Computational & Technology Resources
an online resource for computational,
engineering & technology publications
Computational Science, Engineering & Technology Series
ISSN 1759-3158
CSETS: 23
SOFT COMPUTING IN CIVIL AND STRUCTURAL ENGINEERING
Edited by: B.H.V. Topping, Y. Tsompanakis
Chapter 5

Optimal Seismic Design of Structures using Soft Computing Techniques

E. Salajegheh and S. Gholizadeh

Department of Civil Engineering, University of Kerman, Iran

Full Bibliographic Reference for this chapter
E. Salajegheh, S. Gholizadeh, "Optimal Seismic Design of Structures using Soft Computing Techniques", in B.H.V. Topping, Y. Tsompanakis, (Editors), "Soft Computing in Civil and Structural Engineering", Saxe-Coburg Publications, Stirlingshire, UK, Chapter 5, pp 155-178, 2009. doi:10.4203/csets.23.5
Keywords: seismic loading, particle swarm optimisation, adaptive neuro-fuzzy inference system, wavelet transforms, radial basis function, neural network.

Summary
Optimal design of real structures subject to seismic loading is one of the major concerns in the field of structural engineering. In fact, assuming static response and ignoring the dynamic characteristics of the seismic action during the design optimisation phase may lead to structural configurations that are highly vulnerable to future earthquakes. In order to evaluate the time history responses of structures, seismic design codes specify that the structures should be analyzed against the seven or more pairs of horizontal ground motion components. Therefore, it is clear that the structural optimisation for the full seismic loading is a computationally intensive task and requires prohibitively high computer times for obtaining results from finite element (FE) analyses.

In this study we introduce a soft computing based methodology to significantly reduce the computational burden of the optimal seismic design of realistic steel moment-resisting frame structures.

In order to solve the optimisation problem the particle swarm optimisation (PSO) [1] and a multi-staged PSO (MSPSO) are employed. The proposed MSPSO optimisation algorithm is based on the implementation of PSO in a multi-stage manner and renewing the swarm at the beginning of each stage according to the best solution found in the previous stage. The MSPSO is better than the PSO in terms of search ability and the number of required generations.

The computational effort of the MSPSO algorithm, due to the performing time history analyses, is still extremely large. In order to reduce the computational burden, an efficient neural network system is proposed to accurately predict the necessary structural responses during the optimisation process by hybridization of adaptive neuro-fuzzy inference system (ANFIS) [2], wavelet transforms (WT) [3] and radial basis function (RBF) [4] neural networks.

In this study a ten storey steel moment-resisting frame structure subjected to full seismic loading is designed for optimum weight. For the seismic loading 10 pairs of the appropriate horizontal natural ground motion time-history components are selected and scaled according to the design response spectrum of the Uniform Building Code (UBC) [5]. All the code requirements are considered as the constraints of the optimal design problem. The numerical results demonstrate high computational efficiency of the proposed methodology.

References
[1]
J. Kennedy, "The particle swarm: social adaptation of knowledge", In: Proceedings of the international conference on evolutionary computation, Piscataway (NJ): IEEE, 303-308, 1977.
[2]
J.S.R. Jang, "ANFIS: Adaptive-network-based fuzzy inference systems", IEEE Transactions on Systems Man and Cybernetics, 23, 665-685, 1993. doi:10.1109/21.256541
[3]
I. Daubechies, "The wavelet transform, time-frequency localization and signal analysis", IEEE Transactions on Information Theory, 36, 961-1005, 1990. doi:10.1109/18.57199
[4]
P.D. Wasserman, "Advanced Methods in Neural Computing", Van Nostrand Reinhold, Prentice Hall Company, New York, 1993.
[5]
International Conference of Building Officials, "Uniform Building Code", Volume 2, Whittier, California, 1997.

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

go to the previous chapter
go to the next chapter
return to the table of contents
return to the book description
purchase this book (price £88 +P&P)