The estimation of the electric energy obtained from domestic wastes by using artificial neural networks

Key words: Artificial neural networks, domestic wastes, multiple lineer regression analysis, storage area.Every day, the renewable energy resources gain in importance in meeting the worlds rapidly increasing energy needs. Within this framework, the methane gas accumulated because of the storage of domestic waste, the oldest method known to eliminate domestic waste, has also been considered. The control of methane (C~), with 21 times more greenhouse impact than carbon dioxide (C02) and 7300 times than chlorofluorocarbon (CFC) is an important task related to climate change.In some storage areas in the world and in Turkey, in the premises producing electricity from waste of the Istanbul Metropolitan Municipality?s ISTAC Inc, Kemerburgaz, where this method is applied, 900 daily electric production data were used. Estimations oriented to the future were made by using this data to artificial neural networks (ANNs) and adapting it to the multiple lineer regression technique (MLR). Although the absolute average error percentages of both applications are close to each other in term of results obtained, the best result was provided by the ANNs feed forward back propagation algorithm technique.It is evident that this method is likely to enlighten the path of global warming, the protection of the environment and the future oriented investments related to the production and work programs of energy production from solid wastes.