Missile Attack Zone Fitting Based on K-SAE-SVM

Abstract

In short-range air combat, the UCAV has high maneuverability, and the situation on both sides is changing rapidly. Therefore, it is very important to solve the attack zone in real-time. It can be seen that the calculation of missile attack zone based on the optimal escape strategy of enemy aircraft is complex, and the solution time is about 200s, which completely cannot meet the needs of UCAV air combat. In terms of attack area fitting, the deep neural network has strong nonlinear fitting ability, and has very good real-time after training. It has been effectively applied in many fields, which will be an effective solution for nonlinear attack area fitting. In this paper, k-sparse autoencoder combined with SVM is proposed to build a fast solution network of missile attack zone. Simulation results show that the proposed method has good real-time performance after deep network training, which is verified to meet the decision needs in real time.