Basically, CA represents simple mathematical idealizations of physical systems in which space and time are discrete and physical quantities are taken from a finite set of discrete values. Models based on CA provide an alternative and more general approach to physical modelling rather than an approximation. The CA shows a complex behaviour analogous to that associated with complex differential equations, but in this case complexity emerges from the interaction of simple entities following simple rules.

CGAs are a subclass of GAs in which the population is structured in a specified grid and the genetic operations may only take place in a small neighbourhood of each individual. It has been demonstrated that the CGAs are capable to find better solutions than the traditional GAs. In the field of structural optimisation some of researchers [1,2] have combined the concepts of CA and the GA to create cellular genetic algorithms (CGA). In the proposed CGA, in this paper, a small dimensioned grid is selected and the artificial evolution is evolved by a novel crossover. The proposed cellular crossover operation acts on the design variables and combines the information available at the central site and its immediate neighbours. In each iteration, the cellular crossover operation produces a new design at each site according to the fitness index of neighbouring cells of each site. Also, the real-values of design variables, instead of their binary codes, are considered. In the proposed CGA, the evolutionary rule of the automaton includes the cellular crossover operation and after this the mutation operation is also applied to the sites. The proposed CGA is a multi-staged evolutionary algorithm described in detail in this paper.

In order to investigate the computational performance of the proposed CGA, two numerical examples including a 10-bar planar truss structure and a 72-bar space truss structure subject to static displacement and stress constraints, are optimized and the results are compared with those presented in other papers. The numerical results demonstrate the computational efficiency of the proposed CGA algorithm for structural optimisation.

1

O.E. Canyurt, P. Hajela, "A cellular framework for structural analysis and optimization", Computer Methods in Applied Mechanics and Engineering, 194, 3516-3534, 2005. doi:10.1016/j.cma.2005.01.014

2

S. Rajasekaran, "Optimization of large scale three dimensional reticulated structures using cellular genetics and neural networks", International Journal of Space Structures, 16, 315-324, 2001. doi:10.1260/026635101760832244

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 £145 +P&P)