Model Identification Approach Research Articles

Snow hydrology signatures for model identification within a limits‐of‐acceptability approach

AbstractDischarge simulation from snow‐dominated catchments seems to be an easy task. Any spatially explicit precipitation–runoff model coupled to a temperature‐index snow model generally yields simulations that mimic well the observed daily discharges. The robustness of such models is, however, questionable: in the presence of strong annual discharge cycles, small model residuals do not guarantee high explanatory power of the underlying model. This paper proposes a methodology for snow hydrological model identification within a limits‐of‐acceptability framework, where acceptable model simulations are the ones that reproduce a set of signatures within an a priori specified range. The signatures proposed here namely include the relationship between the air temperature regime and the discharge regime, a new snow hydrology signature that can be readily transferred to other Alpine settings. The discriminatory power of all analysed signatures is assessed with a new measure of their discriminatory power in the model prediction domain. The value of the proposed snow hydrology signatures and of the limits‐of‐acceptability approach is demonstrated for the Dischma river in Switzerland, whose discharge shows a strong temporal variability of hydrologic forcing conditions over the last 30 years. The signature‐based model identification for this case study leads to the surprising conclusion that the observed discharge data contains a multi‐year period that cannot be reproduced with the model at hand. This model‐data mismatch might well result from a yet to be identified problem with the discharge observations, which would have been difficult to detect in a classical residual‐based model identification approach. Overall, the detailed results for this case study underline the robustness of the limits‐of‐acceptability approach in the presence of error‐prone observations if it is applied in combination with relatively robust signatures. Future work will show whether snow hydrology signatures and their limits‐of‐acceptability can be regionalized to ungauged catchments, which would make this model selection approach particularly powerful for Alpine environments. Copyright © 2016 John Wiley & Sons, Ltd.

A neural network-based distributed parameter model identification approach for microcantilever

The microcantilever used in micro–nanomanipulator is a spatially distributed and flexible mechanical system. An accurate model of the microcantilever is essential for the accurate tip positioning and force sensing. Traditional lumped parameter model will lose the spatial dynamics. Though the nominal Euler–Bernoulli model is a distributed parameter model, in practice there are still some unknown nonlinear dynamics. In this study, a neural network-based distributed parameter model identification approach is proposed for modelling the microcantilever. First, a nominal Euler–Bernoulli beam model is derived. To compensate unknown nonlinear dynamics, a nonlinear term that needs to be estimated is added in the nominal model. For finite-dimensional implementation, the infinite-dimensional partial differential equation model is reduced into a finite-dimensional ordinary differential equation model using the Galerkin method. Next, a neural network-based intelligent learning approach is developed to learn the unknown nonlinearities from the input–output data. A radial basis function recurrent neural network observer is designed to estimate the finite-dimensional states from a few sensors of measurements. After that, a general regression neural network model is identified to establish the nonlinear spatiotemporal dynamic model between the inputs and outputs. The effectiveness of the proposed neural network-based distributed parameter modelling approach is verified by the simulations on a typical microcantilever.

Error‐triggered on‐line model identification for model‐based feedback control

In industry, it may be difficult in many applications to obtain a first‐principles model of the process, in which case a linear empirical model constructed using process data may be used in the design of a feedback controller. However, linear empirical models may not capture the nonlinear dynamics over a wide region of state‐space and may also perform poorly when significant plant variations and disturbances occur. In the present work, an error‐triggered on‐line model identification approach is introduced for closed‐loop systems under model‐based feedback control strategies. The linear models are re‐identified on‐line when significant prediction errors occur. A moving horizon error detector is used to quantify the model accuracy and to trigger the model re‐identification on‐line when necessary. The proposed approach is demonstrated through two chemical process examples using a model‐based feedback control strategy termed Lyapunov‐based economic model predictive control (LEMPC). The chemical process examples illustrate that the proposed error‐triggered on‐line model identification strategy can be used to obtain more accurate state predictions to improve process economics while maintaining closed‐loop stability of the process under LEMPC. © 2016 American Institute of Chemical Engineers AIChE J, 63: 949–966, 2017

Intelligent Flight Control Systems Evaluation for Loss-of-Control Recovery and Prevention

Recent developments in the field of loss-of-control recovery and prevention included improved pilot training, cockpit automation, and fault-tolerant control. The Flight Mechanics Action Group on Fault-Tolerant Control of the Group for Aeronautical Research and Technology in Europe demonstrated the advantages of fault-tolerant “intelligent,” flight control systems. The research enabled the improvement of the technology readiness level of these systems by evaluating one of them in realistic operational scenarios. The handling qualities results of a piloted flight simulator assessment with a damaged aircraft model showed that an online physical model identification approach contributed to improved pilot performance following potentially catastrophic structural and flight critical system failures. After the failures and subsequent control reconfiguration by the intelligent flight control system, airline and engineering test pilots experienced no difficulties in conducting a safe approach and landing.

A compressive sensing-based approach for Preisach hysteresis model identification**This work was supported by the National Science Foundation (CMMI 1301243).

The Preisach hysteresis model has been adopted extensively in magnetic and smart material-based systems. Fidelity of the model hinges on accurate identification of the Preisach density function. Existing work on the identification of the density function usually involves applying an input that provides sufficient excitation and measuring a large set of output data. In this paper, we propose a novel compressive sensing-based approach for Preisach model identification that requires fewer measurements. The proposed approach adopts the discrete cosine transform of the output data to obtain a sparse vector, where the order of all the output data is assumed to be known. The model parameters can be efficiently reconstructed using the proposed scheme. For comparison purposes, a constrained least-squares scheme using the same number of measurements is also considered. The root-mean-square error is adopted to examine the model identification performance. The proposed identification approach is shown to have better performance than the least-squares scheme through both simulation and experiments involving a vanadium dioxide ()-integrated microactuator.

A New Approach of Modeling an Ultra-Super-Critical Power Plant for Performance Improvement

A suitable model of coordinated control system (CCS) with high accuracy and simple structure is essential for the design of advanced controllers which can improve the efficiency of the ultra-super-critical (USC) power plant. Therefore, with the demand of plant performance improvement, an improved T-S fuzzy model identification approach is proposed in this paper. Firstly, the improved entropy cluster algorithm is applied to identify the premise parameters which can automatically determine the cluster numbers and initial cluster centers by introducing the concept of a decision-making constant and threshold. Then, the learning algorithm is used to modify the initial cluster center and a new structure of concluding part is discussed, the incremental data around the cluster center is used to identify the local linear model through a weighted recursive least-square algorithm. Finally, the proposed approach is employed to model the CCS of a 1000 MW USC one-through boiler power plant by using on-site measured data. Simulation results show that the T-S fuzzy model built in this paper is accurate enough to reflect the dynamic performance of CCS and can be treated as a foundation model for the overall optimizing control of the USC power plant.

Development and validation of grey-box models for forecasting the thermal response of occupied buildings

Building thermal wall mass provides a flexible heat capacity which can be effectively used for load shifting, thus enabling demand side management (DSM). The most crucial barrier for a practical application of buildings as short term heat storage is the lack of knowledge about the building physical properties. This study presents a model identification approach for forecasting the building thermal response based on grey-box models. The model parameters are estimated by optimizing the model output to historical data under the consideration of plausible physical constraints. The thermal flexibility of a building and therewith its demand side management potential can be determined based on the assessed model parameters. This study uses measurements from buildings in normal operation to evaluate the thermal prediction of grey-box models despite the stochastic events within the training data. In order to find the best level of model complexity, four grey-box models were compared in their ability to forecast the building indoor temperature behaviour. The analysis revealed that a two-capacity model structure with an additional consideration of the indoor air as a mass-less node (4R2C-model) enables the most accurate qualitative prediction of the indoor temperature. Further, the general validity of the model structure allows for its application on different buildings types.

A multiwavelet-based time-varying model identification approach for time–frequency analysis of EEG signals

An efficient multiwavelet-based time-varying modeling scheme is proposed for time–frequency analysis (TFA) of electroencephalogram (EEG) data. In the new multiwavelet-based parametric modeling framework, the time-dependent parameters in the time-varying model are locally represented using a novel multiwavelet decomposition scheme. An effective orthogonal least squares (OLS) algorithm aided by mutual information criterion is then applied for sparse model selection and parameter estimation. The resultant estimation of time-dependent spectral density in the signal can simultaneously achieve a high resolution in both time and frequency, which is a powerful TFA technique for nonstationary biomedical signals including EEG. Two examples, one for an artificial EEG signal and another for a real EEG signal are included to show the effectiveness and applicability of the new proposed approach. Simulation studies and applications to real EEG data elucidate that the proposed wavelet approach is capable of achieving a high time–frequency representation for nonstationary processes.

Handling Plant Variation via Error-Triggered On-line Model Identification: Application to Economic Model Predictive Control

Abstract: In the present work, an error-triggered on-line model identification approach is introduced for closed-loop systems that is able to detect and compensate for significant disturbances and variations in the plant. A moving horizon error detector is used to quantify the model accuracy and to trigger the model re-identification on-line when necessary. The proposed approach is applied in the context of economic model predictive control (EMPC), which is a feedback control approach that optimizes plant economics on-line by utilizing a process model. The on-line model identification scheme coupled with the EMPC system was applied to a benchmark catalytic chemical reactor example where the reaction rates decrease with time due to catalyst deactivation. In the presence of such plant variation, more accurate state predictions were made under the proposed EMPC scheme with on-line model identification than under an EMPC for which the model was not updated, which significantly impacted the plant economics.

Estimating structural deformations for inferential control: a disturbance observer approach

Abstract: Increasingly stringent requirements for motion systems lead to a situation where the positioning performance can often not be measured directly and therefore has to be estimated. A typical example is a wafer stage, where the performance is desired at the point-of-exposure on the wafer but the sensors are located at the edge of the wafer stage. Increasingly stringent performance requirements necessitate taking structural deformations, caused by actuation or disturbance forces, into account. The aim of this paper is to develop a disturbance observer approach including an observer relevant model identification approach and to experimentally validate this approach on a prototype motion system. The experimental results confirm that the proposed disturbance observer approach leads to an improved estimation of the unmeasured performance variables.

Adaptive power oscillation damping controller of superconducting magnetic energy storage device for interarea oscillations in power system

Abstract This paper presents an adaptive power oscillation damping (APOD) scheme for the superconducting magnetic energy storage (SMES) device to suppress the interarea oscillation in the inter-connected power system. The APOD scheme is designed based on the generalized predictive control (GPC) and model identification approaches. A recursive least-squares algorithm (RLSA) with a varying forgetting factor is utilized to identify a reduced-order model of the power system online. Based on this identified model, the GPC scheme considering control output constraints can yield an optimal control action by performing an optimization procedure over a prediction horizon. Owing to the usage of the RLSA, the proposed APOD controller can effectively adapt to the variations of operating conditions and parameter uncertainties of the power system. Case studies are undertaken on the New England 10-machine 39-bus power system. Simulation results verify the proposed APOD can consistently provide better damping performance than that of the conventional lead–lag POD, over a wide range of operating conditions and different disturbances.

Controller design with model identification approach in wide area power system

Using wide area monitoring systems (WAMS) offers a possibility for an integrated measurement-based and model-based control, which suits to the operation of large electric power system (EPS), along with online analysis. This study presents studies on fixed-order controller design through model identification approach with use of synchronous measurement data. Firstly, in the study, the coherent generator in each area of large EPS is determined by the mutual information theory. Then, state-space two-input two-output model is identified for the generator that has highest participation factor and thus referred as coherent generator. The model identification algorithms; least-square, instrumental variable and subspace state-space based generalised Poisson moment function are used. Next, WAMS level model is identified between the input controllable variable and speed deviation difference of coherent generator of each area. Finally, a local controller (decentralised) in each coherent area and a centralised controller at WAMS level between two coherent areas are designed by optimisation of the several design functions; H∞ norm, H 2 norm, spectral abscissa and complex stability radius, as much as possible. These controllers feed supplementary control signal in addition to one fed by local conventionally tuned power system stabiliser. The centralised controller at WAMS level is demonstrated to stabilise the speed deviations of each generator between any two areas in the large EPS. The study is investigated with different input signal variables; ΔV ref, ΔPm excited by different pattern of disturbances.

Electrochemical impedance analysis of TiO2 nanotube porous layers based on an alternative representation of impedance data

A new direct approach for model identification is proposed. It allows a more accurate survey of the different relaxation processes yielding distinct time constants in electrochemical impedance spectroscopy. With the help of self-organized TiO2 nanotubes layers grown in pure Ti substrates, it is shown that the conventional impedance data representation methods are susceptible to miss some information that is made accessible by the frequency survey of αeff, the effective CPE exponent parameter. In the specific case of the present work, the alternative representation proposed here shows that the electrochemical behavior of the TiO2 layers exposed to Ringer physiological solution is characterized by up to three time constants related to a simple double-layered structure composed by a dense barrier and the porous nanotubes one, instead of the ubiquitous two time constants mentioned in the literature for the same systems.

Fault Detection and Fault Tolerant Operation of a Five Phase PM Motor Drive Using Adaptive Model Identification Approach

Nowadays, the use of multiphase fault tolerant permanent magnet (PM) machines is a suitable solution for increasing the reliability of high-performance motion control applications. This solution requires fault detection, isolation, and control adaptation. For this reason, a new open switch fault detection scheme in voltage-source inverter supplying a PM motor is presented in this paper. The detection method is based on model identification of the motor phase currents. The detection scheme is fast, robust, general, and capable of detecting multiple faults. The control algorithm, detection method, isolation, and reconfiguration strategies are discussed. After that, the proposed method is implemented in the fault tolerant control of a five-phase brushless direct current (BLDC) motor drive. Simulation results in MATLAB/Simulink are shown to demonstrate the effectiveness of the proposed detection technique. Experimental results on a five-phase fault tolerant BLDC motor validate the theoretical developments.

Identification of pseudo-State Space Models for Batch Processes using Multivariate Statistical Methods

A new methodology to identify models in a pseudo-state space form for batch/fed-batch processes is proposed. The methodology employs historical data from previous batch runs, where a few intermittent measurements of product quality were made, and multivariate statistical methods in order to identify data-based models. Multivariate statistical methods, such as principal components analysis (PCA) and partial least squares (PLS), are being increasingly employed for batch processes model identification due to the advantages they offer over more difficult and time-consuming first-principle modelling techniques. In the proposed model identification approach, predictors are obtained employing PCA and PLS algorithms. Then, after a new vector of pseudo-states is defined, a pseudo-state space model is identified by performing an algebraic manipulation of the PCA and PLS statistical models. The ability of the pseudo-state space models to accurately predict future process variable trajectories is demonstrated by means of a simulation benchmark for penicillin production.

Internal model based robust inversion feedforward and feedback 2DOF control for LPV system with disturbance

To reduce the adverse effects on the control performance and disturbance rejection caused by system uncertainty, a novel internal model based robust inversion feedforward and feedback 2DOF control approach was proposed for LPV system with disturbance. The proposed control approach combines the internal model control and robust inversion based 2DOF control, it utilizes internal model based control to reject external disturbance, utilizes robust inversion 2DOF control to enhance the control resolution and guarantee the system control performance. At first, a LMI synthesis approach for LPV system model identification and a disturbance compensator optimization design method which could minimize H∞ norm of output error caused by disturbance are presented. Then, combined with internal loop for disturbance compensation, a robust inversion feedforward controller is designed by robust inversion approach and the feedback controller which could render the requirements of reference signal tracking performance and robustness satisfied is obtained by the H∞ mixed sensitivity synthesis approach. Finally, atomic force microscopy (AFM) vertical positioning simulation experiments are conducted and the experiment results showed that the proposed control approach could achieve better output performance and disturbance rejection compared with conventional internal model based control and robust inversion based 2DOF control approach.

Fate and transport modelling of urban highway contaminants by a multi-objective evolutionary method

A hybrid mechanistic data-driven approach was used to identify optimal structures for the processes of a mechanistic build-up wash-off model for predicting the continuous pollutographs of various constituents from urban highway surfaces. The mechanistic model is based on mass balance and the advective–dispersive transport of pollutants in runoff. Using the pollutograph data of seven constituents (TSS, DOC, Cr, Cu, Ni, Pb and Zn) collected from highly urbanized highway sites in California, we applied the two-layer model identification approach to find unique optimum functional forms that represent the processes and their optimum parameter values. The comparison of the model results and observed data indicate acceptable agreement for the examined constituents and rain events. The build-up and wash-off model developed using this approach honours the physical processes involved and is a reliable tool for predicting constituent pollutographs as well as understanding the physical and underlying processes.

A model identification scheme for driver-following dynamics in road traffic

The driver-following, or car-following, model is one of the most fundamental driver behavior models that are applied in intelligent transport applications. Its fidelity determines the applicability of microscopic traffic simulators, where the model is often implemented to mimic real traffic. Meanwhile, the behavioral model is fundamental to the development of advanced driving assistance systems (ADAS). This paper develops a dynamic model identification approach based on iterative usage of the extended Kalman Filtering (EKF) algorithm. Among other things, this allows to carry out model identification using a rather general optimization objective on the whole physical states of the following vehicle. In particular, the method is established on the basis of the equivalence between the Kalman filter and the recursive least squares (RLS) method in a specific context of parameter identification. To illustrate the method, two car-following models are studied in numerical experiments using real car-following data. The method has shown advantages in replication and prediction of vehicle dynamics in car-following over the conventional approaches. It has also the potential to be further extended for building tactical driving controllers in intelligent transportation applications.

PID controller design for micro gas turbines using experimental frequency-response data and a linear identification technique

This paper discusses the identification process of engine dynamics and presents derived transfer function models including; thrust, shaft speed, compressor exit pressure and turbine exit temperature in relation to the fuel flow. Model identification approach, presented in this paper, is the first to consider frequency-sweep signals to excite dynamics, as well as to utilise windowing and smoothing techniques to reduce random errors in the spectral estimates. Here, frequency sweep provides a fairly uniform spectral excitation and chirp-z transform warrants the exact determination of the frequency responses that are robust to the uncertainties. The identified models are also validated with the engine responses to doublet input as well as the engine comprehensive thermodynamic model. In addition, a PID controller is designed based on the identified models to effectively control the non-linear system. The results show great performance of the controller, in simulation environment, implemented to the non-linear system.

An industrial tool for identification and control of nonlinear valve characteristics

This work presents an industrial system identification and control design tool in the context of nonlinear valve control as a typical problem in the refinery industry. A parametric SISO Hammerstein model structure is assumed. Two Hammerstein model identification approaches are implemented and integrated into an efficient tool: Prediction Error Minimization (PEM) as well as a closed-form method involving a Least-Squares and a Singular Value Decomposition part. Parameter uncertainty analysis methods are developed and presented. The second focus of the tool is to give comprehensive support in nonlinear control design and validation. For invertible static nonlinear maps, the Hammerstein structure can be transformed to a globally linear control design problem. This is exploited to design scheduled PID control laws. A series of tools for validation, closed-loop analysis, and tuning is developed and shown. The tool implementation is outlined and the functionality is demonstrated by a series of simulation results highlighting each major feature, based on artificially generated test data. The user is thus equipped with an efficient, state-of-the-art numerical tool to support control engineering for this class of nonlinear control problems.