Magnesium(Mg)alloys have attracted considerable attention as promising implant materials for biodegradable medical devices.In this study,we focused on investigating the effect of macroscopic environmental heterogeneit...Magnesium(Mg)alloys have attracted considerable attention as promising implant materials for biodegradable medical devices.In this study,we focused on investigating the effect of macroscopic environmental heterogeneity due to the degradation of Mg on its corrosion behavior.The immersion experiments using pure Mg plates,which were placed vertically in a culture medium(Dulbecco’s Modified Eagle’s Medium(DEME)+10%fetal bovine serum(FBS))for 1,5,and 10 days,were conducted.Surface analyses for the corrosion product layers and the measurements of the pH values and concentrations of eluted ions in the immersion medium around the upper and lower areas of the Mg plate were performed.The significant effect of the macroscopic environmental heterogeneity derived from Mg degradation on the corrosion behavior was demonstrated by in vitro tests.Additionally,the in vivo tests were carried out by implanting the pure Mg plates in the femur of rabbits.The in vivo results exhibited macroscopically heterogeneous Mg degradation,with areas of more severe corrosion compared to the in vitro test;it is especially noticeable during the early stage of degradation,even though the average corrosion rate was lower.展开更多
Carbon nanotubes(CNTs)hold immense promise for a wide array of applications due to their exceptional physical and chemical properties.Understanding and controlling their structural characteristics,particu-larly the di...Carbon nanotubes(CNTs)hold immense promise for a wide array of applications due to their exceptional physical and chemical properties.Understanding and controlling their structural characteristics,particu-larly the diameter and number of walls,is crucial for harnessing their full potential.We investigated the relationship between these parameters for both commercially available and lab-scale CNTs,spanning a wide range of outer diameters(1-13 nm)and numbers of walls(1-13).Our findings revealed a com-monality among the structural diversity,rather than a random distribution,as evidenced by a piecewise linear relationship between the outer diameter and number of walls,with an inflection point occurring at approximately 4 nm in diameter.This observation is unexpected,as the CNTs were synthesized using different approaches and growth conditions;yet,as a group,they exhibit a“structural scaling”.Addi-tionally,we made an intriguing observation:despite increases in outer diameter and number of walls,the inner diameters remained relatively constant(4-5 nm)for thicker CNTs with more than three walls.These results suggest that structural properties can be estimated based on diameter,which not only ad-vances our fundamental understanding of CNT synthesis but also provides practical insights for tailoring CNT properties for various applications.展开更多
The biotreatment of mine drainage containing dissolved manganese(Mn)using Mn(II)-oxidizing bacteria is challenging.Sequencing-batch(SBRs)and continuous-flow reactors(CFRs)packed with limestones and inoculated with the...The biotreatment of mine drainage containing dissolved manganese(Mn)using Mn(II)-oxidizing bacteria is challenging.Sequencing-batch(SBRs)and continuous-flow reactors(CFRs)packed with limestones and inoculated with the mine-drainage microbial communitywere compared to determine the removal efficiency of Mn(II)from mine drainage.Mn(II)removal in CFRs was 11.4%±0.0%(mean±standard deviation)in the first two weeks and;it slightly increased to 13.6%±0.0%after four weeks,and more than 94%of Mn(II)was removed under the steady-state treatment phase.The performance of SBRs was more effective,wherein 24.4%±0.1%of Mn was removed in the first two weeks,and in four weeks,surpassed 66.6%±0.2%.Rapid Mn(II)removal observed in the start-up of SBR resulted from higher microbial metabolic activities.The adenosine triphosphate(ATP)content of the microbial community was four-fold more than in CFR,but comparable during the steady-state phase.The Mn-oxide deposits occurring in the SBR and CFR at steady-state were mixed phases of birnessite and woodruffite,and the average Mn oxidation valence in the SBR(+3.73)was slightly higher than that in the CFR(+3.54).During the start-up treatment,the closest relatives of Methyloversatilis,Methylibium,and Curvibacter dominated the SBR,whereas putative Mn oxidizers were associated with Hyphomicrobium,Pedobacter,Pedomicrobium,Terricaulis sp.,Sulfuritalea,and Terrimonas organisms.The growth of potential Mnoxidizing genera,including Mesorhizobium,Rhodococcus,Hydrogenophaga,Terricaulis sp.,and‘Candidatus Manganitrophus-noduliformans’was observed under the steady state.The SBR operation was effective as a prior start-up treatment for mine drainage-containing Mn(II),through which the CFR performed well as posterior bio-treatment.展开更多
This study presents the use of an innovative population-based algorithm called the Sine Cosine Algorithm and its metaheuristic form,Quasi Oppositional Sine Cosine Algorithm,to automatic generation control of a multipl...This study presents the use of an innovative population-based algorithm called the Sine Cosine Algorithm and its metaheuristic form,Quasi Oppositional Sine Cosine Algorithm,to automatic generation control of a multiple-source-based interconnected power system that consists of thermal,gas,and hydro power plants.The Proportional-Integral-Derivative controller,which is utilized for automated generation control in an interconnected hybrid power systemwith aDClink connecting two regions,has been tuned using the proposed optimization technique.An Electric Vehicle is taken into consideration only as an electrical load.The Quasi Oppositional Sine Cosinemethod’s performance and efficacy have been compared to the Sine Cosine Algorithm and optimal output feedback controller tuning performance.Applying the QOSCA optimization technique,which has only been shown in this study in the context of an LFC research thus far,makes this paper unique.The main objective has been used to assess and compare the dynamic performances of the recommended controller along with QOSCA optimisation technic.The resilience of the controller is examined using two different system parameters:B(frequency bias parameter)and R(governor speed regulation).The sensitivity analysis results demonstrate the high reliability of the QOSCA algorithm-based controller.Once optimal controller gains are established for nominal conditions,step load perturbations up to±10%&±25%in the nominal values of the systemparameters and operational load condition do not require adjustment of the controller.Ultimately,a scenario is examined whereby EVs are used for area 1,and a single PID controller is used rather than three.展开更多
The integration of renewable energy sources into electricity markets presents significant challenges due to the inherent variability and uncertainty of power generation from wind,solar,and other renewables.Accurate fo...The integration of renewable energy sources into electricity markets presents significant challenges due to the inherent variability and uncertainty of power generation from wind,solar,and other renewables.Accurate forecasting is crucial for ensuring grid stability,optimizing market operations,and minimizing economic risks.This paper introduces a hybrid forecasting framework incorporating fractional-order statistical models,fractal-based feature enginering,and deep learning architectures to improve renewable energy forecasting accuracy.Fractional autoregressive integrated moving average(FARIMA)and fractional exponential smoothing(FETS)models are explored for capturing long-memory dependencies in energy time-series data.Additionally,multifractal detrended fluctuation analysis(MFDFA)is used to analyze the intermittency of renewable energy generation.The hybrid approach further integrates wavelet transforms and convolutional long short-term memory(CNN-LSTM)networks to model shortand long-term dependencies effectively.Experimental results demonstrate that fractional and fractal-based hybrid forecasting techniques significantly outperform traditional models in terms of accuracy,reliability,and adaptability to energy market dynamics.This research provides insights for market participants,policymakers,and grid operators to develop more robust forecasting frameworks,ensuring a more sustainable and resilient electricity market.展开更多
Neural architecture search(NAS)optimizes neural network architectures to align with specific data and objectives,thereby enabling the design of high-performance models without specialized expertise.However,a significa...Neural architecture search(NAS)optimizes neural network architectures to align with specific data and objectives,thereby enabling the design of high-performance models without specialized expertise.However,a significant limitation of NAS is that it requires extensive computational resources and time.Consequently,performing a comprehensive architectural search for each new dataset is inefficient.Given the continuous expansion of available datasets,there is an urgent need to predict the optimal architecture for the previously unknown datasets.This study proposes a novel framework that generates architectures tailored to unknown datasets by mapping architectures that have demonstrated effectiveness on the existing dataset into a latent feature space.As NAS is inherently represented as graph structures,we employed an encoder-decoder transformation model based on variational graph auto-encoders to perform this latent feature mapping.The encoder-decoder transformation model demonstrates strong capability in extracting features from graph structures,making it particularly well-suited for mapping NAS architectures.By training variational graph auto-encoders on existing high-quality architectures,the proposed method constructs a latent space and facilitates the design of optimal architectures for diverse datasets.Furthermore,to effectively define similarity amongarchitectures,wepropose constructing the latent spaceby incorporatingbothdataset andtaskfeatures.Experimental results indicate that our approach significantly enhances search efficiency and outperforms conventional methods in terms of model performance.展开更多
Quasi-one-dimensional(quasi-1D)van der Waals(vdWs)materials,such as ZrTe_(5),exhibit unique elec-trical properties and quantum phenomena,making them attractive for advanced electronic applications.However,large-scale ...Quasi-one-dimensional(quasi-1D)van der Waals(vdWs)materials,such as ZrTe_(5),exhibit unique elec-trical properties and quantum phenomena,making them attractive for advanced electronic applications.However,large-scale growth of ZrTe_(5) thin films presents challenges.We address this by employing sput-tering,a common semiconductor industry technique.The as-deposited ZrTe_(5) film is amorphous,and post-annealing induces a crystallization process akin to transition-metal dichalcogenides.Our study in-vestigates the electrical and optical properties during this amorphous-to-crystalline transition,reveal-ing insights into the underlying mechanism.This work contributes to the fundamental understanding of quasi-1D materials and introduces a scalable fabrication method for ZrTe_(5) which offers the possibility of fabricating unique future electronic and optical devices.展开更多
Exploring earth-abundant,highly active bifunctional electrocatalysts for efficient hydrogen and oxygen evolution is crucial for water splitting.However,due to their distinct free energies and conducting behaviors(elec...Exploring earth-abundant,highly active bifunctional electrocatalysts for efficient hydrogen and oxygen evolution is crucial for water splitting.However,due to their distinct free energies and conducting behaviors(electron/hole),balancing the catalytic efficiency between hydrogen and oxygen evolution remains challenging for achieving bifunctional electrocatalysts.Here,we report a locally-doped MoS_(2)monolayer with an in-plane heterostructure acting as a bifunctional electrocatalyst and apply it to the overall water splitting.In this heterostructure,the core region contains Mo/S vacancies,while the ring region was doped by Fe atoms(in two substitution configurations:1FeMo and 3FeMo-VS clusters)with a p-type conductive characteristic.Our micro-cell measurements,combined with density functional theory(DFT)calculations,reveal that the vacancies-rich core region presents remarkable hydrogen evolution reaction(HER)activity while the Fe-doped ring gives an excellent oxygen evolution reaction(OER)activity,thus forming an in-plane bifunctional electrocatalyst.Finally,as a proof-of-concept for overall water splitting,we constructed a full-cell configuration based on a locally-doped MoS_(2)monolayer,which achieved a cell voltage of 1.87 V at 10 mA·cm^(-2),demonstrating outstanding performance in strong acid electrolytes.Our work provides insight into the hetero-integration of bifunctional electrocatalysts at the atomic level,paving the way for designing transition metal dichalcogenide catalysts with activity-manipulated regions capable of multiple reactions.展开更多
Background:Accurate classification of normal blood cells is a critical foundation for automated hematological analysis,including the detection of pathological conditions like leukemia.While convolutional neural networ...Background:Accurate classification of normal blood cells is a critical foundation for automated hematological analysis,including the detection of pathological conditions like leukemia.While convolutional neural networks(CNNs)excel in local feature extraction,their ability to capture global contextual relationships in complex cellular morphologies is limited.This study introduces a hybrid CNN-Transformer framework to enhance normal blood cell classification,laying the groundwork for future leukemia diagnostics.Methods:The proposed architecture integrates pre-trained CNNs(ResNet50,EfficientNetB3,InceptionV3,CustomCNN)with Vision Transformer(ViT)layers to combine local and global feature modeling.Four hybrid models were evaluated on the publicly available Blood Cell Images dataset from Kaggle,comprising 17,092 annotated normal blood cell images across eight classes.The models were trained using transfer learning,fine-tuning,and computational optimizations,including cross-model parameter sharing to reduce redundancy by reusing weights across CNN backbones and attention-guided layer pruning to eliminate low-contribution layers based on attention scores,improving efficiency without sacrificing accuracy.Results:The InceptionV3-ViT model achieved a weighted accuracy of 97.66%(accounting for class imbalance by weighting each class’s contribution),a macro F1-score of 0.98,and a ROC-AUC of 0.998.The framework excelled in distinguishing morphologically similar cell types demonstrating robustness and reliable calibration(ECE of 0.019).The framework addresses generalization challenges,including class imbalance and morphological similarities,ensuring robust performance across diverse cell types.Conclusion:The hybrid CNN-Transformer framework significantly improves normal blood cell classification by capturing multi-scale features and long-range dependencies.Its high accuracy,efficiency,and generalization position it as a strong baseline for automated hematological analysis,with potential for extension to leukemia subtype classification through future validation on pathological samples.展开更多
The Dahurian larch forest in northeast China is important due to its vastness and location within a transitional zone from boreal to temperate and at the southern distribution edge of the vast Siberian larch forest. T...The Dahurian larch forest in northeast China is important due to its vastness and location within a transitional zone from boreal to temperate and at the southern distribution edge of the vast Siberian larch forest. The continuous carbon fluxes were measured from May 2004 to April 2005 in the Dahurian larch forest in Northeast China using an eddy covariance method. The results showed that the ecosystem released carbon in the dormant season from mid-October 2004 to April 2005, while it assimilated CO2 from the atmosphere in the growing season from May to September 2004. The net carbon sequestration reached its peak of 112 g.m^-2.month ^-1 in June 2004 （simplified expression of g （carbon）.m^-2.month^-1） and then gradually decreased. Annually, the larch forest was a carbon sink that sequestered carbon of 146 g-m^-2.a^-1 （simplified expression of g （carbon）.m^-2.a^-1） during the measurements. The photosynthetic process of the larch forest ecosystem was largely affected by the vapor pressure deficit （VPD） and temperature. Under humid conditions （VPD 〈 1.0 kPa）, the gross ecosystem production （GEP） increased with increasing temperature. But the net ecosystem production （NEP） showed almost no change with increasing temperature because the increment of GEP was counterbalanced by that of the ecosystem respiration. Under a dry environment （VPD 〉 1.0 kPa）, the GEP decreased with the increasing VPD at a rate of 3.0 μmol.m^-2.s^-1kPa -1 and the ecosystem respiration was also enhanced simultaneously due to the increase of air temperature, which was linearly correlated with the VPD. As a result, the net ecosystem carbon sequestration rapidly decreased with the increasing VPD at a rate of 5.2 μmol.m^-2.s-1.kPa^-1. Under humid conditions （VPD 〈 1.0 kPa）, both the GEP and NEP were obviously restricted by the low air temperature but were insensitive to the high temperature because the observed high temperature value comes within the category of the optimum range.展开更多
基金supported by JSPS KAKENHI Grant Number 22K12903.
文摘Magnesium(Mg)alloys have attracted considerable attention as promising implant materials for biodegradable medical devices.In this study,we focused on investigating the effect of macroscopic environmental heterogeneity due to the degradation of Mg on its corrosion behavior.The immersion experiments using pure Mg plates,which were placed vertically in a culture medium(Dulbecco’s Modified Eagle’s Medium(DEME)+10%fetal bovine serum(FBS))for 1,5,and 10 days,were conducted.Surface analyses for the corrosion product layers and the measurements of the pH values and concentrations of eluted ions in the immersion medium around the upper and lower areas of the Mg plate were performed.The significant effect of the macroscopic environmental heterogeneity derived from Mg degradation on the corrosion behavior was demonstrated by in vitro tests.Additionally,the in vivo tests were carried out by implanting the pure Mg plates in the femur of rabbits.The in vivo results exhibited macroscopically heterogeneous Mg degradation,with areas of more severe corrosion compared to the in vitro test;it is especially noticeable during the early stage of degradation,even though the average corrosion rate was lower.
文摘Carbon nanotubes(CNTs)hold immense promise for a wide array of applications due to their exceptional physical and chemical properties.Understanding and controlling their structural characteristics,particu-larly the diameter and number of walls,is crucial for harnessing their full potential.We investigated the relationship between these parameters for both commercially available and lab-scale CNTs,spanning a wide range of outer diameters(1-13 nm)and numbers of walls(1-13).Our findings revealed a com-monality among the structural diversity,rather than a random distribution,as evidenced by a piecewise linear relationship between the outer diameter and number of walls,with an inflection point occurring at approximately 4 nm in diameter.This observation is unexpected,as the CNTs were synthesized using different approaches and growth conditions;yet,as a group,they exhibit a“structural scaling”.Addi-tionally,we made an intriguing observation:despite increases in outer diameter and number of walls,the inner diameters remained relatively constant(4-5 nm)for thicker CNTs with more than three walls.These results suggest that structural properties can be estimated based on diameter,which not only ad-vances our fundamental understanding of CNT synthesis but also provides practical insights for tailoring CNT properties for various applications.
基金funded by the JOGMEC Research Grant and JSPS KAKENHI(No.JP21H03636).
文摘The biotreatment of mine drainage containing dissolved manganese(Mn)using Mn(II)-oxidizing bacteria is challenging.Sequencing-batch(SBRs)and continuous-flow reactors(CFRs)packed with limestones and inoculated with the mine-drainage microbial communitywere compared to determine the removal efficiency of Mn(II)from mine drainage.Mn(II)removal in CFRs was 11.4%±0.0%(mean±standard deviation)in the first two weeks and;it slightly increased to 13.6%±0.0%after four weeks,and more than 94%of Mn(II)was removed under the steady-state treatment phase.The performance of SBRs was more effective,wherein 24.4%±0.1%of Mn was removed in the first two weeks,and in four weeks,surpassed 66.6%±0.2%.Rapid Mn(II)removal observed in the start-up of SBR resulted from higher microbial metabolic activities.The adenosine triphosphate(ATP)content of the microbial community was four-fold more than in CFR,but comparable during the steady-state phase.The Mn-oxide deposits occurring in the SBR and CFR at steady-state were mixed phases of birnessite and woodruffite,and the average Mn oxidation valence in the SBR(+3.73)was slightly higher than that in the CFR(+3.54).During the start-up treatment,the closest relatives of Methyloversatilis,Methylibium,and Curvibacter dominated the SBR,whereas putative Mn oxidizers were associated with Hyphomicrobium,Pedobacter,Pedomicrobium,Terricaulis sp.,Sulfuritalea,and Terrimonas organisms.The growth of potential Mnoxidizing genera,including Mesorhizobium,Rhodococcus,Hydrogenophaga,Terricaulis sp.,and‘Candidatus Manganitrophus-noduliformans’was observed under the steady state.The SBR operation was effective as a prior start-up treatment for mine drainage-containing Mn(II),through which the CFR performed well as posterior bio-treatment.
文摘This study presents the use of an innovative population-based algorithm called the Sine Cosine Algorithm and its metaheuristic form,Quasi Oppositional Sine Cosine Algorithm,to automatic generation control of a multiple-source-based interconnected power system that consists of thermal,gas,and hydro power plants.The Proportional-Integral-Derivative controller,which is utilized for automated generation control in an interconnected hybrid power systemwith aDClink connecting two regions,has been tuned using the proposed optimization technique.An Electric Vehicle is taken into consideration only as an electrical load.The Quasi Oppositional Sine Cosinemethod’s performance and efficacy have been compared to the Sine Cosine Algorithm and optimal output feedback controller tuning performance.Applying the QOSCA optimization technique,which has only been shown in this study in the context of an LFC research thus far,makes this paper unique.The main objective has been used to assess and compare the dynamic performances of the recommended controller along with QOSCA optimisation technic.The resilience of the controller is examined using two different system parameters:B(frequency bias parameter)and R(governor speed regulation).The sensitivity analysis results demonstrate the high reliability of the QOSCA algorithm-based controller.Once optimal controller gains are established for nominal conditions,step load perturbations up to±10%&±25%in the nominal values of the systemparameters and operational load condition do not require adjustment of the controller.Ultimately,a scenario is examined whereby EVs are used for area 1,and a single PID controller is used rather than three.
基金funded under research grant from the Research,Development,andInnovation Authority(RDIA),Saudi Arabia,grant No.13010-Tabuk-2023-UT-R-3-1-SE.
文摘The integration of renewable energy sources into electricity markets presents significant challenges due to the inherent variability and uncertainty of power generation from wind,solar,and other renewables.Accurate forecasting is crucial for ensuring grid stability,optimizing market operations,and minimizing economic risks.This paper introduces a hybrid forecasting framework incorporating fractional-order statistical models,fractal-based feature enginering,and deep learning architectures to improve renewable energy forecasting accuracy.Fractional autoregressive integrated moving average(FARIMA)and fractional exponential smoothing(FETS)models are explored for capturing long-memory dependencies in energy time-series data.Additionally,multifractal detrended fluctuation analysis(MFDFA)is used to analyze the intermittency of renewable energy generation.The hybrid approach further integrates wavelet transforms and convolutional long short-term memory(CNN-LSTM)networks to model shortand long-term dependencies effectively.Experimental results demonstrate that fractional and fractal-based hybrid forecasting techniques significantly outperform traditional models in terms of accuracy,reliability,and adaptability to energy market dynamics.This research provides insights for market participants,policymakers,and grid operators to develop more robust forecasting frameworks,ensuring a more sustainable and resilient electricity market.
基金funded by the New Energy and Industrial Technology Development Organization(NEDO),grant number JPNP18002.
文摘Neural architecture search(NAS)optimizes neural network architectures to align with specific data and objectives,thereby enabling the design of high-performance models without specialized expertise.However,a significant limitation of NAS is that it requires extensive computational resources and time.Consequently,performing a comprehensive architectural search for each new dataset is inefficient.Given the continuous expansion of available datasets,there is an urgent need to predict the optimal architecture for the previously unknown datasets.This study proposes a novel framework that generates architectures tailored to unknown datasets by mapping architectures that have demonstrated effectiveness on the existing dataset into a latent feature space.As NAS is inherently represented as graph structures,we employed an encoder-decoder transformation model based on variational graph auto-encoders to perform this latent feature mapping.The encoder-decoder transformation model demonstrates strong capability in extracting features from graph structures,making it particularly well-suited for mapping NAS architectures.By training variational graph auto-encoders on existing high-quality architectures,the proposed method constructs a latent space and facilitates the design of optimal architectures for diverse datasets.Furthermore,to effectively define similarity amongarchitectures,wepropose constructing the latent spaceby incorporatingbothdataset andtaskfeatures.Experimental results indicate that our approach significantly enhances search efficiency and outperforms conventional methods in terms of model performance.
基金supported by the JSPS KAKENHI(Grant Nos.21H05009,22K20474,and 24K00915)the Murata Science Foundation+1 种基金supported by the Commissioned Research(No.JPJ012368C03701)of the National Institute of Information and Communications Technology(NICT),Japansupport from the Hirose Foundation and Iketani Science and Technology Foundation.
文摘Quasi-one-dimensional(quasi-1D)van der Waals(vdWs)materials,such as ZrTe_(5),exhibit unique elec-trical properties and quantum phenomena,making them attractive for advanced electronic applications.However,large-scale growth of ZrTe_(5) thin films presents challenges.We address this by employing sput-tering,a common semiconductor industry technique.The as-deposited ZrTe_(5) film is amorphous,and post-annealing induces a crystallization process akin to transition-metal dichalcogenides.Our study in-vestigates the electrical and optical properties during this amorphous-to-crystalline transition,reveal-ing insights into the underlying mechanism.This work contributes to the fundamental understanding of quasi-1D materials and introduces a scalable fabrication method for ZrTe_(5) which offers the possibility of fabricating unique future electronic and optical devices.
基金supported by the National Natural Science Foundation of China(Nos.22175060 and 22376062)JSPS Grant-in-Aid for Scientific Research(Nos.JP21H05235,JP22H05478 and JP22F22358)+1 种基金China Postdoctoral Science Foundation(No.2022M722867)the Key Research Project of Higher Education Institutions in Henan Province(No.23A530001).
文摘Exploring earth-abundant,highly active bifunctional electrocatalysts for efficient hydrogen and oxygen evolution is crucial for water splitting.However,due to their distinct free energies and conducting behaviors(electron/hole),balancing the catalytic efficiency between hydrogen and oxygen evolution remains challenging for achieving bifunctional electrocatalysts.Here,we report a locally-doped MoS_(2)monolayer with an in-plane heterostructure acting as a bifunctional electrocatalyst and apply it to the overall water splitting.In this heterostructure,the core region contains Mo/S vacancies,while the ring region was doped by Fe atoms(in two substitution configurations:1FeMo and 3FeMo-VS clusters)with a p-type conductive characteristic.Our micro-cell measurements,combined with density functional theory(DFT)calculations,reveal that the vacancies-rich core region presents remarkable hydrogen evolution reaction(HER)activity while the Fe-doped ring gives an excellent oxygen evolution reaction(OER)activity,thus forming an in-plane bifunctional electrocatalyst.Finally,as a proof-of-concept for overall water splitting,we constructed a full-cell configuration based on a locally-doped MoS_(2)monolayer,which achieved a cell voltage of 1.87 V at 10 mA·cm^(-2),demonstrating outstanding performance in strong acid electrolytes.Our work provides insight into the hetero-integration of bifunctional electrocatalysts at the atomic level,paving the way for designing transition metal dichalcogenide catalysts with activity-manipulated regions capable of multiple reactions.
基金the Deanship of Graduate Studies and Scientific Research at Najran University,Saudi Arabia,for their financial support through the Easy Track Research program,grant code(NU/EFP/MRC/13).
文摘Background:Accurate classification of normal blood cells is a critical foundation for automated hematological analysis,including the detection of pathological conditions like leukemia.While convolutional neural networks(CNNs)excel in local feature extraction,their ability to capture global contextual relationships in complex cellular morphologies is limited.This study introduces a hybrid CNN-Transformer framework to enhance normal blood cell classification,laying the groundwork for future leukemia diagnostics.Methods:The proposed architecture integrates pre-trained CNNs(ResNet50,EfficientNetB3,InceptionV3,CustomCNN)with Vision Transformer(ViT)layers to combine local and global feature modeling.Four hybrid models were evaluated on the publicly available Blood Cell Images dataset from Kaggle,comprising 17,092 annotated normal blood cell images across eight classes.The models were trained using transfer learning,fine-tuning,and computational optimizations,including cross-model parameter sharing to reduce redundancy by reusing weights across CNN backbones and attention-guided layer pruning to eliminate low-contribution layers based on attention scores,improving efficiency without sacrificing accuracy.Results:The InceptionV3-ViT model achieved a weighted accuracy of 97.66%(accounting for class imbalance by weighting each class’s contribution),a macro F1-score of 0.98,and a ROC-AUC of 0.998.The framework excelled in distinguishing morphologically similar cell types demonstrating robustness and reliable calibration(ECE of 0.019).The framework addresses generalization challenges,including class imbalance and morphological similarities,ensuring robust performance across diverse cell types.Conclusion:The hybrid CNN-Transformer framework significantly improves normal blood cell classification by capturing multi-scale features and long-range dependencies.Its high accuracy,efficiency,and generalization position it as a strong baseline for automated hematological analysis,with potential for extension to leukemia subtype classification through future validation on pathological samples.
基金the Global Environment Research Fund,Ministry of the Environment,Japan (S-1: Integrated Study for Terrestrial Carbon Management of Asia in the 21st Century Based on Scientific Advancements)the Chinese Academy of Sciences (07W70000SZ)+1 种基金the National Natural Science Foundation of China (30300271)the State Key Basic Research and Development Plan of China (2004CCA02700)
文摘The Dahurian larch forest in northeast China is important due to its vastness and location within a transitional zone from boreal to temperate and at the southern distribution edge of the vast Siberian larch forest. The continuous carbon fluxes were measured from May 2004 to April 2005 in the Dahurian larch forest in Northeast China using an eddy covariance method. The results showed that the ecosystem released carbon in the dormant season from mid-October 2004 to April 2005, while it assimilated CO2 from the atmosphere in the growing season from May to September 2004. The net carbon sequestration reached its peak of 112 g.m^-2.month ^-1 in June 2004 （simplified expression of g （carbon）.m^-2.month^-1） and then gradually decreased. Annually, the larch forest was a carbon sink that sequestered carbon of 146 g-m^-2.a^-1 （simplified expression of g （carbon）.m^-2.a^-1） during the measurements. The photosynthetic process of the larch forest ecosystem was largely affected by the vapor pressure deficit （VPD） and temperature. Under humid conditions （VPD 〈 1.0 kPa）, the gross ecosystem production （GEP） increased with increasing temperature. But the net ecosystem production （NEP） showed almost no change with increasing temperature because the increment of GEP was counterbalanced by that of the ecosystem respiration. Under a dry environment （VPD 〉 1.0 kPa）, the GEP decreased with the increasing VPD at a rate of 3.0 μmol.m^-2.s^-1kPa -1 and the ecosystem respiration was also enhanced simultaneously due to the increase of air temperature, which was linearly correlated with the VPD. As a result, the net ecosystem carbon sequestration rapidly decreased with the increasing VPD at a rate of 5.2 μmol.m^-2.s-1.kPa^-1. Under humid conditions （VPD 〈 1.0 kPa）, both the GEP and NEP were obviously restricted by the low air temperature but were insensitive to the high temperature because the observed high temperature value comes within the category of the optimum range.
