The Ninth International Conference on the Application of Artificial Intelligence to Civil, Structural and Environmental Engineering: Soft Computing Tools for Structural Optimisation, Sighghiewi, Malta, 18 - 21 Eylül 2007, ss.1-11
In this paper, the effects of different ratios of crossover and mutation, the effects of different population sizes, and binary and value encodings are studied to see how the results obtained by GA is affected by these parameters for large truss structures. The results of value and binary encodings are compared with each other to show the advantages and disadvantages of them. The results show that the algorithm coded by using value encoding can find less weight and requires less run-time and computer memory than the algorithm coded by using binary encoding, and that when the population size is taken large, the weight of structure can be found to be near optimum.
Keywords: crossover, mutation, population size, large truss structures, value encoding, binary encoding.