Abstract: The disclosure provides a Lorentz constraint angle estimation method and a system in a non-Gaussian environment; the method includes the following steps: constructing an N-time slot received signal model based on a non-Gaussian noise environment to obtain a reflected signal; constructing a cost model based on Lorentz norm by a difference value between an actual received signal and the reflected signal, and performing an angle estimation by combining with an atomic norm to obtain a signal sparse reconstruction model; constructing an augmented Lagrangian function by the signal sparse reconstruction model, and carrying out the iterative update on the augmented Lagrangian function to obtain a reconstructed signal; and analyzing the reconstructed signal and searching spectral peaks globally to obtain spatial spectral peak points, and completing an angle estimation of the reconstructed signal.