Keywords: optimum design, earthquake, self-organizing networks, evolutionary algorithm.

The authors introduced an intelligent neural system (INS) for efficient approximation of time history structural responses in [1]. In INS, the input and target spaces are divided into some subspaces as the data located in each subspace have identical properties. These properties can be taken as some natural periods of the structures. Classification of input space is achieved by using competitive neural networks. Then a distinct radial basis function (RBF) network is trained for each subspace using its assigned training data. Also, the authors incorporated the intelligent neural system in the optimisation process in [2]. The numerical results showed great computational efficiency with a main limitation of difficulties for determining the number of data clusters.

In the present study, self-organizing map (SOM) neural networks are used to classify the input space. The SOM is a neural network algorithm developed by Kohonen [3] that forms a two dimensional presentation from multi dimensional data. The SOM neural networks learn to classify input vectors according to how they are grouped in the input space.

The main aim of this paper is to improve the INS by substituting the competitive neural network with a SOM. The resulting neural system is called the self-organizing neural system (SONS).

In all the examples, the input space includes natural periods of the structures and target space consists of corresponding responses of selected nodes and element stresses against the earthquakes. To provide training data, ANSYS [4] and to train the neural networks, MATLAB [5] are utilized, respectively.

The evolutionary algorithm used here is virtual sub population (VSP) method [6]. We design a 72-bar space truss subjected to the El Centro (S-E 1940) earthquake for optimal weight. The numerical results show that incorporating SONS in the framework of VSP creates a powerful tool for optimum design of structures.

In the full length paper, the details of the SONS, dynamic constraint treatment and numerical results are described.

1

S. Gholizadeh and E. Salajegheh, "An Intelligent Neural System for Predicting Structural Response Subject to Earthquakes", in Proceedings of the Fifth International Conference on Engineering Computational Technology, B.H.V. Topping, G. Montero and R. Montenegro, (Editors), Civil-Comp Press, Stirlingshire, United Kingdom, paper 63, 2006. doi:10.4203/ccp.84.63

2

E. Salajegheh, J. Salajegheh and S. Gholizadeh, "Structural Optimisation for Earthquake Loading Using Neural Networks and Genetic Algorithms", in Proceedings of the Eighth International Conference on Computational Structures Technology, B.H.V. Topping, G. Montero and R. Montenegro, (Editors), Civil-Comp Press, Stirlingshire, United Kingdom, paper 249, 2006. doi:10.4203/ccp.83.249

3

T. Kohonen, "Self-Organization and Associative Memory", 2nd Edition, Springer-Verlag, Berlin, 1987.

4

ANSYS Incorporated, "ANSYS Release 8.1", 2004.

5

The Language of Technical Computing, "MATLAB", Math Works Inc, 2004.

6

E. Salajegheh and S. Gholizadeh, "Optimum Design of Structures by an Improved Genetic Algorithm using Neural Networks", Advances in Engineering Software, 36, 757-767, 2005. doi:10.1016/j.advengsoft.2005.03.022

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £62 +P&P)