基于随机鲁棒优化的无人机着舰控制律及仿真

Autonomy Carrier Landing Control System Design and Simulation Based on Stochastical Robust Optimization

无人机自主着舰是无人机飞行控制和仿真的难点,要求具有极高的控制精度和鲁棒性,需要直接面向被控对象的非线性模型和环境不确定性,给出基于非线性工作点校正的控制律结构设计方法。提出一种基于随机鲁棒优化的无人机自主着舰控制律设计方法,使用蒙特卡罗法对不确定域采样,引入拉丁超立方抽样法对控制律参数进行优化,使控制精度在该不确定域内满足设计指标。设计了用于着舰飞行仿真的小型无人机平台和控制系统,进行了物理飞行验证。实际飞行仿真表明,该方法对飞行器模型的结构和参数不敏感,对干扰有较好的抑制作用,同时控制律结构简单,易于工程实现。

Control and simulation of autonomy carrier loading is the key technology for unmanned air vehicles （UAV）. A new method of control law design for autonomy carrier landing is purposed, which is based on stochasti cally robust optimization technique. Latin Hypercube Sampling （LHS） method is introduced to optimize control pa rameters in which uncertain domain defined by Monte Carlo method. A small UAV platform and its control system is developed for real flight demonstration. The result of flight simulation shows that the new method is insensitive with model structure and parameters, and it has good robustness on outer interference. Furthermore, the control law can has simple structure, and can be easy to implement.