Artificial Neural Network Models to Estimate Resilient Modulus of Cementitiously Stabilized Subgrade Soils

A combined laboratory and modeling study was undertaken to develop a database for cementitiously stabilized subgrade soils in Oklahoma and to develop artificial neural network (ANN) models that could be used to estimate resilient modulus (Mr) from commonly used subgrade soil properties in Oklahoma. An Mr database was developed using laboratory test results on 160 specimens prepared by using four soils stabilized with three cementitious additives, namely, lime (3%, 6% and 9%), class C fly ash (CFA) (5%, 10% and 15%) and cement kiln dust (CKD) (5%, 10% and 15%). One Multi-Layer Perceptrons Network (MLPN) and one Radial Basis Function Network (RBFN) types of ANN models were developed using a development dataset and validated using a different dataset. Overall, MLPN neural network was found to show best acceptable performance for the present evaluation and validation datasets.


  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 01487116
  • Record Type: Publication
  • Files: TRIS
  • Created Date: Jun 5 2013 3:32PM