introduction - control and robotics in medicine · 2020. 9. 15. · course organization the course...
TRANSCRIPT
-
Introduction
Control and Robotics in Medicine
Blanca Larraga & Álvaro GutiérrezSeptember 15th, 2020
[email protected]@etsit.upm.es
www.robolabo.etsit.upm.es
N
-
Orthoses
N
-
Prostheses
N
-
Telesurgery
N
-
Contents
1 What is behind?MechanicsSensors and actuatorsControl
2 ExamplesWalking assistanceExoskeletonProsthesesForce feedbackTelesurgery
3 Course organization
N
-
1 What is behind?MechanicsSensors and actuatorsControl
2 ExamplesWalking assistanceExoskeletonProsthesesForce feedbackTelesurgery
3 Course organization
N
-
1 What is behind?MechanicsSensors and actuatorsControl
2 ExamplesWalking assistanceExoskeletonProsthesesForce feedbackTelesurgery
3 Course organization
N
-
Mechanics
◮ Mechanic structures◮ Forward and inverse
kinematics◮ Forward and inverse dynamics◮ Trajectory planning
N
-
1 What is behind?MechanicsSensors and actuatorsControl
2 ExamplesWalking assistanceExoskeletonProsthesesForce feedbackTelesurgery
3 Course organization
N
-
Sensors
N
-
Sensors
◮ Magnetic◮ Optical◮ Acoustic◮ Inertial◮ Mechanical
N
-
Position
N
-
Velocity
Problems with the velocity
◮ Derive the position
◮ For low speeds: Resolution problems → Small changes onposition
◮ For high speeds: Computational problems → Noise in thederivative
N
-
Force
◮ Resistive◮ Piezoelectric◮ Capacitive◮ Analog◮ Digital◮ Compression◮ Traction◮ Flexion◮ ....
N
-
Current
N
-
Inertial
◮ Accelerometer◮ Gyroscope◮ Compass
N
-
Actuators
◮ Hydraulic◮ Pneumatic◮ Electric
◮ DC motors◮ Brushless motors◮ Step motors
N
-
Transmission
◮ Gears◮ Pulleys◮ Capstan◮ Direct
N
-
1 What is behind?MechanicsSensors and actuatorsControl
2 ExamplesWalking assistanceExoskeletonProsthesesForce feedbackTelesurgery
3 Course organization
N
-
Controllers
GFF (z)
+−
GDI(z) ++
++
+
ZOH Gu(s)
GPA(z)
GFB(z)
T
r(k) ẽ(k) uDI(k) u(k)
w(k)
ũ(k) ũ(t)
y(k)
y(t)
uPA(k)
uFF (k)
uFB(k)
N
-
Control systems
Gc(s) G(s)r(t) u(t) y(t)
◮ The output is the variable to measure and control (y )◮ The input is the variable that is changed to modify the
output (r ).◮ The error is the difference between the input and the
output (e = y − r )
◮ Objective: Make the error be or tend to zero.(limt→∞ e = 0)
N
-
Closed loop control system
+−
Gc(s) G(s)r(t) e(t) u(t) y(t)
◮ Feedback: error signal (e)◮ The sensibility to perturbations is reduced.◮ The sensibility to internal parameters variations is reduced.◮ Stability problems.
N
-
1 What is behind?MechanicsSensors and actuatorsControl
2 ExamplesWalking assistanceExoskeletonProsthesesForce feedbackTelesurgery
3 Course organization
N
-
1 What is behind?MechanicsSensors and actuatorsControl
2 ExamplesWalking assistanceExoskeletonProsthesesForce feedbackTelesurgery
3 Course organization
N
-
Walking assistance
H(s) G(s)+−
U(s) Y(s)E(s)R(s)Ref Gen
Pot
Button
Exoesqueleto
Sensor + Motor
G(s) =A
s(s + p)
H(s) = Kp
(
1 + τds +1τis
)
N
-
Walking assistance
Honda Stride Management Assist.
N
-
Walking assistance
Exoesqueleto
Sensor◮ Inertia matrix (M)◮ Coriolis matrix (C)◮ Gravitational torque vector (P)◮ Motor torque vector (T )◮ Exoskeleton torque vector (d)
M(θ)θ̈ + C(θ, θ̇)θ̇ + P(θ) = T + dT = P̂ + (1 − α1)[M̂(θ)θ̈ + Ĉ(θ, θ̇)θ̇]
M(θ)θ̈+C(θ, θ̇)θ̇+P(θ) = ˆP(θ)+ (1−α1)[M̂(θ)θ̈+ Ĉ(θ, θ̇)θ̇] + d
Kazerooni et al., On the Control of the Berkeley Lower Extremity Exoskeleton(BLEEX), ICRA 2005.
N
-
Walking assistance
Cyberdine Robot Suit HAL.
N
-
1 What is behind?MechanicsSensors and actuatorsControl
2 ExamplesWalking assistanceExoskeletonProsthesesForce feedbackTelesurgery
3 Course organization
N
-
Walking kinematics and dynamics
N
-
Estimators
◮ Center of mass◮ Kinematic stability◮ Dynamic stability◮ Velocity stability◮ Turn, rotations,
accelerations,...
N
-
Estimators
◮ Center of mass◮ Kinematic stability◮ Dynamic stability◮ Velocity stability◮ Turn, rotations,
accelerations,...
N
-
Controller
(KALMANFILTER,
NEURAL NETWORKS,
ADAPTIVE,
LEARNING,...)
SENSOR
FUSION
MotorHip+−
H(s) G(s)
+−
H(s) G(s)
+−
H(s) G(s)
ENCODER
TACHOMETER
INERTIAL
PLANNING
PATH
Motor
Motor
Knee
AnkleFORCE
N
-
Exoskeletons
N
-
Poses
N
-
1 What is behind?MechanicsSensors and actuatorsControl
2 ExamplesWalking assistanceExoskeletonProsthesesForce feedbackTelesurgery
3 Course organization
N
-
Prostheses
N
-
Let’s look inside!!!
◮ Mechanics◮ Modelling◮ Attachment◮ Plasticity
◮ Dynamics◮ Movement◮ Control
◮ Communication◮ Actuation◮ Sensorization
GFF (z)
+−
GDI(z) ++
+ZOH Gu(s)
GPA(z)
GFB(z)
T
r(k) ẽ(k) uDI(k) u(k) u(t)
y(k)
y(t)
uPA(k)
uFF (k)
uFB(k)
N
-
Attachment
N
-
Attachment
◮ Pressure ulcers, distention, dermatitis,...
N
-
Dynamics
◮ Passive elements
N
-
Dynamics
◮ Passive elements
N
-
Dynamics
◮ Passive elements
◮ A Prosthesis must be something else!!!!◮ Movement, elasticity◮ Control, feedback◮ Sensorization, communication
N
-
Dynamics
N
-
Dynamics
N
-
Communication
◮ Actuation: Muscle (Nerve) → Prosthesis
N
-
Upper limb prosthesis
◮ Many legs → Where are the arms?
N
-
Upper limb prosthesis
◮ Many legs → Where are the arms?
◮ Limitations:◮ Degrees of freedom: > 19 ?◮ Excessive weight: (5 Kg) → < 1/16 body weight?◮ Autonomy: several days?◮ High cost maintenance: light and resistant materials◮ High prices: ≈ 30 ke◮ Response time: slow◮ Feedback: tactile and temperature sensations?
N
-
Upper limb prosthesis
◮ Price reduction with 3D printers
N
-
Upper limb prosthesis◮ And the others?
N
-
1 What is behind?MechanicsSensors and actuatorsControl
2 ExamplesWalking assistanceExoskeletonProsthesesForce feedbackTelesurgery
3 Course organization
N
-
Force feedback
◮ An infinitely rigid “stick” used as a tool to work remotely.
◮ Because the “stick” is massless, the operator will feel theobject to push.
N
-
Force feedback
◮ First stage: user action.
◮ Second stage: sampling.
◮ Third stage: transformations on the sampled signal.
◮ Fourth stage: to act on the motor to obtain feedback andreal sensations.
N
-
Examples
N
-
Examples
N
-
Examples
N
-
Writing assistance
N
-
1 What is behind?MechanicsSensors and actuatorsControl
2 ExamplesWalking assistanceExoskeletonProsthesesForce feedbackTelesurgery
3 Course organization
N
-
Important milestones
◮ Anesthesia◮ Laparoscopy◮ Robotic surgery
N
-
Robotic surgery
◮ ∄ virtual model◮ Force sensors◮ Position sensors◮ Speed sensors◮ Contact sensors◮ Cameras◮ ....
N
-
Telesurgery
N
-
1 What is behind?MechanicsSensors and actuatorsControl
2 ExamplesWalking assistanceExoskeletonProsthesesForce feedbackTelesurgery
3 Course organization
N
-
Course organization
There is no time for everything!
N
-
Course organization
◮ The course is structured in 2 blocks:◮ Kinematics and trajectories◮ Haptic devices
◮ Every block will be independently evaluated, although bothblocks are mandatories and must be obtained (> 5.0 outof 10.0) independently.
◮ Individual and group deliverables. Group organizationdepends on number of students and laboratory resources.
◮ Both blocks have practice parts in the laboratory. Theorganization of the laboratory will be discussed in theclassroom.
N
-
Calendar
Day Topic
15/09/2020 Introduction. DOFs, forward and inverse kinematics22/09/2020 Trajectory planning. Laboratory*29/09/2020 Deliverable D1 (08:59)
29/09/2020 Laboratory*06/10/2020 Laboratory*13/10/2020 Deliverable D2 (08:59)**
13/10/2020 Haptic systems. Laboratory*20/10/2020 Laboratory*27/10/2020 Laboratory*03/11/2020 Deliverable D3 (12:59)**
(*) Laboratory meetings will be scheduled outside the pre-assigned hours due to equipment restrictions.(**) To be confirmed.
N
-
Evaluation
◮ Deliverable D1 (20 %, individual)◮ To formulate the forward kinematics problem of the
laboratory robot.◮ To solve the inverse kinematics problem of the laboratory
robot.◮ Deliverable D2 (40 %, group)
◮ To formulate the trajectory planning of the laboratory robot.◮ To implement some trajectories in the laboratory robot.
◮ Deliverable D3 (40 %, group)◮ To design and implement a gravitational compensator
controller◮ To design and implement a virtual spring haptic system
N
-
Remember
◮ 1 ECTS → 30h.◮ ROB=3 ECTS → 90h.◮ Presential hours: 28h.◮ Extra work: 90h. - 28h. = 62h.
◮ 7 weeks → 12.85h./week◮ Presential hours: 4h/week◮ Extra work: 8.85h/week
N
-
Thank you
THANKS FORLISTENING!!
N
-
Thank you
THANKS FORLISTENING!!
N
What is behind?MechanicsSensors and actuatorsControl
ExamplesWalking assistanceExoskeletonProsthesesForce feedbackTelesurgery
Course organization