Optimal Decision Tree Design for Near Real-Time Control of Battery Energy Storage Systems

The increasing numbers of distributed energy resources (DER) incur new challenges for the energy supply due to the volatilities and uncertainties of renewable energies. To utilize the benefits of DER, a combination with storage systems and intelligent controls is necessary. Commonly used control methods are based on Model Predictive Control (MPC) with forecast and optimization, which require computational capabilities at the DER on a level, on which the distribution system operator (DSO) has no access. This paper proposes a control algorithm based on data, which combines a model-based and an artificial intelligence (AI) based approach to utilize benefits from both methods while compensating for their drawbacks. Mixed-Integer Quadratic Programming (MIQP) is used to generate training and testing data for a decision tree (DT). The investigation of the optimal composition of the training data and the optimal architecture of the DT are the main focus of this paper.