PROJECT TITLE: A Robust Model Predictive Control Framework for Optimal Operation of Microgrids under Uncertainty
Collaborator: Aalborg University.
Power system integration in a flexible and smart way where supply side and demand side are simultaneously managed has been on technology roadmaps for almost every utility and Independent System Operators (ISO) over the last few decades. With the high penetration of clean energy resources, the current power system faces significant challenges. A smart grid (SG) includes distributed energy resources (DER), energy storage system (ESS) and adjustable loads, which are expected to maximize flexibility while decreasing the operation cost of such grids if properly coordinated. A way of simplifying this coordination is the introduction of intelligent Microgrids (MGs), which act as an intermediate aggregation entity between the individual units and the overall SG.
The adoption of MGs for the massive integration of DERs, ESS and loads will reduce the need for complex and ramified centralized management. However, all these potential benefits are meaningless if internal MG operation is not optimized. Optimization is extremely important for managing its components cost-efficiently, and to harness all the potential privileges of distributed generation.
In this respect, demand and supply uncertainties in a MG present the significant challenge. Moreover, MGs modelling requires both discrete variables such as ON/OFF states of units (unit commitment) and controllable loads (DSM) and continuous variables like ES charge and discharge rate. Therefore, the overall problem needs to be posed as a mixed integer nonlinear problem (MINLP), for which there is no exact solution. In addition, there are no certain solution techniques for the MGs optimization problem.
This project will investigate the optimal operation of MGs considering uncertainties. The project will aim at addressing the need for efficient and viable MGs operation model in which the overall operation cost to meet forecasted demand will be minimized in a period of time (e.g. one day), and also MG uncertainties will be mitigated.
Publications in journals and conference papers may be found at VBN.