Improving Trajectory Predictions for Short Baseline Line-of-Bearing TrackingSystems

摘要

This report investigates methods for analyzing and improving the trackingperformance of short baseline line-of-bearing (LOB) systems. Such a system might be placed on an armored vehicle in order to track an incoming threat projectile and make an accurate prediction of the threat's future location (at a specific time) so that an appropriate reaction can he initiated. The processes for estimating the trajectory of a projectile over the terminal portion of its ffight are discussed. LOB, range, and radial speed sensor information were considered. Three statistical estimation methods for predicting a target's trajectory are discussed: recursive least squares, weighted least squares, and Kalman filter estimation. Simulation were performed for six sensor/estimator configurations for five attack azimuths and two intercept ranges. It is the inherent geometry of a short-baseline LOB system leads to large errors in the system's ability to determine range, whereas the determination of cross-range is less affected by the geometry. Simulation results are presented for all sensor/estimator combinations, and the relative improvements of each are discussed. One result shows that range sensors can greatly improve the performance of inherently bad LOB sensors but that care must be taken in properly formulating the LOB/range sensor estimation routines so as not to bias the results. Another result shows that it is important to allow the incoming threat projectile to approach as close as is feasible to improve the quality of the LOB observations and thus the overall estimation.