An Input-Constrained Reduced-Order Observer and NARMAX Model-Based Adaptive Tracker with Fault Tolerance for Unknown Systems with an Input-Output Direct Feed-Through Term

  • 許 文騰

Student thesis: Doctoral Thesis

Abstract

This dissertation proposes an input-constrained reduced-order observer-based adaptive tracker with fault tolerance for unknown nonlinear stochastic systems with an input-output direct feed-through term to obtain good tracking performance The major contributions of this dissertation are stated as follows: Firstly realization of causal current output-based optimal full/reduced-order observer and tracker for the linear sampled-data system with a direct transmission term from input to output is newly proposed Furthermore the time derivative of the control input existed in the continuous-time reduced-order observer can be avoided in the proposed one for the continuous-time system with an input-output feed-through term Secondly an active full-order fault tolerance tracker using the modified nonlinear autoregressive moving average with exogenous inputs (NARMAX) model for state-space self-tuning control (STC) of unknown nonlinear stochastic hybrid systems with a direct transmission term is proposed For the system identification process the initial parameters of the modified NARMAX model assigned by the off-line OKID and can speed up the convergence speed of the recursive extended least squares (RELS) method to determine a reliable dynamic model Moreover the modified state-space self-tuning control methodology due to the NARMAX model can quickly make an appropriate reaction to the variation of system parameters when the abrupt input fault and/or the gradual input fault occurs Finally an input-constrained reduced-order observer-based self-tuner with fault tolerance for unknown nonlinear stochastic systems with an input-output direct feed-through term has been proposed Some illustrative examples are given to demonstrate the effectiveness of the proposed methodologies
Date of Award2014 Jul 28
Original languageEnglish
SupervisorJason Sheng-Hon Tsai (Supervisor)

Cite this

'