FAULT DIAGNOSIS AND CONTROL FOR FOUR-ROTOR UAV BASED ON IMPROVED RBF NEURAL NETWORK

Aimed at the actuator failure of quadrotor UAV often affects flight , a fault diagnosis and fault-tolerant control method based on improved neural network is designed. The UAV fault model was established. The RBF neural network was improved by introducing weight vector adaptive law , center vector adaptive law and adjusting parameters. A fault diagnosis and fault-tolerant control method were designed using improved neural network. The simulation results show that the proposed improved method has better stability and accuracy than the traditional fault diagnosis and fault-tolerant control methods. The fault diagnosis maximum error is only 0.01 , and the fault-tolerant control tracking maximum error is only 0.3°, which greatly improves the control effect of UAV.