An evolutionarymethodforcomplex-processoptimization

In thispaperwepresentanewevolutionarymethodforcomplex-processoptimization.Itispartiallybased on theprinciplesofthescattersearchmethodology,butitmakesuseofinnovativestrategiestobemore effective inthecontextofcomplex-processoptimizationusingasmallnumberoftuningparameters.In particular, weintroduceanewcombinationmethodbasedonpathrelinking,whichconsidersabroader area aroundthepopulationmembersthanpreviouscombinationmethods.Wealsouseapopulation- update methodwhichimprovesthebalancebetweenintensificationanddiversification.Newstrategies to intensifythesearchandtoescapefromsuboptimalsolutionsarealsopresented.Theapplicationofthe proposed evolutionaryalgorithmtodifferentsetsofbothstate-of-the-artcontinuousglobaloptimization and complex-processoptimizationproblemsrevealsthatitisrobustandefficientforthetypeofproblems intended tosolve,outperformingtheresultsobtainedwithothermethodsfoundintheliterature.