An intelligent modeling method based on genetic algorithm for partner selection in virtual organizations

The goal of a Virtual Organization is to find the most appropriate partners in terms of expertise, cost wise, quick response, and environment. In this study we propose a model and a solution approach to a partner selection problem considering three main evaluation criteria: cost, time and risk. This multiobjective problem is solved by an improved genetic algorithm (GA) that includes meiosis specific characteristics and step-size adaptation for the mutation operator. The algorithm performs strong exploration initially and exploitation in later generations. It has a high global search ability and a fast convergence rate and also avoids premature convergence. On the basis of the numerical investigations, the incorporation of the proposed enhancements has been successfully proved.