主成分分析法與LabVIEW機器視覺功能模擬酸鹼指示劑吸收光譜圖的開發與研究

Abstract

本研究收集了不同酸鹼值下廣用指示液的吸收光譜和對應的顏色數據，這些顏色數據包括添加指示液後液體樣品的顏色以及樣品經過廣用試紙測試後的顏色。使用LabVIEW編寫資料處理程式，通過主成分分析得到光譜數據的主成分和特徵值，並利用線性迴歸模型分析主成分與顏色數據之間的關聯性，最終實現了根據顏色推斷廣用指示液吸收光譜的目標。當將添加了廣用指示液的未知酸鹼樣品顏色數據匯入程式後，模擬與實際吸收光譜的均方根誤差在2%以內，重疊率均在96%以上；若以廣用試紙的顏色進行計算，均方根誤差在4%以內，重疊率約為90%。使用碳酸鈉溶液作為無色未知樣品，通過指示液和試紙的顏色推斷的吸收光譜，均方根誤差約為1.2%，重疊率約為97%。將茜素溶液作為有色樣品進行測試，結果顯示均方根誤差在5%以內，重疊率在80%以上。這些數據表明，本研究的方法能夠僅憑顏色數據模擬出吸收光譜，且結果具有較高的準確性。

In this study, the absorption spectra and corresponding color data of the universal indicator solution at different pH values were collected, which included the color of the liquid samples after the addition of the indicator solution as well as the color of the samples after testing on pH test strips. The data processing program was written in LabVIEW, and the principal components and eigenvalues of the spectral data were obtained through principal component analysis, and the association between the principal components and the color data was analyzed by linear regression model, which finally realized the goal of inferring the absorption spectrum of the universal indicator solution based on the color.When the color data of the unknown acid-alkali samples spiked with the universal indicator solution were imported into the program, the root-mean-square error between the simulated and actual absorption spectra was within 2%, and the overlap rate was over 96%. When the color of the pH test strips was used in the calculation, the root-mean-square error was within 4%, and the overlap rate was about 90%. Using sodium carbonate solution as a colorless unknown sample, the absorption spectra deduced from the color of the indicator solution and the test strip had a root mean square error of about 1.2% and an overlap rate of about 97%. Alizarin solution was tested as a colored sample and the results showed that the root mean square error was within 5% and the overlap rate was over 80%. These data indicate that the method in this study can simulate the absorption spectra based on the color data only, and the results have a high accuracy.