基于线性迭加法的大气紊流数值仿真

Numerical Simulation of Atmospheric Turbulence Based on Linear Superimposing Method

在大气紊流系统优化的研究中,当前普遍采用采用白噪声通过成形滤波器来得到紊流频谱和紊流信号。传统方法对紊流频谱模型进行简化,因而存在模型偏差,仿真精度提高有限。为此提出利用函数线性迭加法,将大气紊流看作无限多个随机余弦函数迭加的随机过程。以Dryden模型和Von Karman为基础,推导出线性迭加法的计算公式,利用谱估计法对这两种方法进行了验证和比较,通过仿真证明线性迭加算法可信可行,不仅精度高于成形滤波器法,而且兼容不同大气紊流频谱模型,通用性更好,为大气紊流的研究提供了参考。

During the atmospheric turbulence simulation, the atmospheric turbulence sequence is generally crea- ted by Gauss white noise sequence passing through a shaping filter. In this method, the turbulence model is simpli- fied, so the model error is unavoidable and the numerical precision is not high enough. A cosine function linear su- perimposing method was proposed, in this new method, the atmospheric turbulence model is assumed as randomized procedure by an expression of a series of different amplitudes, random phase positions and muhifre-quency cosine functions. On the basis of Dryden＇ s model and Von Karman＇ s model, we deduced a calculation formula of the new method. The spectral estimation simulation results show that the proposed new method is rational and the suggested simulation algorithm is valid. Comparatively, the new method has higher precision and better generality than the tradi- tional method.