Optimal robust motion controller design using multiobjective genetic algorithm

Area of Science:

• Control Systems Engineering
• Robotics
• Optimization Algorithms

Background:

• Robust motion controller design is crucial for systems operating under uncertainties.
• Existing methods often struggle to balance robustness and performance objectives.
• The Reusable Intelligent Controller (RIC) framework offers a structured approach to controller synthesis.

Purpose of the Study:

• To develop a novel multiobjective optimization strategy for robust motion controller design.
• To integrate disturbance observer-based control within the RIC framework.
• To simultaneously assess and optimize robustness, stability, and time-performance metrics.

Main Methods:

• Utilized a multiobjective genetic algorithm (MOGA) for controller optimization.
• Employed a disturbance observer within a Reusable Intelligent Controller (RIC) structure with internal and external feedback loops.
• Implemented regional pole placement for controller structure simplification and arbitrary selection.
• Defined objective functions based on polynomial conditions for robustness and performance deviation.

Main Results:

• Successfully designed controllers optimizing multiple, often conflicting, objectives simultaneously.
• Demonstrated the effectiveness of the MOGA-differential evolution approach in achieving robust stability and desired time-performance.
• The regional pole placement method facilitated simplified and adaptable controller structures.

Conclusions:

• The proposed multiobjective optimization framework effectively addresses robust motion controller design challenges.
• The integration of disturbance observers and RIC offers a systematic approach to enhance system resilience.
• This method provides a powerful tool for simultaneously improving stability and performance in complex control systems.