Ohio State Navbar

The Ohio State UniversityOffice of International Affairs

give a donation

A general framework of decentralized demand response in smart grid

Research Scholar

Di Guo, Electrical and Computer Engineering (China)
Adam Taylor, Co-Reseacher
Anupama Joseph, Co-Researcher
Robert Okyere, Co-Researcher
Besik Kankia, Faculty Mentor


Di Guo was born in 1986 and grew up in Hangzhou, China. He graduated from Zhejiang University at 2009 and has remained with the university as a PhD student in automatic control, electrical engineering. He arrived at Ohio State in May of this year and will go back China on November. His mentor here is Wei Zhang, an assistant professor in the Department of Electrical and Computer Engineering. His research interests include coordination control and optimization of multi-agent systems and its application in smart grids or other cyber physical systems.

What is the issue or problem addressed in your research?

Smart gird faces huge challenges nowadays. Because of unstable new energy generator such as wind or solar generator, the power supply will change accordingly. The problem arises that whether we can regulate the electricity usage of our home appliances e.g. air conditions, washers, etc to meet such change of supply side. One basic idea is that when the power supply is enough we can turn them on; on the other hand, when power is not enough turn them off. However, if all the households behave like that, then it will cause new problem. All turning on will result the original enough power not enough any more and vice verse. So we need some insights of this problem. That's to say we will design some control algorithms for our households to achieve this goal cooperatively.

What methodology did you use in your research?

This work presents a decentralized control algorithm for smart appliances to cooperatively accomplish the demand response in smart gird. A general framework of smart appliances is investigated with the dynamics described by hybrid systems. Based on the tool of potential game theory, the convergence of the proposed decentralized algorithm is promised. By introducing the coordination signal generated in the aggregator, the communication burden is reduced significantly. Although the global optimization is not convinced, the local optimal solution has acceptable performance. Several realistic scenarios are considered including Thermostatic Controlled Loads (TCLs, can be regarded as air condition) and Plug-in Electric Hybrid Vehicles (PHEVs, a kind of electric vehicles) charging which are essentially different. Simulation results both in offline planning and receding time horizon show that the aggregated power response can track a reference wave very well which is especially useful in peak reduction or valley filling.

What are the purpose/rationale and implications of your research?

By using our algorithm, the consumers will feel comfortable (i.e. the air temperature is always in a suitable range or the charging of electric vehicle is finished before the desired time). At the same time, the utility company will save a lot of money because they do not need to buy or generate extra electricity due to the peak consumptions.