Improving Prediction Accuracy of Concrete Compressive Strength via Wavelet Transform


Namli E., Erdal H. I., Erdal H.

JOURNAL OF POLYTECHNIC-POLITEKNIK DERGISI, cilt.19, sa.4, ss.471-480, 2016 (ESCI) identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 19 Sayı: 4
  • Basım Tarihi: 2016
  • Doi Numarası: 10.2339/2016.19.4.471-480
  • Dergi Adı: JOURNAL OF POLYTECHNIC-POLITEKNIK DERGISI
  • Derginin Tarandığı İndeksler: Emerging Sources Citation Index (ESCI), TR DİZİN (ULAKBİM)
  • Sayfa Sayıları: ss.471-480
  • İstanbul Üniversitesi Adresli: Evet

Özet

In recent years, Compressive strength prediction of concrete is being studied with an increasing speed by researchers. Instead of traditional statistical techniques, advanced prediction methods are being used in this area of study. In this study artificial neural network (ANN) and wavelet transform artificial neural network (WTANN) methods' prediction performances were compared on compressive strength of concrete with different mixture ratios and additionally effect of wavelet transform which decomposes dataset into subsets for a stationary situation for prediction was presented. Within this scope dataset trained in four different ways and sixteen different tests performed. The results of tests performed, WTANN achieves higher prediction performance in comparison with ANN. Hence, it's proved that WT could be used by researchers as an effective predictive tool for concrete compressive strength.