Distributed model predictive control with guaranteed performance for reconfigurable power flow systems based on passivity

This paper presents a passivity-based distributed model predictive control algorithm that can provide minimum performance bound on power flow systems. The dynamics of large-scale power flow systems can be described by the transportation, conversion, and storage of energy among and across subsystems. By strategically choosing a passive output for each subsystem and augmenting each local model predictive controller with a special passivity constraint, the passivity property of each closed-loop subsystem can be guaranteed; in addition, the corresponding passivity index can be characterized by the designed parameters of the local controller. The passivity-preserving coupling between subsystems can be utilized to maintain passivity and the stability properties of the overall system. Moreover, the passivity index of the overall system can be characterized by the transition of passivity indices of subsystems based on interconnection topology. Based on the passivity index, the minimum performance bound of the overall power flow system can be characterized. Even if the structure of the power flow system changes, the entire system can remain stable and achieve a certain performance level. Simulation results of a fluid tank system show the effectiveness of the proposed algorithm.