Application of PID Control on a Bionic-Hand

Achieving smooth, lifelike motion in robotic prostheses demands precise regulation of motor torques and positions. For the Advances in Robotics and Control curriculum, I tackled the intricate control problem of actuating a multi-degree-of-freedom Bionic Hand.

System Simulation and Tuning

The project required a deep understanding of classical control theory applied to non-linear physical systems:

• Mujoco Physics Simulation: Before touching hardware, I modeled the complex kinematics and contact dynamics of the hand joints within the Mujoco physics engine to ensure safe algorithmic development.
• PID Loop Formulation: Designed robust Proportional-Integral-Derivative (PID) controllers for individual finger actuators. The primary challenge was tuning the $K_p$, $K_i$, and $K_d$ gains to eliminate steady-state error while heavily dampening potential oscillations—crucial for delicate grasping maneuvers.
• Hardware Deployment: The tuned control algorithms were transitioned from simulated environments to a physical bionic prototype, demonstrating stable, responsive actuation that closely mirrored biological articulation under variable loads.

System Visualization