A machine learning based stochastic optimization framework for a wind and storage power plant participating in energy pool market

Renewable energy plants can participate in the energy pool market including day-ahead, intraday and balancing markets. The aim of this work is to develop a decision-making framework for a Wind and Storage Power Plant participating in the pool market to handle the uncertainty associated with the parameters of energy price and available wind energy, which are not known when decisions are to be made. Thus, the problem of maximizing the net income of such a plant participating in the pool market is formulated as a two-stage convex stochastic program. A novel hybrid approach using multivariate clustering technique and recurrent neural network is used to derive scenarios to handle the uncertainty associated with the energy price. Lastly, simulation experiments are carried out to show the effectiveness of the proposed methods using a real-world case study. Operators of variable renewable resource generators could use the proposed framework to make robust decisions and better manage their operations to gain competitive advantage.