An improved scheme for eliminating the coupled motion of surgical instruments used in laparoscopic surgical robots

Abstract

Considering the nonlinear characteristics such as backlash hysteresis and coupled motion commonly exist in cable-driven mechanism of laparoscopic surgical robot end-effector, it is a great challenge to control the motion of robotic end-effector precisely during the surgical procedure. Due to the effects of coupled motion, the surgical end-effector will not move accurately as surgeons expected. Previous studies mostly focused on the design of special compensation mechanisms and software compensation algorithms to solve coupled motion problem. However, these approaches are limited because the backlash hysteresis is ignored and the mechanism of end-effector is restricted. This paper shows an improved scheme to eliminate the coupled motion of end-effector and reduce the position tracking error. The proposed decoupling scheme is conducted in three stages. Firstly, the time and frequency domain information of the driving motor current and the motion information of surgical instrument are extracted in real-time. Thereafter, a feedforward neural network is designed to identify the movement stage of end-effector. Finally, a prediction model is designed to predict the coupling error, after that the coupling error can be eliminated by using feedforward compensation control. An experimental platform was set up to verify the effectiveness of the proposed control scheme, and the results of corresponding comparative experiments revealed that the proposed strategy can substantially improve the tracking accuracy.