Robust Trajectory Estimation for Maneuvering Targets Using an Adaptive Interacting Multiple Model Extended Kalman Filter

Accurately tracking maneuvering targets remains a significant challenge in fields such as autonomous navigation and surveillance. This paper presents a robust solution using an Interacting Multiple Model Extended Kalman Filter (IMM-EKF). The proposed architecture adaptively combines three distinct kinematic models: a Near Constant Velocity (NCV) model for linear motion, a Coordinated Turn (CT) for constant turn rates, and a Coordinated Turn with Rate and Acceleration (CTRA) to handle aggressive maneuvers. The IMM framework dynamically weights each model's contribution based on the measurement likelihood, producing a fused state estimate that is more reliable than any single-model filter. The algorithm's performance was rigorously validated against ground truth data, demonstrating high precision with a position Root Mean Square Error (RMSE) of ۰.۳۱۱۷ m and a yaw RMSE of ۲.۱۶۱۴ degrees. Furthermore, the filter's statistical integrity was confirmed through consistency tests, with ۹۴.۱۶% of the Normalized Innovation Squared (NIS) values falling within the ۹۵% confidence interval. These results underscore the effectiveness of the proposed multi-model approach for complex and dynamic trajectory estimation.

کلیدواژه ها:

State estimation ، Trajectory Estimation ، Maneuvering Target Tracking ، Extended Kalman filter ، Interacting Multiple Model ، Motion Models