In this paper we consider a gradual-impulsive control problem for continuous-time Markov decision processes (CTMDPs) with total cost criteria and constraints. We develop a simple and useful method, which reduces the concerned problem to a standard CTMDP problem with gradual control only. This allows us to derive straightforwardly and under a minimal set of conditions the optimality results (sufficient classes of control policies, as well as the existence of stationary optimal policies) for the original constrained gradual-impulsive control problem.
|Number of pages||23|
|Journal||SIAM Journal on Control and Optimization|
|Publication status||Published - 2020|
Bibliographical noteFunding Information:
∗Received by the editors March 6, 2019; accepted for publication (in revised form) October 14, 2019; published electronically January 16, 2020. https://doi.org/10.1137/19M1248303 Funding: This work was partially supported by the Royal Society, grant IE160503. †Department of Mathematical Sciences, University of Liverpool, Liverpool, L69 7ZL, UK (email@example.com). ‡Corresponding author. Department of Mathematical Sciences, University of Liverpool, Liverpool, L69 7ZL, UK (firstname.lastname@example.org).
© 2020 Society for Industrial and Applied Mathematics.
ASJC Scopus subject areas
- Control and Optimization
- Applied Mathematics