Authors: Abiola, O.S., Owolabi, A.O., Sadiq, O.M. and Aiyedun, P.O.
ARPN Journal of Engineering and Applied Sciences, Vol. 7, No. 8, August 2012 ISSN 1819-6608
Abstract
One of the most common distresses on Lagos (the economic nerves centre of Nigeria) and Ibadan the largest city in West Africa Expressway pavement is surface rutting.
Rutting makes the road surface uneven, patchy and bumpy and subsequently affects the handling of vehicles which can lead to safety problems. The ability to predict the amount and growth of rutting in flexible pavements is an important aspect of pavement design. This paper presents the results of a research aimed at developing reliable and time – dependent Artificial Neural Network (ANN) based rut depth prediction model for Lagos – Ibadan Expressway. The model incorporate relevant variables such as pavement distresses, pavement layer thickness, pavement roughness, cumulative equivalent single axle load, sub grade California Bearing Ratio (CBR) and overlay asphalt concrete characteristics. The results showed that the forecasting accuracy of the 11-24-1 architecture is high compared with other tested architecture in terms of both average absolute error (AAE) and root mean square error (RMSE). The usage of the model will allow the road agencies to obtain reliable and accurate predictions of the future rut depth of the flexible pavements based on the given input variables.
Keywords: rut depth, artificial neural network, pavement, architecture, forecast.