Optimal Tracking in Switched Systems With Free Final Time and Fixed Mode Sequence Using Approximate Dynamic...

Area of Science:

• Control Systems Engineering
• Applied Mathematics
• Computational Intelligence

Background:

• Switched systems present challenges in optimal control due to mode changes.
• Determining optimal switching times and final time is crucial for performance.
• Traditional methods can be computationally intensive.

Purpose of the Study:

• To develop an efficient method for optimal tracking in switched systems.
• To address the complexity of optimizing switching times and final time.
• To reduce the computational burden of Approximate Dynamic Programming (ADP).

Main Methods:

• Formulating the optimal control problem with switching times and final time as parameters.
• Utilizing Approximate Dynamic Programming (ADP) with an inner loop for policy convergence.
• Introducing a novel approach using evolving suboptimal policies to learn the optimal solution.

Main Results:

• The proposed method effectively converges to the optimal solution for switched systems.
• The use of evolving suboptimal policies significantly decreases computational requirements.
• Numerical simulations validate the effectiveness of the introduced technique.

Conclusions:

• The novel ADP-based method provides an efficient solution for optimal tracking in switched systems.
• This approach offers a practical way to handle free final time and parameter optimization.
• The findings contribute to advancements in computational optimal control.