Application of artificial neural networks in the back-calculation of flexible pavement layer moduli from deflection measurements

Abstract

Artificial Neural Networks (ANN) are introduced in this paper with an example application given demonstrating the back calculation of layer elastic moduli (E-moduli) form Falling Weight Deflection (FWD) measurements. An ANN is a network of interconnected elements with the function to produce an output pattern when presented with an input pattern. Deflection measurements with a FWD are meaningful indicators of pavement strength. Back-calculation is used to back-calculate E-moduli of pavement layers that can be used in a mechanistic approach to estimate remaining pavement life from pavement response. It is shown that ANNs can back-calculate E-moduli, but with different degrees of success. The main problem identified is the basis on which the dataset used to train ANNs, is generated using linear elastic theory. ANNs can back-calculate E-moduli very successfully from ideal deflection basins at lightning fast speeds. Improvements in the data generation process using a theory that accommodates non-linear and stress dependent behaviour of materials may result in improved performance of the ANNs. It is also shown that it is very difficult to design a single ANN that can be successfully used on all the possible pavement types. It is better to identify representative pavement types and train ANNs for each of these pavements.