EKF Based Trajectory Tracking and Integrity Monitoring of AIS Data

Abstract

This work presents a novel approach for integrity monitoring of AIS data. Currently, the AIS is a valuable source for maritime traffic situation assessment but not suited for collision avoidance, as it is prone to failures and not capable of indicating the level of data integrity. To tackle this, an EKF was designed to track vessel trajectories, which allows for failure detection based on residual monitoring. For the latter, two methods for hypotheses testing were implemented, namely chi-squared and GLR tests. In addition, the IMM framework was adopted for mixing the state estimates of two different process models, the CV and CTRV. The designed filter will be validated on behalf of simulated and real-world AIS data.