Abstract
Computation of derivatives (gradient and Hessian) of a fidelity function is one of the most crucial steps in many optimization algorithms. Having access to accurate methods for computing these derivatives is even more desirable where the optimization process requires propagation of these computations over many steps, which is particularly important in optimal control of spin systems. Here we propose a novel numerical approach, ESCALADE (Efficient Spin Control using Analytical Lie Algebraic Derivatives), that offers the exact first and second derivatives of the fidelity function by taking advantage of the properties of the Lie group of 2 × 2 unitary matrices, SU(2), and its Lie algebra, the Lie algebra of skew-Hermitian matrices, su(2). A full mathematical treatment of the proposed method along with some numerical examples are presented.
| Original language | English |
|---|---|
| Article number | 109611 |
| Journal | Automatica |
| Volume | 129 |
| Early online date | 10 Apr 2021 |
| DOIs | |
| Publication status | Published - 31 Jul 2021 |
Funding
MF thanks the Royal Society, UK for a University Research Fellowship and a University Research Fellow Enhancement Award (grant numbers URF\R1\180233 and RGF\EA\181018 ). PS thanks Trinity College Oxford for a Junior Research Fellowship and Mathematical Institute, Oxford, where most of this research was carried out.
| Funders | Funder number |
|---|---|
| Trinity College Oxford | |
| Royal Society | URF\R1\180233, RGF\EA\181018 |
Keywords
- Derivatives
- ESCALADE
- Lie algebra
- Numerical algorithms
- Optimal control
ASJC Scopus subject areas
- Electrical and Electronic Engineering
- Control and Systems Engineering