qutip_qoc._goat module

This module contains functions that implement the GOAT algorithm to calculate optimal parameters for analytical control pulse sequences.

class qutip_qoc._goat._GOAT(objective, time_interval, time_options, control_parameters, alg_kwargs, guess_params, **integrator_kwargs)

Bases: object

Class for storing a control problem and calculating the fidelity error function and its gradient wrt the control parameters, according to the GOAT algorithm.

_prepare_generator_dia()

Combines the scaled and parameterized Hamiltonian elements on the diagonal of the coupled system (X, dX) Hamiltonian, with associated pulses: [[ H, 0, 0, …], [[ X], _[d1H, H, 0, …], [d1X], _[d2H, 0, H, …], [d2X], _[…, ]] […]] Additionlly, if the time is a parameter, the time-dependent parameterized Hamiltonian without scaling

_prepare_generator_off_dia()

Combines the scaled and parameterized Hamiltonian off-diagonal elements for the coupled system (X, dX) with associated pulses: [[ H, 0, 0, …], [[ X], _[d1H, H, 0, …], [d1U], _[d2H, 0, H, …], [d2U], _[…, ]] […]] The off-diagonal elements correspond to the derivative elements

_prepare_state()

inital state (t=0) for coupled system (X, dX): [[ X(0)], -> [[1], _[d1X(0)], -> [0], _[d2X(0)], -> [0], _[ … ]] -> [0]]

_solve_EOM(evo_time, params)

Calculates X, and dX i.e. the derivative of the evolution operator X wrt the control parameters by solving the Schroedinger operator equation returns X as Qobj and dX as list of dense matrices

gradient(params)

Calculates the gradient of the fidelity error function wrt control parameters by solving the Schroedinger operator equation

infidelity(params)

returns the infidelity to be minimized store intermediate results for gradient calculation the normalized overlap, the current unitary and its gradient