A gradient algorithm for Hamiltonian identification of open quantum systems
It addresses Hamiltonian identification for open quantum systems, a problem relevant to quantum control and characterization, but the method is incremental (gradient descent applied to a known master equation framework).
The paper proposes a gradient algorithm to identify unknown Hamiltonian parameters in open quantum systems from time traces of local observables, demonstrating effectiveness on a circuit QED system and for learning non-Markovian environment spectra.
In this paper, we present a gradient algorithm for identifying unknown parameters in an open quantum system from the measurements of time traces of local observables. The open system dynamics is described by a general Markovian master equation based on which the Hamiltonian identification problem can be formulated as minimizing the distance between the real time traces of the observables and those predicted by the master equation. The unknown parameters can then be learned with a gradient descent algorithm from the measurement data. We verify the effectiveness of our algorithm in a circuit QED system described by a Jaynes-Cumming model whose Hamiltonian identification has been rarely considered. We also show that our gradient algorithm can learn the spectrum of a non-Markovian environment based on an augmented system model.