Article
Automation & Control Systems
Gemma Massonis, Julio R. Banga, Alejandro F. Villaverde
Summary: Mechanistic dynamic models of biological systems often suffer from over-parameterization, resulting in nonidentifiability and nonobservability. AutoRepar is a methodology that automatically corrects these structural deficiencies, producing reparameterized models with improved identifiability and observability. This approach increases the applicability of mechanistic models, providing reliable information about their parameters and dynamics.
INTERNATIONAL JOURNAL OF ROBUST AND NONLINEAR CONTROL
(2023)
Article
Automation & Control Systems
Alexey Ovchinnikov, Anand Pillay, Gleb Pogudin, Thomas Scanlon
Summary: The parameter identifiability is crucial for recovering parameter values from data. The authors introduce a new algorithm that computes all identifiable functions without additional assumptions, providing a novel approach to address the issue of nonidentifiability.
SYSTEMS & CONTROL LETTERS
(2021)
Article
Automation & Control Systems
Zhe Li, Xun Wang, Uwe Kruger
Summary: Kernel Partial Least Squares (KPLS) is an effective nonlinear modeling technique used in control engineering applications, capable of handling small sample sizes and noisy, highly correlated variable sets. By mapping input variables to a feature space, it produces an optimal prediction model for process output variables. However, the computational intensity of the procedure can be a challenge, especially when dealing with large data sets. The proposed Efficient Kernel Partial Least Squares (EKPLS) aims to reduce computational complexity significantly compared to the traditional approach.
CONTROL ENGINEERING PRACTICE
(2021)
Review
Biochemical Research Methods
Veronica Porubsky, Lucian Smith, Herbert M. Sauro
Summary: The article highlights the importance of publishing repeatable and reproducible computational models in computational biology, specifically focusing on issues in the systems biology field. It discusses the current landscape in terms of software tools, model repositories, standards, and best practices, as well as potential future remedies to improve reproducibility in scientific research.
BRIEFINGS IN BIOINFORMATICS
(2021)
Article
Automation & Control Systems
Jean-Marie Guihal, Francois Auger, Nicolas Bernard, Emmanuel Schaeffer
Summary: This article presents the application of the extended Kalman filter in its continuous-discrete form and introduces several new efficient integration methods, comparing them with classical methods. The experimental results show that the choice of integration method has different effects on the accuracy and convergence of the EKF under different systems and sampling frequencies.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2022)
Article
Computer Science, Artificial Intelligence
Youming Ge, Zitong Chen, Weiyang Kong, Yubao Liu, Raymond Chi-Wing Wong, Sen Zhang
Summary: This study analyzes the computation of Group Steiner Tree (GST) in temporal graphs and proposes an efficient solution based on dynamic programming algorithm. Optimization techniques are adopted to reduce algorithm search space and experimental results verify the efficiency and effectiveness of the proposed algorithms.
INTERNATIONAL JOURNAL OF INTELLIGENT SYSTEMS
(2023)
Article
Automation & Control Systems
Yanling Wei, Hamid Reza Karimi
Summary: This article addresses the dynamic sliding mode control problem for unmatched nonlinear parameter-varying systems. By constructing a linear sliding surface function and deriving the synthesis procedure of the sliding manifold based on a parameter-dependent Lyapunov function, the asymptotic stability of the system is guaranteed. Furthermore, a dynamic SMC law is proposed to guide the closed-loop system towards the sliding manifold in finite time, with both the sliding surface and control law depending on time-varying and measurable parameters. Simulation studies are provided to demonstrate the validity of the proposed method.
INTERNATIONAL JOURNAL OF ROBUST AND NONLINEAR CONTROL
(2021)
Article
Biochemical Research Methods
Alejandro F. Villaverde, Dilan Pathirana, Fabian Frohlich, Jan Hasenauer, Julio R. Banga
Summary: Ordinary differential equation models are widely used for describing biological processes, but their parameter calibration process faces challenges. We provide a protocol to guide users through the calibration of dynamic models, while also providing model code and a way to reproduce the results.
BRIEFINGS IN BIOINFORMATICS
(2022)
Article
Food Science & Technology
Jesus Miguel Zamudio Lara, Laurent Dewasme, Hector Hernandez Escoto, Alain Vande Wouwer
Summary: Two dynamic models of beer fermentation are proposed in this study, and their parameters are estimated using experimental data. The structural identifiability of the measurement configuration and kinetic model structure is analyzed, and the predictive capability of the model is investigated. The model can be used for monitoring and controlling the beer fermentation process.
Article
Biochemical Research Methods
Kate E. Dray, Joseph J. Muldoon, Niall M. Mangan, Neda Bagheri, Joshua N. Leonard
Summary: Mathematical modeling is crucial for understanding and designing synthetic biological systems. However, the model development process is complex and nonintuitive, requiring iteration and comparison with experimental data. To address these challenges, we introduce the GAMES workflow, which combines automated and human-in-the-loop processes. This workflow enables biologists to more easily build and analyze models for various applications.
ACS SYNTHETIC BIOLOGY
(2022)
Article
Engineering, Environmental
Thomas Krumpolc, D. W. Trahan, D. A. Hickman, L. T. Biegler
Summary: Applications of fixed-effects models for kinetic parameter estimation assume independence among batches, but biased residuals often exist in multiple longitudinal batch experiments with time series data. Nonlinear mixed-effects models provide an alternative approach to address the two types of random experimental variation resulting from longitudinal experiments: measurement error for each data point and random batch-to-batch variation. In our case study, implementing a mixed-effects model using nonlinear programming for a batch reactor system yields parameter estimates with less bias compared to a fixed-effects model. Additionally, the Bayesian notion of probability shares is applied to discriminate between several candidate mixed-effects models, demonstrating the ability to elucidate additional model information when fixed-effects models are inappropriate.
CHEMICAL ENGINEERING JOURNAL
(2022)
Article
Automation & Control Systems
Jarrad Courts, Adrian G. Wills, Thomas B. Schon, Brett Ninness
Summary: This paper addresses the challenging problem of parameter estimation for nonlinear state-space models by employing a variational inference approach, which provides deterministic and tractable estimates through an optimization problem. A specialized method for systems with additive Gaussian noise is also presented. Numerical experiments demonstrate the robustness of the proposed method in parameter initialization and favorable comparisons against state-of-the-art alternatives.
Article
Multidisciplinary Sciences
Dominique Joubert, J. D. Stigter, Jaap Molenaar
Summary: Structural identifiability is crucial in model development, and problematic initial values may lead to unidentifiability, which can be resolved by changing these values.
SCIENTIFIC REPORTS
(2021)
Article
Automation & Control Systems
Nicolas Vanspranghe, Francesco Ferrante, Christophe Prieur
Summary: This article addresses a one-dimensional wave equation with a nonlinear dynamic boundary condition and boundary control, focusing on nonlinear stabilizing feedbacks dependent on the controlled extremity's velocity. It investigates exponential decay of mechanical energy and establishes stability and attractivity of suitable invariant sets.
IEEE TRANSACTIONS ON AUTOMATIC CONTROL
(2022)
Article
Biochemical Research Methods
Carlos Sequeiros, Irene Otero-Muras, Carlos Vazquez, Julio R. Banga
Summary: Mechanistic dynamic models are important for understanding biomolecular networks and biological systems. Stochastic dynamic models should be used when dealing with low copy numbers and biochemical stochasticity. This article presents a novel strategy for parameter estimation in stochastic dynamic models, employing global optimization and stochastic simulation techniques.
IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS
(2023)