Radar track prediction method based on BP neural network

Taking into account the complex electromagnetic environment in which the radar target is located, it is difficult for the traditional track prediction method to adapt for the high complexity, randomness, and uncertainty of the manoeuvring target track. This study proposes a radar track prediction method based on backpropagation (BP) neural network. According to the historical track of the radar target, this method uses BP neural network to model its movement law and obtain the target's predicted track. Finally, the track prediction experiment was conducted with measured data and compared with the Kalman filter track. The results show that the proposed method has higher track prediction accuracy and can be used for radar track prediction.