Please use this identifier to cite or link to this item: http://hdl.handle.net/11455/1534
標題: 類神經網路運用於型材矯直之研究
A Study of Profiles Material Straightness by Applying Artificial Neural Networks
作者: 楊景欽
yang, jimching
關鍵字: straighten;矯直;neural networks;error-back-propagation method;finite element method;類神經網路;誤差逆傳遞法;有限元素法
出版社: 機械工程學系
摘要: 
本研究採用類神經網路之誤差逆傳遞法,配合有限元素模擬分析結果,建立型材矯直預測系統,以達到只須於矯直前量測型材平直度就可藉型材矯直預測系統,預測得知矯直位置、方向、衝頭下壓量,並於矯直後達到一定平直度精度。此方法是先以有限元素模擬矯直過程並將分析結果做為類神經網路訓練樣本,以建立類神經網路型材矯直預測系統。本研究中實際將型材矯直預測系統運用於半自動化壓機矯直機,以驗證說明本研究所提出類神經網路運用於型材矯直預測可行性。本研究所嘗試使用類神經網路運用於型材矯直預測,其結果應可提供業界在型材矯直應用開發上的參考。

The application of artificial neural networks of error-back-propagation method and finite element method for simulation results of straightening process are to build prediction system of profile materials straighten. This system used to measure straightness of profile materials can obtain the straightening position, direction, and punch travel and profile materials finally reach required straightness. This system serves results of finite element method simulated straighten process as training patterns for neural networks and build prediction system of profile materials straighten. Actually use this system on straighten machine of automation to straighten profile materials and verify mention of neural networks that could be available for predict process of profile materials straighten. This study attempted to apply neural networks to predict process of profile materials straighten and the results can be provide the reference to develop of profile materials straighten in industry.
URI: http://hdl.handle.net/11455/1534
Appears in Collections:機械工程學系所

Show full item record
 
TAIR Related Article

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.