LAUSR

ABSTRACT The impact of measurement noise and tomographic methods on reconstruction quality are investigated via simulation studies under a hypothesis harsh flow field environment. Computer tomographic methods based on direct absorption spectroscopy and wavelength modulation spectroscopy are conducted under different measurement noise levels to assess the noise effect quantitatively. It turns out that the wavelength modulation spectroscopy-based tomographic method shows efficient immunity to measurement noise. However, the direct absorption spectroscopy-based tomographic method is found to be susceptible to the measurement noise, which makes the reconstruction quality decline with the increasing noise level. In addition, two kinds of tomographic methods including linear and nonlinear method are investigated under a complex multimodal temperature distribution environment with limited projection data. The reconstructed distribution for the nonlinear tomographic method reveals great fidelity to the original value with a maximum deviation of 51.9 K, while the reconstructed distribution for the linear tomographic method under parallel beam arrangement shows an obvious distortion.