V3I7P36

Adaptive Interacting Multiple Model Cubature Kalman Filter for Tracking Highly Maneuvering UAV Targets in Air Defense Radar Systems

Do Ngoc Thuong*¹

Abstract

Accurate tracking of highly maneuvering unmanned aerial vehicles (UAVs) remains a major challenge for modern air defense radar systems due to nonlinear motion, abrupt maneuvering, and measurement uncertainty. Conventional Kalman-based tracking algorithms often suffer from performance degradation when the target rapidly changes its velocity or heading. This paper proposes an Adaptive Interacting Multiple Model Cubature Kalman Filter (AIMM-CKF) for robust UAV target tracking. The proposed approach combines multiple motion models, including constant velocity (CV), constant acceleration (CA), and coordinated turn (CT), with an adaptive process-noise covariance estimation mechanism based on innovation statistics. The adaptive strategy enables automatic adjustment of the process noise according to the target maneuvering intensity while the IMM framework dynamically selects the most suitable motion model. Simulation results under various UAV maneuvering scenarios demonstrate that the proposed method significantly reduces position and velocity estimation errors compared with EKF, UKF, CKF, and conventional IMM-CKF.

Keywords:

UAV tracking, Air defense radar, IMM, Cubature Kalman Filter, Adaptive filtering, Target tracking.