基于自适应与神经网络滑模的航空器主动控制  

Aircraft Active Control Based on Adaptive and Neural Network Sliding Mode

作　　者：贾格非 陈荣杰 钟福明 刘锡烨 JIA Ge-fei;CHEN Rong-jie;ZHONG Fu-ming;LIU Xi-ye(Northeastern University,Shenyang 110000,China)

机构地区：[1]东北大学,辽宁沈阳110000

出　　处：《航空计算技术》2025年第2期77-82,共6页Aeronautical Computing Technique

摘　　要：聚焦于应用压电驱动器实现对夹紧矩形膜结构的大幅非线性振动的主动控制。基于膜结构非线性动力学模型,采用自适应控制策略和引入滑模控制器与径向基函数神经网络的结合,通过Matlab数值仿真验证了控制方法的有效性。研究结果表明,自适应控制和变结构神经网络控制成功抑制了膜结构振动,在面对不同激励条件下均能快速趋近参考模型的动态响应。并引入卡尔曼观测器有效抑制了测量噪声,降低了控制成本。为航空航天领域中薄膜结构振动控制提供了可靠的解决途径。This paper focuses on the application of piezoelectric actuator to realize the active control of large amplitude nonlinear vibration of clamped rectangular membrane structures.Based on the nonlinear dynamic model of the membrane structure,the adaptive control strategy and the combination of sliding mode controller and radial basis function neural network are adopted to verify the effectiveness of the control method through Matlab numerical simulation.Study results show that adaptive control and variable structure neural network control successfully suppress the vibration of the membrane structure,and can rapidly approach the dynamic response of the reference model under different excitation conditions.The Kalman observer is introduced to effectively suppress the measurement noise and reduce the control cost.This provides a reliable solution for the vibration control of thin film structures in the aerospace field.

关 键 词：大振幅振动 滑模控制 自适应控制 径向基函数神经网络 卡尔曼观测器 

分 类 号：TP13[自动化与计算机技术—控制理论与控制工程]