Air-to-Air Missile Launch Envelope Computation Based on Neural Network

A fast algorithm for computing air-to-air missile launch envelope based on neural network is proposed, which can be applied to an air combat scenario for choosing the optimal missile launch time. Firstly, the simplified six-degree-of-freedom motion models of missile and fighter are built. Secondly, the missile-fighter pursuit-evasion engagement is simulated, by designing the traditional guidance strategy of missile and the common evasion maneuver of fighter. After that, the golden section method is used to search the launch envelope, and the envelope database is generated by simulating various air combat scenarios. Based on the database, the neural network is trained, and its optimal structure is determined by comparing the network performance with different number of layers and nodes. Lastly, the simulation results show that the proposed algorithm has a smaller computation error, compared with the traditional method of interpolation method.