On the pole-shifting problem for non-commutative rings (Q1177250): Difference between revisions

If \(A\) is an \(n\times n\) matrix and \(B\) is an \(m\times n\) matrix over \(R\), then \((A,B)\) is called \(n\)-dimensional system over \(R\). Two systems \((A_ 1,B_ 1)\) and \((A_ 2,B_ 2)\) are called feedback equivalent if \(B_ 2=S^{-1} B_ 1 T\) and \(A_ 2=T^{-1} A_ 1 T+A_ 1 T+LB_ 1 T\) for some matrices \(S\), \(T\), \(L\). The system \((A,B)\) is controllable if \(R^ n=\text{Im}(B)+\text{Im}(BA)+\text{Im}(BA^ 2)+\dots\) (here \(B\) is considered as a map from \(R^ m\) to \(R^ n\)). An \(m\times n\) matrix \(B\) is good if there exists an \(n\times n\) matrix \(A\) such that \((A,B)\) is controllable. The system \((A,B)\) is triangularizable if for each \(r_ 1,r_ 2,\dots,r_ n\in R\) there exists a system \((F,G)\) feedback equivalent to \((A,B)\) such that \[ F=\left[\begin{matrix} r_ 1 &0 &\cdots &0\\ \ast &r_ 2 &\cdots &0\\ \cdot &\cdot &\cdots &\cdot\\ \ast &\ast &\cdots &r_ n\end{matrix}\right]. \] Let \((A,B)\) be an \(n\)-dimensional controllable system and \(r_ 1,r_ 2,\dots,r_ n\in R\). The pole- shifting problem is the following: to find a system \((F,G)\) feedback equivalent to \((A,B)\) such that the eigenvalues of \(F\) are \(r_ 1,r_ 2,\dots,r_ n\). Over a noncommutative ring the pole-shifting problem translates into the question of whether or not each controllable system can be triangularized. The main result of this paper is the following Theorem: Let \(R\) be a ring satisfying the property: for each good matrix \(B\) over \(R\) there exists an invertible matrix \(P\) such that \(vBP=[00\cdots 01]\) for some vector \(v\). Then each controllable system over \(R\) is triangularizable (therefore the pole-shifting problem has the positive solution). Corollaries show some particular cases when the condition of the theorem is satisfied.