Batteries play a critical role in electric vehicles and distributed energy generation.With the growing demand for energy storage solutions,new battery materials and systems are continually being developed.In this proc...Batteries play a critical role in electric vehicles and distributed energy generation.With the growing demand for energy storage solutions,new battery materials and systems are continually being developed.In this process,molecular dynamics(MD)simulations can reveal the microscopic mechanisms of battery processes,thereby boosting the design of batteries.Compared to other MD simulation techniques,the machine learning force field(MLFF)holds the advantages of first-principles accuracy along with large spatial and temporal scale,offering opportunities to uncover new mechanisms in battery systems.This review presents a detailed overview of the fundamental principles and model types of MLFFs,as well as their applications in simulating the structure,transport properties,and chemical reaction properties of bulk battery materials and interfaces.Notably,we emphasize the long-range interaction corrections and constant-potential methods in the model design of MLFFs.Finally,we discuss the challenges and prospects of applying MLFF models in the research of batteries.展开更多
Large-scale molecular dynamics(MD) simulations of crosslinked epoxy with quantum-level accuracy while capturing complex reactivity is a compelling yet unrealized challenge. In this work, through the construction of a ...Large-scale molecular dynamics(MD) simulations of crosslinked epoxy with quantum-level accuracy while capturing complex reactivity is a compelling yet unrealized challenge. In this work, through the construction of a chemical-environment-directing dataset, a reactive machine learning force field that accurately captures both reactive events and thermos-mechanical properties is developed. The force field achieves energy and force root-mean-square errors of 1.3 meV/atom and 159 meV/A, respectively, and operates approximately 1200 times faster than ab initio molecular dynamics. MD simulations demonstrate excellent predictive capabilities across multiple critical thermos-mechanical properties(radial distribution function, density, and elastic modulus), with results being well consistent with experimental values. In particular, the force field can provide accurate prediction of the bond dissociation energies for typical bonds with a mean absolute error of 7.8 kcal/mol(<8%), which enables the simulation of tensile-induced failure caused by chemical bond breaking. Our work demonstrates the capability of the machine learning force field to handle the extraordinary complexity of crosslinked epoxy systems, providing a valuable blueprint for future development of more generalized reactive force fields applicable to most polymers.展开更多
Bimetallic surfaces play a pivotal role in heterogeneous catalysis,yet their theoretical modeling has long been hindered by the computational chal-lenges of capturing configurational disorder,a critical feature govern...Bimetallic surfaces play a pivotal role in heterogeneous catalysis,yet their theoretical modeling has long been hindered by the computational chal-lenges of capturing configurational disorder,a critical feature governing their catalytic properties.Tradition-al approaches rely on oversimplified ordered surface models or restrict dis-order to a few atomic layers,limiting their predictive power.Here,we Cu_(1-x)Zn_(x)Cu_(1-x)Zn_(x)present an accurate and efficient computational framework that integrates machine learning force fields(MLFFs)with the cluster expansion(CE)method to study configurationally dis-ordered bimetallic surfaces at finite temperatures.We have developed an efficient workflow in which the MLFF is first trained iteratively via an active learning protocol,and then used to generate accurate energetic data for thousands of configurations that enable robust CE model construction.By treating bulk and surface clusters separately,we can build CE models for surface slabs with an arbitrary number of layers.Using as a case study,our CE-based Monte Carlo simulations reveal key structural insights that are relevant to the under-standing of catalytic properties of surfaces.This work demonstrates how MLFF-aided CE can overcome traditional limitations in theoretical modeling of bimetallic surfaces and highlights pathways toward more realistic modeling of heterogeneous catalysts.展开更多
Phase unwrapping is one of the key roles in fringe projection three-dimensional(3D)measurement technology.We propose a new method to achieve phase unwrapping in camera array light filed fringe projection 3D measuremen...Phase unwrapping is one of the key roles in fringe projection three-dimensional(3D)measurement technology.We propose a new method to achieve phase unwrapping in camera array light filed fringe projection 3D measurement based on deep learning.A multi-stream convolutional neural network(CNN)is proposed to learn the mapping relationship between camera array light filed wrapped phases and fringe orders of the expected central view,and is used to predict the fringe order to achieve the phase unwrapping.Experiments are performed on the light field fringe projection data generated by the simulated camera array fringe projection measurement system in Blender and by the experimental 3×3 camera array light field fringe projection system.The performance of the proposed network with light field wrapped phases using multiple directions as network input data is studied,and the advantages of phase unwrapping based on deep learning in light filed fringe projection are demonstrated.展开更多
Land plants in natural soil form intimate relationships with the diverse root bacterial microbiota. A growing body of evidence shows that these microbes are important for plant growth and health. Root microbiota compo...Land plants in natural soil form intimate relationships with the diverse root bacterial microbiota. A growing body of evidence shows that these microbes are important for plant growth and health. Root microbiota composition has been widely studied in several model plants and crops; however, little is known about how root microbiota vary throughout the plant's life cycle under field conditions. We performed longitudinal dense sampling in field trials to track the time-series shift of the root microbiota from two representative rice cultivars in two separate locations in China. We found that the rice root microbiota varied dramatically during the vegetative stages and stabilized from the beginning of the reproductive stage, after which the root microbiota underwent relatively minor changes until rice ripening. Notably, both rice genotype and geographical location influenced the patterns of root microbiota shift that occurred during plant growth. The relative abundance of Deltaproteobacteria in roots significantly increased overtime throughout the entire life cycle of rice, while that of Betaproteobacteria, Firmicutes, and Gammaproteobacteria decreased. By a machine learning approach, we identified biomarker taxa and established a model to correlate root microbiota with rice resident time in the field(e.g., Nitrospira accumulated from 5 weeks/tillering in field-grown rice). Our work provides insights into the process of rice root microbiota establishment.展开更多
The adoption of the Fifth Generation(5G)and beyond 5G networks is driving the demand for learning approaches that enable users to co-exist harmoniously in a multi-user distributed environment.Although resource-constra...The adoption of the Fifth Generation(5G)and beyond 5G networks is driving the demand for learning approaches that enable users to co-exist harmoniously in a multi-user distributed environment.Although resource-constrained,the Cognitive Radio(CR)has been identified as a key enabler of distributed 5G and beyond networks due to its cognitive abilities and ability to access idle spectrum opportunistically.Reinforcement learning is well suited to meet the demand for learning in 5G and beyond 5G networks because it does not require the learning agent to have prior information about the environment in which it operates.Intuitively,CRs should be enabled to implement reinforcement learning to efficiently gain opportunistic access to spectrum and co-exist with each other.However,the application of reinforcement learning is straightforward in a single-agent environment and complex and resource intensive in a multi-agent and multi-objective learning environment.In this paper,(1)we present a brief history and overview of reinforcement learning and its limitations;(2)we provide a review of recent multi-agent learning methods proposed and multi-agent learning algorithms applied in Cognitive Radio(CR)networks;and(3)we further present a novel framework for multi-CR reinforcement learning and conclude with a synopsis of future research directions and recommendations.展开更多
基金funding support from the National Natural Science Foundation of China(92472109,T2325012)the Program for HUST Academic Frontier Youth Team+1 种基金support from the Fundamental Research Funds for the Central Universities(HUST,5003120083)supported by the Postdoctoral Fellowship Program of CPSF(GZC20240532)。
文摘Batteries play a critical role in electric vehicles and distributed energy generation.With the growing demand for energy storage solutions,new battery materials and systems are continually being developed.In this process,molecular dynamics(MD)simulations can reveal the microscopic mechanisms of battery processes,thereby boosting the design of batteries.Compared to other MD simulation techniques,the machine learning force field(MLFF)holds the advantages of first-principles accuracy along with large spatial and temporal scale,offering opportunities to uncover new mechanisms in battery systems.This review presents a detailed overview of the fundamental principles and model types of MLFFs,as well as their applications in simulating the structure,transport properties,and chemical reaction properties of bulk battery materials and interfaces.Notably,we emphasize the long-range interaction corrections and constant-potential methods in the model design of MLFFs.Finally,we discuss the challenges and prospects of applying MLFF models in the research of batteries.
基金supported by the National Natural Science Foundation of China(Nos.52303116,52403125)the Natural Science Foundation of Hunan Province(No.2024JJ6461)+2 种基金the Science and Technology Innovation Program of Hunan Province(Nos.2022RC1080,2023RC3006)the Innovation Research Foundation of NUDT(Nos.22-ZZCX-076 and 23-ZZCX-ZZGC-01-10)the Key Research and Development Program of Hunan Province of China(No.2023ZJ1040).
文摘Large-scale molecular dynamics(MD) simulations of crosslinked epoxy with quantum-level accuracy while capturing complex reactivity is a compelling yet unrealized challenge. In this work, through the construction of a chemical-environment-directing dataset, a reactive machine learning force field that accurately captures both reactive events and thermos-mechanical properties is developed. The force field achieves energy and force root-mean-square errors of 1.3 meV/atom and 159 meV/A, respectively, and operates approximately 1200 times faster than ab initio molecular dynamics. MD simulations demonstrate excellent predictive capabilities across multiple critical thermos-mechanical properties(radial distribution function, density, and elastic modulus), with results being well consistent with experimental values. In particular, the force field can provide accurate prediction of the bond dissociation energies for typical bonds with a mean absolute error of 7.8 kcal/mol(<8%), which enables the simulation of tensile-induced failure caused by chemical bond breaking. Our work demonstrates the capability of the machine learning force field to handle the extraordinary complexity of crosslinked epoxy systems, providing a valuable blueprint for future development of more generalized reactive force fields applicable to most polymers.
基金supported by the National Natural Science Foundation of China(No.22273002)the National Key Research and Development Program of China(No.2022YFB4101401).We acknowledge the High-performance Computing Platform of Peking University for providing the computational facility.
文摘Bimetallic surfaces play a pivotal role in heterogeneous catalysis,yet their theoretical modeling has long been hindered by the computational chal-lenges of capturing configurational disorder,a critical feature governing their catalytic properties.Tradition-al approaches rely on oversimplified ordered surface models or restrict dis-order to a few atomic layers,limiting their predictive power.Here,we Cu_(1-x)Zn_(x)Cu_(1-x)Zn_(x)present an accurate and efficient computational framework that integrates machine learning force fields(MLFFs)with the cluster expansion(CE)method to study configurationally dis-ordered bimetallic surfaces at finite temperatures.We have developed an efficient workflow in which the MLFF is first trained iteratively via an active learning protocol,and then used to generate accurate energetic data for thousands of configurations that enable robust CE model construction.By treating bulk and surface clusters separately,we can build CE models for surface slabs with an arbitrary number of layers.Using as a case study,our CE-based Monte Carlo simulations reveal key structural insights that are relevant to the under-standing of catalytic properties of surfaces.This work demonstrates how MLFF-aided CE can overcome traditional limitations in theoretical modeling of bimetallic surfaces and highlights pathways toward more realistic modeling of heterogeneous catalysts.
基金the National Natural Science Foundation of China(No.61905178)the Science&Technology Development Fund of Tianjin Education Commission for Higher Education(No.2019KJ021)the Natural Science Foundation of Tianjin(No.18JCQNJC71100)。
文摘Phase unwrapping is one of the key roles in fringe projection three-dimensional(3D)measurement technology.We propose a new method to achieve phase unwrapping in camera array light filed fringe projection 3D measurement based on deep learning.A multi-stream convolutional neural network(CNN)is proposed to learn the mapping relationship between camera array light filed wrapped phases and fringe orders of the expected central view,and is used to predict the fringe order to achieve the phase unwrapping.Experiments are performed on the light field fringe projection data generated by the simulated camera array fringe projection measurement system in Blender and by the experimental 3×3 camera array light field fringe projection system.The performance of the proposed network with light field wrapped phases using multiple directions as network input data is studied,and the advantages of phase unwrapping based on deep learning in light filed fringe projection are demonstrated.
基金supported by the“Strategic Priority Research Program”of the Chinese Academy of Sciences(XDB11020700)CPSF-CAS Joint Foundation for Excellent Postdoctoral Fellows(2016LH00012)+1 种基金Strategic Priority Research Program of the Chinese Academy of Sciences(QYZDB-SSW-SMC021)the National Natural Science Foundation of China(31772400)
文摘Land plants in natural soil form intimate relationships with the diverse root bacterial microbiota. A growing body of evidence shows that these microbes are important for plant growth and health. Root microbiota composition has been widely studied in several model plants and crops; however, little is known about how root microbiota vary throughout the plant's life cycle under field conditions. We performed longitudinal dense sampling in field trials to track the time-series shift of the root microbiota from two representative rice cultivars in two separate locations in China. We found that the rice root microbiota varied dramatically during the vegetative stages and stabilized from the beginning of the reproductive stage, after which the root microbiota underwent relatively minor changes until rice ripening. Notably, both rice genotype and geographical location influenced the patterns of root microbiota shift that occurred during plant growth. The relative abundance of Deltaproteobacteria in roots significantly increased overtime throughout the entire life cycle of rice, while that of Betaproteobacteria, Firmicutes, and Gammaproteobacteria decreased. By a machine learning approach, we identified biomarker taxa and established a model to correlate root microbiota with rice resident time in the field(e.g., Nitrospira accumulated from 5 weeks/tillering in field-grown rice). Our work provides insights into the process of rice root microbiota establishment.
文摘The adoption of the Fifth Generation(5G)and beyond 5G networks is driving the demand for learning approaches that enable users to co-exist harmoniously in a multi-user distributed environment.Although resource-constrained,the Cognitive Radio(CR)has been identified as a key enabler of distributed 5G and beyond networks due to its cognitive abilities and ability to access idle spectrum opportunistically.Reinforcement learning is well suited to meet the demand for learning in 5G and beyond 5G networks because it does not require the learning agent to have prior information about the environment in which it operates.Intuitively,CRs should be enabled to implement reinforcement learning to efficiently gain opportunistic access to spectrum and co-exist with each other.However,the application of reinforcement learning is straightforward in a single-agent environment and complex and resource intensive in a multi-agent and multi-objective learning environment.In this paper,(1)we present a brief history and overview of reinforcement learning and its limitations;(2)we provide a review of recent multi-agent learning methods proposed and multi-agent learning algorithms applied in Cognitive Radio(CR)networks;and(3)we further present a novel framework for multi-CR reinforcement learning and conclude with a synopsis of future research directions and recommendations.
