导弹自动驾驶仪改进遗传模糊控制器设计

Fuzzy Logic Controllers Using Modified Genetic Algorithms for Missile Autopilot

针对导弹飞行控制系统具有参考模型不精确,非线性时变的特点,设计了一种新型的导弹自动驾驶仪的模拟退火遗传模糊PID控制器。该控制器是利用模糊控制器的模糊推理能力,在线整定PID控制器参数,再采用遗传算法与模拟退火算法结合,离线搜索寻优模糊控制器中的隶属函数参数集。Matlab仿真实验结果表明,所设计的控制器能够使系统具有良好的动态性能、鲁棒性以及全弹道性能,利用模糊查询表可满足系统的实时性要求,工程应用前景较好。

The missile flight control system has the characteristics of model inaccuracy, nonlinearity and time-varying. This paper designs a new Fuzzy-PID controller optimized by simulated annealing genetic algorithm for missile autopilot. The new controller is to use ability of the inference of Fuzzy-PID Controller to set the parameters of PID controller,use genetic algorithm and simulated annealing algorithm to optimize the parameters of membership function of Fuzzy-PID Controller. Matlab simulation results show that the controller has good dynamic performance,robustness and good performance of every point on the trajectory. By using fuzzy lookup table,the controller can meet the real-time requirement and has a good engineering application prospect.