Parameters Determination for Optimum Design by Evolutionary Algorithm

The finding a maximum or a minimum function problem under some constraint conditions are called optimization problem. Almost every engineering design problem can be formulated as optimum problems. Solving the optimum problem requires the computation of the global maxima or minima of the objection function. Many heuristic intelligent algorithms had been developed and adapted to several optimal design problems. Heuristic algorithms have merit that they can search the global optimum with a higher probability than deterministic ones. Evolutionary algorithms are stochastic search methods that mimic the metaphor of natural biological evolution and/or the social behavior of species (Shyr, 2008). Obviously, in order to reach this goal makes the search process complicate and the selection of an optimum technique critical. It is a challenge for engineers to design efficient and cost-effect systems without compromising the integrity of the system. The conventional design process depends on the designer’s intuition, experience, and skill. There are several kinds of numerical optimization methods such as neural network, gradient-based search, genetic algorithm, etc (Rao, 1979). Neural network can simulate the relation between the input and output. But it needs samples for training the network first. Sometimes the gradient-based search is fast and efficient, but it is easy to get stuck in local extreme. Compared with them, the genetic algorithm has its special characteristics. No sample is needed for the implementation of genetic algorithm and the most important is that genetic algorithm can derive a global optimum by mutation and crossover technique so as to avoid being trapped in local optima. It is considered as a technique the most suitable for combinatorial optimization design. The concept of Genetic Algorithm (GA) was first established by Holland (Holland, 1975), based on the mechanism of nature selection and evolutionary genetics. The purpose of the genetic algorithm is to find a better function via some simulation artificial operation process, which includes evaluation, selection, crossover and mutation. Genetic algorithm is used in optimum design because of its efficient optimum capabilities. The genetic algorithm is an efficient tool in the field of engineering education (Bütün, 2005).