实现负载均衡的电力通信网络优化策略分析-analysis of power communication network optimization strategy for load balancing.docx

算法通过对选择、交叉、变异三种遗传算子的优化改进，有效提升了VECMP模型的求解效率。仿真实验表明，本文提出的IGA-VECMP求解算法能够快速得到网络的最佳权重配置方案，实现网络负载的有效均衡，降低最大链路利用率，提升网络资源的利用效率，在电力通信网络路由优化领域具有广泛的应用价值。关键词：电力通信网络；负载均衡；拓扑优化；流量分析和预测；ARIMA；等价多路径ABSTRACTTheprofoundchangesineconomyandenergyfieldathomeandabroadhavebroughtnewopportunitiesandchallengestothedevelopmentofelectricpowercommunicationnetwork.Inrecentyears,withthecontinuousconstructionofsmartgrid,moderngrid’sfeaturesofinformation,automationandinteractionarebecomingincreasinglyapparent,andthepowercommunicationnetwork,whichplaysastrongsupportingroleforthegrid,isalsoundergoingchanges.Withthedeeppenetrationofinformationcommunicationtechnologytoelectricpowerproductionandmanagement,thepowergridisbecomingincreasinglydependentoncommunicationnetworks.Thereliabilityandtransmissionperformanceofthecommunicationnetworkwillplayakeyroleforthesecurityandstabilityofthegrid.Withtheincreasingcomplexityofnetworktopologyandthegraduallyenrichedbusinessdemands,theloadimbalanceproblemsinelectricpowercommunicationnetworkbecomemoreserious.Certaincorenodesorlinkscarryalargeamountofbusinessdata,whichincreasesthevulnerabilityofthenetwork,andalsoreducestheutilizationefficiencyofnetworkresource.Traditionaloptimizationstrategiesdidn’tconsiderthetopologycharacteristicsandbusinessflowattributesinelectricpowercommunicationnetwork,sotheywerealwaysdifficulttousedirectly.Basedontopologicalcharacteristicsandflowattributesofelectricpowercommunicationnetwork,thisthesisstudiedtheproblemsoftopologyoptimization,flowanalysisandpredictionandroutingoptimizationinelectricpowercommunicationnetwork.Themajorcontributionofthisdissertationisspecificallystatedasfollows:Basedontherobustnessandvulnerabilityanalysis,oneoftopologyoptimizationalgorithmbasedonthematrixofnodebetweennesspluswasproposed,whichisusedfortopologyoptimization.Theoptimizationalgorithmfirstlygeneratedasetofoptimizedlinksbasedonmatrixofnodebetweennessplus,andthenachievedoptimizationgoalsbyaddingalimitednumberoflinks.Italsousedevasivemechanismstoavoidexcessiv