Research on Airborne Radar Multi-target Continuous Tracking Algorithm on Sea Surface Based on Deep Kalman Filter

It is difficult for airborne radar to track multiple targets on the sea surface because of the large number of targets, high density and various types of targets. The application of traditional tracking algorithm is limited by operation, especially in the case of airborne radar tracking of sea target, the amount of tracking calculation will increase explosively with the increase of target track and radar echo number. In this paper, a multi-target continuous tracking algorithm based on deep Kalman filter is used to predict the state matrix through slicing recurrent neural network, combined with linear Kalman filter, which can improve the tracking accuracy of the target and improve the computing efficiency. Compared with the traditional tracking algorithm, the tracking accuracy of the proposed method is improved by about 10 m, and the convergence time is reduced by about 25 s. Simulation results verify the effectiveness of the proposed multi-target continuous tracking algorithm, and it has good performance.