Wavelet-based intelligent optimization for doppler velocity estimation in the presence of celestial spectral distortion

The variations in instrument status, along with the absorption and reflectance of planetary bodies, can cause distortions in celestial spectra that affect the accuracy of Doppler velocity estimation. High-precision Doppler velocity estimation can enhance the accuracy of the celestial velocimetry navigation. To address this issue, we propose a wavelet-based intelligent optimization for Doppler velocity estimation (WIODVE), considering that wavelet coefficients encapsulate spectral distortion signals. During the training phase, the WIODVE utilizes a weighted factor set derived from the wavelet coefficients of celestial spectra to construct the position of horned lizards, using the Doppler velocity error as the fitness function. The horned lizard optimization algorithm (HLOA) is employed to optimize the weight factor set, allowing for the reconstruction of spectral distortions. In the testing phase, the optimized weight factor set and wavelet transform are used to dynamically reconstruct the distortion of the celestial spectra. Subsequently, the reconstructed distortions are employed to correct the observed celestial spectra, with Doppler velocity estimated by the Taylor method. Additionally, we derive the Cramér-Rao lower bound (CRLB) for Doppler velocity estimation in the presence of celestial spectral distortion. Experimental results demonstrate that the WIODVE outperforms both the template enhanced radial velocity reanalysis application (TERRA) and the Taylor methods, approaching the CRLB, and exhibits strong robustness to spectral distortions in the estimation of the Doppler velocity. Furthermore, the WIODVE significantly enhances the accuracy of the celestial velocimetry navigation.