High-precision Prediction of Ships Including Speed and Orientation

Veracity and instantaneity when identifying a vessel's position are the main factors for safe navigation. Information obtained through the traditional automatic identification system (AIS) is imprecise and hysteretic, which restricts early warning capacities. The main purpose of this paper is to solve two problems. First, we test the precision of locating a ship's position, speed, and orientation from AIS and find that position precision fluctuates while speed and orientation precision are stable for an epoch. We propose that the position of next epoch can be calculated by the position, speed, and orientation of present epoch. So, we can compare the calculated results with observation results, and eliminate gross errors. Second, we combine position, speed, and orientation and establish a high-order polynomial to predict the position of ship at the next epoch, which can meet the need for real-time acquisition of dynamic information. Results show that if there is no gross error, the precision of prediction can reach 1 m.