简介:Wepresentasimulationrunallocationschemeforimprovingefficiencyinsimulationexperimentsfordecisionmakingunderuncertainty.ThisschemeiscalledOptimalComputingBudgetAllocation(OCBA).OCBAadvancesthestate-of-the-artbyintelligentlyallocatingacomputingbudgettothecandidatealternativesunderevaluation.Thebasicideaistospendlesscomputationaleffortonsimulatingnon-criticalalternativestosavecomputationcost.Inparticular,OCBAisemployedtointelligentlyprovidethesmallestnumberofsimulationrunsforadesiredaccuracy.Inthispaper,wepresentanewandmoregeneralOCBAschemewhichcanconsidercasesthatusersareinterestednotonlythebestdesign,butalsoanyoneinagooddesignset.Inaddition,thispaperalsopresentstheapplicationofourOCBAtoadesignprobleminUSairtrafficmanagement.ThenationalairtrafficsysteminUSismodeledasalarge,complex,andstochasticnetwork.Thenumericalexamplesshowthatthecomputationtimecanbereducedby54%to88%withtheuseofOCBA.