On reachable families of the Loewner differential equation in several complex variables (Q2634233): Difference between revisions

For the space \(L(\mathbb C^n)\) of linear operators \(A:\mathbb C^n\to\mathbb C^n\), the notation \[ \mathcal A:=\big\{A\in L(\mathbb C^n):k_+<2m(A)\big\} \] is used, where \(m(A):=\min\big\{\text{Re}\langle Az,z\rangle:z\in\mathbb C^n,\|z\|=1\big\}\), \(k_+(A):=\max\big\{\text{Re}\;\alpha:\alpha\in\sigma(A)\big\}\) and \(\sigma(A)\) is the spectrum of \(A\). Let \(\mathbb B_r^n\) denote the Euclidean ball centered at 0 with radius \(r>0\), \(\mathbb B_1^n=\mathbb B^n\), and for every \(A\in L(\mathbb C^n)\) with \(m(A)>0\), let \[ \mathcal N_A:=\Big\{h\in\mathcal H(\mathbb B^n):h(0)=0,\quad Dh(0)=A\quad\text{and}\quad\text{Re}\langle h(z),z\rangle>0\text{ for all }z\in\mathbb B^n\setminus\{0\}\Big\}. \] \(D\) stands for the differential with respect to the complex variables and \(\mathcal H(\mathbb B^n)\) is the space of holomorphic functions on \(\mathbb B^n\) with values in \(\mathbb C^n\). For every \(\Omega\subseteq\mathcal H(\mathbb B^n)\), \(\overline{\Omega}\) denotes the closure of \(\Omega\) with respect to the compact-open topology, and \(\mathrm{co\,}\Omega\) is the convex hull of \(\Omega\). Denote by \(\mathcal C(I,\Omega)\) the family of mappings \(h:\mathbb B^n\times I\to\mathbb C^n\), where \(I\subset[0,\infty)\) is an interval, \(h(\cdot,t)\in\Omega\) for all \(t\in I\), and \(h(z,\cdot)\) is measurable on \(I\) for all \(z\in\mathbb B^n\). Denote by \(v(\cdot,\cdot;h):\mathbb B^n\times I\to\mathbb B^n\), where \(h\in\mathcal C(I,\mathcal N_A)\), \(A\in L(\mathbb C^n)\) and \(m(A)>0\), the unique locally absolutely continuous solution on \(I\) of the initial value problem \[ \frac{\partial v}{\partial t}(z,t;h)=-h(v(z,t;h),t),\;\;v(z,0;h)=z. \] For every \(A\in\mathcal A\) and \(T\in[0,\infty)\), let \[ \tilde R_T(\mathrm{id}_{\mathbb B^n},\Omega):=\Big\{e^{TA}v(\cdot,T;h):h\in\mathcal C([0,T]),\Omega\Big\}. \] For \(h\in\mathcal C(I,\mathcal N_A)\), denote \(\tilde h(z,t):=e^{tA}\big(h(e^{-tA}z,t)-Ae^{-tA}z\big)\) and \(\tilde{\mathcal C}(I,\Omega):=\big\{\tilde h:h\in\mathcal C(I,\Omega)\big\}\). Let \(\mathcal A^q(\mathbb B_r^n):=\mathcal H(\mathbb B_r^n)\cap L^q(\mathbb B_r^n,\mathbb C^n)\), and let \(cl^w_{L_r^{p,q}}\) denote the closure with respect to the weak topology of \(L_p(I,\mathcal A^q(\mathbb B_r^n))\). The main result is given in the following theorem. Theorem 3.3. For every \(r\in(0,1)\), \(p,q\in(1,\infty)\), \(T\in[0,\infty]\), \(A\in\mathcal A\) and \(\Omega\subseteq\mathcal N_A\), we have \[ \overline{\tilde R_T(\mathrm{id}_{\mathbb B^n},\tilde C(I,\Omega))}=\tilde R_T\big(\mathrm{id}_{\mathbb B^n},cl^w_{L_r^{p,q}}\tilde C(I,\Omega)\big)=\tilde R_T\big(\mathrm{id}_{\mathbb B^n},\tilde C(I,\overline{\mathrm{co\,}\Omega})\big), \] where \(I\) is \([0,T]\) if \(T\in[0,\infty)\) and \([0,\infty)\) if \(T=\infty\).