Study of a multi-parameter fusion Gaussian a priori reconstruction method for hypersonic aero-radiation spectroscopy

Abstract

The extreme thermodynamic environment created by the high-temperature gas at the optical window in the wake of a hypersonic vehicle leads to strong aero-radiation, which causes significant background noise in star sensor observations. The study of spectral detection of aero-radiation contributes to suppress the background noise it introduces due to its spectral band characteristics. This study analyzes the spectral band characteristics of the aero-radiation spectra. Based on the theoretical framework of compressed sensing, the spectral band features are mapped to a multiple Gaussian distribution, thereby constructing a multi-parameter fused Gaussian prior model. A multi-parameter fused Gaussian prior adaptive reconstruction method is proposed. The method integrates spectral characteristics extracted from physical experimental results with mathematical modeling, enabling non-blind reconstruction of aero-radiation spectra, in which prior information—such as peak positions, widths, and intensities—is explicitly incorporated to improve reconstruction accuracy and robustness. Furthermore, a spectral goodness-of-fit evaluation criterion is proposed in this study to assess the fusion results of physical experimental spectra and simulated reconstructed spectra, serving as a standard for evaluating the quality of spectral reconstruction. The experimental results prove the effectiveness of the proposed method.