Naslov
Predicting the Tempering Curve of Tool Steel Using Neural Networks

Autori
Filetin, Tomislav ; Majetić, Dubravko ; Žmak, Irena

Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni

Izvornik
2. International Conference on Industrial Tools : ICIT'99 / Kuzman, Karl ; Balič, Jože - Maribor : TECOS Slovenian Tool and Die Development Centre, 1999, 424-427

Skup
2. International Conference on Industrial Tools

Mjesto i datum
Rogaška Slatina, Slovenija, 18.04.1999. - 22.04.1999

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
tool steels; neural network; prediction of tempering curve

Sažetak
The purpose of this paper is to demonstrate how a neural network approach is used for predicting steel properties in case when some of the relevant influence factors and their relations are unknown. An attempt has been made to establish a nonlinear statis discrete-time neuron model, the so-called Static Elemntary Processor (SEP). Based on SEP neuopns, a static multi layer forward neural network is proposed to predict s tempering curve from chemical composition (%C, CR, NI, MO, W, V, Co) and austenizing temperature. In the neural network learning procedure datasets with 18 tool steel grades with different compositions are used. The mean error between hardness data from catalogue and predicted hardness data as well as standard deviation for testing dataset (8 steel grades) are very small and acceptable. These preliminary results encourage further investigation.

Izvorni jezik
Engleski

Znanstvena područja
Strojarstvo

POVEZANOST RADA