查詢結果分析
相關文獻
- An Artificial Neural Network System Process Control
- 乾膜光阻生產製程中塗佈作業之ANN-SPC-EPC製程管制系統建立
- 應用類神經網路於球閘陣列基板之自動化表面瑕疵檢測
- Neural Network Procedures for Taguchi's Dynamic Problems
- A Fast and Efficient Competitive Learning Design Algorithm Based on Weight Vector Training in Transform Domain
- 專家系統振動訊號圖型判別之研究
- 反傳遞模糊類神經網路於流量推估之應用
- 類神經網路(Neural Networks)的種類及其在影像處理上的應用
- C++Fuzzy類神經網路物件導向發展系統之建立
- 臺灣汽保費率之估計--對數線性費率模式與類神經網路之比較
頁籤選單縮合
題 名 | An Artificial Neural Network System Process Control=類神經網路於製程控制 |
---|---|
作 者 | 陳智瑋; 黃祥熙; | 書刊名 | 國立屏東科技大學學報 |
卷 期 | 9:1 2000.03[民89.03] |
頁 次 | 頁43-58 |
分類號 | 448.94 |
關鍵詞 | 類神經網路; 屬性分類; 變數控制; 線上製程控制; 視覺系統; Artificial neural network; Attribute classification; Variable control; On-line process control; Visual system; |
語 文 | 英文(English) |
中文摘要 | 一個類神經網路雛型系統已經開發成功且應用於製程控制,此系統包含視覺系統 及類神經網路分類器。該視覺系統乃用來獲得類神經網路分類器之輸入資料,該系統成功的 將金屬元件的邊界轉換數字,並以矩陣方式輸入類神經網路分類器之輸入單元。此視覺系統 的實施提供了製造環境中減少人為干預的一個可行方案,主要乃是人為疏忽通常是生產系統 中最不容易控制的因素。 本研究展示類神經網路系統於製程控制的可行性,探討內容包含屬性分類及變數控制。 類神經網路系統經過訓練,已經成功的辨識訓練過及未訓練過之資料,即使已經遭受扭曲之 資料,該類神經網路系統亦能高準確的辨識。 |
英文摘要 | A prototype artificial neural network (ANN) system consisting of a visual system (VS) and an ANN classifier for process control has been successfully developed. The VS was used for obtaining inputs data for the ANN classifier. The implementation of the VS suggests a feasible alternative for reducing the amount of human intervention in a manufacturing environment because human errors are usually the most uncontrollable factors in a production system. This study also demonstrates the feasibility of using the ANN classifier in the domain of process control. The implementation of process control includes two major aspects, which are attribute classification and variable control. The ANN shows that the classification tasks can be accomplish. For attribute classification, the ANN classifier shows the capability of classifying trained patterns (parts). In the domain of variable control, the ANN also shows the capability of classifying trained and partially untrained unnatural patterns. However, the structures of the ANN including the numbers of input units, hidden units, and output units, and random seeds selected are very important criteria and must be considered thoughtfully for successful implementation. |
本系統中英文摘要資訊取自各篇刊載內容。