Parameter Optimization Of ARM Control System Using Elitist Selection Genetic Algorithm

Our main work is researching of the Anti-Radiation Missile system modeling and parameter optimization.The guide system of Anti-Radiation Missile is multi-variable, multi-loop, nonlinear and nonconstant.The loops have coupling each other. So the system is really complicate.First the mathematic model, which includes body move model, dynamics model, Autopilot model, and guided loop model, are built. And different kinds of disturbances are taken into account. Some defined parameters are given.Second, A mix algorithm, which is a mixer of Adams algorithm and Runge-Kutta algorithm, are used to get the numerical solver of the differential equations. We got the values of the state variables of missile flight. We wrote the program and at last drew the trajectory curves.Because the system is complex, the traditional optimal method doesn't work well. The genetic algorithm is used to solve the problem.Two type of genetic algorithm are compared. They are Elitist Selection Genetic Algorithm and Steady State Genetic Algorithm. They are both real-code algorithm. The De Jong's five-function platform is used to test. The result shows that Elitist Selection Genetic Algorithm is fitter for the optimization of Anti-Radiation Missile system.Finally we used Elitist Selection Genetic Algorithm to optimization. The problems on the implement of Elitist Selection Genetic Algorithm are discussed. In the end, the simulating results are given. Compared with another method— the random radial line algorithm. The Elitist Selection Genetic Algorithm hasadvantage in many fields, such as the precision and the miss distance. Especially The Elitist Selection Genetic Algorithm runs much fast than the random radial line algorithm.