Akademik Veri Yönetim Sistemi
International Journal of Global Energy Issues, cilt.20, sa.3, ss.302-315, 2003 (Scopus)
Two methods are currently used to model residential energy consumption at the national or regional level: the engineering method and the conditional demand analysis (CDA) method. One of the major difficulties associated with the use of engineering models is the inclusion of consumer behaviour and socioeconomic factors that have significant effects on the residential energy consumption. The CDA method can handle socioeconomic factors if they are included in the model formulation. However, the multicollinearity problem and the need for a very large amount of data make the use of CDA models very difficult. It is shown in this paper that the neural network (NN) method can be used to model the residential energy consumption with the inclusion of socioeconomic factors. The appliances, lighting, and cooling component of the NN based energy consumption model developed for the Canadian residential sector is presented here and the effects of some socioeconomic factors on the residential energy consumption are examined using the model.