Small-time local controllability of the bilinear Schr\"odinger equation, despite a quadratic obstruction, thanks to a cubic term

TL;DR

This paper demonstrates that for a 1D bilinear Schrödinger equation, small controls in less regular spaces can achieve local controllability despite quadratic obstructions, thanks to the dominance of cubic nonlinear terms.

Contribution

It introduces a novel method leveraging cubic terms to recover controllability lost due to quadratic drifts in PDEs, extending control theory for quantum systems.

Findings

Cubic terms can compensate for quadratic drifts in control.

Controllability is achievable with small controls in less regular spaces.

A new tangent vector concept aids in infinite-dimensional control analysis.

Abstract

We consider a 1D linear Schr{\"o}dinger equation, on a bounded interval, with Dirichlet boundary conditions and bilinear control. We study its controllability around the ground state when the linearized system is not controllable. More precisely, we study to what extent the nonlinear terms of the expansion can recover the directions lost at the first order.In previous works, for any positive integer $n$, assumptions have been formulated under which the quadratic term induces a drift in the nonlinear dynamics, quantified by the $H^{-n}$ norm of the control. This drift is an obstruction to the small-time local controllability (STLC) under a smallness assumption on the controls in regular spaces. In this paper, we prove that for controls small in less regular spaces, the cubic term can recover the controllability lost at the linear level, despite the quadratic drift. The proof is inspired by Sussman's method to prove the sufficiency of the $\mathcal{S}(\theta)$ condition for STLC of ODEs. However, it uses a different global strategy relying on a new concept of tangent vector, better adapted to the infinite-dimensional setting of PDEs. Given a target, we first realize the expected motion along the lost direction by using control variations for which the cubic term dominates the quadratic one. Then, we correct the other components exactly, by using a STLC in projection result, with simultaneous estimates of weak norms of the control. These estimates ensure that the new error along the lost direction is negligible, and we conclude with the Brouwer fixed point theorem.