Data sources Load models
Impact assessment models
Expert systems and artificial neural networks can be used for the prediction of energy demand. Expert systems ‘make decisions’ based on an interpretation of data and a selection among alternatives. Neural networks are used to make predictions based on a set of data.
The approach is suited for load modelling and energy estimations because it is able to handle incomplete data which might result form measured energy data and climatic parameters. Neural networks are trained in relation to a set of data until the network recognizes the patterns presented, and then it is capable on making predictions based on new patterns.
Published August 3, 2012
Contact: Gerd Kjølle, SINTEF Energy Research