Stress‐enhanced clonal selection algorithm for structural topology optimization

Recently, numerous modified versions of immune algorithms (IAs) have been adopted in both theoretical and practical applications. However, few have been proposed for solving structural topology optimization problems. In addition, the design connectivity handling and one‐node connected hinge prevention, which are vital in the application of population‐based methods with binary representation for structural topology optimization, have not been applied to IAs in the literature. A stress‐enhanced clonal selection algorithm (SECSA) incorporating an IA with a dominance‐based constraint‐handling technique and a new stress‐enhanced hypermutation operator is proposed to rectify those deficiencies. To demonstrate the high viability of the presented method, comparisons between the presented SECSA and genetic algorithm‐based methods were made on minimum compliance and minimum weight benchmark structural topology design problems in two‐dimensional, three‐dimensional, and multiloading cases. In each case, SECSA was shown to be competitive in terms of convergence speed and solution quality. The main goal of this study is not only to further explore the capabilities of IAs, but also to show that an IA with appropriate enhancements can lead to the development of attractive computational tools for global search in structural topology optimization. Copyright © 2011 John Wiley & Sons, Ltd.