基于涡结构的高速湍流光学成像退化模型及其统计特性研究

项目名称： 基于涡结构的高速湍流光学成像退化模型及其统计特性研究

项目编号： No.60878051

项目类型： 面上项目

立项/批准年度： 2009

项目学科： 金属学与金属工艺

项目作者： 许东

作者单位： 北京航空航天大学

项目金额： 30万元

中文摘要： 带有光学成像探测系统的飞行器在大气层内高速飞行时，光学头罩周围会形成剧烈变化的激波湍流，导致目标图像出现闪烁、抖动、偏移和模糊等退化现象，极大影响着系统的成像质量和探测能力，是大气层内执行高精度航空测绘、光学成像制导等任务所面临的瓶颈问题之一。 本项目针对这一科学问题，系统研究了高超声速湍流对光学成像的影响机理和影响规律。在湍流大涡模拟和风洞测试试验的基础上，研究了高速激波湍流的涡结构特征，提出了基于典型相关分析与本征正交分解、高阶奇异值分解和小波包分析的多尺度流场结构提取方法；针对不同尺度的流场结构，提出了基于CFD网格的光线追迹方法和基于Mie散射理论的多尺度流场结构光传输计算方法；并根据模拟计算和风洞测试试验数据，利用统计理论和退化模型辨识方法，分析了湍流不同尺度的涡结构对光传输的影响规律，给出了光学成像退化的统计特性；在此基础上，提出了基于参数估计、最大似然估计和非负支持域递归逆滤波技术的湍流退化图像校正算法。 相关的研究成果对实现高速激波湍流成像退化的准确校正，提高我国大气层内高速飞行器光学成像探测的能力，具有重要的理论意义和实际价值。

中文关键词： 湍流；光学成像；涡结构；光传输；退化模型

英文摘要： There are violent turbulence flow fields out of the optical hoods when aircrafts with optical imaging detection system fly with very high speed in atmosphere. It will cause object image degradation of flicker, dethering, shifting and blurring, which reduce the imaging abilities and qualities of the systems. This problem is one of the bottlenecks in the task of high-precision aerial remote sensing and optical imaging guidance in atmosphere. This project systematically studied the impact mechanism and patterns in optical imaging of high hyperacousitic turbulence. Based on large eddy simulation (LES) and wind tunnel test experiments, we studied the characteristics of high speed turbulence, and introduced the canonical correlation analysis (CCA) and proper orthogonal decomposition (POD), high-order singular decomposition (HOSVD) and wavelet package analysis to extract the vortexes with different scales. We proposed to use computational fluid dynamics (CFD) based ray tracing approach and the Mie's scattering theory to perform the optical transmission computation for different scale strutures of turbulences. And we also analysed the impact patterns in optical imaging of different scale strutures of turbulences with statistic theory and degradation model indentification approches, and gave the statistic characteristics of the optical imaging degradation. Finally, the approches based on parameter estimation, maximum likelihood estimation and nonegativity support constrains were proposed to correct the image gradation cause by turbulence. The results of these researches will offer great supports to the correction of degraded images caused by high-speed turbulence, and have great theoretical and pratical values to improve the detecting abilities of optcial imaging in atmosphere.

英文关键词： turbulence; optical imaging; vortex structure; optical transmission; degradation model