Force Feedback in the da Vinci Surgical Robot
Author
Bordalba Llaberia, Ricard
Term
4. term
Education
Publication year
2016
Submitted on
2016-06-01
Pages
141
Abstract
Robotic surgery has become important in minimally invasive procedures, yet limited or absent haptic feedback remains a barrier to wider adoption. This thesis investigates how to provide force feedback on the da Vinci Surgical Robot without adding force sensors. A kinematic and dynamic model of the AAU da Vinci manipulator is developed, including friction, and a simulation environment with PD control is set up. Robot parameters are identified using designed excitation trajectories, signal processing, and both linear least squares and particle swarm optimization. Based on the identified model, an extended Kalman filter is designed to estimate external contact forces from joint measurements and map them to end-effector forces through the robot’s Jacobian. The approach is evaluated in simulation and on the AAU robot, and a haptic device is used to verify the estimated forces when the robot is in contact with tissue and to convey force cues to the operator. The work also addresses haptic system stability and implementation in a ROS-based environment. Overall, the results indicate a feasible path to recover contact force information for surgical teleoperation without dedicated force sensors.
Robotkirurgi er blevet vigtig i minimal invasiv kirurgi, men begrænset eller manglende haptisk feedback hæmmer udbredelsen. Dette speciale undersøger, hvordan kraftfeedback kan tilføjes da Vinci-operationsrobotten uden at montere kraftsensorer. En kinematisk og dynamisk model af AAU’s da Vinci-manipulator udvikles, inklusive friktion, og der etableres et simulationsmiljø med PD-regulering. Robotparametre estimeres via designede eksperimenter, signalbehandling samt både lineær mindste kvadraters-estimation og particle swarm optimization. Med udgangspunkt i den identificerede model designes et udvidet Kalman-filter, der estimerer eksterne kontaktkræfter ud fra ledmålinger og omregner dem til endeeffektorkræfter via robotens Jacobian. Tilgangen afprøves i simulation og på AAU-robotten, og en haptisk enhed bruges til at verificere de estimerede kræfter, når robotten er i kontakt med væv, samt til at give operatøren kraftindtryk. Arbejdet omfatter også stabilitetsanalyse af det haptiske system og implementering i et ROS-baseret miljø. Samlet peger resultaterne på en gennemførlig vej til at genskabe kontaktkraftinformation i kirurgisk teleoperation uden dedikerede sensorer.
[This apstract has been generated with the help of AI directly from the project full text]
Keywords