This study addresses the engineering problem of underwater tracked dredging robots sliding due to insufficient adhesion when dredging caissons in bridge construction. Therefore, a trajectory tracking controller based on the barrier Lyapunov function (BLF) is designed.

The algorithm considers the non-coincidence of the centroid and geometric center, influence of unknown bounded disturbances and system dynamic uncertainty to establish kinematic and dynamic motion models. A terminal sliding mode observer (TSMO) is used to approximate external disturbances and system dynamic uncertainty in finite time. The stability of the control system is verified by time-varying symmetric finite time BLF stability analysis, and control failure is prevented by limiting the velocity state of the system.

The simulation results show that the robot reaches the desired trajectory smoothly and quickly under the control of the designed controller.

The proposed algorithm can limit the velocity state of the robot system to an interval that conforms to the actual engineering practice.