A Concept of Distributed Energy Management Systems with Fully Decentralized Primary Control Strategies for Microgrids

Due to environmental and economic incentives the penetration of distributed energy resources (DERs) is increasing worldwide. The concept microgrid (MG) is expected as a promising solution to deal with the huge amount of DERs. A MG contains usually interconnected loads, storages and DERs, which can operate both in grid-connected (GC) and islanded mode. In this thesis a control concept for MGs in both GC and islanded mode is presented. This control concept consists of a distributed energy management system (EMS) and a fully decentralized primary control system. The distributed EMS is based on a price-based model predictive control approach, where a so-called market-maker (MM) is introduced to coordinate the electricity consumers, resources and storages within a MG. The decisions on consumption, production and charging strategy are distributed to each electricity end-user, whereas the MM gathers the power demand from each end-user and initiates a negotiation process if the network requirements or the power balance would be violated. The electricity price would converge to a new point/sequence depending on the network conditions and a compromise for all parties would be found. The decentralized primary control system is designed based on the H-infinity-optimal control approach. To that purpose a network model with quick dynamics is set up. The H-infinity-norm of the performance channels mapping from the disturbances to the predefined performance outputs is minimized by tuning the control parameters collectively. Especially in islanded mode, the droop control law for inverters with a first-order low-pass filter has been interpreted as a mimic of classical synchronous generator dynamics with inertia. By tuning the virtual inertias of inverters a better performance can be achieved.