An effective mutation operator to deal with multi-objective constrained problems: SPM

In this paper, a novel mutation operator for Evolutionary Multi-objective Algorithms (MOEAs), named as Subspace Polynomial Mutation (SPM) is presented. This specialized mutation operator is particularly designed to deal with constrained continuos problems. As a variation operator, SPM ensures the production of suitable candidate solutions which has not only the chance to improve their survival rate, but that fulfills feasibility also-saving in this way a considerable amount of function evaluations when avoiding unnecessary trials. This feature coupled with the ability of SPM for performing movements along the constrained Pareto set improves the efficiency of the mutation process in a MOEA.