Axiomatisation of Fully Probabilistic Design Revisited

0525231 - ÚTIA 2021 RIV NL eng J - Journal Article
Kárný, Miroslav
Axiomatisation of Fully Probabilistic Design Revisited.
Systems and Control Letters. Roč. 141, č. 1 (2020), č. článku 104719. ISSN 0167-6911. E-ISSN 1872-7956
R&D Projects: GA MŠMT(CZ) LTC18075
Grant - others:The European Cooperation in Science and Technology (COST)(XE) CA16228
Institutional support: RVO:67985556
Keywords : Closed-loop control * Control theory * Stochastic control * Stochastic modelling * Performance indices
OECD category: Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)
Impact factor: 2.804, year: 2020
Method of publishing: Limited access
http://library.utia.cas.cz/separaty/2020/AS/karny-0525231.pdf https://www.sciencedirect.com/science/article/pii/S0167691120301006

Fully probabilistic design (FPD) of control strategies models both the closed control loop and control objectives by joint probabilities of involved variables. It selects the optimal strategy as the minimiser of Kullback–Leibler (KL) divergence of the closed-loop model to its ideal counterpart expressing the control objectives. Since its proposal (Kárný, 1996) and general algorithmisation (Kárný and Guy, 2006), FPD has been axiomatised (Kárný and Kroupa, 2012) and successfully applied both theoretically (Kárný and Guy, 2012) and practically (Quinn et al., 2003. Kárný et al., 2006)[1]. This paper refines the FPD axiomatisation and bridges FPD to standard stochastic control theory, which it encompasses, in a better way. This enhances applicability of both as well as of its popular, independently proposed, special case known as KL control (Guan et al., 2014).
Permanent Link: http://hdl.handle.net/11104/0309413