Towards application of linear genetic programming for indirect estimation of the resilient modulus of pavements subgrade soils

Abstract

The success of a flexible pavement design depends on the accuracy of determining the structural response of the pavement to dynamic loads, known as resilient modulus (M_R). This paper explores the potential of a branch of the computational intelligence techniques, namely linear genetic programming (LGP), for indirect estimation of the M_R of pavements subgrade soils. Furthermore, a pre-design model is proposed which characterises M_R factor in terms of subgrade soil properties and applied stress states utilising a selected database which comprises of several test results conducted on cohesive Ohio A-6 soils. In order to assess the degree of accuracy and reliability of the obtained model, various statistical criteria and verification study phases are conducted. Finally, better results of the obtained model in comparison with traditional models prove the robustness and capability of LGP approach for indirect estimation of M_R of pavement subgrade soils.