diplomska naloga
Nataša Zavrtanik (Avtor), Goran Turk (Mentor), Dejan Zupan (Komentor)

Povzetek

Uporaba umetnih nevronskih mrež za oceno trdnosti lesenih elementov

Ključne besede

gradbeništvo;diplomska dela;UNI;Konstrukcijska smer;umetne nevronske mreže;raztros podatkov;karakteristike lesa;trdnost lesa;

Podatki

Jezik: Slovenski jezik
Leto izida:
Izvor: Ljubljana
Tipologija: 2.11 - Diplomsko delo
Organizacija: UL FGG - Fakulteta za gradbeništvo in geodezijo
Založnik: [N. Šinkovec]
UDK: 004:624.011.1:624.07(043.2)
COBISS: 3973985 Povezava se bo odprla v novem oknu
Št. ogledov: 2015
Št. prenosov: 920
Ocena: 0 (0 glasov)
Metapodatki: JSON JSON-RDF JSON-LD TURTLE N-TRIPLES XML RDFA MICRODATA DC-XML DC-RDF RDF

Ostali podatki

Sekundarni jezik: Angleški jezik
Sekundarni naslov: The use of artificial neural networks for timber strenght estimation
Sekundarni povzetek: In design of various ingeneering structures structural material together with proper planning and construction realisation is of key importance. Structural material should have satisfactory strength that enables a proper load carrying capacity. To determine the characteristics of wooden elements destructive and nondestructive methods are used. In destructive methods the tested materials are loaded until failure occurs. Artificial neural networks have been used in this diploma thesis to estimate the strength of structural timber. 293 wooden speciments were taken into consideration. For all specimens the density and seven elastic modules were measured by indestructive testing methods and the strength was based on the destructive method. 250 specimens were randomly chosen for the training of artificial networks and 43 were left for testing procedure. Five examples, in which different input data is used, were taken into closer consideration, the fifth example also considers different output data. For generating and learning of neural networks two programs were used: (i) a fortran code NRT2003 and (ii) code based on Neural Network Library in Matlab. The efficiency of various naural networks is compared through several statistical quantities. The effect of data dissipation is also considered and a method for data integration is presented.
Sekundarne ključne besede: artificial neural networks;data dissipation;characteristics of timber;strength of structural timber;
Vrsta datoteke: application/pdf
Vrsta dela (COBISS): Diplomsko delo
Komentar na gradivo: Univ. v Ljubljani, Fak. za gradbeništvo in geodezijo
Strani: X f., 81 str., priloge na priloženem CD
Vrsta dela (ePrints): thesis
Naslov (ePrints): Uporaba umetnih nevronskih mrež za oceno trdnosti lesenih elementov
Ključne besede (ePrints): gradbeništvo;diplomska dela;UNI;Konstrukcijska smer;umetne nevronske mreže;raztros podatkov;karakteristike lesa;trdnost lesa;
Ključne besede (ePrints, sekundarni jezik): artificial neural networks;data dissipation;characteristics of timber;strength of structural timber;
Ključne besede (ePrints, sekundarni jezik): artificial neural networks;data dissipation;characteristics of timber;strength of structural timber;
ID: 8311103