Abstract: Computed tomography of chemiluminescence (CTC) can achieve transient three-dimensional measurement of the chemiluminescence field of the excited state particles in the combustion chamber. To further improve the measurement accuracy of CTC， an improved optimization algorithm of the projection difference minimization is proposed based on the characteristics of flames. Then， using the Sine simulation model to evaluate the reconstruction performance of the ART algorithm， the ASD-POCS algorithm and the PDM algorithm at different angles. Finally， the reconstruction of OH chemiluminescence field in a stable two-dimensional premixed flame is performed. The performance of the reconstruction algorithm is further evaluated by comparing the reconstructed results with the actual imaging results. The results show that reconstructed results using the medical algorithm cannot reveal the internal flame structures and therefore the algorithm is not suitable for the reconstruction of flames. The PDM algorithm can improve the reconstruction quality and speed effectively. Even at sparse angles， the algorithm can still achieve accurate reconstructed results and capture the main structure of the flame.

Key words: Chemiluminescence；Tomography；Reconstruction algorithm；Sparse angle；Flame image reconstruction

摘要： 化学发光计算断层成像技术(CTC)能够实现对燃烧室内激发态粒子化学发光场的瞬时三维测量。为进一步提高CTC的测量精度，依据火焰的结构特性提出一种基于投影残差最小化的优化重建算法(PDM算法)，利用Sine仿真模型对ART算法、ASD-POCS算法和PDM算法在不同角度下的重建性能进行评估。实现对一组稳定二维预混火焰的OH化学发光场重建，通过重建结果与实际拍摄结果的对比，进一步评估重建算法的性能。结果表明:传统医学重建算法丢失火焰内部结构特征，不适用于火焰图像重建；PDM算法可以加快重建速度，提高重建质量，即使在稀疏角度下的重建，仍能得到较高精度的重建结果，有效捕捉火焰的主要结构特征。

关键词: 化学发光；断层成像；重建算法；稀疏角度；火焰图像重建