Component selection using hybrid optimization algorithms

The objective of this paper is to outline a computer-aided component selection methodology that will be available to the systems designer in the conceptual design phase to help him/her realize optimum design solutions. In this study, a modified technique for order preference by similarity to the ideal solution (TOPSIS) has been used first to select a set of components (for a given system layout) among approved lists of manufacturers' catalogs so that all design specifications are satisfied. After generating the initial list of the desired component, some components which are clearly the best in the catalog list, are separated out. A genetic algorithm is then used to successfully select an optimal solution. The proposed hybrid optimization algorithms have been implemented in a tube selection problem and it has been presented in the paper as a case study.