Very short term load forecasting of residential electricity consumption using the Markov-chain mixture distribution (MCM) model

This study utilizes the Markov-chain mixture distribution model (MCM) for very short term load forecasting of residential electricity consumption. The is used to forecast one step ahead half hour resolution consumption data from Australia. results are compared with Quantile Regression (QR) and Persistence Ensemble (PeEn) as advanced simple benchmark models. were in terms reliability, reliability mean absolute error (rMAE), prediction interval normalized average width (PINAW) continuous ranked probability score (nCRPS). For 10 steps conditioning QR PeEn, MCM on par QR, superior PeEn. As a sensitivity analysis, simulations performed where number points PeEn was varied output, which based only point conditioning. It shown that nCRPS rMAE converged towards lower included QR. never reached results, but rMAE, above 24, most reliable. Based sparse complexity design MCM, high computational speed competitive performance, it suggested candidate probabilistic