With the increased uptake of  renewable energy in urban communities, there has been considerably more  discussion around the topic of microgrids. Small groups would operate an  electrical grid for their community at a lower cost than major  retailers, based on the use of renewable energy. Research conducted at  Southern University of Science and Technology (SUSTech) has determined  some unique energy management strategies that could be applied for smart  microgrids across the globe.

Assistant Professor Youwei Jia (Electrical and Electronic Engineering)  has led his research group to publish their research results on smart  microgrid operations in the high-impact academic journal, IEEE Transactions on Smart Grid (IF = 10.486). Their articles were titled “A Novel Retrospect-inspired Regime for Microgrid Real-time Energy Scheduling with Heterogeneous Sources” and “A Multi-agent Reinforcement Learning based Data-driven Method for Home Energy Management.” A third paper received the Best Paper Award at the virtual 2020 IEEE Power & Energy Society (PES) General Meeting. That paper was titled, “Real-Time Operation Optimization of Islanded Microgrid with Battery Energy Storage System.”

More specifically, smart microgrids represent an emerging paradigm of  distributed power systems. It integrates advanced techniques in  electricity transmission, monitoring & control, distributed  generation, and energy storage. In remote areas, there is an urgent need  for improved resilience in power networks, energy security, and  electrification. Smart microgrids will continue to play a vital role in  global energy transition, but there are significant economic and  technological challenges associated with

The first paper published in IEEE Transactions on Smart Grid,  titled “A Novel Retrospect-inspired Regime for Microgrid Real-time  Energy Scheduling with Heterogeneous Sources,” proposed a unique online  energy scheduling algorithm to deal with heterogeneous generators in  microgrids. The algorithm is unique because it is prediction-oblivious.  Hence, it pushes the operation of the microgrid towards economic  optimality faster. The paper shows the enormous potential for practical  applications in areas including networked microgrids

Fig.1 Intelligent microgrid operation

Fig.2 Hardware-in-the-loop Microgrid Platform

SUSTech is the corresponding unit for the paper. Assistant Professor  Youwei Jia is the first and corresponding author. Additional  contributions came from The Hong Kong Polytechnic University (PolyU) and City University of Hong Kong (CityU).

The second paper, titled “A Multi-agent Reinforcement Learning based  Data-driven Method for Home Energy Management,” assessed at smart home  energy management (HEM) systems. They are commonly used to integrate  into smart microgrids to maximize energy efficiency throughout the  entire system for all users. HEM systems are now faced with unique  challenges with the increased use of rooftop photovoltaic cells for  energy generation and the use of electric vehicles. The research team  took a data-driven approach to HEM systems to effectively manage  different household appliances through the use of smart “agents.” Big  data analytics would allow smart homes to assist in the demand-side  response to the electricity market. It provides enormous potential for  household savings on energy consumption.

Fig. 3 Smart home energy management

SUSTech is the corresponding unit. Visiting Scholar Dr. Xu Xu is the first author. Assistant Professor Youwei Jia is the correspondent author. Additional contributions came from Nanyang Technological University (NTU), PolyU, and Brunel University London.

The paper published at 2020 IEEE PES that received the Best Paper  Award was titled, “Real-Time Operation Optimization of Islanded  Microgrid with Battery Energy Storage System.” Energy storage systems  (ESS) are essential to provide the flexibility needed for microgrids. As  an essential component in the microgrid system, ESS can effectively  complement renewable generation and enhance the operational stability.  However, quantitatively modeling the cost of the degradation of  lithium-ion based ESS becomes a crucial problem in analyzing the  economic operation of islanded microgrids.

The paper proposed a real-time energy scheduling approach for  islanded microgrid by considering schedulable ESS with detailed  degradation cost modeling. Their hypothesis is dedicated to utilizing  ESS to achieve power smoothing in renewable energy embedded microgrid  operation. It would also avoid battery over-discharge while maximizing  energy efficiency and sustainability.

SUSTech is the corresponding unit. Visiting Ph.D. student Mr. Cheng Lyu is the first author. Assistant Professor Youwei Jia is the correspondent author. Additional contributions came from PolyU.