分块矩阵drazin逆和电阻距离算法及符号模式分析-algorithm of block matrix dra zin inverse sum resistance distance and symbolic pattern analysis.docxVIP

摘要矩阵广义逆理论在数学和工程等领域有重要的理论和应用价值，用广义逆研究拉普拉斯矩阵理论和应用问题是国际上的前沿课题.拉普拉斯矩阵理论在图的连通性、图的谱理论中都有重要应用，是国际上一个热门的研究领域.本文介绍了本文的研究意义和相关基础知识.在不附加任何条件的情况下给出了分块加权拉普拉斯矩阵的群逆表达式，给出网络电阻距的群逆算法.这为电阻距计算、网络鲁棒性分析提供了新的有效算法.主要结果如下1.给出了加权图的2??2块拉普拉斯矩阵的子块正交和广义Schur补性质；2.给出加权图的2??2块拉普拉斯矩阵的群逆表达式；3.给出了加权图任意两点之间的电阻距使用拉普拉斯矩阵群逆的计算公式和算法；4.给出了连通二分图使用无符号拉普拉斯矩阵的电阻距计算公式；5.给出加权图的2??2块无符号拉普拉斯矩阵的群逆表达式.在一定条件下，对一类有向图的拉普拉斯矩阵给出了新的Drazin表达式，并且对其进行符号分析.主要结果如下1.在条件rank?BCB????rank?B??，B??AB??0下，给出了M=??AB??(A?Cn?n)的??C0???Drazin表达式；??2.给出了M=??AB??的Drazin表达式；??B*0?*3.在条件(E*)??A??0下，给出了M????AE???Cn?n的Drazin表达式；??E0???4.给出一类分块矩阵有符号Drazin逆的必要条件.关键词：拉普拉斯矩阵；电阻距离；矩阵群逆；Drazin逆；符号Drazin逆AbstractTheGeneralizedinversetheoryalsohasmuchimportanttheoreticalsignificanceandapplicationvalue.ItisaninternationaladvancedtopicthatusingGeneralizedinversetheorytoresearchthetheoryandtheapplicationsofLaplacianmatrix.TheLaplacianmatrixplaysanimportantroleonthetheoryofconnectiitvityandspectrumofgraph.Itisahotresearchfield.Inthispaper,someresearchsignificanceandbasicknowledgeofourtopicareintroduced.Next,wegiveblockrepresentationsforgroupinverseoftheLaplacianmatrix,whichtherearenotattachedanyconditions.Andanalgorithmforresistancedistanceisalsogiven.Thusweprovideaneweffectivealgorithmtocalculationresistancedistanceandanalysisrobustnessofthenetwork.ThemainresultsarelistedbelowSomepropertiesthattheorthogonallyofsubblockandtheSchurcomplementofa2??2blockmatrixofaweighteddigrapharegiven;Therepresentationforthegroupinverseofthe2??2weightedgrapharegiven;blockLaplacianmatrixoftheLet?ijbetheresistancedistancebetweenanytwoverticesi,jinaweightedgraph.Wegiveaformulaandalgorithmtocalculationmatrix;?ijusingthegroupinverseofLaplacianItisshownthatacalculationformulaofresistancedistanceusingthesignlessLaplacianmatrixofconnectbipartitegraph;LetMbea2??2blockmatrixforthesignlessLaplacianmatrixofaweightedgraph.TherepresentationsofMisgiven.Finally,newrepresentationsforDrazininverseoftheLaplacianmatricesareshownundersomeconditions,andweanalysisthesignoftherepresentations.Th