本論文旨在建立一套光電檢測基重系統,此技術開發以非接觸、線上、多點區域、穿透式與驗證之技術,藉以改善量測不織布準確性和製程能力精密度。基本上,不織布基重系統由光源發射器穿透不織布,光傳輸平行於光接收器,利用比爾-朗伯定律,經訊號轉換輸出入介面裝置。 我們首先架設並進行光電檢測系統調整補償,實驗結果顯示經溫度補償後之系統相當穩定,再利用最小平方法找到一可靠有效的電壓值與基重之參數關係方程式,經使用已知不織布基重之量測,結果顯示所量測獲得之基重值與已知值相符。找出適合電壓、基重參數關係方程式,藉由選取三種原料與不同組織,再利用標準試片的實驗模型,求轉換關係方程式,並進行找出適合方程式,並求修正方程式,接著製程精密度與驗證,提昇生產製程穩定性和預測準確度,且有效改善品質與減少基重均勻性的不良率,針對織布檢測使用監督式學習演算法與分段式的方程式,並準確率誤差的比較,進而提出有效自動監督式學習演算法,找出準確量測值。 經由光學近紅外線檢測系統得到光訊號轉換電壓訊號,求修正的關係方程式,再進行基重檢測,並求出適合指數型、連續式最小平方法、分段式最小平方法、監督式學習法、修正方程式等,經求出適合方程式與驗證。其實驗結果監督式學習法,可達良好線上測試,接著定義製程品質能力分析與常態分佈,可降低不織布原料的成本,不織布生產製程速度在120 m/min以下,以1000組樣本,經實驗製程精密度Cp 是1.66,並滿足其條件。因此,應用監督學習法可有效的提升不織布製程之生產能力品質穩定性。 ;The purpose of this thesis study is to develop a method of optical transmission inspection of the basis weight on-line, by combining the modified least squares and optical processing technique. A near infrared light transmission inspection is applied production of equipment nonwoven fabrics to detect the basis weight and support the producing quality. Using least squares method, the parameter transfer equations of the piecewise polynomials functions between the measured voltage and the nonwoven basis weight are found. Supervised learning method is adopted to improve the producing capability. Obvious, the equations and supervised learning method is effective to improve measures the range, producing capability and support the producing quality. This process is developed to significantly target toward improving the mass quality analysis of the nonwoven material. The real-time scanning width piecewise least squares method and area-based strategy for determining based on the process quality of nonwoven manufacturing. To avoid the influence of ambient factors, the compensation controls device are adopted and successfully showed. Subsequently, the modified least squares method is used to obtain the suitable parameter transformation between the measured voltage and the nonwoven fabrics basis weight. The piecewise least squares method was obtained as the parameter transfer equation. We consider estimating and testing Cp with the presence of on-line basis weight measurement errors. To obtain the true process precision Cp are presented to practitioners for their factory applications. In this study, a NIR transmission-based inspection for the basis weight of to improve quality production process, to avoid production flaws and to reduce the production costs are the major issues of the manufacturing industry. The apparatus basically consists of a light emitter mounted parallel to a light receiver. The light is emitted from the light emitter. Residual light is received by the receiver after being transmitted through the nonwoven fabric. An equation acquired by using the Beer–Lambert law, the parameter transfer equations of the equation functions between the measured voltage and the nonwoven basis weight are found. Optical inspection techniques can be also used in which the optical modulated to find a modified equation, then obtain the non-woven basis weight inspected on-line and verified by quality capability of process. A modified equation that can be used to reduce the uniformity and decrease the basis weight density. The potential of an optical sensor with increased sensitivity the range for finding the equations, near infrared light detecting the basis weight for a nonwoven material, to predict quality capability of nonwoven fabrics. In the proposed algorithm the supervised learning for finding polynomials the equations between the measured voltage and the nonwoven basis weight are found, the error deviation inspection works online and accuracy prediction. The verification accuracy prediction has been conducted to illustrate the performance of the proposed inspection algorithm by a dynamic of experiments nonwoven fabrics.Parameter transfer equations, is adopted to improve the producing capability. It is shown that the capability index of process Cp is over 1.66 under 1000 testing samples when the supervised learning algorithm is used.