Multiple nonlinear parameter estimation using PI feedback control

Lith, Pascal F. van; Witteveen, Heidi J.B.; Betlem, Bernardus H.L.; Roffel, Brian

doi:10.1016/s0967-0661(01)00010-7

Cited by 10 publications

(6 citation statements)

References 7 publications

(5 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It is noted that the state space can be multi-dimensional depending on the embedding dimension m*. Once the input and state spaces are both discretized, they can be combined to form the discretized augmented input space Ã ¼ { (1) , (2) …”

Section: Remarkmentioning

confidence: 99%

“…Several data-driven techniques have been reported in literature for fault detection and health monitoring in dynamical systems, which include statistical linearization [1], Kalman filtering [2], unscented Kalman filtering (UKF) [3,4], particle filtering (PF) [5], Markov chain Monte Carlo (MCMC) [6], Bayesian networks [7], artificial neural networks (ANN) [8], maximum likelihood estimation (MLE) [9], wavelet-based tools [10], and genetic algorithms (GA) [11]. However, fault detection in single components is only a small part of the health monitoring problem in its entirety.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Symbolic identification for fault detection in aircraft gas turbine engines

Chakraborty

Sarkar

Ray

2011

Proceedings of the Institution of Mechanical Engineers, Part G:

View full text Add to dashboard Cite

This article presents a robust and computationally inexpensive technique of component-level fault detection in aircraft gas-turbine engines. The underlying algorithm is based on a recently developed statistical pattern recognition tool, symbolic dynamic filtering (SDF), that is built upon symbolization of sensor time series data. Fault detection involves abstraction of a language-theoretic description from a general dynamical system structure, using state space embedding of output data streams and discretization of the resultant pseudo-state and input spaces. System identification is achieved through grammatical inference based on the generated symbol sequences. The deviation of the plant output from the nominal estimated language yields a metric for fault detection. The algorithm is validated for both singleand multiple-component faults on a simulation test-bed that is built upon the NASA C-MAPSS model of a generic commercial aircraft engine.

show abstract

Section: Remarkmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Symbolic identification for fault detection in aircraft gas turbine engines

Chakraborty

Sarkar

Ray

2011

Proceedings of the Institution of Mechanical Engineers, Part G:

View full text Add to dashboard Cite

show abstract

“…In such cases, parameter estimation approaches can be used. In this work, a PIestimator will be used (Van Lith, Witteveen, Betlem & Roffel, 2001). The PI-estimator is structurally similar to a PI-feedback controller.…”

Section: Hybrid Modelingmentioning

confidence: 99%

Combining prior knowledge with data driven modeling of a batch distillation column including start-up

Lith

Betlem

Roffel

2003

Computers & Chemical Engineering

View full text Add to dashboard Cite

show abstract

“…These rates are nonlinear and time varying. One could use simple PIfeedback control techniques [9] or state estimation approaches such as Kalman filtering to obtain parameter estimates. An extended Kalman filter was designed and two additional state equations were introduced; one for and one for q p .…”

Section: Subprocess Behavior Estimationmentioning

confidence: 99%

“…The optimization results were comparable with the results obtained with Eq. (9). The gradient information that the optimization algorithm uses indicates the importance of these parameters (and thus of the parameters of the fuzzy model) with respect to the states.…”

Section: Submodel Integrationmentioning

confidence: 99%

A structured modeling approach for dynamic hybrid fuzzy-first principles models

Lith

Betlem

Roffel

2002

Journal of Process Control

View full text Add to dashboard Cite

Hybrid fuzzy-first principles models can be attractive if a complete physical model is difficult to derive. These hybrid models consist of a framework of dynamic mass and energy balances, supplemented with fuzzy submodels describing additional equations, such as mass transformation and transfer rates. In this paper, a structured approach for designing this type of model is presented. The modeling problem is reduced to several simpler problems, which are solved independently: hybrid model structure and subprocess determination, subprocess behavior estimation, identification and integration of the submodels to form the hybrid model. The hybrid model is interpretable and transparent. The approach is illustrated using data from a (simulated) fed-batch bioreactor.

show abstract

Multiple nonlinear parameter estimation using PI feedback control

Cited by 10 publications

References 7 publications

Symbolic identification for fault detection in aircraft gas turbine engines

Symbolic identification for fault detection in aircraft gas turbine engines

Combining prior knowledge with data driven modeling of a batch distillation column including start-up

A structured modeling approach for dynamic hybrid fuzzy-first principles models

Contact Info

Product

Resources

About