Modeling high-dimensional aerodynamic data presents a significant challenge in aero-loads prediction, aerodynamic shape optimization, flight control, and simulation. This article develops a machine learning approach b...Modeling high-dimensional aerodynamic data presents a significant challenge in aero-loads prediction, aerodynamic shape optimization, flight control, and simulation. This article develops a machine learning approach based on a convolutional neural network (CNN) to address this problem. A CNN can implicitly distill features underlying the data. The number of parameters to be trained can be significantly reduced because of its local connectivity and parameter-sharing properties, which is favorable for solving high-dimensional problems in which the training cost can be prohibitive. A hypersonic wing similar to the Sanger aerospace plane carrier wing is employed as the test case to demonstrate the CNN-based modeling method. First, the wing is parameterized by the free-form deformation method, and 109 variables incorporating flight status and aerodynamic shape variables are defined as model input. Second, more than 7000 sample points generated by the Latin hypercube sampling method are evaluated by performing computational fluid dynamics simulations using a Reynolds-averaged Navier-Stokes flow solver to obtain an aerodynamic database, and a CNN model is built based on the observed data. Finally, the well-trained CNN model considering both flight status and shape variables is applied to aerodynamic shape optimization to demonstrate its capability to achieve fast optimization at multiple flight statuses.展开更多
The simulation techniques of hardware-in-loop simulation(HLS) of homing antitank missile based on the personal computer (PC) are discussed. The PC and MCS-96 chip controller employ A/D and D/A boards (with photoelectr...The simulation techniques of hardware-in-loop simulation(HLS) of homing antitank missile based on the personal computer (PC) are discussed. The PC and MCS-96 chip controller employ A/D and D/A boards (with photoelectricity isolation) to transfer measur ment and control information about homing head, gyro and rudder and utilize the digital hand shaking board to build correct communication communication protocol. In order to satisfy the real-time requirement of HLS, this paper first simplifies the aerodynamic data file reasonably, then builds a PC software with C language. The program of the controller part is made with PL/M language. The simulation of HLS based on PC is done with the same sampling period of 10ms as that of YH-F1 and the experiment results are identical to those of digital simulation of the homing anti-tank guided missile.展开更多
The fast-growing demand of computational fluid dynamics（CFD） application for computing resources stimulates the development of high performance computing（HPC） and meanwhile raises new requirements for the technolo...The fast-growing demand of computational fluid dynamics（CFD） application for computing resources stimulates the development of high performance computing（HPC） and meanwhile raises new requirements for the technology of parallel application performance monitor and analysis.In response to large-scale and long-time running for the application of CFD,online and scalable performance analysis technology is required to optimize the parallel programs as well as to improve their operational efficiency.As a result,this research implements a scalable infrastructure for online performance analysis on CFD application with homogeneous or heterogeneous system.The infrastructure is part of the parallel application performance monitor and analysis system（PAPMAS） and is composed of two modules which are scalable data transmission module and data storage module.The paper analyzes and elaborates this infrastructure in detail with respect to its design and implementation.Furthermore,some experiments are carried out to verify the rationality and high efficiency of this infrastructure that could be adopted to meet the practical needs.展开更多
基金National Numerical Wind Tunnel Project(grant No.NNW2019ZT6-A12)Science Fund for Distinguished Young Scholars of Shaanxi Province of China(grant No.2020JC-31)Natural Science Foundation of Shaanxi Province(grant No.2020JM-127).
文摘Modeling high-dimensional aerodynamic data presents a significant challenge in aero-loads prediction, aerodynamic shape optimization, flight control, and simulation. This article develops a machine learning approach based on a convolutional neural network (CNN) to address this problem. A CNN can implicitly distill features underlying the data. The number of parameters to be trained can be significantly reduced because of its local connectivity and parameter-sharing properties, which is favorable for solving high-dimensional problems in which the training cost can be prohibitive. A hypersonic wing similar to the Sanger aerospace plane carrier wing is employed as the test case to demonstrate the CNN-based modeling method. First, the wing is parameterized by the free-form deformation method, and 109 variables incorporating flight status and aerodynamic shape variables are defined as model input. Second, more than 7000 sample points generated by the Latin hypercube sampling method are evaluated by performing computational fluid dynamics simulations using a Reynolds-averaged Navier-Stokes flow solver to obtain an aerodynamic database, and a CNN model is built based on the observed data. Finally, the well-trained CNN model considering both flight status and shape variables is applied to aerodynamic shape optimization to demonstrate its capability to achieve fast optimization at multiple flight statuses.
文摘The simulation techniques of hardware-in-loop simulation(HLS) of homing antitank missile based on the personal computer (PC) are discussed. The PC and MCS-96 chip controller employ A/D and D/A boards (with photoelectricity isolation) to transfer measur ment and control information about homing head, gyro and rudder and utilize the digital hand shaking board to build correct communication communication protocol. In order to satisfy the real-time requirement of HLS, this paper first simplifies the aerodynamic data file reasonably, then builds a PC software with C language. The program of the controller part is made with PL/M language. The simulation of HLS based on PC is done with the same sampling period of 10ms as that of YH-F1 and the experiment results are identical to those of digital simulation of the homing anti-tank guided missile.
基金Aeronautical Science Foundation of China(2010ZA04001)National Natural Science Foundation of China (61073013,90818024)
文摘The fast-growing demand of computational fluid dynamics（CFD） application for computing resources stimulates the development of high performance computing（HPC） and meanwhile raises new requirements for the technology of parallel application performance monitor and analysis.In response to large-scale and long-time running for the application of CFD,online and scalable performance analysis technology is required to optimize the parallel programs as well as to improve their operational efficiency.As a result,this research implements a scalable infrastructure for online performance analysis on CFD application with homogeneous or heterogeneous system.The infrastructure is part of the parallel application performance monitor and analysis system（PAPMAS） and is composed of two modules which are scalable data transmission module and data storage module.The paper analyzes and elaborates this infrastructure in detail with respect to its design and implementation.Furthermore,some experiments are carried out to verify the rationality and high efficiency of this infrastructure that could be adopted to meet the practical needs.
