Dynamic Multiobjective Clonal Selection Algorithm for Engineering Design

We propose a Multiobjective Clonal Selection Algorithm (MCSA) with dynamic variation of its main parameters for the solution of engineering design problems. The MCSA performs a cloning process using different probability distributions, in which the mutation strengths are guided based on a logarithmic rule and on information implicitly created by a simple differential evolution technique. This feature results in a self-adapting search in the algorithm. The efficiency of the MCSA is studied comparing its performance with the Nondominated Sorting Genetic Algorithm II (NSGA-II) in analytical test problems and also in the design of a microwave heating device. The MCSA has outperformed the NSGA-II in all problems investigated.