Prediction of Hybrid Fiber Reinforced Concrete Strength on Base of RBF & BP

Abstract:

The complex nonlinear problem exists between the factors which influence the strength of hybrid fiber reinforced concrete. The nonlinear relationship between input variables as the factors and output variables as the strength can be obtained artificial neural network which have the feature of self-adapting, self-studying and nonlinear mapping. RBF and BP neural network models are established in MATLAB in this paper based on experimental values. Compressive strength and flexural strength of hybrid fiber reinforced concrete are predicted respectively and comparing with the measured values, predicted values are analyzed. The results show that the predicted values based on RBF tally with the experimental values, and the strength prediction ability of RBF is higher than BP neural network. The needs of engineering can be meet by this method with accurate prediction ability, a new method for the study of hybrid fiber reinforced concrete strength is provided.

• Periodical: Advanced Materials Research (Volume 1035)
• Pages: 180-185
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.1035.180
• Online since: October 2014
• Authors: Nan Li*, Jun Hai Zhao, Juan Wang, Qian Zhu
• Keywords: Concrete Strength, Hybrid Fiber, Nonlinear Mapping, Prediction, RBF Neural Network

References

[1] Qiaowei Bi. Experimental research on the micro-structure of basalt fiber reinforced concrete and the oblique section bearing capacity of the BFRP bars reinforced fiber concrete beams[D]. Da Lian University of Tecnology Doctoral Dissertation, 2012. (in chinese).

DOI: 10.1016/j.compstruct.2022.115443

[2] Yi Liu. Study on hybrid fiber concrete property and prodict the concrete strength based on BP neural network[D]. Chong Qing University Master Dissertation, 2008. (in chinese).

[3] Tianqi Lv, Guofan Zhao, Zhishen Lin. Engineering Mechanics, Vol. 20-6(2003), p.52. (in chinese).

[4] Bingwei Xu, Xiliang Jiang. Engineering Mechanics, Vol. 26 (2009), p.163. (in chinese).

[5] Qiang Xue, Jiping Hao, Yue Zhen. Xi'an Univ. of Arch. &Tech. (Natural Science Edition), Vol. 43-2 (2011), p.192. (in chinese).

[6] Lijun Wang, Tiecheng Wang. Journal of Harbin Institute of Technology, Vol. 41-2 (2009), p.196. (in chinese).

[7] J. Amani, R. Moeini. Scientia Iranica A, Vol. 19-2 (2012), p.242.

[8] Defeng Zhang. MATLAB simulation and application of neural network, Beijing, 2009. (in chinese).

[9] Moody J, Darken C. Learning with Localized Receptive Fields. In Proc 1988 Connectionist Models Summer School, DTouretzky, G Hinton, and T Sejnowski (Eds. ), Carnegie Mellon University, Morgan Kaufmann Publishers, (1988).

DOI: 10.1145/1056754.1056760

[10] J, Darken C. Fast Learning in Networks of Locally-tuned Processing Units. Neural Computation, Vol. 1 (1989), p.281.

DOI: 10.1162/neco.1989.1.2.281

[11] Jing Liu, Qing Ye. Journal of Huaqiao University(Natural Science), Vol. 201 (2013), p.83. (in chinese).

[12] Shaofei Jiang. Joural of Harbin University of C. E. &Architecture, Vol. 32-5 (1999), p.24. (in chinese).