Double-Closed-Loop Robust Optimal Control for Uncertain Nonlinear Systems.

Area of Science:

• Control Engineering
• Nonlinear System Dynamics
• Optimization Theory

Background:

• Optimal control is crucial for nonlinear systems but struggles with uncertainties and complex objectives.
• Solving optimal reference trajectories is challenging, often hindering robust performance.
• Existing methods may not adequately address uncertainties in complex optimization problems.

Purpose of the Study:

• To propose a novel double-closed-loop robust optimal control (DCL-ROC) method.
• To enhance the reliability and performance of optimal control for uncertain nonlinear systems.
• To address the difficulties in solving optimal reference trajectories and control laws under uncertainty.

Main Methods:

• A double-closed-loop scheme separates trajectory optimization and control law computation.
• A closed-loop robust optimization (CL-RO) algorithm uses data-driven forms for uncertain objectives.
• An adaptive controller tracks optimal trajectories without requiring exact system dynamics.

Main Results:

• The DCL-ROC method effectively divides the control process into distinct optimization and control loops.
• The CL-RO algorithm improves the optimality of reference trajectories under uncertainty.
• The adaptive controller ensures adaptivity and reliability of control laws, even with unknown system dynamics.
• Experimental results show superior performance compared to other optimal control techniques.

Conclusions:

• The proposed DCL-ROC method offers a robust solution for optimal control of uncertain nonlinear systems.
• The approach enhances both the generation of optimal trajectories and the derivation of reliable control laws.
• DCL-ROC demonstrates improved adaptivity, reliability, and overall performance in complex control scenarios.