✦ LIBER ✦

Towards more precise recursive prediction error algorithms

✍ Scribed by Roman Weinfeld

Publisher: Elsevier Science
Year: 1990
Tongue: English
Weight: 328 KB
Volume: 26
Category: Article
ISSN: 0005-1098
DOI: 10.1016/0005-1098(90)90011-6

No coin nor oath required. For personal study only.

✦ Synopsis

This paper presents a variant of the Recursive Gauss-Newton algorithm for identification of the model

To derive the algorithm the formulae for the predictor and its derivatives have been transformed and then applied to build up more exact schema for the evaluation of the gradient and the Hessian. The tests show that this algorithm has an important advantage in that it converges, in general, substantially faster than the conventional one. It turns out that the presented idea, as a tool, is of a very general nature and might be applied to other algorithms.

📜 SIMILAR VOLUMES

Recursive prediction error algorithms wi

Recursive prediction error algorithms without a stability test

✍ Haim Weiss; John B. Moore 📂 Article 📅 1980 🏛 Elsevier Science 🌐 English ⚖ 507 KB

In this paper, recursive prediction error identification schemes are studied which incorporate Kalman gain calculations to ensure exponential stability of the predictor and related recursions. Thus there is an alternative to the trial and error projection techniques and stability test calculations o

Analysis of a general recursive predicti

Analysis of a general recursive prediction error identification algorithm

✍ Lennart Ljung 📂 Article 📅 1981 🏛 Elsevier Science 🌐 English ⚖ 979 KB

A general recursive identification algorithm is found to have the same convergence properties as a family of off-line prediction error methods.