上海佳实DH-9900蛋白仪的近红外分析算法解析

上海佳实电子科技有限公司研制的DH-9900蛋白检测系统的软件分析系统具有自学习的能力。

近红外（NIR）检测的优点：样品无需预处理，近红外区内光穿透深度大，使得近红外光谱技术可以用漫反射技术对样品直接测定，同时还具有分析具有非破坏性、分析速度快、远距离测定和实时分析、测定重现性好、适用的样品范围广、分析成本较低、对操作人员的要求较低。简而言之，使用方便，无需培训。

NIR用于物质成份的定量测定时，谱带较宽并且容易重叠，需要有化学计量学方法进行分析（主成分分析PCA、主成分回归PCR、多元线性回归、偏小二乘法PLS、ANN），应用多的为PCA、PLS。

DH-9900蛋白测量仪采用的算法是PCA和PLS，目前，上海佳实研发部的软件工程师正在设计多的测量算法如PCR, ANN等以升仪器，提。

测定固体样品的NIR光谱时，一般要测定样品不同面的光谱以减少测定误差并获得可靠的信息，有时可以进行光谱的重复测定以提光谱图的信噪比。在得到NIR光谱图后，一般还需要对光谱进行预处理（基线校正、一阶导数、二阶导数和正交信号校正）。

感谢广大客户的信赖和支持！

Analysis of Near Infrared Analysis Algorithm of Shanghai Jiashi DH-9900 Protein Meter

The software analysis system of DH-9900 protein detection system developed by Shanghai Jiashi Electronic Technology Co., Ltd. has self-learning ability.

Advantages of Near Infrared (NIR) Detection: The sample does not require pretreatment, and the depth of light penetration in the near-infrared region is large, so that near-infrared spectroscopy can directly measure the sample by diffuse reflection technology, and also has non-destructive analysis and analysis. Fast, long-distance measurement and real-time analysis, good reproducibility, wide range of applicable samples, low cost of analysis, and low operator requirements.

When NIR is used for the quantitative determination of material components, the bands are wide and easy to overlap, and chemometric methods are needed for analysis (principal component analysis PCA, principal component regression PCR, multiple linear regression, partial least squares PLS, ANN). The most widely used are PCA and PLS.

The algorithms used in the DH-9900 Protein Meter are PCA and PLS. Currently, software engineers in the R&D department are developing more measurement algorithms such as PCR, ANN, etc. to upgrade the instrument and improve accuracy.

When measuring the NIR spectrum of a solid sample, it is common to measure the spectrum of different faces of the sample to reduce the measurement error and obtain reliable information. Sometimes, the repeated measurement of the spectrum can be performed to improve the signal-to-noise ratio of the spectrum. After obtaining the NIR spectrum, it is generally necessary to preprocess the spectrum (baseline correction, first derivative, second derivative and quadrature signal correction).

Thank you for your trust and support!