直接力控制导弹的拟人智能自动驾驶仪设计

Reaction-jet Control Missile Autopilot Design Based on Human-imitating Intelligent Control

拟人智能控制是一种基于被控对象的物理模型以及人类相应的知识模拟人类处理控制问题的思路和一般方法解决问题的控制方法。基于拟人智能控制思想提出一种针对采用直接力控制导弹的姿态控制的控制逻辑和非线性控制律的设计方法。控制逻辑决定了气动力与直接力两种控制方式的权限分配。然后,利用拟人思想设计定性控制律。利用广义归约的方法将复杂的导弹姿态控制问题分解成若干个可直接解决的本原问题,然后依据各本原问题的主次关系以及耦合关系通过分析导弹的飞行动力学模型设计出非线性控制律结构,最后利用遗传算法将控制律中的各权值优化定量得到定量控制律。所设计出的控制逻辑和控制律在导弹进行大过载机动时保证导弹的稳定性。仿真结果表明复合控制方法可以使导弹具有更高的精度。

Based on the human-imitating control, which imitated human thinking way and general manner of dealing with control problem depending on the physical model of the object and corresponding human knowledge, a control logic and a nonlinear attitude control law for reaction-jet control missile were proposed. The control logic determined the sequence between aerodynamic and reaction-jet multi-controller. Then, the qualitative control law was designed by human imitating. The problem of missiles’ attitude control could be decomposed into several primitive problems by generalized reduction, which could be easily resolved. Then, according to the importance of the primitive problems and the coupling relationship among them, the structure of a nonlinear control law could be given by analyzing the flight dynamic model of the missile. Finally, the quantitative control law was optimized using the genetic algorithm. The present control is suitable for the attitude control for missiles which are characteristic of high-maneuvering, system nonlinearities and strong coupling. The control logic and control law can acquire large maneuver, keeping the missile stabile. The simulation result shows that the compound control method brings higher precision.