An Agile Solution to Estimate the Capacity of Virtual Energy of Refrigerator Using ANN

Integrating a renewable energy source (RES) into a distribution network is challenging due to its intermittent nature. One of the methods to address the issue is a smart energy storage (SES) system, which is achieved by a stationary battery system (SBS). SES can be reached by using demand response management (DRM) when the loads work with long thermal constants. However, the SBS faces life cycle issues, unsteady charging and discharging, high cost, and sizing complexities. In lieu of SBS, temperature-controlled electrical loads (TCLs) like air conditioners, refrigerators, water heaters, and similar appliances with virtual energy storage (VES) capacity can be considered in the context of SES. During peak demand periods, VES and DRM are implemented to ensure the grid's stability and bring equilibrium between supply and demand. Utilizing artificial neural network (ANN) models to capture their energy capacity and self-discharge, residential refrigerators analogously hold promise in predicting energy consumption and performance. This research addresses while SBS energy capacity is insufficient to meet peak demand, by utilizing residential refrigerators as virtual energy storage systems (VESS), enabling more efficient and reliable energy management in the context of smart grid systems and the intermittent challenge of RES integration.