Neural solution to multitarget tracking in clutter with velocity measurements

The problem of tracking multitarget in clutter environment in a track while scan (TWS) system is considered. In such an environment, missed detections and false alarms make it impossible to decide, with certainty, the origin of received report. The two main functions of the TWS algorithm are : 1) plot-to-track (state predicted estimate) association, and 2) the filtering algorithm that uses the associated plot for recursive estimation. In this paper, ANN (artificial-neural-network based data association method approximate to NNPDA (nearest neighbour probabilistic data association) is used to realize plot-to-track association, while filtering algorithm suitable to tracking with radial velocity measurements is introduced. Simulation results included in this paper show a great improvement of tracking performance when radial velocity measurements are used