Adaptive Fault Diagnosis And Fault Tolerant Control And Its Applications To Unmanned Aerial Vehicle

With the development of science and technology,the scale of modern control system is becoming larger and larger,and the number of system components is becoming more and more.At the same time,the operational environment of system is becoming more and more complex as well.In the system components,such as actuators,sensors and other components,the failure possibility will also increase.In order to eliminate the impact of the fault in time,the fault-tolerant control methods are proposed.Fault-tolerant control is divided into two types:active fault-tolerant control and passive fault-tolerant control.Compared with the passive fault-tolerant method,the active fault-tolerant control has less conservatism.In active fault-tolerant control,fault diagnosis and estimation are two important parts,and are also the basis of fault-tolerant control.In recent years,fault diagnosis and fault-tolerant control of dynamic systems have attracted wide attention.Considering various failures,researchers put forward a variety of fault diagnosis and fault-tolerant control methods for various systems,respectively.Although there are abundant results in fault diagnosis and fault-tolerant control,the fault diagnosis and fault-tolerant control of the dynamic systems including the complex nonlinear uncertain UAV system studied in this article should be further investigated.In this article,we investigated the fault diagnosis and fault-tolerant control for several uncertain nonlinear systems with actuator or sensor faults.Considering the influence of unmodeled dynamics and external disturbance on the controlled system,taking advantage of the approximation ability of neural networks,and combining sliding mode control,adaptive control and other control techniques the corresponding fault detection methods,decision mechanisms are designed,the fault estimation algorithms are proposed to estimated the faults.And the corresponding adaptive fault-tolerant control schemes are proposed.Then,the UAV pitching subsystem is used to show of the efficiency of the proposed scheme.The research in this article is summarized as follows:Firstly,for a class of uncertain nonlinear strict-feedback systems with actuator faults and unknown disturbance,we consider the actuator fault as a bias fault and investigate the fault detection and estimation problems.By combining the equivalent output sliding mode observer and nonlinear fault diagnosis theory,based on adaptive sliding mode observer,a fault diagnosis scheme is designed using the approximation ability of neural network.First,an adaptive sliding mode observer is constructed to generate residual to detect faults.Then,a fault estimation algorithm is designed to estimate the fault.By theoretical analysis,it is proven that the proposed algorithms effectively detect and estimate the fault and ensure the observation error and fault estimation errors converge to zero in finite time.Finally,the simulation results of UAV pitching subsystem verify the effectiveness of the designed scheme.Secondly,for a class of nonlinear systems with unmodeled dynamics and sensor failure,the problem of fault detection and estimation is investigated.Combining the unmodeled dynamics with nonlinear fault diagnosis theory,an adaptive fault diagnosis scheme is proposed for the system.First,a dynamic signal is introduced to control the unmodeled dynamics in the system.Then,two algorithms are designed to realize fault detection and estimation,respectively.Through theoretical analysis,it is proved that the algorithm can detect and estimate faults effectively,and ensure that the observation error and fault estimation error converge to zero in finite time.Finally,the UAV pitching subsystem is used for simulation study.The simulation results show the effectiveness of the proposed scheme.Thirdly,for a class of nonlinear systems with output feedback and actuator failures,the fault-tolerant control problem is investigated.By constructing K filter to reconstruct system state,an adaptive backstepping fault-tolerant control scheme is proposed by using the approximation ability of neural network.The Nussbaum function is introduced to solve the problem of unknown symbol of control gain.Through the Lyapunov stability theory,it is proved that the closed loop control system is semi globally uniformly terminated bounded,and the tracking error converges to zero.Finally,UAV pitching subsystem in the cruise phase is given to verify the effectiveness of the proposed scheme.