Loading...
Please wait, while we are loading the content...
Nonparametric identification
| Content Provider | Scilit |
|---|---|
| Author | Bosch, P. P. J. Van Den Klauw, A. C. Van Der |
| Copyright Year | 2020 |
| Description | This chapter discusses the estimation of the parameters of nonparametric models. It also discusses three methods of identification of the frequency function: frequency response analysis, correlation analysis and Fourier analysis. An estimate of the impulse response of a process can be obtained by correlation analysis. The basic idea is to suppress the noise term by correlating the data with appropriate signals. A computationally simple method to obtain an estimate of the frequency function is frequency response analysis or sine-wave testing. The graphical approach to frequency response analysis is a noise-sensitive method. The chapter describes two frequency-domain identification techniques, that both have the disadvantage of long experiment periods, and requiring sinusoidal inputs. The method is based on discrete Fourier analysis of input and output signals. The resulting estimate of the frequency function is called the Empirical Transfer Function Estimate (ETFE). The ETFE can be improved by making use of the uncorrelatedness of the observations at different frequencies. Book Name: Modeling, Identification and Simulation of Dynamical Systems |
| Related Links | https://content.taylorfrancis.com/books/download?dac=C2006-0-15802-X&isbn=9780138746872&format=googlePreviewPdf |
| Ending Page | 88 |
| Page Count | 16 |
| Starting Page | 73 |
| DOI | 10.1201/9780138746872-4 |
| Language | English |
| Publisher | Informa UK Limited |
| Publisher Date | 2020-12-17 |
| Access Restriction | Open |
| Subject Keyword | Book Name: Modeling, Identification and Simulation of Dynamical Systems Automotive Engineering Identification Models Noise Input Nonparametric Etfe Fourier Frequency Function |
| Content Type | Text |
| Resource Type | Chapter |
