Approximate Optimal Adaptive Prescribed Performance Fault-Tolerant Control for Autonomous Underwater Vehicle Based on Self-Organizing Neural Networks

This paper investigates the problem of performance-optimized fault-tolerant tracking control of autonomous underwater vehicles (AUVs) with unknown disturbances and actuator faults. To ensure the prescribed tracking accuracy and save the consumption of control inputs, the prescribed performance function is used to constrain the tracking error, and a performance optimization strategy is designed to plan the virtual control inputs based on the optimized cost function. To deal with the uncertainties caused by unknown disturbances and additive faults, the self-organizing neural network (NN) is employed to estimate these composite unknown dynamics. The structure of the NN can be dynamically adapted according to the complexity of the fitted function, which avoids excessive computational burden. A fault-tolerant estimation law is designed to minimize the impact of AUVs when they are subject to actuator partial multiplicative faults, and the actual fault-tolerant controller is designed based on the dynamic event triggering strategy, which reduces the updating frequency of the control inputs while guaranteeing the safe operation of the system. The proposed control framework enables the AUV to ensure the prescribed tracking accuracy during the tracking process, improving tracking safety and reducing energy consumption for accomplishing the task. Finally, a simulation experiment verifies the effectiveness of the designed control scheme.