Distributed Supervisory Predictive Control of Distributed Wind and Solar Energy Systems

In this work, we design a distributed supervisory model predictive control (MPC) system for optimal management and operation of distributed wind and solar energy generation systems integrated into the electrical grid to facilitate the development of the so-called “smart electrical grid”. We consider a topology in which two spatially distributed energy generation systems, a wind subsystem and a solar subsystem, are integrated in a DC power grid, providing electrical power to a local area, and each subsystem is coupled with an energy storage device. A supervisory MPC optimization problem is first formulated to take into account optimality considerations on system operation and battery maintenance; then a sequential and an iterative distributed supervisory MPC architectures are developed to coordinate the actions of the subsystems accordingly. Simulations of 24-hour system operation are carried out under the different control architectures to demonstrate the applicability and effectiveness of the distributed supervisory predictive control design.

I. Introduction

In a traditional electrical grid, electrical power generated by large, centralized power plants is transmitted to end users using one-directional power flows. In recent years there have been many calls for the development of the so-called “smart electrical grid” (e.g., [1]–[3]) by expanding the traditional electrical grid with distributed, medium-scale renewables-based energy generation systems to better meet the increasing energy demand and environmental regulations. A “smart electrical grid” is expected to be more reliable, more energy efficient and more environmentally friendly. Specifically, it is possible to integrate distributed energy resources and generation systems, like wind and solar systems, into the electrical grid.