QuantumControl.jl

HEADS UP: this repo was experimental and functionality will be moved to my new repo: QubitControl.jl

this package aims to provide an interface between the python package QuTiP and the Altro.jl trajectory optimization package.

the goal is to do multi-qubit quantum optimal control quickly and robustly.

installation

to use this package, clone it, and then, from a julia REPL in the cloned directory run

(@v1.7) pkg> activate .

(QuantumControl) pkg> instantiate

one should have a python environment containing scipy and qutip (this can be achieved with conda, by activating the environment before running the julia scripts.

usage

see experiments directory for examples including:

• single qubit gates with a single quantum state
• single qubit gates with multiple quantum states
• bootstrapped single qubit, multi-state solves

quantum optimal control

this problem involves taking advantage of an experimental time-dependent qubit Hamiltonian that has a controllable parameter, e.g.

$$ \begin{equation} H(t, a(t)) = f_q {\sigma_z \over 2} + a(t) {\sigma_x \over 2} \end{equation} $$

which governs the dynamics of a qubit state via the Schroedinger equation:

$$ \begin{equation} {d \over dt} \ket{\psi} = -iH(t, a(t)) \ket{\psi} \end{equation} $$

The goal is then to find $a(t)$ s.t. $\ket{\psi(t_0)} \to \ket{\psi(t_f)} = X\ket{\psi(t_0)}$ with the dynamics satisfying (1) and (2).

example: single fluxonium qubit

The above plot shows the dynamics for the wavefunction and control $a(t)$ s.t.

$$ \ket{0} \to X \ket{0} = \ket{1} $$

TODO:

• multistate single qubit script
• scripts for X, Y, Z gates
• add functionality to define wavefunctions as complex vectors
• add plotting utilities
• two qubit problem