A novel ARX-based discretization method for linear non-rational systems
Apstrakt
This paper presents a novel, simple, flexible and effective discretization method for linear
non-rational systems including arbitrary linear fractional order systems (LFOS). The
discretization algorithm relies on the direct integration in the complex domain and
application of ARX (AutoRegressive eXogenous) model. Parameters of ARX-model are
obtained by numerical inversion of Laplace transform from the set of input/output data
from recorded step response to model of non-rational system. Numerical simulations of
several representatives of LFOS (e.g. fractional order PID controller, fractional
logarithmic filter, fractional oscillator etc.) are used to demonstrate the effectiveness of
the proposed discretization method, both in the time and frequency domains. The
obtained results indicate that the proposed ARX-based discretization method is adequate
technique for obtaining digital approximation of LFOS.
Ključne reči:
discretization / fractional order systems / AutoRegressive eXogenous model / model reduction / frequency and time domainIzvor:
Proceedings of International Conference on Fractional Differentiation and its Application ICFDA16, 18-20 July 2016, Novi Sad, Serbia, 2016, 343-352Izdavač:
- Belgrade: Serbian Society of Mechanics
- Faculty of Technical Sciences Novi Sad
Finansiranje / projekti:
- Održivost i unapređenje mašinskih sistema u energetici i transportu primenom forenzičkog inženjerstva, eko i robust dizajna (RS-MESTD-Technological Development (TD or TR)-35006)
- Povećanje energetske efikasnosti HE i TE EPS-a razvojem tehnologije i uređaja energetske elektronike za regulaciju i automatizaciju (RS-MESTD-Technological Development (TD or TR)-33020)
- Inteligentni sistemi upravljanja klimatizacije u cilju postizanja energetski efikasnih režima u složenim uslovima eksploatacije (RS-MESTD-Technological Development (TD or TR)-33047)
- Inteligentni nadzorno upravljački sistem za rano otkrivanje i eliminaciju neželjenih stanja i promena na uređajima, opremi i procesima procesne industrije (RS-MESTD-Technological Development (TD or TR)-32018)
- Razvoj inteligentnog nadzorno upravljačkog sistema za povećanje energetske efikasnosti zgrada (RS-MESTD-Technological Development (TD or TR)-33013)
Kolekcije
Institucija/grupa
Mašinski fakultetTY - CONF AU - Bošković, Marko AU - Šekara, Tomislav AU - Rapaić, Milan AU - Lazarević, Mihailo AU - Mandić, Petar PY - 2016 UR - https://machinery.mas.bg.ac.rs/handle/123456789/4187 AB - This paper presents a novel, simple, flexible and effective discretization method for linear non-rational systems including arbitrary linear fractional order systems (LFOS). The discretization algorithm relies on the direct integration in the complex domain and application of ARX (AutoRegressive eXogenous) model. Parameters of ARX-model are obtained by numerical inversion of Laplace transform from the set of input/output data from recorded step response to model of non-rational system. Numerical simulations of several representatives of LFOS (e.g. fractional order PID controller, fractional logarithmic filter, fractional oscillator etc.) are used to demonstrate the effectiveness of the proposed discretization method, both in the time and frequency domains. The obtained results indicate that the proposed ARX-based discretization method is adequate technique for obtaining digital approximation of LFOS. PB - Belgrade: Serbian Society of Mechanics PB - Faculty of Technical Sciences Novi Sad C3 - Proceedings of International Conference on Fractional Differentiation and its Application ICFDA16, 18-20 July 2016, Novi Sad, Serbia T1 - A novel ARX-based discretization method for linear non-rational systems EP - 352 SP - 343 UR - https://hdl.handle.net/21.15107/rcub_machinery_4187 ER -
@conference{ author = "Bošković, Marko and Šekara, Tomislav and Rapaić, Milan and Lazarević, Mihailo and Mandić, Petar", year = "2016", abstract = "This paper presents a novel, simple, flexible and effective discretization method for linear non-rational systems including arbitrary linear fractional order systems (LFOS). The discretization algorithm relies on the direct integration in the complex domain and application of ARX (AutoRegressive eXogenous) model. Parameters of ARX-model are obtained by numerical inversion of Laplace transform from the set of input/output data from recorded step response to model of non-rational system. Numerical simulations of several representatives of LFOS (e.g. fractional order PID controller, fractional logarithmic filter, fractional oscillator etc.) are used to demonstrate the effectiveness of the proposed discretization method, both in the time and frequency domains. The obtained results indicate that the proposed ARX-based discretization method is adequate technique for obtaining digital approximation of LFOS.", publisher = "Belgrade: Serbian Society of Mechanics, Faculty of Technical Sciences Novi Sad", journal = "Proceedings of International Conference on Fractional Differentiation and its Application ICFDA16, 18-20 July 2016, Novi Sad, Serbia", title = "A novel ARX-based discretization method for linear non-rational systems", pages = "352-343", url = "https://hdl.handle.net/21.15107/rcub_machinery_4187" }
Bošković, M., Šekara, T., Rapaić, M., Lazarević, M.,& Mandić, P.. (2016). A novel ARX-based discretization method for linear non-rational systems. in Proceedings of International Conference on Fractional Differentiation and its Application ICFDA16, 18-20 July 2016, Novi Sad, Serbia Belgrade: Serbian Society of Mechanics., 343-352. https://hdl.handle.net/21.15107/rcub_machinery_4187
Bošković M, Šekara T, Rapaić M, Lazarević M, Mandić P. A novel ARX-based discretization method for linear non-rational systems. in Proceedings of International Conference on Fractional Differentiation and its Application ICFDA16, 18-20 July 2016, Novi Sad, Serbia. 2016;:343-352. https://hdl.handle.net/21.15107/rcub_machinery_4187 .
Bošković, Marko, Šekara, Tomislav, Rapaić, Milan, Lazarević, Mihailo, Mandić, Petar, "A novel ARX-based discretization method for linear non-rational systems" in Proceedings of International Conference on Fractional Differentiation and its Application ICFDA16, 18-20 July 2016, Novi Sad, Serbia (2016):343-352, https://hdl.handle.net/21.15107/rcub_machinery_4187 .