The traditional ventilation mode of subway vehicles adopts the form that the inlets and outlets are placed on the upper part of the cabin.The air distribution formed in this mode often cause serious problems of therma...The traditional ventilation mode of subway vehicles adopts the form that the inlets and outlets are placed on the upper part of the cabin.The air distribution formed in this mode often cause serious problems of thermal comfort and energy consumption.In order to solve these problems caused by the traditional ventilation mode,a new hybrid ventilation mode was proposed.The hybrid ventilation mode uses both upper and underside air supply inlets.A method for evaluating the air distribution performance of subway air conditioning was developed.The method applies non-uniformity coefficients,maximum temperature difference,air diffusion performance index,modified energy utilization coefficient and Air short-circuit comprehensive coefficient.Air short-circuit comprehensive coefficient was a new index to evaluate the degree of air short-circuit of supply air.Based on the airflow simulation,the air distribution performance for the hybrid ventilation mode was evaluated using these indexes,and compared with the traditional ventilation mode.The results show that compared with the traditional ventilation mode,the hybrid ventilation mode has more uniform temperature distribution,better thermal comfort,higher energy utilization efficiency and lower degree of air short-circuit of supply air.展开更多
The effects of optical losses oil a directly-modulated radio-over-fiber （RoF） system used for distributed antenna networks are determined. The results show that with a properly designed bidirectional amplifier, the ...The effects of optical losses oil a directly-modulated radio-over-fiber （RoF） system used for distributed antenna networks are determined. The results show that with a properly designed bidirectional amplifier, the RoF link can tolerate over 20 and 16 dB of optical losses for down- and up-links, respectively. Simulation results are also consistent with the experimental data. These findings can contribute to tile design of RoF distributed antenna systems with different topologies.展开更多
Cross-silo federated learning(FL),which benefits from relatively abundant data and rich computing power,is drawing increasing focus due to the significant transformations that foundation models(FMs)are instigating in ...Cross-silo federated learning(FL),which benefits from relatively abundant data and rich computing power,is drawing increasing focus due to the significant transformations that foundation models(FMs)are instigating in the artificial intelligence field.The intensified data heterogeneity issue of this area,unlike that in cross-device FL,is caused mainly by substantial data volumes and distribution shifts across clients,which requires algorithms to comprehensively consider the personalization and generalization balance.In this paper,we aim to address the objective of generalized and personalized federated learning(GPFL)by enhancing the global model’s cross-domain generalization capabilities and simultaneously improving the personalization performance of local training clients.By investigating the fairness of performance distribution within the federation system,we explore a new connection between generalization gap and aggregation weights established in previous studies,culminating in the fairness-guided federated training for generalization and personalization(FFT-GP)approach.FFT-GP integrates a fairness-aware aggregation(FAA)approach to minimize the generalization gap variance among training clients and a meta-learning strategy that aligns local training with the global model’s feature distribution,thereby balancing generalization and personalization.Our extensive experimental results demonstrate FFT-GP’s superior efficacy compared to existing models,showcasing its potential to enhance FL systems across a variety of practical scenarios.展开更多
文摘The traditional ventilation mode of subway vehicles adopts the form that the inlets and outlets are placed on the upper part of the cabin.The air distribution formed in this mode often cause serious problems of thermal comfort and energy consumption.In order to solve these problems caused by the traditional ventilation mode,a new hybrid ventilation mode was proposed.The hybrid ventilation mode uses both upper and underside air supply inlets.A method for evaluating the air distribution performance of subway air conditioning was developed.The method applies non-uniformity coefficients,maximum temperature difference,air diffusion performance index,modified energy utilization coefficient and Air short-circuit comprehensive coefficient.Air short-circuit comprehensive coefficient was a new index to evaluate the degree of air short-circuit of supply air.Based on the airflow simulation,the air distribution performance for the hybrid ventilation mode was evaluated using these indexes,and compared with the traditional ventilation mode.The results show that compared with the traditional ventilation mode,the hybrid ventilation mode has more uniform temperature distribution,better thermal comfort,higher energy utilization efficiency and lower degree of air short-circuit of supply air.
基金supported by the National "863" Program of China(No.2011AA010306)the National "973" Program of China(No.2012CB315705)+2 种基金the National Natural Science Foundation of China(Nos.61107058,61120106001,60932004,61001121,and 60837004)the Beijing Excellent Doctoral Thesis Project(No.YB20101001301)the Cooperation Project between Province and Ministries(No.2011A090200025)
文摘The effects of optical losses oil a directly-modulated radio-over-fiber （RoF） system used for distributed antenna networks are determined. The results show that with a properly designed bidirectional amplifier, the RoF link can tolerate over 20 and 16 dB of optical losses for down- and up-links, respectively. Simulation results are also consistent with the experimental data. These findings can contribute to tile design of RoF distributed antenna systems with different topologies.
基金Project supported by the National Key R&D Program of China(No.2022ZD0160702)the STCSM(Nos.22511106101,18DZ2270700,and 21DZ1100-100)+1 种基金the 111 Plan(No.BP0719010)the State Key Laboratory of UHD Video and Audio Production and Presentation。
文摘Cross-silo federated learning(FL),which benefits from relatively abundant data and rich computing power,is drawing increasing focus due to the significant transformations that foundation models(FMs)are instigating in the artificial intelligence field.The intensified data heterogeneity issue of this area,unlike that in cross-device FL,is caused mainly by substantial data volumes and distribution shifts across clients,which requires algorithms to comprehensively consider the personalization and generalization balance.In this paper,we aim to address the objective of generalized and personalized federated learning(GPFL)by enhancing the global model’s cross-domain generalization capabilities and simultaneously improving the personalization performance of local training clients.By investigating the fairness of performance distribution within the federation system,we explore a new connection between generalization gap and aggregation weights established in previous studies,culminating in the fairness-guided federated training for generalization and personalization(FFT-GP)approach.FFT-GP integrates a fairness-aware aggregation(FAA)approach to minimize the generalization gap variance among training clients and a meta-learning strategy that aligns local training with the global model’s feature distribution,thereby balancing generalization and personalization.Our extensive experimental results demonstrate FFT-GP’s superior efficacy compared to existing models,showcasing its potential to enhance FL systems across a variety of practical scenarios.
