Robust least-squares bias estimation for radar detecting biases and attitude biases

To focus of this paper is on the estimation for the ship-borne radar detecting systematic or registration errors. According to the ship-borne radar data processing, the types of bias are divided into four main categories: radar measurement biases, ship-position biases, attitude biases and baseline transform biases. In this paper, we present an algorithm that uses detecting data for estimation of equivalent biases. Our approach is unique for two reasons. Firstly, we explicitly avoid the use of individual biases and use equivalent biases model the four main class biases, This leads to a highly nonlinear bias model that contains 12 unknown parameters. Secondly, we use the singular value decomposition (SVD) within least-squares estimator to automatically handle the issue of parameter observability. Finally, according to two different simulation scenes, we demonstrate that our algorithm can improve track accuracy, especially for ship-borne radar.