System Identification Approach Research Articles

Dynamic modeling of Boiler drum using nonlinear system identification approach

Thermal power generation plays a major role in meeting present-day power demands. Efficient and economic controls of crucial elements like steam boiler section need superior control mechanism to ensure better efficiency. Identification of model, dynamics is a formidable task due to the strong coupling of process variables, nonlinearities in the dynamics and constraints on the control inputs. This paper presents the application of nonlinear modeling techniques to estimate boiler drum dynamics in terms of nonlinear model framework which is one of the important requirements for the design of drum water level controls schemes. The historical plant data for estimation is collected from boiler of 210 MW, Stage-II, Boiler-4 of Dr. Narla Tata Rao Thermal power Corporation (NTTPC) Ibrahimpatnam, Vijayawada. Different operating conditions are selected to estimate the nonlinear dynamics of the boiler to check with the existing Hammerstein, Wiener, model frame works. Later the nonlinear framework is modified to estimate other nonlinearities present in the system. The estimated model responses are presented to check the boiler dynamics.

Exergy Analysis of a Passive Thermal System Using Hammerstein–Wiener Estimation

Abstract A nonlinear system identification approach was used to exploit the nonlinearly in the exergy of the system and reduce it into two or more interconnected elements. The Hammerstein–Wiener (H-W) methodology was adopted to describe the dynamics of a passive thermal system using a combination of nonlinear and linear blocks. Here, the linear block is a discrete transfer function, which symbolizes the dynamic component of the model. The proposed model was validated using the state functions measured for the solar air collector. The mean absolute percentage error (MAPE) for enthalpy changes falls in the domain of −0.01% to 0.01%, whereas it varied from −0.06% to 0.02% as the entropy of the system changed with time. Similarly, the MAPE encountered while evaluating the exergy of the system was in the closed interval of −0.066% to −0.0017%. The average exergy gain by the H-W model across the Ist and IInd passages was, respectively, 0.90 kJ · kg−1 (8.10 g · s−1), 0.61 kJ · kg−1 (10.10 g · s−1) and 0.46 kJ · kg−1 (12.10 g · s−1); and 0.57 kJ · kg−1 (8.10 g · s−1), 0.48 kJ · kg−1 (10.10 g · s−1), and 0.79 kJ · kg−1 (12.10 g · s−1). The proposed model exhibited good fitting with the validation data.

Orientation Control Design of a Telepresence Robot: An Experimental Verification in Healthcare System

Automation in the modern world has become a necessity for humans. Intelligent mobile robots have become necessary to perform various complex tasks in healthcare and industry environments. Mobile robots have gained attention during the pandemic; human–robot interaction has become vibrant. However, there are many challenges in obtaining human–robot interactions regarding maneuverability, controllability, stability, drive layout and autonomy. In this paper, we proposed a stability and control design for a telepresence robot called auto-MERLIN. The proposed design simulated and experimentally verified self-localization and maneuverability in a hazardous environment. A model from Rieckert and Schunck was initially considered to design the control system parameters. The system identification approach was then used to derive the mathematical relationship between the manipulated variable of robot orientation control. The theoretical model of the robot mechanics and associated control were developed. A design model was successfully implemented, analyzed mathematically, used to build the hardware and tested experimentally. Each level takes on excellent tasks for the development of auto-MERLIN. A higher level always uses the services of lower levels to carry out its functions. The proposed approach is comparatively simple, less expensive and easily deployable compared to previous methods. The experimental results showed that the robot is functionally complete in all aspects. A test drive was performed over a given path to evaluate the hardware, and the results were presented. Simulation and experimental results showed that the target path is maintained quite well.

Bridging similar ships’ dynamics for safeguarding the system identification of maneuvering models

System identification (SI) approach identifies ship maneuvering models using data from free-running maneuvers. As it is impossible to comprehensively validate model performance, a reliable model must be built within a consistent range with similar ships’ models. However, in the SI approach, models are greatly influenced by the dataset design and feature selection, leading to the instability of model identification and the failure to ensure such consistency. To address this issue, our new idea introduces similar ship’s model into model identification as knowledge foundation. First, we select a similar ship with three-step procedure, then build a new model by refining model parameters for a similar ship. Such a refinement is conducted with additional l2 regularization term in ridge regression, which balances knowledge and data-driven refinement. Designing the safe range of such refinement with hyperparameters, this study helps designers find a robust-and-accurate model within designed safe zones. In simulation experiments, we built knowledge connection between 161 m and 175 m container ships. By using the well-validated model of the former, a robust-and-accurate model for the latter was easily built by using a limited dataset, resulting in excellent model performance. This study makes the SI approach more promising by incorporating knowledge connections between similar ships.

Fuzzy Fractional Order PID Tuned via PSO for a Pneumatic Actuator with Ball Beam (PABB) System

This study aims to improve the performance of a pneumatic positioning system by designing a control system based on Fuzzy Fractional Order Proportional Integral Derivative (Fuzzy FOPID) controllers. The pneumatic system’s mathematical model was obtained using a system identification approach, and the Fuzzy FOPID controller was optimized using a PSO algorithm to achieve a balance between performance and robustness. The control system’s performance was compared to that of a Fuzzy PID controller through real-time experimental results, which showed that the former provided better rapidity, stability, and precision. The proposed control system was applied to a pneumatically actuated ball and beam (PABB) system, where a Fuzzy FOPID controller was used for the inner loop and another Fuzzy FOPID controller was used for the outer loop. The results demonstrated that the intelligent pneumatic actuator, when coupled with a Fuzzy FOPID controller, can accurately and robustly control the positioning of the ball and beam system.

On the combined estimation of the parameters and the states of fractional-order systems

In this article, we propose an approach for system identification for a class of discrete-time fractional-order Hammerstein systems using only input–output data. Using a combined state and parameter estimation approach, we develop an algorithm serving to estimate simultaneously, the system parameters, the system orders, and the system states. By minimizing the defining criterion, which is non-convex and nonlinear in the parameters, the model parameters are estimated using the recursive least squares and the Levenberg–Marquardt algorithms. Next, the system states are estimated using the Luenberger observer. Then, the convergence analysis of the proposed algorithm is proved. Finally, the Monte Carlo simulation analysis and an application to Ultracapacitor system are used to demonstrate the effectiveness of the suggested method.

Data-driven system identification of hydrodynamic maneuvering coefficients from free-running tests

A data-driven system identification approach was developed to identify the hydrodynamic coefficients of a mathematical maneuvering model. The method, developed primarily for use in the context of autonomous shipping, solved the ship motion equations using measurements from free-running model tests, whereby an efficient recently developed Euler equation-based numerical approach determined the zero-frequency added masses. The method is simple and robust and incorporates the physical properties of hydrodynamic forces to enforce a physically realistic solution. The method was verified and validated with free-running maneuver tests. The predicted ship kinematics and trajectories compared favorably with the measurements. The potential of the method was demonstrated.

A Loewner-Based System Identification and Structural Health Monitoring Approach for Mechanical Systems

Data-driven structural health monitoring (SHM) requires precise estimates of the target system behaviour. In this sense, SHM by means of modal parameters is strictly linked to system identification (SI). However, existing frequency-domain SI techniques have several theoretical and practical drawbacks. This paper proposes using an input-output system identification technique based on rational interpolation, known as the Loewner framework (LF), to estimate the modal properties of mechanical systems. Pioneeringly, the Loewner framework mode shapes and natural frequencies estimated by LF are then applied as damage-sensitive features for damage detection. To assess its capability, the Loewner framework is validated on both numerical and experimental datasets and compared to established system identification techniques. Promising results are achieved in terms of accuracy and reliability.

Simulation-based study on active tilt controlled narrow tilting three-wheeler considering road bank angle

The issues like traffic congestion, parking space and vehicular emission can be reduced by adopting electric narrow tilting three-wheelers, which can also provide a cost-effective and safe mode of transport for low-income people. According to energy technology perspectives 2017, well-to-wheel global greenhouse gas emissions can be reduced up to 1.3 GtCO2-eq by 2060 by adopting electric mobility. This study presents a new and modified SDTC strategy and performance parameters that takes Road Bank Angle (RBA) into account so as to exhibit the adverse effects on rollover safety, ride comfort in terms of the perceived lateral acceleration and desired trajectory. Besides, this study identifies the importance of the RBA consideration in active tilt controlled vehicle for the scenario of parking on sloped surface. Simple electronic control architecture of the SDTC approach with real-time RBA consideration is proposed to address the limitations of SDTC and its supplementary advantages. This study evaluates the scenario of constant turn manoeuvre and vehicle parking on the banked roads. A multi-body model is used to capture the dynamic characteristics of the vehicle, whereas a characteristic of the Tilt Actuation System (TAS) and Active Steering (AS) system is incorporated in the form of a Transfer Function (TF) estimated using a system identification approach. The results indicate that the inclusion of a real-time RBA in the SDTC enables it to maintain steady-state torque and perceived lateral acceleration close to zero with the proposed controlled strategy, and it also keeps the vehicle upright on the sloped surface. It may be concluded that RBA compensation improves the ease of handling the vehicle while parking or riding on a sloped surface without significant control system alterations, which also makes it a viable alternative for physically disabled people.

A Single Objective Crow Search for Modelling of Horizontal Flexible Plate Structure

The magnificent features of flexible plate structure, including lightweight and high-speed response, resulted in additional market demand, especially in the automotive and manufacturing industries. Nevertheless, the structure may incur structural damage and performance degradation when the system encounters excessive vibration. Therefore, a system identification approach utilising a metaheuristic algorithm via crow search to develop a horizontal flexible plate (HFP) model for vibration control is introduced in this paper. Crow search (CS) is a modern algorithm inspired by a crow’s intellectual operation to store additional food and memorise the food storage location. In this study, CS is employed to optimise the objective function, which is the mean squared error for accomplishing a precise predicted model in replicating the dynamic response of the actual structure. Hence, the preliminary action for modelling using this approach is designing and fabricating an HFP rig for experimentally gathering the real input-output vibration data. After that, the mathematical modelling utilising the CS algorithm was implemented using a parametric model structure. Finally, the best-fit model is chosen for the representation of the HFP based on the lowest mean squared error, correlation test within a 95% confidence level and stability in a pole-zero plot. The simulation result reveals that the CS algorithm with a second-order estimated model accomplished a minimum MSE of 1.1168 × 10-5, an unbiased correlation test and excellent stability for the HFP structure.

Adaptive Infinite Impulse Response System Identification Using Elitist Teaching-Learning- Based Optimization Algorithm

Infinite Impulse Response (IIR) systems identification is complicated by traditional learning approaches. When reduced-order adaptive models are utilised for such identification, the performance suffers dramatically. The IIR system is identified as an optimization issue in this study. For system identification challenges, a novel population-based technique known as Elitist teacher learner-based optimization (ETLBO) is used to calculate the best coefficients of unknown infinite impulse response (IIR) systems. The MSE function is minimised and the optimal coefficients of an unknown IIR system are found in the system identification problem. The MSE is the difference between an adaptive IIR system's outputs and an unknown IIR system's outputs. For the unknown system coefficients of the same order and decreased order cases, exhaustive simulations have been performed. In terms of mean square error, convergence speed, and coefficient estimation, the results of actual and reduced-order identification for the standard system using the novel method outperform state-of-the-art techniques. For approximating the same-order and reduced-order IIR systems, four benchmark functions are examined utilizing GA, PSO, CSO, and BA. To demonstrate the improvements, the approach is evaluated on three conventional IIR systems of 2nd, 3rd, and 4th order models. On the basis of computing the mean square error (MSE) and fitness function, the suggested ETLBO approach for system identification is proven to be the best among others. Furthermore, it is confirmed that the suggested ETLBO method outperforms some of the other known system identification strategies. Finally, the efficiency of the dynamic nature of the control parameters of DE, TLBO, and BA in finding near parameter values of unknown systems is demonstrated through comparison data. The simulation results show that the suggested system identification approach outperforms the current methods for system identification.

A Two-Step FE Model Updating Approach for System and Damage Identification of Prestressed Bridge Girders

This study presents a two-step FE model updating approach for health monitoring and damage identification of prestressed concrete girder bridges. To reduce the effects of modeling error in the model updating process, in the first step, modal-based model updating is used to estimate linear model parameters mainly related to the stiffness of boundary conditions and material properties. In the second step, a time-domain model updating is carried out using acceleration data to refine parameters accounting for the nonlinear response behavior of the bridge. In this step, boundary conditions are fixed at their final estimates using modal-based model updating. To prevent the convergence of updating algorithm to local solutions, the initial estimates for nonlinear material properties are selected based on the first-step model updating results. To validate the applicability of the two-step FE model updating approach, a series of forced-vibration experiments are designed and carried out on a pair of full-scale decommissioned and deteriorated prestressed bridge I-girders. In the first step, parameters related to boundary conditions, including stiffness of supports and coupling beams, as well as material properties, including initial stiffness of concrete material, are estimated. In the second step, concrete compressive strength and damping properties are updated. The final estimates of the concrete compressive strength are used to infer the extent of damage in the girders. The obtained results agree with the literature regarding the extent of reduction in concrete compressive strength in deteriorated concrete structures.

System Identification Approach for eVTOL Aircraft Demonstrated Using Simulated Flight Data

This paper describes a system identification method for electric vertical takeoff and landing (eVTOL) aircraft. The approach merges fixed-wing and rotary-wing modeling techniques with new strategies to develop a modeling method for eVTOL vehicles using flight-test data. The eVTOL aircraft system identification approach is demonstrated through application to the NASA Langley Aerodrome No. 8 tandem tilt-wing, distributed electric propulsion aircraft using a high-fidelity flight dynamics simulation. Orthogonal phase-optimized multisine inputs are applied to each control surface and propulsor at numerous flight conditions throughout the flight envelope to collect informative flight data. An aero-propulsive model is identified at each flight condition using the equation-error method in the frequency domain. The local model parameters are then blended to create a global model across the nominal flight envelope. The identified models are shown to provide a good fit to modeling data with good prediction capability. The methodology is developed with a discussion of unique eVTOL vehicle aerodynamic characteristics and practical strategies intended to inform future flight-based system identification efforts for eVTOL aircraft.

Parsimonious Identification of Continuous-Time Systems: A Block-Coordinate Descent Approach⋆

The identification of electrical, mechanical, and biological systems using data can benefit greatly from prior knowledge extracted from physical modeling. Parametric continuous-time identification methods can naturally incorporate this knowledge, which leads to inter-pretable and parsimonious models. However, some applications lead to model structures that lack parsimonious descriptions using unfactored transfer functions, which are commonly used in standard direct approaches for continuous-time system identification. In this paper we characterize this parsimony problem, and develop a block-coordinate descent algorithm that delivers parsimonious models by sequentially estimating an additive decomposition of the transfer function of interest. Numerical simulations show the efficacy of the proposed approach.

Closed-Loop System Identification Approach of the Inertial Models

The mathematical model that approximates the dynamics of the industrial process is essential for the efficient synthesis of control algorithms in industrial applications. The model of the process can be obtained according to the identification procedures in the open-loop, or in the closed-loop. In the open-loop, the identification methods are well known and offer good process approximation, which is not valid for the closed-loop identification, when the system provides the feedback output and doesn’t permit it to be identified in the open-loop. This paper offers an approach for experimental identification in the closed-loop, which supposes the approximation of the process with inertial models, with or without time delay and astatism. The coefficients are calculated based on the values of the critical transfer coefficient and period of the underdamped response of the closed-loop system with P controller, when system achieves the limit of stability. Finally, the closed-loop identification was verified by the computer simulation and the obtained results demonstrated, that the identification procedure in the closed-loop offers good results in process of estimation of the model of the process.

Neural network and Sparse identification of Nonlinear Dynamics Integrated Algorithm for Digital Twin identification

Digital twins play a critical role in simulating industrial manufacturing systems to increase productivity and reduce time spent on troubleshooting. Owing to the complexity of real-world industrial systems, automatic sparse identification has emerged as an attractive approach to perform digital twin modelling. The sparse identification of nonlinear dynamics (SINDy) is a machine learning algorithm that performs feature engineering by generating a model term library and then solves a sparse regression problem between the objective outputs and the generated features. By solving a linear-in-parameter sparse regression problem, SINDy provides automatic discovery of system governing equations. However, the performance of SINDy-based algorithms may decline dramatically when applied to identify complex nonlinear relationships, such as implicit relationships. The substantial number of input variables for a real industrial process may further complicate the modelling procedure. We therefore propose the neural network and SINDy integrated algorithm to automatically select the critical features from a model term library and utilize the neural network to capture the process nonlinearity that cannot be captured by a linear-in-parameter model. SINDy performs feature generation considering both numerical methods and first-principles knowledge, making the proposed algorithm a hybrid system identification approach. A diesel hydrotreating unit case study with 37 input variables is analyzed in this paper to demonstrate the advantages of the proposed algorithm for nonlinear digital twin identification. By combining the advantages from both SINDy and neural networks, the proposed algorithm is able to improve the output prediction accuracy for all the three objectives.

Estimation of Soil–Structure Model Parameters for the Millikan Library Building Using a Sequential Bayesian Finite Element Model Updating Technique

We present a finite element model updating technique for soil–structure system identification of the Millikan Library building using the seismic data recorded during the 2002 Yorba Linda earthquake. A detailed finite element (FE) model of the Millikan Library building is developed in OpenSees and updated using a sequential Bayesian estimation approach for joint parameter and input identification. A two-step system identification approach is devised. First, the fixed-base structural model is updated to estimate the structural model parameters (including effective elastic modulus of structural components, distributed floor mass, and Rayleigh damping parameters) and some uncertain components of the foundation-level motion. Then, the identified structural model is used for soil–structure model updating wherein the Rayleigh damping parameters, the stiffness and viscosity of the soil subsystem (modeled using a substructure approach), and the foundation input motions (FIMs) are estimated. The identified model parameters are compared with state-of-practice recommendations. While a specific application is made for the Millikan Library, the present work offers a framework for integrating large-scale FE models with measurement data for model inversion. By utilizing this framework for different civil structures and earthquake records, key structural model parameters can be estimated from the real-world recorded data, which can subsequently be used for assessing and improving, as necessary, state-of-the-art seismic analysis and structural modeling techniques. This paper presents an effort towards using real-world measurements for large-scale FE model updating in the soil and structure, uniform soil time domain for joint parameter and input estimation, and thus paves the way for future applications in system identification, health monitoring, and diagnosis of civil structures.

The Thermal Network Approach to Model Occupants’ Heat and CO2 Generation Interactions: A Case Study in an Office Building in Panama

This investigation proposes a methodology to predict indoor air temperature and CO2 levels. For this, a two-occupant office inside a building in the Technological University of Panama is taken as a case study and modeled in Designbuilder simulation software validated via experimental data. Here, a mathematical model that considers internal heat gains by the occupants and CO2 emissions, including physical characteristics and activities developed, is constructed via the thermal network (RC) and system identification approaches. Three linear grey-box models are identified: a 4R2C for cooling system mode, a 3R2C for natural ventilation conditions, and a 1R1C for CO2 model. The results showed that the identified model is useful for estimating the indoor air temperature under both modes: “natural ventilation on” and “cooling system on,” in separated situations. Thus, it is determined that by incorporating the internal heat gain generated by the occupant in the model identification process, the data set is closer to real values than implementing a standard value as suggested by the literature. On the contrary, the CO2 model allowed an approximation between estimated and real data, but this prediction must be developed in a non-linear model for better results.

Data-Driven Unsteady Aeroelastic Modeling for Control

Aeroelastic structures, from insect wings to wind turbine blades, experience transient unsteady aerodynamic loads that are coupled to their motion. Effective real-time control of flexible structures relies on accurate and efficient predictions of both the unsteady aeroelastic forces and airfoil deformation. For rigid wings, classical unsteady aerodynamic models have recently been reformulated in state space for control and extended to include viscous effects. Here, we further extend this modeling framework to include the deformation of a flexible wing in addition to the quasi-steady, added mass, and unsteady viscous forces. We develop low-order linear models based on data from direct numerical simulations of flow past a flexible wing at a low Reynolds number. We demonstrate the effectiveness of these models to track aggressive maneuvers with model predictive control while constraining maximum wing deformation. This system identification approach provides an interpretable, accurate, and low-dimensional representation of an aeroelastic system that can aid in system and controller design for applications where transients play an important role.

System identification using autoregressive Bayesian neural networks with nonparametric noise models

System identification is of special interest in science and engineering. This article is concerned with a system identification problem arising in stochastic dynamic systems, where the aim is to estimate the parameters of a system along with its unknown noise processes. In particular, we propose a Bayesian nonparametric approach for system identification in discrete time nonlinear random dynamical systems assuming only the order of the Markov process is known. The proposed method replaces the assumption of Gaussian distributed error components with a flexible family of probability density functions based on Bayesian nonparametric priors. Additionally, the functional form of the system is estimated by leveraging Bayesian neural networks, which leads to flexible uncertainty quantification. Hamiltonian Monte Carlo sampler within a Gibbs sampler for posterior inference is proposed and its effectiveness is illustrated in real time series.