检索结果-南通市图书馆

Robust facial expression recognition via lightweight reinforcement learning for rehabilitation robotics

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Optoelectronics Letters 2025年第2期21卷 97-104页

作者： CHEN Yifan FAN Weiming GAO Hongwei yu jiahui JU Zhaojie School of Computing University of Portsmouth School of Automation and Electrical Engineering Shenyang Ligong University Department of Biomedical Engineering Zhejiang University

This paper proposes a lightweight reinforcement network (LRN) and auxiliary label distribution learning (ALDL)based robust facial expression recognition (FER) *** designed representation reinforcement (RR) network mainly comprises two modules,i.e.,the RR module and the auxiliary label space construction (ALSC) *** RR module highlights key feature messaging nodes in feature maps,and ALSC allows multiple labels with different intensities to be linked to one ***,LRN has a more robust feature extraction capability when model parameters are greatly reduced,and ALDL is proposed to contribute to the training effect of LRN in the condition of ambiguous training *** tested our method on FER-Plus and RAF-DB datasets,and the experiment demonstrates the feasibility of our method in practice during rehabilitation robots.

关键词： Reinforcement learning

Theoretical Predition of Two-dimensional SiGeP_(2)by the Global Optimization Method

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Journal of Wuhan University of Technology(Materials Science) 2023年第5期38卷 1010-1016页

作者： XUE Xiao yu jiahui ZHOU Dawei PU Chunying School of Information Engineering Nanyang Institute of TechnologyNanyang 473004China School of Mathematics and Physics Nanyang Institute of TechnologyNanyang 473004China School of Physics and Electronic Engineering Nanyang Normal UniversityNanyang473061China

The well-developed particle-swarm optimization method together with density functional theory calculations were employed to search lowest-energy geometric structures of two-dimensional(2D)SiGeP_(2).Two newly found structures(P3m1 and Pmm2)are *** unbiased global search reveals that the two lowest-energy structures are honeycomb lattices with robust dynamical stabilities.A more accurate Heyd-Scuseria-Ernzerhof(HSE06)hybrid functional is used to estimate the band structures of SiGeP_(2),which indicates that both the structures are semiconductors with indirect band-gap energies 1.80 e V for P3m1 and1.93 e V for Pmm^(2),*** the deformation potential theory,the P3m1-SiGeP_(2)is predicted to have high electron mobilities(6.4×10^(4)along zigzag direction and 2.9×10^(3)cm^(2)·V^(-1)·s^(-1)along armchair direction,respectively)and hole electron mobilities(1.0×10^(3)along zigzag direction and 2.5×10^(3)cm^(2)·V^(-1)·s^(-1)along armchair direction,respectively),which can be comparable with that of phosphorene and show anisotropic character *** addition,to estimate the elastic limit of SiGeP_(2),we also calculated the surface tension of SiGeP_(2)as a function of tensile *** results show that the 2D SiGeP_(2)may be good candidaticates for applications in nanoelectronic devices.

关键词： two dimensional SiGeP_(2) the first-principles carrier mobility

Study of biomass gasification in an industrial-scale dual circulating fluidized bed(DCFB)using the Eulerian-Lagrangian method

在线全文

学校读者我要写书评

暂无评论

Particuology 2023年第12期83卷 156-168页

作者： jiahui yu Shuai Wang Kun Luo Debo Li Jianren Fan State Key Laboratory of Clean Energy Utilization Zhejiang UniversityHangzhou310027China Shanghai Institute for Advanced Study of Zhejiang University Shanghai200120China Guangdong Diankeyuan Energy Technology Co. Ltd.Guangzhou510080China

Dual circulating fluidized bed(DCFB)has emerged as an efficient reactor for biomass gasification due to its unique feature of high gas-solid contact efficiency and separated reactions in two reactors,yet the understanding of complex in-furnace phenomena is still *** this study,biomass gasification in an industrial-scale DCFB system was numerically studied using a multiphase particle-in-cell(MP-PIC)method featuring thermochemical sub-models(e.g.,heat transfer,heterogeneous reactions,and homogeneous reactions)under the Eulerian-Lagrangian *** model validation,the hydrodynamics and thermochemical characteristics(i.e.,pressure,temperature,and species)in the DCFB are comprehensively *** results show that size-/density-induced segregation makes solid fuels concentrate on the bed *** momentum exchange leads to the continuous decrease of the gas pressure *** the gasifier and combustor,the lower heating value(LHV)of the gas products is 5.56 MJ/Nm^(3)and 0.2 MJ/Nm^(3)and the combustible gas concentration(CGC)is 65.5%and 1.86%,*** temperature in the combustor is about 100 K higher than that in the gasifier.A higher solid concentration results in a smaller value of particle heat transfer coefficient(HTC).The HTCs range from 50 to 150 W/(m^(2) K)for a solid concentration larger than 0.3 in the combustor while the HTCs range from 100 to 200 W/(m^(2 )K)in the *** Reynolds number of biomass particles is two orders of magnitude larger than that of the sand *** numerical results shed light on the reactor design and process optimization of biomass gasification in DCFBs.

关键词： Dual fluidized bed Biomass gasification Numerical simulation MP-PIC Heat transfer coefficient

Multi-scale numerical simulation of fluidized beds: Model applicability assessment

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Particuology 2023年第9期80卷 11-41页

作者： Shuai Wang Chenshu Hu Kun Luo jiahui yu Jianren Fan State Key Laboratory of Clean Energy Utilization Zhejiang UniversityHangzhou310027China Shanghai Institute for Advanced Study of Zhejiang University Shanghai200120China

In the past few decades,multi-scale numerical methods have been developed to model dense gas-solidflow in fluidized beds with different resolutions,accuracies,and ***,ambiguity needsto be clarified in the multi-scale numerical simulation of fluidized beds:(i)the selection of the submodels,parameters,and numerical resolution;(ii)the multivariate coupling of operating conditions,bed configurations,polydispersity,and additional ***,a state-of-the-art review is performed to assess the applicability of multi-scale numerical methods in predicting dense gas-solid flow influidized beds at specific fluidization regimes(e.g.,bubbling fluidization region,fast fluidization regime),with a focus on the inter-particle collision models,inter-phase interaction models,collision parameters,and polydispersity effect.A mutual restriction exists between resolution and *** methods need more computational resources and thus are suitable for smaller-scale simulations to provide a database for closure ***-resolution methods require fewercomputational resources and thus underpin large-scale simulations to explore macro-scale *** validations need to be further conducted under multiple flow conditions and comprehensivemetrics(e.g.,velocity profiles at different heights,bubbles,or cluster characteristics)for furtherimprovement of the applicability of each numerical method.

关键词： Fluidized bed Dense gas-solid flows Multi-scale numerical methods Drag model Inter-particle collisions

Impact of Surface Potential Vorticity Density Forcing over the Tibetan Plateau on the South China Extreme Precipitation in January 2008. Part Ⅱ:Numerical Simulation

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Meteorological Research 2019年第3期33卷 416-432页

作者： jiahui yu Yimin LIU Tingting MA Guoxiong WU State Key Laboratory of Numerical Modeling for Atmospheric Sciences and Geophysical Fluid Dynamics (LASG) Institute of Atmospheric Physics (IAP) Chinese Academy of Sciences Beijing 100029 Tianjin Meteorological Service Center Tianjin 300074 College of Earth Science University of Chinese Academy of Sciences Beijing 100049

The surface air convergence on the eastern flank of the Tibetan Plateau (TP) can increase the in situ surface potential vorticity density (PVD). Since the elevated TP intersects with the isentropic surfaces in the lower troposphere, the increased PVD on the eastern flank of TP thus forms a PVD forcing to the intersected isentropic surface in the boundary layer. The influence of surface PVD forcing over the TP on the extreme freezing rain/snow over South China in January 2008 is investigated by using numerical experiments based on the Finite-volume Atmospheric Model of the IAP/LASG (FAMIL). Compared with observations, the simulation results show that, by using a nudging method for assimilating observation data in the initial flow, this model can reasonably reproduce the distribution of precipitation, atmospheric circulation, and PVD propagation over and downstream of the TP during the extreme winter precipitation period. In order to investigate the impact of the increased surface PVD over the TP on the extreme precipitation in South China, a sensitivity experiment with surface PVD reduced over the TP region was performed. Compared with the control experiment, it is found that the precipitation in the TP downstream area, especially in Southeast China, is reduced. The rainband from Guangxi Region to Shandong Province has almost disappeared. In the lower troposphere, the increase of surface PVD over the TP region has generated an anomalous cyclonic circulation over southern China, which plays an important role in increasing southerly wind and the water vapor transport in this area;it also increases the northward negative absolute vorticity advection. In the upper troposphere, the surface PVD generated in eastern TP propagates on isentropic surface along westerly wind and results in positive absolute vorticity advection in the downstream areas. Consequently, due to the development of both ascending motion and water vapor transport in the downstream place of the TP, extre

关键词： extreme weather event Tibetan Plateau potential vorticity density (PVD) surface PVD forcing numerical simulation

Image analysis considering textual correlations enables accurate user switching tendency prediction

在线全文

学校读者我要写书评

暂无评论

Optoelectronics Letters 2023年第8期19卷 498-505页

作者： WANG Jianbin SHI Shuyuan WANG Xuna yu jiahui Zhejiang University Hangzhou 310027China China Telecom Zhejiang Branch Hangzhou 310014China School of Automation and Electrical Engineering Shenyang Ligong UniversityShenyang 110159China School of Computing University of PortsmouthPortsmouth PO13HEUK

Predicting likely-to-churn users employing surveys is a challenging *** with different personalities may make different choices in the same situation,so we introduced social media avatars that reflect the user’s psychological state when analyzing their churn *** this paper,we propose a multimodal framework that jointly learns image and text features to establish correlations among users with low net promoter score(NPS) and those likely to *** conducted experiments on actual data,and the results show that our proposed method can identify NPS-degraded users in advance,promoting the commercial development of the operator.

关键词： Image likely enable

Impact of Surface Potential Vorticity Density Forcing over the Tibetan Plateau on the South China Extreme Precipitation in January 2008. Part Ⅰ:Data Analysis

在线全文

学校读者我要写书评

暂无评论

Journal of Meteorological Research 2019年第3期33卷 400-415页

作者： Tingting MA Guoxiong WU Yimin LIU Zhihong JIANG jiahui yu Key Laboratory of Meteorological Disaster of Ministry of Education/Joint International Research Laboratory of Climate and Environmental Change Nanjing University of Information Science & Technology Nanjing 210044 State Key Laboratory of Numerical Modeling for Atmospheric Sciences and Geophysical Fluid Dynamics Institute of Atmospheric Physics Chinese Academy of Sciences Beijing 100029 College of Earth Science University of Chinese Academy of Sciences Beijing 100049 Center for Excellence in Tibetan Plateau Earth Sciences Chinese Academy of Sciences Beijing 100101

The external source/sink of potential vorticity (PV) is the original driving force for the atmospheric circulation. The relationship between surface PV generation and surface PV density forcing is discussed in detail in this paper. Moreover, a case study of the extreme winter freezing rain/snow storm over South China in January 2008 is performed, and the surface PV density forcing over the eastern flank of the Tibetan Plateau (TP) has been found to significantly affect the precipitation over South China in this case. The TP generated PV propagated eastward in the middle troposphere. The associated zonal advection of positive absolute vorticity resulted in the increasing of cyclo-nic relative vorticity in the downstream region of the TP. Ascending air and convergence in the lower troposphere developed, which gave rise to the development of the southerly wind. This favored the increasing of negative meridio-nal absolute vorticity advection in the lower troposphere, which provided a large-scale circulation background conducive to ascending motion such that the absolute vorticity advection increased with height. Consequently, the ascending air further strengthened the southerly wind and the vertical gradient of absolute vorticity advection between the lower and middle troposphere in turn. Under such a situation, the enhanced ascending, together with the moist air transported by the southerly wind, formed the extreme winter precipitation in January 2008 over South China.

关键词： potential vorticity density (PVD) absolute vorticity advection Tibetan Plateau winter precipitation

Enhanced Sensitivity and Selectivity of Chemosensor for Malonate by Anchoring on Gold Nanoparticles

在线全文

学校读者我要写书评

暂无评论

Chinese Journal of Chemistry 2011年第3期29卷 531-538页

作者： Liu, Yanyun Liu, yun Liang, Zhongshi Li, Xiangyang Liu, Shunying yu, jiahui Institute for Advanced Interdisciplinary Research East China Normal University Shanghai 200062 China

A novel organic-inorganic nanohybrided receptor 4 functionalized with bis-thiourea arms and then assembled on gold nanoparticles was synthesized. The preorganized system possesses phenylthiourea units for the spectropho- tometric sensing of dicarboxylates, especially malonate, based on changes in the surface plasmon absorption of the gold nanoparticles （GNPs）. The intensity of the absorbance band increases gradually with the concentration of di- carboxylates increasing. But such an ion-selective change in the plasmon band was not observed in control tests carried out by mono-thiourea-modified GNPs, receptor 3 and the free receptor 2. As it is shown from the association constants, derived from quantitative titrations, receptor 4 can selectively recognize dicarboxylate anions of shorter carbon chain, and has the highest affinity to malonate. The interaction properties for anions of receptor 4 were evaluated by lH NMR and UV-vis spectroscopic methods.

关键词： thiourea gold nanoparticle assembly manolate sensor

Kinome-wide polypharmacology profiling of small molecules by multi-task graph isomorphism network approach

维普期刊数据库评论

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Acta Pharmaceutica Sinica B 2023年第1期13卷 54-67页

作者： Lingjie Bao Zhe Wang Zhenxing Wu Hao Luo jiahui yu yu Kang Dongsheng Cao Tingjun Hou Innovation Institute for Artificial Intelligence in Medicine of Zhejiang University College of Pharmaceutical SciencesZhejiang UniversityHangzhou 310058China State Key Lab of CAD&CG Zhejiang UniversityHangzhou 310058China Xiangya School of Pharmaceutical Sciences Central South UniversityChangsha 410013China

Prediction of the interactions between small molecules and their targets play important roles in various applications of drug development,such as lead discovery,drug repurposing and elucidation of potential drug side ***,a variety of machine learning-based models have been developed to predict these *** this study,a model called auxiliary multi-task graph isomorphism network with uncertainty weighting(AMGU)was developed to predict the inhibitory activities of small molecules against 204 different kinases based on the multi-task Graph Isomorphism Network(MT-GIN)with the auxiliary learning and uncertainty weighting *** calculation results illustrate that the AMGU model outperformed the descriptor-based models and state-of-the-art graph neural networks(GNN)models on the internal test ***,it also exhibited much better performance on two external test sets,suggesting that the AMGU model has enhanced generalizability due to its great transfer learning ***,a naÏve model-agnostic interpretable method for GNN called edges masking was devised to explain the underlying predictive mechanisms,and the consistency of the interpretability results for 5typical epidermal growth factor receptor(EGFR)inhibitors with their structure-activity relationships could be ***,a free online web server called KIP was developed to predict the kinomewide polypharmacology effects of small molecules(http://***/kip).

关键词： Kinome-wide polypharmacology Machine learning Kinases Graph neural networks Artificial intelligence