Research article

Global error bounds for the extended vertical LCP of SDD matrices

  • Received: 22 May 2024 Revised: 19 July 2024 Accepted: 26 July 2024 Published: 16 August 2024
  • MSC : 65H14

  • An error bound for the solution of the SDD matrix extended vertical linear complementarity problem is given when the SDD matrix satisfies the row W-property. It is shown by the illustrative example that the new bound is better than those in [1] in some cases.

    Citation: Mengting Gan. Global error bounds for the extended vertical LCP of SDD matrices[J]. AIMS Mathematics, 2024, 9(9): 24326-24335. doi: 10.3934/math.20241183

    Related Papers:

    [1] Yuanjie Geng, Deshu Sun . Error bounds for linear complementarity problems of strong $ SDD_{1} $ matrices and strong $ SDD_{1} $-$ B $ matrices. AIMS Mathematics, 2023, 8(11): 27052-27064. doi: 10.3934/math.20231384
    [2] Lanlan Liu, Pan Han, Feng Wang . New error bound for linear complementarity problem of $ S $-$ SDDS $-$ B $ matrices. AIMS Mathematics, 2022, 7(2): 3239-3249. doi: 10.3934/math.2022179
    [3] Yingxia Zhao, Lanlan Liu, Feng Wang . Error bounds for linear complementarity problems of $ SD{{D}_{1}} $ matrices and $ SD{{D}_{1}} $-$ B $ matrices. AIMS Mathematics, 2022, 7(7): 11862-11878. doi: 10.3934/math.2022662
    [4] Lanlan Liu, Yuxue Zhu, Feng Wang, Yuanjie Geng . Infinity norm bounds for the inverse of $ SDD_1^{+} $ matrices with applications. AIMS Mathematics, 2024, 9(8): 21294-21320. doi: 10.3934/math.20241034
    [5] Xiaoyong Chen, Yating Li, Liang Liu, Yaqiang Wang . Infinity norm upper bounds for the inverse of $ SDD_1 $ matrices. AIMS Mathematics, 2022, 7(5): 8847-8860. doi: 10.3934/math.2022493
    [6] Xiaodong Wang, Feng Wang . Infinity norm upper bounds for the inverse of $ {SDD_k} $ matrices. AIMS Mathematics, 2023, 8(10): 24999-25016. doi: 10.3934/math.20231276
    [7] Hongmin Mo, Yingxue Dong . A new error bound for linear complementarity problems involving $ B- $matrices. AIMS Mathematics, 2023, 8(10): 23889-23899. doi: 10.3934/math.20231218
    [8] Xinnian Song, Lei Gao . CKV-type $ B $-matrices and error bounds for linear complementarity problems. AIMS Mathematics, 2021, 6(10): 10846-10860. doi: 10.3934/math.2021630
    [9] Dizhen Ao, Yan Liu, Feng Wang, Lanlan Liu . Schur complement-based infinity norm bounds for the inverse of $ S $-Sparse Ostrowski Brauer matrices. AIMS Mathematics, 2023, 8(11): 25815-25844. doi: 10.3934/math.20231317
    [10] Deshu Sun . Note on error bounds for linear complementarity problems involving $ B^S $-matrices. AIMS Mathematics, 2022, 7(2): 1896-1906. doi: 10.3934/math.2022109
  • An error bound for the solution of the SDD matrix extended vertical linear complementarity problem is given when the SDD matrix satisfies the row W-property. It is shown by the illustrative example that the new bound is better than those in [1] in some cases.



    The extended vertical linear complementarity problem is to find a vector xRn such that

    r(x):=min(A0x+q0,A1x+q1,,Akx+qk)=0,

    or prove that there is no such vector, where the min operater works componentwise for both vectors and matrices. Here, it is abbreviated by EVLCP(A, q), where

    A=(A0,A1,,Ak),AlRn×n,l=0,1,,k,

    is a block matrix and

    q=(q0,q1,,qk),qlRn,l=0,1,,k,

    is a block vector.

    When k=1,A0=I,q0=0, the EVLCP(A, q) reduces to linear Complementarity problems (denoted by LCP (A1,q1), and when A0=I,q0=0, the EVLCP(A, q) comes back to vertical linear complementarity problems (can be found in [2,3]).The results proposed by Gohda and Sznajder (can be found in [4,5,6]) for the extended vertical linear complementarity problem.

    Some experts and scholars have extended the theories of existence, uniqueness, and error bound of the linear complementarity problem to the extended vertical linear complementarity problem. For example, for any vector q, in the LCP (A, q) has unique solutions if and only if matrix A is the P-matrix (can be found in [7,8]). Gowda and Sznajder extended this to the extended vertical linear complementarity to be replaced by problem [2] in 1994, which became the theoretical basis for the later study of the extended vertical linear complementarity problem.

    In [9,10,11,12], the algorithms for solving the extended vertical linear complementarity problem are proposed, but the solutions obtained by using these algorithms often have errors; thus, the error estimation of the extended vertical linear complementarity problem is worth studying.

    Next, let us review some relevant symbols, concepts, theorems, and lemmas.

    When the B=(bij)Rn×n matrix satisfies bij0, for any ij, B is called Z-matrices; if the principal and sub equations of B are all positive, then B is a P-matrix; if B is a Z-matrix and B1>0, then B is an M-matrix; if ˜B=(˜bij) and if the comparison matrix of B is an M-matrix, when i=j,˜bij=|bij|, when ij,˜bij=|bij|, then B is an H-matrix (can be found in [12]).

    Definition 1.1. [13]   B matrix is B=(bij)Rn×n, if for any i,jN,|bii|>ri(B), is called a strictly diagonally dominant (SDD) matrix.

    Definition 1.2. [1]   If for any iN,

    (Aj)i.=(Aji)i.{(A0)i.,(A1)i.,,(Al)i.}={(A0)i.,(A1)i.),,(Al)i.},

    where (Aj)i. represents the i-th row of matrix (Aj), then the block matrix A=(A0,A1,,Ak) is called row rearrangement. Similarly, block vectors q and q also satisfy the above relationship, respectively and here use (M) and (q) to represent the set of rearranged rows of A and q.

    Theorem 1.1. [13]   For any block vector q=(q0,q1,,qk), the EVLCP(A, q) has a unique solution if and only if the block matrix A=(A0,A1,,Ak) has the row W-property, i.e.,

    min(A0x,A1x,,Akx)0max(A0x,,A1x,,Aix)x=0,

    where, the max and min operators work componentwise for both matrices and vectors, 0 represents a zero vector.

    In 2009, Zhang et al. (can be found in [1]) applied the property of row rearrangement in block matrices to provid necessary and sufficient conditions for block matrices to have the row properties. The specific expression is as follows:

    Lemma 1.1. [1]   The block matrix A=(A0,A1,,Ak) has the row W-property if and only if (ID)Aj+DAl is nonsingular for any two blocks Aj,Al of A(A) and for any D=diag(di),di[0,1],iN.

    Furthermore, a global error estimation formula for the EVLCP(A, q) is also provided, as follows:

    Let x be the solution of the EVLCP(A, q). If the block matrix A=(A0,A1,,Ak) has the row W-property, then for any xRn,

    xxα(A)r(x), (1.1)

    where

    α(A):=maxM(M)maxj<l{0,1,,k}maxd[0,1]n[(ID)Aj+DAl]1.

    Apparently, due to the difficulty to compute [(ID)Aj+DAl]1 exactly, it is upper bound (1.1) for α(A) and cannot be computed easily. Thus, some computable upper bounds for α(A) are given under various matrix norms by considering the structured property for matrices Aj,j=0,1,,k.

    For example, Zhang et al. provided an upper bound on α(A) when all Aj are SDD matrices in [1], and also provided an upper bound on another α(A) for a special matrix A under certain conditions.

    Theorem 1.2. [1]   Suppose that matrices A0,A1,,Ak has the positive diagonals, with the spectral radius

    ρ(max(Λ10|Q0|,Λ11|Q1|,,Λ1k|Qk|))<1,

    then A=(A0,A1,,Ak) has the row W-property and

    α(A)[Imaxi=0,1,,k(Λ1i|Qi|)]1maxi=0,1,,k(Λ1i),

    where Λi is the diagonal part of Ai, Qi=ΛiAi, for i=0,1,,k.

    In addition, Zhang et al. presented a computable upper bound for α(A) under the infinity norm in another rspecial class of block matrices.

    Theorem 1.3. [1]   Suppose that A0,A1,,Ak are SDD matrices and A=(A0,A1,,Ak) has the row W-property, and for each iN,(Aj)ii(Al)ii>0, with any j<l{0,1,,k}, then

    α(A)1miniN{(min(A0e,A1e,,Ake))i},

    where ˜Ai is the comparison matrix of Ai, i.e., (˜Ai)ττ=|(Ai)ττ|,(˜Ai)τj=|(Ai)τj| for τj,(Ai)τj is the element in the τ-th row and j-th column of Ai, and ~(Ai)τj is the element in the τ-th row and j-th column of ~(A)i.

    However, the error estimation formula provided is applicable only to a certain type of special matrix, and it is not easy to verify. Therefore, it is necessary to explore the error estimation formula for solutions of EVLCP(A, q) for other special matrix classes.

    In 2021, Wang et al. provided error estimation formulas for the EVLCP(A, q) of BRπ-matrices and B-matrices in reference [14]. In 2023, Zhao et al. provided error estimation formulas for the EVLCP(A, q) of SSDDSB matrix and SSDDS matrix in reference [15].

    In 2013, García Esnaola and Peña first proposed a special class of H-matrices: -SDD matrices (see [16] for details) and provided error estimates for their linear complementarity problems with parameters. Its definition is as follows:

    Definition 1.3. [16]   Let matrix B=(bij)Cn×n if there is a non empty subset S such that the following two conditions hold:

    (I) |bii|>rSi(B),iS;

    (II) (|bii|rsi(B))(|bjj|rsj(B))>rˉsi(B)rsj(B),iS,jˉS.

    then B is called the SDD matrix.

    Theorem 1.4. [16]   If B is an SDD matrix and S is a nonempty subset of N, then

    B1maxiS,jˉSmax{ρSij(B),ρˉSji(B)},

    where

    ρSij(B)=|bii|rSi(B)+rSj(B)(|bii|rSi(B))(|bjj|rˉSj(B))rˉSi(B)rSj(B),
    ρˉSji(B)=|bij|rˉSj(B)+rˉSi(B)(|bii|rSi(B))(|bjj|rˉSj(B))rˉSi(B)rSj(B).

    Proposition 2.1. Let B=(bij)Cn×n and M=(mij)Cn×n be SDD matrices with positive main diagonal elements, and all have the same set SN, If iS,jˉS, satisfies bijmij>0 (or bjimji>0) and

    (|bii|rSi(B))(|mjj|rˉSj(M))>rˉSi(B)rSj(M),
    (|mii|rSi(M))(|bjj|rˉSj(B))>rˉSi(M)rSj(B),

    then (ID)B+DM is a SDD matrices, where D=diag(di),di[0,1],iN.

    Proof. Since both B and M are SDD matrice, so for any iS,jˉS, the following hold:

    |bii|>rSi(B),(|bii|rSi(B))(|bjj|rSj(B))>rˉSi(B)rSj(B),
    |mii|>rSi(M),(|mii|rSi(M))(|mjj|rSj(M))>rˉSi(M)rSj(M).

    Note that di[0,1], hence 1di0 and di0, they are not equal to 0 at the same time. Let (ID)B+D=C=(cij). Thus for any iS,jˉS, we have

    |cii|rSi(C)=(1di)|bii|+di|mii|jS/{i}(1di)|bij|jS/{i}di|mii|=(1di)(|bii|rSi(B))+di(|mii|rSi(M))>0,

    i.e.,

    |cii|>rSi(C),

    and for any iS,jˉS, we have

    (|cii|rSi(C))(|cjj|rˉSj(C))=[(1di)(|bii|rSi(B))+di(|mii|rSi(M))]×[(1dj)(|bjj|rˉSj(B))+dj(|mjj|rˉSj(M))]=(1di)(1dj)(|bii|rSi(B))(|bjj|rˉSj(B))+(1di)dj(|bii|rSi(B))(|mjjrˉSj(M))+di(1dj)(|mii|rSi(M))(|bjj|rˉSj(B))+didj(|mii|rSi(M))(|mjj|rˉSj(M))>(1di)(1dj)rˉSi(B)rSj(B)+(1di)djrˉSi(B)rSj(M)+di(1dj)rˉSi(M)rSj(B)+didjrˉSi(M)rSj(M)=[(1di)rˉSi(B)+dirˉSi(M)][(1dj)rSj(B)+djrSj(M)]=rˉSi(C)rSj(C),

    form Definition 1.3 the conclusion follows.

    According to Proposition 2.1, Lemma 1.1 and the fact that a SDD matrix is nonsingular(can be found in [17] for), block matrix composed of SDD matrix has the row W-property.

    Proposition 2.2. A0,A1,,Ak are SSD matrices, and each Ali(li=0,1,,k) satisfies the condition of Proposition 2.1, then each Al in A(A) is a SDD matrix, A=(A0,A1,,Ak) has the row W-property.

    Proof. From Definition 1.2, for the i-th row (Al)i.(iN) of Al, there exist li{0,1,,k}, such that (Al)i.=(Aji)i.. Since Ali is a SDD matrix, for any iS,jˉS, we have

    |(Ali)ii|>rsi(Ali),(|(Ali)ii|rsi(Ali))(|(Ali)jj|rˉsj(Ali))>rˉSi(Ali)rSj(Ali),

    i.e.,

    |(Al)ii|>rsi(Al),(|(Al)ii|rsi(Al))(|(Al)jj|rˉsj(Al))>rˉSi(Al)rSj(Al),

    from Definition 1.3, Al is a SDD matrix.

    Let Aj,Al be any two blocks in AR(A), then Aj,Al are all SDD matrices. According to Proposition 2.1, (ID)Aj+DAl is a SDD matrix for any D=diag(di),di[0,1],iN, and thus (ID)Aj+DAl is nonsingular. From Lemma 1.1, block matrix A=(A0,A1,,Ak) has the row W-property.

    Next, we provide an upper bounds for α(A) with each Al,l=0,1,2,,k being a SDD matrix.

    Theorem 2.1. Let A=(A0,A1,,Ak), if each Ali(li=0,1,,k) is a SDD matrix and satisfies the condition of Proposition 2.1, then

    maxM(M)maxj<l{0,1,,k}maxd[0,1]n[(ID)Aj+DAl]1maxmaxiS,τˉS{4maxp=0,1,,k{(αpmax)ρSmax}maxp=0,1,,k{(βpmin)ρSminAp}minp=0,1,,k{(βpmin)ρSminAp}2,4maxp=0,1,,k{(αpmax)ρˉSmax}maxp=0,1,,k{(βpmin)ρSminAp}minp=0,1,,k{(βpmin)ρSminAp}2},

    where

    (αpmax)ρSmin=maxiS,τˉS{(αpiτ)ρSiτ},(αpiτ)ρSτi=|(Ap)ττ|rˉSτ(Ap)+rˉSi(Ap),(βpmin)ρSminAp=miniS,τˉS(βpiτ)ρSiτAp.

    Proof. For any two blocks Aj,Al in AR(A), and any D=diag(di),di[0,1],iN. According to Propositions 2.1 and 2.2, it can be inferred that Aj,Al and AD all are SDD matrices with positive main diagonal elements. According to Theorem 1.4,

    A1D=((ID)Aj+DAl)1maxiSmaxτˉS{ρSiτ(AD),ρˉSτi(AD)},

    where any iS,τˉS,

    ρSiτ(AD)=|aii|rSi(AD)+rSτ(AD)(|aii|rSi(AD))(|aττ|rˉSτ(AD))rˉSi(AD)rSτ(AD),
    ρˉSτi(AD)=|aττ|rˉSτ(AD)+rˉSτ(AD)(|aii|rSi(AD))(|aττ|rˉSτ(AD))rˉSi(AD)rSτ(AD).

    We have

    |aii|rSi(AD)+rSτ(AD)=|(1di)(Aj)ii+di(Al)ii|(1di)rSi(Aj)dirSi(Al)+(1dτ)rSτ(Aj)+dirSi(Al)<(|(Aj)ii|rSi(Aj)+rSτ(Aj))+(|(Al)ii|rSi(Al)+rSτ(Al))=(αjiτ)ρSiτ+(αliτ)ρSiτ2maxt=j,l(αtmax)ρSmax,

    where t=i,l, and

    (αtmax)ρSmax=max{(αtiτ)ρSiτ},(αtiτ)ρSiτ=|(At)ii|rSi(At)+rSτ(At).

    Similarly, we have

    |aττ|rˉSτ(AD)+rˉSi(AD)=|(1dτ)(Aj)ττ+dτ(Al)ii|(1dτ)rˉSτ(Aj)dτrˉSτ(Al)+(1di)rˉSi(Aj)+dirˉSi(Al)<(|(Aj)ττ|rˉSτ(Aj)+rSτ(Aj))+(|(Al)ττ|rˉSτ(Al)+rˉSi(Al))=(αjτi)ρˉSiτ+(αliτ)ρˉSτi2maxt=j,l(αtmax)ρˉSmax,

    where t=j,l, and

    (αtmax)ρˉSmax=max{(αtτi)ρˉSiτ},(αtτi)ρˉSτi=|(At)ττ|rˉSi(At)+rˉSi(At).

    Therefore, it can be concluded that

    (|aii|rSi(AD))(|aττ|rˉSτ(AD))rˉSi(AD)rSτ(AD)=(|(1di)(Aj)ii+di(Al)ii|(1di)rSi(Aj)dirSi(Al))×(|(1dτ)(Aj)ττ+dτ(Al)ττ|(1dτ)rτˉS(Aj)dττrˉSτ(Al))((1di)rˉSi(Aj)+dirˉSi(Al))×((1dτ)rSτ(Aj)+dτrSτ(Al))=(1di)(1dτ)(|(Aj)ii|rSi(Aj))(|(Aj)ττ|rˉSτ(Aj))+(1di)dτ(|(Aj)ii|rSi(Aj))(|(Al)ττ|rˉSτ(Al))+di(1dτ)(|(Al)ii|rSi(Aj))(|(Al)ττ|rˉSτ(Aj))+didτ(|(Al)ii|rSi(Al))(|(Al)ττ|rˉSτ(Al))(1di)(1dτ)rˉSi(Aj)rSτ(Aj)(1di)dτrˉSi(Aj)rSτ(Al)di(1dτ)rˉSi(Al)rSτ(Aj)didτrˉSi(Al)rSτ(Al)
    >(1di)(1dτ)(|(Aj)ii|rSi(Aj))(|(Aj)ττ|rˉSτ(Aj))+(1di)dτ(|(Aj)ii|rSi(Aj))(|(Al)ττ|rˉSτ(Al))+didτ(|(Al)ii|rSi(Al))(|(Al)ττ|rˉSτ(Al))(1di)(1dτ)rˉSi(Aj)rSτ(Aj)(1di)dτrˉSi(Aj)rSτ(Al)di(1dτ)rˉSi(Al)rSτ(Aj)didτrˉSi(Al)rSτ(Al)=(1di)(1dτ)[(|(Aj)ii|rSi(Aj))(|(Aj)ττ|rˉSτ(Aj))rˉSi(Aij)rSτ(Aj)]+didτ[(|(Al)ii|rSi(Al))(|(Al)ττ|rˉSτ(Al))rˉSi(Al)rSτ(Al)]=(1di)(1dτ)(βjiτ)ρSiτAj+didτ(βliτ)ρSiτAl(1di)(1dτ)(βjmin)ρSminAj+didτ(βlmin)ρSminAl>0,

    where t=j,l, and

    (βtmax)ρSminAt=miniS,τS{(βtiτ)ρSiτAt},
    (βtτi)ρSiτMt=(|(At)ii|rSi(At))(|(At)ττ|rˉSτ(At))rˉSi(At)rSτ(At).

    Therefore,

    ρSiτ(AD)=|aii|rSi(AD)+rSτ(AD)(|aii|rSi(AD))(|aττ|rˉSτ(AD))rˉSi(AD)rSτ(AD)<2maxt=j,l(αtmax)ρSmax(1di)(1dτ)(βjmin)ρSminMj+didτ(βlmin)ρSminMl2maxt=j,l(αtmax)ρSmax[min{(βjmin)ρSminAj,(βlmin)ρSminAl}]22max{(βjmin)ρSminAj,(βlmin)ρSminAl}=4maxt=j,l(αtmax)ρSmaxmax{(βjmin)ρSminAj,(βlmin)ρSminAl}[min{(βjmin)ρSminAj,(βlmin)ρSminAl}]2.

    Similarly, it can be inferred that

    ρˉSiτ(AD)4maxt=j,l(αtmax)ρˉSmaxmax{(βjmin)ρSminAj,(βlmin)ρSminAl}[min{(βjmin)ρSminAj,(βlmin)ρSminAl}]2,

    then

    maxM(M)maxj<l{0,1,,k}maxd[0,1]n[(ID)Aj+DAl]1maxmaxiS,τˉS{4maxt=j,l(αtmax)ρSmaxmax{(βjmin)ρSminAj,(βlmin)ρSminAl}[min{(βjmin)ρSminAj,(βlmin)ρSminAl}]2,4maxt=j,l(αtmax)ρˉSmaxmax{(βjmin)ρSminAj,(βlmin)ρSminAl}[min{(βjmin)ρSminAj,(βlmin)ρSminAl}]2}.

    Furthermore, from Definition 1.2, we can regard Aj and Al as two blocks in a row rearrangement fo A=(A0,A1,,Ak), and thus for t=j or t=l and for any iN, there exists ti{0,1,,k} such that

    (αtiτ)ρSiτ=(αtiiτ)ρSiτ,(αtτi)ρSτi=(αtiτi)ρˉSτi,(βtiτ)ρSiτAt=(βtiiτ)ρSiτAti,

    we have

    maxt=j,l(αtmax)ρSmax=maxt=j,l{maxi=S,τˉS(αtiτ)ρSiτ}=maxiS,τS{maxt=j,l(αtiiτ)ρSiτ}maxiS,τˉS{maxp=0,1,,k(αpiτ)ρSiτ}=maxp=0,1,,k{maxiS,τˉS(αpiτ)ρSiτ}=maxp=0,1,,k{(αpmax)ρSmax},

    and

    min{(βjmin)ρSminAj,(βlmin)ρSminAl}=mint=j,l{miniS,τˉS(βtiτ)ρSiτAt}=mint=j,l{miniS,τˉS(βtiiτ)ρSiτAti}=miniS,τS{mint=j,l(βtiiτ)ρSiτAti}miniS,τˉS{minp=0,1,,k(βpiτ)ρSiτAp}=minp=0,1,,k{miniS,τˉS(βpiτ)pSiτAp}=minp=0,1,,k{(βpmin)ρSminAp},

    and

    max{(βjmin)ρSminAj,(βlmin)ρSminAl}maxp=0,1,,k{(βpmin)ρSminAp},

    thus

    4maxt=j,l(αtmax)ρSmaxmax{(βjmin)ρSminAj,(βlmin)ρSminAl}[min{(βjmin)ρSminAj,(βlmin)ρSminAl}]24maxp=0,1,k{(αpmax)ρSmax}maxp=0,1,,k{(βpmin)ρSminAp}minp=0,1,,k{(βpmin)ρSminAp}2.

    Similarly, it can be concluded that

    4maxt=j,l(αtmax)ρˉSmaxmax{(βjmin)ρSminAj,(βlmin)ρSminAl}[min{(βjmin)ρSminAj,(βlmin)ρSminAl}]24maxp=0,1,k{(αpmax)ρˉSmax}maxp=0,1,,k{(βpmin)ρSminAp}minp=0,1,,k{(βpmin)ρSminAp}2,

    then for any Aj and Al for AR(A), satisfy

    maxM(M)maxj<l{0,1,,k}maxd[0,1]n[(ID)Aj+DAl]1maxmaxiS,τˉS{4maxp=0,1,k{(αpmax)ρSmax}maxp=0,1,,k{(βpmin)ρSminAp}minp=0,1,,k{(βpmin)ρSminAp}2,4maxp=0,1,k{(αpmax)ρˉSmax}maxp=0,1,,k{(βpmin)ρSminAp}minp=0,1,,k{(βpmin)ρSminAp}2},

    from the arbitrariness of Aj and Al, the conclusion follows.

    Next, we give a numerical example to illustrate that our results have advantages over some existing results.

    Example 1. Let A=(A0,A1,A2), where

    A0=(311161116),A1=(310151217),A2=(31.512631.913),

    are all SDD matrices and satisfies the condition of Proposition 2.1, thus A=(A0,A1,A2)A=(A0,A1,A2) has the row W-property. Then by Theorem 2.1 we can get

    α(A)2.0378.

    By Theorem 1.2, Since ρ(max(Λ10|Q0|,Λ11|Q1|,,Λ1k|Qk|))=0.8760<1, we get

    α(A)10.

    By Theorem 1.3 we get

    α(A)2.5210.

    According to the calculation example, it can be seen that new bound in Theorem 2.1 is sharper than those in Theorems 1.2 and 1.3 given by Zhang et al. in [1] in some cases.

    In this paper, I first apply the properties of SDD matrices to prove that block matrices composed of SDD matrices have row W-properties under certain conditions and obtains the extension of SDD matrices under these conditions for the error bound of the solution to the vertical linear complementarity problem. In the process of research, it is found that the error bounds of extended vertical linear complementarity problems for other types of matrices need to be further studied and explored, such as N-type matrices and CKV-type matrices.

    The author declares he/she has not used Artificial Intelligence (AI) tools in the creation of this article.

    The author declares no conflict of interest.



    [1] C. Zhang, X. J. Chen, N. H. Xiu, Global error bounds for the extended vertical LCP, Comput. Optim. Appl., 42 (2009), 335–352. http://dx.doi.org/10.1007/s10589-007-9134-9 doi: 10.1007/s10589-007-9134-9
    [2] M. S. Gowda, R. Sznajder, The generalized order linear complementarity problem, SIAM J. Matrix Anal. Appl., 15 (1994), 779–795. http://dx.doi.org/10.1137/S0895479892237859 doi: 10.1137/S0895479892237859
    [3] R. W. Cottle, G. B. Dantzig, A generalization of the linear complementarity problem, J. Combinat. Theory Series A, 8 (1970), 79–90. http://dx.doi.org/10.1016/S0021-9800(70)80010-2 doi: 10.1016/S0021-9800(70)80010-2
    [4] M. Sun, Monotonicity of Mangasarian's iterative algorithm for generalized linear complementarity problems, J. Math. Anal. Appl., 144 (1989), 474–485. http://dx.doi.org/10.1016/0022-247X(89)90347-8 doi: 10.1016/0022-247X(89)90347-8
    [5] M. Sun, Singular stochastic control problems in bounded intervals, Stochastics, 21 (1987), 303–344. http://dx.doi.org/10.1287/mnsc.17.9.612 doi: 10.1287/mnsc.17.9.612
    [6] M. Sun, Singular stochastic control problems solved by a sparse simplex method, Ima J. Math. Contro. Inf., 6 (1989), 27–38. http://dx.doi.org/10.1093/imamci/6.1.27 doi: 10.1093/imamci/6.1.27
    [7] L. L. Zhang, Z. R. Ren, Convergence of multi splitting iterative methods for M-Matrix linear complementarity problems, J. Math., 60 (2017), 547–556. http://dx.doi.org/10.3969/j.issn.0583-1431.2017.04.002 doi: 10.3969/j.issn.0583-1431.2017.04.002
    [8] R. W. Cottle, J. S. Pang, R. E. Stone, The linear complementarity problem, San Diego: Academic Press, 1992. http://dx.doi.org/10.1137/1.9780898719000
    [9] M. Z. Wang, M. M. Ali, On the ERM formulation and a stochastic approximation algorithm of the stochastic-R0 EVLCP, J. Math., 217 (2014), 513–534. http://dx.doi.org/10.1007/s10479-014-1575-9 doi: 10.1007/s10479-014-1575-9
    [10] J. Zhang, W. B. Shan, N. Shi, Smoothing SAA method for solving a special class of stochastic generalized vertical linear complementarity problems, J. Liaoning Normal Univ., 40 (2017), 18–23. http://dx.doi.org/CNKI:SUN:LNSZ.0.2017-03-004
    [11] L. P. Zheng, Z. Y. Gao, Global linear and quandratic one-step smoothing newton method for vertcal linear complementarity problems, Appl. Math. Mech., 24 (2003), 738–746. http://dx.doi.org/10.1007/BF02437876 doi: 10.1007/BF02437876
    [12] F. Q. Zhu, Iterative algorithms and related research for two types of linear complementarity problems, University of Electronic Science and Technology of China, 2007. http://dx.doi.org/10.7666/d.Y1105908
    [13] G. N. Chen, Matrix theory and applications, Science Press, 2007.
    [14] H. H. Wang, H. B. Zhang, C. Q. Li, Global error bounds for the extended vertical LCP of B-type matrices, Comput. Appl. Math., 40 (2021), 1–15. http://dx.doi.org/10.1007/s40314-021-01528-0 doi: 10.1007/s40314-021-01528-0
    [15] Y. X. Zhao, Error estimation of solutions for several types of structural matrix extended vertical linear complementarity problems, Guizhou Minzu University, 2023. http://dx.doi.org/10.27807/d.cnki.cgzmz.2023.000516
    [16] M. García-Esnaola, J. M. Peña, Error bounds for linear complementarity problems with a SDD matrices, Linear Algebra Appl., 438 (2013), 1329–1346. http://dx.doi.org/10.1016/j.laa.2012.09.018 doi: 10.1016/j.laa.2012.09.018
    [17] N. Moraca, Upper bounds for the innity norm of the inverse of SDD and S-SDD matrices, Comput. Appl. Math., 206 (2007), 666–678. http://dx.doi.org/10.1016/j.cam.2006.08.013 doi: 10.1016/j.cam.2006.08.013
  • Reader Comments
  • © 2024 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Metrics

Article views(824) PDF downloads(71) Cited by(0)

Other Articles By Authors

/

DownLoad:  Full-Size Img  PowerPoint
Return
Return

Catalog