基于GPU加速的高流速尾焰仿真算法
High Velocity Tail Flame Simulation Algorithm Based on GPU Acceleration

在空军训练系统等一些特殊应用场景，需要使用由多个计算机组成的计算机组对数十架飞机进行实时虚拟仿真，模拟不同天气环境下战斗机的尾焰特效。战斗机尾焰和普通火焰不同，具有高流速、焰色特殊等特点。传统仿真方法在模拟高流速火焰时会出现撕裂现象，同时仿真需要的粒子数偏高，在模拟大规模战斗机群和其它粒子特效时会造成场景中总粒子数过高，需要较高的显卡性能。为了达到较好的仿真效果，同时降低对硬件的需求，本文提出了一种基于GPU加速的大规模战斗机高速喷射尾焰仿真算法，使用多个绘制管线加速火焰粒子运动计算，采取高斯白噪声来扰动尾焰，最后根据战斗机尾焰燃烧特性设计了热值衰变函数来进一步提高火焰的仿真效果。计算机仿真实验结果验证了本文提出算法的可行性和有效性，在满足真实感的前提下减小了对硬件性能的要求。
In some special application scenarios such as the Air Force training system, it is necessary to use a computer group composed of multiple computers to simulate dozens of aircraft in real time, espe-cially to simulate the tail flame effects of fighters in different weather conditions. The tail flame of the aircraft is different from the ordinary flame, and has the characteristics of high flow velocity and special flame color. The traditional simulation method will tear when simulating high-flow flame, and requires a large number of particles. When simulating large-scale fighter groups and other particle effects, the total number of particles in the scene will be too high, which requires a high graphics card performance. In order to achieve better simulation results and reduce the hardware requirements, this paper proposes a high-speed jet tail flame simulation algorithm based on GPU acceleration for large-scale fighters. It uses multiple rendering pipelines to accelerate the calculation of flame particle motion and disturbs the flame by white Gaussian noise. Finally, the heat value decay function is designed according to the characteristics of the tail flame of the fighter to further improve the simulation effect of the flame. The computer simulation results verify the feasibility and effectiveness of the proposed algorithm, and reduce the hardware performance requirements while satisfying the realism.