A saturation based self-tuned robust control design for Euler Lagrange systems.

Area of Science:

• Robotics and Control Systems
• Mechanical Engineering
• Applied Mathematics

Background:

• Controlling mechanical systems in Euler-Lagrange (EL) form is challenging due to parameter uncertainties and external disturbances, often leading to performance degradation and stability issues.
• Existing robust and high-gain control methods can suffer from chattering, impacting system performance and reliability.
• Model-free approaches are desirable for systems where precise models are unavailable or difficult to obtain.

Purpose of the Study:

• To design and analyze a novel, continuous, model-free robust controller for mechanical systems described by Euler-Lagrange equations.
• To address parameter estimation errors and external disturbances without requiring an exact system model.
• To ensure closed-loop stability and improve control performance, while avoiding controller chattering.

Main Methods:

• A saturation function-based, continuous robust controller is designed for EL systems.
• Lyapunov-based arguments are employed to rigorously prove the stability of the closed-loop system.
• An adaptive gain-tuning algorithm is developed as an add-on to simplify controller tuning.
• The controller's effectiveness is validated through simulations on a twin rotor multi-input-multi-output (TRMS) system and experimental testing on a mobile robotic platform.

Main Results:

• The proposed model-free robust controller effectively manages parameter uncertainties and disturbances in EL systems.
• The controller utilizes continuously differentiable terms to prevent chattering, ensuring smooth control inputs.
• Simulations on the TRMS model and experiments on a mobile robot demonstrated satisfactory performance.
• Experimental results showed less than 0.5° error in roll/pitch and less than 1° error in yaw for the mobile robot.

Conclusions:

• The developed saturation function-based, model-free robust controller offers a stable and effective solution for controlling complex mechanical systems.
• The adaptive gain-tuning algorithm simplifies controller implementation and tuning.
• The controller's practical feasibility and performance are confirmed through both simulation and real-world robotic applications.