GPU并行计算驱动的大规模无人机集群仿真研究

doi:10.3778/j.issn.1002-8331.2505-0315

摘要/Abstract

摘要： 随着无人机集群仿真规模的不断扩大，传统基于CPU的串行计算方式已难以满足大规模仿真高效率、高实时性和低成本的实际需求。提出一种GPU并行计算驱动的大规模无人机集群仿真方法，从方法设计层面出发，系统构建契合GPU执行特性的并行仿真机制。该方法以CUDA为并行计算基础，结合仿真对象的独立性特征，采用细粒度的数据并行策略，依据静态负载均衡原则与线程映射模型，将仿真任务均匀分配至并行线程，实现计算资源的高效利用。为验证所提出方法的有效性，构建面向大规模集群的并行仿真系统，并与传统CPU仿真方法进行性能对比测试，从仿真实时性与系统扩展性两个维度深入开展性能评估。实验结果表明，该方法在保障仿真精度和系统稳定性的同时，能实现在NVIDIA GeForce GTX 1650显卡上至多2 322架无人机模型的实时仿真，不仅为无人机集群的高效仿真提供了可行路径，也为其他领域的大规模并行计算问题提供了技术参考。

关键词: GPU计算, 无人机集群, 并行仿真, 实时性

Abstract: With the continuous expansion of unmanned aerial vehicle (UAV) swarm simulation scales, traditional CPU-based serial computing methods have become increasingly inadequate in meeting the high-efficiency, real-time, and low-cost requirements of large-scale simulations. To address this challenge, this paper proposes a GPU-parallel-computing-driven method for large-scale UAV swarm simulation. From a methodological perspective, a parallel simulation framework compatible with GPU execution characteristics is systematically developed. Leveraging CUDA as the foundation for parallel computation, the method exploits the inherent independence of simulation objects by adopting a fine-grained data-parallel strategy. Simulation tasks are evenly distributed across parallel threads based on a static load-balancing principle and thread mapping model, thereby maximizing computational resource utilization. To validate the effectiveness of the proposed method, a parallel simulation system for large-scale UAV swarms is constructed, and performance benchmarking against conventional CPU-based simulation approaches is conducted. The evaluation is systematically performed from two critical dimensions： simulation real-time performance and system scalability. Experimental results demonstrate that while ensuring simulation accuracy and system stability, this method can achieve real-time simulation of up to 2?322 drone models on an NVIDIA GeForce GTX 1650 graphics card. This work provides a practical solution for efficient UAV swarm simulation and offers valuable technical insights for addressing large-scale parallel computing challenges in other domains.

Key words: GPU computing, UAV swarm, parallel simulation, real-time capability

余玲, 郭川东, 王国庆, 刘艳菊, 曹立佳. GPU并行计算驱动的大规模无人机集群仿真研究[J]. 计算机工程与应用, 2025, 61(22): 114-122.

YU Ling, GUO Chuandong, WANG Guoqing, LIU Yanju, CAO Lijia. Research on Large-Scale Unmanned Aerial Vehicle Swarms Simulation Driven by GPU Parallel Computing[J]. Computer Engineering and Applications, 2025, 61(22): 114-122.

参考文献

[1] 刘祖均, 何明, 马子玉, 等. 基于分布式一致性的无人机编队控制方法[J]. 计算机工程与应用, 2020, 56(23): 146-152.
LIU Z J, HE M, MA Z Y, et al. UAV formation control method based on distributed consistency[J]. Computer Engineering and Applications, 2020, 56(23): 146-152.
[2] 邹立岩, 张明智, 柏俊汝, 等. 无人机集群作战建模与仿真研究综述[J]. 战术导弹技术, 2021(3): 98-108.
ZOU L Y, ZHANG M Z, BAI J R, et al. A survey of modeling and simulation of UAS swarm operation[J]. Tactical Missile Technology, 2021(3): 98-108.
[3] BURRAGE K. Parallel methods for initial value problems[J]. Applied Numerical Mathematics, 1993, 11(1/2/3): 5-25.
[4] 李文光, 王强, 曹严. 基于GPU的大规模无人机编队控制并行仿真方法[J]. 弹箭与制导学报, 2019, 39(4): 118-122.
LI W G, WANG Q, CAO Y. Parallel simulation method for multi-UAV formation with GPU[J]. Journal of Projectiles, Rockets, Missiles and Guidance, 2019, 39(4): 118-122.
[5] LING H F, LUO H C, CHEN H S, et al. Modelling and simulation of distributed UAV swarm cooperative planning and perception[J]. International Journal of Aerospace Engineering, 2021(1): 9977262.
[6] GARCíA J, MOLINA J M. Simulation in real conditions of navigation and obstacle avoidance with PX4/Gazebo platform[J]. Personal and Ubiquitous Computing, 2022, 26(4): 1171-1191.
[7] YANG Y L, MENG W, ZHU S Q. A digital twin simulation platform for multi-rotor UAV[C]//Proceedings of the 2020 7th International Conference on Information, Cybernetics, and Computational Social Systems. Piscataway: IEEE, 2021: 591-596.
[8] WU M M, XIAO Y X, BI Q. Software design of monitoring and flight simulation for UAV swarms based on OSGEarth[J]. International Journal of Computational Science and Enginee-ring, 2020, 21(3): 346-354.
[9] TANGIRALA N T, ABRAHAM A, VYAS P, et al. Distributed implementation of network simulations for large-scale vehicle-to-everything applications[J]. International Journal of Vehicle Autonomous Systems, 2022, 16(2/3/4): 246-258.
[10] FUJIMOTO R M. Development of the parallel and distributed simulation field[J]. Simulation, 2024, 100(12): 1197-1223.
[11] 杨慧杰, 肖桃顺, 武晨, 等. 无人集群虚实混合仿真试验环境集成构建研究[J]. 系统仿真学报, 2024, 36(4): 825-833.
YANG H J, XIAO T S, WU C, et al. Research on integration of virtual-real hybrid simulation experiment environment for unmanned swarm[J]. Journal of System Simulation, 2024, 36(4): 825-833.
[12] 钟林钢, 降晶晶, 叶超宇. 基于Unity3D的无人机集群仿真平台设计[J]. 计算机科学与应用, 2021, 11(9): 2242-2251.
ZHONG L G, JIANG J J, YE C Y. Design of drone swarm simulation platform based on Unity3D[J]. Computer Science and Application, 2021, 11(9): 2242-2251.
[13] D’URSO F, SANTORO C, SANTORO F F. An integrated fra-mework for the realistic simulation of multi-UAV applications[J]. Computers & Electrical Engineering, 2019, 74: 196-209.
[14] 蒋化南, 张帅, 林宇斐, 等. 基于MPI的分布式并行Gazebo仿真优化与测试[J]. 计算机科学, 2021, 48(S2): 672-677.
JIANG H N, ZHANG S, LIN Y F, et al. Simulation optimization and testing based on gazebo of MPI distributed parallelism[J]. Computer Science, 2021, 48(S2): 672-677.
[15] 汤恒先, 姚远, 康浩翔. 基于容器的规模化无人机集群仿真引擎研究与实现[J]. 计算机应用, 2025, 45(8): 2704-2711.
TANG H X, YAO Y, KANG H X. Research and implementation of large-scale unmanned aerial vehicle swarm simulation engine based on container[J]. Journal of Computer Applications, 2025, 45(8): 2704-2711.
[16] ALAEI M, YAZDANPANAH F. A survey on heterogeneous CPU GPU architectures and simulators[J]. Concurrency and Computation: Practice and Experience, 2025, 37(1): e8318.
[17] 陆天宇, 徐湛, 崔红元, 等. 大幅宽SAR图像嵌入式舰船实时检测系统设计[J]. 计算机工程与应用, 2024, 60(1): 301-309.
LU T Y, XU Z, CUI H Y, et al. Design of embedded real-time large size SAR image ship detection system[J]. Computer Engineering and Applications, 2024, 60(1): 301-309.
[18] JAISWAL D, KUMAR P. Real-time implementation of moving object detection in UAV videos using GPUs[J]. Journal of Real-Time Image Processing, 2020, 17(5): 1301-1317.
[19] NIEPCERON B, NAIT-SIDI-MOH A, GRASSIA F. Moving medical image analysis to GPU embedded systems: application to brain tumor segmentation[J]. Applied Artificial Intelligence, 2020, 34(12): 866-879.
[20] 孙卡, 俞俗强. 面向GPU的地形遮蔽探测并行算法[J]. 计算机工程与应用, 2024, 60(15): 66-76.
SUN K, YU S Q. GPU-oriented parallel algorithm for terrain occlusion detection[J]. Computer Engineering and Applications, 2024, 60(15): 66-76.
[21] RIBOLDI C E D, TOMASONI M. Formation flight of fixed-wing UAVs: dynamic modeling, guidance design, and testing in realistic scenarios[J]. Aerospace, 2025, 12(3): 260.
[22] PARK S, DEYST J, HOW J. A new nonlinear guidance logic for trajectory tracking[C]//Proceedings of the 2004 AIAA Guidance, Navigation, and Control Conference and Exhibit. Reston: AIAA, 2004: 4900.
[23] YOSHITANI N. A flight-path control of aircraft based on requ-ired acceleration vector[J]. Journal of the Japan Society for Aeronautical and Space Sciences, 2007, 55: 111-116.
[24] MELLINGER D, MICHAEL N, KUMAR V. Trajectory genera-tion and control for precise aggressive maneuvers with quadrotors[J]. The International Journal of Robotics Research, 2012, 31(5): 664-674.
[25] ZUO J W, FAN J F, LI K, et al. Parallel CUDA-based optimization of the intersection calculation process in the Greiner Hormann algorithm[J]. Algorithms, 2025, 18(3): 147.
[26] 左军涛, 朱恩成, 黄四牛, 等. 基于GPU的弹道快速计算方法[J]. 火力与指挥控制, 2012, 37(9): 193-197.
ZUO J T, ZHU E C, HUANG S N, et al. Fast calculation method base on GPU for trajectory[J]. Fire Control & Command Control, 2012, 37(9): 193-197.
[27] TANG Y, QIAO L B, YIN L J, et al. Training large-scale language models with limited GPU memory: a survey[J]. Frontiers of Information Technology & Electronic Engineering, 2025, 26(3): 309-331.
[28] TALLADA M G, MORANCHO E. Heterogeneous programming using OpenMP and CUDA/HIP for hybrid CPU-GPU scientific applications[J]. The International Journal of High Performance Computing Applications, 2023, 37(5): 626-646.
[29] AL-HAYANNI M A N, XIA F, RAFIEV A, et al. Amdahl’s law in the context of heterogeneous many-core systems a survey[J]. IET Computers & Digital Techniques, 2020, 14(4): 133-148.