Observer-based differential evolution constrained control for safe reference tracking in robots

Area of Science:

• Robotics
• Control Systems Engineering
• Optimization Algorithms

Background:

• High torque inputs in robotic controls can lead to increased energy consumption.
• Inaccurate perturbation estimations in observers may cause damage to robotic devices.
• Safe reference tracking and perturbation estimation are critical for robotic system reliability.

Purpose of the Study:

• To propose a constrained control method for safe reference tracking in robots.
• To develop a constrained observer for safe perturbation estimation in robotic systems.
• To introduce a method for optimizing control gains to balance safety and energy efficiency.

Main Methods:

• An observer-based differential evolution constrained control approach is proposed.
• A constrained observer is designed for estimating perturbations safely.
• A constrained control strategy is implemented for safe reference tracking.
• Differential evolution is employed to determine optimal control gains.
• Robust stability of the proposed control system is mathematically assured.

Main Results:

• The proposed constrained observer effectively estimates perturbations, preventing device damage.
• The constrained control ensures safe reference tracking, even with significant external disturbances.
• The differential evolution optimization successfully identifies control gains that balance safe tracking and energy consumption.
• The observer-based differential evolution constrained control demonstrates robust stability.
• The method was successfully applied to safe reference tracking in two distinct robotic applications.

Conclusions:

• The developed observer-based differential evolution constrained control offers a robust solution for safe and energy-efficient robotic operations.
• This approach enhances robotic safety by preventing damage from excessive torque and inaccurate estimations.
• The optimization technique provides a systematic way to tune control parameters for improved performance in real-world robotic applications.