基于投影序阵和压缩传感的脉动燃烧四维光偏折层析研究

项目名称： 基于投影序阵和压缩传感的脉动燃烧四维光偏折层析研究

项目编号： No.61205005

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 信息四处

项目作者： 张斌

作者单位： 青岛科技大学

项目金额： 26万元

中文摘要： 脉动燃烧因其优异的燃烧和排放特性在能源、动力等领域具有重要的应用价值，由于脉动燃烧具有复杂的物性和结构特点，其全场显示与测量是仍未完全解决的热点问题。项目旨在研究脉动燃烧的动态变化显示与测量问题，构建新的光偏折四维层析(4D-CT)理论和方法体系。项目拟结合数字微镜器件(DMD)技术原理，建立投影有序阵列可视化方法，结合高速摄像系统实现单CCD、多视角、同时、同光路条件动态采样。以基于光偏折成像原理、极度欠采样的特定优化目标、融合流场先验知识的迭代算法为重建策略，推导非完全投影光偏折CT的压缩传感(CS)重建算法。研究燃烧火焰结构和流场参量变化的4D显示方法，实现脉动火焰三维表面和温度、密度等参量场随时间周期性变化的动态影像。本项目的开展将为脉动燃烧理论研究和新型燃烧器设计提供基础性科学依据，有助于解决能源、环境、航空航天等领域的相关热点问题，同时将推进CT和CS新理论的发展。

中文关键词： 光偏折层析；燃烧测量；压缩传感；重建算法；4D可视化

英文摘要： Pulsating combustion has been recognized as an important applied technology since it is characterized by high combustion load, reduced pollutant emission, and augmented heat transfer. These characteristics also make it difficult to visualize and measure the combustion flow field. In this project, a new system of theories and methodologies of four-dimensional deflection tomography is constructed to visualize and measure the dynamic changes of combustion flow field. Based on the principle of digital micromirror device, a deflection tomographic setup is presented for obtaining the chronological arrays of multidirectional deflectograms. Deflection projections in different angles of view can be captured synchronously in same optical path condition by high speed camera system. An iterative reconstruction strategy is designed based on ray deflection theory, optimization objective for a very few views tomography, and a priori knowledge fusion. An algorithm is derived for deflection tomographic reconstruction from incomplete data using the compressed sensing framework. The reconstruction method is tested using simulated data for incompleteness conditions similar to those found in the experimental data. Four-dimensional visualization is researched for the reconstruction of flame surface and the variations of parameter dis

英文关键词： deflection tomography；combustion measurement；compressed sensing；reconstruction algorithm；four-dimensional visualization