Robust Temperature and Gas Concentration Imaging by LAS Tomography With Adaptive Basis Function Fitting and Artifact Removal

Xiaoqian Zhang,Lijun Xu,Jinting Wen,Kai Zhao,Zhang Cao

DOI: https://doi.org/10.1109/tim.2024.3400350

IF: 5.6

2024-05-30

IEEE Transactions on Instrumentation and Measurement

Abstract:Laser absorption spectroscopy (LAS) tomography is an effective method for cross-sectional imaging of temperature and gas concentration distributions in combustion diagnosis. In this article, the adaptive basis function fitting and artifact removal method is proposed to realize robust imaging in LAS tomography. Modified Mexican hat functions are introduced as basis functions to depict continuous distributions in the region of interest (ROI). Adaptive basis function fitting is realized by introducing a second reconstruction with core parameters of basis functions, the scale factors, and center points, determined adaptively from the first reconstruction. After obtaining the integral absorbance densities of two different spectral lines by adaptive basis function fitting, the similarity between their ratio and one of them is used to locate and remove the artifacts in reconstructed temperature images. The proposed method yields less artifacts and shows stronger noise immunity. The temperature image error can decrease by 12% at high noise levels. Dynamic flames of a McKenna burner were measured and temperature images with less artifacts were achieved. In an acoustically excited Bunsen burner, the proposed method precisely extracted the fundamental frequency and acoustical excitation frequency and yielded a higher structural similarity index. For a high-temperature wind tunnel, the maximum relative error of temperature in the center of the imaging region was 5.43% and the robustness of the proposed method in practical applications was verified.

engineering, electrical & electronic,instruments & instrumentation