Formulation and numerical solution of finite-level quantum optimal control problems (Q2483345)

The authors consider localized finite-level quantum systems modeled by a Schrödinger equation for an \(n\)-component wave function \(\psi\) as follows \[ i\dot\psi(t)= H(\varepsilon(t))\psi(t)\text{ for }t\in[0,T]\text{ and }\psi(0)= \psi_0, \] where \(\varepsilon(t)\) is the control. The optimality criteria is realized in the cost functional \[ J(\psi,\varepsilon)= \tfrac{1}{2}|\psi(T)- \psi_d|^2_{C^n}+ \tfrac{\gamma}{2}\|\varepsilon\|^2_{L^2(0,T;C)}+ \tfrac{\mu}{2}\|\dot\varepsilon\|^2_{L^2(0,T;C)}+ \tfrac{1}{2} \sum_{j\in I}\alpha_j\|\psi_j\|^2_{L^2(0,T;C)}. \] This optimal control problem can be written as the following abstract minimization problem \[ \min J(x)\text{ subject to }x= (\psi,\varepsilon)\in X\text{ and }e(x)= 0\quad\text{in }Y. \] For these problems, first-order necessary optimality conditions and second-order sufficient optimality conditions are investigated. For solving these problems, a cascadic nonlinear conjugate gradient scheme and a monotonic scheme are discussed. Numerical experiments are presented.