Missile Trajectory Defense Planning based on Distributed Neural Network

Abstract

In this article, the development status, models and learning algorithms of distributed neural networks are investigated, conventional BP algorithms are analyzed, some improved algorithms are proposed, and the principle and structure of neural network PID controllers are discussed. In addition, a distributed neural network is used to predict the trajectory of the missile and develop an appropriate attack prevention plan. At the same time, the construction of the virtual world will be introduced by constructing the virtual sky and terrain in the simulation process, changing the perspective in the virtual world, and completing the ballistic counterattack plan of the missile. The purpose of the anti-attack plan is to provide technical reserves for the further development of the National Missile Defense Program, so that the Ministry of Defense can update the technology demonstration system of the existing bases, reduce risks and accelerate technological updates. In order to solve the problem of peculiar or drastic changes in the existing trajectory tracking deviation, we introduced a new method that takes the trajectory arc as the independent variable, and built a linear variable parameter system for trajectory tracking on this basis, and calculated a new automatic Tracking deviation under variable.