Medium-Term Forecasts of Load Profiles in Polish Power System including E-Mobility Development

The main objective of this study was to conduct multi-stage and multi-variant prognostic research assess the impact e-mobility development on Polish power system for period 2022–2027. steps were as follows: forecast number electric vehicles (using seven methods), annual demand arising solely out operation vehicles, with without growth six daily profiles typical days three methods). For purpose research, we developed a unique Growth Dynamics Model in Poland. application Multi-Layer Perceptron (MLP) extrapolation non-linear functions (to growth) is our original, proposal use Artificial Neural Network (ANN). Another unique, innovative include Networks (Multi-Layer Long short-term memory (LSTM)) an Ensemble simultaneous 24 taking into account. This determined system, both terms within particular (hourly distribution) two (summer winter). Under (most likely) balanced variant power, due e-mobility, such would grow by more than 4%, almost 7% under optimistic variant. Percentage variation according time day determined. instance, variant, largest percentage share evening “peak” (about 6%), smallest night “valley” 2%).