In recent years,with the continuous development of deep learning and knowledge graph reasoning methods,more and more researchers have shown great interest in improving knowledge graph reasoning methods by inferring mi...In recent years,with the continuous development of deep learning and knowledge graph reasoning methods,more and more researchers have shown great interest in improving knowledge graph reasoning methods by inferring missing facts through reasoning.By searching paths on the knowledge graph and making fact and link predictions based on these paths,deep learning-based Reinforcement Learning(RL)agents can demonstrate good performance and interpretability.Therefore,deep reinforcement learning-based knowledge reasoning methods have rapidly emerged in recent years and have become a hot research topic.However,even in a small and fixed knowledge graph reasoning action space,there are still a large number of invalid actions.It often leads to the interruption of RL agents’wandering due to the selection of invalid actions,resulting in a significant decrease in the success rate of path mining.In order to improve the success rate of RL agents in the early stages of path search,this article proposes a knowledge reasoning method based on Deep Transfer Reinforcement Learning path(DTRLpath).Before supervised pre-training and retraining,a pre-task of searching for effective actions in a single step is added.The RL agent is first trained in the pre-task to improve its ability to search for effective actions.Then,the trained agent is transferred to the target reasoning task for path search training,which improves its success rate in searching for target task paths.Finally,based on the comparative experimental results on the FB15K-237 and NELL-995 datasets,it can be concluded that the proposed method significantly improves the success rate of path search and outperforms similar methods in most reasoning tasks.展开更多
Through the analysis of macro-and micro-plant remains,food residues and the rice-field like features from the mid-Neolithic site of Hanjing in the Huai River region,we propose an early beginning of rice cultivation at...Through the analysis of macro-and micro-plant remains,food residues and the rice-field like features from the mid-Neolithic site of Hanjing in the Huai River region,we propose an early beginning of rice cultivation at Hanjing.The presence of non-shattering rice spikelet bases and the increasing percentages of rice phytoliths confirm the appearance of domesticated rice in the Hanjing archaeobotanical assemblage.However,as indicated by the different prediction rates of rice domestication shown by morphometric of the double-peaked Oryza-type glum cells and fish-scale decorations on the Oryza-type bulliform cells from different cultural phases before 7,000 a BP,rice cultivation was at an early stage of development.Our findings provide new and significant evidence towards the establishment of the Huai River as another important center for early rice cultivation and domestication in prehistoric China.展开更多
The present study proposes a predictive model to explore the effect of partially filled porous media on the con-jugate heat transfer characteristic of phase change material(PCM)with interfacial coupling conditions bet...The present study proposes a predictive model to explore the effect of partially filled porous media on the con-jugate heat transfer characteristic of phase change material(PCM)with interfacial coupling conditions between pure fluid region and porous region.The enthalpy-porosity method,local thermal non-equilibrium model and Darcy-Forchheimer law are comprehensively considered to describe the convective heat transfer process in porous media.The modified model is then validated by benchmark data provided by particle image velocimetry(PIV)ex-periments.The phase change behavior,heat transfer efficiency and energy storage performance are numerically investigated for different partial porous filling strategies in terms of filling content,position,height of porous foam and inclination angles of cavity.The results indicate that due to the resistance in porous region,the shear stress exerted by the main vortex(natural convection)in pure fluid region and the momentum transferred,a secondary vortex phenomenon appears in the porous region near the fluid/porous interface.Moreover,such dis-continuity of permeability and fluid-to-porous thermal conductivity results in the cusp of phase change interface at the horizontal fluid/porous boundary.Among four partial porous filling cases,the lower porous filling one has more desirable heat transfer performance,and the 3/4H lower porous filling configuration is the best solution for optimization of the latent heat thermal energy storage(LHTES)systems.For tilted cavity,the increase of inclination angle positively affects the heat transfer efficiency as well as the energy storage rate of the LHTES system,where the performance of 3/4H lower porous filling configuration is further highlighted.展开更多
基金supported by Key Laboratory of Information System Requirement,No.LHZZ202202Natural Science Foundation of Xinjiang Uyghur Autonomous Region(2023D01C55)Scientific Research Program of the Higher Education Institution of Xinjiang(XJEDU2023P127).
文摘In recent years,with the continuous development of deep learning and knowledge graph reasoning methods,more and more researchers have shown great interest in improving knowledge graph reasoning methods by inferring missing facts through reasoning.By searching paths on the knowledge graph and making fact and link predictions based on these paths,deep learning-based Reinforcement Learning(RL)agents can demonstrate good performance and interpretability.Therefore,deep reinforcement learning-based knowledge reasoning methods have rapidly emerged in recent years and have become a hot research topic.However,even in a small and fixed knowledge graph reasoning action space,there are still a large number of invalid actions.It often leads to the interruption of RL agents’wandering due to the selection of invalid actions,resulting in a significant decrease in the success rate of path mining.In order to improve the success rate of RL agents in the early stages of path search,this article proposes a knowledge reasoning method based on Deep Transfer Reinforcement Learning path(DTRLpath).Before supervised pre-training and retraining,a pre-task of searching for effective actions in a single step is added.The RL agent is first trained in the pre-task to improve its ability to search for effective actions.Then,the trained agent is transferred to the target reasoning task for path search training,which improves its success rate in searching for target task paths.Finally,based on the comparative experimental results on the FB15K-237 and NELL-995 datasets,it can be concluded that the proposed method significantly improves the success rate of path search and outperforms similar methods in most reasoning tasks.
基金supported by the National Social Science Foundation of China (Grant No. 18CKG002)
文摘Through the analysis of macro-and micro-plant remains,food residues and the rice-field like features from the mid-Neolithic site of Hanjing in the Huai River region,we propose an early beginning of rice cultivation at Hanjing.The presence of non-shattering rice spikelet bases and the increasing percentages of rice phytoliths confirm the appearance of domesticated rice in the Hanjing archaeobotanical assemblage.However,as indicated by the different prediction rates of rice domestication shown by morphometric of the double-peaked Oryza-type glum cells and fish-scale decorations on the Oryza-type bulliform cells from different cultural phases before 7,000 a BP,rice cultivation was at an early stage of development.Our findings provide new and significant evidence towards the establishment of the Huai River as another important center for early rice cultivation and domestication in prehistoric China.
基金support from the National Natural Science Foundation of China(Grant No.:52006039)Natural Science Foundation of Guangdong Province(Grant No.:2022A1515010602)+1 种基金Guangzhou Science and Technology Plan Project(Grant No.:202201010575)Guangdong Provincial Key Laboratory of Distributed Energy Systems(Grant No.:2020B1212060075).
文摘The present study proposes a predictive model to explore the effect of partially filled porous media on the con-jugate heat transfer characteristic of phase change material(PCM)with interfacial coupling conditions between pure fluid region and porous region.The enthalpy-porosity method,local thermal non-equilibrium model and Darcy-Forchheimer law are comprehensively considered to describe the convective heat transfer process in porous media.The modified model is then validated by benchmark data provided by particle image velocimetry(PIV)ex-periments.The phase change behavior,heat transfer efficiency and energy storage performance are numerically investigated for different partial porous filling strategies in terms of filling content,position,height of porous foam and inclination angles of cavity.The results indicate that due to the resistance in porous region,the shear stress exerted by the main vortex(natural convection)in pure fluid region and the momentum transferred,a secondary vortex phenomenon appears in the porous region near the fluid/porous interface.Moreover,such dis-continuity of permeability and fluid-to-porous thermal conductivity results in the cusp of phase change interface at the horizontal fluid/porous boundary.Among four partial porous filling cases,the lower porous filling one has more desirable heat transfer performance,and the 3/4H lower porous filling configuration is the best solution for optimization of the latent heat thermal energy storage(LHTES)systems.For tilted cavity,the increase of inclination angle positively affects the heat transfer efficiency as well as the energy storage rate of the LHTES system,where the performance of 3/4H lower porous filling configuration is further highlighted.