基于并行学习鲁棒自适应的多操纵面飞机故障估计方法  

A Robust Adaptive Fault Estimation Method for Multi-control Surface Aircraft Based on Parallel Learning

作　　者：刘静[1] 丛岩[2] 赵迪 LIU Jing;CONG Yan;ZHAO Di(Hunan Vocational College of Engineering,Changsha 410151,China;Aviation University of Air Force,Changchun 130000,China;Hunan First Normal University,Changsha 410000,China)

机构地区：[1]湖南工程职业技术学院,长沙410151 [2]空军航空大学,长春130000 [3]湖南第一师范学院,长沙410000

出　　处：《火力与指挥控制》2021年第1期50-55,共6页Fire Control & Command Control

基　　金：国家“十三五”重点研发计划项目(2017YFD0701103);湖南省教育厅科学研究资助项目(18C1311)。

摘　　要：针对多操纵面飞机执行机构多发、频发故障的严峻问题,提出了一种鲁棒自适应故障估计方法,以得到精确故障信息。通过低通滤波,结合并行学习技术,引入了一组系统变量的滤波变量;基于上述滤波变量,构建了故障估计直接误差向量,进而设计了鲁棒自适应故障估计器;基于Lyapunov稳定性理论证明了所设计估计器的收敛性和鲁棒性。以ADMIRE飞机为对象,在无/有扰动两种仿真情形下,所提方法对单个故障的估计时间为1.6 s,对并发故障的估计时间为3 s,验证了本设计故障估计方法的鲁棒性和快速性。A robust adaptive fault estimation method is proposed to solve the severe problem of multiple and frequent faults of aircraft actuators with multiple control surfaces in order to obtain accurate fault information.Firstly,a set of filtering variables of system variables is introduced by low-pass filtering combined with parallel learning technology.Secondly,based on the above mentioned filtering variables,a direct error vector of fault estimation is constructed,and then a robust adaptive fault estimator is designed.Finally,the convergence and robustness of the designed estimator are proved based on Lyapunov stability theory.Taking ADMIRE aircraft as an object,the proposed method can estimate the time of a single fault as 1.6 seconds and a parallel fault as 3 seconds in two simulation scenarios without or with disturbance,which verifies the robustness and rapidity of the proposed fault eseimation method.

关 键 词：多操纵面飞机 执行机构 故障估计 并行学习 自适应 

分 类 号：TJ85[兵器科学与技术—武器系统与运用工程] V249.1[航空宇航科学与技术—飞行器设计]