System Identification Tools Research Articles

Elucidating Xylose Metabolism of Scheffersomyces stipitis for Lignocellulosic Ethanol Production

The conversion of pentose to ethanol is one of the major barriers of industrializing the lignocellulosic ethanol processes. As one of the most promising native strains for pentose fermentation, Scheffersomyces stipitis (formerly known as Pichia stipitis) has been widely studied for its xylose fermentation. In spite of the abundant experimental evidence regarding ethanol and byproducts production under various aeration conditions, the mathematical descriptions of the processes are rare. In this work, a constraint-based metabolic network model for the central carbon metabolism of S. stipitis was reconstructed by integrating genomic (S. stipitis v2.0, KEGG), biochemical (ChEBI, PubChem), and physiological information available for this microorganism and other related yeast. The model consists of the stoichiometry of metabolic reactions, biosynthetic requirements for growth, and other constraints. Flux balance analysis is applied to characterize the phenotypic behavior of S. stipitis grown on xylose. The model predictions are in good agreement with published experimental results. To understand the effect of redox balance on xylose fermentation, we propose a system identification-based metabolic analysis framework to extract biological knowledge embedded in a series of designed in silico experiments. In the proposed framework, we first design in silico experiments to perturb the metabolic network in order to investigate the interested properties and then perform system identification, whereby applying principal component analysis (PCA) to the data generated by the designed in silico experiments. By combining the in silico perturbation experiments with system identification tools, biologically meaningful information contained in the complex network structure can be decomposed and translated into easily interpretable information that is useful for biologist. The PCA analysis identifies the phenotypic changes caused by oxygen supply and reveals key metabolic reactions related to redox homeostasis in different phenotypes. In addition, the influence of the cofactor preference of key enzyme (xylose reductase) in xylose metabolism is investigated using the proposed approach, and the results provide important insights on cofactor engineering of xylose metabolism.

Modeling the Biological Nanopore Instrument for Biomolecular State Estimation

The nanopore is a powerful tool for probing biomolecular interactions at the single-molecule level, and shows great promise commercially as a next-generation deoxyribonucleic acid (DNA) sequencing technology. Coupling active voltage control with the nanopore has expanded its capabilities, for example, by allowing precise manipulation of DNA-enzyme complexes at millisecond timescales. However, any change in voltage excites capacitance in the system and results in masking the molecule's contribution to the measured current. To improve active control capabilities, a method is needed for continuous monitoring of the molecule's contribution to the current during voltage-varying experiments. The method must be able to separate the capacitive effects from the channel conductance, which is the parameter that can be used to infer the state of the molecule in the pore. The contributions of this paper are: 1) to develop a dynamic model of the nanopore instrument which includes capacitance and conductance parameters and 2) to develop model-based algorithms for estimating the conductance parameter during voltage varying experiments. First, grey- and black-box state-space models are estimated and compared using nanopore experimental data and system identification tools. Next, a validated grey-box model is used to derive two methods for estimating the channel conductance under voltage-varying conditions: one based on least-squares, and one based on the extended Kalman filter. In simulations and experiments, the Kalman filter outperforms the simpler least-squares method.

Robust Control Analysis Techniques Applied to a Mini Aircraft

This paper presents a methodology for the development of an UAV Autopilot System, based on modeling, instrumentation and a control law, with the purpose of stabilizing the airplane attitude. Firstly, were made experimental test flights by manual control for obtaining the longitudinal model parameters using the Matlab toolbox called "System Identification Tool". The model is represented by a transfer function of the pitch angle with respect to the elevator into an open-loop system. The system is stabilized using a PD controller (derivative-proportional) tuned with poles placement method. The zero exclusion principle is used to compute the robustness margin of the closed-loop system. Different experiments were made in order to validate theoretical results which are presented at the end of this work.

Artificial Neural Network–Based System Identification for a Single-Shaft Gas Turbine

During recent decades, artificial intelligence has been employed as a powerful tool for identification of complex industrial systems with nonlinear dynamics, such as gas turbines (GT). In this study, a methodology based on artificial neural network (ANN) techniques was developed for offline system identification of a low-power gas turbine. The processed data was obtained from a SIMULINK model of a gas turbine in matlab environment. A comprehensive computer program code was generated and run in matlab for creating and training different ANN models with feed-forward multilayer perceptron (MLP) structure. The code consisted of various training functions, different number of neurons as well as a variety of transfer (activation) functions for hidden and output layers of the network. It was shown that the optimal model for a two-layer network with MLP structure consisted of 20 neurons in its hidden layer and used trainlm as its training function, as well as tansig and logsid as its transfer functions for the hidden and output layers. It was also observed that trainlm has a superior performance in terms of minimum mean squared error (MSE) compared with each of the other training functions. The resulting model could predict performance of the system with high accuracy. The methodology provides a comprehensive view of the performance of over 18,720 ANN models for system identification of the single-shaft gas turbine. One can use the optimal ANN model from this study when training from real data obtained from this type of GT. This is particularly useful when real data is only available over a limited operational range.

A Novel Fault Diagnostics and Prediction Scheme Using a Nonlinear Observer With Artificial Immune System as an Online Approximator

In this paper, an observer-based fault diagnostics and prediction (FDP) scheme for a class of nonlinear discrete-time systems via output measurements is introduced by using artificial immune system (AIS) and a robust adaptive term. Traditionally, AIS was considered as an offline tool for system identification and pattern recognition whereas here AIS is utilized as an online approximator in discrete-time (OLAD) in a fault detection (FD) observer. A fault is detected when the output residual exceeds a predefined threshold. Upon detection, the OLAD is initiated to learn the unknown fault dynamics online while the robust adaptive term ensure asymptotic convergence of the output residual for a state fault whereas a bounded result for an output fault. Additionally, a mathematical equation is introduced to estimate the time-to-failure (TTF) by using the output residual and the estimated fault parameters. Finally, the performance of the proposed FDP scheme is demonstrated on an axial piston pump hardware test-bed.

Multi-model identification of a fractional non linear system

Takagi-Sugeno fuzzy models have proved to be a powerful tool for non-linear system identification. Due to their simplicity, linear integer sub-models are mainly used in such structure. In this paper multi-model approach, based on Takagi-Sugeno form is used to perform identification of fractional non-linear systems, applying fractional local models. In this purpose, an output error method is used, and the algorithm is extended to the fractional case. Numerical simulations are performed in order to analyze the two sub-models: integer and fractional, fitting ability to approximate the non linear system behaviour. The validation of the method is carried out on real data of the heat diffusion.

Real time optimization for quality control of batch thermal sterilization of prepackaged foods

In this contribution, we present a distributed decision-making architecture for control to optimally command thermal sterilization, despite process uncertainty or unexpected process disturbances. The control structure combines in a synchronous way modeling and simulation environments with efficient system identification and dynamic optimization tools and methods. Process simulation provides a complete dynamic description of the current status of the operation, including the evolution of temperature and pressure in the retort unit as well as temporal and spatial distribution of temperature and quality or safety parameters within the product. Such virtual representation will be regularly confronted with plant measurements to quantify the degree of discrepancy (uncertainty) between real plant and models and react accordingly when such discrepancy becomes unacceptable by re-estimating plant parameters, either during the cycle or from batch to batch. The virtual plant will be also accessed by the regulatory system as well as the dynamic optimization module. In the first instance to estimate unmeasured states related with the product status (e.g. temperature in the product or lethality) under feed-back control. In the second, to continuously recompute optimal cycle profiles so to respond to unexpected disturbances or deviations from the prescribed safety constraints while maximizing quality attributes. Experimental evidence of the complete control system performance will be given on the operation of a pilot plant prototype.

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.

A novel artificial bee colony algorithm with space contraction for unknown parameters identification and time-delays of chaotic systems

This paper is concerned with the uncertain parameters and time-delay of nonlinear chaotic systems, which is of vital significance in chaos control and synchronization. In this article, a novel artificial bee colony algorithm (ABC), with space contraction and optimization technique based on the foraging behavior of honeybees, is newly proposed to solve the estimation problem via a nonnegative multi-modal nonlinear optimization, which finds a best combination of parameters such that an objective function is minimized. The illustrative examples, in Lorénz, Chen, Lü chaos systems and Mackey–Glass time-delay chaos system, are given though ABC and ABC with space contraction self-adaptively (ABCSC) respectively. Simulation are done and comparisons to some existing results by a recent version of ABC, HTCMIABC demonstrate that ABCSC is superior to ABC and HTCMIABC for unknown parameters and time-delays of the chaotic systems accurately and effectively. And it is a promising tool for chaotic system identification as well as other numerical optimization problems in mathematics.

A simple algorithm for stable order reduction of z-domain Laguerre models

Discrete-time Laguerre series are a well known and efficient tool in system identification and modeling. This paper presents a simple solution for stable and accurate order reduction of systems described by a Laguerre model.

Online Bayesian Time-varying Parameter Estimation of HIV-1 Time-series

Nonlinear Bayesian filtering offers various online tools for system identification of (parametric) ordinary differential equation models. Since parameters may change with time, it is a relevant question to assess how well time-varying parameters can be estimated. For this purpose we tested two filtering methods, Extended Kalman Filter and Particle Filter for joint state and time-varying parameter estimation on a dynamic model of HIV-1 virus immune response. After evaluating the methods on simulated time-series we applied them to clinical datasets. Estimated time-varying parameters on clinical data are consistent with previously reported results with offline algorithms.

A new method for aquifer system identification and parameter estimation

AbstractThe standard practice for assessing aquifer parameters is to match groundwater drawdown data obtained during pumping tests against theoretical well function curves specific to the aquifer system being tested. The shape of the curve derived from the logarithmic time derivative of the drawdown data is also very frequently used as a diagnostic tool to identify the aquifer system in which the pumping test is being conducted. The present study investigates the incremental area method (IAM) to serve as an alternative diagnostic tool for the aquifer system identification as well as a supplement to the aquifer parameter estimation procedure. The IAM based diagnostic curves for ideal confined, leaky, bounded and unconfined aquifers have been derived as part of this study, and individual features of the plots have been identified. These features were noted to be unique to each aquifer setting, which could be used for rapid evaluation of the aquifer system. The effectiveness of the IAM methodology was investigated by analyzing field data for various aquifer settings including leaky, unconfined, bounded and heterogeneous conditions. The results showed that the proposed approach is a viable method for use as a diagnostic tool to identify the aquifer system characteristics as well as to support the estimation of the hydraulic parameters obtained from standard curve matching procedures. Copyright © 2012 John Wiley & Sons, Ltd.

Dynamic State and Parameter Estimation Applied to Neuromorphic Systems

Neuroscientists often propose detailed computational models to probe the properties of the neural systems they study. With the advent of neuromorphic engineering, there is an increasing number of hardware electronic analogs of biological neural systems being proposed as well. However, for both biological and hardware systems, it is often difficult to estimate the parameters of the model so that they are meaningful to the experimental system under study, especially when these models involve a large number of states and parameters that cannot be simultaneously measured. We have developed a procedure to solve this problem in the context of interacting neural populations using a recently developed dynamic state and parameter estimation (DSPE) technique. This technique uses synchronization as a tool for dynamically coupling experimentally measured data to its corresponding model to determine its parameters and internal state variables. Typically experimental data are obtained from the biological neural system and the model is simulated in software; here we show that this technique is also efficient in validating proposed network models for neuromorphic spike-based very large-scale integration (VLSI) chips and that it is able to systematically extract network parameters such as synaptic weights, time constants, and other variables that are not accessible by direct observation. Our results suggest that this method can become a very useful tool for model-based identification and configuration of neuromorphic multichip VLSI systems.

I-pIDtune: An interactive tool for integrated system identification and PID control

This paper describes i-pIDtune, an interactive software tool that integrates system identification and PID controller design. The tool supports experimental design and execution under plant-friendly conditions, high-order ARX estimation, and control-relevant model reduction leading to models that comply with the IMC-PID tuning rules. All these stages are depicted simultaneously and interactively in one screen. Thus, i-pIDtune allows to display both open- and closed-loop responses of the estimated models and important control-relevant validation criteria, what enables the user to readily assess how design variable choices, control performance requirements and model error can impact the achievable closed-loop performance from a restricted complexity model estimated under noisy conditions.

On the Parallelization of Vector Fitting Algorithms

The so-called vector fitting (VF) algorithm has gained much popularity over the last few years. This technique provides a very effective system identification tool that, starting from input-output responses of a linear and time-invariant system, computes a rational approximation of its transfer matrix. The latter is routinely used to synthesize compact broadband equivalent circuits or state-space models of possibly complex interconnects at the chip, package, board, or even system level. The VF algorithm is based on a combination of iterative linear least squares solutions and eigensolutions, and proves robust and reliable. A potential weak point of VF is its relatively poor scalability with the complexity of the structure under modeling. When the number of input-output ports is very large (one hundred or more, as in the case of power buses or packages), the excessive computational requirements may hinder VF performance and prevent its successful application. In this paper, we address these issues by first presenting a detailed analysis of the computational cost of all the algorithm parts. The results show a very good potential for VF parallelization for multicore hardware, and suggest a few alternative parallelization strategies. Each of these strategies is described in detail. Finally, numerical results and comparisons are provided on a large set of industrial benchmarks. These results demonstrate excellent scalability and speedup factors for the parallel sections of the algorithm, leading to a drastic reduction in overall runtime.

Design of Robust Digital PID Controller for H-Bridge Soft-Switching Boost Converter

In this paper, a robust digital proportional-integral-derivative (PID) controller is proposed for the H-bridge soft-switching boost converter (HSBC). This digital PID controller is designed to ensure load voltage regulation as well as to give robust performance with step loads and source rejection. The mathematical models of the H-bridge boost converter are formulated, using the system identification tool, and then used in digital PID design. Here, this compensator is designed in the direct digital domain according to a pole placement approach that uses sensitivity function shaping in order to ensure closed-loop converter system stability as well as robust performance against converter parameter uncertainties. To confirm this, design simulations have been carried out on a 60-W 24-42-V HSBC. The experimental results are provided to validate the robust controller design concept.

Control-Oriented Model for Intake Shock Position Dynamics in Ramjet Engine

R AMJET-POWERED engines have a history of over 100 years [1]. The secret to efficiency, safety, and performance of ramjet combustion systems has been the correct location and control of the terminal shock in the intake duct [2]. The position of the intake shock is affected by perturbations propagating upstream from the combustor [3] and from disturbances in the freestream [4]. These can lead to the familiar instability problems of unstart or buzz [5,6]. At the same time, these very instabilities can be controlled by pressure perturbations injected by suitably manipulating the exhaust nozzle throat area [7]. To properly evaluate such a control, it is necessary to obtain amodel for the ramjet engine including the intake, combustor, and exhaust nozzle. The issue of shock position control has always been an interesting one [8], but accurately sensing the position of the intake shock for the purpose of active control has been a challenge. In recent years, though, the problemhas attractedmuch attention [9,10], and building on the earlier work on the dynamics of shocks in ducts [3,11], models that may be used for designing intake shock position controllers have been obtained [12,13].However, itmay be noted that all thesemodels were limited to the intake alone and hence could not be directly used to study the effect of the exhaust nozzle throat area variation on the intake shock location. A model that couples the intake with the combustor and exhaust nozzle for a ramjet was first realized recently by our coworkers [14], and it was used [15] to derive a control law for the intake shock location using the nozzle throat area as input. A significant feature of the model in [14] was the use of time lags to capture the physics of upstream anddownstreampropagatingwaves between the intake and combustor. However, these time lags were applied to the primitive variables for simplicity instead of the specific pressure or entropy waves [12]. Second, despite the relatively low order of the global model in [14], the model for each component was multiparameter and nonlinear, in order to capture the various physical phenomena in the intake and combustor. Thus, in deriving the control law in [15], local linearized reduced-order models at several operating points had to be obtained using a system identification tool. The system identification is, unfortunately, a mathematical procedure that results in a set of states that cannot be easily related to the physical variables, thus masking the physical relationships inherent in the system. Control of the terminal shock positionwas obtained indirectly in [15] by defining a related parameter called intake backpressure margin. The present Note differs from these previous works in three significantways. First, wewrite an explicit equation for the dynamics of the intake shock location in a coupled model of the intake and the combustor plus nozzle. Second, a single time-lag parameter is obtained numerically, and applied directly to the shock position variable. Third, the various component models are written using standard quasi-one-dimensional flow relations in such a manner that the physical relationships they represent are apparent. These features make the present model more suitable for controller design with the objective of regulating the intake shock position; hence, it is called a control-oriented model.

Fast Multi-Order Stochastic Subspace Identification⋆

AbstractStochastic subspace identification methods are an efficient tool for system identification of mechanical systems in Operational Modal Analysis (OMA), where modal parameters are estimated from measured vibrational data of a structure. System identification is usually done for many successive model orders, as the true system order is unknown and identification results at different model orders need to be compared to distinguish true structural modes from spurious modes in so-called stabilization diagrams. In this paper, this multi-order system identification with the subspace-based identification algorithms is studied and an efficient algorithm to estimate the system matrices at multiple model orders is derived.

DYNAMIC SYSTEM IDENTIFICATION USING TIME-DELAY FEEDFORWARD NEURAL NETWORKS: APPLICATION TO DC MOTOR

The universal function approximation capabilities of multilayer feedforward neural networks make it a popular choice for modeling dynamic systems. In this paper, identification of dynamic system using time-delay feedforward neural networks with application to DC motor as a case study has been developed. The developed neural network model is a three layer network with nonlinear (sigmoid) activation functions in the hidden layer and linear output layer with input-output delays. Simulation results showed that the neural networks are promising tool for dynamic system identification

Dipteran insect flight dynamics. Part 2: Lateral–directional motion about hover

The purpose of this study is to determine computationally tractable models describing the lateral–directional motion of a Drosophila-like dipteran insect, which may then be used to estimate the requirements for flight control and stabilization. This study continues the work begun in Faruque and Humbert (2010) to extend the quasi-steady aerodynamics model via inclusion of perturbations from the hover state. The aerodynamics model is considered as forcing upon rigid body dynamics, and frequency-based system identification tools used to derive the models. The analysis indicates two stable real poles, and two very lightly damped and nearly unstable complex poles describing a decoupling of roll/sideslip oscillatory motion from a first order subsidence yaw behavior. The results are presented with uncertainty variation for both a smaller male and larger female phenotype.