Robust adaptive one-step ahead predictor

DOI10.1093/IMAMCI/18.4.491MaRDI QIDQ2777930zbMATH OpenOpenAlexFDO

Authors Vojislav Filipović

Publication date 4 December 2002

Published in IMA Journal of Mathematical Control and Information (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1093/imamci/18.4.491

zbMATH Keywords

adaptive predictionARMAX modeldiscrete-time systemglobal convergencenon-Gaussian noiseone-step ahead predictor

Mathematics Subject Classification ID

Design techniques (robust design, computer-aided design, etc.) (93B51) Adaptive control/observation systems (93C40) Discrete-time control/observation systems (93C55) Estimation and detection in stochastic control theory (93E10)

Review text

A robust adaptive one-step ahead predictor of a linear single-imput single-output discrete-time dynamic system, when a priori disturbance statistics are incomplete, is analysed. The given algorithm is robust with respect to uncertainty in the noise distribution. Using martingale theory, the global convergence of the adaptive prediction algorithm is proved in the situation when the input is not persistently exciting, an upper bound of the system order is known, arbitrary feedback is allowed between output and input, and the noise is non-Gaussian. The results can be extended to the case of a d-step-ahead predictor.

Recommendations

• Optimal design of robust predictors for linear discrete-time systems
• scientific article; zbMATH DE number 775740
• scientific article; zbMATH DE number 696813
• A global convergent outlier robust adaptive predictor for MIMO Hammerstein models
• Adaptive prediction in the presence of unmodelled dynamics

Cited in

(4)

• Reflections on `Adaptive prediction'
• One-step-ahead implementation
• A global convergent outlier robust adaptive predictor for MIMO Hammerstein models
• Optimal design of robust predictors for linear discrete-time systems

This page was built for publication: Robust adaptive one-step ahead predictor

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2777930)