Structural optimization

, Volume 8, Issue 4, pp 242–250 | Cite as

Neurocomputing strategies in structural design — decomposition based optimization

  • Z. P. Szewczyk
  • P. Hajela
Technical Papers


The present paper introduces a scheme utilizing neurocomputing strategies for a decomposition approach to large scale optimization problems. In this scheme the modelling capabilities of a backpropagation neural network are employed to detect weak couplings in a system and to effectively decompose it into smaller, more tractable subsystems. When such partitioning of a design space is possible (decomposable systems), independent optimization in each subsystem is performed with a penalty term added to an objective function to eliminate constraint violations in all other subsystems. Dependencies among subsystems are represented in terms of global design variables, and since only partial information is needed, a neural network is used to map relations between global variables and all system constraints. A featuresensitive network (a variant of ahierarchical vector quantization technique, referred to as the HVQ network) is used for this purpose as it offers easy training, approximations of an arbitrary accuracy, and processing of incomplete input vectors. The approach is illustrated with applications to minimum weight sizing of truss structures with multiple design constraints.


Vector Quantization Constraint Violation Scale Optimization Truss Structure Quantization Technique 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. Bloebaum, C.L.; Hajela, P. 1991: Heuristic decomposition for non-hierarchic systems.Proc. 32-nd AIAA/ASME/ASCE/ASC SDM Meeting (held in Baltimore, MD)Google Scholar
  2. Bloebaum, C.L.; Hajela, P.; Sobieski, J. 1990: Non-hierarchic system decomposition in structural optimization.Proc. 3-rd Air Force/NASA Symp. on Recent Advances in Multidisciplinary Analysis and Optimization (held in San Francisco)Google Scholar
  3. Haftka, R.T. 1984: An improved computational approach for multilevel optimum design.J. Struct. Mech. 12, 245–261Google Scholar
  4. Hajela, P.; Berke, L. 1991: Neural network based decomposition in optimal structural synthesis.Computing Systems in Engineering 3, 1–9Google Scholar
  5. Hajela, P.; Berke, L. 1992: Neural networks in engineering analysis and design: an overview.Computing Systems in Engineering 3, 525–538Google Scholar
  6. Kirsch, U. 1983: Multilevel optimal design of reinforced concrete structure.Eng. Optimiz. 6, 207–212Google Scholar
  7. Kirsch, U. 1985: An improved multilevel structural synthesis.J. Struct. Mech. 13, 123–144Google Scholar
  8. Rumelhart, D.E.; McClelland, J.L. 1988:Parallel distributed processing. Cambridge: The MIT PressGoogle Scholar
  9. Sobieski-Sobieszczanski, J. 1988: On the sensitivity of complex, internally coupled systems.Proc. 29-th SDM Conf. (held in Williamsburg)Google Scholar
  10. Sobieski-Sobieszczanski, J.; Barthelemy, J.F.; Riley, K.M. 1982: Sensitivity of optimum solutions of problem parameters.AIAA J. 20, 1291–1299Google Scholar
  11. Sobieski-Sobieszczanski, J.; James, B.B.; Dovi, A.R. 1985: Structural optimization by multilevel decomposition.AIAA J. 23, 1775–1782Google Scholar
  12. Szewczyk, Z. 1993:Neurocomputing based approximate models in structural analysis and optimal design. Ph.D. Thesis, Rensselaer Polytechnic Institute, Troy, N.Y.Google Scholar
  13. Szewczyk, Z.; Hajela, P. 1992: Feature-sensitive neural networks in structural response estimation. In: Dagli, C.; Berke, L. (eds.)Intelligent systems in engineering through neural networks, pp. 85–89. ASME PressGoogle Scholar

Copyright information

© Springer-Verlag 1994

Authors and Affiliations

  • Z. P. Szewczyk
    • 1
  • P. Hajela
    • 1
  1. 1.Department of Mechanical Engineering, Aerospace Engineering & MechanicsRensselaer Polytechnic InstituteTroyUSA

Personalised recommendations