The document describes research to increase the workspace of a Stewart platform using an optimized movable base. It introduces Stewart platforms and their applications. It then discusses limitations of workspace due to actuator length and joint constraints. Two questions are posed regarding platform architecture and guaranteeing continuous motion. Solutions explored include a hybrid Stewart platform with a movable middle plate, optimization of platform geometry using genetic algorithms and adaptive simulated annealing, and inverse kinematics analysis. The optimized hybrid platform is shown to increase workspace coverage compared to a standard Stewart platform.
1. Workspace Increase of Stewart
Platform Using Movable Base
By: Marzieh Nabi
Supervisor: prof. Malaek
2. Introduction of Stewart Platform
Applications:
Simulators Aircraft
Car
Motorcycle
Manufacturing Tools
Medicine Physio-Trappy
Eye-Surgery
Entertainment
Entertainment
Skate-Learning
4. Two Questions
1. What should be the architecture of the Stewart platform
for an specific application?
2. How to guarantee continues motion of the Stewart
platform at each step simulation considering the joints
and the actuators constraints?
11. Translational Workspace
Results of the Simulation
3.5
Desired Workspace 3
2.5
Z
2
20 20
1.5
2
1 2
0
-1
L 0
1
-1
-2 -2
Y X
Rotational Workspace
20 20
20 20
12. Translational Workspace
Result of the Simulation
3.5
Desired Workspace 3
2.5
Z
2
1.5
Rotational Workspace
2
1 2
0
-1
L 0
1
-1
20 -2 -2
Y X
10
0
-10
-20
20
10 20
0 10
-10 0
-10
-20 -20
13. Case 1 & 2
No Joint Angle Limitation &
min 2.7 m Covered Workspace = 96 %
20 20
min 2.3 m Covered Workspace = 82.1 %
14. 0.25
0.2
Fitness Value
0.15
Convergence of GA
0.1
Case 1
0.05
0
0 5 10 15 20 25 30 35 40
Generation
0.5
0.45
0.4
Fitness Value
0.35 Convergence of GA
0.3
Case 2
0.25
0.2
0 10 20 30 40 50 60
Generation
15. Case 3
No Joint Angle Limitation &
min 16
min 25 As Optimization Variables
min 34
min 16 3.0 m
min 25 3.0
Covered Workspace = 99.65 %
m
min 34 3.0
m
16. 0.45
0.4
0.35
0.3
Fitness Value
0.25
Convergence of GA
0.2
0.15
Case 3
0.1
0.05
0
0 10 20 30 40 50
Generation
4
3
Schematic of SP 2
Case 3 1
0
2 4
1 2
0 0
-1 -2
-2 -4
17. Case 4
max Pla tfo rm
160
Joint Angle Limitations
max Ba see
160
min 2.3 m
Covered Workspace = 45.21 %
18. Case 5
max Pla tfo rm
160
Joint Angle Limitations
max Ba see
160
min 16
min 25 As Optimization Variables
min 34
19. min 16 3.0 m
min 25 2.92 m Covered Workspace = 94.47 %
min 34 2.48 m
1
0.9
0.8
0.7
Convergence of GA 0.6
Fitness Value
0.5
Case 5 0.4
0.3
0.2
0.1
0
0 5 10 15 20 25 30 35 40 45 50
Generation
20. Optimization With Adaptive Simulated Annealing
1. Running time is more than the genetic algorithm
(GA)
2. Percentile of the Covered Workspace is Less than
GA
3. Slower Convergence rate compare to GA
21. Rate of Convergence: GA and ASA
1
0.9
0.8 ASA
0.7
ASA
Fitness Value
0.6
0.5
0.4
0.3 GA
GA
0.2
0.1
0
0 5 10 15 20 25 30 35 40
Iteration
Case 1
22. Rate of Convergence: GA and ASA
1
0.9
0.8
ASA
ASA
ASA
0.7
Fitness Value
0.6
0.5
GA
0.4
GA
0.3
0.2
0.1
0
0 10 20 30 40 50 60
Iteration
Case 2
23. Rate of Convergence: GA and ASA
1
0.95
ASA
0.9
ASA
0.85
Fitness Value
0.8
0.75
0.7
GA GA
0.65
0.6
0.55
0.5
0 5 10 15 20 25 30 35 40 45 50
Iteration
Case 4
24. Next Step of the Comparison Between SP and HSP
Length of SP actuators
Inverse Kinematic of
OPTIMIZED SP
Coordinate of
End Effector Comparison
Inverse Kinematic of
HSP
Length of HSP actuators
So we need to solve the inverse kinematic of HSP
26. The DOF of Lower Robot
Lower Robot
Screw Theory 3 Translational DOF
27. Dividing Strategies
1. Translational Coordinates of the Middle Plate k 0 1
Translational Coordinates of End Effector
k 0.5, 0.7
28. k 0.7 x1 y1 z1 k x y z
Actuator 1&2(m) 4.5
4
3.5
3 Actuator 1 & 2
2.5
0 5 10 15 20 25 30
time (s)
k 0.7
2.6
2.4
Actuator 3(m)
2.2
2
Actuator 3
1.8
1.6
0 5 10 15 20 25 30
time (s)
29. 5 4
Stewart Platform Stewart Platform
Hybrid Stewart Platform Act. 4 & 9 Hybrid Stewart Platform
4.5
Act. 5 & 8
3.5
Actuator 5&8(m)
SP
Actuator 4&9 (m)
4
3.5
SP 3
HSP
HSP
3
2.5
2.5 0 5 10 15 20 25 30
0 5 10 15 20 25 30
time (s)
time (s)
3.6
Stewart Platform 6
HSP
3.4
Hybrid Stewart Platform
Act. 6 & 7 5
3.2
4
3
3
Actuator 6&7(m)
2.8
2.6
SP 2
2.4 1
2.2
HSP 0
4
2 2
0
1.8
-2
4 6
1.6 0 2
0 5 10 15 20 25 30 -4 -4 -2
-6
time (s)
30. k 0.5 x1 y1 z1 k x y z
4
Actuator 1&2(m)
3.5
3
Actuator 1 & 2
2.5
0 5 10 15 20 25 30
time (s)
2
1.8
Actuator 3(m)
1.6
1.4 Actuator 3
0 5 10 15 20 25 30
time (s)
31. 5 4
Stewart Platform Stewart Platform
Hybrid Stewart Platform Act. 4 & 9 3.8
Hybrid Stewart Platform
4.5
3.6
Act. 5 & 8
3.4
HSP
Actuator 4&9 (m)
SP
Actuator 5&8(m)
4
3.2
SP 3
3.5 HSP
2.8
3 2.6
2.4
2.5
0 5 10 15 20 25 30 2.2
0 5 10 15 20 25 30
time (s)
3.6 time (s)
Stewart Platform
3.4
Hybrid Stewart Platform Act. 6 & 7 HSP
5
3.2
4.5
3 4
3.5
Actuator 6&7(m)
2.8
2.6
SP 3
2.5
2.4 HSP 2
1.5
2.2 1
0.5
2
0
4
1.8 2
0 6
4
-2 2
1.6 0
-2
0 5 10 15 20 25 30 -4 -4
-6
time (s)
32. l lmin
Minimization of f1
lmax
2.6
Actuator 1&2(m)
2.4
2.2
2
Actuator 1 & 2
1.8
0 5 10 15 20 25 30
time (s)
1.3
1.2
Actuator 3(m)
1.1
1
0.9
Actuator 3
0.8
0 5 10 15 20 25 30
time (s)
33. 4.8 4
Stewart Platform Stewart Platform
4.6
Hybrid Stewart Platform Act. 4 & 9 Hybrid Stewart Platform
3.5
Act. 5 & 8
4.4
4.2
Actuator 5&8(m)
Actuator 4&9 (m)
4
3
SP HSP
3.8 SP
HSP 2.5
3.6
3.4 2
3.2
1.5
3 0 5 10 15 20 25 30
0 5 10 15 20 25 30 time (s)
3.6 time (s)
Stewart Platform
4.5
Hybrid Stewart Platform
3.4
Act. 6 & 7 4 HSP
3.2
3.5
3 3
Actuator 6&7(m)
2.8 2.5
2.6 SP 2
1.5
2.4
1
2.2
0.5
2
0
1.8 HSP 5
0
1.6
0 5 10 15 20 25 30
-5
-6 -4 -2 0 2 4 6
time (s)
34. l
Minimization of f2
lmax
3
Actuator 1&2(m)
2.8
2.6
Actuator 1 & 2
2.4
2.2
0 5 10 15 20 25 30
time (s)
2
1.8
Actuator 3(m)
1.6
1.4
Actuator 3
1.2
1
0 5 10 15 20 25 30
time (s)
35. 5 4
Stewart Platform Stewart Platform
Hybrid Stewart Platform Act. 4 & 9 Hybrid Stewart Platform
3.5 Act. 5 & 8
4.5
3
SP
Actuator 5&8(m)
Actuator 4&9 (m)
4
HSP
SP 2.5
3.5
HSP 2
3 1.5
1
2.5 0 5 10 15 20 25 30
0 5 10 15 20 25 30
time (s)
4 time (s)
Stewart Platform
Hybrid Stewart Platform 5
HSP
3.5 Act. 6 & 7 4.5
4
3.5
3
SP
Actuator 5&8(m)
3
2.5
2.5
2
1.5
2 1
0.5
1.5 0
HSP 4
2
0
-2 4 6
1 -2 0 2
0 5 10 15 20 25 30 -4 -4
-6
time (s)
36. Comparison Between Different 4.5
Hybrid Stewart Platform (K=0.7)
Hybrid Stewart Platform (K=0.5)
Hybrid Stewart Platform (|(l-l )/l |)
min max
Actuator 1&2(m)
4
K=0.7 Hybrid Stewart Platform (|l/lmax|)
Strategies 3.5
l lmin
3 f1
lmax
2.5
0 5 10 15 20 25 30
time (s)
2.5
K=0.7
2
Actuator 3(m)
1.5
l lmin
1 f1
lmax
5
Stewart Platform 0.5
0 5 10 15 20 25 30
Hybrid Stewart Platform (K=0.7)
time (s)
Hybrid Stewart Platform (K=0.5)
Hybrid Stewart Platform ( ||l-lmin/lmax || )
4.5
Hybrid Stewart Platform ( ||l/lmax || )
SP
Actuator 4 & 9
Actuator 4&9 (m)
4
l lmin
f1
lmax
3.5
3
K=0.7
2.5
0 5 10 15 20 25 30
time (s)
37. 4
Stewart Platform
Hybrid Stewart Platform (K=0.7)
3.8 Hybrid Stewart Platform (K=0.5)
Hybrid Stewart Platform ( || ( l-l ) /l || )
SP
min max
3.6 Hybrid Stewart Platform ( || l/lmax || )
l lmin
3.4 f1 Actuator 5 & 8
lmax
Actuator 5&8(m)
3.2
3
2.8
K=0.7
2.6
2.4
3.6
Stewart Platform
2.2 Hybrid Stewart Platform (K=0.7)
0 5 10 15 20 25 3.4 30
Hybrid Stewart Platform (K=0.5)
time (s)
Hybrid Stewart Platform ( || ( l-l ) /l || )
3.2 min max
Hybrid Stewart Platform ( || l/l || )
max
3
SP
Actuator 6&7(m)
2.8
l lmin
2.6 f1
Actuator 6 & 7 2.4
lmax
2.2
2
1.8 K=0.7
1.6
0 5 10 15 20 25 30
time (s)
38. Summery
Optimization of the Stewart platform
Genetic Algorithm
Adaptive Simulated Annealing
Inverse Kinematic of Hybrid Stewart Platform
Comparison between the optimized SP and HSP
39. Submitted papers
1. S.M. Malaek, M. Nabi-A, “Optimal Design of Stewart Platform Using
Genetic Algorithm”, ICINCO, 9-12 May 2007, Angers, France
2. S.M. Malaek, M. Nabi-A, “Optimal Design of Stewart Platform Using
Adaptive Simulated Annealing”, ICINCO, 9-12 May 2007, Angers,
France