Lectures on stochastic control and nonlinear filtering. Lectures delivered at the Indian Institute of Science, Bangalore, under the T.I.F.R.-I.I.Sc. programme in applications of mathematics. Notes by K. M. Ramachandran (Q760404)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Lectures on stochastic control and nonlinear filtering. Lectures delivered at the Indian Institute of Science, Bangalore, under the T.I.F.R.-I.I.Sc. programme in applications of mathematics. Notes by K. M. Ramachandran
scientific article

    Statements

    Lectures on stochastic control and nonlinear filtering. Lectures delivered at the Indian Institute of Science, Bangalore, under the T.I.F.R.-I.I.Sc. programme in applications of mathematics. Notes by K. M. Ramachandran (English)
    0 references
    1984
    0 references
    Part I of this book describes the martingale theory and the stochastic calculus of jump processes in order to treat optimal control problems for a special class of jump processes, namely piecewise deterministic processes (PDP). Roughly speaking such processes, which had been introduced by the author earlier, are deterministic between the random jumps. So, PDP's stand in the middle between stochastic and deterministic control theory, and none of the standard results from both theories apply to the control of PDP's. As such processes include as special cases nearly all non-diffusion continuous-time processes, the author's contribution fills a gap in control theory which will certainly turn out to be most important for several applications. Using the stochastic calculus developed in the first section Markov properties for PDP's are derived, a Dynkin formula is presented, and (extended) infinitesimal generators for such processes are studied. The control theory for PDP's makes use of generalized dynamic programming results as developped by \textit{R. B. Vinter} and \textit{R. M. Lewis} [see e.g. SIAM J. Control Optimization 16, 571-583 (1978; Zbl 0392.49011)]. Part II of the book separately treats problems in filtering theory. The well known results as Kalman filters for linear systems, the contributions of Fujisaki, Kallianput, Kunita, and Zakai to filtering theory are presented. The main part, however, is devoted to the derivation of robust filtering, i.e. the derivation of filters which depend continuously on the observation. Mainly the contributions of the author to the theory of robust filtering are presented. The book is written in a most readable way, and may be seen as an excellent introduction into the theory of PDP's and into filtering theory. As, furthermore, it gives most of the known results in the two theories the book will turn out to be a useful reference book.
    0 references
    piecewise deterministic processes
    0 references
    Dynkin formula
    0 references
    generalized dynamic programming
    0 references
    robust filtering
    0 references
    0 references

    Identifiers

    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references