摘要
针对高超声速再入飞行器非线性程度高、参数不确定性大、快时变等特点,提出一种基于神经网络特征模型的自适应滑模姿态控制方案。首先,采用现有特征建模方法,将对象模型中的非线性、时变不确定性压缩至特征参量中;进一步,结合模糊神经网络,将快时变特征显式地表征在特征模型中,使得待估计的特征参量具有时不变特性,从而易于其自适应律的设计。然后,在该神经网络特征模型的框架下,设计递推形式的自适应滑模控制律,以进一步提高飞行控制系统的鲁棒性。最后,通过仿真校验了所提出控制方法的正确性和有效性。
An adaptive sliding mode control scheme based on the characteristic model with a neural network is proposed for a class of hypersonic reentry vehicles with strong nonlinearities and time-varying uncertainties. Based on the existing characteristic modeling method,the nonlinearities and time-varying uncertainties are integrated into several characteristic parameters which inherit the time-variation property from the aerodynamics. Further, the time-variation property is expressed explicitly in the characteristic model using the neural network,which obtains the time invariant characteristic parameters and makes their adaptive law design feasible. And then the adaptive sliding mode control law with the recursive form is constructed to enhance the system robustness. The numerical simulations have demonstrated the effectiveness of the control scheme proposed.
作者
常亚菲
姜甜甜
CHANG Ya-fei;JIANG Tian-tian(Beijing Institute of Control Engineering,Beijing 100190,China;Seienee and Teehnology on Space Intelligent Control Laboratory,Beijing 100094,China)
出处
《宇航学报》
EI
CAS
CSCD
北大核心
2018年第8期889-899,共11页
Journal of Astronautics
基金
国家自然科学基金(61333008
61603038)
关键词
高超声速再入飞行器
不确定性
时变
特征建模
自适应递推滑模控制
Hypersonic reentry vehicle
Uncertainty
Time-variation
Characteristic modeling
Adaptive recursive sliding mode control
