Common eigenvectors of two matrices (Q761518)

The following questions are studied. Given a matrix \(A\in C^{n,n}\), and a subspace \({\mathcal U}\), described by \(Bx=0\), \(B\in C^{m,n}\), when has A an eigenvector in \({\mathcal U}?\) Given two matrices \(A,B\in C^{n,n}\), when exists a common eigenvector? The main results are Theorem 2.1. There exists an eigenvector x of A with \(Bx=0\) iff \(\cap^{q-1}_{k=0}\ker (BA^ k)\neq \{0\},\) where q is any integer greater or equal to the degree of the minimal polynomial of A. This result can be related to the properties of ''observability'' and ''controllability'' in system theory. Theorem 3.1. Given A, \(B\in C^{n,n}\). There exists \(x\neq 0\), \(\lambda\),\(\mu\in C\) such that \(Ax=\lambda x\), \(Bx=\mu x\) iff \(\cap^{n- 1}_{k,\ell =1}\ker [A^ k,B^{\ell}]\neq \{0\}.\) Here \([A,B]=AB-BA\).