Model-based Parameter Estimation Research Articles

Nonlinear parameter estimation with physics-constrained spectral–spatial priors for highly accelerated chemical exchange saturation transfer MRI

Objective.To develop a nonlinear, model-based parameter estimation method directly from incomplete measurements ink - wspace for robust spectral analysis in highly accelerated chemical exchange saturation transfer (CEST) magnetic resonance imaging (MRI).Approach. A CEST-specific, separable nonlinear model, which describes spectral decomposition using multi-pool Lorentzian functions (conventional magnetization transfer (MT), direct saturation of water signals (DS), amide, amine, and nuclear Overhauser effect) derived from the steady-state Bloch McConnel equation, is incorporated into a measurement model in CEST MRI. Furthermore, signal drop in saturation transfer experiments is formulated by an additional, separable nonlinear spectral prior indicating that the symmetric z-spectra synthesized using conventional MT and DS always remain higher or equal to the whole z-spectra with all pools. Given the above considerations, linear and nonlinear parameters in the proposed method are estimated in an alternating fashion directly from highly incomplete measurements ink - wspace by solving a constrained optimization problem with the physics-constrained spectral priors while imposing additional sparsity priors on spatial parameter maps.Main results.Compared with conventional methods, the proposed method yields clearer delineation of tumor-specific CEST maps without apparent artifact and noise.Significance.We successfully demonstrated the feasibility of the proposed method for CEST MRI with highly incomplete measurements thus enabling high-resolution whole brain CEST MRI in clinically reasonable imaging time.

Model-Centric Integration of Uncertain Expert Knowledge into Importance Sampling-Based Parameter Estimation

This study presents a model-based parameter estimation method for integrating and validating uncertainty in expert knowledge and simulation models. The parameters of the models of complex systems are often unknown due to a lack of measurement data. The experience-based knowledge of experts can substitute missing information, which is usually imprecise. The novelty of the present paper is a method based on Monte Carlo (MC) simulation and importance sampling (IS) techniques for integrating uncertain expert knowledge into the system model. Uncertain knowledge about the model parameters is propagated through the system model by MC simulation in the form of a discrete sample, while IS helps to weight the sample elements regarding imprecise knowledge about the outputs in an iterative circle. Thereby, the consistency of expert judgments can be investigated as well. The contributions of this paper include an expert knowledge-based parameter estimation technique and a method for the evaluation of expert judgments according to the estimation results to eliminate incorrect ones. The applicability of the proposed method is introduced through a case study of a Hungarian operating waste separation system. The results verify that the assessments of experts can be efficiently integrated into system models, and their consistency can be evaluated.

Differential Model-Based Parameter Estimation of IPMSMs From Multi-State Measurements

Differential Model-Based Parameter Estimation of IPMSMs From Multi-State Measurements

Parameter Estimation of Cellular Communication Systems Models in Computational MATLAB Environment: A Systematic Solver-based Numerical Optimization Approaches

Model-based parameter estimation, identification, and optimisation play a dominant role in many aspects of physical and operational processes in applied sciences, engineering, and other related disciplines. The intricate task involves engaging and fitting the most appropriate parametric model with nonlinear or linear features to experimental field datasets priori to selecting the best optimisation algorithm with the best configuration. Thus, the task is usually geared towards solving a clear optimsation problem. In this paper, a systematic-stepwise approach has been employed to review and benchmark six numerical-based optimization algorithms in MATLAB computational Environment. The algorithms include the Gradient Descent (GRA), Levenberg-Marguardt (LEM), Quasi-Newton (QAN), Gauss-Newton (GUN), Nelda-Meald (NEM), and Trust-Region-Dogleg (TRD). This has been accomplished by engaging them to solve an intricate radio frequency propagation modelling and parametric estimation in connection with practical spatial signal data. The spatial signal data were obtained via real-time field drive test conducted around six eNodeBs transmitters, with case studies taken from different terrains where 4G LTE transmitters are operational. Accordingly, three criteria in connection with rate of convergence Results show that the approximate hessian-based QAN algorithm, followed by the LEM algorithm yielded the best results in optimizing and estimating the RF propagation models parameters. The resultant approach and output of this paper will be of countless assets in assisting the end-users to select the most preferable optimization algorithm to handle their respective intricate problems.

An integration method of a hybrid genetic algorithm and the Levenberg–Marquardt algorithm for ultrasonic testing

The accurate measurement of time-of-flight (TOF) is essential in ultrasonic testing. Further, noise interference is the key factor affecting the measurement accuracy. Therefore, to develop a reliable computational method of TOF for test pieces working in noisy environments, an integration method of a hybrid genetic algorithm and the Levenberg–Marquardt algorithm (GA–LM) for ultrasonic thickness measurement is proposed in the present research. A Gaussian model is first established for an echo signal. Further, the model-based parameter estimation is converted into a nonlinear optimization problem by applying the least square method. As the parameter estimation methods are easily affected by the initial value, an integrating innovation of the GA–LM algorithm is proposed. The initial values of the model parameters are selected by GA to obtain an approximate global optimal solution. Subsequently, this approximate solution is used as the initial value for the LM algorithm to perform iterations. The accurate global optimal solution of the Gaussian model is obtained through these iterations. Finally, the measuring accuracy and robustness of the GA–LM algorithm for TOF computation are verified by both numerical simulation and experiment data

Compensating voltage waveform distortions using a practical topology of series active power filters

Addressing harmonics, voltage sags, voltage swells, and asymmetrical variations is essential to seamlessly integrate renewable energy sources, electric vehicle charging stations, and power electronics devices into electric power grids. These issues can significantly impact the sinusoidal symmetry of voltage waveforms. Series Active Power Filters (APFs) present a promising solution that can dramatically improve power quality (PQ) by effectively addressing voltage distortions. This paper first identifies several topology-related impracticalities in the existing literature on series APFs, such as using nonlinear loads or linear loads with a current source. Secondly, to understand the motivations behind these impracticalities, commonly used reference extraction methods, namely, Instantaneous Reactive Power Theory (IRPT) and Synchronous Reference Frame (SRF), are tested on a practical topology consisting of a distorted voltage source and a linear RL load. Results conducted in the MATLAB/Simulink environment show that IRPT fails to extract the reference signal under the practical topology, and SRF leads to unacceptable THDs surpassing the 5% threshold mandated by relevant standards set forth for such applications. Thirdly, the matrix pencil method (MPM), a model-based parameter estimation technique that exploits a voltage waveform's underlying exponential signal model to extract the series APF's reference voltage, is proposed. Extensive simulations showcase the superior performance of the MPM-based series APF. It successfully reduces voltage total harmonic distortion (THD) to below 1.13% across various scenarios, including situations involving harmonic-polluted voltage sources, incorporating nonlinear loads, sag and swell phenomenon, and capacitor utilization in DC link implementation.

Reinforcement Learning Based Parameter Lookup Table Generating Method for Optimal Torque Control of Induction Motors

In the optimal control of induction motors, it is a challenging task to maintain the optimal torque over the varying operation conditions. This paper proposes a parameter look-up table generating method, that can achieve an optimal torque over a wide range of currents and speeds, even though the commands of current are not set correctly. Based on the motor’s testing data, this method uses a reinforcement-learning algorithm to generate parameter look-up tables iteratively. Experimental results show that the proposed method can learn appropriate parameters from the running data to output an optimal torque. The comparative studies show that the proposed method can generate 5%-25% more torque than traditional model-based parameter estimation methods, over a wide range of currents and speeds. Furthermore, the proposed method has a faster convergence feature and a higher identification resolution than many conventional search-based methods.

A Filter-Less Time-Domain Method for Reference Signal Extraction in Shunt Active Power Filters

Current distortion degrades power quality and affects system performance, especially for sensitive loads that require pure sinusoidal waves. Owing to its excellent dynamic response, a well-designed active power filter (APF) can achieve a total harmonic distortion (THD) within the acceptable limits defined by IEEE 3002 standards through compensating harmonic distortions. The APF consists of two main modules: a reference signal extraction module and a modulation module. This paper adapts the matrix pencil method, a well-known model-based parameter estimation technique, to the problem of reference extraction. Contrary to conventional time-domain methods such as the synchronous reference frame (SRF) and the reactive power theory (PQ theory) that rely on low-pass filters in their implementations, the proposed method does not use a filter. However, it extracts the reference signal by first decomposing the load current into its constituent frequency components, and then subtracting the pure sine wave synthesized from the obtained fundamental component from the load current. Results on simulated data from MATLAB/Simulink confirm the higher accuracy and fast response time of the proposed method in extracting the reference signal.

Contact parameters evolution of bolted joint interface under transversal random vibrations

Bolt connection plays an important role in assembly structures. Under long-term vibrations, bolted joint interface will change due to fretting-induced friction and wear behavior. This leads to changes in the dynamic characteristics of assembly structures, which increases the difficulty of control of the structure. In this paper, a new random fretting test method was developed to reproduce realistic vibration situations as much as possible and to study the evolution of contact parameters (friction coefficient and contact stiffness) of bolted joint interfaces with increasing wear. A recently developed test rig was used to perform the fretting wear tests under random oscillation. During the tests, interface hysteresis response and bolt preload were recorded. A model-based parameter estimation method was developed to estimate the friction coefficient and tangential contact stiffness. The evolution curves of contact parameters were obtained, and the effects of initial bolt preload, excitation level, and surface roughness were discussed. To the authors' knowledge, no studies on the fretting behavior of bolted joint interfaces under random vibration have been published. This research can promote a better understanding of the fretting wear behavior of bolted joints and provide a basis for the dynamic analysis of joint-dominated structures.

A Broadband Model-Based Parameter Estimation Method for Analyzing Multilayer Periodic Structures

A new model-based parameter estimation (MBPE) method for multilayer periodic EM surfaces is discussed in this article. By applying the proposed method to predict the generalized scattering matrix (GSM) of each unit layer section, the response of a composite structure can be determined by cascading GSMs while allowing the salient behavior of each section to be accurately captured. Based on the spectral properties of Green's functions with high-order wavenumbers and chosen basis functions, an advanced interpolation model is developed to estimate the broadband response of periodic structures in a robust and efficient manner. It needs fewer samples in the interpolation compared to the state of the art, and no restriction for the sample selection is required. Numerical analysis has been carried out for two types of multilayer structures, and a good agreement between the conventional periodic method of moments (PMM) and the proposed method has been obtained.

FBG-Based Estimation of External Forces Along Flexible Instrument Bodies.

A variety of medical treatment and diagnostic procedures rely on flexible instruments such as catheters and endoscopes to navigate through tortuous and soft anatomies like the vasculature. Knowledge of the interaction forces between these flexible instruments and patient anatomy is extremely valuable. This can aid interventionalists in having improved awareness and decision-making abilities, efficient navigation, and increased procedural safety. In many applications, force interactions are inherently distributed. While knowledge of their locations and magnitudes is highly important, retrieving this information from instruments with conventional dimensions is far from trivial. Robust and reliable methods have not yet been found for this purpose. In this work, we present two new approaches to estimate the location, magnitude, and number of external point and distributed forces applied to flexible and elastic instrument bodies. Both methods employ the knowledge of the instrument’s curvature profile. The former is based on piecewise polynomial-based curvature segmentation, whereas the latter on model-based parameter estimation. The proposed methods make use of Cosserat rod theory to model the instrument and provide force estimates at rates over 30 Hz. Experiments on a Nitinol rod embedded with a multi-core fiber, inscribed with fiber Bragg gratings, illustrate the feasibility of the proposed methods with mean force error reaching 7.3% of the maximum applied force, for the point load case. Furthermore, simulations of a rod subjected to two distributed loads with varying magnitudes and locations show a mean force estimation error of 1.6% of the maximum applied force.

On the use of fixed point translations as input variable for digital twins in deep drawing compared to current methods

Recent work aims at the inverse parameter estimation in deep drawing using pretrained surrogate models for the detection of the current process, material or tool parameters. The use of the methodology requires the definition of state variables to describe the current process state. Whereas our recent work makes use of draw-ins and local blankholder forces, other approaches from the literature also use skid-lines measured after the deep drawing process. For the future, the solution with even higher information content would be to detect the global strain distribution on the final part and use it as a state variable for process detection, which has not been documented in the literature to the best knowledge of the authors.In this work, we present a first step into this direction by comparing the surrogate model based parameter estimation by using draw-ins and by using the movement of material fixed points on the blank over the deep drawing process. The result shows that the mathematical methods used for parameter prediction based on draw-ins can directly be used for the prediction with fixed point translations as reference. For the investigations, a cup drawing process is used.

Modeling, estimation, and control in large-scale urban road networks with remaining travel distance dynamics

City-scale control of urban road traffic poses a challenging problem. Dynamical models based on the macroscopic fundamental diagram (MFD) enable development of model predictive perimeter control methods for large-scale urban networks, representing an advanced control solution carrying substantial potential for field implementation. In this paper we develop a multi-region approximation of the trip-based model, which describes in more details the trip length characteristics compared to existing accumulation-based models. The proposed M-model includes effects of the total remaining travel distance on the transfer flows driving the vehicle accumulation dynamics, potentially yielding improved accuracy over the standard production-over-trip length approximation of the outflow MFD considered in many works on MFD-based modeling and control. We explain that to properly perform perimeter control, boundary queue dynamics have to be integrated. Furthermore, model-based parameter estimation (MBPE), nonlinear moving horizon observer (MHO), and model predictive control (MPC) formulations for the proposed models are presented, forming an integrated traffic control framework. Microsimulation-based case studies, considering an urban network with 1500 links, where the model parameters obtained by MBPE method are used in MHO and MPC design, demonstrate the efficient operation of the proposed framework.

A systematic approach for COVID-19 predictions and parameter estimation.

The world is currently facing a pandemic called COVID-19 which has drastically changed our human lifestyle, affecting it badly. The lifestyle and the thought processes of every individual have changed with the current situation. This situation was unpredictable, and it contains a lot of uncertainties. In this paper, the authors have attempted to predict and analyze the disease along with its related issues to determine the maximum number of infected people, the speed of spread, and most importantly, its evaluation using a model-based parameter estimation method. In this research the Susceptible-Infectious-Recovered model with different conditions has been used for the analysis of COVID-19. The effects of lockdown, the light switch method, and parameter variations like contact ratio and reproduction number are also analyzed. The authors have attempted to study and predict the lockdown effect, particularly in India in terms of infected and recovered numbers, which show substantial improvement. A disease-free endemic stability analysis using Lyapunov and LaSalle’s method is presented, and novel methods such as the convalescent plasma method and the Who Acquires Infection From Whom method are also discussed, as they are considered to be useful in flattening the curve of COVID-19.

A novel data-driven model based parameter estimation of nonlinear systems

In practice, it is usually difficult to estimate the characteristic parameters of a system due to the system nonlinearity and uncertainty. To address this problem, in this study, the characteristic parameters of a nonlinear system are identified by using a data driven method based on the best discretization of the Nonlinear Differential Equation (NDE) model of the system. The best discretization of a NDE model is firstly determined, and then the discretized model, known as the Nonlinear Auto Regressive with eXegenous input (NARX) model, is determined by using the Least Squares (LS) algorithm from the input and output data of system. A case study is discussed to validate the proposed system parameter identification method, where the characteristic parameters of a rotating blade-casing system are evaluated under a bandwidth rub impact in the horizontal direction with noise. The result shows that the identified model can be used to describe the characteristics of the underlying system accurately, which provides a reliable model for the dynamic analysis, control of rotating blade-casing system.

Best linear unbiased estimation for varying probability with and without replacement sampling

Abstract When sample survey data with complex design (stratification, clustering, unequal selection or inclusion probabilities, and weighting) are used for linear models, estimation of model parameters and their covariance matrices becomes complicated. Standard fitting techniques for sample surveys either model conditional on survey design variables, or use only design weights based on inclusion probabilities essentially assuming zero error covariance between all pairs of population elements. Design properties that link two units are not used. However, if population error structure is correlated, an unbiased estimate of the linear model error covariance matrix for the sample is needed for efficient parameter estimation. By making simultaneous use of sampling structure and design-unbiased estimates of the population error covariance matrix, the paper develops best linear unbiased estimation (BLUE) type extensions to standard design-based and joint design and model based estimation methods for linear models. The analysis covers both with and without replacement sample designs. It recognises that estimation for with replacement designs requires generalized inverses when any unit is selected more than once. This and the use of Hadamard products to link sampling and population error covariance matrix properties are central topics of the paper. Model-based linear model parameter estimation is also discussed.

Rapid Radial T1 and T2 Mapping of the Hip Articular Cartilage With Magnetic Resonance Fingerprinting.

Quantitative MRI can detect early changes in cartilage biochemical components, but its routine clinical implementation is challenging. To introduce a novel technique to measure T1 and T2 along radial sections of the hip for accurate and reproducible multiparametric quantitative cartilage assessment in a clinically feasible scan time. Reproducibility, technical validation. A seven-compartment phantom and three healthy volunteers. A novel MR pulse sequence that simultaneously measures proton density (PD), T1 , and T2 at 3 T was developed. Automatic positioning and semiautomatic cartilage segmentation were implemented to improve consistency and simplify workflow. Intra- and interscanner variability of our technique was assessed over multiple scans on three different MR scanners. For each scan, the median of cartilage T1 and T2 over six radial slices was calculated. Restricted maximum likelihood estimation of variance components was used to estimate intrasubject variances reflecting variation between results from the two scans using the same scanner (intrascanner variance) and variation among results from the three scanners (interscanner variance). The estimation error for T1 and T2 with respect to reference standard measurements was less than 3% on average for the phantom. The average interscanner coefficient of variation was 1.5% (1.2-1.9%) and 0.9% (0.0-3.7%) for T1 and T2 , respectively, in the seven compartments of the phantom. Total scan time in vivo was 7:13 minutes to obtain PD, T1 , and T2 maps along six radial hip sections at 0.6 × 0.6 × 4.0 mm3 voxel resolution. Interscanner variability for the in vivo study was 1.99% and 5.46% for T1 and T2 , respectively. in vivo intrascanner variability was 1.15% for T1 and 3.24% for T2 . Our method, which includes slice positioning, model-based parameter estimation, and cartilage segmentation, is highly reproducible. It could enable employing quantitative hip cartilage evaluation for longitudinal and multicenter studies. 1 Technical Efficacy: Stage 1 J. Magn. Reson. Imaging 2019;50:810-815.

Parameter estimation of a six-lump kinetic model of an industrial fluid catalytic cracking unit

In this work a simulation of detailed steady state model of an industrial fluid catalytic cracking (FCC) unit with a newly proposed six-lumped kinetic model which cracks gas oil into diesel, gasoline, liquefied petroleum gas (LPG), dry gas and coke. Frequency factors, activation energies and heats of reaction for the catalytic cracking kinetics and a number of model parameters were estimated using a model based parameter estimation technique along with data from an industrial FCC unit in Sudan. The estimated parameters were used to predict the major riser fractions; diesel as 0.1842 kg-lump/kg-feed with a 0.81% error while gasoline as 0.4863 kg-lump/kg-feed with a 2.71% error compared with the plant data. Thus, with good confidence, the developed kinetic model is able to simulate any type of FCC riser with six-lump model as catalyst-to-oil (C/O) ratios were varied and the results predicted the typical riser profiles.

Rotor fault diagnosis in induction motors by the matrix pencil method and support vector machine

This paper proposes a high-resolution motor-current-signature-analysis method and a supervised learning algorithm for the detection and classification of broken rotor bars and broken end-ring connectors in 3-phase induction motors. Signature analysis relies on the matrix pencil method (MPM), a well-known model-based parameter estimation technique, to extract representative features or signatures from stator current signals. Extracted feature vectors are subsequently used to train off-line a support vector machine classifier. Once trained, the classifier is tested on a benchmark dataset of simulated stator current signals representing healthy and faulty rotors with the aim of classifying the underlying motor condition. The obtained results validate the matrix pencil method as a feature extraction method and show that the trained classifier achieves a 100% success rate in identifying the number of broken bars and connectors. Moreover, the advantages of the matrix pencil method over fast Fourier transform are demonstrated using experimental data.

Fast Wideband Solutions Obtained Using Model Based Parameter Estimation with Method of Moments

Integration of the Model Based Parameter Estimation (MBPE) technique into Method of Moments (MOM) provides fast solutions over a wide frequency band to solve radiation and scattering problems. The MBPE technique uses the Padé rational function to approximate solutions over a wide frequency band from a solution at a fixed frequency. In this paper, the MBPE technique with MOM is applied to a thin-wire antenna. The solutions obtained by repeated simulations of MOM agree very well with the solutions obtained by MBPE technique in a single simulation. Therefore, MBPE technique according to MOM provides a remarkable saving in the computation time. Computed results show that solutions at a wider frequency band of interest are achieved in a single simulation.