SHIP WASTE FORECASTING AT THE BOTAS LNG PORT USING ARTIFICIAL NEURAL NETWORKS


Satır T., Demir H., Alkan G. , Ucan O. N. , Bayat C.

FRESENIUS ENVIRONMENTAL BULLETIN, cilt.17, ss.2064-2070, 2008 (SCI İndekslerine Giren Dergi) identifier identifier

  • Cilt numarası: 17
  • Basım Tarihi: 2008
  • Dergi Adı: FRESENIUS ENVIRONMENTAL BULLETIN
  • Sayfa Sayıları: ss.2064-2070

Özet

Cargo and passenger vessels are required to give their waste to reception facilities when at port, and due to new regulations Turkish ports need to establish or reconstruct these facilities. It is thus very important for ports to be able to predict the quantity of waste. In this study, the authors use Artificial Neural Networks (ANNs) to model four years of data on the reception of ship's waste at the Botas LNG Port in Marmara Ereglisi, Turkey. Satisfactory results are obtained by the ANN outputs. and confirmed by classical approaches. This ANN forecasting model can be used by waste management companies to plan new ports.