A geometric characterization of the slow space of the Hamiltonian system arising from the singular LQR problem
Author(s) -
Imrul Qais,
Debasattam Pal,
Chayan Bhawal
Publication year - 2020
Publication title -
ifac-papersonline
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.308
H-Index - 72
eISSN - 2405-8971
pISSN - 2405-8963
DOI - 10.1016/j.ifacol.2020.12.342
Subject(s) - linear subspace , eigenvalues and eigenvectors , linear quadratic regulator , mathematics , hamiltonian (control theory) , matrix pencil , hamiltonian matrix , quadratic equation , matrix (chemical analysis) , pencil (optics) , hamiltonian system , mathematical analysis , pure mathematics , symmetric matrix , mathematical optimization , optimal control , physics , geometry , quantum mechanics , materials science , composite material
In this paper we first characterize the slow space of a given state-space system. We provide this characterization in terms of an eigenspace of the corresponding Rosenbrock matrix pair. We also characterize the “good” slow space in terms of a stable eigenspace of the Rosenbrock matrix pair. Moreover, we show how the dimensions of these subspaces can be calculated from the determinant of the Rosenbrock matrix pencil. Then, we apply these results to the Hamiltonian system arising from the singular linear quadratic regulator (LQR) problem and explore a few interesting properties of the good slow space of this Hamiltonian system. Finally, we provide a feedback law to achieve the smooth optimal solutions.
Accelerating Research
Robert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom
Address
John Eccles HouseRobert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom