Feature- and process-based optimal control of quantum dynamics

Preparing desired quantum states and quantum operations (processes) is essential for numerous tasks in quantum computation. Several approaches have been developed for optimal control of quantum states, whereas optimal strategies for preparation of a given quantum process have remained fairly less explored. For some applications, rather than a specific desired state or process, it may suffice to obtain states or processes with specific desired features such as (high) coherence or purity. In such cases, fidelity-based measures alone are inadequate for evaluating the performance of a quantum control strategy, and hence proper feature-based figures of merit should be employed. Here we develop a feature-based optimal coherent control formalism for open quantum processes under Markovian evolutions which demonstrate high coherence and purity features. In particular, we observe that performance of the feature-based Krotov optimization algorithm can be improved by choosing educated initial guess fields. In addition, in a model for a qutrit Rydberg system, it is shown that the convex overlap-based fidelity outperforms other process-based as well as state-based fidelities. This analysis underscores the utility of feature-based optimal control strategies for quantum processes.