Spectrogram data reanalysis

1．foreword

When using chromatography or wavelength technology to detect mixture samples, a series of spectra are usually obtained. Depicting the data in the coordinate system is a curve. Through the comparison with the standard sample, the content of a substance in the mixture can be obtained, and the detection is over. However, this group of spectrogram data is the embodiment of some properties of the mixture, which is related to some properties of the mixture (such as viscosity, density, etc.). According to this principle, when testing the sample spectrogram, we can test some properties of the mixture (such as viscosity, density, etc.) at the same time. After such multiple tests, a large number of spectrogram data and property data can be obtained, So as to obtain many data samples related to them.

These samples are used to create a mathematical relationship algorithm between spectral data and properties. After the detection of sample spectral data, the property value can be calculated by the algorithm.

2．principle

、

Figure 1 Schematic diagram

2.1. In the dotted line, the same sample is subjected to spectral analysis and property analysis respectively, the analysis data is matched into one sample, and multiple samples are associated into an algorithm that uses spectral data to calculate properties.

2.2. After algorithm is established, the spectrum of similar samples can be analyzed. The content of the mixture can be calculated from the spectrum data, and the properties of the mixture can be calculated by the algorithm (the red part in Figure 1).

3．Summarize

Usually the spectral analysis is clean and simple, and the original property analysis needs to use bottles and jars to operate tediously. Through the above process, when using the spectral analysis to do the content of the mixture, the properties of the mixture are also analyzed, which can replace the original properties testing. Analysis is convenient and quick.