Inversion Analysis on the Initial Damage of Concrete

Article Preview

Abstract:

The inverse analysis method based on genetic algorithm has been widely applied, but the general genetic algorithm is not applicable to the objective function as a non-analytic formula, so that the clear target function formula is often difficultly been obtained in the practical engineering. In order to solve the problem, the historical data are studied by neural network with the neural network embedded into the genetic algorithm, so as to establish the effective neural network model to replace the target function formula. At the same time, the corresponding fitness function is constructed for the optimal calculation. This method has been used for the inversion analysis on the initial damage of concrete, and the mechanical parameter inversion method is put forward based on the uniform design-genetic algorithm-neural network ensemble theory. Moreover, the method is verified by a calculation example, the results show that the initial damage value of concrete can effectively be obtained.

You might also be interested in these eBooks

Info:

Periodical:

Pages:

4048-4051

Citation:

Online since:

March 2014

Export:

Price:

Permissions CCC:

Permissions PLS:

Сopyright:

© 2014 Trans Tech Publications Ltd. All Rights Reserved

Share:

Citation:

* - Corresponding Author

[1] Qiu Ling, Xu Daoyuan. The experimental study on initial damage and damage evolution of concrete compression. Journal of HeFei University of Technology, vol. 24 (2011), p.1061.

Google Scholar

[2] Zhao Xinming, Liu Ning, Zhang Jian. The method of numerical inversion of rock and soil mechanics inverse analysis. Advances in Water Science, vol. 23 (2003), p.55.

Google Scholar

[3] Gao Wei, Zheng Yingren. The evolution inversion of geotechnical parameters. Journal of Hydraulic Engineering, vol. 23 (2012), p.47.

Google Scholar

[4] Zheng Lijun. The study on multi-objective optimization and decision-making method based on genetic algorithm. Journal of National University of Defense Technology, Vol. 12 (2013), p.64.

Google Scholar

[5] He Haoxiang, Yan Wei-ming, Wang Zhuo. The recursive model updating method based on the substructure and genetic neural network. Engineering Mechanics, vol. 25 (2008), p.99.

Google Scholar