Content uploaded by Vyankatesh Ashtekar
Author content
All content in this area was uploaded by Vyankatesh Ashtekar on Jan 24, 2018
Content may be subject to copyright.
Measured joint angle variation
The Optimization Problem
Sit-to-Stand Experiment
Validation of Results and Conclusions
•Simulated joint angle variation emulates the natural variation
•Deviation in the joint angles builds up after the body has
gained vertical momentum. Optimization criterion may be
different once the gross motion is completed
•Vertical GRF calculated using the predicted trajectory tracks
the experimental one closely over the complete motion
•Derive inputs for simulation
•Hands folded close to chest
•Five repetitions performed on
a single person
•3D Motion capture system used to
track body segments at 490Hz
•Force plate is placed under the feet,
sampling data at 980Hz
Comparison:
predicted & measured GRFy
Why Study Sit-to-Stand Motion?
•An example of excellent neuro-muscular coordination
•Ubiquitous activity, prerequisite to a lot of daily tasks,
performed ≥ 30 times in a day even by elderly people
•Defines independence of a person in daily life
•Decide motion strategy for the unfit
•Motion planning of humanoid robots
Robotic Model of Human
•Motion observed in sagittal plane only
•Human simplified to three link serial robot (3-R planar)
•Body segment parameters adapted from literature
•Feet are considered to be fixed to the ground, but
predicted motion validated by checking for upward
ground reaction force
Motion Prediction Algorithm
Trajectory Optimization Based
Prediction of Sit-to-Stand Motion
Vyankatesh Ashtekar1, Sourav Rakshit1
1Department of Mechanical Engineering, Indian Institute of Technology Madras, Chennai 600036
Email: vyankatesh0811@gmail.com, srakshit@iitm.ac.in
Paper No: 136
•# B-Spline segments
•Initial & final values of
joint angles
•Initial & final values of
joint angular velocities
•Duration of motion (T)
Start
Joint angle convention,
link lengths, positions of
c.o.m., mass and inertia
of the links
Generate
Equations of Motion
Define cost function
Minimize cost function
Optimized
B-spline
control points
•Plot joint angle trajectories
•Calculate vertical ground reaction
force by inverse recursive dynamics
•Validation of results
End
(Simulation Parameters)
Comparison of simulated trajectories with the experimental ones
−𝜏3
𝜏2
−𝜏2
𝐺𝑅𝐹
y𝐺𝑅𝐹
𝑥
𝑚1𝑔
𝑚2𝑔
𝜏3
𝑚3𝑔
ሷ
𝒑c3
ሷ
𝒑c2
𝑓
3𝑦
𝑓
3𝑥
𝑓
2𝑦
𝑓
2𝑥 ሷ
𝒑𝑐1
A snap during
STS experiment FBD of Human for calculation of GRFy