Published: 2012-04-01

Identification of number and thickness of new test section pavement model layers using artificial intelligence methods

Andrzej Pożarycki

Abstract

The paper presents methodology based on feedforward Artificial Neural Networks (ANN) techniques to estimate the most probable number and thickness of new test section flexible pavement layers. There is a consideration in the described method that only results from FWD measurements on surface of each asphalt concrete layer are known. Based on pavement mechanics theory and making a reference to typical flexible pavements used in Poland, the deflection basins were calculated. Theoretically determined deflection basins were used to train ANN. Finally an artificial neural network approach is used to estimate the wanted parameters of analyzed test section pavement layers. Comparing the ANN’s results with the real test section pavement construction it was found that parameters obtained with ANN can be used for further standard backcalculation procedure.

Keywords:

backcalculation, feedforward artificial neural networks, pavement layers thickness identification

Similar Articles

<< < 9 10 11 12 13 14 15 16 > >> 

You may also start an advanced similarity search for this article.

Download files

PDF (Język Polski)

Citation rules

Pożarycki, A. (2012). Identification of number and thickness of new test section pavement model layers using artificial intelligence methods. Roads and Bridges - Drogi I Mosty, 11(2), 123–149. Retrieved from https://rabdim.pl/index.php/rb/article/view/v11n2p123

Cited by / Share


This website uses cookies for proper operation, in order to use the portal fully you must accept cookies.