基于RBF神经网络的挠性航天器分数阶滑模姿态控制  

Fractional order sliding mode attitude control of flexible spacecraft based on RBF neural network

作　　者：管萍[1] 罗宗樾 戈新生[1] GUAN Ping;LUO Zongyue;GE Xinsheng(School of Automation,Beijing Information Science and Technology University,Beijing 100192,China)

机构地区：[1]北京信息科技大学自动化学院,北京100192

出　　处：《飞行力学》2021年第6期68-74,共7页Flight Dynamics

基　　金：国家自然科学基金资助(11472058;11732005)。

摘　　要：针对挠性航天器大角度姿态机动问题,将神经网络、分数阶微积分与滑模控制相结合,构造基于神经网络的分数阶滑模控制并应用于挠性航天器的姿态控制中。首先,设计了分数阶滑模控制器,在滑模面的设计中采用分数阶微积分算子,利用分数阶微积分算子的记忆性和遗传特性,使挠性航天器的姿态角能够快速达到稳定;然后,将外界干扰项和挠性附件振动构成复合扰动,设计径向基(Radial Basis Function,RBF)神经网络逼近复合扰动,有效抑制了外界扰动和挠性振动对姿态产生的影响;最后,通过仿真证明了所设计控制系统的稳定性。仿真结果表明,所设计的基于RBF神经网络的分数阶滑模控制器能够使挠性附件振动迅速衰减,从而可使挠性航天器的姿态角快速达到期望值。Considering large angle attitude maneuver of flexible spacecraft,neural network,fractional calculus and sliding mode control are combined to constitute the fractional sliding mode control based on neural network,and applied to the attitude control of flexible spacecraft.Firstly,the fractional order sliding mode controller is designed.The fractional calculus operator is introduced into the design of the sliding mode surface,so that the attitude of flexible spacecraft can arrive the desired angle quickly by employing the memory and heredity of fractional calculus operator.Then,external disturbance and the flexible appendage vibration are regarded as compound disturbances.The RBF neural network is used to approximate the compound disturbance,so as to effectively suppress the external disturbance and the flexible vibration.Finally,the stability of the designed control system is proved.Simulation results show that the proposed fractional sliding mode controller can rapidly suppress the vibration of the flexible appendages,so that the attitude angles of flexible spacecraft fast track the desired value.

关 键 词：挠性航天器 振动抑制 姿态控制 分数阶滑模 RBF神经网络 

分 类 号：V448.22[航空宇航科学与技术—飞行器设计] TP183[自动化与计算机技术—控制理论与控制工程]