Fast clonal algorithm

Abstract

The aim of this paper is to design an efficient and fast clonal algorithm for solving various numerical and combinatorial real-world optimization problems effectively and speedily, irrespective of its complexity. The idea is to accurately read the inherent drawbacks of existing immune algorithms (IAs) and propose new techniques to resolve them. The basic features of IAs dealt in this paper are: hypermutation mechanism, clonal expansion, immune memory and several other features related to initialization and selection of candidate solution present in a population set. Dealing with the above-mentioned features we have proposed a fast clonal algorithm (FCA) incorporating a parallel mutation operator comprising of Gaussian and Cauchy mutation strategy. In addition, a new concept has been proposed for initialization, selection and clonal expansion process. The concept of existing immune memory has also been modified by using the elitist mechanism. Finally, to test the efficacy of proposed algorithm in terms of search quality, computational cost, robustness and efficiency, quantitative analyses have been performed in this paper. In addition, empirical analyses have been executed to prove the superiority of proposed strategies. To demonstrate the applicability of proposed algorithm over real-world problems, Machine-loading problem of flexible manufacturing system (FMS) is worked out and matched with the results present in literature.