Fault Tolerant Control of Quadrotor UAV Based on Support Vector Machine

Abstract

Aiming at the problem of partial failure of Quadrotor uav actuator, a collaborative design method of optimizing PID parameters of controller, fault diagnosis and fault-tolerant control is proposed. Firstly, the off-line optimization of PID parameters of four-rotor uav controller is carried out by using ant colony algorithm. According to the severity of actuator fault, the fault is partitioned and the corresponding fault tolerant controller is designed. In the case of actuator failure, support vector machine (SVM) is used to diagnose the fault, and then the fault of the fault tolerant controller is compensated by online scheduling. Finally, the feasibility and effectiveness of the proposed method are verified on the matlab experimental platform.