recursive estimator is demonstrated on a bivariate dataset, where it is shown to provide a speedup of two orders of magnitude over an ordinary least squares batch method. Additionally, it is demonstrated that inclusion of differential constraints in the least squares optimization scheme can prevent polynomial divergence close to edges of the model domain where local data coverage may be insufficient, a situation often encountered with global recursive data approximation.

Online Model Identification for Set-valued State Estimators With Discrete-Time Measurements

AbstractAlthough Kalman Filters provide an optimal solution to the state estimation problem for linear systems, they require knowledge of an accurate description of the model of the system. More robust approaches to Kalman Filtering for systems with uncertain models have been developed, such as set-valued state estimators, which identify the set of possible states that the system may be in, given a description of its uncertainties. The theory behind this class of filters contemplates the possibility of determining if a given system model is feasible with respect to its measurements, but does not explore the possibility of explicitly identifying the system online. This work provides insight into the advantages of performing simultaneous model identification for the class of set-valued estimators, by introducing an Adaptive Set-Valued Estimator, and presenting a practical example of its usage. The results of this robust estimator are then compared to those obtained when using a classical set-valued estimator.

Recursive clustering based on a Gustafson–Kessel algorithm

In this paper an on-line fuzzy identification of Takagi Sugeno fuzzy model is presented. The presented method combines a recursive Gustafson–Kessel clustering algorithm and the fuzzy recursive least squares method. The on-line Gustafson–Kessel clustering method is derived. The recursive equations for fuzzy covariance matrix, its inverse and cluster centers are given. The use of the method is presented on two examples. First example demonstrates the use of the method for monitoring of the waste water treatment process and in the second example the method is used to develop an adaptive fuzzy predictive functional controller for a pH process. The results for the Mackey–Glass time series prediction are also given.

Online identification of nonlinear multivariable processes using self-generating RBF neural networks

This paper addresses the problem of online model identification for multivariable processes with nonlinear and time-varying dynamic characteristics. For this purpose, two online multivariable identification approaches with self-organizing neural network model structures will be presented. The two adaptive radial basis function (RBF) neural networks are called as the growing and pruning radial basis function (GAP-RBF) and minimal resource allocation network (MRAN). The resulting identification algorithms start without a predefined model structure and the dynamic model is generated autonomously using the sequential input-output data pairs in real-time applications. The extended Kalman filter (EKF) learning algorithm has been extended for both of the adaptive RBF-based neural network approaches to estimate the free parameters of the identified multivariable model. The unscented Kalman filter (UKF) has been proposed as an alternative learning algorithm to enhance the accuracy and robustness of nonlinear multivariable processes in both the GAP-RBF and MRAN based approaches. In addition, this paper intends to study comparatively the general applicability of the particle filter (PF)-based approaches for the case of non-Gaussian noisy environments. For this purpose, the Unscented Particle Filter (UPF) is employed to be used as alternative to the EKF and UKF for online parameter estimation of self-generating RBF neural networks. The performance of the proposed online identification approaches is evaluated on a highly nonlinear time-varying multivariable non-isothermal continuous stirred tank reactor (CSTR) benchmark problem. Simulation results demonstrate the good performances of all identification approaches, especially the GAP-RBF approach incorporated with the UKF and UPF learning algorithms. Copyright © 2010 John Wiley and Sons Asia Pte Ltd and Chinese Automatic Control Society

Model identification of a twin screw extruder for thermoplastic vulcanizate (TPV) applications

AbstractMulti‐input multi‐output (MIMO) models of a twin‐screw co‐rotating extruder for thermoplastic vulcanizate (TPV) are developed using the process identification techniques. The process inputs are screw speed (SS) and barrel temperature (WT). The three outputs are motor load (ML), melt temperature (MT), and melt pressure (MP). Two appropriate rubbers for TPV applications with different physical and mechanical properties are used for the experimentation. The process model is obtained from the experimental input–output data using various identification techniques such as least squares and prediction error. Recursive online model identification is performed on the process to update the model parameters in real time. To perform the identification studies, the process data was transferred via OPC server from the local PLC (Programmable Logic Controller) to the Advanced Control and Identification toolbox in MATLAB software. The effect of rubber properties and two curative agents (Peroxide and Phenolic) in the TPV experiment are studied on the final identified models. This comprehensive model identification study provided sufficient accurate models for further model based process analysis and control for TPV applications. POLYM. ENG. SCI., 2010. © 2009 Society of Plastics Engineers

Affine modeling of nonlinear multivariable processes using a new adaptive neural network-based approach

This paper presents a new method for on-line identification of exact affine model for multivariable processes with nonlinear and time-varying behaviors. A self-generating radial basis function (RBF) neural network trained by growing and pruning algorithm for RBF (GAP–RBF) is utilized for deriving the affine model. The extended Kalman filter (EKF) is used for parameter adaptation in the GAP–RBF neural network. The growing and pruning criteria of the original GAP–RBF have been modified with the objective to enhance its performance in on-line identification. Simulation results on two nonlinear multivariable CSTR benchmark problems show an excellent performance of the proposed approach, incorporated with the modified GAP–RBF (MGAP–RBF) neural network, for affine modeling.

Online transition matrix identification of the stateevolution model for the extended Kalman filter inelectrical impedance tomography

One of the electrical impedance tomography objectives is to estimate the electricalresistivity distribution in a domain based only on contour electrical potential measurementscaused by an imposed electrical current distribution into the boundary. In biomedicalapplications, the random walk model is frequently used as evolution model and, under thisconditions, it is observed poor tracking ability of the Extended Kalman Filter (EKF). Ananalytically developed evolution model is not feasible at this moment. The present workinvestigates the possibility of identifying the evolution model in parallel to the EKF and updatingthe evolution model with certain periodicity. The evolution model is identified using the historyof resistivity distribution obtained by a sensitivity matrix based algorithm. To numericallyidentify the linear evolution model, it is used the Ibrahim Time Domain Method, normallyused to identify the transition matrix on structural dynamics. The investigation was performedby numerical simulations of a time varying domain with the addition of noise. Numericaldificulties to compute the transition matrix were solved using a Tikhonov regularization. TheEKF numerical simulations suggest that the tracking ability is significantly improved.

Direct and Indirect Stable Adaptive Control for Suspension Systems with MR Damper

The paper is concerned with stable adaptive schemes for semi-active control of suspension systems, which can deal with uncertainties in a nonlinear model of MR damper. To compensate for unknown nonlinear hysteresis dynamics of the MR damper, an adaptive inverse controller is implemented by indirect and direct methods. In the indirect method, on-line identification of a forward model of MR damper is executed. The direct method updates the inverse controller directly. Then a linear control scheme is designed and applied to the nonlinearly compensated system to give the desired damping force by taking into account the passivity of the MR damper. The effectiveness of the proposed adaptive semiactive control scheme is validated and the stability of the obtained total adaptive system is analyzed.

On Use of a Nominal Internal Model to Detect a Loss of Balance in a Maximal Forward Reach

We hypothesize that the CNS detects a loss of balance by comparing outputs predicted by a nominal, forward internal model with actual sensory outputs. When the resulting control error signal reaches an anomalously large value, this control error anomaly (CEA) signals a loss of balance and precedes any observable compensatory response. To test this hypothesis, a multi-input, multi-output internal model of a standing forward reach task was developed that incorporated on-line model identification and a Gaussian failure detection algorithm. Eleven healthy young women were then asked to stand and reach forward to a target positioned from 95 to 125% of their maximum reach distance. Kinematic and kinetic data were recorded at 100 Hz unilaterally from the upper body, leg, and foot. Evidence of successful CEA detection was a compensatory step between 100 ms and 2 s later. The results show that use of a threshold, set at 3 SD from the mean, on error in the control of leg segment acceleration detected a CEA and correctly predicted a compensatory response in 92.6% of 108 trials. Leg acceleration control error was a better predictor than upper body or foot acceleration control error (P = 0.000). CEA detection performed more reliably than loss of balance detection algorithms based on kinematic thresholds (P = 0.000). The results support the hypothesis that a loss of balance may be identified via the use of a nominal forward internal model and probabilistic error monitoring.

Active vibration control of machine tool structures - Part 1: DSP algorithms

Ths paper describes the first stage of a project to actively cancel vibration in vertical milling machmes and thereby improve the work-piece surface finish. Parallels are drawn between continuous methods of vibration modelling structures, and digital methods that can be implemented in real-time. The timevariant nature of machine structures when cutting led to an investigation of adaptive digital filter algorithms. Results are shown for experiments carried out with two adaptive schemes used for on-line model identification and control.

Advanced process control of hot isostatic pressing. Application to Astroloy HIP

For a HIP operation, conventional control of the process variables (T, P) is the current mode used. The process variables follow a prespecified trajectory (schedule) assumed to lead to the final quality of product. A second mode is advanced control, where on-line measurements of a product state variable enable to monitor the evolution of the product variables and if necessary to compute corrections to the schedule in order to obtain the desired quality of the final product. HIP advanced process control requires continuous monitoring of the sample density during all stages of the process and the on-line identification a process model. To this end, a dilatometer has been built for in situ monitoring densification during a HIP operation, the Ashby process model has been validated and on-line identification has been demonstrated. It is shown that advanced process control of a HIP operation can be performed in two steps : on-line identification of the process model and subsequent process path replanning. The first step can only be performed during the process variables plateau of the preset temperature- pressure cycle, and process path replanning consists of a new temperature- pressure plateau.

Predictive fuzzy PID control: theory, design and simulation

Predictive fuzzy PID control theory is developed in this paper, which offers a new approach for robust control of time-delay systems. The paper describes the functional structure, design principle, and stability analysis of a new predictive fuzzy PID controller. Sufficient computer simulations are provided for illustration and verification. First, the structure of the controller is derived from both the fuzzy PID control and the generalized predictive control concepts. Then, on-line model identification, optimal cost index, fuzzification , rule base, and defuzzification of the representative predictive fuzzy PD+I controller are discussed in detail. Lyapunov asymptotic stability analysis is conducted. Then, many computer simulations are performed to compare with several closely related controllers such as the fuzzy PD+I controller and the Smith-type predictive fuzzy PD+I controller. In the simulations, second-order linear systems with/without time delays, nonlinear systems with/without time delays, uncertain linear systems with time delays, and uncertain nonlinear systems with time delays are used to confirm the advantages of the new predictive fuzzy PD+I controller. Finally, this method is applied to control some chaotic systems with success. This predictive fuzzy control method provides a new way for controlling uncertainty and complex linear and nonlinear systems, even with significant time delay.

Modeling and Compensation of Hysteresis in Piezoceramic Transducers for Vibration Control

Hysteretic behavior in piezoceramic transducers is investigated theoretically and experimentally. The applicability of the rate-independent generalized Maxwell resistive capacitor (MRC) hysteresis model is established. Methods for MRC and inverse MRC online model identification are developed by first establishing that the MRC and its inverse are the same particular cases of the classical Preisach hysteresis model. This enables use of the extensive mathematical framework that has been developed for Preisach models. A method of incorporating the MRC model in a feed forward control scheme for hysteresis compensation is also presented. Experimental studies on a 1-3 piezoceramic composite support the theoretical developments and their applicability to piezoceramics.

A comparison of some adaptive-predictive fuzzy-control strategies

A comparison of some fuzzy model-based adaptive-predictive control strategies is provided. The considered strategies are based on the minimization of a predefined long-range cost functional reflecting the tracking error between the reference to be followed and the plant's output. They begin at the current time instant and continue up to a given time horizon in the future. A fuzzy model of the plant is used for forecasting. A particular control strategy is specified by selecting the type of fuzzy model to be used as a plant description, the algorithm used for model parameter estimation and the assumptions required for the long-range cost minimization. The simulation results presented show that the control performance depends not only on the assumption required for multi-step optimization, but most notably it depends on the convergence rate of the parameter estimation algorithm used in the online model identification.