Abstarct: This thesis presents two adaptive fuzzy control method baxsed on indirect approaches on underwater robots. In this method, the minimum run time, improvement adaptation parameters and the minimum error with respect to the purpose and direction is desirable. In the first method, due to undesired effects caused by the presence of external disturbances, a robust adaptive fuzzy controller has been used to implement the laws. Evaluation of this method on the depth of an underwater robot shows the tracking error and convergence speed parameter is improved. It also presents a new Lyapunov function to extract the implementation of the proposed rules, the system stability is shown. In the second approach, the control law to guarantee tracking performance and faster convergence structures in the presence of parameter uncertainty and external disturbances is defined. The series–parallel identification model to obtain the prediction error is introduced. In this way, the composite adaptive laws utilize both tracking and prediction errors. Overall stability of the closed-loop system is proved by Lyapunov analysis. The control system ensures that all signals involved are bounded. Simulation results on the depth of underwater robot with respect to its position in the XY plane has shown good performance.