Improving Aided Track Performance During Periods of Tracker Sensor Failure by Utilizing a Target Body Fixed Coordinate System.

摘要

A new target acceleration model capable of improving the performance of an airborne aided track system during periods of tracker sensor failure was investigated. An aided track target prediction algorithm utilizing range and angle rate measurements was modified so that the target state estimates were propagated by integrating the estimated constant, body fixed acceleration. The modified algorithm performance was compared to the original by simulating three target paths. At specified times, all tracking sensors were turned off to simulate sensor failure. Each algorithm attempted to keep the target centered on the tracker boresight by extrapolating target position estimates without processing any measurements. RMS pointing errors were computed for the one second periods following each failure. During normal sensor operations, the performance of the modified algorithm was equivalent to the original. When the sensors were failed during the tracking of targets following a constant, body fixed acceleration, the modified algorithm generally produced smaller RMS errors. (After 2.8 seconds of filter operating time, the modified algorithm produced one third the RMS error of the original.) When sensor failure occured during the tracking of targets following a rapidly oscillating path or traveling in a straight line under large linear accelerations, both algorithms produced equivalent RMS errors. The modified algorithm demonstrated the same sensitivity to measurement uncertainties as the original. (Author)