A Dual-user Teleoperation System with Adaptive Authority Adjustment for Haptic Training Fei LIU*, Arnaud Lelevé, Damien Eberard, Tanneguy Redarce fei.liu@insa-lyon.fr Laboratory Ampère University of Lyon, INSA de Lyon July 23, 2015 1 / 16
Motivation: dual-user haptic training system Traditional Training Medical experiments : phantoms, cadavers, animals → costly and lack realism from human body Block box haptic device : difficult for performance evaluation Virtual Reality based simulators : no haptic feedback Our Motivation Dual-user : trainer and trainee together Haptic feedback 3D haptic rendering of the environment 2 / 16
Background Existing Architectures No full haptic feedback [Nudehi-2005] Linear dynamics [Khademian-2011] Nonlinear dynamics , no transparency analysis [Ghorbanian-2013] No delay [Khademian-2011, Jian-2014] Stability/Passivity Method H ∞ [Nudehi-2005] Llewellyn’s unconditional stability criterion [Khademian-2011] Zeheb-Walach criteria [Razi-2014] Lyapunov-Krasovski function [Ghorbanian-2013] Raisbeck’s passivity criterion [Jian-2014] Limitation for: Nonlinear dynamics, Delays (data losses), Mode Switching Stability 3 / 16
Background Port-Hamiltonian Modeling Approach [B. Maschke-1992, S. Stramigioli-2001] Energy based approach : energy variable pairs → force, velocity Hamiltonian function : formed as a Lyapunov function → passivity IPC controller : intrinsically passive controller, modeled within PH form Composability : inherits Hamiltonian structure by interconnection (Dirac structure) 4 / 16
Our Topic Modeling System interconnection : Port-Hamiltonian Approach ⇒ simplifies passive system design Communication channel : Scattering Transformation ⇒ to keep delayed transmission passive Control Controller : Intrinsically Passive Controller (IPC) ⇒ modeled within port-Hamiltonian form, passive Authority management : Online Authority Adjustment (OAA) Transparency : with/without delays 5 / 16
Contributions Contributions 6 / 16
Overall System Architecture Bilateral teleoperation [C. Secchi-2003] Dual-user teleoperation: dirac structures 7 / 16
Overall System Architecture Bilateral teleoperation [C. Secchi-2003] Dual-user teleoperation: dirac structures 7 / 16
Overall System Architecture 7 / 16
Blue: trajectory of trainee Trainer Control Authority Adaptive Virtual Boundary Virtual Boundary Force Adaptive Control Authority Position 1 0 Red: Adaptive Virtual Boundary Black: trajectory of trainee Adaptive Authority Adjustment (AAA) Adaptive Authority Adjustment (AAA) Overrule Function � α a , overrule = False 8 / 16 α = α o , overrule = True
Blue: trajectory of trainee Trainer Control Authority Adaptive Virtual Boundary Virtual Boundary Force Adaptive Control Authority Position 1 0 Red: Adaptive Virtual Boundary Black: trajectory of trainee Adaptive Authority Adjustment (AAA) Adaptive Authority Adjustment (AAA) Overrule Function � α a , overrule = False 8 / 16 α = α o , overrule = True
Master 1 Master 2 Slave Communication Channel ? ? Passivity Analysis Passivity Propositions [F . Liu, 2015-MESROB] Modulated flow sources MSf → passive with bounded energy generation Variation control authority → closed loop switching system passive 9 / 16
HAPTIC DEVICE SIMULINK Experiment Setup 10 / 16
Experiment Results: AAA Behavior Trainer Trainee Slave Wall Target 0.8 A B C D E F G H I 0.6 Joint Angle (rad) 0.4 t =50 : 9 s 0.2 t =37 : 2 s t =54 : 8 s t =22 : 4 s t =18 : 4 s t =43 : 5 s 0.0 t =59 : 0 s t =13 : 8 s −0.2 0 10 20 30 40 50 60 70 Time(s) 1.2 A B C D E F G H I 1.0 0.8 Alpha 0.6 0.4 0.2 0.0 0 10 20 30 40 50 60 70 Time(s) 11 / 16
Experiment Results: AAA Behavior Period A (Training Mode) : Trainer full authority (overruled) Trainer Trainee Slave Wall Target 0.8 0.6 Joint Angle (rad) 0.4 0.2 0.0 t =13 : 8 s −0.2 0 2 4 6 8 10 12 14 Time(s) 1.2 1.0 0.8 Alpha 0.6 0.4 0.2 0.0 0 2 4 6 8 10 12 14 Time(s) 11 / 16
Experiment Results: AAA Behavior Period B (Evaluation Mode) : Trainee full authority Trainer Trainee Slave Wall Target 0.8 0.6 Joint Angle (rad) 0.4 0.2 t =18 : 4 s 0.0 t =13 : 8 s −0.2 14 15 16 17 18 19 Time(s) 1.2 1.0 0.8 Alpha 0.6 0.4 0.2 0.0 14 15 16 17 18 19 Time(s) 11 / 16
Experiment Results: AAA Behavior Period C (Guidance Mode) : Adaptive Authority Adjustment (AAA) Trainer Trainee Slave Wall Target 0.8 0.6 Joint Angle (rad) 0.4 0.2 t =22 : 4 s t =18 : 4 s 0.0 −0.2 18 19 20 21 22 23 Time(s) 1.2 1.0 0.8 Alpha 0.6 0.4 0.2 0.0 18 19 20 21 22 23 Time(s) 11 / 16
Experiment Results: AAA Behavior Period G (Wall Contact Case) : Adaptive Authority Adjustment (AAA) Trainer Trainee Slave Wall Target 0.8 0.6 Joint Angle (rad) 0.4 t =50 : 9 s 0.2 t =54 : 8 s 0.0 −0.2 50.5 51.0 51.5 52.0 52.5 53.0 53.5 54.0 54.5 55.0 Time(s) 1.2 1.0 0.8 Alpha 0.6 0.4 0.2 0.0 50.5 51.0 51.5 52.0 52.5 53.0 53.5 54.0 54.5 55.0 Time(s) 11 / 16
Experiment Results: AAA Behavior Period H (Wall Off-contact Case) : Adaptive Authority Adjustment (AAA) Trainer Trainee Slave Wall Target 0.8 0.6 Joint Angle (rad) 0.4 0.2 t =54 : 8 s 0.0 t =59 : 0 s −0.2 54 55 56 57 58 59 Time(s) 1.2 1.0 0.8 Alpha 0.6 0.4 0.2 0.0 54 55 56 57 58 59 Time(s) 11 / 16
0 10 10 15 20 25 30 0 5 15 Environment 20 25 30 Trainer Trainee Slave Wall 5 -0.2 Trainee 0.1 Joint Angle (rad) 0.7 0.6 0.5 0.4 0.3 0.2 0.0 -0.1 -0.1 Hand/Environment Torque (N*m) 0.5 0.4 0.3 0.2 0.1 0.0 Trainer Experiment Results: Wall Contact Transparency Wall Contact Force Tracking 12 / 16
0 10 10 15 20 25 30 0 5 15 Environment 20 25 30 Trainer Trainee Slave Wall 5 -0.2 Trainee 0.1 Joint Angle (rad) 0.7 0.6 0.5 0.4 0.3 0.2 0.0 -0.1 -0.1 Hand/Environment Torque (N*m) 0.5 0.4 0.3 0.2 0.1 0.0 Trainer Experiment Results: Wall Contact Transparency Wall Contact Force Tracking 12 / 16
0 10 10 15 20 25 30 0 5 15 Environment 20 25 30 Trainer Trainee Slave Wall 5 -0.2 Trainee 0.1 Joint Angle (rad) 0.7 0.6 0.5 0.4 0.3 0.2 0.0 -0.1 -0.1 Hand/Environment Torque (N*m) 0.5 0.4 0.3 0.2 0.1 0.0 Trainer Experiment Results: Wall Contact Transparency Wall Contact Force Tracking 12 / 16
Current Works Constant/Varying time delays: passivity and transparency analysis 13 / 16
Current Works Constant/Varying time delays: passivity and transparency analysis 13 / 16
Current Works Sampled data system Related works Continuous-discrete coupling: [S. Stramigioli-2005] Discrete IPC controller: [S. Aoues-2013] 14 / 16 Discrete scattering: [S. Stramigioli-2005]
Future Works and Perspectives Future Works 3 d.o.f and 3D haptic rending : ROS (Robotic Operating System), CHAI3D End-user feedback : medical doctors Perspectives Gesture analysis : evaluation of gesture behaviors Human(trainee)-robot semi-autonomous shared control 3D immersion : integration with virtual reality Extension to multi-masters/multi-slaves system 15 / 16
Future Works and Perspectives Future Works 3 d.o.f and 3D haptic rending : ROS (Robotic Operating System), CHAI3D End-user feedback : medical doctors Perspectives Gesture analysis : evaluation of gesture behaviors Human(trainee)-robot semi-autonomous shared control 3D immersion : integration with virtual reality Extension to multi-masters/multi-slaves system 15 / 16
The end Thank you for your attention! 16 / 16
Recommend
More recommend
