The existing fault-tolerant control methods mostly use pseudo inverse reconstruction for fault-tolerant controller design, but in complex underwater environments, they are easily affected by factors such as water flow, water temperature, and water pressure, resulting in the controller being unable to respond to faults in a timely manner. Research on fault-tolerant control method for underwater robot thrusters based on MEKF algorithm. Firstly, utilizing the recursive and state estimation capabilities of the MEKF algorithm, an accurate coordinate system suitable for underwater robots is constructed. Subsequently, the hydrodynamic effects experienced by underwater robots during motion were analyzed, including fluid resistance, lift, lateral forces, and their dynamic changes. Finally, design a thrust allocation strategy for AUV thruster modeling criteria. This strategy predicts the changes in the robot"s navigation position and dynamically adjusts the thrust output of each thruster by combining fault-tolerant control ideas, in order to achieve effective control of the robot"s deviation angle. The experimental results show that the application of the above-mentioned fault-tolerant control method can effectively control the deviation angle of the robot"s motion in underwater direct flight, heave motion, and lateral motion. The power effect of the thruster can be effectively guaranteed, which helps to maintain the mobility of the underwater robot.