Property Prediction for Complex Compounds Using Structure-Free Mendeleev Encoding and Machine Learning DOI
Zixin Zhuang, Amanda S. Barnard

Journal of Chemical Information and Modeling, Год журнала: 2024, Номер unknown

Опубликована: Дек. 12, 2024

Predicting the properties for unseen materials exclusively on basis of chemical formula before synthesis and characterization has advantages research resource planning. This can be achieved using suitable structure-free encoding machine learning methods, but additional processing decisions are required. In this study, we compare a variety encodings algorithms to predict structure/property relationships battery materials. It was found that physical units used measure property labels have an important impact predictive ability models, regardless computational approach. Property with respect weight give excellent performance, volume cannot predicted confidence only information, even when underlying characteristics same. These results contrast previous studies unsupervised classification, where excelled, highlight how structural features or represented plays role in models.

Язык: Английский

Progress of machine learning in materials design for Li-Ion battery DOI Creative Commons

Prasshanth C.V.,

Arun Kumar Lakshminarayanan,

R. Brindha

и другие.

Next Materials, Год журнала: 2024, Номер 2, С. 100145 - 100145

Опубликована: Янв. 1, 2024

The widespread adoption of lithium-ion batteries has ushered in a transformative era across industries, powering an array devices from portable electronics to electric vehicles. This review explores recent advancements machine learning tools tailored for improving battery materials, management strategies, and system-level optimization. It provides comprehensive overview the current landscape, emphasizing less-explored evolution algorithms materials. Machine integration enhances our understanding material properties, accelerates discovery efficient compositions, contributes development more durable batteries. article also delves into learnings role predicting State Health remaining useful life, crucial proactive maintenance. highlights how integrating field potential revolutionize design accelerate energy storage technology, promising sustainable technologically advanced future.

Язык: Английский

Процитировано

10

Future of battery thermal management systems (BTMS): Role of advanced technologies, artificial intelligence and sustainability DOI Creative Commons
M. M. Quazi, Farzad Jaliliantabar, K. Sudhakar

и другие.

Next Sustainability, Год журнала: 2025, Номер 6, С. 100114 - 100114

Опубликована: Янв. 1, 2025

Язык: Английский

Процитировано

1

Layered double hydroxides: next promising materials for energy storage and conversion DOI Creative Commons

Kui Fan,

Pengcheng Xu, Zhenhua Li

и другие.

Next Materials, Год журнала: 2023, Номер 1(4), С. 100040 - 100040

Опубликована: Сен. 27, 2023

Layered double hydroxides (LDHs) are a family of two-dimensional (2D) layered materials with controllable supramolecular structure and unique physicochemical properties, making them highly attractive in the fields energy storage conversion. Considering intense interest LDHs family, this review aims to provide comprehensive summary their development history, synthesis strategies, energy-related applications. Special attention is given distinctive properties LDHs, such as oriented assembly topological transformation, which can serve systematic guidance for preparation LDHs-based nanostructures. Furthermore, outlines both classical cutting-edge applications electrocatalysis. Of particular interest, emerging coupling system based on electrocatalytic water splitting thoroughly analyzed. Finally, prospects potential challenges discussed, aiming raise awareness among researchers stimulate further progress material development.

Язык: Английский

Процитировано

22

Compilation and deciphering MoS2’s physical properties: Accurate benchmark DFT simulations and assessment of advanced quantum methods DOI
Najeh Rekik, Ibrahim Isah, Norah A. M. Alsaif

и другие.

Chemical Physics, Год журнала: 2024, Номер 580, С. 112229 - 112229

Опубликована: Фев. 8, 2024

Язык: Английский

Процитировано

6

The Contribution of Artificial Intelligence to Phase Change Materials in Thermal Energy Storage: From Prediction to Optimization DOI
Shuli Liu,

Junrui Han,

Yongliang Shen

и другие.

Renewable Energy, Год журнала: 2024, Номер unknown, С. 121973 - 121973

Опубликована: Ноя. 1, 2024

Язык: Английский

Процитировано

6

Enhancing energy materials with data-driven methods: A roadmap to long-term hydrogen energy sustainability using machine learning DOI
Cheng Li, Jianjun Ma, D. R. Gibson

и другие.

International Journal of Hydrogen Energy, Год журнала: 2025, Номер 119, С. 108 - 125

Опубликована: Март 21, 2025

Язык: Английский

Процитировано

0

Emergent fullerene nanocomposites with conjugated matrices—An overview DOI Creative Commons
Ayesha Kausar, Ishaq Ahmad

Next Materials, Год журнала: 2024, Номер 2, С. 100131 - 100131

Опубликована: Янв. 1, 2024

Fullerene, an intimate zero dimensional nanocarbon, has been frequently adopted as nanocomposite reinforcement. Imperative utilizations of fullerene and derived nanocomposites have observed in energy, optical, electronics devices sectors. Conjugated or conductive polymers π-conjugation backbone system leading to semiconducting features. In addition fine electron conduction, these advantages low weight, facile processing, chemical thermal robustness. Including nanocarbon dopants conjugated enhanced several technical features matrices. Accordingly, this progressive overview highlights design, properties, potential reinforced conducting polymer nanocomposites. The structural multiplicity polymer/fullerene enabled advanced potential. Consequently, ensuing revealed efficient structural, morphological, physical characteristics, along with revolts technological fields like photovoltaics, supercapacitors, sensors, etc. Forthcoming research on ground-breaking may daze design/performance challenges towards large scale practical applications.

Язык: Английский

Процитировано

3

Inorganic–organic modular silicon and dye-sensitized solar cells and predicted role of artificial intelligence towards efficient and stable solar chargers based on supercapacitors DOI Creative Commons
Ireneusz Plebankiewicz, Krzysztof Artur Bogdanowicz, Paweł Kwaśnicki

и другие.

Scientific Reports, Год журнала: 2024, Номер 14(1)

Опубликована: Март 13, 2024

Abstract Appropriate and rational management of the energy produced by renewable sources is one most urgent challenges for global sector. This paper devoted to systematic experimental theoretical studies a modular solar charger based on silicon dye-sensitized cells as an source, supercapacitor bank. Using MathCAD program, I–V characteristics were plotted both single cell photovoltaic module various series-to-parallel connections. To assess surface quality modules, additional tests using thermal imaging camera carried out well. The charging (two series-connected with capacity 300 F), determined depending parameters well considering influence voltage balancing system control system. charge, discharge, recharge carefully analyzed optimize operating conditions, i.e. number cells. evaluate stability operation time, their temperature dependence (17–65 °C), modules tested ten days under Central European weather conditions. Importantly, comparative analysis chargers different configurations showed increase in electrical proposed inorganic–organic concept compared alone rigid substrate. Finally, preliminary assumptions (requirements) developed regarding optical new that could be used innovative instead along predicted role artificial intelligence (AI) these devices.

Язык: Английский

Процитировано

2

Estimating best nanomaterial for energy harvesting through reinforcement learning DQN coupled with fuzzy PROMETHEE under road-based conditions DOI Creative Commons

Sekar Kidambi Raju,

Ganesh Karthikeyan Varadarajan,

Amal H. Alharbi

и другие.

Scientific Reports, Год журнала: 2024, Номер 14(1)

Опубликована: Окт. 14, 2024

Energy harvesters based on nanomaterials are getting more and popular, but their way to commercial availability, some crucial issues still need be solved. The objective of the study is select an appropriate nanomaterial. Using features Reinforcement Deep Q-Network (DQN) in conjunction with Fuzzy PROMETHEE, proposed model, we present this work a hybrid fuzzy approach selecting materials for vehicle-environmental-hazardous substance (EHS) combination that operates roadways under traffic conditions. DQN able accumulate useful experience operating dynamic environment, accordingly deliver highest energy output at same time bring consideration factors such as durability, cost, environmental impact. PROMETHEE allows participation human experts during decision-making process, going beyond quantitative data typically learned by through inclusion qualitative preferences. Instead, method unites strength individual approaches, result providing highly resistant adjustable material selection real EHS. pointed out can give high efficiency reference years service, price, effects. model provides 95% accuracy computational 300 s, application hypothesis practical testing chosen showed selected harvest fluctuating conditions proved concept True Vehicle Environmental High-risk Substance scenarios.

Язык: Английский

Процитировано

1

Property Prediction for Complex Compounds Using Structure-Free Mendeleev Encoding and Machine Learning DOI
Zixin Zhuang, Amanda S. Barnard

Journal of Chemical Information and Modeling, Год журнала: 2024, Номер unknown

Опубликована: Дек. 12, 2024

Predicting the properties for unseen materials exclusively on basis of chemical formula before synthesis and characterization has advantages research resource planning. This can be achieved using suitable structure-free encoding machine learning methods, but additional processing decisions are required. In this study, we compare a variety encodings algorithms to predict structure/property relationships battery materials. It was found that physical units used measure property labels have an important impact predictive ability models, regardless computational approach. Property with respect weight give excellent performance, volume cannot predicted confidence only information, even when underlying characteristics same. These results contrast previous studies unsupervised classification, where excelled, highlight how structural features or represented plays role in models.

Язык: Английский

Процитировано

1