Reinforcement Learning(RL)techniques are being studied to solve the Demand and Capacity Balancing(DCB)problems to fully exploit their computational performance.A locally gen-eralised Multi-Agent Reinforcement Learning...Reinforcement Learning(RL)techniques are being studied to solve the Demand and Capacity Balancing(DCB)problems to fully exploit their computational performance.A locally gen-eralised Multi-Agent Reinforcement Learning(MARL)for real-world DCB problems is proposed.The proposed method can deploy trained agents directly to unseen scenarios in a specific Air Traffic Flow Management(ATFM)region to quickly obtain a satisfactory solution.In this method,agents of all flights in a scenario form a multi-agent decision-making system based on partial observation.The trained agent with the customised neural network can be deployed directly on the corresponding flight,allowing it to solve the DCB problem jointly.A cooperation coefficient is introduced in the reward function,which is used to adjust the agent’s cooperation preference in a multi-agent system,thereby controlling the distribution of flight delay time allocation.A multi-iteration mechanism is designed for the DCB decision-making framework to deal with problems arising from non-stationarity in MARL and to ensure that all hotspots are eliminated.Experiments based on large-scale high-complexity real-world scenarios are conducted to verify the effectiveness and efficiency of the method.From a statis-tical point of view,it is proven that the proposed method is generalised within the scope of the flights and sectors of interest,and its optimisation performance outperforms the standard computer-assisted slot allocation and state-of-the-art RL-based DCB methods.The sensitivity analysis preliminarily reveals the effect of the cooperation coefficient on delay time allocation.展开更多
The study of changes in rocks due to interaction with hydrothermal fluids at active volcanoes provides insights into wall rock alteration associated with ore deposits formed in the geological past. Therefore, the natu...The study of changes in rocks due to interaction with hydrothermal fluids at active volcanoes provides insights into wall rock alteration associated with ore deposits formed in the geological past. Therefore, the nature of mineral alteration and chemical changes experienced by wall rocks can be investigated at eruptive sites on active volcanoes and the results used to better constrain ore-forming processes. In this study, we investigated the alteration at eruptive sites at Mount Cameroon volcano. These eruptive vents lie along NE-SW-trending fissures that define the Mount Cameroon rift. The vents are surrounded by cones composed largely of pyroclastic materials and to a lesser extent lava. Fumaroles (volcanic gases) rising through the vents during and after the 1999 eruption have resulted in the alteration of the pyroclastic robble along the fissures and the inner walls of the cones. Consequently, altered basaltic materials are covered with thin films of reddish, yellowish to whitish secondary minerals. These coatings result from an interaction between the surfaces of the basaltic glass with volcanically-derived acidic fluids. Petrographic investigations and XRD analysis of the basalts have identified primary mineral phases, such as olivine, pyroxene (mainly augite) and feldspars. Alteration products revealed include ubiquitous silica as well as gypsum, magnetite, feldspar, alunite and jarosite. Jarosite occurrence indicates that SO2 is the primary volcanically-derived acid source involved in coating formation. High contents of sulfur identified in the basalts indicate that melts at Mount Cameroon can be sulfur saturated as backed by previous melt inclusion data. Whole rock geochemical analysis shows a gain in silica in the altered samples and this ties with the mass balance calculations although minor gains of Al2O3, , MgO, MnO, CaO and K2O are shown by some samples.展开更多
In this paper, a parallel solution framework for the linear static analysis of large structures on PC clusters is presented. The framework consists of two main steps: data preparation and parallel solution. The parall...In this paper, a parallel solution framework for the linear static analysis of large structures on PC clusters is presented. The framework consists of two main steps: data preparation and parallel solution. The parallel solution is performed by a substructure based method with direct solvers. The aim of the data preparation step is to create the best possible substructures so that the parallel solution time is minimized. An actual structural model was solved utilizing both homogeneous and heterogeneous PC clusters to illustrate the performance and applicability of the presented framework.展开更多
With the dramatic development of spatial data in- frastructure, CyberGIS has become significant for geospatial data sharing. Due to the large number of concurrent users and large volume of vector data, CyberGIS faces ...With the dramatic development of spatial data in- frastructure, CyberGIS has become significant for geospatial data sharing. Due to the large number of concurrent users and large volume of vector data, CyberGIS faces a great chal- lenge in how to improve performance. The real-time visual- ization of vector maps is the most common function in Cyber- GIS applications, and it is time-consuming especially when the data volume becomes large. So, how to improve the effi- ciency of visualization of large vector maps is still a signif- icant research direction for GIScience scientists. In this re- search, we review the existing three optimization strategies, and determine that the third category strategy （i.e., parallel optimization） is appropriate for the real-time visualization of large vector maps. One of the key issues of parallel optimiza- tion is how to decompose the real-time visualization tasks into balanced sub tasks while taking into consideration the spatial heterogeneous characteristics. We put forward some rules that the decomposition should conform to, and design a real-time visualization framework for large vector maps. We focus on a balanced decomposition approach that can assure efficiency and effectiveness. Considering the spatial hetero- geneous characteristic of vector data, we use a ＂horizontal grid, vertical multistage＂ approach to construct a spatial point distribution information grid. The load balancer analyzes the spatial characteristics of the map requests and decomposes the real-time viewshed into multiple balanced sub viewsheds.Then, all the sub viewsheds are distributed to multiple server nodes to be executed in parallel, so as to improve the real- time visualization efficiency of large vector maps. A group of experiments have been conducted by us. The analysis results demonstrate that the approach proposed in this research has the ability of balanced decomposition, and it is efficient and effective for all geometry types of vector data.展开更多
基金co-funded by the National Natural Science Foundation of China(No.61903187)the National Key R&D Program of China(No.2021YFB1600500)+2 种基金the China Scholarship Council(No.202006830095)the Natural Science Foundation of Jiangsu Province(No.BK20190414)the Jiangsu Province Postgraduate Innovation Fund(No.KYCX20_0213).
文摘Reinforcement Learning(RL)techniques are being studied to solve the Demand and Capacity Balancing(DCB)problems to fully exploit their computational performance.A locally gen-eralised Multi-Agent Reinforcement Learning(MARL)for real-world DCB problems is proposed.The proposed method can deploy trained agents directly to unseen scenarios in a specific Air Traffic Flow Management(ATFM)region to quickly obtain a satisfactory solution.In this method,agents of all flights in a scenario form a multi-agent decision-making system based on partial observation.The trained agent with the customised neural network can be deployed directly on the corresponding flight,allowing it to solve the DCB problem jointly.A cooperation coefficient is introduced in the reward function,which is used to adjust the agent’s cooperation preference in a multi-agent system,thereby controlling the distribution of flight delay time allocation.A multi-iteration mechanism is designed for the DCB decision-making framework to deal with problems arising from non-stationarity in MARL and to ensure that all hotspots are eliminated.Experiments based on large-scale high-complexity real-world scenarios are conducted to verify the effectiveness and efficiency of the method.From a statis-tical point of view,it is proven that the proposed method is generalised within the scope of the flights and sectors of interest,and its optimisation performance outperforms the standard computer-assisted slot allocation and state-of-the-art RL-based DCB methods.The sensitivity analysis preliminarily reveals the effect of the cooperation coefficient on delay time allocation.
文摘The study of changes in rocks due to interaction with hydrothermal fluids at active volcanoes provides insights into wall rock alteration associated with ore deposits formed in the geological past. Therefore, the nature of mineral alteration and chemical changes experienced by wall rocks can be investigated at eruptive sites on active volcanoes and the results used to better constrain ore-forming processes. In this study, we investigated the alteration at eruptive sites at Mount Cameroon volcano. These eruptive vents lie along NE-SW-trending fissures that define the Mount Cameroon rift. The vents are surrounded by cones composed largely of pyroclastic materials and to a lesser extent lava. Fumaroles (volcanic gases) rising through the vents during and after the 1999 eruption have resulted in the alteration of the pyroclastic robble along the fissures and the inner walls of the cones. Consequently, altered basaltic materials are covered with thin films of reddish, yellowish to whitish secondary minerals. These coatings result from an interaction between the surfaces of the basaltic glass with volcanically-derived acidic fluids. Petrographic investigations and XRD analysis of the basalts have identified primary mineral phases, such as olivine, pyroxene (mainly augite) and feldspars. Alteration products revealed include ubiquitous silica as well as gypsum, magnetite, feldspar, alunite and jarosite. Jarosite occurrence indicates that SO2 is the primary volcanically-derived acid source involved in coating formation. High contents of sulfur identified in the basalts indicate that melts at Mount Cameroon can be sulfur saturated as backed by previous melt inclusion data. Whole rock geochemical analysis shows a gain in silica in the altered samples and this ties with the mass balance calculations although minor gains of Al2O3, , MgO, MnO, CaO and K2O are shown by some samples.
基金the Scientific Research Project Foundation of METU (No. BAP-2007-03-03-09)
文摘In this paper, a parallel solution framework for the linear static analysis of large structures on PC clusters is presented. The framework consists of two main steps: data preparation and parallel solution. The parallel solution is performed by a substructure based method with direct solvers. The aim of the data preparation step is to create the best possible substructures so that the parallel solution time is minimized. An actual structural model was solved utilizing both homogeneous and heterogeneous PC clusters to illustrate the performance and applicability of the presented framework.
文摘With the dramatic development of spatial data in- frastructure, CyberGIS has become significant for geospatial data sharing. Due to the large number of concurrent users and large volume of vector data, CyberGIS faces a great chal- lenge in how to improve performance. The real-time visual- ization of vector maps is the most common function in Cyber- GIS applications, and it is time-consuming especially when the data volume becomes large. So, how to improve the effi- ciency of visualization of large vector maps is still a signif- icant research direction for GIScience scientists. In this re- search, we review the existing three optimization strategies, and determine that the third category strategy （i.e., parallel optimization） is appropriate for the real-time visualization of large vector maps. One of the key issues of parallel optimiza- tion is how to decompose the real-time visualization tasks into balanced sub tasks while taking into consideration the spatial heterogeneous characteristics. We put forward some rules that the decomposition should conform to, and design a real-time visualization framework for large vector maps. We focus on a balanced decomposition approach that can assure efficiency and effectiveness. Considering the spatial hetero- geneous characteristic of vector data, we use a ＂horizontal grid, vertical multistage＂ approach to construct a spatial point distribution information grid. The load balancer analyzes the spatial characteristics of the map requests and decomposes the real-time viewshed into multiple balanced sub viewsheds.Then, all the sub viewsheds are distributed to multiple server nodes to be executed in parallel, so as to improve the real- time visualization efficiency of large vector maps. A group of experiments have been conducted by us. The analysis results demonstrate that the approach proposed in this research has the ability of balanced decomposition, and it is efficient and effective for all geometry types of vector data.
