Article
Multidisciplinary Sciences
V. Balaji, Fleur Couvreux, Julie Deshayes, Jacques Gautrais, Frederic Hourdinf, Catherine Rio
Summary: Traditional general circulation models (GCMs) have been a main tool in climate research, but they have limitations. Future models can address these limitations through higher resolution and machine learning techniques. Calibration is key to understanding the inner workings of complex systems. GCMs will continue to play a central role in future climate research.
PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA
(2022)
Article
Engineering, Mechanical
Qian Zhang, Yihan Zhong, Zizhe Wang, Ahmad B. H. Kueh, Jianguo Cai, Jian Feng
Summary: This paper proposes a mechanical model with the prestress and deflection based on Foppl-von Karman equations, and introduces nonlinear buckling finite element analysis and experiments to determine key parameters. The proposed model shows accurate predictive ability for stress and configuration parameters of wrinkles, and a general calculation procedure for wrinkle parameters of complex membranes is established.
INTERNATIONAL JOURNAL OF MECHANICAL SCIENCES
(2022)
Article
Mechanics
Weibing Deng, Rongrong Xie, Shegfeng Deng, Armen E. Allahverdyan
Summary: By comparing the first half of a text to its second half, it is possible to distinguish meaningful texts from meaningless sets of symbols based on statistical features. These features include more diverse and rare words in the first half, as well as less homogeneous distribution of words. The asymmetry in temporal organization of meaningful texts is revealed through these differences.
JOURNAL OF STATISTICAL MECHANICS-THEORY AND EXPERIMENT
(2021)
Article
Mathematics
Xing-Bin Pan
Summary: This paper studies a nonlinear magneto-static model on a general domain in R-3 with multiple connections and holes, with a nonlinear relation between magnetic induction B and magnetic field H. The existence of solutions of the boundary value problems and a more general nonlinear Maxwell-Stokes system with topological parameters are proven, demonstrating the effects of domain topology on electromagnetic fields and nonlinear systems involving curl.
JOURNAL OF DIFFERENTIAL EQUATIONS
(2021)
Article
Multidisciplinary Sciences
Xiaotie Deng, Ningyuan Li, David Mguni, Jun Wang, Yaodong Yang
Summary: We prove that solving Markov perfect equilibria of general-sum stochastic games is PPAD-complete, establishing the algorithmic complexity for multi-agent reinforcement learning methodology. Similar to Markov decision processes in reinforcement learning, Markov games (or stochastic games) form the basis for studying multi-agent reinforcement learning and sequential agent interactions. We introduce approximate Markov perfect equilibrium as a solution to the computational problem of finite-state stochastic games repeated in the infinite horizon and prove its PPAD-completeness. This solution concept maintains the Markov perfect property and enables the extension of successful multi-agent reinforcement learning algorithms from static two-player games to dynamic multi-agent games, expanding the scope of the PPAD-complete class.
NATIONAL SCIENCE REVIEW
(2023)
Article
Automation & Control Systems
Lijuan Li, Xingyu Chen, Shipin Yang
Summary: This article introduces a method to assess controller parameter performance based on disturbance characteristic variation in an industrial control system. The Markov parameters of a disturbance model are used to design the assessing index of controller parameters. Simulation results show the validity of the proposed index in both the Wood-Berry model and TE process.
INTERNATIONAL JOURNAL OF ADAPTIVE CONTROL AND SIGNAL PROCESSING
(2021)
Article
Mathematics
Andrey Borisov, Alexey Bosov, Gregory Miller
Summary: This paper presents an optimal control problem for a partially observable stochastic differential system driven by an external Markov jump process. The filter optimization problem is solved using the Wonham filter, while the control problem is solved by formulating an equivalent control problem and applying dynamic programming. The applicability of the theoretical results is then illustrated through a numerical example.
Article
Automation & Control Systems
Feng Li, Shengyuan Xu, Hao Shen, Zhengqiang Zhang
Summary: This article discusses the extended dissipativity-based control problem for singularly perturbed systems with Markov jump parameters, considering partial information on the Markov chain. A comprehensive hidden Markov model is established to address the partial information issues, leading to a criterion for analyzing the extended stochastic dissipativity of the systems. The theoretical results are validated through an illustrative example and a vehicle active suspension system.
IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS
(2021)
Article
Engineering, Chemical
Yunyan Huang, Abraham Sagiv, Raphael Semiat, Hilla Shemer
Summary: Donnan Dialysis (DD) uses ion exchange membranes to selectively transport target ions, and a computational model has been developed to predict ion concentration profiles and transport mechanisms in DD processes. The model has been verified and validated with experimental data, demonstrating its accuracy and reliability.
JOURNAL OF MEMBRANE SCIENCE
(2022)
Article
Geosciences, Multidisciplinary
Kees Nederhoff, Jasper Hoek, Tim Leijnse, Maarten van Ormondt, Sofia Caires, Alessio Giardino
Summary: A new method for generating synthetic tropical cyclone tracks has been proposed, which can help overcome data scarcity limitations and reduce reliance on historical data for coastal risk assessment. The TCWiSE tool, based on empirical track modeling using Markov chains, can simulate thousands of synthetic TC tracks and wind fields in any oceanic basin for reliable assessment of coastal hazards. Validation results in the Gulf of Mexico show that TC patterns and extreme wind speeds can be accurately reproduced by TCWiSE.
NATURAL HAZARDS AND EARTH SYSTEM SCIENCES
(2021)
Article
Mathematics
Alexander Zhdanok
Summary: In this paper, the authors study general Markov chains in an arbitrary measurable space with discrete time. They analyze the properties of Cesaro means of powers of Markov operators on the set of finitely additive probability measures. The authors prove that the set of limit measures of such sequences in the weak topology is non-empty, weakly compact, and all of them are invariant for the operator. They also establish the equivalence between the well-known Doeblin condition and the condition that all invariant finitely additive measures of the Markov chain are countably additive.
Article
Energy & Fuels
Baoting Huang, Naseem S. Hayek, Guanjie Sun, Sogol Mottaghi-Tabar, David S. A. Simakov, Oz M. Gazit
Summary: Identifying key catalyst parameters that govern catalytic performance is a major challenge. Through a simplified methodology, this study highlights the importance of highly dispersed Mn2O3 particles in a dispersed Na2WO4 melt for promoting the oxidative coupling of methane. Experimental results and literature data show consistent correlations, providing insights into the behavior of the catalyst under OCM conditions.
Article
Engineering, Civil
Chen Fang, Hong-Jun Liu, Heung-Fai Lam, Mujib Olamide Adeagbo, Hua-Yi Peng
Summary: The Bayesian model updating framework is a reliable method for constructing high-fidelity finite element models and provides a practical solution for efficient model updating of large-scale civil engineering structures. The framework was applied to update the finite element model of the Ting Kau Bridge in Hong Kong using measured modal parameters. The results demonstrate that the framework accurately updated the finite element model and can facilitate structural health monitoring of large-scale civil engineering structures.
ENGINEERING STRUCTURES
(2022)
Article
Automation & Control Systems
Jianqing Fan, Bai Jiang, Qiang Sun
Summary: This paper establishes Hoeffding's lemma and inequality for bounded functions of general-state space and not necessarily reversible Markov chains, showing the necessity of boundedness of functions for such results. The optimality of the ratio between variance proxies in the Markov-dependent and independent settings is characterized. The new results are applied to various practical problems to showcase their usefulness.
JOURNAL OF MACHINE LEARNING RESEARCH
(2021)
Article
Ecology
Timothy A. Gowan, Michael D. Tringali, Jeffrey A. Hostetler, Julien Martin, Leslie I. Ward-Geiger, Jennifer M. Johnson
Summary: This paper introduces a hidden Markov model for estimating age-specific survival when age is unknown, which is evaluated through simulations and implemented in commonly used software. The model is applied to genetic capture-recapture data of Florida manatees to estimate survival probabilities, demonstrating its broad applicability in studies aiming to quantify age-specific effects on population dynamics.
