An LSTM-SAE-Based Behind-the-Meter Load Forecasting Method

Nowadays, modern technologies in power systems have been attracting more attention, and households can supply a portion of or all their electricity based on on-site generation at location. This be challenging for utilities terms monitoring recording the data because households’ facilities generate consume energy without passing it through meter, increasing complexity distribution network. The speed transferring to is another important concern. There necessity send smart meter (SM) each house management system (DMS) analysis shortest possible time. paper presents novel deep learning framework collaborating with sequence-to-sequence (seq2seq), long short-term memory (LSTM), stacked autoencoders (SAEs) forecast residential load profiles considering photovoltaic (PV), battery storage (BESS), electric vehicle (EV) loads capability pre-defined patterns. Experimental results show that proposed method achieves outstanding performance forecasting process comparison other algorithms. Also, transformer help receive instantly via wireless communication, which reduce transfer duration every minute make prediction manageable different combinations distributed resources (DERs) locations.