简介：ForasimpleundirectedgraphG,denotebyA(G)the(0,1)-adjacencymatrixofG.LetthematrixS(G)=J-I-2A(G)beitsSeidelmatrix,andletSG(λ)=det(λI-S(G))beitsSeidelcharacteristicpolynomial,whereIisanidentitymatrixandJisasquarematrixallofwhoseentriesareequalto1.IfalleigenvaluesofSG(λ)areintegral,thenthegraphGiscalledS-integral.Inthispaper,ourmaingoalistoinvestigatetheeigenvaluesofSG(λ)forthecompletemultipartitegraphsG=Kn1,n2,...,nt.AnecessaryandsufficientconditionforthecompletetripartitegraphsKm,n,tandthecompletemultipartitegraphsKm,...,ms,n,...,nttobeS-integralisgiven,respectively.

简介：LetthelinearsystemAx=bwherethecoefficientmatrixA=(aij)∈Rm,nisanL-ma-trix(thatis,aij>0（?）iandaij≤0（?）i≠j),A=I-L-U,Iistheidentitymatrix,-Land-Uare,respectively,strictlylowerandstrictlyuppertriangularpartsofA.In[1]theauthorsconsideredtwopreconditionedlinearsystems?x=（?）and?x=（?）

简介：AnasynchronousparallelmultisplittingnonlinearGauss-SeideliterativemethodisestablishedfortheparticularlystructuredsystemofnonlinearequationsAφ(x)+Bφ(x)=bwithA,B∈(R^n)φ,φtR^n→R^nbeingdiagonalmappingsandb∈R^n,andtheglobalconvergenceofitisproved.

简介：在计算线性方程组时,我们有时会遇到其系数矩阵A是严格次对角占优及次正定的次对称的情形,对于这样的方程组,我们不能直接应用Jacobi、Gauss—Seidel及超松驰迭代法进行求解.在文[2]中,利用了JA是严格对角占优(占A是严格次对角占优)及JA是正定对称(当A是次正定的次对称)的性质,对方程AX=b作用J得方程JAX=Jb,对此方程我们再使用以上的方法进行求解,然而JA是对A作一条列的行变换得到的,当n是偶数时,至少要作n/2次行对换,在计算机上将A经行变换变成JA至少要进行3/2n~2次赋值,当n是奇数时,至少要进行3/2n(n-1)次赋值.并且在这个过程中还要增加n个单元的内

简介：ThepreconditionedGauss-Seideltypeiterativemethodforsolvinglinearsystems,withtheproperchoiceofthepreconditioner,ispresented.ConvergenceofthepreconditionedmethodappliedtoZ-matricesisdiscussed.Alsotheoptimalparameterispresented.NumericalresultsshowthattheproperchoiceofthepreconditionercanleadtoeffectivebythepreconditionedGauss-Seideltypeiterativemethodsforsolvinglinearsystems.

简介：Inthispaper,weproposeaparallelGauss-Seideltypeiterativemethodforsolvingthelarge-scalesystemofnonlinearalgebraicequationsAφ(x)+Bψ(x)=b,whichisanasynchronousvariantofthesynchronousparallelnonlinearGauus-SeideltypemethodgivenbyR.E.White.Withalmostthesamebutsomewhatmorerelaxedconstrainteonthemultiplesplittings,weprovetheconvergenceandestimatetheconvergencerateofthenewmethod.

简介：TheJacobiandGauss-Seidelalgorithmsareamongthestationaryiterativemethodsforsolvinglinearsystemofequations.Theyarenowmostlyusedasprecondition-ersforthepopulariterativesolvers.Inthispaperageneralizationofthesemethodsareproposedandtheirconvergencepropertiesarestudied.Somenumericalexperimentsaregiventoshowtheefficiencyofthenewmethods.

简介：有限元模型修正是一类特殊的二次反特征值问题．我们将有限元模型修正看成二次规划问题来解决，并采用非线性Gauss－Seidel方法来求解其相应的Lagrange对偶函数．最后，给山的数值文验说明方法的有效性．

简介：文章利用求解线性方程组的Gauss-Seidel迭代法推导出其＂反方法＂,正反两种方法相匹配生成预报-校正系统,给出了它们收敛的条件,并运用这三种不同的公式求解实例,根据其结果,说明这些公式的优缺点。