In this paper, we present a unified approach to decomposing a special class of block tridiagonal matrices <i>K</i> (<i>α</i> ,<i>β</i> ) into block diagonal matrices using similar...In this paper, we present a unified approach to decomposing a special class of block tridiagonal matrices <i>K</i> (<i>α</i> ,<i>β</i> ) into block diagonal matrices using similarity transformations. The matrices <i>K</i> (<i>α</i> ,<i>β</i> )∈ <i>R</i><sup><i>pq</i>× <i>pq</i></sup> are of the form <i>K</i> (<i>α</i> ,<i>β</i> = block-tridiag[<i>β B</i>,<i>A</i>,<i>α B</i>] for three special pairs of (<i>α</i> ,<i>β</i> ): <i>K</i> (1,1), <i>K</i> (1,2) and <i>K</i> (2,2) , where the matrices <i>A</i> and <i>B</i>, <i>A</i>, <i>B</i>∈ <i>R</i><sup><i>p</i>× <i>q</i></sup> , are general square matrices. The decomposed block diagonal matrices <img src="Edit_00717830-3b3b-4856-8ecd-a9db983fef19.png" width="15" height="15" alt="" />(<i>α</i> ,<i>β</i> ) for the three cases are all of the form: <img src="Edit_71ffcd27-6acc-4922-b5e2-f4be15b9b8dc.png" width="15" height="15" alt="" />(<i>α</i> ,<i>β</i> ) = <i>D</i><sub>1</sub> (<i>α</i> ,<i>β</i> ) ⊕ <i>D</i><sub>2</sub> (<i>α</i> ,<i>β</i> ) ⊕---⊕ <i>D</i><sub>q</sub> (<i>α</i> ,<i>β</i> ) , where <i>D<sub>k</sub></i> (<i>α</i> ,<i>β</i> ) = <i>A</i>+ 2cos ( <i>θ<sub>k</sub></i> (<i>α</i> ,<i>β</i> )) <i>B</i>, in which <i>θ<sub>k</sub></i> (<i>α</i> ,<i>β</i> ) , k = 1,2, --- q , depend on the values of <i>α</i> and <i>β</i>. Our decomposition method is closely related to the classical fast Poisson solver using Fourier analysis. Unlike the fast Poisson solver, our approach decomposes <i>K</i> (<i>α</i> ,<i>β</i> ) into <i>q</i> diagonal blocks, instead of <i>p</i> blocks. Furthermore, our proposed approach does not require matrices <i>A</i> and <i>B</i> to be symmetric and commute, and employs only the eigenvectors of the tridiagonal matrix <i>T</i> (<i>α</i> ,<i>β</i> ) = tridiag[<i>β b</i>, <i>a</i>,<i>αb</i>] in a block form, where <i>a</i> and <i>b</i> are scalars. The transformation matrices, their inverses, and the explicit form of the decomposed block diagonal matrices are derived in this paper. Numerical exampl展开更多
文摘In this paper, we present a unified approach to decomposing a special class of block tridiagonal matrices <i>K</i> (<i>α</i> ,<i>β</i> ) into block diagonal matrices using similarity transformations. The matrices <i>K</i> (<i>α</i> ,<i>β</i> )∈ <i>R</i><sup><i>pq</i>× <i>pq</i></sup> are of the form <i>K</i> (<i>α</i> ,<i>β</i> = block-tridiag[<i>β B</i>,<i>A</i>,<i>α B</i>] for three special pairs of (<i>α</i> ,<i>β</i> ): <i>K</i> (1,1), <i>K</i> (1,2) and <i>K</i> (2,2) , where the matrices <i>A</i> and <i>B</i>, <i>A</i>, <i>B</i>∈ <i>R</i><sup><i>p</i>× <i>q</i></sup> , are general square matrices. The decomposed block diagonal matrices <img src="Edit_00717830-3b3b-4856-8ecd-a9db983fef19.png" width="15" height="15" alt="" />(<i>α</i> ,<i>β</i> ) for the three cases are all of the form: <img src="Edit_71ffcd27-6acc-4922-b5e2-f4be15b9b8dc.png" width="15" height="15" alt="" />(<i>α</i> ,<i>β</i> ) = <i>D</i><sub>1</sub> (<i>α</i> ,<i>β</i> ) ⊕ <i>D</i><sub>2</sub> (<i>α</i> ,<i>β</i> ) ⊕---⊕ <i>D</i><sub>q</sub> (<i>α</i> ,<i>β</i> ) , where <i>D<sub>k</sub></i> (<i>α</i> ,<i>β</i> ) = <i>A</i>+ 2cos ( <i>θ<sub>k</sub></i> (<i>α</i> ,<i>β</i> )) <i>B</i>, in which <i>θ<sub>k</sub></i> (<i>α</i> ,<i>β</i> ) , k = 1,2, --- q , depend on the values of <i>α</i> and <i>β</i>. Our decomposition method is closely related to the classical fast Poisson solver using Fourier analysis. Unlike the fast Poisson solver, our approach decomposes <i>K</i> (<i>α</i> ,<i>β</i> ) into <i>q</i> diagonal blocks, instead of <i>p</i> blocks. Furthermore, our proposed approach does not require matrices <i>A</i> and <i>B</i> to be symmetric and commute, and employs only the eigenvectors of the tridiagonal matrix <i>T</i> (<i>α</i> ,<i>β</i> ) = tridiag[<i>β b</i>, <i>a</i>,<i>αb</i>] in a block form, where <i>a</i> and <i>b</i> are scalars. The transformation matrices, their inverses, and the explicit form of the decomposed block diagonal matrices are derived in this paper. Numerical exampl
