izpis_h1_title_alt

Identifikacija dinamičnih sistemov z globokimi Gaussovimi procesi
Jančič, Mitja (Author), Govekar, Edvard (Mentor) More about this mentor... This link opens in a new window, Kocijan, Juš (Co-mentor)

.pdfPDF - Presentation file, Download (6,72 MB)
MD5: 4B446F743B0F60919D2719EA09F5C590

Abstract
Zaradi naraščajoče kompleksnosti obravnavanih sistemov in posledično zahtevnega matematičnega modeliranja, v praksi pogosto uporabimo empirične modele ali modele črne škatle, s katerimi modeliramo le povezave med vhodno-izhodnimi vrednostmi, ne pa tudi fizikalnih zakonitosti, ki se jim sistem podreja. Za modeliranje oziroma identifikacijo zveze med vhodnimi in izhodnimi vrednostmi sistema se uporabljajo tudi globoki Gaussovi procesi. Ti za opis kompleksnejših procesov uporabljajo gnezdenje in hierarhično strukturo. Z identificirano zvezo med vhodno-izhodnimi vrednostmi z uporabo Gaussovih procesov lahko za dane vhodne vrednosti napovemo vrednost izhoda in pripadajočo negotovost, kar lahko s pridom uporabimo. V okviru magistrskega dela predstavimo teoretične osnove modeliranja z globokimi Gaussovimi procesi in njihove prednosti. V ta namen v ilustrativnem primeru uspešno identificiramo dinamični sistem nelinearnega nihanja mase, v bolj praktičnem primeru pa obravnavamo bistveno kompleksnejši sistem napovedovanja temperature v prizemni plasti atmosfere.

Language:Slovenian
Keywords:identifikacija dinamičnih sistemov, globoki Gaussovi procesi, autoregresivni model, enokoračna napoved, neparametričen model, nelinearni sistem
Work type:Master's thesis/paper (mb22)
Organization:FS - Faculty of Mechanical Engineering
Year:2017
Views:1033
Downloads:565
Metadata:XML RDF-CHPDL DC-XML DC-RDF
 
Average score:(0 votes)
Your score:Voting is allowed only to logged in users.
:
Share:AddThis
AddThis uses cookies that require your consent. Edit consent...

Secondary language

Language:English
Title:Identification of dynamic systems using deep Gaussian Processes
Abstract:
Mathemathical and physical modelling only provide approximate description of the true nature of a dynamic system. The higher the precision of the model the more likely it becomes analytically intractable and, therefore, empirical models or black box models are used. When dynamic systems are considered as black box models, almost no prior knowledge about the system is considered. Deep Gaussian Processes, which use hierarchical structure to provide adequate identification of very complex systems, can be used to identify the mapping between the system input and output values. With the given mapping function we can then provide a one-step ahead prediction of the system output values, together with its uncertainty, which can be advantageously used. In this paper we use deep Gaussian Processes to identify a dynamic system and present its advantages by studying two cases. In the first illustrative case we successfully identify the dynamic properties of a nonlinear oscillating mass, while in the second, more realistic and complex case, we study one-step ahead prediction of air temperature in the atmospheric surface layer.

Keywords:identification of dynamic systems, deep Gaussian Processes, one-step ahead prediction, nonparametric model, nonlinear systems

Similar documents

Similar works from RUL:
Similar works from other Slovenian collections:

Comments

Leave comment

You have to log in to leave a comment.

Comments (0)
0 - 0 / 0
 
There are no comments!

Back