The invention provides a method and an apparatus for reinforcement learning based energy bidding, which can optimize a profit of the aggregator, a profit of the power producer of the renewable energy sources, and an expense of the green energy user.  In the method, a supply amount of each of the energy suppliers and a demand amount of each of the energy users are acquired.  A total demand amount of the energy users is calculated and replied to each energy supplier, and a total supply amount of the energy suppliers is calculated and replied to each energy user.  An electricity purchase quotation determined by each energy user according to its own demand amount and the total supply amount, and an electricity sale quotation determined by each energy supplier according to its own supply amount and the total demand amount are received.  A linear programming (LP) method is adopted to determine the energy supply configuration between the energy suppliers and the energy users.

The invention provides a method and an apparatus for reinforcement learning based energy bidding, which can optimize a profit of the aggregator, a profit of the power producer of the renewable energy sources, and an expense of the green energy user. The proposed framework is a hybrid in the sense that the advantages of energy trading using peer-to-peer or aggregation schemes can be preserved. Our analysis based on real-world data shows that the developed technology outperformed the  iterative double auction scheme: 1.11% increase in revenue for renewable generators and 0.61% decrease in costs for power users per hour, which is equivalent to  26.56% increase in revenue for renewable generators and 14.69% decrease in costs for power users per day.

This technology provides a hybrid bidding scheme that can take advantage of a peer-to-peer bidding scheme and an aggregate energy trading scheme. therefore, privacy of end users (EUs) and maintenance of communications infrastructure, among other things, can be ensured, facilitating energy service based business models. This hybrid bidding scheme allows power suppliers and EUs to meet their needs by adjusting desired prices sent to the aggregator, and the aggregator determines the amount of power bought from suppliers and the amount of power sold to EUs. Furthermore, the technology is suitable for unbundled RECs, which is the mainstream standard. It facilitates the interactions between renewable power generators, users, and aggregators of all sizes, boosting the energy market substantially.