Data-Centric Hierarchical Distributed Model Predictive Control for Smart Grid Energy Management

The smart grid energy management with variable renewable energy resources presents many challenges to the grid operation. An optimized solution to manage the available resources is necessary to achieve reliable operation. This paper presents the hierarchical distributed model predictive control (HDMPC) to solve the energy management problem in the multitime frame and multilayer optimization strategy. The HDMPC combines the concept of enabling the optimization over long time-horizon for a centralized supervisory management (SM) layer and another short time-horizon during high-power variability for a distributed coordination management (CM) layer. The information exchange and interoperability between different layers are provided through the data-centric communication approach. The SM (upper layer) works to present the grid operator with certain operational plans and gives the guidelines to the CM (lower layer). The CM has the responsibility to coordinate the relationship between the centralized optimization objectives and the physical power system layer. The proposed HDMPC control was verified both numerically and experimentally. The obtained simulation results show that the control strategy proposed here is successful and combines the benefits of both the centralized and distributed control for a global solution of the grid operation problem. The experimental results demonstrate the feasibility of the real-time implementation of the proposed system for deployment to control future smart grid assets.