Federated learning has emerged as a distributed learning paradigm by training at each client and aggregat-ing at a parameter server.System heterogeneity hinders stragglers from responding to the server in time with hu...Federated learning has emerged as a distributed learning paradigm by training at each client and aggregat-ing at a parameter server.System heterogeneity hinders stragglers from responding to the server in time with huge com-munication costs.Although client grouping in federated learning can solve the straggler problem,the stochastic selection strategy in client grouping neglects the impact of data distribution within each group.Besides,current client grouping ap-proaches make clients suffer unfair participation,leading to biased performances for different clients.In order to guaran-tee the fairness of client participation and mitigate biased local performances,we propose a federated dynamic client selec-tion method based on data representativity(FedSDR).FedSDR clusters clients into groups correlated with their own lo-cal computational efficiency.To estimate the significance of client datasets,we design a novel data representativity evalua-tion scheme based on local data distribution.Furthermore,the two most representative clients in each group are selected to optimize the global model.Finally,the DYNAMIC-SELECT algorithm updates local computational efficiency and data representativity states to regroup clients after periodic average aggregation.Evaluations on real datasets show that FedS-DR improves client participation by 27.4%,37.9%,and 23.3%compared with FedAvg,TiFL,and FedSS,respectively,tak-ing fairness into account in federated learning.In addition,FedSDR surpasses FedAvg,FedGS,and FedMS by 21.32%,20.4%,and 6.90%,respectively,in local test accuracy variance,balancing the performance bias of the global model across clients.展开更多
YTB block in Sichuan basin is a favorable area to exploit oil and gas in shallow tight rock. 3D seismic project of this zone has two characteristics. Firstly, it has high requirements for the tolerance rate of the con...YTB block in Sichuan basin is a favorable area to exploit oil and gas in shallow tight rock. 3D seismic project of this zone has two characteristics. Firstly, it has high requirements for the tolerance rate of the construction process and the acquisition of high signal-to-noise ratio seismic data;Second, there are widely obstacles and noises that lead to difficult acquisition construction organization. In acquisition practice, high signal-to-noise ratio seismic data was obtained by reasonable design of construction scheme, optimization of excitation parameters, improvement of receiving conditions and optimization of obstacle crossing observation system. .展开更多
The pandemic of respiratory diseases enlightened people that monitoring respiration has promising prospects in averting many fatalities by tracking the development of diseases.However,the response speed of current opt...The pandemic of respiratory diseases enlightened people that monitoring respiration has promising prospects in averting many fatalities by tracking the development of diseases.However,the response speed of current optical fiber sensors is still insufficient to meet the requirements of high-frequency respiratory detection during respiratory failure.Here,a scheme for a fast and stable tachypnea monitor is proposed utilizing a water-soluble C_(60)-Lys ion compound as functional material for the tracking of humidity change in the progression of breath.The polarization of C_(60)-Lys can be tuned by the ambient relative humidity change,and an apparent refractive index alteration can be detected due to the small size effect.In our experiments,C_(60)-Lys is conformally and uniformly deposited on the surface of a tilted fiber Bragg grating(TFBG)to fabricate an ultra-fast-response,high-sensitivity,and long-term stable optical fiber humidity sensor.A relative humidity(RH)detecting sensitivity of 0.080 dB/%RH and the equilibrium response time and recovery time of 1.85 s and 1.58 s are observed,respectively.Also,a linear relation is detected between the resonance intensity of the TFBG and the environment RH.In a practical breath monitoring experiment,the instantaneous response time and recovery time are measured as 40 ms and 41 ms,respectively,during a 1.5 Hz fast breath process.Furthermore,an excellent time stability and high repeatability are exhibited in experiments conducted over a range of 7 days.展开更多
All-dielectric metasurfaces are usually limited because of their static functionality and small scale.In this paper,we use an easy nanofabrication technique to fabricate all-dielectric metasurfaces with the advantages...All-dielectric metasurfaces are usually limited because of their static functionality and small scale.In this paper,we use an easy nanofabrication technique to fabricate all-dielectric metasurfaces with the advantages of having dynamic tunability and a large area.Using an anodized aluminum oxide(AAO)template as an evaporation mask,a large-area metasurface embedded in polydimethylsiloxane(PDMS)(>2 cm^(2))is fabricated.The metasurface exhibits remarkable electric dipole(ED)and magnetic dipole(MD)resonances.Based on the solvent-swelling effect of PDMS in 20%toluene,the ED/MD resonance peak shifts dynamically∼40 nm to red.So far,to the best of our knowledge,a large-area metasurface embedded in PDMS and achieved by using the AAO template method has not appeared.展开更多
The practice of exploration and production has proved that explosives are excited in different surrounding rocks and the seismic wavelets collected have different characteristics. In this paper, by establishing a nume...The practice of exploration and production has proved that explosives are excited in different surrounding rocks and the seismic wavelets collected have different characteristics. In this paper, by establishing a numerical model of the explosion in the well, using finite element analysis technology for numerical simulation, the simulation calculated the stress structure in the near-source area of the earthquake excitation, and extracted the seismic wavelet. The results show that the simulation seismic wavelet characteristics of different thin interbedded sand and mudstone structures have changed significantly. Through excitation simulation, the amplitude and spectrum information of seismic wavelets can be compared and analyzed, and the excitation parameters can be optimized. .展开更多
With the rapid development of the UHV power grid, evaluation of the economic and social benefits of the UHV power grid is conducive to guiding the planning and construction of UHV power grid. At present economic benef...With the rapid development of the UHV power grid, evaluation of the economic and social benefits of the UHV power grid is conducive to guiding the planning and construction of UHV power grid. At present economic benefit evaluation system of the UHV power grid is driving to maturity stage on the whole at home and abroad, but it invariably tends to regard social benefits as part of economic benefits, without evaluating social benefit separately. The social benefit evaluation model of UHV power grid is built in case of sufficient investigation. The differentials between social benefit and social cost are calculated respectively by three kinds of solutions according to the constructed social cost evaluation index system and social benefits evaluation index system, conclusion that UHV power grid transmission has better social benefits can be reached by contrastively analyzing the social three kinds of solutions corresponding to benefits. At last, the evaluation model and method are verified and analyzed through the analysis of engineering projects.展开更多
<div style="text-align:justify;"> In view of the shallow buried depth of the target layer and it is difficult to obtain high-quality seismic data of HCX gas storage project, based on the acquisition pa...<div style="text-align:justify;"> In view of the shallow buried depth of the target layer and it is difficult to obtain high-quality seismic data of HCX gas storage project, based on the acquisition parameters of the adjacent area, the 3D acquisition technical scheme is designed for focusing on the imaging of the top of HCX high-steep anticline. By optimizing the acquisition parameters and the excitation and receiving conditions, the acquisition of high-resolution and high signal-to-noise ratio seismic data is guaranteed to meet the demand of fracture prediction and reservoir prediction. </div>展开更多
As an important class of resource allocation approaches, decentralised job scheduling in large-scale grids has to deal with the difficulties in acquiring timely model information and improving performance by autonomou...As an important class of resource allocation approaches, decentralised job scheduling in large-scale grids has to deal with the difficulties in acquiring timely model information and improving performance by autonomous coordination. In this study, a gossip-based reinforcement learning （GRL） method is proposed for decentralised job scheduling in grids. In the GRL method, a decentralised scheduling architecture based on multi-agent reinforcement learning is presented to improve the scalability and adaptability of job scheduling. A gossip mechanism is designed to realise autonomous coordination among the decentralised schedulers. Simulation results show that the proposed GRL-based schedulers can complete the task of grid job scheduling effectively and achieve load balancing efficiently.展开更多
基金This work is supported by the National Key Research and Development Program of China under Grant No.2022YFC3005401the Key Research and Development Program of Yunnan Province of China under Grant No.202203AA080009+1 种基金the Transformation Program of Scientific and Technological Achievements of Jiangsu Province of China under Grant No.BA2021002the Key Research and Development Program of Jiangsu Province of Chin under Grant No.BE2020729.
文摘Federated learning has emerged as a distributed learning paradigm by training at each client and aggregat-ing at a parameter server.System heterogeneity hinders stragglers from responding to the server in time with huge com-munication costs.Although client grouping in federated learning can solve the straggler problem,the stochastic selection strategy in client grouping neglects the impact of data distribution within each group.Besides,current client grouping ap-proaches make clients suffer unfair participation,leading to biased performances for different clients.In order to guaran-tee the fairness of client participation and mitigate biased local performances,we propose a federated dynamic client selec-tion method based on data representativity(FedSDR).FedSDR clusters clients into groups correlated with their own lo-cal computational efficiency.To estimate the significance of client datasets,we design a novel data representativity evalua-tion scheme based on local data distribution.Furthermore,the two most representative clients in each group are selected to optimize the global model.Finally,the DYNAMIC-SELECT algorithm updates local computational efficiency and data representativity states to regroup clients after periodic average aggregation.Evaluations on real datasets show that FedS-DR improves client participation by 27.4%,37.9%,and 23.3%compared with FedAvg,TiFL,and FedSS,respectively,tak-ing fairness into account in federated learning.In addition,FedSDR surpasses FedAvg,FedGS,and FedMS by 21.32%,20.4%,and 6.90%,respectively,in local test accuracy variance,balancing the performance bias of the global model across clients.
文摘YTB block in Sichuan basin is a favorable area to exploit oil and gas in shallow tight rock. 3D seismic project of this zone has two characteristics. Firstly, it has high requirements for the tolerance rate of the construction process and the acquisition of high signal-to-noise ratio seismic data;Second, there are widely obstacles and noises that lead to difficult acquisition construction organization. In acquisition practice, high signal-to-noise ratio seismic data was obtained by reasonable design of construction scheme, optimization of excitation parameters, improvement of receiving conditions and optimization of obstacle crossing observation system. .
基金supported by the National Natural Science Foundation of China(Nos.12274386 and 52002365)Zhejiang Provincial Natural Science Foundation of China(Nos.LQ23F050006,LY21F050006,and LQ21E020005)+1 种基金Key R&D Project of Zhejiang Province(No.2021C01179)National Key R&D Program of China(No.2021YFF0600203)。
文摘The pandemic of respiratory diseases enlightened people that monitoring respiration has promising prospects in averting many fatalities by tracking the development of diseases.However,the response speed of current optical fiber sensors is still insufficient to meet the requirements of high-frequency respiratory detection during respiratory failure.Here,a scheme for a fast and stable tachypnea monitor is proposed utilizing a water-soluble C_(60)-Lys ion compound as functional material for the tracking of humidity change in the progression of breath.The polarization of C_(60)-Lys can be tuned by the ambient relative humidity change,and an apparent refractive index alteration can be detected due to the small size effect.In our experiments,C_(60)-Lys is conformally and uniformly deposited on the surface of a tilted fiber Bragg grating(TFBG)to fabricate an ultra-fast-response,high-sensitivity,and long-term stable optical fiber humidity sensor.A relative humidity(RH)detecting sensitivity of 0.080 dB/%RH and the equilibrium response time and recovery time of 1.85 s and 1.58 s are observed,respectively.Also,a linear relation is detected between the resonance intensity of the TFBG and the environment RH.In a practical breath monitoring experiment,the instantaneous response time and recovery time are measured as 40 ms and 41 ms,respectively,during a 1.5 Hz fast breath process.Furthermore,an excellent time stability and high repeatability are exhibited in experiments conducted over a range of 7 days.
基金financially supported by the National Natural Science Foundation of China(No.62075040)National Key Research and Development Program of China(No.2021YFF0701100)。
文摘All-dielectric metasurfaces are usually limited because of their static functionality and small scale.In this paper,we use an easy nanofabrication technique to fabricate all-dielectric metasurfaces with the advantages of having dynamic tunability and a large area.Using an anodized aluminum oxide(AAO)template as an evaporation mask,a large-area metasurface embedded in polydimethylsiloxane(PDMS)(>2 cm^(2))is fabricated.The metasurface exhibits remarkable electric dipole(ED)and magnetic dipole(MD)resonances.Based on the solvent-swelling effect of PDMS in 20%toluene,the ED/MD resonance peak shifts dynamically∼40 nm to red.So far,to the best of our knowledge,a large-area metasurface embedded in PDMS and achieved by using the AAO template method has not appeared.
文摘The practice of exploration and production has proved that explosives are excited in different surrounding rocks and the seismic wavelets collected have different characteristics. In this paper, by establishing a numerical model of the explosion in the well, using finite element analysis technology for numerical simulation, the simulation calculated the stress structure in the near-source area of the earthquake excitation, and extracted the seismic wavelet. The results show that the simulation seismic wavelet characteristics of different thin interbedded sand and mudstone structures have changed significantly. Through excitation simulation, the amplitude and spectrum information of seismic wavelets can be compared and analyzed, and the excitation parameters can be optimized. .
基金supported by the National Key Research and Development Program of China(2022YFB3705201)the National Natural Science Foundation of China(51971018 and U20B2025)+3 种基金the Funds for Creative Research Groups of National Natural Science Foundation of China(51921001)111 Project(BP0719004)the Program for Changjiang Scholars and Innovative Research Team in University of China(IRT_14R05)the Fundamental Research Funds for the Central Universities of China(FRF-MP-20-43Z)。
基金supported by the National Natural Science Foundation of China (No. 51775416)National Thousand Talents Program of China (No. WQ2017610446)supported by the Project of Innovation Team of Xi’an Jiaotong University, China (No. XTR0118008)
基金the Top Talent Support Program for Medical Experts Team and for Young and Middle-Aged People of Wuxi Health Committee(202109 and 202014)the National Key R&D Program of China(2021YFC2501100 and 2020YFA0803700)the National Natural Science Foundation of China(82071296,81801158,and 81970425)。
文摘With the rapid development of the UHV power grid, evaluation of the economic and social benefits of the UHV power grid is conducive to guiding the planning and construction of UHV power grid. At present economic benefit evaluation system of the UHV power grid is driving to maturity stage on the whole at home and abroad, but it invariably tends to regard social benefits as part of economic benefits, without evaluating social benefit separately. The social benefit evaluation model of UHV power grid is built in case of sufficient investigation. The differentials between social benefit and social cost are calculated respectively by three kinds of solutions according to the constructed social cost evaluation index system and social benefits evaluation index system, conclusion that UHV power grid transmission has better social benefits can be reached by contrastively analyzing the social three kinds of solutions corresponding to benefits. At last, the evaluation model and method are verified and analyzed through the analysis of engineering projects.
文摘<div style="text-align:justify;"> In view of the shallow buried depth of the target layer and it is difficult to obtain high-quality seismic data of HCX gas storage project, based on the acquisition parameters of the adjacent area, the 3D acquisition technical scheme is designed for focusing on the imaging of the top of HCX high-steep anticline. By optimizing the acquisition parameters and the excitation and receiving conditions, the acquisition of high-resolution and high signal-to-noise ratio seismic data is guaranteed to meet the demand of fracture prediction and reservoir prediction. </div>
基金This work was supported by the National Natural Science Foundation of China under grant U1564214.
文摘As an important class of resource allocation approaches, decentralised job scheduling in large-scale grids has to deal with the difficulties in acquiring timely model information and improving performance by autonomous coordination. In this study, a gossip-based reinforcement learning （GRL） method is proposed for decentralised job scheduling in grids. In the GRL method, a decentralised scheduling architecture based on multi-agent reinforcement learning is presented to improve the scalability and adaptability of job scheduling. A gossip mechanism is designed to realise autonomous coordination among the decentralised schedulers. Simulation results show that the proposed GRL-based schedulers can complete the task of grid job scheduling effectively and achieve load balancing efficiently.
