Control of a teleoperating robotic arm subject to time delays.

摘要

Robotics in Minimally Invasive Surgeries (MIS) can be explored to its full potential if there is complete transparency between the surgeon and the virtual environment. In the setup, haptic device which conveys force from the surgeon to the surgery site and vice verse acts as the master station and robotic arm at the surgery site acts as the slave robot. In surgeries carried out remotely, time delays in the communication channels can cause commanded velocity/force to overshoot before the surgeon can even pull out during a task causing temporary or permanent damage during the surgery.;Since the physical capability of the transmission channels cannot be improved beyond a certain point there will always be time delays no matter how small. Therefore the work presented attempts to minimize the effects of time delays for robot manipulators having more than single degree of freedom. This will allow more delicate and complex surgeries to be performed with increased precision, dexterity and control. The controller uses a stable Lyapunov based backstepping design with tuning functions to filter out high frequency signals and smooth out the control forces. Neural networks will be used for estimating the rate of change of external forces and nonlinear system dynamics. This design guarantees stability and inherently adds robustness to account for unknown environment.;Simulations are carried out for a 1-DOF and 2-DOF robot manipulators and the results show a substantial improvement in the performance over the direct force application. The controller also performs reasonably well with higher time delays compared to direct force method where large delays destabilize the robot.