Energy Demand Analysis and Forecast

摘要: Sustainable energy systems are necessary to save the natural resources avoiding environmental impacts which would compromise development of future generations. Delivering sustainable will require an increased efficiency generation process including demand side. The architecture supply can be characterized by a combination conventional centralized power plants with increasing number distributed resources, cogeneration and renewable systems. Thus efficient forecast tools predicting for operation planning role forecasting in deregulated markets is essential key decision making, such as purchasing generating electric power, load switching, side management. This chapter describes data analysis basics mathematical modeling demand. problem discussed context management Because large influence factors their uncertainty it impossible build up ‘exact’ physical model Therefore calculated on basis statistical models describing climate operating conditions consumption. Additionally artificial intelligence used. A variety methods ideas have been used (see Hahn et al., 2009, or Fischer, 2008). quality depends significantly availability historical consumption well knowledge about main parameters These also determine selection best suitable tool. Generally there no 'best' method. very important proof available exact application Within this algorithm building treatment methods. results interpreted tests. focus investigation lies regression neural networks heat It shown that similar applied both tasks. described demonstrated real district heating system containing different units.