Ramalan Cirian Reologi Campuran Berasfalt Menggunakan Rangkaian Saraf Tiruan
DOI:
https://doi.org/10.11113/jt.v65.1822Keywords:
rangkaian saraf tiruan, rangkaian saraf suap-depan pelbagai lapisan, rangkaian fungsi asas jejarian, modulus kompleks (E*) dan sudut fasa (δ)Abstract
Objektif utama kajian ini ialah untuk membangunkan dua jenis model rangkaian saraf tiruan iaitu rangkaian saraf suap-depan pelbagai lapisan dan rangkaian fungsi asas jejarian bagi meramal sifat reologi campuran berasfalt daripada segi i) modulus kompleks, E* dan ii) sudut fasa, δ. Kajian juga dilakukan bertujuan untuk menilai ketepatan kedua-dua jenis model tersebut dengan penentuan nilai parameter-parameter statistik seperti pekali penentuan (R2), ralat mutlak min (MAE), ralat kuasa dua min (MSE) dan ralat punca kuasa dua min (RMSE) bagi setiap model yang dibangunkan. Model-model peramalan di dalam kajian ini dibangunkan dengan menggunakan data-data E* dan δ daripada kajian terdahulu oleh sekumpulan penyelidik dari Pusat Kejuruteraan Pengangkutan Nottingham. Berdasarkan kepada analisis model rangkaian saraf tiruan tersebut, didapati bahawa kedua-dua model mampu meramal sifat reologi campuran berasfalt dengan sangat baik dengan nilai R2 yang melebihi 0.99. Perbandingan antara kedua-dua jenis model menunjukkan bahawa model rangkaian fungsi asas jejarian mempunyai ketepatan yang lebih baik daripada model rangkaian saraf suap-depan pelbagai lapisan dengan nilai R2 yang lebih tinggi dan nilai MAE, MSE dan RMSE yang lebih rendah. Dapat disimpulkan bahawa model rangkaian saraf tiruan yang tidak menggunakan ungkapan matematik ini boleh digunakan sebagai satu kaedah alternatif untuk meramal sifat reologi sesuatu campuran berasfalt.
Â
Â
References
Xiao, F. & Amirkhanian, S.N. 2009. Artificial Neural Network Approach Toestimating Stiffness Behavior of Rubberized Asphalt Concrete Containingreclaimed Asphalt Pavement. Journal of Transportation Enginering. DOI:10.1061/(ASCE)TE.1943–5436.0000014.
Federal Highway Administration. 2000. Temperature Predictions Andadjustment Factors for Asphalt Pavement. FHWA-RD-98-085.
Giuliana, G., Nicolosi, V. & Festa, B. 2012. Predictive Formulas of Complex Modulus for High Air Void Content Mixes. Transportation Research Board 91stAnnual Meeting.
Epps, A., Harvey, J.T., Kim, Y.R. & Roque, R. 2002. Structural Requirements of Bituminous Paving Mixtures. Transportation Research Board –Transportation in the New Millennium.
Ceylan, H., Gopalakrishnan, K. & Kim, S. 2009. Looking to the Future: Thenext-Generation Hot Mix Asphalt Dynamic Modulus Prediction Models. International Journal of Pavement Engineering. 10(5): 341–352.
Zeghal, M. 2008. Visco-elastic Portrayal of Bituminous Materials: Artificial Neural Network Approach. Proceedings of GeoCongress 2008. NewOrleans, Louisiana, 9–12 Mac.
Garcia, G. & Thompson, M. 2007. HMA Dynamic Modulus Predictive Models: A Review. Illinois Center for Transportation, University of Illinois. Report No.FHWA-ICT-07-005.
Freeman, J. A. 1994. Simulating Neural Networks with Mathematica. USA: Addison-Wasley Publishing Company, Inc.
Tasdemir, Y. 2009. Artificial Neural Network for Predicting Low Temperature Performances of Modified Asphalt Mixtures. Indian Journal of Engineering & Materials Sciences. 16: 237–244.
Hagan, M. T., Demuth, H. B. & Beale, M. H. 1996. Neural network Design. PWS Publishing. ISBN: 0-9717321-0-8.
Hassan, H. F. 2010. Artificial Neural Network Technique for Rainfall Forecasting Applied to Alexandria, Egypt. Tesis Master, Jabatan Kejuruteraan Awam & Struktur, Universiti Kebangsaan Malaysia.
Svozil, D., KvasniÄka, V. & PospÃchal, J. 1997. Introduction to Multi-layer Feed-forward Neural Networks. Chemometrics and Intelligent Laboratory Systems. 39(1997): 43–62.
Sulaiman, M. & El-Shafie, A., Karim, O. & Basri, H. 2011. Improved Water Level Forecasting Performance by Using Optimal Steepness Coefficients in an Artificial Neural Network. Water Resources Management. DOI: 10.1007/s11269-011-9824-z.
Hornik, K., Stinchcombe, M. & White, H. 1989. Multilayer feedforward networks are universal approximators. Neural Networks 2: 359–366.
Leshno, M., Lin, V. Y., Pinkus, A. &Schocken, S. 1993. Multilayer Feedforward Networks with a Non-Polynomial Activation Function Can Approximate Any Function. Neural Networks. 6(6): 861–867.
Saltan, M., Tiğdemir, M. & Karaşahin, M. 2002. Artificial Neural Network Application for Flexible Pavement Thickness Modeling. Turkish Journal Engineering & Environmental Sciences. 26: 243–248.
Lin, G. & Chen, L. 2004. A non-linear Rainfall-runoff Model Using Radial Basis Function Network. Journal of Hydrology. 289: 1–8.
Jain, A. K., Mao, J. & Mohiuddin, K. M. 1996. Artificial Neural Networks: A Tutorial. Computer–Special Issue: Neural Computing: Companion Issue to Spring1996 IEEE Computational Science & Engineering. 29(3): 31–44. DOI: 10.1109/2.485891.
Panas, A., Pantouvakis, J. & Lambropoulos, S. 2012. Non-linear Analysis of Concrete Pavement Construction by the Use of Artificial Neural Networks. Procedia- Social and Behavioral Sciences. 48: 3671–3680.
Huang, S., H, C., Lee, C. & Chang, C. 2005. Application of Neural Network For Selection of Airport Rigid Pavement Maintenance Strategies. Journal of Marine Science and Technology. 13(2): 125–132.
Downloads
Published
Issue
Section
License
Copyright of articles that appear in Jurnal Teknologi belongs exclusively to Penerbit Universiti Teknologi Malaysia (Penerbit UTM Press). This copyright covers the rights to reproduce the article, including reprints, electronic reproductions, or any other reproductions of similar nature.
