Total Variation Regularized TDLAS Tomography for Temperature Imaging

Tunable Diode Laser Absorption Spectroscopy (TDLAS) tomography is a well-accepted technique for noninvasive combustion diagnosis, which can generate images for the distribution of critical flow parameters in the combustion field, such as temperature and species concentration. Many computational tomographic algorithms have been proposed to reconstruct the absorbance distributions from limited line-of-sight TDLAS measurements. However, due to the ill-posedness of this inverse problem, the reconstruction accuracy of traditional computational tomographic algorithms is usually not high, which leads to severe artefacts in the reconstructed temperature images. In this work, a new Total Variation (TV) regularized computational TDLAS tomographic algorithm is proposed for temperature imaging. The TV regularization is introduced to formulate the inverse problem for the reconstruction of absorbance distribution, and the modified Landweber algorithm is employed to solve the resulting optimization problem. With the well-reconstructed absorbance distributions, temperature distribution is calculated by the two-line thermometry. Simulation results show that the proposed TV regularized modified Landweber algorithm improves the quality of the reconstructed temperature images in comparison to reconstruction algorithm with Tikhonov regularization, Landweber algorithm and algebraic reconstruction technique with TV regularization.