Optimal Power Allocation of BESS in Microgrid using Machine Learning Techniques

Renewable energy sources (RES) are often considered unreliable due to their inherent fluctuations. However, the installation of a Battery Energy Storage System (BESS) can effectively address this issue. RES are often considered unreliable due to their inherent fluctuations. However, the installation of a Battery Energy Storage System (BESS) can effectively address this issue. This paper proposes a technique to attain the optimal allocation of a BESS where the optimal solution is decided by using the Long Short-Term Memory Algorithm (LSTM). The objective function is formulated to minimize the operational cost in the distribution grid while ensuring power balance and stability. Two cases are investigated in this paper the first case focuses on predicting the real-time power demand and generation patterns, along with BESS critical factors such as the state of charge, the availability of renewable energy sources (like solar intensity, and wind speed), and load patterns while the second case involves the optimal allocation of BESS in day-ahead market. The proposed technique has been applied to a generic distribution network and the results show that LSTM algorithm can provide the optimal allocation of the BESS in the distribution network, thus reducing the total system losses. This approach can be applied to various microgrid systems such as residential, commercial, and industrial, contributing to the development of smart grid technology and providing a practical solution for improving the reliability and sustainability of microgrids.