Robust Quantum Control: Analysis & Synthesis via Averaging
An approach is presented for robustness analysis and quantum (unitary) control synthesis based on the classic method of averaging. The result is a multicriterion optimization competing the nominal (uncertainty-free) fidelity with a well known robustness measure: the size of an interaction (error) Hamiltonian, essentially the first term in the Magnus expansion of an interaction unitary. Combining this with the fact that the topology of the control landscape at high fidelity is determined by the null space of the nominal fidelity Hessian, we arrive at a new two-stage algorithm. Once the nominal fidelity is sufficiently high, we approximate both the nominal fidelity and robustness measure as quadratics in the control increments. An optimal solution is obtained by solving a convex optimization for the control increments at each iteration to keep the nominal fidelity high and reduce the robustness measure. Additionally, by separating fidelity from the robustness measure, more flexibility is available for uncertainty modeling.