Short-term hourly load forecasting of a hospital using an artificial neural network
DOI:
https://doi.org/10.24084/repqj09.355Abstract
Electricity demand forecasting is important for utilities and for some costumers. It allows to balance energy production and consumption. From the costumer point of view, it is essential for efficient operation, sizing of installation, maintenance scheduling, to name just a few. Load forecasting is a difficult task and there are many tools available to perform it. Among them, Artificial Neural Networks are receiving a lot of attention because it is not needed to know any relationship between the involved variables. But, they are constructed as black boxes, what is one their drawbacks. In this paper, some results of the load demand forecasting of a hospital are shown. It is important the previous statistical analysis of the load curves and how the results are improved adding new information to the training data set, as maximum and minimum daily temperatures.