Abstract：Modern aircrafts, such as, missiles and rockets have a large slenderness ratio and their first several order natural frequencies are lower, so influences of elastic deformation and vibration on their trajectory simulation navigation, guidance and thrust module can’t be ignored. Here, for problems of elastic module’s slower calculation speed and being unable to realize real-time simulation in full trajectory simulation of slender elastic vehicles, through analyzing time proportion occupied of each calculation step under different computing scales, the dynamic parallel construction of octree was used to represent aerodynamic parameters table under the environment of single and multi GPU. The performance optimization of aerodynamic data index was achieved through adaptive hardware resources and rational use of shared memory. At the same time, the asynchronous computing architecture of CPU-side task queue was designed to realize the parallel task calculation with different granularities of CPU-GPU. The numerical results showed that the speedup of about 20 times under single GPU condition can be obtained, and the parallel computing of dual-GPU can obtain the speedup of at least 30 times; the real-time simulation of an elastic aircraft with 1 200 station points and 40 truncated orders is realized within 5 ms.
胡斌星1,李新国1,常武权2. 细长弹性飞行器飞行动力学并行计算及优化研究[J]. 振动与冲击, 2019, 38(11): 42-47.
HU Binxing1 LI Xinguo1 CHANG Wuquan2. Parallel computation and optimization for flight dynamics of slender elastic vehicles. JOURNAL OF VIBRATION AND SHOCK, 2019, 38(11): 42-47.
[1] 陈士橹. 近代飞行器飞行力学[M]. 西安: 西北工业大学出版社, 1991.
[2] 李惠峰, 肖进, 张冉. 高超声速飞行器刚体/弹性体耦合动力学建模[J]. 北京航空航天大学学报, 2012(02):160-165.
LI Huifeng, XIAO Jin, ZHANG Ran. Hypersonic vehicle rigid /elastic coupled dynamic modeling[J]. Journal of Beijing University of Aeronautics and Astronautics, 2012(02):160-165
[3] 赖剑奇, 李桦, 张冉等. 多GPU并行可压缩流求解器及其性能分析[J]. 航空学报, 2018:1-10.
LAI Jianqi, LI Hua, ZHANG Ran, et al. Multi-GPU parallel compressible flow solver and performance analysis[J]. Acta Aeronautica et Astronautica Sinica, 2018:1-10
[4] 韩志伟, 刘志刚, 鲁晓帆等. 基于CUDA的高速并行小波算法及其在电力系统谐波分析中的应用[J]. 电力自动化设备, 2010(01):98-101.
HAN Zhiwei, LIU Zhigang, LU Xiaofan, et al. High-speed parallel wavelet algorithm based on CUDA and its application in power system harmonic analysis[J]. Electric Power Automation Equipment, 2010(01):98-101
[5] Michael M. J. Quinn. MPI与OpenMP并行程序设计: C语言版[M]. 陈文光, 武永卫, 译. 北京: 清华大学出版社, 2004.
[6] BARTEZZAGHI A., CREMONESI M., PAROLINI N., et al. An explicit dynamics GPU structural solver for thin shell finite elements[J]. Computers and Structures, 2015,154:29-40.
[7] 李红豫, 滕军, 李祚华. 基于CPU-GPU异构平台的高层结构地震响应分析方法研究[J]. 振动与冲击, 2014(13):86-91.
LI Hongyu, TENG Jun, LI Zuohua. Analysis method for seismic response of high-rise structure based on CPU-GPU heterogeneous platform[J]. Journal of Vibration and Shock, 2014(13):86-91
[8] 李红豫, 滕军, 李祚华. 钢筋混凝土框架结构非线性静、动力分析的高效计算平台HSNAS(GPU)——Ⅰ程序开发[J]. 振动与冲击, 2016(14):47-53.
LI Hongyu, TENG Jun, LI Zuohua. An efficient platform HSNAS (GPU) for nonlinear static and dynamic analysis of reinforced concrete frames[J]. Journal of Vibration and Shock, 2016(14):47-53
[9] FLEISCHMANN Dominique, LONE Mudassir. Fast Computational Aeroelastic Analysis of Helicopter Rotor Blades[C]//: 2018 AIAA Aerospace Sciences Meeting, 2018.
[10] CHENG Tangpei. Accelerating universal Kriging interpolation algorithm using CUDA-enabled GPU[J]. Computers & Geosciences, 2013,54:178-183.
[11] 李涛, 董前琨, 张帅等. 基于线程池的GPU任务并行计算模式研究[J]. 计算机学报, 2017,40:1-19.
LI Tao, DONG Qiankun, ZHANG Shuai. GPU Task Parallel Computing Paradigm Based on Thread Pool Model[J]. Chinese Journal of Computers, 2017,40:1-19
[12] GAO Jiaquan, ZHOU Yuanshen, HE Guixia, et al. A multi-GPU parallel optimization model for the preconditioned conjugate gradient algorithm[J]. Parallel Computing, 2017,63:1-16.
[13] LIN Shaozhong, XIE Zhiqiang. A Jacobi_PCG solver for sparse linear systems on multi-GPU cluster[J]. The Journal of Supercomputing, 2017,73(1):433-454.
[14] HARRIS Mark. Optimizing parallel reduction in CUDA[R].2009.
[15] CHENG John, GROSSMAN Max, MCKERCHER Ty. Professional CUDA C Programming[M]. 10475 Crosspoint Boulevard Indianapolis,IN: John Wiley&Sons, 2014.
[16] 王倩. 基于Delaunay的三维快速克里金插值[D]. 电子科技大学, 2015.
[17] 苏统华, 李东, 李松泽等. CUDA并行程序设计——GPU编程指南[M]. 北京: 机械工业出版社, 2014.
[18] 张健飞, 沈德飞. 基于GPU的稀疏线性系统的预条件共轭梯度法[J]. 计算机应用, 2013(03):825-829.
ZHANG Jianfei, SHEN Defei. GPU-based preconditioned conjugate gradient method for solving sparse linear systems[J]. Journal of Computer Applications, 2013(03):825-829
[19] NVIDIA. CUDA C Programming Guide[S]. America: NVIDIA, 2017.
[20] NVIDIA. CUDA Occupancy Calculator[S]. America: 2016.
