Target Tracking of Navigation Radar for Unmanned Surface Vehicle Based on an Improved Adaptive Kalman Filtering

In order to solve the problem of accurate navigation radar target tracking for a class of unmanned surface vehicle (USV) in highly dynamic and complex navigation environments, an improved adaptive Kalman filtering algorithm for tracking navigation radar targets of USV is proposed, considering that the observation noise of USV navigation radar target tracking is time-varying. The algorithm uses the USV attitude sensor to obtain real-time bow angle rate information, and adaptively updates the observation noise covariance in response to the attitude change. The method for determining the value of the Kalman filter observation noise covariance has been enhanced, resulting in a tighter match between the observation noise covariance and the actual value in the Kalman filter algorithm, thereby boosting target tracking accuracy. The effectiveness and superiority of the proposed method are verified by the ship experiments using the USV navigation radar of “Lanxin” to track the actual targets at sea.