A genetic algorithm and its application in engineering structural optimization are developed in this paper. Developed genetic operators are adopted based on former working. Regarding to the restricted optimization problem, two methods are developed to deal with constraint, including scale adjustment and quasi-exactness penalization function. According to the characteristic of structural optimum problem, an energy criterion is used to evaluation function. Based on the parallel characteristic of genetic algorithm, a parallel genetic algorithm with a larger grain is presented. Using this method, speedup and efficiency of parallel computation increased.Based on these improvements, numerical examples are given to show the reliability and efficiency of genetic algorithm in structural optimization. |