As the micromanipulator of surgical robots works in a narrow space, it is difficult to install any position sensors at the end, so the position control and position detection cannot be accurately performed. A position estimator based on the parameter autonomous selection model is proposed to estimate the end position indirectly. Firstly, a single joint principle prototype and a position estimator model are established through the 4-Dof driving scheme of the micromanipulator and the cable-driven model. Secondly, the proposed parameter change model is combined with the parameter selection method to form a parameter autonomous selection model. Finally, a position estimator based on the parameter autonomous selection model is established. The experimental results show the maximum estimation error of the position estimator is 0.1928°. Compared with other position estimation methods, the position estimator proposed in this paper has higher accuracy and better robustness, which lays a foundation for the full closed-loop control of micromanipulator position.