Artificial Neural Network-Based New Methodology for Modeling of Asphalt Mixtures and Comparison with IKE Method

Abstract

Artificial Neural Networks (ANNs) are the most adopted approach in modeling of engineering problems. In this paper, we have developed ANN-based a novel modeling approach for asphalt mixtures. The Flow, Stability and MQ of the mixtures have been modeled and predicted by the introduced ANN-based approach. The legibility, comprehensibility, consistency, estimation performance, standard deviation etc. of the presented approach has been compared with the previous study. The experimental studies have shown that the proposed approach provides robustness, stability and a high accuracy ratio for estimation the Flow, Stability and MQ. While this paper has presented a novel approach to modeling the asphalt mixtures, it has also verified the results of literature. Thus, powerful, efficient and alternative approaches were presented to the literature for modeling the asphalt mixtures.

Keywords

• Intuitive k-nearest neighbor estimator (IKE),Artificial neural networks (ANN),Asphalt mixtures,Marshall stability test

References

1. Ibrahim M. A. Performance evaluation of SUPERPAVE and Marshall asphalt mix designs to suite Jordan climatic and traffic conditions. Constr Build Mater 2007;21:1732–1740.
2. Tapkın S, Cevik A, Usar U. Prediction of Marshall test results for polypropylene modified dense bituminous mixtures using neural networks. Expert Syst Appl 2010;37(6):4660–4670.
3. Ozgan E. Artificial neural network based modeling of the Marshall Stability of asphalt concrete. Expert Syst Appl 2011;38:6025–6030.
4. Hınıslıoglu S, Agar E. Use of waste high density polyethylene as bitumen modifier in asphalt concrete mix. Mater Lett 2004;58:267–271.
5. Kuloglu N. Effect of astragalus on characteristics of asphalt concrete. J Mater Civil Eng 1999;11(4):283-286.
6. Robertus C, Mulder EA, Koole RC. SBS modified bitumen for heavy duty asphalt pavements. In: Second international conference on roads and airfield pavement technology, Singapore; September 1995.
7. Zoorob SE, Suparma LB. Laboratory design and investigation of the properties of continuously graded asphaltic concrete containing recycled plastics aggregate replacement (plastiphalt). Cem Concr Compos 2000;22:233–242.
8. Nijboer LW. Some considerations of the Marshall test method for investigating bituminous masses. Strasse Autobahn 1957:210–4.

9. Haddadi S, Ghorbel E, Laradi N. Effects of the manufacturing process on the performances of the bituminous binders modified with EVA. Constr Build Mater 2008;22:1212–9.
10. Mirzahosseini MR, Aghaeifar A, Alavi AH, Gandomi AH, Seyednour R. Permanent deformation analysis of asphalt mixtures using soft computing techniques. Expert Syst Appl 2011;38: 6081–6100.
11. Ozsahin TS, Oruc S. Neural network model for resilient modulus of emulsified asphalt mixtures, Constr Build Mater 2008;22:1436–1445.
12. Gopalakrishnan K, Manik A. Co-variance matrix adaptation evolution strategy for pavement backcalculation. Constr Build Mater 2010;24:2177–2187.
13. Attoh-Okine NO. Grouping pavement condition variables for performance modeling using self-organizing maps. Comput-Aided Civ Inf 2001;16(2):112–125.
14. Lee BJ, Lee HD. Position-invariant neural network for digital pavement crack analysis. Comput-Aided Civ Inf 2004; 19(2):105–118.
15. Mei X, Gunaratne M, Lu JJ, Dietrich B. Neural network for rapid evaluation of shallow cracks in asphalt pavements. Comput-Aided Civ Inf 2004;19(3):223–230.
16. Ceylan H, Guclu A, Tutumluer E, Thompson MR. Backcalculation of full-depth asphalt pavement layer moduli considering nonlinear stressdependent subgrade behavior. The International Journal of Pavement Engineering 2005;6(3):171–182.
17. Attoh-Okine NO. Modeling incremental pavement roughness using functional network. Can J Civil Eng 2005;32(5):805–899.
18. Saltan M, Sezgin H. Hybrid neural network and finite element modeling of sub-base layer material properties in flexible pavements. Mater Design 2007;28(5):1725-1730.
19. Tapkın S, Cevik A, Usar U. Accumulated strain prediction of polypropylene modified marshall specimens in repeated creep test using artificial neural networks. Expert Syst Appl 2009;36(8):11186-11197.
20. Banzhaf W, Nordin P, Keller R, Francone FD. Genetic programming–An introduction on the automatic evolution of computer programs and its application. Heidelberg/San Francisco: dpunkt/Morgan Kaufmann; 1998.
21. Koza J, Genetic programming, on the programming of computers by means of natural selection. Cambridge, MA: MIT Press; 1992.
22. Cevik A, Cabalar AF. Modeling damping ratio and shear modulus of sand–mica mixtures using genetic programming. Expert Syst Appl 2009;36(4):7749–7757.
23. Cevik A. A new formulation for web crippling strength of cold-formed steel sheeting using genetic programming. J Constr Steel Res 2007;63(7):867–883.
24. Gandomi AH, Alavi AH, Kazemi S, Alinia MM. Behavior appraisal of steel semirigid joints using linear genetic programming. J Constr Steel Res 2009;65(8–9):1738–1750.
25. Johari A, Habibagahi G, Ghahramani A. Prediction of soil–water characteristic curve using genetic programming. J Geotech Geoenviron Eng – ASCE 2006;132(5):661–665.
26. Aksoy, A. Iskender, E. Kahraman. HT. Application of the intuitive k-NN Estimator for prediction of the Marshall Test (ASTM D1559) results for asphalt mixtures. Constr Build Mater 2012;561–569.
27. Kahraman HT, Bayindir R, Sagiroglu S. A new approach to predict the excitation current and parameter weightings of synchronous machines based on genetic algorithm-based k-NN estimator. Energ Convers Manage, doi.10.1016/j.enconman.2012.05.004.
28. Kelly JD, Davis L. A hybrid genetic algorithm for classification, IJCAI-91, Morgan Kaufmann Publishers Inc. San Francisco CA USA 1991;645-650.
29. Goldberg DE. Genetic Algorithms in Search Optimization and Machine Learning. Addison Wesley, ISBN 0201157675. The University of Alabama USA 1989.
30. Mitchell TM. Machine Learning, McGraw-Hill Science/Engineering/Math 1997; 154-184.
31. Thodesen C, Xiao F, Amirkhanian SN. Modeling viscosity behavior of crumb rubber modified binders Constr Build Mater 2009;23:3053–3062.
32. Hao W, Lu Z, Wei P, Feng J, Wang B. A new method on ANN for variance based importance measure analysis of correlated input variables. Structural Safety 2012;38:56–63.
33. Ferreira IML, Gil PJS. Application and performance analysis of neural networks for decision support in conceptual design. Expert Syst Appl 2012;39:7701–7708.