Hi,
I am currently researching reflexes control models for 3D gait within SCONE. I am particularly interested in gaining a deeper understanding of the following models:
Gait3D - H1922 - Hyfydy
Gait3D - H1922 - Treadmill - Hyfydy
Gait3D - H1922 - Uneven Terrain - Hyfydy
Specifically, I would greatly appreciate it if you could provide detailed information on:
Reflex Mechanisms: A breakdown of the specific reflex mechanisms implemented in each model, such as stretch reflexes, CPGs, or other relevant reflexes.
Parameterization and Tuning: How are these reflex mechanisms parameterized and tuned for each specific model? Please elaborate on any differences in parameter values or tuning strategies between the models.
Control Strategy Variations: How do the control strategies differ between the three models to address the unique challenges of level walking, treadmill walking, and uneven terrain?
Furthermore, could you please direct me to any publications or documentation that provide an in-depth explanation of these models? Any links to relevant resources would be extremely helpful.
Thank you in advance for your time and assistance!
Control Models for Gait3D
-
Thomas Geijtenbeek
- Posts: 436
- Joined: Wed Mar 12, 2014 8:08 am
Re: Control Models for Gait3D
All the examples you mention use the same reflex-based control strategy, which consists of a combination of force, length and velocity reflexes, with a constant excitation. The difference between the examples is that the control parameters are optimized for different objectives, hence the difference in behavior.
The control strategy is based on: Geyer, Herr. A muscle-reflex model that encodes principles of legged mechanics produces human walking dynamics and muscle activities. IEEE Trans Neural Syst Rehabil Eng. 2010 Jun;18(3):263-73. Many of the reflexes are described and motivated in this paper.
The main difference is that the controller is adapted for 3D models instead of 2D models, and that it differentiates between only two states instead of three (stance and swing). It also has additional reflexes for hip abductors and adductors. The controller is not yet published, but all the reflexes are described in the file:
SCONE/Examples/controllers/H1922RS2v3.scone.
In this file, you can see an overview of all reflexes that are used, and during which phase they are active. The meaning of the individual settings is documented here:
https://scone.software/doku.php?id=ref:muscle_reflex
Finally, the different gait phases that are used are documented here:
https://scone.software/doku.php?id=ref: ... controller
The control strategy is based on: Geyer, Herr. A muscle-reflex model that encodes principles of legged mechanics produces human walking dynamics and muscle activities. IEEE Trans Neural Syst Rehabil Eng. 2010 Jun;18(3):263-73. Many of the reflexes are described and motivated in this paper.
The main difference is that the controller is adapted for 3D models instead of 2D models, and that it differentiates between only two states instead of three (stance and swing). It also has additional reflexes for hip abductors and adductors. The controller is not yet published, but all the reflexes are described in the file:
SCONE/Examples/controllers/H1922RS2v3.scone.
In this file, you can see an overview of all reflexes that are used, and during which phase they are active. The meaning of the individual settings is documented here:
https://scone.software/doku.php?id=ref:muscle_reflex
Finally, the different gait phases that are used are documented here:
https://scone.software/doku.php?id=ref: ... controller