Gates using the Krotov method
Luis Miguel Fabra
Hello everyone,
I am trying to reproduce the results of an article that uses the Krotov method for quantum control, but I have some questions.
The thing is, in the article, there is a part about quantum gates, where they use three initial states in particular with weights “wi” to reproduce the Hadamard gate using JT_re(), and I understand that Krotov supports different ways of defining objectives, for example: krotov.objectives. Objective(*, initial_state, H, target, c_ops=None) to define separate objectives and krotov.objectives.gate_objectives(basis_states, gate, H, c_ops=None, local_invariants=False, liouville_states_set=None, weights=None, normalize_weights=True), which is ideal for gates.
In principle, I could use gate_objectives() since it is specifically for gates, but I can't find a way to use the initial states of the article precisely, given that this class only receives the base and, in more special cases, the Liouville base (‘3states’, ‘d+1’, ‘full’), which are different from those in the article. On the other hand, if I define them using the Objective() class, this time I can use those of the article, but as you can see, this class does not allow me to add the weights of each state.
How do I define those particular states with their respective weights in Krotov so that they are well defined and the function works?On the other hand, in previous tests, the article uses parameter values of lambda=1e-3 and 5e-4, but in my results, the control that Krotov returns for those values is very noisy with very large amplitudes and is indeed different from that in the article. Why is that? What should I do to prevent Krotov from returning such an unstable control with those lambda values?
Nathan Shammah
--
You received this message because you are subscribed to the Google Groups "QuTiP: Quantum Toolbox in Python" group.
To unsubscribe from this group and stop receiving emails from it, send an email to qutip+un...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/qutip/CAEK8qsTULFK7B2bqyN72wcRDo3U%2BGNjngy2VrOv5LK5SWCdCPA%40mail.gmail.com.