Date of Award
Open Access Thesis
With the environment protection and energy usage efficiency promotion issues concerned more and more by governments and industries these days, smart grid technologies have been in rapid development. On this new technology stage, residential site power management system has become a star. This paper introduces the concept, development of two high level control methods for “Smart Green Power Node” (SGPN), solutions of demand side energy management system. One control algorithm is optimal model method which uses the predictive data of a battery model with electrical parameters and pricing methods, to keep scheduling the optimal control input on the receding time horizon continuously. The other one is rule-based method which determines each time instant’s control inputs by logical judgment upon current and past battery electrical parameters and price states without model and predictive data in an easier implementing and faster calculating manner. Both algorithm systems incorporate necessary data such as photovoltaic solar power generation, user input, hardware states and load power information, as well as a TOU pricing schedule as inputs to the novel system control algorithm to achieve money saving for both electricity supplier and homeowner. Compared to commercial energy management systems which exist, this system can effectively eliminate the need for human interaction and intelligently regulate house power flow without complicated inputs from homeowner except some basic preferences.
According to the simulation of SGPN models comparison results carried out in MATLAB environment, the rule-based method can provide an annual total credit earn of more than $1000, based on reasonable industrial electric price and residential side feed-in tariff, yet guarantee a micro-second level calculation rate, which is valuable for real-time use scenario. And the load management during peak time is able to limit the average grid power to utility acceptable level. Moreover, the standalone mode DC bus voltage maintenance capability empowers the system’s self-sufficiency during grid outage period. On the other hand, the optimal model method is highly dependent on the accurate predictive model, which could not be practically economically efficient at present to implement but shows the optimality than rule-based method in theory and the potential of guarantee more residential profit in practice if such accurate model can be derived. These results show SGPN’s potential market value for both electricity consumers and providers.
Xu, H.(2014). High Level Control Implementation for a Residential Smart Grid Power System. (Master's thesis). Retrieved from http://scholarcommons.sc.edu/etd/2953