High-Resolution Load Forecasting on Multiple Time Scales Using Long Short-Term Memory and Support Vector Machine

Electricity load prediction is an essential tool for power system planning, operation and management. The critical information it provides can be used by energy providers to maximise efficiency minimise costs. Long Short-Term Memory (LSTM) Support Vector Machine (SVM) are two suitable methods that have been successfully analysing time series problems. In this paper, the algorithms explored further prediction; developed verified using half-hourly data from University of Warwick campus centre with four different horizons. novelty lies in comparing accuracy intelligent multiple scales exploring better scenarios their applications. High-resolution forecasting over a long range also conducted paper. MAPE values LSTM 2.501%, 3.577%, 25.073% 69.947% horizons delineated. For SVM, 2.531%, 5.039%, 7.819% 10.841%, respectively. It found both shorter horizon predictions. results show more capable ultra-short short-term forecasting, while SVM has higher medium-term long-term forecasts. Further investigation performed via blind tests test consistent.