Control Algorithms for Lower Limb Exosksleton

Contributions

• The project involves to design a controller to track a predefined gait trajectory.

• We are using Tmotor for actuating the joints of the exoskeleton.

• Control algorithms tested on:

   a) PID
   b) Torque based PD
   c) iLQR
   d) MPC
  

Tmotor CAN bus read/write

• A Python module for controlling the AK-series Tmotor Actuator AK10-9 and AK 80-64 through the CAN bus.

• TmotorCAN: Python framework for CAN read/write.

• Channel_config: To initialize the CAN communication with Kvaser USB to CAN.

The commands will be updated soon.

GUI using PyQt

Arduino/Jetson Nano

• Code for control using these two boards will be updated soon

Issuses addressed in the Tmotor firmware

• The motor sudenly jerks after the first CAN start command, a inherent problem in the Tmotor firmware.

• Zero positioning of the motor fails at random instants, build a safety filter to address it.

• Establishing a robust CAN communcation with the motors.

Deployment

1. clone the repository

https://github.com/Mowbray-R-V/Exoskeleton_control_Tmotor.git

2. For GUI based control

 python3 final_app.py 

3. To control your Tmotor using python use the below as module

 tmotorCAN.py

GUI Instructions

refer

Phase 1

Phase 2

Phase 3

Working Video

• https://drive.google.com/file/d/1yswbQRFnBGRXAdPrOnonQIoeQ4WLLH7M/view?usp=sharing

Authors

If you use this code in your project please cite us as:

@misc{Exoskeleton_control_Tmotors,
  author = {Mowbray RV, Satvik, Haricharan},
  title = {Control Algorithms for Lower Limb Exosksleton},
  year = {2023},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/Mowbray-R-V/Exoskeleton_control_Tmotor/tree/main}},
}