State estimator design using Jordan based long short-term memory networks

Avneet Kaur a and Kirsten Morris a · 2025

Read State estimator design using Jordan based long short-term memory networks by Avneet Kaur a and Kirsten Morris a (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-04518.txt.

Description

Avneet Kaur and Kirsten Morris design a neural estimator that infers a system's hidden state from noisy readings, in a 2025 paper. State estimation is easy for simple linear systems, where the Kalman filter is optimal, but hard for nonlinear ones. The authors build the estimator from a long short-term memory network whose recurrent wiring follows the older Jordan network design.

They compare their Jordan-based network with the standard Elman-style network and with Kalman-type filters on linear and nonlinear test systems. For nonlinear problems, the learning-based estimators cut both error and computation time. The Jordan design also needed less training to match the standard network's accuracy. The results suggest a practical path toward near-optimal estimation for complex, high-order systems.

Catalog record

CREATORAvneet Kaur a and Kirsten Morris a
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE44 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2502-04518

Browse the full library · About the Tower of Babel