Neural Network-Based Approach for Analysis of Rigid Pavement Systems Using Deflection Data

Abstract

This paper focuses on the development of backcalculation models based on artificial neural networks (ANNs) for predicting the layer moduli of the jointed plain concrete pavements, that is, the elastic modulus of the portland cement concrete (PCC) layer and the coefficient of subgrade reaction for the pavement foundation. The ANN-based models were trained to predict the layer moduli by using the falling-weight deflectometer (FWD) deflection basin data and the thickness of the concrete pavement structure. The ISLAB2000 finite element program, extensively tested and validated for more than 20 years, has been employed as an advanced structural model for solving the responses of the rigid pavement systems and generating a knowledge database. ANN-based backcalculation models trained with the results from the ISLAB2000 solutions have been found to be viable alternatives for rapid assessment (capable of analyzing 100,000 FWD deflection profiles in a single second) of the rigid pavement systems. The trained ANN-based models are capable of predicting the concrete pavement parameters with very low (<0.4%) average absolute error values. The ANN model predictions and closed-form solutions were compared through the use of the FWD deflection data, and the results are summarized in the paper. In addition, a sensitivity study was conducted to verify the significance of the layer thicknesses and the effect of bonding between the PCC and the base layer in the backcalculation procedure. The results of this study demonstrated that the ANN-based models are capable of successfully predicting the rigid pavement layer moduli with high accuracy.