| Summary: | Understanding the use of electrical appliances in households is crucial for improving the accuracy of electricity and energy loads forecasts. In particular, bottom-up techniques provide a powerful tool, not only for predicting demands considering socio-demographic characteristics of the occupants, but also to better resolve and implement demand side management strategies in homes.
With this purpose, a study of the temporal energy use of low-load appliances (meaning those whose annual energy share is individually negligible but relevant when considered as a group) has been carried out, with the longer term objective of finding a parsimonious approach to modelling them, and which considers an appropriate aggregation of appliances. In this work, a discrete-time stochastic process has been implemented for a specific classification of low-load appliances. More precisely, a time-inhomogeneous Markov chain has been used to model energy variations over time for four different categories of appliances and its prediction capabilities have been tested and compared.
|
|---|