A correlation study between the chromium returns-ratio and the content of silicon/aluminum for 15-5PH(V) in the electric are furnace using the Artificial Neural Network method
| Autor: | 薛彬佑 |
|---|---|
| Rok vydání: | 2011 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 99 This study applies the artificial neural network (ANN) to analyze the influence on the recovery of silicon and aluminum components of high and low levels after the melting of 15-5PH(V) in EAF to look for the best recovery of chromium. First, the experiment is to take scrap to measure chrome content before EAF’s melting. After the melting, the recovery is achieved by measuring chrome in steel water, and the experimental data are trained by using Back propagation (BP). Also, the content of silicon and aluminum are used as variables to obtain the best model. The accuracy of ANN is confirmed by using root-mean-square (RMS) and mean relative error to reach the best chrome recovery by predicting the control value of silicon and aluminum components. The result shows that by using ANN, the RMS is 1.51%, and the mean relative error is 1.43%, which proves that the ANN model can predict the recovery effectively. Moreover, the best chrome recovery during the EAF process can be achieved through controlling the content of silicon. Keywords: Artificial neural network, Electric arc furnace, Root mean square error, Mean relative error |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|