Towards Reliable Navigation for Intelligent Unmanned Surface Vehicles

This paper presents a novel navigation system developed to tackle the challenge of unreliable navigation data during unmanned surface vehicles (USVs) operation. The system incorporates a method to determine the reliability index of the navigation system, which helps identify system uncertainties, and an Unscented Kalman Filter (UKF) based algorithm for USV trajectory recovery from the unreliable sensor data. Based on the determined reliability index, the system can quickly identify unreliable data, accurately assess risks associated with imprecise data, which enables it to generate a more dependable guidance route. The effectiveness and capabilities of this reliable navigation system have been validated through simulations and practical trials, highlighting its practical value and significance.