• 作者： 林繼雄；林理智
• 作者服務機構： 國立成功大學化學系
• 中文摘要： 本論文提出一新光譜數?分析法，可用以求出完全未知內容的多成分混合物中之成分組成。此項定量分析，由對高度情報空間中的光譜向量之相對關係之探討而達成。以ASI法分離所得的最適成分光譜向量，用?本法中之暫設成分向量。在預備估量中，所得的不合理組成?，用以轉動這些成分向量。結果可得較佳成分光譜及合理的近似組成?。文中亦描述一完整的電腦程式，此程式單由混合物光譜之輸入，則自動執行一切所需的判斷與演算，而後輸出混合物內各成分物質之光譜圖及其組成
• 英文摘要： A new numerical spectrophotometric method for estimating the composition of multi-component mixtures of totally unknown nature is presented. The estimation is made by studying the relationship among spectral vectors in the high-dimensional information space. Optimal spectral vectors of component substances are first found using the ASI (Automated Spectral Isolation) technique, and are used as temporary basis for the vector analysis. Irrational composition data obtained in the initial estimation is utilized to shift those optimal vectors. Better spectral shapes and composition are resulted from this operation. A computer program which outputs both the spectra of component substances and the composition data from the input of a set of mixture spectra is also described.
• 中文關鍵字： --
• 英文關鍵字： --