1 Introduction Current continual learning methods[1–4]can utilize labeled data to alleviate catastrophic forgetting effectively.However,obtaining labeled samples can be difficult and tedious as it may require expert ...1 Introduction Current continual learning methods[1–4]can utilize labeled data to alleviate catastrophic forgetting effectively.However,obtaining labeled samples can be difficult and tedious as it may require expert knowledge.In many practical application scenarios,labeled and unlabeled samples exist simultaneously,with more unlabeled than labeled samples in streaming data[5,6].Unfortunately,existing class-incremental learning methods face limitations in effectively utilizing unlabeled data,thereby impeding their performance in incremental learning scenarios.展开更多
This study examines the long-term impact of historical clan strength on contemporary human capital development in China.Our findings reveal that historical clan strength had a significant positive effect on individual...This study examines the long-term impact of historical clan strength on contemporary human capital development in China.Our findings reveal that historical clan strength had a significant positive effect on individual human capital,particularly in rural households,even afier accounting for confounders and robustness checks.We observe that clan culture complemented the implementation of compulsory education in shaping contemporary educational outcomes.Mechanism analysis suggests that clan strength motivated parents to prioritize their children's education and increased the likelihood of low-income individuals receiving financial assistance from relatives and friends.However,we find no evidence that clan strength affected the allocation of public educational resources.These findings highlight the enduring role of cultural transmission mechanisms in human capital formation and underscore the importance of incorporating cultural factors into the design of educational policies.展开更多
Mixed insulation oil is considered as a desired substitute for traditional mineral oil.A novel 3-element mixed insulation oil has been developed in our previous works,whose parameters satisfy the mineral oil standard(...Mixed insulation oil is considered as a desired substitute for traditional mineral oil.A novel 3-element mixed insulation oil has been developed in our previous works,whose parameters satisfy the mineral oil standard(IEC 60296:2012).In the application of transformer oil,dissolved gas analysis is an effective method for insulation fault diagnosis.Thermal fault is a common and typical fault in transformers;however,the pyrolysis process and gas production characteristics of mixed insulation oil under thermal fault remain unknown.In this study,the pyrolysis process and gas production characteristics of 3-element mixed insulation oil have been studied through simulation and experiments,and mineral oil is equally treated for comparison.Results show that the gas production characteristics of mixed oil are closely related to its base oils,and 3-element mixed insulation oil generates more H_(2)and C_(2)H_(6)under thermal fault,in which the percentage of H_(2)in the mixed oil is always three times that in the mineral oil.More CO_(2)is produced in mixed oil under thermal fault because of ester bonds;C_(2)H_(4)is also a main characteristic gas for mixed insulation oil under thermal fault,and its proportion increases continuously with the increase of temperature.For mineral oil,the weight loss temperature in thermogravimetric(TG)curve corresponds to evaporation rather than decomposition.Natural esters have poor thermal stability according to simulation results,but they possess high flash point and weight loss temperature in the TG curve primarily because of their large average molecular weight.展开更多
Insulation oil is an important dielectric in power devices,and many studies on mixed insulation oil have been conducted in recent years to improve the performance of insulation oils.To replace mineral oil directly,the...Insulation oil is an important dielectric in power devices,and many studies on mixed insulation oil have been conducted in recent years to improve the performance of insulation oils.To replace mineral oil directly,the authors previously developed a novel three-element mixed insulation oil successfully;the main parameters of which satisfy the IEC 60296-2012 standard for mineral oil.In the present study,the AC breakdown properties of insulation paper(pressboard)immersed by the new mixed oil and naphthenic mineral oil were compared.For both insulation oils,the increase in temperature cannot significantly reduce the breakdown strength of oil-immersed insulation paper(pressboard)at a low moisture content,and the increment of moisture content cannot reduce the breakdown voltage at a low temperature(25℃).The breakdown voltage decreases only when the two factors increase simultaneously.For the mixed and mineral oils,the AC breakdown voltage of oilimmersed paper(pressboard)with different thicknesses has significant difference.The mineral oil has a high breakdown voltage for the thin insulation paper,whereas the mixed oil has a high breakdown voltage when the thickness of the insulation paper(pressboard)exceeds 0.2 mm.This phenomenon is mainly caused by the breakdown field strength that varies with the increase of dielectric thickness and the different change trends of paper(pressboard)immersed with the mixed and the mineral oils.Moreover,the stack of thin multilayer insulation paper enables the mixed oil-immersed paper to have a higher breakdown strength than the mineral oil-immersed paper.For the AC breakdown voltage of oil-immersed pressboard with an oil-gap structure,the mixed oil is comprehensively superior to the mineral oil due to its larger relative permittivity.展开更多
Furfural content in oil is considered as an effective parameter to evaluate the ageing state of oil-immersed transformers.In previous thermal ageing experiments at 130℃,we found that moisture increment in oil-paper s...Furfural content in oil is considered as an effective parameter to evaluate the ageing state of oil-immersed transformers.In previous thermal ageing experiments at 130℃,we found that moisture increment in oil-paper system makes more furfural distribute in oil.However,the microcosmic mechanism of moisture effect on furfural diffusion from paper to oil is still unknown,and the furfural equilibrium law at transformers'typical operating temperatures is also unclear,resulting in the unreliability of the existed furfural evaluation model.To elucidate the above problems,molecular dynamics simulation and furfural equilibrium experiments at 60℃ were carried out at different moisture contents.The results show that high moisture content makes part of furfural diffuse from paper to oil indeed.Supposing moisture content increases from 1% to 2%,mass fraction of furfural in oil may increase from 32.98% to 49.83%,leading to furfural content in oil increases from 0.76 mg/L to 1.06 mg/L when the degree of polymerisation of insulation paper is about 524,which increased by 39.5%.The increase in moisture content makes cellulose form hydrogen bonds with H_(2)O preferentially,decreasing the hydrogen bonds interaction between cellulose and furfural.Thus,partial furfural falls off cellulose and diffuses into oil,leading to an increase in furfural content in oil.To eliminate the moisture interference,a corrected furfural assessment model based on the moisture equilibrium curve has been constructed,which can accurately evaluate the ageing state of insulation paper regardless of the moisture influence.展开更多
Reinforcement learning(RL)has become a dominant decision-making paradigm and has achieved notable success in many real-world applications.Notably,deep neural networks play a crucial role in unlocking RL’s potential i...Reinforcement learning(RL)has become a dominant decision-making paradigm and has achieved notable success in many real-world applications.Notably,deep neural networks play a crucial role in unlocking RL’s potential in large-scale decision-making tasks.Inspired by current major success of Transformer in natural language processing and computer vision,numerous bottlenecks have been overcome by combining Transformer with RL for decision-making.This paper presents a multiangle systematic survey of various Transformer-based RL(TransRL)models applied in decision-making tasks,including basic models,advanced algorithms,representative implementation instances,typical applications,and known challenges.Our work aims to provide insights into problems that inherently arise with the current RL approaches,and examines how we can address them with better TransRL models.To our knowledge,we are the first to present a comprehensive review of the recent Transformer research developments in RL for decision-making.We hope that this survey provides a comprehensive review of TransRL models and inspires the RL community in its pursuit of future directions.To keep track of the rapid TransRL developments in the decision-making domains,we summarize the latest papers and their open-source implementations at https://github.com/williamyuanv0/Transformer-in-Reinforcement-Learning-for-Decision-Making-A-Survey.展开更多
This study assesses the effect of human capital expansion on China's export product quality.It employs the difference-in-differences(DID)framework based on a quasi-natural experiment investigating the 1999 higher ...This study assesses the effect of human capital expansion on China's export product quality.It employs the difference-in-differences(DID)framework based on a quasi-natural experiment investigating the 1999 higher education enrollment expansion as the exogenous policy shock.The empirical results confirm that human capital expansion appreciably improved the quality of China's export products.Human capital expansion promoted the transformation and upgrading of old products and the development of new products in term of intensive margin;it strengthened the endowment advantages of incumbent high-quality export enterprises while preventing low-quality enterprises from entering the market through price competition on the extensive margin.This prevented quality decline.Further,the study reveals that the improvement effect driven by human capital came from both innovation-induced and managerial efficiency improvement channels and was more prominent for large or foreign-funded enterprises in the eastern region.Our findings highlight the role of human capital in China's remarkable export performance from an endogenous growth perspective.展开更多
文摘1 Introduction Current continual learning methods[1–4]can utilize labeled data to alleviate catastrophic forgetting effectively.However,obtaining labeled samples can be difficult and tedious as it may require expert knowledge.In many practical application scenarios,labeled and unlabeled samples exist simultaneously,with more unlabeled than labeled samples in streaming data[5,6].Unfortunately,existing class-incremental learning methods face limitations in effectively utilizing unlabeled data,thereby impeding their performance in incremental learning scenarios.
基金supported by the National Natural Science Foundation of China(No.72373057)and the National Social Sciences Foundation of China(No.23AZD085).
文摘This study examines the long-term impact of historical clan strength on contemporary human capital development in China.Our findings reveal that historical clan strength had a significant positive effect on individual human capital,particularly in rural households,even afier accounting for confounders and robustness checks.We observe that clan culture complemented the implementation of compulsory education in shaping contemporary educational outcomes.Mechanism analysis suggests that clan strength motivated parents to prioritize their children's education and increased the likelihood of low-income individuals receiving financial assistance from relatives and friends.However,we find no evidence that clan strength affected the allocation of public educational resources.These findings highlight the enduring role of cultural transmission mechanisms in human capital formation and underscore the importance of incorporating cultural factors into the design of educational policies.
基金Natural Science Foundation of Hebei Province,Grant/Award Number:E2021202065。
文摘Mixed insulation oil is considered as a desired substitute for traditional mineral oil.A novel 3-element mixed insulation oil has been developed in our previous works,whose parameters satisfy the mineral oil standard(IEC 60296:2012).In the application of transformer oil,dissolved gas analysis is an effective method for insulation fault diagnosis.Thermal fault is a common and typical fault in transformers;however,the pyrolysis process and gas production characteristics of mixed insulation oil under thermal fault remain unknown.In this study,the pyrolysis process and gas production characteristics of 3-element mixed insulation oil have been studied through simulation and experiments,and mineral oil is equally treated for comparison.Results show that the gas production characteristics of mixed oil are closely related to its base oils,and 3-element mixed insulation oil generates more H_(2)and C_(2)H_(6)under thermal fault,in which the percentage of H_(2)in the mixed oil is always three times that in the mineral oil.More CO_(2)is produced in mixed oil under thermal fault because of ester bonds;C_(2)H_(4)is also a main characteristic gas for mixed insulation oil under thermal fault,and its proportion increases continuously with the increase of temperature.For mineral oil,the weight loss temperature in thermogravimetric(TG)curve corresponds to evaporation rather than decomposition.Natural esters have poor thermal stability according to simulation results,but they possess high flash point and weight loss temperature in the TG curve primarily because of their large average molecular weight.
文摘Insulation oil is an important dielectric in power devices,and many studies on mixed insulation oil have been conducted in recent years to improve the performance of insulation oils.To replace mineral oil directly,the authors previously developed a novel three-element mixed insulation oil successfully;the main parameters of which satisfy the IEC 60296-2012 standard for mineral oil.In the present study,the AC breakdown properties of insulation paper(pressboard)immersed by the new mixed oil and naphthenic mineral oil were compared.For both insulation oils,the increase in temperature cannot significantly reduce the breakdown strength of oil-immersed insulation paper(pressboard)at a low moisture content,and the increment of moisture content cannot reduce the breakdown voltage at a low temperature(25℃).The breakdown voltage decreases only when the two factors increase simultaneously.For the mixed and mineral oils,the AC breakdown voltage of oilimmersed paper(pressboard)with different thicknesses has significant difference.The mineral oil has a high breakdown voltage for the thin insulation paper,whereas the mixed oil has a high breakdown voltage when the thickness of the insulation paper(pressboard)exceeds 0.2 mm.This phenomenon is mainly caused by the breakdown field strength that varies with the increase of dielectric thickness and the different change trends of paper(pressboard)immersed with the mixed and the mineral oils.Moreover,the stack of thin multilayer insulation paper enables the mixed oil-immersed paper to have a higher breakdown strength than the mineral oil-immersed paper.For the AC breakdown voltage of oil-immersed pressboard with an oil-gap structure,the mixed oil is comprehensively superior to the mineral oil due to its larger relative permittivity.
基金National Natural Science Foundation of China,Grant/Award Number:52207152Opening Foundation of MOE Key Laboratory of Special Machine and High Voltage Apparatus,Grant/Award Number:KFKT202203。
文摘Furfural content in oil is considered as an effective parameter to evaluate the ageing state of oil-immersed transformers.In previous thermal ageing experiments at 130℃,we found that moisture increment in oil-paper system makes more furfural distribute in oil.However,the microcosmic mechanism of moisture effect on furfural diffusion from paper to oil is still unknown,and the furfural equilibrium law at transformers'typical operating temperatures is also unclear,resulting in the unreliability of the existed furfural evaluation model.To elucidate the above problems,molecular dynamics simulation and furfural equilibrium experiments at 60℃ were carried out at different moisture contents.The results show that high moisture content makes part of furfural diffuse from paper to oil indeed.Supposing moisture content increases from 1% to 2%,mass fraction of furfural in oil may increase from 32.98% to 49.83%,leading to furfural content in oil increases from 0.76 mg/L to 1.06 mg/L when the degree of polymerisation of insulation paper is about 524,which increased by 39.5%.The increase in moisture content makes cellulose form hydrogen bonds with H_(2)O preferentially,decreasing the hydrogen bonds interaction between cellulose and furfural.Thus,partial furfural falls off cellulose and diffuses into oil,leading to an increase in furfural content in oil.To eliminate the moisture interference,a corrected furfural assessment model based on the moisture equilibrium curve has been constructed,which can accurately evaluate the ageing state of insulation paper regardless of the moisture influence.
基金Project supported by the National Natural Science Foundation of China(No.62376280)。
文摘Reinforcement learning(RL)has become a dominant decision-making paradigm and has achieved notable success in many real-world applications.Notably,deep neural networks play a crucial role in unlocking RL’s potential in large-scale decision-making tasks.Inspired by current major success of Transformer in natural language processing and computer vision,numerous bottlenecks have been overcome by combining Transformer with RL for decision-making.This paper presents a multiangle systematic survey of various Transformer-based RL(TransRL)models applied in decision-making tasks,including basic models,advanced algorithms,representative implementation instances,typical applications,and known challenges.Our work aims to provide insights into problems that inherently arise with the current RL approaches,and examines how we can address them with better TransRL models.To our knowledge,we are the first to present a comprehensive review of the recent Transformer research developments in RL for decision-making.We hope that this survey provides a comprehensive review of TransRL models and inspires the RL community in its pursuit of future directions.To keep track of the rapid TransRL developments in the decision-making domains,we summarize the latest papers and their open-source implementations at https://github.com/williamyuanv0/Transformer-in-Reinforcement-Learning-for-Decision-Making-A-Survey.
基金This study is supported by the Projects of National Natural Science Foundation of China(Nos.72163016 and 71963017).
文摘This study assesses the effect of human capital expansion on China's export product quality.It employs the difference-in-differences(DID)framework based on a quasi-natural experiment investigating the 1999 higher education enrollment expansion as the exogenous policy shock.The empirical results confirm that human capital expansion appreciably improved the quality of China's export products.Human capital expansion promoted the transformation and upgrading of old products and the development of new products in term of intensive margin;it strengthened the endowment advantages of incumbent high-quality export enterprises while preventing low-quality enterprises from entering the market through price competition on the extensive margin.This prevented quality decline.Further,the study reveals that the improvement effect driven by human capital came from both innovation-induced and managerial efficiency improvement channels and was more prominent for large or foreign-funded enterprises in the eastern region.Our findings highlight the role of human capital in China's remarkable export performance from an endogenous growth perspective.
