We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

math.OC

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Mathematics > Optimization and Control

Title: On cost design in applications of optimal control

Abstract: A new approach to feedback control design based on optimal control is proposed. Instead of expensive computations of the value function for different penalties on the states and inputs, we use a control Lyapunov function that amounts to be a value function of an optimal control problem with suitable cost design and then study combinations of input and state penalty that are compatible with this value function. This drastically simplifies the role of the Hamilton-Jacobi-Bellman equation, since it is no longer a partial differential equation to be solved, but an algebraic relationship between different terms of the cost. The paper illustrates this idea in different examples, including $\mathcal{H}_{\infty}$ control and optimal control of coupled oscillators.
Comments: 6 pages, 4 figures
Subjects: Optimization and Control (math.OC)
DOI: 10.1109/LCSYS.2021.3079642
Cite as: arXiv:2104.00917 [math.OC]
  (or arXiv:2104.00917v2 [math.OC] for this version)

Submission history

From: Taouba Jouini [view email]
[v1] Fri, 2 Apr 2021 07:24:32 GMT (375kb,D)
[v2] Fri, 3 Sep 2021 08:38:11 GMT (444kb,D)

Link back to: arXiv, form interface, contact.