Onboard reentry landing footprints generation based on neural network

In order to meet the real-time requirement of onborad landing footprint generation, a fast onboard method based on neural network is proposed. We have analyzed the characteristics of the footprint of differernt vehicles and found that the boundary under downrange-crossrange description is similar to the ellipse. Then, the footprint samples are obtained by the trajectory optimization algorithm off-line and the multi neural network is trained based on scheduling management method. The trained neural network is used to generate the footprint onboard. The simulation results show that the method is effective and has high accuracy.