Low-altitude Wind Shear Speed Estimation Method Based on KL Divergence Orthogonal Projection of Matrix Information Geometry

To address the issue of poor low-altitude wind shear speed estimation caused by an insufficient number of Independent and Identically Distributed (IID) training samples, a method for estimating low-altitude wind shear speed based on KL Divergence Orthogonal Projection (KLD-OP) of matrix information geometry is proposed in this paper. First, the method constructs a matrix manifold based on the intrinsic distribution of radar echo data. Then, the KL divergence on the manifold is employed to estimate the Clutter Covariance Matrix (CCM), which is subsequently used to partition the clutter subspace. Next, an orthogonal projection matrix is constructed for clutter suppression, and finally, the estimation of low-altitude wind shear speed is achieved. Simulation results validate the effectiveness of the KLD-OP method.