基于增强型差分进化算法求解广义Nash均衡问题

  • 打印
  • 收藏
收藏成功


打开文本图片集

Solving generalized Nash equilibrium problem based on enhanced differential evolutionary algorithm

Wang Kaia,b,Jia Wenshenga,bt (a.Collgeofts&istis,rocalybotofeecisioaing&tolSeuzUesit 550025,China)

Abstract:Addressng theproblems thatclassicalmathematical methods forsolvingthegeneralizedNashequilibriumproblem face,suchasrelianceoninitialpointseletion,highdiferentiabilityrequirements,informationlossduringproblemtrasfor mation,andinsuffcientperformanceofmeta-heuristicalgorithms,thispaperproposedanenhanceddiferentialevolutionalg rithmtodirectlysolvethegeneralizedNashequilibriumproblemusingtheNikaido-Isoda function.Firstly,toimprovethediversityandconvergence speedofthediferentialevolutionalgorithm,itintroducedtheideasof tent chaotic mapping,adaptive coeficients,andtheslme mouldalgorithmtodesignanimprovedversionofthediferentialevolutionalgorithm.Italsoprovied atheoretical proofofthealgorithm’sconvergence.Secondly,itdefinedadominancestrategyandarelativefitnessfunction using the Nikaido-Isoda function toenhancepopulationvariationand selectioninthediferentialevolutionalgorithm.Finaly, theresultsofarithmeticcases indiferentdimensionsdemonstratethatthealgorithmsuccessullyresolves thegeneralized Nash equilibrium problem.Therefore,theproposedmethodforsolvingthegeneralizedNashequilibriumproblemdoesnotrelyoninitialpointslectinorfereiabilityanditvidsiformationlossduringprblmtansforatineringcrtaindvntges and effectiveness.

Keywords:generalized Nashequilibrium;Nikaido-Isoda function;bimodal variants;dominance strategies;meta-heuristic algorithms

0 引言

Nash均衡是非合作博弈中非常重要的概念,它在经济学、管理学、人工智能等[1\~3]领域都有广泛的应用。(剩余16041字)

目录
monitor