TARGET TRACKING WITH THE USE OF NEURAL NETWORKS
Ilke Titi, Ahmet Kaplan · 2000
In this work, a simple method for target tracking, based on Artificial Neural Networks (ANN), is presented. The backpropagation algorithm is used to train the networks by using the measured position, velocity and acceleration data sets obtained from six different types of aircraft radars such as cargo, bomber, fighter and commercial aircrafts. The test results of ANNs are in very good agreement with the measured results. The results of ANNs are also compared with the results of Kalman filter which is widely used in target tracking. It is shown that the results predicted by using ANNs are better than those predicted by Kalman filter.