Sampling-based learning control of quantum systems with bounded inputs and uncertainties via path planning

Robust control design is a central problem for quantum systems in practical implementation and applications. In this paper, we formulate the control problem of a quantum system with bounded uncertainties as the problem of steering this system to a target state with bounded controls via an optimized evolution path to achieve a satisfactory level of fidelity. To find the optimized path (controls), we present a combined design method of sampling-based learning control and path planning. The numerical results on an example of a four-level quantum system show the effectiveness of the proposed learning control design method.