Computational & Technology Resources
an online resource for computational,
engineering & technology publications
Civil-Comp Proceedings
ISSN 1759-3433
CCP: 84
PROCEEDINGS OF THE FIFTH INTERNATIONAL CONFERENCE ON ENGINEERING COMPUTATIONAL TECHNOLOGY
Edited by: B.H.V. Topping, G. Montero and R. Montenegro
Paper 45

An Artificial Immune System Algorithm for Structural Optimization

C.J. Shih and B.S. Chen

Department of Mechanical and Electro-Mechanical Engineering, Tamkang University, Tamsui, Taiwan R.O.C.

Full Bibliographic Reference for this paper
C.J. Shih, B.S. Chen, "An Artificial Immune System Algorithm for Structural Optimization", in B.H.V. Topping, G. Montero, R. Montenegro, (Editors), "Proceedings of the Fifth International Conference on Engineering Computational Technology", Civil-Comp Press, Stirlingshire, UK, Paper 45, 2006. doi:10.4203/ccp.84.45
Keywords: immune system, optimum design, artificial intelligence, structural optimization, biological computation.

Summary
A biological immune system [1] can precisely detect and eliminate toxic substances and pathogens so that the body can by itself remain secure and healthy. Once pathogens have entered the body, they are dealt with by the innate immune system and then by the adaptive immune system. This paper utilizes such a concept to develop an efficient and accurate zero-order optimization approach. The affinity maturation theory in the immune system performs as the basis to build the main frame in this development.

Learning, memory, diversity, hyper-mutation, differentiation and natural selection are the primary operators to ensure the proposed approach has the global as well as the local search potential. The technique of hyper-mutation is extensively modified to fulfil the primary search that shows effectively powerful. Numerically, one can replicate the adaptive immune system to proliferate diversity immune cells for bind pathogens. An algorithm of simulating the affinity maturation is as follows:

  1. Recognize the pathogens.
  2. Start to activate the memory cells of lymphocytes (called B-cells) and produce the required diversity with a pseudo-random process.
  3. B-cell cloning is subject to a somatic hyper-mutation.
  4. The new B-cell clones succeed to bind pathogenic, they will differentiate into plasma B-cells or memory B-cells. Plasma B-cells secrete a soluble form of their receptors, called antibodies.
  5. Repeat the cycle of affinity maturation (step 2 to 4) until high-affinity B-cells are obtained.
The proposed algorithm has been applied to general structural optimization problems yielding efficient, robust and true optimum results.

For example, a 72-bar space truss is required to support two load conditions given in reference [2] and is to be designed with allowable constraints and the allowable displacements. This problem was solved by the proposed approach in which the total population is 100, the maximum generations is 300. Other prescribed number of group population are , , , and . Using the proposed immune based approach to solve the problem ten times to find the optimum cross-sectional area of 16 members by minimizing the total structural weight. Figure 1 shows the iteration history of the ten searches for the optimum. It is obvious to see that the proposed method is robust and gives accurate results.

Figure 1: Iteration history of 72-bar truss design.

References
1
Hofmeyr, S.A., An Interpretative Introduction to the Immune System, Design principles for the Immune System and Other Distributed Autonomous Systems, Eds. Ohen, I. and Segel, L., Oxford University Press, 2000.
2
Haftka, R. T. and Gurdal, Z., Elements of structural optimization, Kluwer Academic publishers, pp. 245-248, 1992.

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