Prediction of resilient modulus for subgrade soils based on ANN approach

The resilient modulus (MR) of subgrade soils is usually used to characterize the stiffness and a crucial parameter in pavement design. In order determine compacted quickly accurately, an optimized artificial neural network (ANN) approach based on multi-population genetic algorithm (MPGA) was proposed this study. MPGA overcomes problems traditional ANN such as low efficiency, local optimum over-fitting. developed method consists ten input variables, twenty-one hidden neurons, one output variable. physical properties (liquid limit, plastic plasticity index, 0.075 mm passing percentage, maximum dry density, moisture content), state variables (degree compaction, content) stress (confining pressure, deviatoric stress) were selected variables. MR directly Then, adopting large amount experimental data from existing literature, compared with representative estimation methods. results show that has advantages fast speed, strong generalization ability good accuracy estimation.