NEURO-FUZZY APPROXIMATION FOR PREDICTION OF SIGNIFICANT WAVE HEIGHTS: THE CASE OF FILYOS REGION


TÜR R., BALAS C. E.

JOURNAL OF THE FACULTY OF ENGINEERING AND ARCHITECTURE OF GAZI UNIVERSITY, cilt.25, sa.3, ss.505-510, 2010 (SCI-Expanded) identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 25 Sayı: 3
  • Basım Tarihi: 2010
  • Dergi Adı: JOURNAL OF THE FACULTY OF ENGINEERING AND ARCHITECTURE OF GAZI UNIVERSITY
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, TR DİZİN (ULAKBİM)
  • Sayfa Sayıları: ss.505-510
  • Anahtar Kelimeler: Wave Prediction, Neuro-Fuzzy approximation, ANFIS, Filyos, Coastal structures, WIND, PARAMETERS, NETWORKS, MODEL
  • Akdeniz Üniversitesi Adresli: Evet

Özet

First and the most important phase of designing high cost coastal structures is to forecast the significant wave height (H-s). In this study, an alternative approach to the deterministic and stochastic methods found in literature is proposed. Neuro-fuzzy is a method in which advantages of Artificial Neural Networks (ANN) and Fuzzy Logic (FL) are combined. In the Neuro-Fuzzy models developed in this study, daily significant wave height can be estimated using daily average H-s and significant wave period (T-s) data. H-s and T-s data recorded at different time intervals were used in the proposed models. In this study, H-s and T-s measured in Filyos region of the Black Sea was estimated by the Adaptive Network-based Fuzzy Inference System (ANFIS). Predicted results from proposed models were compared with the measured wave data and it is found that high correlation values are obtained. It is thus concluded that the proposed models can efficiently be used to estimate missing wave data.

First and the most important phase of designing high cost coastal structures is to forecast the significant wave height (H(s)). In this study, an alternative approach to the deterministic and stochastic methods found in literature is proposed. Neuro-fuzzy is a method in which advantages of Artificial Neural Networks (ANN) and Fuzzy Logic (FL) are combined. In the Neuro-Fuzzy models developed in this study, daily significant wave height can be estimated using daily average H(s) and significant wave period (T(s)) data. H(s) and T(s) data recorded at different time intervals were used in the proposed models. In this study, H(s) and T(s) measured in Filyos region of the Black Sea was estimated by the Adaptive Network-based Fuzzy Inference System (ANFIS). Predicted results from proposed models were compared with the measured wave data and it is found that high correlation values are obtained. It is thus concluded that the proposed models can efficiently be used to estimate missing wave data.