As the only energy source in the solar system, the sun is a rich treasure of spectral information for having a very wide continuous spectrum and tens of thousands of absorption and emission lines. The energy of solar electromagnetic radiation is mainly concentrated in the visible and infrared regions, among which the solar infrared spectrum with Doppler redshift characteristics can be used as the information source for astronomical velocity measurement and navigation. As an important part of astronomical velocity measurement navigation, Solar spectral Doppler redshift velocity measurement can deduce the relative radial velocity between spacecraft and the sun by calculating the Doppler redshift of the received solar spectrum relative to the standard solar spectrum. However, the spectral distortion caused by such solar activities as sunspots, corona, or flares will lead to the instability of the solar spectrum, which will affect the velocity measurement accuracy of the solar spectrum and in turn, the navigation accuracy. In order to improve the navigation performance of solar spectral velocity measurement, based on the principle of solar spectral velocity measurement, the signal processing method of solar spectral Doppler redshift velocity measurement is explored in this paper. This paper proposes an adaptive EMD-NDFT Doppler redshift velocity measurement method for solar spectral velocity measurement navigation. By this method, the redshift is calculated according to the Doppler effect of the solar spectrum and the radial velocity of the spacecraft relative to the light source is derived. The method consists of EMD processing, NDFT and correlation matching. First, the non-stationary received solar spectral signals are stratified adaptively by using the EMD algorithm, and the adaptive threshold filtering and noise reduction are carried out according to each layer of intrinsic mode signal to obtain a stable reconstructed signal. Second, according to the characteristics of non-uniform sampling of the solar spectrum, the standard solar spectrum and the received spectrum respectively are transformed by NDFT to convert the spectrum from the time domain to the frequency domain. Thirdly, Taylor matching is performed on the low-frequency characteristic spectral lines of the two spectra and the phase difference to obtain the radial velocity of spacecraft relative to the Sun. This method combines time-domain denoising and frequency-domain sparsity to obtain radial velocity more quickly and accurately. This paper analyses the spectral changes of sunspot activity in different years within a cycle, and their doppler redshift velocities are calculated and analyzed. The simulation results show that the adaptive EMD-NDFT method can effectively improve the accuracy of velocity measurement and greatly reduce the computational complexity for the solar spectral data in different periods and under different noises.