1: \begin{abstract}

2:   The non-linear optimization method developed by Konnov and

3:   Krotov [Automation and Remote Control \textbf{60}, 1427 (1999)] has

4:   been used previously to extend the capabilities of optimal control

5:   theory from the linear to

6:   the non-linear Schr\"odinger equation 

7:   [Sklarz and Tannor, Phys. Rev. A \textbf{66},

8:   053619 (2002)]. Here we show that based on the Konnov-Krotov method, 

9:   monotonically convergent algorithms are obtained for a 

10:   large class of quantum control problems. It includes, 

11:   in addition to non-linear equations of motion, control

12:   problems that are characterized by non-unitary time evolution, 

13:   non-linear dependencies of the Hamiltonian on the control, 

14:   time-dependent targets  and

15:   optimization functionals that depend to higher than second order on

16:   the time-evolving states. 

17:   We furthermore show that the non-linear (second order) contribution

18:   can be estimated  either analytically or

19:   numerically, yielding readily applicable optimization algorithms.

20:   We demonstrate monotonic convergence for an optimization functional

21:   that is an eighth-degree polynomial in the states.

22:   For the 'standard' quantum control problem of a convex

23:   final-time functional, linear equations of motion and linear

24:   dependency of the Hamiltonian on the field, the second-order

25:   contribution is not required for monotonic convergence but can be

26:   used to speed up convergence.  We demonstrate this by comparing 

27:   the performance of first and second order

28:   algorithms for two examples.

29: \end{abstract}

30: