Enhancing USVs navigation based on minimum error entropy of GPS vector tracking

Abstract

In recent years, Unmanned Surface Vehicles (USVs) have increasingly been used for river monitoring and hydrological surveys. USVs rely on global navigation satellite systems (GNSS) for navigation. However, signal blocking can cause the traditional GNSS vector tracking loop to increase the code phase and carrier frequency errors, leading to higher positioning errors that do not meet USVs’ requirements. To address this problem, we propose a vector tracking method based on the minimum error entropy (MEE) in the signal tracking module. The minimum error entropy Kalman filter (MEEKF) is adopted as the loop filter to mitigate code phase and carrier frequency errors, reduce non-Gaussian noise and random errors generated by signal blocking, and enhance the positioning accuracy and robustness of USV navigation. The measurement noise covariance of the loop filter was adjusted adaptively using the signal carrier-to-noise ratio (CNR). A field experiment was conducted using a commercial GNSS receiver as reference. The results demonstrate a 19.3% improvement in positioning accuracy compared with the traditional method in an open environment. Moreover, the proposed method maintains stable operation and achieves a 79.4% improvement in positioning accuracy during signal blocking. This novel algorithm offers a new concept for USV navigation systems to cope with signal blocking.