Hierarchical Distributed Model Predictive Control of Standalone Wind/Solar/Battery Power System

Abstract

A microgrid is a distributed networked generation system, which can effectively integrate various sources of distributed generation, especially renewable energy sources into the information network. The standalone wind/solar/battery power system is a typical standalone microgrid, in which the wind and solar power generations are the intermittent systems with complex dynamics and multiconstraints. Coordinated optimization between the wind power and solar power generations can effectively meet the load demand, reduce wear and tear of generating units, prolong the lifetime and, thus, guarantee the safety of the power grid. Regarding the large-scale, geographically dispersed standalone wind/solar/battery power generation system, this paper constituted a hierarchical distributed model predictive control (HDMPC). In this HDMPC, the upper layer utilizes an iterative distributed control strategy to realize the coordination of the power dispatch. It thus reaches the economic object, e.g., reducing the torsional shaft torque transmitted to gearbox in wind turbine system. The lower layer utilizes the supervisory predictive control to realize both the economic and tracking property. Under this hierarchical structure, the back-calculation from the lower control layer to the upper layer is utilized to keep the consistency of constraints. Through coordinated optimization among the subsystems, the proposed HDMPC realizes the plug and play of distributed energy. The simulation and the experiment validate the advantages of the proposed method in that it can realize the reliability, high efficiency, flexibility, and interactivity for the microgrid control.