The research team has developed numerous control algorithms for microgrid applications that rely on a distributed decision-making approach wherein control decisions pertaining to individual assets in the microgrid are based on information acquired locally and other information obtained by exchanges with other nearby assets. Through these works, the team has shown that the distributed decision making approach has the potential for realizing, in a scalable fashion, all the control functions that are necessary for stable, reliable, and resilient microgrid operation. The objectives of this project are then:
- To integrate the distributed control algorithms into a hardware platform;
- To demonstrate the effectiveness of this platform and the companion algorithms in performing standard microgrid control functions, as defined in IEEE 1547, IEEE 2030.7, and IEEE 2030.8;
- To demonstrate the effectiveness of this platform for (i) networking multiple microgrids, and (ii) integrating additional critical loads, generation, and storage assets into an existing microgrid.
Instead of having a centrally-located computing device implementing all control functions, the proposed solution is distributed, i.e., a microgrid is endowed with multiple, geographically dispersed, computing devices referred to as control nodes. The control nodes use the information they acquire locally, e.g., from measurements, and via exchanges with nearby control nodes as inputs to a suite of distributed algorithms implementing different control functions (both for grid-connected and islanded operation), e.g., primary and secondary frequency control, voltage control, and optimal generation asset dispatch, as well as handling transitions from grid-connected to islanded operation and vice versa. The final project deliverable is an industrial-grade control node prototype that integrates the aforementioned distributed control algorithms. By using controller hardware in the loop (C-HIL) testing and quantitative reliability analysis, the research team will demonstrate that using a collection of such devices to control the assets of a microgrid provides a plug-and-play control platform that is robust and resilient. In the process, the team will also demonstrate that such a control platform provides a cost-effective solution for promoting a seamless interconnection and interoperability of multiple microgrids, as well as enabling a plug-and-play expansion of an existing microgrid.
The research team believes that the developed technology can be adopted by all Department of Defense (DoD) military microgrids to improve energy resilience. Unlike centralized counterparts, a control architecture based on a distributed platform enables managing and processing real-time measurement data locally at control nodes, and obviates the need for moving the data to a central processing unit. This enables the development of algorithmic solutions that are far more scalable than those based on a centralized architecture, and results in a reduction of communication requirements. Having multiple control nodes further enhances resiliency of the system to node misbehavior due to cyber attacks or hardware failures, and minimizes the risk of disrupting the operation of vital control functions. This also implies that generation assets can seamlessly connect or disconnect from the system, thus enabling plug-and-play functionality. Also, the risk of a system wide failure is far less likely to be caused by a control node being compromised by a cyber attack. Hence, the execution of the control tasks in a distributed way boosts the level of cyber and physical security, which is essential for DoD microgrids. A distributed architecture also enables high-scale penetration of renewables, since it facilitates the design of control strategies that are capable of quickly computing the set-points for a large population of renewable-based generation resources, and provide a quick response to rapid and large fluctuations that are typical of renewable generation due to cloud or wind transients.