Covariance Control for Sensor Management in Cluttered Tracking Environments
Michael Kalandros, Lucy Y. Pao · Journal of Guidance Control and Dynamics · 2004
A simple structure for a high-performance three-dimensional tracking filter is derived in the rotating sensor frame assuming high measurement rates and small angular deviations during target tracking. The resulting three-dimensional estimator consists of three independent filters, whose state includes position, velocity, and acceleration along each of the axes in the rotating sensor frame. Because line-of-sight models for the target acceleration are identical in all three axes and independent, the replacement of the basic exponentially correlated acceleration model by more complicated models preserves the filter structure. Such sophisticated models are based on the multiple model approach.