Output fault-tolerant control of underwater vehicle based on adaptive sliding mode control under multiple constraints

Abstract

In order to improve the stability of the underwater vehicle, the output fault-tolerant control method of the underwater vehicle based on adaptive sliding mode control is designed on the basis of setting multiple constraints. After the dynamic analysis of the underwater vehicle and its thruster, the output dimension constraint, input amplitude constraint, and state amplitude constraint were set. Based on these three constraints, the recursive cerebellar neural network is used to identify the time-varying and nonlinear unknown faults in the underwater vehicle online, and a deep convolution neural network is used to detect the known faults of the underwater vehicle. The error correction output code is then combined with a support vector machine to classify the detected faults. Finally, the output fault-tolerant control of an underwater vehicle is realized based on adaptive sliding mode control. The experiment shows that after the application of this method, because the recursive cerebellar neural network and the deep convolution neural network can estimate the size of the fault information in time, the sliding mode fault-tolerant controller can readjust the output of the controller according to the result of the fault identification to offset the effect of the fault on the robot, so that the robot can continue to maintain the original stable operation state. Even if the thruster has faults, this method can adjust the output of the thruster in time and reconstruct and recover the total control output, so that the actual output thrust is very close to the expected value, to achieve fault-tolerant control of the underwater vehicle output.