Research On Fault Diagnosis And Fault Tolerant Control Of Quadrotor UAV Flight Control System

With the rapid development of aerospace science and technology,spacecraft represented by unmanned aerial vehicles are becoming more and more complex.People have also put forward higher requirements for the stability,safety and reliability of spacecraft.The fault diagnosis and fault-tolerant control technology of the UAV flight control system can improve the autonomous operation capability of spacecraft.In addition,the complex and changeable space environment will inevitably affect the UAV,and its long-term operation will also cause the system to produce different types of failures,of which 44% are caused by actuator faults.Therefore,studying fault diagnosis of actuator and the faulttolerant control of flight control system is of great significance for extending service life of the equipment and improving ability of system to operate stably.The unique advantages of vertical take-off and landing,free hovering,compactness and flexibility of rotary-wing UAVs are widely used in military and civilian fields.In this doctoral dissertation,Quadrotor UAV as a representative,in view of the influence of external disturbance,model uncertainty and actuator failure on the flight control system operation safety,the problem of fault diagnosis,robust tracking control and fault-tolerant control in the flight control system is studied.The main contents are as follows:Based on the influence of external disturbances and actuator faults on system modeling,the dynamic model principles of typical UAVs are analyzed in depth and the mathematical model of UAV flight control system is optimized.Firstly,analyze the types of actuator faults and establish a mathematical model of typical failures.Then establish a dynamic model of UAV under failure mode,which lays a theoretical foundation for follow-up research of UAV fault diagnosis and fault-tolerant control.Aiming at the problem of actuator fault diversity and strong noise interference,a fault feature extraction algorithm based on variational modal decomposition and hierarchical fuzzy entropy is proposed.Realize feature extraction of fault signals at different frequencies,theoretically improve the accuracy of fault feature extraction.By optimizing the Alexnet neural network algorithm to classify and identify the bearing signal fault characteristics,improve the accuracy of fault diagnosis.The test results show that,compared with convolutional neural network and Googlenet network,this method can not only identify the type of failure with an accuracy rate of 97.92%,but also identify the severity of failure with an accuracy rate of 94.73%.Aiming at the tracking control problem of Quadrotor UAV under external disturbance and model uncertainty,the tracking control of Quadrotor UAV under fixed disturbance and random disturbance is realized.Aiming at the situation that the disturbance is a fixed value,an improved backstepping sliding mode tracking control algorithm is proposed,which solves chattering problem of sliding mode control itself and improves the anti-disturbance characteristics of system.Furthermore,considering that disturbance and model uncertainty are random values,a new type of robust tracking control optimization algorithm combining fuzzy control and PID control is proposed,and the random part of system is observed and compensated in real time through the extended state observer.The results show that the response time of Quadrotor UAV yaw control subsystem is reduced by nearly 50% compared with the traditional PID control algorithm,and the overshoot is also greatly reduced.This method improves the response speed and tracking performance of system.Based on the analysis of influence of external disturbance on system,the problem of fault-tolerant control of actuator for the Quadrotor UAV is studied.Aiming at the fault-tolerant control problem of additive faults such as actuator deviation,an improved fractional-order power sliding mode fault-tolerant control algorithm is proposed,which solves the problem of traditional fault-tolerant control in realizing system asymptoticity stable and makes the system reach the balance point in a limited time.And adaptive control technology compensates for influence of interference on system,which improves stability of system.Aiming at the fault-tolerant control problem of multiplicative faults such as actuator failure,a new non-singular terminal sliding mode control algorithm is designed.Experiments show that this method not only eliminates the singular phenomena in attitude and position control,but also makes the flight control system reach a stable state within a limited time.In addition,an adaptive control strategy is used to compensate for system disturbances and uncertainties to further improve the performance of controller.