Federated learning(FL)is a distributed machine learning paradigm for edge cloud computing.FL can facilitate data-driven decision-making in tactical scenarios,effectively addressing both data volume and infrastructure ...Federated learning(FL)is a distributed machine learning paradigm for edge cloud computing.FL can facilitate data-driven decision-making in tactical scenarios,effectively addressing both data volume and infrastructure challenges in edge environments.However,the diversity of clients in edge cloud computing presents significant challenges for FL.Personalized federated learning(pFL)received considerable attention in recent years.One example of pFL involves exploiting the global and local information in the local model.Current pFL algorithms experience limitations such as slow convergence speed,catastrophic forgetting,and poor performance in complex tasks,which still have significant shortcomings compared to the centralized learning.To achieve high pFL performance,we propose FedCLCC:Federated Contrastive Learning and Conditional Computing.The core of FedCLCC is the use of contrastive learning and conditional computing.Contrastive learning determines the feature representation similarity to adjust the local model.Conditional computing separates the global and local information and feeds it to their corresponding heads for global and local handling.Our comprehensive experiments demonstrate that FedCLCC outperforms other state-of-the-art FL algorithms.展开更多
[Objective]Real-time monitoring of cow ruminant behavior is of paramount importance for promptly obtaining relevant information about cow health and predicting cow diseases.Currently,various strategies have been propo...[Objective]Real-time monitoring of cow ruminant behavior is of paramount importance for promptly obtaining relevant information about cow health and predicting cow diseases.Currently,various strategies have been proposed for monitoring cow ruminant behavior,including video surveillance,sound recognition,and sensor monitoring methods.How‐ever,the application of edge device gives rise to the issue of inadequate real-time performance.To reduce the volume of data transmission and cloud computing workload while achieving real-time monitoring of dairy cow rumination behavior,a real-time monitoring method was proposed for cow ruminant behavior based on edge computing.[Methods]Autono‐mously designed edge devices were utilized to collect and process six-axis acceleration signals from cows in real-time.Based on these six-axis data,two distinct strategies,federated edge intelligence and split edge intelligence,were investigat‐ed for the real-time recognition of cow ruminant behavior.Focused on the real-time recognition method for cow ruminant behavior leveraging federated edge intelligence,the CA-MobileNet v3 network was proposed by enhancing the MobileNet v3 network with a collaborative attention mechanism.Additionally,a federated edge intelligence model was designed uti‐lizing the CA-MobileNet v3 network and the FedAvg federated aggregation algorithm.In the study on split edge intelli‐gence,a split edge intelligence model named MobileNet-LSTM was designed by integrating the MobileNet v3 network with a fusion collaborative attention mechanism and the Bi-LSTM network.[Results and Discussions]Through compara‐tive experiments with MobileNet v3 and MobileNet-LSTM,the federated edge intelligence model based on CA-Mo‐bileNet v3 achieved an average Precision rate,Recall rate,F1-Score,Specificity,and Accuracy of 97.1%,97.9%,97.5%,98.3%,and 98.2%,respectively,yielding the best recognition performance.[Conclusions]It is provided a real-time and effective method for monitoring cow ruminant behavior,and the proposed federated edge intelligence model can be ap‐plied in practical settings.展开更多
A novel approach to compute the high frequency radar cross-section (RCS) of complex targets is described in this paper.From the three views or the sectional views of the target, target is geometrically modeled by non-...A novel approach to compute the high frequency radar cross-section (RCS) of complex targets is described in this paper.From the three views or the sectional views of the target, target is geometrically modeled by non-uniform rational B-spline (NURBS) parametric surfaces using the software CNFEOV developed by oneself which constructs NURBS representation of complex target from engineering orthographic views. RCS is obtained through PO, PTD, MEC and IBC techniques. When calculating RCS of the target, it is necessary to get the unit normal vector to surface illumi- nated by radar and the value Z which is the distance from the point on the surface to radar. ln this novel approach, the unit normal vector to the surface can be obtained either by the Phong rendering model, in which the color components (RGB) of every pixel on the image are equal to the coordinate components of the normal, or by the NURBS expressions. The value Z can be achieved by software or hardware Z-buffer. The effects of the size of image on the RCS of target are discussed and the correct method is recommended. The RCS of the perfect conducting sphere, cylinder and dihedral as well as the coated cylinder, as some examples, are computed. The accuracy of the method is verified by comparing the numerical results with those obtained by using other methods.展开更多
Several computer packages have been developed to accomplish improved programs for animal breeding and genetic selection. This paper described most of the currant software and provided suggestions for improved software...Several computer packages have been developed to accomplish improved programs for animal breeding and genetic selection. This paper described most of the currant software and provided suggestions for improved software. Khon Kaen University, Thailand, will provide free of charge the new software developed at Khon Kaen University by the author of this paper. The contact for requesting the software is listed: monchai@kku.ac.th.展开更多
Artificial bone with porous structure is crucial for tissue scaffold and clinic implants.Scaffold provides structure support for cells and guides tissues regeneration for final tissue structure.A computational aided p...Artificial bone with porous structure is crucial for tissue scaffold and clinic implants.Scaffold provides structure support for cells and guides tissues regeneration for final tissue structure.A computational aided process of porous bone modeling was developed which described the design and fabrication of tissue scaffolds by considering intricate architecture,porosity and pore size.To simulate intricate bone structure,different constructive units were presented.In modeling process,bone contour was gotten from computed tomography(CT)images and was divided into two levels.Each level was represented by relatively reconstructive process.Pore size distribution was controlled by using mesh generation.The whole hexahedral mesh was reduced by unit structure,when a 3D mesh with various hexahedral elements was provided.The simulation results show that constructive structure of porous scaffold can meet the needs of clinic implants in accurate and controlled way.展开更多
Ubiquitous computing must incorporate a certain level of security. For the severely resource constrained applications, the energy-efficient and small size cryptography algorithm implementation is a critical problem. H...Ubiquitous computing must incorporate a certain level of security. For the severely resource constrained applications, the energy-efficient and small size cryptography algorithm implementation is a critical problem. Hardware implementations of the advanced encryption standard （AES） for authentication and encryption are presented. An energy consumption variable is derived to evaluate low-power design strategies for battery-powered devices. It proves that compact AES architectures fail to optimize the AES hardware energy, whereas reducing invalid switching activities and implementing power-optimized sub-modules are the reasonable methods. Implementations of different substitution box （S-Boxes） structures are presented with 0.25μm 1.8 V CMOS （complementary metal oxide semiconductor） standard cell library. The comparisons and trade-offs among area, security, and power are explored. The experimental results show that Galois field composite S-Boxes have smaller size and highest security but consume considerably more power, whereas decoder-switch-encoder S-Boxes have the best power characteristics with disadvantages in terms of size and security. The combination of these two type S-Boxes instead of homogeneous S-Boxes in AES circuit will lead to optimal schemes. The technique of latch-dividing data path is analyzed, and the quantitative simulation results demonstrate that this approach diminishes the glitches effectively at a very low hardware cost.展开更多
Multiple QoS modeling and algorithm in grid system is considered. Grid QoS requirements can be formulated as a utility function for each task as a weighted sum of its each dimensional QoS utility functions. Multiple Q...Multiple QoS modeling and algorithm in grid system is considered. Grid QoS requirements can be formulated as a utility function for each task as a weighted sum of its each dimensional QoS utility functions. Multiple QoS constraint resource scheduling optimization in computational grid is distributed to two subproblems： optimization of grid user and grid resource provider. Grid QoS scheduling can be achieved by solving sub problems via an iterative algorithm.展开更多
The on line computational burden related to model predictive control (MPC) of large scale constrained systems hampers its real time applications and limits it to slow dynamic process with moderate number of inputs....The on line computational burden related to model predictive control (MPC) of large scale constrained systems hampers its real time applications and limits it to slow dynamic process with moderate number of inputs. To avoid this, an efficient and fast algorithm based on aggregation optimization is proposed in this paper. It only optimizes the current control action at time instant k , while other future control sequences in the optimization horizon are approximated off line by the linear feedback control sequence, so the on line optimization can be converted into a low dimensional quadratic programming problem. Input constraints can be well handled in this scheme. The comparable performance is achieved with existing standard model predictive control algorithm. Simulation results well demonstrate its effectiveness.展开更多
A computer model has been developed to simulate the distribution behaviors of Ni, Co. Sn. Ph,Zn, As, Sb, Bi, An and Ag in copper smelting process. The model assumes that the copper smelting furnaceis in thermodynamic ...A computer model has been developed to simulate the distribution behaviors of Ni, Co. Sn. Ph,Zn, As, Sb, Bi, An and Ag in copper smelting process. The model assumes that the copper smelting furnaceis in thermodynamic equilibrium. As many as 21 elements (Cu. S, Fe. Ni, Co. Sn, As, Sb. Bi, Ph. Zn.An. Ag. O, N, C, H, Ca, Mg, Al, and St) and 73 compounds are considered. This model accounts forphysical entrainment in the melts. The predictions by the present computer model are compared with theknown commercial data from Guixi Smelter in China, Home Smelter in Canada and Naoshima Smelter inJapan. The agreements between the computer predictions and the commercial data are excellent, so that thepresent computer model can be used to monitor and optimize the actual industrial operations of copper smelting. It is applicable to simulation of almost all copper pyrometallurgical processes.展开更多
Milling Process Simulation is one of the important re search areas in manufacturing science. For the purpose of improving the prec ision of simulation and extending its usability, numerical algorithm is more and more ...Milling Process Simulation is one of the important re search areas in manufacturing science. For the purpose of improving the prec ision of simulation and extending its usability, numerical algorithm is more and more used in the milling modeling areas. But simulative efficiency is decreasin g with increase of its complexity. As a result, application of the method is lim ited. Aimed at above question, high-efficient algorithm for milling process sim ulation is studied. It is important for milling process simulation’s applicatio n. Parallel computing is widely used to solve the large-scale computation question s. Its advantages include system flexibility, robust, high-efficient computing capability and high ratio of performance to price. With the development of compu ter network, utilizing the computing resource in the Internet, a virtual computi ng environment with powerful computing capability can be consisted by microc omputers, and the difficulty of building hardware environment which is used to s upport parallel computing is reduced. How to use network technology and parallel algorithm to improve simulative effic iency for milling forces simulation is investigated in the paper. In order to pr edict milling forces, a simplified local milling forces model is used in the pap er. End milling cutter is assumed to be divided by r number of differential elem ents along the axial direction of the cutter. For a given time, the total cuttin g forces can be obtained by summarizing the resultant cutting force produced by each differential cutter disc. Divide the whole simulative time into some segmen ts, send these program’s segments to microcomputers in the Internet and obtain the result of the program’s segments, all of the result of program’s segments a re composed the final result. For implementing the algorithm, a distributed Parallel computing framework is de signed in the paper. In the framework, web server plays a role of controller. Us ing Java RMI(remote method interface), the computing processes in computing serv er are called by web server. There are lots of control processes in web server a nd control the computing servers. The codes of simulative algorithm can be dynam ic sent to the computing servers, and milling forces at the different time are c omputed through utilizing the local computer’s resource. The results that are ca lculated by every computing servers are sent to the web server, and composed the final result. The framework can be used by different simulative algorithm. Comp ared with the algorithm running single machine, the efficiency of provided algor ithm is higher than that of single machine.展开更多
A deep understanding of the internal ballistic process and the factors affecting it is of primary importance to efficiently design a gun system and ensure its safe management. One of the main goals of internal ballist...A deep understanding of the internal ballistic process and the factors affecting it is of primary importance to efficiently design a gun system and ensure its safe management. One of the main goals of internal ballistics is to estimate the gas pressure into the combustion chamber and the projectile muzzle velocity in order to use the propellant to its higher efficiency while avoiding over-pressure phenomena. Dealing with the internal ballistic problem is a complex undertaking since it requires handling the interaction between different constituents during a transient time lapse with very steep rise of pressure and temperature. Several approaches have been proposed in the literature, based on different assumptions and techniques. Generally, depending on the used mathematical framework, they can be classified into two categories: computational fluid dynamics-based models and lumped-parameter ones. By focusing on gun systems, this paper offers a review of the main contributions in the field by mentioning their advantages and drawbacks. An insight into the limitations of the currently available modelling strategies is provided,as well as some considerations on the choice of one model over another. Lumped-parameter models, for example, are a good candidate for performing parametric analysis and optimisation processes of gun systems, given their minimum requirements of computer resources. Conversely, CFD-based models have a better capacity to address more sophisticated phenomena like pressure waves and turbulent flow effects. The performed review also reveals that too little attention has been given to small calibre guns since the majority of currently available models are conceived for medium and large calibre gun systems.Similarly, aspects like wear phenomena, bore deformations or projectile-barrel interactions still need to be adequately addressed and our suggestion is to dedicate more effort on it.展开更多
针对动车组故障预测与健康管理(Prognostics and Health Management,PHM)实时海量数据解析处理与模型计算问题,提出一种基于流计算的动车组PHM模型处理框架。首先分析动车组车载数据处理流程,然后基于Spark Streaming给出动车组PHM模型...针对动车组故障预测与健康管理(Prognostics and Health Management,PHM)实时海量数据解析处理与模型计算问题,提出一种基于流计算的动车组PHM模型处理框架。首先分析动车组车载数据处理流程,然后基于Spark Streaming给出动车组PHM模型处理的总体框架。针对实时海量数据解析处理,首先分析解析前的车载数据结构,定义解析后的车载数据结构,然后设计通用化数据解析组件,给出流计算实现方式。针对模型计算,详细给出PHM模型的形式化定义,包括模型的基本信息、输入、输出和逻辑主体等,根据此定义设计模型通用组件,实现模型的快速研发、高效计算和统一应用。通过动车组PHM系统的有效应用,证明了该框架可以很好地满足海量数据的实时计算需求。展开更多
基金supported by the Natural Science Foundation of Xinjiang Uygur Autonomous Region(Grant No.2022D01B 187)。
文摘Federated learning(FL)is a distributed machine learning paradigm for edge cloud computing.FL can facilitate data-driven decision-making in tactical scenarios,effectively addressing both data volume and infrastructure challenges in edge environments.However,the diversity of clients in edge cloud computing presents significant challenges for FL.Personalized federated learning(pFL)received considerable attention in recent years.One example of pFL involves exploiting the global and local information in the local model.Current pFL algorithms experience limitations such as slow convergence speed,catastrophic forgetting,and poor performance in complex tasks,which still have significant shortcomings compared to the centralized learning.To achieve high pFL performance,we propose FedCLCC:Federated Contrastive Learning and Conditional Computing.The core of FedCLCC is the use of contrastive learning and conditional computing.Contrastive learning determines the feature representation similarity to adjust the local model.Conditional computing separates the global and local information and feeds it to their corresponding heads for global and local handling.Our comprehensive experiments demonstrate that FedCLCC outperforms other state-of-the-art FL algorithms.
文摘[Objective]Real-time monitoring of cow ruminant behavior is of paramount importance for promptly obtaining relevant information about cow health and predicting cow diseases.Currently,various strategies have been proposed for monitoring cow ruminant behavior,including video surveillance,sound recognition,and sensor monitoring methods.How‐ever,the application of edge device gives rise to the issue of inadequate real-time performance.To reduce the volume of data transmission and cloud computing workload while achieving real-time monitoring of dairy cow rumination behavior,a real-time monitoring method was proposed for cow ruminant behavior based on edge computing.[Methods]Autono‐mously designed edge devices were utilized to collect and process six-axis acceleration signals from cows in real-time.Based on these six-axis data,two distinct strategies,federated edge intelligence and split edge intelligence,were investigat‐ed for the real-time recognition of cow ruminant behavior.Focused on the real-time recognition method for cow ruminant behavior leveraging federated edge intelligence,the CA-MobileNet v3 network was proposed by enhancing the MobileNet v3 network with a collaborative attention mechanism.Additionally,a federated edge intelligence model was designed uti‐lizing the CA-MobileNet v3 network and the FedAvg federated aggregation algorithm.In the study on split edge intelli‐gence,a split edge intelligence model named MobileNet-LSTM was designed by integrating the MobileNet v3 network with a fusion collaborative attention mechanism and the Bi-LSTM network.[Results and Discussions]Through compara‐tive experiments with MobileNet v3 and MobileNet-LSTM,the federated edge intelligence model based on CA-Mo‐bileNet v3 achieved an average Precision rate,Recall rate,F1-Score,Specificity,and Accuracy of 97.1%,97.9%,97.5%,98.3%,and 98.2%,respectively,yielding the best recognition performance.[Conclusions]It is provided a real-time and effective method for monitoring cow ruminant behavior,and the proposed federated edge intelligence model can be ap‐plied in practical settings.
文摘A novel approach to compute the high frequency radar cross-section (RCS) of complex targets is described in this paper.From the three views or the sectional views of the target, target is geometrically modeled by non-uniform rational B-spline (NURBS) parametric surfaces using the software CNFEOV developed by oneself which constructs NURBS representation of complex target from engineering orthographic views. RCS is obtained through PO, PTD, MEC and IBC techniques. When calculating RCS of the target, it is necessary to get the unit normal vector to surface illumi- nated by radar and the value Z which is the distance from the point on the surface to radar. ln this novel approach, the unit normal vector to the surface can be obtained either by the Phong rendering model, in which the color components (RGB) of every pixel on the image are equal to the coordinate components of the normal, or by the NURBS expressions. The value Z can be achieved by software or hardware Z-buffer. The effects of the size of image on the RCS of target are discussed and the correct method is recommended. The RCS of the perfect conducting sphere, cylinder and dihedral as well as the coated cylinder, as some examples, are computed. The accuracy of the method is verified by comparing the numerical results with those obtained by using other methods.
文摘Several computer packages have been developed to accomplish improved programs for animal breeding and genetic selection. This paper described most of the currant software and provided suggestions for improved software. Khon Kaen University, Thailand, will provide free of charge the new software developed at Khon Kaen University by the author of this paper. The contact for requesting the software is listed: monchai@kku.ac.th.
基金Project(2011DFB70230)supported by State International Cooperation Program of ChinaProject(N110403003)supported by Basic Research Foundation of Education Ministry of China
文摘Artificial bone with porous structure is crucial for tissue scaffold and clinic implants.Scaffold provides structure support for cells and guides tissues regeneration for final tissue structure.A computational aided process of porous bone modeling was developed which described the design and fabrication of tissue scaffolds by considering intricate architecture,porosity and pore size.To simulate intricate bone structure,different constructive units were presented.In modeling process,bone contour was gotten from computed tomography(CT)images and was divided into two levels.Each level was represented by relatively reconstructive process.Pore size distribution was controlled by using mesh generation.The whole hexahedral mesh was reduced by unit structure,when a 3D mesh with various hexahedral elements was provided.The simulation results show that constructive structure of porous scaffold can meet the needs of clinic implants in accurate and controlled way.
基金the "863" High Technology Research and Development Program of China (2006AA01Z226)the Scientific Research Foundation of Huazhong University of Science and Technology (2006Z011B)the Program for New Century Excellent Talents in University (NCET-07-0328).
文摘Ubiquitous computing must incorporate a certain level of security. For the severely resource constrained applications, the energy-efficient and small size cryptography algorithm implementation is a critical problem. Hardware implementations of the advanced encryption standard （AES） for authentication and encryption are presented. An energy consumption variable is derived to evaluate low-power design strategies for battery-powered devices. It proves that compact AES architectures fail to optimize the AES hardware energy, whereas reducing invalid switching activities and implementing power-optimized sub-modules are the reasonable methods. Implementations of different substitution box （S-Boxes） structures are presented with 0.25μm 1.8 V CMOS （complementary metal oxide semiconductor） standard cell library. The comparisons and trade-offs among area, security, and power are explored. The experimental results show that Galois field composite S-Boxes have smaller size and highest security but consume considerably more power, whereas decoder-switch-encoder S-Boxes have the best power characteristics with disadvantages in terms of size and security. The combination of these two type S-Boxes instead of homogeneous S-Boxes in AES circuit will lead to optimal schemes. The technique of latch-dividing data path is analyzed, and the quantitative simulation results demonstrate that this approach diminishes the glitches effectively at a very low hardware cost.
基金the National Natural Science Foundation of China (60402028, 60672137) Wuhan Yonger Dawning Foundation (20045006071-15)China Specialized Research Fund for the Doctoral Program of Higher Eduction (20060497015).
文摘Multiple QoS modeling and algorithm in grid system is considered. Grid QoS requirements can be formulated as a utility function for each task as a weighted sum of its each dimensional QoS utility functions. Multiple QoS constraint resource scheduling optimization in computational grid is distributed to two subproblems： optimization of grid user and grid resource provider. Grid QoS scheduling can be achieved by solving sub problems via an iterative algorithm.
文摘The on line computational burden related to model predictive control (MPC) of large scale constrained systems hampers its real time applications and limits it to slow dynamic process with moderate number of inputs. To avoid this, an efficient and fast algorithm based on aggregation optimization is proposed in this paper. It only optimizes the current control action at time instant k , while other future control sequences in the optimization horizon are approximated off line by the linear feedback control sequence, so the on line optimization can be converted into a low dimensional quadratic programming problem. Input constraints can be well handled in this scheme. The comparable performance is achieved with existing standard model predictive control algorithm. Simulation results well demonstrate its effectiveness.
文摘A computer model has been developed to simulate the distribution behaviors of Ni, Co. Sn. Ph,Zn, As, Sb, Bi, An and Ag in copper smelting process. The model assumes that the copper smelting furnaceis in thermodynamic equilibrium. As many as 21 elements (Cu. S, Fe. Ni, Co. Sn, As, Sb. Bi, Ph. Zn.An. Ag. O, N, C, H, Ca, Mg, Al, and St) and 73 compounds are considered. This model accounts forphysical entrainment in the melts. The predictions by the present computer model are compared with theknown commercial data from Guixi Smelter in China, Home Smelter in Canada and Naoshima Smelter inJapan. The agreements between the computer predictions and the commercial data are excellent, so that thepresent computer model can be used to monitor and optimize the actual industrial operations of copper smelting. It is applicable to simulation of almost all copper pyrometallurgical processes.
文摘Milling Process Simulation is one of the important re search areas in manufacturing science. For the purpose of improving the prec ision of simulation and extending its usability, numerical algorithm is more and more used in the milling modeling areas. But simulative efficiency is decreasin g with increase of its complexity. As a result, application of the method is lim ited. Aimed at above question, high-efficient algorithm for milling process sim ulation is studied. It is important for milling process simulation’s applicatio n. Parallel computing is widely used to solve the large-scale computation question s. Its advantages include system flexibility, robust, high-efficient computing capability and high ratio of performance to price. With the development of compu ter network, utilizing the computing resource in the Internet, a virtual computi ng environment with powerful computing capability can be consisted by microc omputers, and the difficulty of building hardware environment which is used to s upport parallel computing is reduced. How to use network technology and parallel algorithm to improve simulative effic iency for milling forces simulation is investigated in the paper. In order to pr edict milling forces, a simplified local milling forces model is used in the pap er. End milling cutter is assumed to be divided by r number of differential elem ents along the axial direction of the cutter. For a given time, the total cuttin g forces can be obtained by summarizing the resultant cutting force produced by each differential cutter disc. Divide the whole simulative time into some segmen ts, send these program’s segments to microcomputers in the Internet and obtain the result of the program’s segments, all of the result of program’s segments a re composed the final result. For implementing the algorithm, a distributed Parallel computing framework is de signed in the paper. In the framework, web server plays a role of controller. Us ing Java RMI(remote method interface), the computing processes in computing serv er are called by web server. There are lots of control processes in web server a nd control the computing servers. The codes of simulative algorithm can be dynam ic sent to the computing servers, and milling forces at the different time are c omputed through utilizing the local computer’s resource. The results that are ca lculated by every computing servers are sent to the web server, and composed the final result. The framework can be used by different simulative algorithm. Comp ared with the algorithm running single machine, the efficiency of provided algor ithm is higher than that of single machine.
基金the support provided by the Royal Higher Institute for Defence (RHID) of the Belgian Defence, which has contributed to the progress of this ongoing research.
文摘A deep understanding of the internal ballistic process and the factors affecting it is of primary importance to efficiently design a gun system and ensure its safe management. One of the main goals of internal ballistics is to estimate the gas pressure into the combustion chamber and the projectile muzzle velocity in order to use the propellant to its higher efficiency while avoiding over-pressure phenomena. Dealing with the internal ballistic problem is a complex undertaking since it requires handling the interaction between different constituents during a transient time lapse with very steep rise of pressure and temperature. Several approaches have been proposed in the literature, based on different assumptions and techniques. Generally, depending on the used mathematical framework, they can be classified into two categories: computational fluid dynamics-based models and lumped-parameter ones. By focusing on gun systems, this paper offers a review of the main contributions in the field by mentioning their advantages and drawbacks. An insight into the limitations of the currently available modelling strategies is provided,as well as some considerations on the choice of one model over another. Lumped-parameter models, for example, are a good candidate for performing parametric analysis and optimisation processes of gun systems, given their minimum requirements of computer resources. Conversely, CFD-based models have a better capacity to address more sophisticated phenomena like pressure waves and turbulent flow effects. The performed review also reveals that too little attention has been given to small calibre guns since the majority of currently available models are conceived for medium and large calibre gun systems.Similarly, aspects like wear phenomena, bore deformations or projectile-barrel interactions still need to be adequately addressed and our suggestion is to dedicate more effort on it.
文摘针对动车组故障预测与健康管理(Prognostics and Health Management,PHM)实时海量数据解析处理与模型计算问题,提出一种基于流计算的动车组PHM模型处理框架。首先分析动车组车载数据处理流程,然后基于Spark Streaming给出动车组PHM模型处理的总体框架。针对实时海量数据解析处理,首先分析解析前的车载数据结构,定义解析后的车载数据结构,然后设计通用化数据解析组件,给出流计算实现方式。针对模型计算,详细给出PHM模型的形式化定义,包括模型的基本信息、输入、输出和逻辑主体等,根据此定义设计模型通用组件,实现模型的快速研发、高效计算和统一应用。通过动车组PHM系统的有效应用,证明了该框架可以很好地满足海量数据的实时计算需求。
