New doubling algorithm for the discrete periodic Riccati equation (Q1316136)

From MaRDI portal
Revision as of 01:29, 27 June 2024 by Daniel (talk | contribs) (‎Created claim: Wikidata QID (P12): Q126437416, #quickstatements; #temporary_batch_1719448105926)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
scientific article
Language Label Description Also known as
English
New doubling algorithm for the discrete periodic Riccati equation
scientific article

    Statements

    New doubling algorithm for the discrete periodic Riccati equation (English)
    0 references
    30 October 1994
    0 references
    A new doubling algorithm (NDA) is introduced for the recursive discrete periodic Riccati equation for the filtering error variance with initial condition. Doubling algorithms have been previously proposed by \textit{D. G. Lainiotis} [IEEE Trans. automatic Control AC-21, 677-689 (1976; Zbl 0344.93063)] for time invariant models only. It is shown by numerical tests that the NDA is numerically efficient. Its computational requirements (time and memory) are established for the sequential implementation and comparisons are made to classical algorithms (Kalman filter and Lainiotis filter). The time required by the NDA is shorter when the length of the a priori unknown time required to reach the steady state solution is not very small. The NDA is also numerically stable and possesses very good parallelism efficiency.
    0 references
    Kalman filter
    0 references
    doubling algorithm
    0 references
    recursive discrete periodic Riccati equation
    0 references
    numerical tests
    0 references
    Lainiotis filter
    0 references
    0 references
    0 references
    0 references

    Identifiers

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