DISTRIBUTED MPC STRATEGIES FOR AUTOMATIC GENERATION CONTROL

The paper considers distributed model predictive control (MPC) strategies that are appropriate for controlling large-scale systems such as power systems. The overall system is decomposed into subsystems, each with its own MPC controller. To achieve performance equivalent to centralized MPC, these distributed regulators must work iteratively and cooperatively towards satisfying a common, systemwide control objective. Automatic generator control (AGC) provides a practical example for contrasting the performance of centralized and decentralized controllers.