https://github.com/ami-iit/paper_sorrentino_2024_humanoids_friction_estimation
https://github.com/ami-iit/paper_sorrentino_2024_humanoids_friction_estimation
Science Score: 36.0%
This score indicates how likely this project is to be science-related based on various indicators:
-
○CITATION.cff file
-
✓codemeta.json file
Found codemeta.json file -
✓.zenodo.json file
Found .zenodo.json file -
○DOI references
-
✓Academic publication links
Links to: arxiv.org, ieee.org -
○Committers with academic emails
-
○Institutional organization owner
-
○JOSS paper metadata
-
○Scientific vocabulary similarity
Low similarity (12.2%) to scientific vocabulary
Repository
Basic Info
- Host: GitHub
- Owner: ami-iit
- Language: Python
- Default Branch: main
- Size: 25.2 MB
Statistics
- Stars: 8
- Watchers: 4
- Forks: 2
- Open Issues: 1
- Releases: 0
Metadata Files
README.md
Physics-Informed Learning for the Friction Modeling of High-Ratio Harmonic Drives
Installation
This repository requires to install idyntree library and MATLAB.
Use the requirements.txt file to recreate the environment:
conda create --name <new_environment_name> --file requirements.txt
Repo usage
The application for acquiring data for friction identification can be found in https://github.com/LoreMoretti/bipedal-locomotion-framework/tree/add/MotorCurrentTrackingApplication/utilities/motor-current-tracking. You can follow instruction in the repo to install and use it.
Datasets used for this paper for the training can be found at https://huggingface.co/datasets/ami-iit/sensorless-torque-control/tree/main.
After taking data, the first step is data post-processing. Run the bash script postprocess_data.sh. Example usage for parsing data for the r_ankle_pitch joint.
bash postprocess_data.sh -f '/home/isorrentino/dev/dataset/friction/r_ankle_pitch/sinusoid' -j 'r_ankle_pitch' -a 'torso_pitch torso_roll torso_yaw l_hip_pitch l_hip_roll l_hip_yaw l_knee l_ankle_pitch l_ankle_roll r_hip_pitch r_hip_roll r_hip_yaw r_knee r_ankle_pitch r_ankle_roll'
Find the Stribeck-Coulomb-Viscous model for the physics information used by the PINN. Change the joint to model in the script simple_friction_modeling.py.
python simple_friction_modeling.py
Before running the PINN training you need to specify the configuration file for the join to model. The config folder contains an example for the r_ankle_roll joint. After creating the configuration file you can run the training by means of wight&biases tool:
python feedforwardNN_wandb.py --joint_name "r_ankle_roll"
The trained networks are saved in the results forlder and can be converted in a onnx model by using the script convert_to_onnx.py.
The onnx model is loaded by the device JointTorqueControlDevice running on the robot torso computer.
Maintainer
This repository is maintained by:
| | |
|:---:|:---:|
| 
| @inessorrentino |
Owner
- Name: Artificial and Mechanical Intelligence
- Login: ami-iit
- Kind: organization
- Location: Italy
- Website: https://ami.iit.it/
- Repositories: 111
- Profile: https://github.com/ami-iit
GitHub Events
Total
- Watch event: 9
- Issue comment event: 6
- Push event: 5
- Fork event: 2
Last Year
- Watch event: 9
- Issue comment event: 6
- Push event: 5
- Fork event: 2
Committers
Last synced: about 1 year ago
Top Committers
| Name | Commits | |
|---|---|---|
| Ines Sorrentino | 4****o | 10 |
Issues and Pull Requests
Last synced: 11 months ago
All Time
- Total issues: 1
- Total pull requests: 0
- Average time to close issues: N/A
- Average time to close pull requests: N/A
- Total issue authors: 1
- Total pull request authors: 0
- Average comments per issue: 6.0
- Average comments per pull request: 0
- Merged pull requests: 0
- Bot issues: 0
- Bot pull requests: 0
Past Year
- Issues: 1
- Pull requests: 0
- Average time to close issues: N/A
- Average time to close pull requests: N/A
- Issue authors: 1
- Pull request authors: 0
- Average comments per issue: 6.0
- Average comments per pull request: 0
- Merged pull requests: 0
- Bot issues: 0
- Bot pull requests: 0
Top Authors
Issue Authors
- hak7un (1)