Based on the problems in the dynamic modeling of the robot arm and the algorithm itself in the actual industry, the robot arm jitters and the repeated positioning accuracy is not high, and it is difficult to achieve precise control. This paper proposes a robot arm based on two degrees of freedom (2DOF). The research method of adaptive fuzzy backstepping control. Firstly, model the kinematics and dynamics of the manipulator, and then design the adaptive fuzzy backstepping controller to observe the tracking of the manipulator under disturbance. Experiments show that with the passage of simulation time, the actual trajectory of the end of the robotic arm keeps approaching the desired trajectory. When the simulation time is 5s, the error on the x-axis is reduced to 6.09*10 −3 m, and the error on the y-axis is reduced to 8.28*10 −4 . Until the end of the simulation, the error on the x-axis is reduced to 2.94*10 −3 m, and the error on the y-axis is reduced to 4.42*10 −4 . It can be seen that the algorithm has a good control effect for the 2DOF manipulator, and shows the superiority of the algorithm.