Energy Management Strategy to Enhance a Smart Grid Station Performance: A Data Driven Approach

This paper proposes an energy management strategy (EMS) to enhance the power quality (PQ) parameters, i.e., voltage unbalance, power factor, and frequency deviation, of a smart grid station (SGS). Here, the SGS is represented as grid-connected multi-microgrids (MMGs), which are equipped with distributed generators (DGs), i.e., solar photovoltaic (PV) and wind turbines (WTs), battery energy storage systems (BESs), electric vehicle charging stations, capacitor banks, chillers, and building load power demand (LPD). Maintaining the PQ parameters of the SGS within the threshold limits is challenging due to the stochastic nature of building LPD and the dynamic behaviors of chiller operations. Furthermore, reactive power compensation with capacitor banks and robust control of DGs with BESs might not be a straightforward solution to improve the PQ parameters due to the nonlinearity of building LPD, the intermittent nature of DG power, and the limited capacity of BESs. In this context, an artificial neural network approach is used to predict the future values of building LPD, DG power, and cooling power demand. The SGS energy sources, converters, and grid connections are modeled at the system level. PQ index models are developed based on PQ parameter threshold limits. A fuzzy-based peerto-peer (P2P) energy-sharing strategy is developed based on a unique identification index, an energy-sharing index, and DGs' energy supplying, sharing, or buying to and/or from the neighborhood building. The BESs' charging and discharging control strategy is implemented based on the available energy in BESs. Furthermore, a cooling energy demand (CED) reduction strategy is implemented based on the predicted mean voltage and building CED index. Finally, an EMS is implemented for the SGS, which consists of existing and proposed EMS. The existing EMS is the baseline strategy, which provides available DG energy to the building, and deficit energy is supplied from the grid. The proposed EMS is the PQ parameter mitigation strategy, which maintains the PQ parameters within the threshold limits through the fuzzy-based P2P energy-sharing strategy. Measured data from the SGS are used to demonstrate the effectiveness of the proposed EMS. Through simulation studies, it is shown that the proposed EMS is capable of maintaining the PQ parameters within the threshold limits and reducing CED by concurrently enabling SGS energy.