Model validation: a connection between robust control and identification
Roy S. Smith,John Doyle +1 more
TLDR
The model validation problem addressed is: given experimental data and a model with both additive noise and norm-bounded perturbations, is it possible that the model could produce the observed input-output data?Abstract:
The gap between the models used in control synthesis and those obtained from identification experiments is considered by investigating the connection between uncertain models and data. The model validation problem addressed is: given experimental data and a model with both additive noise and norm-bounded perturbations, is it possible that the model could produce the observed input-output data? This problem is studied for the standard H/sub infinity // mu framework models. A necessary condition for such a model to describe an experimental datum is obtained. For a large class of models in the robust control framework, this condition is computable as the solution of a quadratic optimization problem. >read more
Citations
More filters
Journal ArticleDOI
Probabilistic Estimates for Mixed Model Validation Problems With ${\cal H}_{\infty}$ Type Uncertainties
Wenguo Liu,Jie Chen +1 more
TL;DR: Bounds on this probability are computable based on the distribution of Chi-square random variables when the noise is a Gaussian variable, and solvable as an LMI problem when only statistical information such as the expectation and covariance of the noise are known.
ReportDOI
Enlightened Multiscale Simulation of Biochemical Networks. Core Theory, Validating Experiments, and Implementation in Open Software
John Doyle,Michael Hucka +1 more
TL;DR: A distinguishing theme of this work is its focus on scalable methods of robustness and model validation and invalidation with data, as opposed to relying purely on simulation.
Proceedings ArticleDOI
The experimental validation of robust control models for a heat experiment: a linear matrix inequality approach
Geir E. Dullerud,Roy S. Smith +1 more
TL;DR: Theoretical developments in time domain, sampled-data, model validation are applied to a radiant heat experiment as mentioned in this paper, where a continuous time robust control model, including unknown perturbations and signals, using a discrete datum of finite length.
Proceedings ArticleDOI
Tuned l/sub 1/ identification from impulse response data: application to a fluid dynamics problem
TL;DR: Three tuned, convergent identification algorithms which compute a model and an l/sub 1/ error bound from the impulse response experimental data, are presented and are applied to the identification of a Taylor-Couette hydrodynamic instability process.
Proceedings ArticleDOI
Probabilistic bounds for model invalidation assessment
Wenguo Liu,Jie Chen +1 more
TL;DR: An additive uncertain model is considered, in which the modelling uncertainty is characterized in time domain by the l/sub 1/ induced system norm, and the probability for no uncertainty to exist that may satisfy the prescribed bound and match the input-output measurements is computed.
References
More filters
Analysis of feedback systems with structured uncertainties
TL;DR: In this article, a general approach for analysing linear systems with structured uncertainty based on a new generalised spectral theory for matrices is introduced, which naturally extend techniques based on singular values and eliminate their most serious difficulties.
Proceedings ArticleDOI
Structured uncertainty in control system design
TL;DR: This paper reviews control system analysis and synthesis techniques for robust performance with structured uncertainty in the form of multiple unstructured perturbations and parameter variations in the case where parameter variations are known to be real.
Journal ArticleDOI
Control oriented system identification: a worst-case/deterministic approach in H/sub infinity /
TL;DR: The authors formulate and solve two related control-oriented system identification problems for stable linear shift-invariant distributed parameter plants, each involving identification of a point sample of the plant frequency response from a noisy, finite, output time series obtained in response to an applied sinusoidal input.