Article
Automation & Control Systems
Yawen Mao, Chen Xu, Jing Chen, Yan Pu
Summary: This article proposes an efficient Cholesky CMA-ES method for parameter estimation of nonlinear systems with colored noise, which addresses the computational complexity issue caused by matrix decomposition in traditional CMA-ES algorithm and improves the search accuracy. Simulation examples of parameter estimation problems in Hammerstein nonlinear systems demonstrate the effectiveness of the proposed method.
INTERNATIONAL JOURNAL OF ROBUST AND NONLINEAR CONTROL
(2023)
Article
Biochemical Research Methods
Neythen J. Treloar, Nathan Braniff, Brian Ingalls, Chris P. Barnes
Summary: The field of optimal experimental design utilizes mathematical techniques to identify experiments with maximal information content. In this study, a reinforcement learning technique is applied to achieve optimal experimental design for maximizing confidence in model parameter estimates. Results demonstrate that reinforcement learning outperforms other algorithms, such as one-step ahead optimization and model predictive control, in inferring bacterial growth parameters in a simulated chemostat. Moreover, reinforcement learning is shown to be robust to parametric uncertainty, as it can be trained over a distribution of parameters.
PLOS COMPUTATIONAL BIOLOGY
(2022)
Article
Engineering, Environmental
Qian-Kun Wang, Jia-Ni Shen, Zi-Feng Ma, Yi-Jun He
Summary: This study proposes a modeling strategy that couples a pseudo two-dimensional electrochemical model and a three-dimensional thermal model to describe the electrical and thermal dynamics of commercial LIBs, with the adoption of a variable solid state diffusion concept to enhance prediction ability. By introducing a decoupling estimation strategy and surrogate model optimization approach, the efficiency and accuracy of parameter estimation are improved significantly.
CHEMICAL ENGINEERING JOURNAL
(2021)
Article
Automation & Control Systems
Yancheng Zhu, Huaiyu Wu, Zhihuan Chen, Yang Chen, Xiujuan Zheng
Summary: This study investigates the parameter identification issues of bilinear-in-parameter systems through fractional adaptive algorithms. The proposed algorithms, based on auxiliary model identification idea and convergence index, provide improved accuracy and computational efficiency for parameter estimation. Numerical simulations further verify the effectiveness and accuracy of the proposed algorithms.
INTERNATIONAL JOURNAL OF ADAPTIVE CONTROL AND SIGNAL PROCESSING
(2022)
Article
Biochemistry & Molecular Biology
Tim Litwin, Jens Timmer, Clemens Kreutz
Summary: This paper presents a two-dimensional profile likelihood approach for identifying optimal experimental designs, which effectively estimates the uncertainty of model parameters. By using experimental design, it is possible to reduce the amount of data and resources required, and the method is validated in real-world applications.
FRONTIERS IN MOLECULAR BIOSCIENCES
(2022)
Article
Neurosciences
Jan Morez, Filip Szczepankiewicz, Arnold J. den Dekker, Floris Vanhevel, Jan Sijbers, Ben Jeurissen
Summary: Tensor-valued diffusion encoding allows disentangling variations in diffusivity from microscopic anisotropy, orientation dispersion, and mixtures of multiple isotropic diffusivities. Two precision-optimized acquisition schemes are created to estimate the diffusion tensor distribution (DTD) parameters accurately and precisely. The weighted linear least squares (WLLS) estimator with the squared reciprocal of the predicted signal as weights outperforms conventional estimators in terms of accuracy and precision when appropriate constraints are used.
HUMAN BRAIN MAPPING
(2023)
Article
Computer Science, Artificial Intelligence
Zhantao Chen, Cheng Peng, Alexander N. Petsch, Sathya R. Chitturi, Alana Okullo, Sugata Chowdhury, Chun Hong Yoon, Joshua J. Turner
Summary: Advanced experimental measurements play a crucial role in driving theoretical developments and uncovering novel phenomena in condensed matter and materials physics. However, the scarcity of large-scale facility resources often poses limitations. In this study, we introduce a methodology that utilizes the Bayesian optimal experimental design paradigm to efficiently extract key quantum spin fluctuation parameters from XPFS data. Our method is compatible with existing theoretical simulation pipelines and can also be combined with fast machine learning surrogate models to accelerate data collection and scientific discoveries.
MACHINE LEARNING-SCIENCE AND TECHNOLOGY
(2023)
Article
Computer Science, Interdisciplinary Applications
Lu Zhang, Junyao Xie, Stevan Dubljevic
Summary: This manuscript proposes moving horizon control and state/parameter estimation designs for pipeline networks modeled by partial differential equations (PDEs) with boundary actuation. The effectiveness of the proposed controller and estimator designs is demonstrated via numerical examples.
COMPUTERS & CHEMICAL ENGINEERING
(2023)
Article
Multidisciplinary Sciences
Ammara Mehmood, Muhammad Asif Zahoor Raja
Summary: The application of evolutionary computing paradigm-based heuristics for system modeling and parameter estimation of complex nonlinear systems has been widely explored. This study investigates the use of weighted differential evolution (WDE) in estimating the parameters of Hammerstein-Wiener model (HWM) and compares it with state-of-the-art methods. The HWM parameters are estimated using the WDE and genetic algorithms (GAs) heuristics, and the worth and value of the designed WDE algorithm is demonstrated through extensive graphical and numerical comparisons.
JOURNAL OF ADVANCED RESEARCH
(2023)
Article
Computer Science, Interdisciplinary Applications
Weifeng Chen, Baojia Wang, Lorenz T. Biegler
Summary: Parameter estimation is a crucial step in system modeling. However, in practice, the complexity of chemical engineering models and the interplay between parameters make it difficult to estimate all parameters accurately. This study proposes a parameter estimation approach based on a reduced Hessian matrix and statistical criteria, which improves model prediction by considering the influence of initial parameter values.
COMPUTERS & CHEMICAL ENGINEERING
(2022)
Article
Computer Science, Artificial Intelligence
Kechao Xu, Bo Meng, Zhen Wang
Summary: This study proposes an adaptive data-driven control protocol design scheme for nonaffine multi-agent systems with unknown nonlinearity. The control protocol uses higher order parameter estimation and iterative learning to solve the consistency tracking problem of multi-agent systems with fixed and switching topology communications. It achieves consensus tracking by using only the input/output data of the agents, without requiring knowledge of their dynamical systems. The control protocol has a modular design that combines iterative learning and higher order parameter estimation algorithms. Its effectiveness is demonstrated through simulation experiments.
KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Automation & Control Systems
Ling Xu, Feng Ding, Erfu Yang
Summary: This article proposes an auxiliary model stochastic gradient identification approach based on gradient optimization and develops an auxiliary model multiinnovation stochastic gradient estimation method to enhance estimation accuracy by introducing more observed data dynamically. The numerical simulation results demonstrate the effectiveness of the proposed auxiliary model identification method for nonlinear sandwich systems.
INTERNATIONAL JOURNAL OF ROBUST AND NONLINEAR CONTROL
(2021)
Article
Computer Science, Information Systems
Youngjoon Hong, Bongsuk Kwon, Byung-Jun Yoon
Summary: This paper discusses the optimal experimental design problem for an uncertain system described by coupled ODEs, aiming to reduce model uncertainty within a limited experimental budget through the development of an OED strategy based on MOCU. The main objective is to identify the optimal experiment that maximally reduces the uncertainty cost, demonstrating the importance of quantifying potential experiments' operational impact in designing optimal experiments.
Article
Automation & Control Systems
Ning Xu, Feng Ding
Summary: This article focuses on the identification of time-varying systems. Unlike conventional polynomial approximation approaches, the changing laws of the time-varying parameters are taken into account to establish the identification model. The concept of the invariant matrix is introduced to characterize the time-varying parameters and establish the state-space model. Two state estimation algorithms, stacked and detached, are proposed to estimate the time-varying parameters and enhance computational efficiency. Numerical simulation examples are provided to demonstrate the effectiveness of the proposed algorithms.
INTERNATIONAL JOURNAL OF ROBUST AND NONLINEAR CONTROL
(2023)
Article
Engineering, Chemical
Gustavo Lunardon Quillo, Satyajeet Bhonsale, Alain Collas, Christos Xiouras, Jan F. M. Van Impe
Summary: The design of the crystallization process relies on predictive solubility models, but their calibration is resource-intensive. Optimal experimental design techniques can reduce the experimental burden and augment existing datasets to improve model prediction power.
CHEMICAL ENGINEERING RESEARCH & DESIGN
(2022)
Article
Biochemical Research Methods
Andrew D. Mathis, Bradley C. Naylor, Richard H. Carson, Eric Evans, Justin Harwell, Jared Knecht, Eric Hexem, Fredrick F. Peelor, Benjamin F. Miller, Karyn L. Hamilton, Mark K. Transtrum, Benjamin T. Bikman, John C. Price
MOLECULAR & CELLULAR PROTEOMICS
(2017)
Review
Engineering, Chemical
C. Anthony Hunt, Ahmet Erdemir, William W. Lytton, Feilim Mac Gabhann, Edward A. Sander, Mark K. Transtrum, Lealem Mulugeta
Article
Thermodynamics
Lee D. Hansen, Nieves Barros, Mark K. Transtrum, Jose A. Rodriguez-Anon, Jorge Proupin, Veronica Pineiro, Ander Arias-Gonzalez, Nahia Gartzia
THERMOCHIMICA ACTA
(2018)
Article
Automation & Control Systems
Philip E. Pare, David Grimsman, Alma T. Wilson, Mark K. Transtrum, Sean Warnick
IEEE TRANSACTIONS ON AUTOMATIC CONTROL
(2019)
Article
Engineering, Electrical & Electronic
Vanja G. Svenda, Mark K. Transtrum, Benjamin L. Francis, Andrija T. Saric, Aleksandar M. Stankovic
Summary: This paper presents a method for estimating system state by analyzing system observability. The method utilizes information geometry to detect unidentifiable system parameters and states, and simplifies the model by removing reference to unidentifiable state variables. The effectiveness of the method is tested through co-simulation of the physical and cyber system layers.
IEEE TRANSACTIONS ON POWER SYSTEMS
(2022)
Article
Chemistry, Physical
Yonatan Kurniawan, Cody L. Petrie, Kinamo J. J. Williams, Mark K. Transtrum, Ellad B. Tadmor, Ryan S. Elliott, Daniel S. Karls, Mingjian Wen
Summary: This paper investigates the quantification of parametric uncertainty in classical empirical interatomic potentials using Bayesian and frequentist methods. It reveals that these potentials are typically insensitive and parameters are unidentifiable. Information geometry is used to explain the underlying cause and suggest new parameterizations and simplified models.
JOURNAL OF CHEMICAL PHYSICS
(2022)
Review
Physics, Multidisciplinary
Katherine N. Quinn, Michael C. Abbott, Mark K. Transtrum, Benjamin B. Machta, James P. Sethna
Summary: Complex models in various fields often have parameter ambiguity, where the parameters of the model are not well determined by the predictions for collective behavior. This review uses information geometry to explore the concept of sloppiness and its connection to emergent theories. The review discusses the structure of the model manifold and how it can explain why only a few parameter combinations matter for behavior. It also introduces methods for finding simpler models on nearby boundaries of the model manifold and discusses Bayesian priors that favor simpler models.
REPORTS ON PROGRESS IN PHYSICS
(2023)
Article
Computer Science, Information Systems
Aleksandar A. Saric, Mark K. Transtrum, Andrija T. Saric, Aleksandar M. Stankovic
Summary: This article explores the analysis of transient phenomena in large-scale power systems subjected to major disturbances from the aspect of interleaving, coordinating, and refining physics- and data-driven models. The study proposes a framework that enables coordinated and seamlessly integrated use of the two types of models in engineered systems.
IEEE SYSTEMS JOURNAL
(2023)
Article
Chemistry, Multidisciplinary
Katrina Pedersen, Ryan R. Jensen, Lucas K. Hall, Mitchell C. Cutler, Mark K. Transtrum, Kent L. Gee, Shane V. Lympany
Summary: Applying machine learning methods to geographic data helps in understanding spatial patterns and interpreting environments. This study used k-means clustering to analyze 51 geospatial layers and identified 8 clusters with distinct characteristics. The results can guide data collection for modeling outdoor acoustic environments.
APPLIED SCIENCES-BASEL
(2023)
Proceedings Paper
Computer Science, Interdisciplinary Applications
Yonatan Kurniawan, Cody L. Petrie, Mark K. Transtrum, Ellad B. Tadmor, Ryan S. Elliott, Daniel S. Karls, Mingjian Wen
Summary: Atomistic simulations are important in materials modeling, and the choice of interatomic potentials (IPs) greatly affects the accuracy of the predictions. Uncertainty quantification (UQ) is a new tool for assessing the reliability of these simulations. The OpenKIM project aims to standardize the study of IPs and enable transparent research, and the KLIFF Python package provides tools for fitting IP parameters. This paper introduces a UQ extension to KLIFF, focusing on parameter variations and inadequacy of the IP's functional form.
2022 IEEE 18TH INTERNATIONAL CONFERENCE ON E-SCIENCE (ESCIENCE 2022)
(2022)
Article
Engineering, Electrical & Electronic
Andrija T. Saric, Mark K. Transtrum, Aleksandar M. Stankovic
Summary: This paper presents a manifold learning-based algorithm for big data classification and reduction, as well as parameter identification in real-time operation of a power system. The algorithm examines both black-box and gray-box settings for SCADA- and PMU-based measurements, and uses improved data-informed metric construction for partition trees in data classification. Demonstrations are made on a measurement tensor example of calculated transient dynamics between two SCADA refreshing scans.
IEEE TRANSACTIONS ON POWER SYSTEMS
(2021)
Article
Engineering, Electrical & Electronic
Andrija T. Saric, Aleksandar A. Saric, Mark K. Transtrum, Aleksandar M. Stankovic
Summary: This paper presents a data-driven symbolic regression identification method designed for power systems, which extends the SINDy modeling procedure to include exogenous signals and nonlinear trigonometric terms. The resulting framework is shown to require minimal data, be computationally efficient, and robust to noise, making it a feasible option for online identification in response to rapid system changes. The proposed method is illustrated on a real-world benchmark example, demonstrating its effectiveness in reducing the differential-algebraic equations-based SG dynamic models.
IEEE TRANSACTIONS ON POWER SYSTEMS
(2021)
Article
Materials Science, Multidisciplinary
Nathan S. Sitaraman, Michelle M. Kelley, Ryan D. Porter, Matthias U. Liepe, Tomas A. Arias, Jared Carlson, Alden R. Pack, Mark K. Transtrum, Ravishankar Sundararaman
Summary: The electronic free energy has a profound effect in systems with a high-temperature threshold for kinetics and a high Fermi-level density of states. Antisite defects disrupt the high Fermi-level density of states and locally reduce electronic free energy, affecting superconductivity. The study on Nb3Sn reveals the key role of electronic free energy in determining the behavior of antisite defects, their interactions, and their impact on superconductivity.
Article
Materials Science, Multidisciplinary
Jared Carlson, Alden Pack, Mark K. Transtrum, Jaeyel Lee, David N. Seidman, Danilo B. Liarte, Nathan S. Sitaraman, Alen Senanian, Michelle M. Kelley, James P. Sethna, Tomas Arias, Sam Posen
Summary: Our study investigates the mechanisms of vortex nucleation in Nb3Sn superconducting cavities, revealing Sn segregation at grain boundaries which may affect the local superconducting properties. Using ab initio calculations and time-dependent Ginzburg-Landau theory, simulations show that grain boundaries can act as both nucleation sites and pinning sites for vortices. We estimate the superconducting losses due to vortices filling grain boundaries and compare with experimental observations of cavity heating for performance evaluation.
Article
Materials Science, Multidisciplinary
Alden R. Pack, Jared Carlson, Spencer Wadsworth, Mark K. Transtrum