Nonconvex Optimization and Its Applications, 2005, Volume 79, Part 2, 995-1006, DOI: 10.1007/0-387-24276-7_58

On the Convergence of the Matrices Associated to the Adjugate Jacobians

C. Sbordone

View Related Documents

Abstract

For any n × n matrix D ∈ ℝ n×n let adjD denote the transpose of its cofactors. If det D > 0 then there exists a symmetric matrix $ \mathcal{A} = \mathcal{A}\left( \mathcal{D} \right) $ \mathcal{A} = \mathcal{A}\left( \mathcal{D} \right) with det $ \mathcal{A} $ \mathcal{A} = 1 such that
$ adjD = \left( {detD} \right)^{\tfrac{{n - 2}} {n}} \mathcal{A}D^l $ adjD = \left( {detD} \right)^{\tfrac{{n - 2}} {n}} \mathcal{A}D^l
Where D’ is the transpose of D.
For n = 2, K ≥ 1, the set of K -quasiconformal matrices is denoted by
$ Q_2 \left( K \right) = \left\{ {D \in \mathbb{R}^{2 \times 2} :\left\| D \right\|^2 \leqslant K det D} \right\} $ Q_2 \left( K \right) = \left\{ {D \in \mathbb{R}^{2 \times 2} :\left\| D \right\|^2 \leqslant K det D} \right\}
Furthermore define
$ \varepsilon _2 \left( K \right) = \left\{ {\mathcal{A} \in \mathbb{R}^{2 \times 2} :\frac{I} {K} \leqslant \mathcal{A}^l = A \leqslant KI, det \mathcal{A} = 1} \right\} $ \varepsilon _2 \left( K \right) = \left\{ {\mathcal{A} \in \mathbb{R}^{2 \times 2} :\frac{I} {K} \leqslant \mathcal{A}^l = A \leqslant KI, det \mathcal{A} = 1} \right\}
For variable matrices D = D(x)∈ ɛ 2(K) for a.e. x ∈ Ω ⊂ ℝ2×2, Ω a simply connected and bounded domain, a natural question is to see how does $ \mathcal{A} = \mathcal{A}\left( {x,\mathcal{D}} \right) $ \mathcal{A} = \mathcal{A}\left( {x,\mathcal{D}} \right) (x,D) change with D(x).

Fulltext Preview

Image of the first page of the fulltext document