Fault-tolerant controller based on artificial intelligence combined with terminal sliding mode and pre-described performance function applied on ship dynamic position stabilization systems

Abstract

The presented study, develops an intelligent fault-tolerant controller to automatically control and maintain the position of a ship. The controller possesses the capabilities of holding the ship steady at a fixed position, automatically transferring the ship from one point to another, and tracking a target in the presence of various disturbances, uncertainties in the ship model, and thruster faults. A backstepping control approach based on the Terminal Sliding Mode method is employed to improve the accuracy of the controller. The aforesaid method guarantees time-limited convergence without relying on the initial conditions of system. A pre-described performance function (PPF) is utilized to ensure the performance in transient states and the calculations, which leads to the specific equations achievement for the study, which guarantees the system's stability. Environmental disturbances, uncertainties, and thruster faults are estimated using a Radial Basis Neural Network, allowing the controller to adapt to changes in the ship's environment and system uncertainties. The designed controller is proven to ensure the tracking error remains within a predefined range and can stabilize the ship system within a constant time without being influenced by the initial conditions. This stability is achieved in the presence of environmental disturbances, model uncertainties, and thruster faults, and the tracking error converges to a neighborhood near the origin. The controller is verified using a ship mode in the MATLAB environment, and the results show that the designed controller effectively controls the ship's position. The theoretical findings are successfully verified through these simulation tests.