Iterative identification and control redesign via unfalsified sets of models: A basic scheme

DOI10.1080/002071799220623MaRDI QIDQ4485270zbMATH OpenOpenAlexFDO

Authors S. M. Veres

Publication date 12 June 2000

Published in International Journal of Control (Search for Journal in Brave)

Full work available at URL https://doi.org/10.1080/002071799220623

zbMATH Keywords

convergencemodel structure selectionparameter identificationparametric perturbationsrobust control designstability

Mathematics Subject Classification ID

System identification (93B30) Design techniques (robust design, computer-aided design, etc.) (93B51)

Review text

This paper introduces a new worst-case iterative identification and control scheme which is a combination of parameter bounding and control design. The plant model is given by \[ y_t +a_1 y_{t-1} +\cdots +a_n y_{t-n}= \] \[ = b_0u_{t-d} + b_1 u_{t-d-1} +\cdots +b_m u_{t- m-d} + P(y_t) + U(u_t)+e_t, \] where \( P,U \) represent parametric perturbations and unmodelled dynamics respectively; \( e_t \) represents an unstructured additive disturbance. The present scheme ensures global convergence to an ideal best performance that can be achieved under the given set of model and controller structures. Model structure selection is contained in the iterative identification and control design scheme proposed. The scheme makes an effort to modify the model only if the new experiment strongly contradicts the current model. The controller used in an experiment depends on the model parameter obtained from the previous experiment by choosing the most favourable one. Computation of the controller parameters is based on a fixed control design to obtain a linear controller of the form proposed. The same controller will be used in consecutive experiments until the input-output data contradict those of the previous model used and hence the controller has to be modified. There is a time after which no further modification of the controller will be necessary. The main result of this paper is the proof of convergence of the iterative identification and control scheme. Several numerical results of identification and feedback control are presented and discussed.

Recommendations

• Convergence of control performance by unfalsification of models --- levels of ence
• Suboptimal feedback control by a scheme of iterative identification and control design
• A combined iterative scheme for identification and control redesigns
• Iterative modelling in \(H_{\infty}\) framework

Cited in

(11)

• Predictive self-tuning control by parameter bounding and worst-case design
• From experiment design to closed-loop control
• A model reference approach to safe controller changes in iterative identification and control
• Convergence of control performance by unfalsification of models --- levels of ence
• Iterative modelling in \(H_{\infty}\) framework
• Iterative identification and two step control design for partially unknown unstable plants
• Iterative identification and control design based on Vinnicombe distance
• Iterative identification and control design using finite-signal-to-noise models
• Self-tuning control by model unfalsification (Part II)
• A combined iterative scheme for identification and control redesigns
• An iterative identification based on the Youla-Kucera parameterisation without model reduction

This page was built for publication: Iterative identification and control redesign via unfalsified sets of models: A basic scheme

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4485270)