2020•IEEE Control SystemsRequires access

Iterative Model Identification of Nonlinear Systems of Unknown Structure: Systematic Data-Based Modeling Utilizing Design of Experiments

Patrick Schrangl, Pavlo Tkachenko, Luigi del Re

Open publisher page 34 citations

Abstract

High-quality models are essential to the performance of many control-related tasks [1]-[3]. If the structure of the system is known, first principle models can be created (which constitutes the best choice for most uses), especially if they should be used as design tools for parametric studies without having to build the corresponding hardware. However, first principle modeling is hardly possible for many real systems, either because the detailed knowledge of the system structure is not available or the model would be too complex to be useful for control design or to be parameterized. It has become common to use data-driven models, that is, correctly reproducing the input-output behavior of the system without trying to correctly describe its physics. For linear systems, data-driven modeling has been intensively studied, and powerful tools exist [4].

About this research paper

What this paper is about

High-quality models are essential to the performance of many control-related tasks [1]-[3]. If the structure of the system is known, first principle models can be created (which constitutes the best choice for most uses), especially if they should be used as design tools for parametric studies without having to build the corresponding hardware. However, first principle modeling is hardly possible for many real systems, either because the detailed knowledge of the system structure is not available or the model would be too complex to be useful for control design or to be parameterized. It has become common to use data-driven models, that is, correctly reproducing the input-output behavior of the system without trying to correctly describe its physics. For linear systems, data-driven modeling has been intensively studied, and powerful tools exist [4].

Why it matters

OpenAlex reports 34 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

High-quality models are essential to the performance of many control-related tasks [1]-[3]. If the structure of the system is known, first principle models can be created (which constitutes the best choice for most uses), especially if they should be used as design tools for parametric studies without having to build the corresponding hardware. However, first principle modeling is hardly possible for many real systems, either because the detailed knowledge of the system structure is not available or the model would be too complex to be useful for control design or to be parameterized. It has become common to use data-driven models, that is, correctly reproducing the input-output behavior of the system without trying to correctly describe its physics. For linear systems, data-driven modeling has been intensively studied, and powerful tools exist [4].

Key concepts: Computer science, System identification, Identification (biology), Control engineering, Parametric statistics, Parameterized complexity, Parametric model, Nonlinear system

Related papers

Back to paper searchBrowse research topicsOriginal source
Iterative Model Identification of Nonlinear Systems of Unknown Structure: Systematic Data-Based Modeling Utilizing Design of Experiments — Research Paper | ScholarLens