= Solution
<Hamiltonian engineering> chooses accessible interactions so that their actual or effective <Quantum Hamiltonian> produces the desired evolution. A typical model is $H_f(t)=H_0+\sum_kf_k(t)H_k$; the controls may change frequencies, amplitudes, phases or coupling strengths. Three standard model-based <open-loop control> approaches illustrate different mechanisms.
In <resonant quantum control>, select a transition with a near-resonant field. In a rotating frame, after neglecting rapidly oscillating terms in a justified regime, an effective two-level drive can be $H_{\mathrm{eff}}(t)=\Omega(t)\sigma_x/2$. When its axis is fixed,
$$
U(T)=\exp\left[-\frac{i\sigma_x}{2}\int_0^T\Omega(t)\,dt\right].
$$
The pulse area sets the rotation angle and the drive phase selects the transverse axis. A resonant pulse of area $\pi$ swaps the two level populations. This requires knowledge of transition frequencies, couplings and the validity of the selective-drive approximation.
In <adiabatic quantum control>, design a slowly varying <Quantum Hamiltonian> whose desired instantaneous eigenstate connects the initial and final states. For a nondegenerate eigenstate separated by a gap, the <quantum adiabatic theorem> explains why sufficiently slow evolution follows that state up to a phase. The relative size of matrix elements of $\dot H$ and squared spectral gaps controls the approximation; degeneracies or rapid changes invalidate simple following. This engineers a state path rather than a prescribed resonant rotation.
In <average Hamiltonian engineering>, apply a predetermined sequence of fast pulses $P_j$ and let the system evolve for intervals $\tau_j$ in the corresponding frames. For cycle length $T_c=\sum_j\tau_j$, the leading effective generator is
$$
\overline H^{(0)}=\frac1{T_c}\sum_j\tau_j P_j^\dagger H_0P_j.
$$
Choose the sequence to cancel unwanted terms or retain desired couplings. For instance, equal intervals under $H_0=\omega\sigma_z/2$ and its $\sigma_x$ conjugate cancel the leading generator because $\sigma_x\sigma_z\sigma_x=-\sigma_z$. Noncommuting terms produce higher-order <commutator> corrections, so fast cycling, finite pulse errors and the accuracy of the model matter. All three methods determine a waveform or sequence before its execution; none automatically corrects an unknown mismatch during that run.
Back to article page
