摘要
In this paper, a decentralized fault-tolerant cooperative control scheme is developed for multiple unmanned aerial vehicles(UAVs) in the presence of actuator faults and a directed communication network. To counteract in-flight actuator faults and enhance formation flight safety, neural networks(NNs) are used to approximate unknown nonlinear terms due to the inherent nonlinearities in UAV models and the actuator loss of control effectiveness faults. To further compensate for NN approximation errors and actuator bias faults, the disturbance observer(DO) technique is incorporated into the control scheme to increase the composite approximation capability.Moreover, the prediction errors, which represent the approximation qualities of the states induced by NNs and DOs to the measured states, are integrated into the developed fault-tolerant cooperative control scheme. Furthermore,prescribed performance functions are imposed on the attitude synchronization tracking errors, to guarantee the prescribed synchronization tracking performance. One of the key features of the proposed strategy is that unknown terms due to the inherent nonlinearities in UAVs and actuator faults are compensated for by the composite approximators constructed by NNs, DOs, and prediction errors. Another key feature is that the attitude synchronization tracking errors are strictly constrained within the prescribed bounds. Finally, simulation results are provided and have demonstrated the effectiveness of the proposed control scheme.
In this paper, a decentralized fault-tolerant cooperative control scheme is developed for multiple unmanned aerial vehicles(UAVs) in the presence of actuator faults and a directed communication network. To counteract in-flight actuator faults and enhance formation flight safety, neural networks(NNs) are used to approximate unknown nonlinear terms due to the inherent nonlinearities in UAV models and the actuator loss of control effectiveness faults. To further compensate for NN approximation errors and actuator bias faults, the disturbance observer(DO) technique is incorporated into the control scheme to increase the composite approximation capability.Moreover, the prediction errors, which represent the approximation qualities of the states induced by NNs and DOs to the measured states, are integrated into the developed fault-tolerant cooperative control scheme. Furthermore,prescribed performance functions are imposed on the attitude synchronization tracking errors, to guarantee the prescribed synchronization tracking performance. One of the key features of the proposed strategy is that unknown terms due to the inherent nonlinearities in UAVs and actuator faults are compensated for by the composite approximators constructed by NNs, DOs, and prediction errors. Another key feature is that the attitude synchronization tracking errors are strictly constrained within the prescribed bounds. Finally, simulation results are provided and have demonstrated the effectiveness of the proposed control scheme.
基金
the National Natural Science Foundation of China (Nos. 61833013, 61573282, and 61473229)
the Natural Science Foundation of Shaanxi Province, China (No. 2015JZ020)
the Natural Sciences and Engineering Research Council of Canada.
