Load forecasting for energy communities: a novel LSTM-XGBoost hybrid model based on smart meter data

Energy Informatics

Table 2 Features for XGBoost datasets for peak load \(P_{max}\) and peak time \(t_{P_{max}}\) forecasting for each day d, based on values from previous day (\(d-1\)) or previous days. Features are either based on the aggregated energy community load (agg) or smart meter data of households N

Feature	\(t_{P_{max}}\) input data	\(P_{max}\) input data
\(P_{max,d-1,n}\) \(\forall\) n \(\in\) N	Yes	Yes
\(P_{min,d-1,n}\) \(\forall\) n \(\in\) N	Yes	Yes
\(P_{mean,d-1,n}\) \(\forall\) n \(\in\) N	Yes	Yes
\(P_{median,d-1,n}\) \(\forall\) n \(\in\) N	Yes	Yes
\(P_{\sigma ,d-1,n}\) \(\forall\) n \(\in\) N	Yes	Yes
\(t_{P_{max,d-1, agg}}\), \(\ldots\), \(t_{P_{max,d-21,agg}}\)	Yes	No
\(t_{P_{max,d-1,n}}\), \(\ldots\), \(t_{P_{max,d-21,n}}\) \(\forall\) n \(\in\) \(N_{large}\)	Yes	No
\(P_{max,d-1, agg}\), \(\ldots\), \(P_{max,d-21,agg}\)	No	Yes
\(P_{max,d-1,agg}\)	No	Yes
\(t_{P_{max,d-1,agg}}\)	Yes	No

Input features include day-before maximum loads, minimum loads, mean loads, median loads and load standard deviation (\(P_{max,d-1,n}\), \(P_{min,d-1,n}\), \(P_{mean,d-1,n}\), \(P_{median,d-1,n}\), \(P_{\sigma ,d-1,n}\)). For the peak time forecast, especially day-before peak load times \(t_{P_{max,d-1}}\) are relevant