A simulation-based study for optimizing proportional-integral-derivative controller gains for different control...

Area of Science:

• Biomechanics
• Human body modeling
• Control systems engineering

Background:

• Accurate human body models are crucial for understanding biomechanical responses.
• Simulating active human behavior requires sophisticated control strategies.
• Previous models often simplify muscle activation and sensory feedback.

Purpose of the Study:

• To investigate the influence of Proportional-Integral-Derivative (PID) controller gains, reaction time, and initial muscle activation on active human model behavior.
• To compare three distinct control strategies for simulating human arm movement.
• To optimize model parameters using experimental data from human subjects.

Main Methods:

• Utilized a finite element model of the human left arm from the Global Human Body Models Consortium (GHBMC).
• Modeled major skeletal muscles as 1D beam elements with Hill-type muscle material.
• Employed angular position control, muscle length control, and a combined strategy, optimizing PID gains and reaction delay.

Main Results:

• Controller gains and reaction delays were optimized based on experimental data from five male subjects under varied conditions (muscle state, eye awareness).
• Distinct controller gains and initial activation were necessary for relaxed versus tensed muscle states.
• Increased reaction delay effectively simulated closed-eye conditions, validated by CORA scores ranging from 0.77 to 0.95.

Conclusions:

• The study successfully demonstrated the impact of PID controller parameters and reaction time on active human model behavior.
• Muscle state and awareness significantly influence the required control strategy and parameters.
• The combined control strategy, incorporating optimized PID gains and reaction delay, provided a robust simulation of human arm biomechanics across different conditions.