Knowledge Base Applications to Adaptive Space-Time Processing, Volume 4: Knowledge-Based Tracking

摘要

This report describes a knowledge-based tracking capability that will support a space-time adaptive processing (STAP) environment. The effort described in this report resulted in development of a multiple target tracker and the design and testing of several knowledge-based rules. Three types of tracking filters are described: (1) An uncoupled two-state alpha-beta filter with position and velocity component states, (2) An uncoupled three-state Kalman filter with position, velocity, and acceleration component states, and (3) An extended four-state Kalman filter with both x and y position and velocity component states. The first two filters use a one-dimensional measurement vector containing the report position component. The extended Kalman filters uses a three-dimensional measurement vector containing x and y report position and 'pseudo' Doppler, the latter defined as range times range rate.