输入输出受限的无人机防滑刹车系统容错控制

针对无人机防滑刹车系统工作过程中同时出现系统输出滑移率稳定区域受限、控制输入饱和与刹车执行机构故障的多重约束问题,提出了一种基于障碍Lyapunov形式的自适应神经网络反演容错控制器的设计方法。当刹车执行机构发生故障时,通过自适应神经网络补偿刹车系统中的非线性及不确定项。根据反演设计原理,应用神经网络输出设计相应的容错控制律,同时,在控制器的设计中引入鲁棒切换控制项,优化系统快速容错的暂态性能。首先本文设计的容错控制器无需精确获取执行机构在线故障的重构信息,也能使刹车闭环系统能够快速稳定,然后基于Lyapunov方法分析了系统的稳定性,最后通过数值仿真结果表明,所提出的容错控制算法能够有效地保证刹车执行机构故障时控制系统的稳定性和有效性。

Fault-tolerant control of UAV anti-skid braking system with input and output constraints

In this paper,a method of adaptive neural network backstepping fault-tolerant control,based on barrier Lyapunov function,is proposed for anti-skid braking system in the presence of slip-ratio constraint,control input saturation and partial loss of actuator effectiveness.The neural network can more accurately approximate the unknown nonlinearity in order to compensate the effect of actuator fault,and the great robustness to actuator fault is guaranteed.In this approach,the output of neural network is used to design the backstepping controller to achieve fault-tolerant control and uncertainty compensation,and a robust term is employed to optimize the transient performance of braking system.Firstly,the closed-loop fault-tolerant control system could be stable without the reconfiguration value of actuator fault in real time.Then,the stability of the system is analyzed based on the Lyapunov method.Finally,the numerical simulation results show that the proposed fault-tolerant control scheme can effectively guarantee the stability and effectiveness of the control system when the actuator happens faulty.