中文摘要：

为解决反馈线性化（FL）方法的控制性能过于依赖精确系统模型的问题，提出了一种自适应的非线性控制策略，并应用于升力式飞行器的控制器设计．利用模糊小脑模型神经网络（FCMAC）良好的非线性逼近能力和自学习能力，设计了基于FCMAC的干扰观测器，对模型的不确定性和干扰进行在线估计．其网络权值更新规则采用李亚普诺夫方法设计，保证了闭环系统跟踪误差和干扰观测误差的有界．6自由度仿真结果显示该控制方案可实现姿态角对制导指令的稳定、快速跟踪，具有良好的鲁棒性．

英文摘要：

The performance of feedback linearization （FL） depends largely on the accuracy of systems. Thus, an adaptive FL strategy is proposed to design flight controller for lifting vehicle. Fuzzy cerebella model articulation control （FCMAC） disturbance observer is developed to deal with the uncertainties and disturbance in the nominal model by using nonlinearity approximating and self-learning abilities. Weights Updating laws are derived by Lyapun.ov method, which could guarantee the boundedness of all signals in the entire system. Six freedom degree simulation results show that the improved tracking approach of attitude angles, which demonstrates its good robustness.