Neural Network Prediction of Segment Wear in Stone Sawing

Abstract:

Article Preview

An investigation was carried out to predict the wear performance of diamond sawblade segments by using a Levenberg-Marquardt backpropagation (BP) neural network. The wear of the diamond segments were measured in circular sawing of natural gray granite in order to train the network and examine its validation. Since the depth of cut and workpiece speed are two main variables in the sawing of a specific granite material with a fixed diamond sawblade, a 2-5-1 structure of BP network was found to be capable of predicting the wear performance. In spite of the limited experimental data, the average value of relative errors between the simulated and measured results was found to be around 10%. Since experimentally measuring of segment wear is a time-consuming job, the trained network was also used to predict the wear performance under a very wide range of operating parameters, which can provide a useful guideline for the optimization of stone sawing.

Info:

Periodical:

Materials Science Forum (Volumes 471-472)

Edited by:

Xing Ai, Jianfeng Li and Chuanzhen Huang

Pages:

485-489

DOI:

10.4028/www.scientific.net/MSF.471-472.485

Citation:

X. Xu and Y.F. Zhang, "Neural Network Prediction of Segment Wear in Stone Sawing", Materials Science Forum, Vols. 471-472, pp. 485-489, 2004

Online since:

December 2004

Authors:

Export:

Price:

$35.00

In order to see related information, you need to Login.

In order to see related information, you need to Login.