Research on the bearing estimation of underwater target based on an improved Huber method

In an active sonar vehicle homing system based on split beam model, there is an arbitrarily large error in an underwater target instantaneous bearing estimation if the length of digital signal processing does not match that of target echo, or SNR is not a constant during the target bearing estimation. It will easily cause a wrong target classification. For increasing the performance of an underwater target bearing estimation, this paper presents a time-bearing model of an underwater target. Robust statistics and Huber estimator are introduced. An improved Huber estimator algorithm is proposed and its application is given. The method is characterized by simple algorithm, low computational cast, high estimation quality and accuracy as well as low classification error. The simulation experiment results have verified the efficiency and robust performance of the proposed method.