模糊邏輯模式應用於CVD磊晶沉積製程模式化之研究

Abstract

本論文研究模糊邏輯模式應用於CVD磊晶沉積製程之模式化,其作法乃是 先由訓練資料配合資料群估測法和逆迴遞演算來建立數個模式架構,再根 據以測試資料之互相關係數為基礎的決策法則來求得模糊模式的最佳架 構,藉由這最佳架構並結合訓練和測試資料,再以逆迴遞演算法對最後模式 的參數做微調.我們將此一演算法應用於多種非線性函數和化學氣相沉積 之製程模擬,而從得到的結果充分說明本論文所用模糊邏輯模式,相較於已 提出的模糊邏輯模式和類神經網路模式,有更好的準確性和時效性. A CVD epitaxial deposition process modeling using fuzzy logic models has beenproposed. The algorithm starts with cluster estimation method and back propagationalgorithm to construct a number of modeling structures from the training data.A decision rule based on the multiple correlation factor is used to obtain theoptimum structure of the fuzzy model using the testing data. Upon the optimumstructure has been reached, the gradient-descent method is used to refine theparameters of the final fuzzy model using both training and testing data. Thealgorithm has been applied to various nonlinear functions and a verticl chemicalvapor deposition process. The results demonstrate the efficiency and effective-ness of the proposed fuzzy logic model in comparison with existing fuzzy logicmodels and artificial neural network models.