基于模糊神经网络扰动观测器的RLV动态逆再入姿态控制

Dynamic inversion control for RLV reentry attitude based on fuzzy-neural disturbance observer

针对于RLV再入段飞行参数不确定性大和外界干扰的问题,提出基于B样条模糊神经网络扰动观测器的RLV动态逆控制方法。非线性动态逆控制器面向标称系统设计,保证控制性能。把参数不确定性和外扰看成是系统的总扰动,设计B样条模糊神经网络扰动观测器进行估计,构造补偿控制信号加入到系统中。通过引入气动参数偏差和干扰力矩对所设计的控制方法进行仿真,并与传统动态逆方法的控制器进行对比,研究结果表明:本文的RLV姿态控制方法对于参数不确定性和外扰具有较强的鲁棒性,弥补了动态逆方法要求模型精确这一不足。

In order to minimize the effects of large parameter uncertainties and external disturbance,a method of RLV reentry attitude control based on fuzzy-neural disturbance observer was proposed.Nonlinear dynamic inversion controller was focused on nominal system and ensures control performance.Parameter uncertainties and external disturbance were regarded as total system disturbance,which was estimated by B-spline fuzzy-neural disturbance observer and added to system controller as compensating control.Considering aerodynamic parameter uncertainties and disturbance moments,simulations were made to invalidate the fuzzy-neural dynamic inversion controller compared with dynamic inversion controller.The results show that the proposed controller has strong robustness to parameter uncertainties and external disturbance and makes up for dynamic inversion method.