類神經網路應用於線切割放電加工厚度變化工件之研究
Date Issued
2004
Date
2004
Author(s)
李其政
DOI
zh-TW
Abstract
An algorithm to estimate the stable feed so as to control Wire-EDM process and prevent wire from breaking under varying workpiece height conditions is proposed and implemented. A neural network model related feed rate with workpiece height and relevant machining parameters in Wire-EDM process is established. With this model appropriate feed rate can be computed in accordance with the identified workpiece height and known discharge parameters. Pulse off time is increased immediately to prevent wire from breaking when the process becomes unstable and the rate of abnormal sparks is higher than the predetermined limit. It is reduced gradually once the rate of abnormal sparks is reduced to a safe value. Experiments have shown that satisfactory results are achieved.
Subjects
不正常放電比
間隙電壓
放電休止時間
類神經網路
artificial neural network
pluse off time
rate of abnormal sparks
Type
thesis
File(s)![Thumbnail Image]()
Loading...
Name
ntu-93-R91522821-1.pdf
Size
23.53 KB
Format
Adobe PDF
Checksum
(MD5):489091ac570d5005df92cb7152b5f103