Distributed Economic MPC for Dynamically Coupled Linear Systems With Uncertainties.

Area of Science:

• Control Engineering
• Optimization Theory
• Distributed Systems

Background:

• Complex systems often involve multiple interacting subsystems with individual constraints.
• Optimizing economic objectives in such systems is challenging due to disturbances and decentralized information.
• Existing control methods may struggle with scalability and robustness in distributed environments.

Purpose of the Study:

• To develop a novel distributed economic model-predictive control (MPC) algorithm for interconnected linear systems.
• To address challenges posed by unknown disturbances and subsystem-specific constraints.
• To ensure stability and recursive feasibility of the closed-loop system.

Main Methods:

• A distributed MPC approach is proposed, decomposing control computation into local optimization problems.
• Compatibility constraints are introduced to manage information exchange between subsystems.
• Constraints are tightened to account for state and input uncertainties under disturbances.
• Dissipativity and terminal constraints are incorporated to guarantee stability and feasibility.

Main Results:

• The proposed algorithm effectively optimizes economic objectives for a group of disturbed linear systems.
• Distributed implementation is achieved through local optimization with compatibility constraints.
• Recursive feasibility and closed-loop stability are formally established.
• The method's efficiency is validated through a building temperature control case study.

Conclusions:

• The developed distributed economic MPC algorithm offers an efficient and robust solution for complex control problems.
• The approach successfully handles disturbances and constraints in a decentralized manner.
• This work provides a foundation for advanced distributed control strategies in various applications.