Accurate retrieval of atmospheric vertical profiles is critical for improving weather prediction and climate monitoring.However,the complexity of atmospheric processes in cloudy regions poses challenges compared to th...Accurate retrieval of atmospheric vertical profiles is critical for improving weather prediction and climate monitoring.However,the complexity of atmospheric processes in cloudy regions poses challenges compared to those of clear sky scenarios.This study presents a novel framework that integrates Bayesian optimization and machine learning approaches to retrieve atmospheric vertical profiles—including temperature,humidity,ozone concentration,cloud fraction,ice water content(IWC),and liquid water content(LWC)—from hyperspectral infrared observations.Specifically,a Bayesian method was used to refine ERA5 reanalysis data by minimizing brightness temperature(BT)discrepancies against FY-4B Geostationary Interferometric Infrared Sounder(GIIRS)observations,generating a high-quality profile database(~2.8 million profiles)across diverse weather systems.The optimized profiles improve radiative consistency,reducing BT biases from>40 K to<10 K in cloudy regions.To further overcome the limitations of the Bayesian method,we developed a Transformer-Resnet hybrid model(TERNet),which achieved superior performance with RMSE values of 1.61 K(temperature),5.77%(humidity),and 2.25×10^(–6)/6.09×10^(–6)kg kg^(–1)(IWC/LWC)across the entire vertical levels in all-sky conditions.The TERNet outperforms both ERA5 in cloud parameter retrieval and the GIIRS L2 product in thermodynamic profiling.Independent verification with radiosonde and Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observations(CALIPSO)datasets confirms the framework's reliability across various meteorological regimes.This work demonstrates the capability of combining physics-informed Bayesian methods with data-driven machine learning to fully exploit hyperspectral IR data.展开更多
Artificial intelligence(AI)based models have been used to predict the structural,optical,mechanical,and electrochemical properties of zinc oxide/graphene oxide nanocomposites.Machine learning(ML)models such as Artific...Artificial intelligence(AI)based models have been used to predict the structural,optical,mechanical,and electrochemical properties of zinc oxide/graphene oxide nanocomposites.Machine learning(ML)models such as Artificial Neural Networks(ANN),Support Vector Regression(SVR),Multilayer Perceptron(MLP),and hybrid,along with fuzzy logic tools,were applied to predict the different properties like wavelength at maximum intensity(444 nm),crystallite size(17.50 nm),and optical bandgap(2.85 eV).While some other properties,such as energy density,power density,and charge transfer resistance,were also predicted with the help of datasets of 1000(80:20).In general,the energy parameters were predicted more accurately by hybrid models.The hydrothermal method was used to synthesize graphene oxide(GO)and zinc oxide(ZnO)nanocomposites.The increased surface area,conductivity,and stability of graphene oxide in zinc oxide nanoparticles make the composite an ideal option for energy storage.X-ray diffraction(XRD)confirmed the crystallite size of 17.41 nm for the nanocomposite and the presence of GO(12.8○)peaks.The scanning electron microscope(SEM)showed anchored wrinkled GO sheets on zinc oxide with an average particle size of 2.93μm.Energy-dispersive X-ray spectroscopy(EDX)confirmed the elemental composition,and Fouriertransform infrared spectroscopy(FTIR)revealed the impact of GO on functional groups and electrochemical behavior.Photoluminescence(PL)wavelength of(439 nm)and band gap of(2.81 eV)show that the material is suitable for energy applications in nanocomposites.Smart nanocomposite materials with improved performance in energy storage and related applications were fabricated by combining synthesis,characterization,fuzzy logic,and machine learning in this work.展开更多
At present,the polymerase chain reaction(PCR)amplification-based file retrieval method is the mostcommonly used and effective means of DNA file retrieval.The number of orthogonal primers limitsthe number of files that...At present,the polymerase chain reaction(PCR)amplification-based file retrieval method is the mostcommonly used and effective means of DNA file retrieval.The number of orthogonal primers limitsthe number of files that can be accurately accessed,which in turn affects the density in a single oligo poolof digital DNA storage.In this paper,a multi-mode DNA sequence design method based on PCR file retrie-val in a single oligonucleotide pool is proposed for high-capacity DNA data storage.Firstly,by analyzingthe maximum number of orthogonal primers at each predicted primer length,it was found that the rela-tionship between primer length and the maximum available primer number does not increase linearly,and the maximum number of orthogonal primers is on the order of 10^(4).Next,this paper analyzes themaximum address space capacity of DNA sequences with different types of primer binding sites for filemapping.In the case where the capacity of the primer library is R(where R is even),the number ofaddress spaces that can be mapped by the single-primer DNA sequence design scheme proposed in thispaper is four times that of the previous one,and the two-level primer DNA sequence design scheme can reach [R/2·(R/2-1)]^(2)times.Finally,a multi-mode DNA sequence generation method is designed based onthe number of files to be stored in the oligonucleotide pool,in order to meet the requirements of the ran-dom retrieval of target files in an oligonucleotide pool with large-scale file numbers.The performance ofthe primers generated by the orthogonal primer library generator proposed in this paper is verified,andthe average Gibbs free energy of the most stable heterodimer formed between the orthogonal primersproduced is−1 kcal·(mol·L^(−1))^(−1)(1 kcal=4.184 kJ).At the same time,by selectively PCR-amplifying theDNA sequences of the two-level primer binding sites for random access,the target sequence can be accu-rately read with a minimum of 10^(3) reads,when the primer binding site sequences at different positionsare mutually different.This paper provides a pipeline for orthogonal primer library generation and multi-mode mapping schemes between files and primers,which can help achieve precise random access to filesin large-scale DNA oligo pools.展开更多
This study investigates the prediction of groundwater Storage in the Rabat-Sale-Kenitra region under climate change conditions using advanced machine learning models.A comprehensive dataset encompassing hydrological,m...This study investigates the prediction of groundwater Storage in the Rabat-Sale-Kenitra region under climate change conditions using advanced machine learning models.A comprehensive dataset encompassing hydrological,meteorological,and geological factors was meticulously curated and preprocessed for model training.Six regression models Decision Tree,Random Forest,LightGBM,CatBoost,Extreme Learning Machine(ELM),and Artificial Neural Network(ANN)were employed to predict groundwater Storage,with hyperparameters optimized via grid search.The performance of these models was rigorously evaluated using metrics such as Root Mean Squared Error(RMSE),Mean Absolute Error(MAE),and the coefficient of determination(R^(2)).Results demonstrated that the LightGBM model outperformed the others,achieving an impressive testing RMSE of 3.07 and an R^(2)of 0.9997,indicating its robustness in handling large datasets.The Extreme Learning Machine and ANN showed considerable limitations,highlighting the importance of model selection.This research underscores the critical role of advanced machine learning techniques in enhancing groundwater resource management,providing valuable insights for policymakers in developing sustainable strategies to address groundwater challenges in the face of climate variability.展开更多
High-entropy materials(HEMs)show exceptional mechanical properties,highly adjustable chemical characteristics,and outstanding stability,making them suitable for energy storage.However,the broad compositional space and...High-entropy materials(HEMs)show exceptional mechanical properties,highly adjustable chemical characteristics,and outstanding stability,making them suitable for energy storage.However,the broad compositional space and intricate chemical interactions in HEMs present challenges to traditional trial-and-error research methods,restricting their efficacy in swift screening and synthesis.Hence,the application of machine learning(ML)to the realm of high-entropy materials and energy storage becomes imperative.ML demonstrates its formidable capabilities for navigating the complexity of HEMs,with their diverse metal components,structures and property combinations,to advance energy storage applications.This review comprises the following sections:a concise introduction to the general process of ML in the energy materials field,a summary of HEMs in the energy storage field,a review of the latest achievements of ML in the HEMs and energy storage field,and finally,an exploration of current challenges and prospects in this interdisciplinary arena.With the advent of ML,the precision of its predictions and the efficiency of its screening methods have offered novel perspectives for material research,expediting the discovery and application of new materials.This article contributes to the advancement of research in related fields,hastening the development of novel materials to meet the escalating energy demands and promote sustainable development goals.展开更多
To achieve carbon dioxide(CO_(2))storage through enhanced oil recovery,accurate forecasting of CO_(2) subsurface storage and cumulative oil production is essential.This study develops hybrid predictive models for the ...To achieve carbon dioxide(CO_(2))storage through enhanced oil recovery,accurate forecasting of CO_(2) subsurface storage and cumulative oil production is essential.This study develops hybrid predictive models for the determination of CO_(2) storage mass and cumulative oil production in unconventional reservoirs.It does so with two multi-layer perceptron neural networks(MLPNN)and a least-squares support vector machine(LSSVM),hybridized with grey wolf optimization(GWO)and/or particle swarm optimization(PSO).Large,simulated datasets were divided into training(70%)and testing(30%)groups,with normalization applied to both groups.Mahalanobis distance identifies/eliminates outliers in the training subset only.A non-dominated sorting genetic algorithm(NSGA-II)combined with LSSVM selected seven influential features from the nine available input parameters:reservoir depth,porosity,permeability,thickness,bottom-hole pressure,area,CO_(2) injection rate,residual oil saturation to gas flooding,and residual oil saturation to water flooding.Predictive models were developed and tested,with performance evaluated with an overfitting index(OFI),scoring analysis,and partial dependence plots(PDP),during training and independent testing to enhance model focus and effectiveness.The LSSVM-GWO model generated the lowest root mean square error(RMSE)values(0.4052 MMT for CO_(2) storage and 9.7392 MMbbl for cumulative oil production)in the training group.That trained model also exhibited excellent generalization and minimal overfitting when applied to the testing group(RMSE of 0.6224 MMT for CO_(2) storage and 12.5143 MMbbl for cumulative oil production).PDP analysis revealed that the input features“area”and“porosity”had the most influence on the LSSVM-GWO model's pre-diction performance.This paper presents a new hybrid modeling approach that achieves accurate forecasting of CO_(2) subsurface storage and cumulative oil production.It also establishes a new standard for such forecasting,which can lead to the development of more effective and sustainable solutions for oil recovery.展开更多
In order to evaluate the efficiency of the automated storage/retrieval system(AS/RS)accurately,and compare different layouts of the AS/RS using mean travel time,under randomized storage conditions,an exact,geometry-ba...In order to evaluate the efficiency of the automated storage/retrieval system(AS/RS)accurately,and compare different layouts of the AS/RS using mean travel time,under randomized storage conditions,an exact,geometry-based analytical model is presented.The model can be used to compute the expected single-command and dual-command travel time for a storage/retrieval(S/R)machine which can travel simultaneously horizontally and vertically as it moves along a storage aisle.The rack may be either square in time or non square in time.Additionally,the alternative layouts of the AS/RS and travel-time models are examined.Comparing with setting the I/O point at the left-lower corner of the rack,setting the I/O point at any point at the vertical edge can help enhance the efficiency of the AS/RS.展开更多
Magnesium hydride(MgH_(2)),a promising high-capacity hydrogen storage material,is hindered by slow dehydrogenation kinetics.AIdriven catalyst discovery to address this is often hampered by the laborious extraction of ...Magnesium hydride(MgH_(2)),a promising high-capacity hydrogen storage material,is hindered by slow dehydrogenation kinetics.AIdriven catalyst discovery to address this is often hampered by the laborious extraction of data from unstructured literature.To overcome this,we introduce a transformative“LLM to Agent”framework that synergistically integrates Large Language Models(LLMs)for automated data curation with Machine Learning(ML)for predictive design.We automatically constructed a comprehensive database of 809 MgH_(2)catalysts(6555 data rows)with high fidelity and an~40-fold acceleration over manual methods.The resulting ML models achieved high accuracy(average R^(2)>0.91)in predicting dehydrogenation temperature and activation energy,subsequently guiding a Genetic Algorithm(GA)in an exploratory inverse design that autonomously uncovered key design principles for high-performance catalysts.Encouragingly,a strong alignment was found between these AI-discovered principles and the design strategies of recently reported,state-of-the-art experimental systems,providing substantial evidence for the validity of our approach.The framework culminates in Cat-Advisor,a novel,domain-adapted multi-agent system.Cat-Advisor translates ML predictions and retrieval-augmented knowledge into actionable design guidance,demonstrating capabilities that surpass those of general-purpose LLMs in this specialized domain.This work delivers a practical AI toolkit for accelerated materials discovery and advances the emerging Agent-based paradigm for designing next-generation energy technologies.展开更多
In this paper, we present a novel Support Vector Machine active learning algorithm for effective 3D model retrieval using the concept of relevance feedback. The proposed method learns from the most informative objects...In this paper, we present a novel Support Vector Machine active learning algorithm for effective 3D model retrieval using the concept of relevance feedback. The proposed method learns from the most informative objects which are marked by the user, and then creates a boundary separating the relevant models from irrelevant ones. What it needs is only a small number of 3D models labelled by the user. It can grasp the user＇s semantic knowledge rapidly and accurately. Experimental results showed that the proposed algorithm significantly improves the retrieval effectiveness. Compared with four state-of-the-art query refinement schemes for 3D model retrieval, it provides superior retrieval performance after no more than two rounds of relevance feedback.展开更多
Due to its characteristics distribution and virtualization, cloud storage also brings new security problems. User＇s data is stored in the cloud, which separated the ownership from management. How to ensure the securi...Due to its characteristics distribution and virtualization, cloud storage also brings new security problems. User＇s data is stored in the cloud, which separated the ownership from management. How to ensure the security of cloud data, how to increase data availability and how to improve user privacy perception are the key issues of cloud storage research, especially when the cloud service provider is not completely trusted. In this paper, a cloud storage ciphertext retrieval scheme based on AES and homomorphic encryption is presented. This ciphertext retrieval scheme will not only conceal the user retrieval information, but also prevent the cloud from obtaining user access pattern such as read-write mode, and access frequency, thereby ensuring the safety of the ciphertext retrieval and user privacy. The results of simulation analysis show that the performance of this ciphertext retrieval scheme requires less overhead than other schemes on the same security level.展开更多
E-commerce, as an emerging marketing mode, has attracted more and more attention and gradually changed the way of our life. However, the existing layout of distribution centers can＇t fulfill the storage and picking d...E-commerce, as an emerging marketing mode, has attracted more and more attention and gradually changed the way of our life. However, the existing layout of distribution centers can＇t fulfill the storage and picking demands of e-commerce sufficiently. In this paper, a modified miniload automated storage/retrieval system is designed to fit these new characteristics of e-commerce in logistics. Meanwhile, a matching problem, concerning with the improvement of picking efficiency in new system, is studied in this paper. The problem is how to reduce the travelling distance of totes between aisles and picking stations. A multi-stage heuristic algorithm is proposed based on statement and model of this problem. The main idea of this algorithm is, with some heuristic strategies based on similarity coefficients, minimizing the transportations of items which can not arrive in the destination picking stations just through direct conveyors. The experimental results based on the cases generated by computers show that the average reduced rate of indirect transport times can reach 14.36% with the application of multi-stage heuristic algorithm. For the cases from a real e-commerce distribution center, the order processing time can be reduced from 11.20 h to 10.06 h with the help of the modified system and the proposed algorithm. In summary, this research proposed a modified system and a multi-stage heuristic algorithm that can reduce the travelling distance of totes effectively and improve the whole performance of e-commerce distribution center.展开更多
Support vector machines have met with significant success in the information retrieval field, especially in handling text classification tasks. Although various performance estimators for SVMs have been proposed, thes...Support vector machines have met with significant success in the information retrieval field, especially in handling text classification tasks. Although various performance estimators for SVMs have been proposed, these only focus on accuracy which is based on the leave-one-out cross validation procedure. Information-retrieval-related performance measures are always neglected in a kernel learning methodology. In this paper, we have proposed a set of information-retrieval-oriented performance estimators for SVMs, which are based on the span bound of the leave-one-out procedure. Experiments have proven that our proposed estimators are both effective and stable.展开更多
Permanent magnet homopolar inductor machine(PMHIM) has attracted much attention in the field of flywheel energy storage system(FESS) due to its merits of simple structure,brushless excitation, and rotor flywheel integ...Permanent magnet homopolar inductor machine(PMHIM) has attracted much attention in the field of flywheel energy storage system(FESS) due to its merits of simple structure,brushless excitation, and rotor flywheel integration. However, the air-gap flux generated by the PM cannot be adjusted, which would cause large electromagnetic losses in the standby operation state of FESS. To solve this problem, a novel mechanically adjusted variable flux permanent magnet homopolar inductor machine with rotating magnetic poles(RMP-PMHIM) is proposed in this paper. The permanent magnet poles are rotated by an auxiliary rotating device and the purpose of changing the air-gap flux is achieved. First, the structure and operation principle of the proposed RMP-PMHIM are explained. Second,the flux weakening principle of the RMP-PMHIM is analyzed and the equivalent magnetic circuit models under different flux weakening states are built. Third, the parameters of the PM and its fixed structure are optimized to obtain the good electromagnetic performance. Fourth, the electromagnetic performance, including the air-gap flux density, back-EMF, flux weakening ability, loss, etc. of the proposed RMP-PMHIM are investigated and compared. Compared with the non-rotating state of the PM of RPM-PMHIM, the air-gap flux density amplitude can be weakened by 99.95% when the PM rotation angle is 90 degrees, and the no-load core loss can be suppressed by 99.98%,which shows that the proposed RPM-PMHIM is a good candidate for the application of FESS.展开更多
At present,the traditional blockchain for data storage and retrieval reflects the characteristics of slow data uploading speed,high cost,and transparency,and there are a lot of corresponding problems,such as not suppo...At present,the traditional blockchain for data storage and retrieval reflects the characteristics of slow data uploading speed,high cost,and transparency,and there are a lot of corresponding problems,such as not supporting private data storage,large data operation costs,and not supporting Data field query.This paper proposes a method of data encryption storage and retrieval based on the IOTA distributed ledger,combined with the fast transaction processing speed and zero-value transactions of the IOTA blockchain,through the Masked Authenticated Messaging technology,so that the data is encrypted in the data stream.The form is stored in the distributed ledger,quickly retrieved through the field index mechanism established by the data form,and the data operation is carried out on the chain.Experimental results show that this system has high storage,encryption and retrieval performance,and good practicability.展开更多
We propose a new approach to store and query XML data in an RDBMS basing on the idea of the numbering scheme and inverted list. O ur approach allows us to quickly determine the precedence, sibling and ancestor/ desc...We propose a new approach to store and query XML data in an RDBMS basing on the idea of the numbering scheme and inverted list. O ur approach allows us to quickly determine the precedence, sibling and ancestor/ descendant relationships between any pair of nodes in the hierarchy of XML, and utilize path index to speed up calculating of path expressions. Examples have de monstrated that our approach can effectively and efficiently support both XQuery queries and keyword searches. Our approach is also flexible enough to support X ML documents both with Schema and without Schema, and applications both retrieva l and update. We also present the architecture of middleware for application acc essing XML documents stored in relations, and an algorithm translating a given X ML document into relations effectively.展开更多
In the global transition towards sustainable energy sources,hydrogen energy has emerged as an indispensable pillar in reshaping the energy landscape,owing to its environmental sustainability,zero emissions,and high ef...In the global transition towards sustainable energy sources,hydrogen energy has emerged as an indispensable pillar in reshaping the energy landscape,owing to its environmental sustainability,zero emissions,and high efficiency.Nevertheless,the large-scale deployment of hydrogen energy is confronted with substantial technical barriers in storage and transportation.Although contemporary research has shifted focus to the development of highly efficient hydrogen storage materials,conventional material design concepts remain predominantly empirical,typically relying on trial-and-error methodologies.Importantly,the widespread application of artificial intelligence technologies in accelerating materials discovery and optimization has attracted considerable attention.This review provides a comprehensive overview of the latest advancements in hydrogen storage technologies,with an emphasis on the synergistic application of high-throughput screening and machine learning in solid-state hydrogen storage materials.These approaches demonstrate exceptional potential in accurately predicting hydrogen storage properties,optimizing material performance,and accelerating the development of innovative hydrogen storage materials.Specifically,we discuss in detail the essential role of artificial intelligence in developing hydrogen storage materials such as metal hydrides,alloys,carbon materials,metal–organic frameworks,and zeolites.Moreover,underground hydrogen storage is further explored as a scalable renewable energy storage solution,particularly in terms of optimizing storage parameters and performance prediction.By systematically analyzing the limitations of existing hydrogen storage approaches and the transformative potential of artificial intelligence-driven methods,this review offers insights into the discovery and optimization of high-performance hydrogen storage materials,contributing to sustainable global energy development and technological innovation.展开更多
The development of sustainable electrode materials for energy storage systems has become very important and porous carbons derived from biomass have become an important candidate because of their tunable pore structur...The development of sustainable electrode materials for energy storage systems has become very important and porous carbons derived from biomass have become an important candidate because of their tunable pore structure,environmental friendliness,and cost-effectiveness.Recent advances in controlling the pore structure of these carbons and its relationship between to is energy storage performance are discussed,emphasizing the critical role of a balanced distribution of micropores,mesopores and macropores in determining electrochemical behavior.Particular attention is given to how the intrinsic components of biomass precursors(lignin,cellulose,and hemicellulose)influence pore formation during carbonization.Carbonization and activation strategies to precisely control the pore structure are introduced.Finally,key challenges in the industrial production of these carbons are outlined,and future research directions are proposed.These include the establishment of a database of biomass intrinsic structures and machine learning-assisted pore structure engineering,aimed at providing guidance for the design of high-performance carbon materials for next-generation energy storage devices.展开更多
Ciphertext data retrieval in cloud databases suffers from some critical limitations,such as inadequate security measures,disorganized key management practices,and insufficient retrieval access control capabilities.To ...Ciphertext data retrieval in cloud databases suffers from some critical limitations,such as inadequate security measures,disorganized key management practices,and insufficient retrieval access control capabilities.To address these problems,this paper proposes an enhanced Fully Homomorphic Encryption(FHE)algorithm based on an improved DGHV algorithm,coupled with an optimized ciphertext retrieval scheme.Our specific contributions are outlined as follows:First,we employ an authorization code to verify the user’s retrieval authority and perform hierarchical access control on cloud storage data.Second,a triple-key encryption mechanism,which separates the data encryption key,retrieval authorization key,and retrieval key,is designed.Different keys are provided to different entities to run corresponding system functions.The key separation architecture proves particularly advantageous in multi-verifier coexistence scenarios,environments involving untrusted third-party retrieval services.Finally,the enhanced DGHV-based retrieval mechanism extends conventional functionality by enabling multi-keyword queries with similarity-ranked results,thereby significantly improving both the functionality and usability of the FHE system.展开更多
BACKGROUND Traditional limitations of cold static storage(CSS)on ice at 4℃during lung transplantation have necessitated limiting cold ischemic time(CIT)to 4-6 hours.Ex vivo lung perfusion(EVLP)can extend this preserv...BACKGROUND Traditional limitations of cold static storage(CSS)on ice at 4℃during lung transplantation have necessitated limiting cold ischemic time(CIT)to 4-6 hours.Ex vivo lung perfusion(EVLP)can extend this preservation time through the suspension of CIT and normothermic perfusion.As we continue to further expand the donor pool in all aspects of lung transplantation,teams are frequently traveling further distances to procure organs.AIM To determine the effect of CSS or EVLP on donors with extended travel distance[>750 nautical miles(NM)]to recipient.METHODS Lung transplants,whose donor traveled greater than 750 NM,were identified from the United Network for Organ Sharing Database.Recipients were stratified into either:CSS or EVLP,based on preservation method.Groups were assessed with comparative statistics and survival was assessed by Kaplan-Meier methods.A 3:1 propensity match was then created,and same analysis was repeated.RESULTS Prior to matching,those in the EVLP group had significantly increased postoperative morbidity to include dialysis,ventilator use,acute rejection,and treated rejection in the first year(P<0.05 for all).However,there were no significant differences in midterm survival(P=0.18).Following matching,those in the EVLP group again had significantly increased post-operative morbidity to include dialysis,extracorporeal membrane oxygenation use,ventilator use,and treated rejection in the first year(P<0.05 for all).As before,there were no significant differences in midterm survival following matching(P=0.08).CONCLUSION While there was no significant difference in survival,EVLP patients had increased peri-operative morbidity.With the advent of changes in CSS with 10℃storage further analysis is necessary to evaluate the best methods for utilizing organs from increased distances.展开更多
This paper describes a new method for active learning in content-based image retrieval. The proposed method firstly uses support vector machine (SVM) classifiers to learn an initial query concept. Then the proposed ac...This paper describes a new method for active learning in content-based image retrieval. The proposed method firstly uses support vector machine (SVM) classifiers to learn an initial query concept. Then the proposed active learning scheme employs similarity measure to check the current version space and selects images with maximum expected information gain to solicit user's label. Finally, the learned query is refined based on the user's further feedback. With the combination of SVM classifier and similarity measure, the proposed method can alleviate model bias existing in each of them. Our experiments on several query concepts show that the proposed method can learn the user's query concept quickly and effectively only with several iterations.展开更多
基金supported by the National Natural Science Foundation of China under Grant U2442219Fengyun Satellite Application Pioneer Program(2023)Special Initiative on Numerical Weather Prediction(NWP)Applications,the Civil Aerospace Technology Pre-Research Project(D040405)the Joint Funds of the Zhejiang Provincial Natural Science Foundation of China under Grant No.LZJMZ23D050003。
文摘Accurate retrieval of atmospheric vertical profiles is critical for improving weather prediction and climate monitoring.However,the complexity of atmospheric processes in cloudy regions poses challenges compared to those of clear sky scenarios.This study presents a novel framework that integrates Bayesian optimization and machine learning approaches to retrieve atmospheric vertical profiles—including temperature,humidity,ozone concentration,cloud fraction,ice water content(IWC),and liquid water content(LWC)—from hyperspectral infrared observations.Specifically,a Bayesian method was used to refine ERA5 reanalysis data by minimizing brightness temperature(BT)discrepancies against FY-4B Geostationary Interferometric Infrared Sounder(GIIRS)observations,generating a high-quality profile database(~2.8 million profiles)across diverse weather systems.The optimized profiles improve radiative consistency,reducing BT biases from>40 K to<10 K in cloudy regions.To further overcome the limitations of the Bayesian method,we developed a Transformer-Resnet hybrid model(TERNet),which achieved superior performance with RMSE values of 1.61 K(temperature),5.77%(humidity),and 2.25×10^(–6)/6.09×10^(–6)kg kg^(–1)(IWC/LWC)across the entire vertical levels in all-sky conditions.The TERNet outperforms both ERA5 in cloud parameter retrieval and the GIIRS L2 product in thermodynamic profiling.Independent verification with radiosonde and Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observations(CALIPSO)datasets confirms the framework's reliability across various meteorological regimes.This work demonstrates the capability of combining physics-informed Bayesian methods with data-driven machine learning to fully exploit hyperspectral IR data.
基金extend their gratitude to the Deanship of Scientific Research,Vice Presidency for Graduate Studies and Scientific Research,King Faisal University,Saudi Arabia,for funding the publication of this work under the Ambitious Researcher program(Project No.KFU253806).
文摘Artificial intelligence(AI)based models have been used to predict the structural,optical,mechanical,and electrochemical properties of zinc oxide/graphene oxide nanocomposites.Machine learning(ML)models such as Artificial Neural Networks(ANN),Support Vector Regression(SVR),Multilayer Perceptron(MLP),and hybrid,along with fuzzy logic tools,were applied to predict the different properties like wavelength at maximum intensity(444 nm),crystallite size(17.50 nm),and optical bandgap(2.85 eV).While some other properties,such as energy density,power density,and charge transfer resistance,were also predicted with the help of datasets of 1000(80:20).In general,the energy parameters were predicted more accurately by hybrid models.The hydrothermal method was used to synthesize graphene oxide(GO)and zinc oxide(ZnO)nanocomposites.The increased surface area,conductivity,and stability of graphene oxide in zinc oxide nanoparticles make the composite an ideal option for energy storage.X-ray diffraction(XRD)confirmed the crystallite size of 17.41 nm for the nanocomposite and the presence of GO(12.8○)peaks.The scanning electron microscope(SEM)showed anchored wrinkled GO sheets on zinc oxide with an average particle size of 2.93μm.Energy-dispersive X-ray spectroscopy(EDX)confirmed the elemental composition,and Fouriertransform infrared spectroscopy(FTIR)revealed the impact of GO on functional groups and electrochemical behavior.Photoluminescence(PL)wavelength of(439 nm)and band gap of(2.81 eV)show that the material is suitable for energy applications in nanocomposites.Smart nanocomposite materials with improved performance in energy storage and related applications were fabricated by combining synthesis,characterization,fuzzy logic,and machine learning in this work.
基金supported by the fund from Tianjin Municipal Science and Technology Bureau(22JCYBJC01390).
文摘At present,the polymerase chain reaction(PCR)amplification-based file retrieval method is the mostcommonly used and effective means of DNA file retrieval.The number of orthogonal primers limitsthe number of files that can be accurately accessed,which in turn affects the density in a single oligo poolof digital DNA storage.In this paper,a multi-mode DNA sequence design method based on PCR file retrie-val in a single oligonucleotide pool is proposed for high-capacity DNA data storage.Firstly,by analyzingthe maximum number of orthogonal primers at each predicted primer length,it was found that the rela-tionship between primer length and the maximum available primer number does not increase linearly,and the maximum number of orthogonal primers is on the order of 10^(4).Next,this paper analyzes themaximum address space capacity of DNA sequences with different types of primer binding sites for filemapping.In the case where the capacity of the primer library is R(where R is even),the number ofaddress spaces that can be mapped by the single-primer DNA sequence design scheme proposed in thispaper is four times that of the previous one,and the two-level primer DNA sequence design scheme can reach [R/2·(R/2-1)]^(2)times.Finally,a multi-mode DNA sequence generation method is designed based onthe number of files to be stored in the oligonucleotide pool,in order to meet the requirements of the ran-dom retrieval of target files in an oligonucleotide pool with large-scale file numbers.The performance ofthe primers generated by the orthogonal primer library generator proposed in this paper is verified,andthe average Gibbs free energy of the most stable heterodimer formed between the orthogonal primersproduced is−1 kcal·(mol·L^(−1))^(−1)(1 kcal=4.184 kJ).At the same time,by selectively PCR-amplifying theDNA sequences of the two-level primer binding sites for random access,the target sequence can be accu-rately read with a minimum of 10^(3) reads,when the primer binding site sequences at different positionsare mutually different.This paper provides a pipeline for orthogonal primer library generation and multi-mode mapping schemes between files and primers,which can help achieve precise random access to filesin large-scale DNA oligo pools.
文摘This study investigates the prediction of groundwater Storage in the Rabat-Sale-Kenitra region under climate change conditions using advanced machine learning models.A comprehensive dataset encompassing hydrological,meteorological,and geological factors was meticulously curated and preprocessed for model training.Six regression models Decision Tree,Random Forest,LightGBM,CatBoost,Extreme Learning Machine(ELM),and Artificial Neural Network(ANN)were employed to predict groundwater Storage,with hyperparameters optimized via grid search.The performance of these models was rigorously evaluated using metrics such as Root Mean Squared Error(RMSE),Mean Absolute Error(MAE),and the coefficient of determination(R^(2)).Results demonstrated that the LightGBM model outperformed the others,achieving an impressive testing RMSE of 3.07 and an R^(2)of 0.9997,indicating its robustness in handling large datasets.The Extreme Learning Machine and ANN showed considerable limitations,highlighting the importance of model selection.This research underscores the critical role of advanced machine learning techniques in enhancing groundwater resource management,providing valuable insights for policymakers in developing sustainable strategies to address groundwater challenges in the face of climate variability.
基金supported by the National Natural Science Foundation of China(22005072,21965006)Guiyang Guian science and Technology Personnel Training Project(2024-2-13)+1 种基金Guizhou Provincial Key Technology R&D Program(Qian Ke He support(2023)General 122)Guizhou Provincial Science and Technology Foundation(ZD2025049,KXJZ2024029).
文摘High-entropy materials(HEMs)show exceptional mechanical properties,highly adjustable chemical characteristics,and outstanding stability,making them suitable for energy storage.However,the broad compositional space and intricate chemical interactions in HEMs present challenges to traditional trial-and-error research methods,restricting their efficacy in swift screening and synthesis.Hence,the application of machine learning(ML)to the realm of high-entropy materials and energy storage becomes imperative.ML demonstrates its formidable capabilities for navigating the complexity of HEMs,with their diverse metal components,structures and property combinations,to advance energy storage applications.This review comprises the following sections:a concise introduction to the general process of ML in the energy materials field,a summary of HEMs in the energy storage field,a review of the latest achievements of ML in the HEMs and energy storage field,and finally,an exploration of current challenges and prospects in this interdisciplinary arena.With the advent of ML,the precision of its predictions and the efficiency of its screening methods have offered novel perspectives for material research,expediting the discovery and application of new materials.This article contributes to the advancement of research in related fields,hastening the development of novel materials to meet the escalating energy demands and promote sustainable development goals.
文摘To achieve carbon dioxide(CO_(2))storage through enhanced oil recovery,accurate forecasting of CO_(2) subsurface storage and cumulative oil production is essential.This study develops hybrid predictive models for the determination of CO_(2) storage mass and cumulative oil production in unconventional reservoirs.It does so with two multi-layer perceptron neural networks(MLPNN)and a least-squares support vector machine(LSSVM),hybridized with grey wolf optimization(GWO)and/or particle swarm optimization(PSO).Large,simulated datasets were divided into training(70%)and testing(30%)groups,with normalization applied to both groups.Mahalanobis distance identifies/eliminates outliers in the training subset only.A non-dominated sorting genetic algorithm(NSGA-II)combined with LSSVM selected seven influential features from the nine available input parameters:reservoir depth,porosity,permeability,thickness,bottom-hole pressure,area,CO_(2) injection rate,residual oil saturation to gas flooding,and residual oil saturation to water flooding.Predictive models were developed and tested,with performance evaluated with an overfitting index(OFI),scoring analysis,and partial dependence plots(PDP),during training and independent testing to enhance model focus and effectiveness.The LSSVM-GWO model generated the lowest root mean square error(RMSE)values(0.4052 MMT for CO_(2) storage and 9.7392 MMbbl for cumulative oil production)in the training group.That trained model also exhibited excellent generalization and minimal overfitting when applied to the testing group(RMSE of 0.6224 MMT for CO_(2) storage and 12.5143 MMbbl for cumulative oil production).PDP analysis revealed that the input features“area”and“porosity”had the most influence on the LSSVM-GWO model's pre-diction performance.This paper presents a new hybrid modeling approach that achieves accurate forecasting of CO_(2) subsurface storage and cumulative oil production.It also establishes a new standard for such forecasting,which can lead to the development of more effective and sustainable solutions for oil recovery.
基金The National Key Technology R&D Program of China during the 11th Five-Year Plan Period(No.2006BAH02A06)
文摘In order to evaluate the efficiency of the automated storage/retrieval system(AS/RS)accurately,and compare different layouts of the AS/RS using mean travel time,under randomized storage conditions,an exact,geometry-based analytical model is presented.The model can be used to compute the expected single-command and dual-command travel time for a storage/retrieval(S/R)machine which can travel simultaneously horizontally and vertically as it moves along a storage aisle.The rack may be either square in time or non square in time.Additionally,the alternative layouts of the AS/RS and travel-time models are examined.Comparing with setting the I/O point at the left-lower corner of the rack,setting the I/O point at any point at the vertical edge can help enhance the efficiency of the AS/RS.
基金supported by the Natural Science Foundation of Hebei Province(E2023502006)Fundamental Research Fund for the Central Universities(2025MS131).
文摘Magnesium hydride(MgH_(2)),a promising high-capacity hydrogen storage material,is hindered by slow dehydrogenation kinetics.AIdriven catalyst discovery to address this is often hampered by the laborious extraction of data from unstructured literature.To overcome this,we introduce a transformative“LLM to Agent”framework that synergistically integrates Large Language Models(LLMs)for automated data curation with Machine Learning(ML)for predictive design.We automatically constructed a comprehensive database of 809 MgH_(2)catalysts(6555 data rows)with high fidelity and an~40-fold acceleration over manual methods.The resulting ML models achieved high accuracy(average R^(2)>0.91)in predicting dehydrogenation temperature and activation energy,subsequently guiding a Genetic Algorithm(GA)in an exploratory inverse design that autonomously uncovered key design principles for high-performance catalysts.Encouragingly,a strong alignment was found between these AI-discovered principles and the design strategies of recently reported,state-of-the-art experimental systems,providing substantial evidence for the validity of our approach.The framework culminates in Cat-Advisor,a novel,domain-adapted multi-agent system.Cat-Advisor translates ML predictions and retrieval-augmented knowledge into actionable design guidance,demonstrating capabilities that surpass those of general-purpose LLMs in this specialized domain.This work delivers a practical AI toolkit for accelerated materials discovery and advances the emerging Agent-based paradigm for designing next-generation energy technologies.
基金the National Basic Research Program (973) of China (No. 2004CB719401)the National Research Foundation for the Doctoral Program of Higher Education of China (No.20060003060)
文摘In this paper, we present a novel Support Vector Machine active learning algorithm for effective 3D model retrieval using the concept of relevance feedback. The proposed method learns from the most informative objects which are marked by the user, and then creates a boundary separating the relevant models from irrelevant ones. What it needs is only a small number of 3D models labelled by the user. It can grasp the user＇s semantic knowledge rapidly and accurately. Experimental results showed that the proposed algorithm significantly improves the retrieval effectiveness. Compared with four state-of-the-art query refinement schemes for 3D model retrieval, it provides superior retrieval performance after no more than two rounds of relevance feedback.
基金the National Natural Science Foundation of China under Grant，the Fundamental Research Funds for the Central Universities under Grant No.FRF-TP-14-046A2
文摘Due to its characteristics distribution and virtualization, cloud storage also brings new security problems. User＇s data is stored in the cloud, which separated the ownership from management. How to ensure the security of cloud data, how to increase data availability and how to improve user privacy perception are the key issues of cloud storage research, especially when the cloud service provider is not completely trusted. In this paper, a cloud storage ciphertext retrieval scheme based on AES and homomorphic encryption is presented. This ciphertext retrieval scheme will not only conceal the user retrieval information, but also prevent the cloud from obtaining user access pattern such as read-write mode, and access frequency, thereby ensuring the safety of the ciphertext retrieval and user privacy. The results of simulation analysis show that the performance of this ciphertext retrieval scheme requires less overhead than other schemes on the same security level.
文摘E-commerce, as an emerging marketing mode, has attracted more and more attention and gradually changed the way of our life. However, the existing layout of distribution centers can＇t fulfill the storage and picking demands of e-commerce sufficiently. In this paper, a modified miniload automated storage/retrieval system is designed to fit these new characteristics of e-commerce in logistics. Meanwhile, a matching problem, concerning with the improvement of picking efficiency in new system, is studied in this paper. The problem is how to reduce the travelling distance of totes between aisles and picking stations. A multi-stage heuristic algorithm is proposed based on statement and model of this problem. The main idea of this algorithm is, with some heuristic strategies based on similarity coefficients, minimizing the transportations of items which can not arrive in the destination picking stations just through direct conveyors. The experimental results based on the cases generated by computers show that the average reduced rate of indirect transport times can reach 14.36% with the application of multi-stage heuristic algorithm. For the cases from a real e-commerce distribution center, the order processing time can be reduced from 11.20 h to 10.06 h with the help of the modified system and the proposed algorithm. In summary, this research proposed a modified system and a multi-stage heuristic algorithm that can reduce the travelling distance of totes effectively and improve the whole performance of e-commerce distribution center.
文摘Support vector machines have met with significant success in the information retrieval field, especially in handling text classification tasks. Although various performance estimators for SVMs have been proposed, these only focus on accuracy which is based on the leave-one-out cross validation procedure. Information-retrieval-related performance measures are always neglected in a kernel learning methodology. In this paper, we have proposed a set of information-retrieval-oriented performance estimators for SVMs, which are based on the span bound of the leave-one-out procedure. Experiments have proven that our proposed estimators are both effective and stable.
基金supported in part by the National Natural Science Foundation of China under Grant 52007055in part by the Natural Science Foundation of Hunan Province of China under Grant 2021JJ40099。
文摘Permanent magnet homopolar inductor machine(PMHIM) has attracted much attention in the field of flywheel energy storage system(FESS) due to its merits of simple structure,brushless excitation, and rotor flywheel integration. However, the air-gap flux generated by the PM cannot be adjusted, which would cause large electromagnetic losses in the standby operation state of FESS. To solve this problem, a novel mechanically adjusted variable flux permanent magnet homopolar inductor machine with rotating magnetic poles(RMP-PMHIM) is proposed in this paper. The permanent magnet poles are rotated by an auxiliary rotating device and the purpose of changing the air-gap flux is achieved. First, the structure and operation principle of the proposed RMP-PMHIM are explained. Second,the flux weakening principle of the RMP-PMHIM is analyzed and the equivalent magnetic circuit models under different flux weakening states are built. Third, the parameters of the PM and its fixed structure are optimized to obtain the good electromagnetic performance. Fourth, the electromagnetic performance, including the air-gap flux density, back-EMF, flux weakening ability, loss, etc. of the proposed RMP-PMHIM are investigated and compared. Compared with the non-rotating state of the PM of RPM-PMHIM, the air-gap flux density amplitude can be weakened by 99.95% when the PM rotation angle is 90 degrees, and the no-load core loss can be suppressed by 99.98%,which shows that the proposed RPM-PMHIM is a good candidate for the application of FESS.
基金supported by the National Key Research and Development Program“Biological Information Security and Efficient Transmission”Project,Project Letter No.2017YFC1201204.
文摘At present,the traditional blockchain for data storage and retrieval reflects the characteristics of slow data uploading speed,high cost,and transparency,and there are a lot of corresponding problems,such as not supporting private data storage,large data operation costs,and not supporting Data field query.This paper proposes a method of data encryption storage and retrieval based on the IOTA distributed ledger,combined with the fast transaction processing speed and zero-value transactions of the IOTA blockchain,through the Masked Authenticated Messaging technology,so that the data is encrypted in the data stream.The form is stored in the distributed ledger,quickly retrieved through the field index mechanism established by the data form,and the data operation is carried out on the chain.Experimental results show that this system has high storage,encryption and retrieval performance,and good practicability.
文摘We propose a new approach to store and query XML data in an RDBMS basing on the idea of the numbering scheme and inverted list. O ur approach allows us to quickly determine the precedence, sibling and ancestor/ descendant relationships between any pair of nodes in the hierarchy of XML, and utilize path index to speed up calculating of path expressions. Examples have de monstrated that our approach can effectively and efficiently support both XQuery queries and keyword searches. Our approach is also flexible enough to support X ML documents both with Schema and without Schema, and applications both retrieva l and update. We also present the architecture of middleware for application acc essing XML documents stored in relations, and an algorithm translating a given X ML document into relations effectively.
基金the National Natural Science Foundation of China(No.22273096)the Young Scientists Fund of the National Natural Science Foundation of China(Grant No.22409139).
文摘In the global transition towards sustainable energy sources,hydrogen energy has emerged as an indispensable pillar in reshaping the energy landscape,owing to its environmental sustainability,zero emissions,and high efficiency.Nevertheless,the large-scale deployment of hydrogen energy is confronted with substantial technical barriers in storage and transportation.Although contemporary research has shifted focus to the development of highly efficient hydrogen storage materials,conventional material design concepts remain predominantly empirical,typically relying on trial-and-error methodologies.Importantly,the widespread application of artificial intelligence technologies in accelerating materials discovery and optimization has attracted considerable attention.This review provides a comprehensive overview of the latest advancements in hydrogen storage technologies,with an emphasis on the synergistic application of high-throughput screening and machine learning in solid-state hydrogen storage materials.These approaches demonstrate exceptional potential in accurately predicting hydrogen storage properties,optimizing material performance,and accelerating the development of innovative hydrogen storage materials.Specifically,we discuss in detail the essential role of artificial intelligence in developing hydrogen storage materials such as metal hydrides,alloys,carbon materials,metal–organic frameworks,and zeolites.Moreover,underground hydrogen storage is further explored as a scalable renewable energy storage solution,particularly in terms of optimizing storage parameters and performance prediction.By systematically analyzing the limitations of existing hydrogen storage approaches and the transformative potential of artificial intelligence-driven methods,this review offers insights into the discovery and optimization of high-performance hydrogen storage materials,contributing to sustainable global energy development and technological innovation.
文摘The development of sustainable electrode materials for energy storage systems has become very important and porous carbons derived from biomass have become an important candidate because of their tunable pore structure,environmental friendliness,and cost-effectiveness.Recent advances in controlling the pore structure of these carbons and its relationship between to is energy storage performance are discussed,emphasizing the critical role of a balanced distribution of micropores,mesopores and macropores in determining electrochemical behavior.Particular attention is given to how the intrinsic components of biomass precursors(lignin,cellulose,and hemicellulose)influence pore formation during carbonization.Carbonization and activation strategies to precisely control the pore structure are introduced.Finally,key challenges in the industrial production of these carbons are outlined,and future research directions are proposed.These include the establishment of a database of biomass intrinsic structures and machine learning-assisted pore structure engineering,aimed at providing guidance for the design of high-performance carbon materials for next-generation energy storage devices.
基金supported by the Innovation Program for Quantum Science and technology(2021ZD0301300)supported by the Fundamental Research Funds for the Central Universities(Nos.3282024046,3282024052,3282024058,3282023017).
文摘Ciphertext data retrieval in cloud databases suffers from some critical limitations,such as inadequate security measures,disorganized key management practices,and insufficient retrieval access control capabilities.To address these problems,this paper proposes an enhanced Fully Homomorphic Encryption(FHE)algorithm based on an improved DGHV algorithm,coupled with an optimized ciphertext retrieval scheme.Our specific contributions are outlined as follows:First,we employ an authorization code to verify the user’s retrieval authority and perform hierarchical access control on cloud storage data.Second,a triple-key encryption mechanism,which separates the data encryption key,retrieval authorization key,and retrieval key,is designed.Different keys are provided to different entities to run corresponding system functions.The key separation architecture proves particularly advantageous in multi-verifier coexistence scenarios,environments involving untrusted third-party retrieval services.Finally,the enhanced DGHV-based retrieval mechanism extends conventional functionality by enabling multi-keyword queries with similarity-ranked results,thereby significantly improving both the functionality and usability of the FHE system.
基金Supported by The Jewel and Frank Benson Family Endowmentand The Jewel and Frank Benson Research Professorship.
文摘BACKGROUND Traditional limitations of cold static storage(CSS)on ice at 4℃during lung transplantation have necessitated limiting cold ischemic time(CIT)to 4-6 hours.Ex vivo lung perfusion(EVLP)can extend this preservation time through the suspension of CIT and normothermic perfusion.As we continue to further expand the donor pool in all aspects of lung transplantation,teams are frequently traveling further distances to procure organs.AIM To determine the effect of CSS or EVLP on donors with extended travel distance[>750 nautical miles(NM)]to recipient.METHODS Lung transplants,whose donor traveled greater than 750 NM,were identified from the United Network for Organ Sharing Database.Recipients were stratified into either:CSS or EVLP,based on preservation method.Groups were assessed with comparative statistics and survival was assessed by Kaplan-Meier methods.A 3:1 propensity match was then created,and same analysis was repeated.RESULTS Prior to matching,those in the EVLP group had significantly increased postoperative morbidity to include dialysis,ventilator use,acute rejection,and treated rejection in the first year(P<0.05 for all).However,there were no significant differences in midterm survival(P=0.18).Following matching,those in the EVLP group again had significantly increased post-operative morbidity to include dialysis,extracorporeal membrane oxygenation use,ventilator use,and treated rejection in the first year(P<0.05 for all).As before,there were no significant differences in midterm survival following matching(P=0.08).CONCLUSION While there was no significant difference in survival,EVLP patients had increased peri-operative morbidity.With the advent of changes in CSS with 10℃storage further analysis is necessary to evaluate the best methods for utilizing organs from increased distances.
文摘This paper describes a new method for active learning in content-based image retrieval. The proposed method firstly uses support vector machine (SVM) classifiers to learn an initial query concept. Then the proposed active learning scheme employs similarity measure to check the current version space and selects images with maximum expected information gain to solicit user's label. Finally, the learned query is refined based on the user's further feedback. With the combination of SVM classifier and similarity measure, the proposed method can alleviate model bias existing in each of them. Our experiments on several query concepts show that the proposed method can learn the user's query concept quickly and effectively only with several iterations.
