技术领域technical field
本发明涉及一种添加物含量的辅助测量方法，特别是一种混合体系中添加物直接光谱定量的背景库扩充方法。The invention relates to an auxiliary measurement method for additive content, in particular to a background library expansion method for direct spectral quantification of additives in a mixed system.
背景技术Background technique
现场快速检测中经常会遇到某一类样本是否掺杂或者配比是否符合要求的分析任务，如在产品质量检验和食品药品监督中，需要快速检测添加物的含量。在现有技术中，检测添加物含量采用最多的方法是色谱法和基于特征波长的分光光度法，如专利公开号为“CN102507757A”的《一种高效液相色谱法测定条斑紫菜中抗坏血酸含量的方法》、专利公开号为“CN104297175A”的《采用分光光度法测定食用柠檬黄含量的方法》，有时也采用多变量统计分析建模方法。但依据特征峰定量的分光光度法无法分析缺乏特征峰的样本；多变量统计分析建模方法需要通过模型更新和校准来适应新样本的加入，模型更新，某种程度上是一个重新建模的过程，需要对被测物在新背景下重新定量，然后再汇入原有模型样本集，重新建模，工作量大，成本高；这些方法都无法简便地响应体系变化。On-site rapid detection often encounters the analysis task of whether a certain type of sample is adulterated or whether the ratio meets the requirements. For example, in product quality inspection and food and drug supervision, it is necessary to quickly detect the content of additives. In the prior art, chromatography and spectrophotometry based on characteristic wavelengths are the most widely used methods for detecting the content of additives, such as "A High Performance Liquid Chromatography Method for Determination of Ascorbic Acid Content in Porphyra zebra" with the patent publication number "CN102507757A". The method of ", the patent publication number is " CN104297175A " " the method adopting spectrophotometry to measure edible tartrazine content ", also adopts multivariate statistical analysis modeling method sometimes. However, the spectrophotometric method based on the quantification of characteristic peaks cannot analyze samples lacking characteristic peaks; the multivariate statistical analysis modeling method needs to update and calibrate the model to adapt to the addition of new samples, and the model update is to some extent a remodeling In the process, it is necessary to re-quantify the measured object under the new background, and then import the original model sample set for re-modelling, which requires a lot of work and high cost; these methods cannot easily respond to system changes.
发明内容Contents of the invention
本发明要解决的技术问题是：本发明针对一类混合物中添加物的定量，如果背景库所包含的物种增加，提出一种不需要对新增背景进行重新定量的混合体系中添加物直接光谱定量的背景库扩充方法。The technical problem to be solved by the present invention is: the present invention is aimed at the quantification of additives in a class of mixtures. If the species contained in the background library increases, a direct spectrum of the additives in the mixed system that does not need to be re-quantified for the newly added background is proposed. Quantitative background library augmentation methods.
解决上述技术问题的技术方案是：一种混合体系中添加物直接光谱定量的背景库扩充方法，其特征在于：该方法包括以下步骤：The technical solution for solving the above technical problems is: a background library expansion method for direct spectral quantification of additives in a mixed system, characterized in that the method includes the following steps:
①通过光谱仪测定需要添加的新背景组分光谱bnew；① Measure the spectrum bnew of the new background component that needs to be added by a spectrometer;
②比较新背景组分光谱bnew与原背景矩阵B0的一致性，如果新背景组分光谱bnew与原背景矩阵B0高度一致或差异较大，则背景库不需扩充，操作结束；如果新背景组分光谱bnew与原背景矩阵B0存在一定差异时，进入步骤③；② Compare the consistency between the new background component spectrum bnew and the original background matrix B0. If the new background component spectrum bnew is highly consistent with the original background matrix B0 or if the difference is large, the background library does not need to be expanded, and the operation ends; if the new background component When there is a certain difference between the sub-spectrum bnew and the original background matrix B0, enter step ③;
③将新背景组分光谱 bnew加入到原背景矩阵B0中，构成新背景矩阵B1；③ Add the new background component spectrum bnew to the original background matrix B0 to form a new background matrix B1;
④计算校正样本集S0与新背景矩阵B1的空间夹角系列值的方差值集D0，将该方差值集D0与校正样本集S0中被测添加物含量集C0回归，得到调整后的新标准曲线；④ Calculate the variance value set D0 of the spatial angle series values between the calibration sample set S0 and the new background matrix B1, and regress the variance value set D0 with the measured additive content set C0 in the calibration sample set S0 to obtain the adjusted new standard curve;
⑤对于由新背景矩阵B1中组分与被测添加物X构成的被测样本S，测定其光谱，并计算被测样本S的光谱与新背景矩阵B1的空间夹角系列值的方差值dx，通过调整后的新标准曲线，即可预测背景扩充后的被测样本S中被测添加物X的含量。⑤ For the measured sample S composed of components in the new background matrix B1 and the measured additive X, measure its spectrum, and calculate the variance value of the space angle series between the spectrum of the measured sample S and the new background matrix B1 dx, through the adjusted new standard curve, the content of the tested additive X in the tested sample S after the background expansion can be predicted.
本发明的进一步技术方案是：在步骤②中，所述的比较新背景组分光谱bnew与原背景矩阵B0的一致性采用归属系数ψ判断，如果归属系数ψ达到0.999以上，则新背景组分光谱 bnew与原背景矩阵B0的一致性很高，直接沿用原背景矩阵B0及对应的标准曲线，背景不需扩充；如果归属系数ψ <0.9，则新背景组分光谱 bnew与原背景矩阵B0差异很大，该组分与原有背景不属于同一类，不扩充为新的背景库，应作为新的分类；如果归属系数ψ在0.9至0.999之间，则新背景组分光谱 bnew与原背景矩阵B0存在一定差异，进入步骤③。The further technical solution of the present invention is: in step ②, the consistency between the new background component spectrum bnew and the original background matrix B0 is judged by the belonging coefficient ψ, if the belonging coefficient ψ reaches more than 0.999, the new background component The consistency between the spectrum bnew and the original background matrix B0 is very high, the original background matrix B0 and the corresponding standard curve are directly used, and the background does not need to be expanded; if the belonging coefficient ψ <0.9, the difference between the new background component spectrum bnew and the original background matrix B0 is very large, this component does not belong to the same category as the original background, and should not be expanded into a new background library, it should be used as a new classification; if the belonging coefficient ψ is between 0.9 and 0.999, the spectrum bnew of the new background component There are some differences in matrix B0, go to step ③.
本发明的再进一步技术方案是：所述的原背景矩阵B0为原背景库针对同一类的n个样本b构成的背景矩阵，B0={b1，…，bn}；校正样本集S0为被测添加物X含量已知的p个原有校正样本s0的集合，S0={s01，…，s0p}，被测添加物含量集C0为被测添加物X的p个含量c0的集合，C0={c01，…，c0p}。A further technical solution of the present invention is: the original background matrix B0 is the background matrix formed by the original background library for n samples b of the same class, B0={b1,...,bn}; the corrected sample set S0 is the measured A set of p original calibration samples s0 whose content of additive X is known, S0={s0 1,...,s0 p}, the measured additive content set C0 is the p content c of the measured additive X The set of0 , C0={c0 1,...,c0 p}.
由于采用上述结构，本发明之混合体系中添加物直接光谱定量的背景库扩充方法与现有技术相比，具有以下有益效果：Due to the adoption of the above structure, compared with the prior art, the background library expansion method of direct spectral quantification of additives in the mixed system of the present invention has the following beneficial effects:
1.不需要对新增背景进行重新定量1. No need to re-quantify the new background
由于本发明包括步骤：①测定需要添加的新背景组分光谱bnew；②比较新背景组分光谱bnew与原背景矩阵B0的一致性，如果新背景组分光谱bnew与原背景矩阵B0存在一定差异时，进入步骤③；③将新背景组分光谱 bnew加入到原背景矩阵B0中，构成新背景矩阵B1；④计算校正样本集S0与新背景矩阵B1的空间夹角系列值的方差值集D0，将该方差值集D0与校正样本集S0中被测添加物含量集C0回归，得到调整后的新标准曲线；⑤对于由新背景矩阵B1中组分与被测添加物X构成的被测样本S，测定其光谱，并计算被测样本S的光谱与新背景矩阵B1的空间夹角系列值的方差值dx，通过调整后的新标准曲线，即可预测背景扩充后的被测样本S中被测添加物X的含量。因此，本发明利用被测组分与背景矩阵的独立性，对于整体矩阵空间而言，被测物向量与背景空间夹角系列值的方差只随被测物在整个体系中相对含量变化；而作为背景的其他组分，只要其相对于总体的含量确定，就不会对被测物的定量结果造成影响。这在农产品、食品、精细化学品的配方中是常见的，掺杂造假的辨识任务多数情况下也属于这种情形。因此，本发明不需要对新增背景进行重新定量。Because the present invention includes steps: 1. measure the new background component spectrum bnew that needs to be added; 2. compare the consistency between the new background component spectrum bnew and the original background matrix B0, if there is a certain difference between the new background component spectrum bnew and the original background matrix B0 , enter step ③; ③ add the new background component spectrum bnew to the original background matrix B0 to form a new background matrix B1; ④ calculate the variance value set of the spatial angle series values between the calibration sample set S0 and the new background matrix B1 D0, the variance value set D0 is regressed with the measured additive content set C0 in the calibration sample set S0 to obtain an adjusted new standard curve; Measure the spectrum of the measured sample S, and calculate the variance value dx of the space angle series values between the spectrum of the measured sample S and the new background matrix B1, through the adjusted new standard curve, you can predict the background expansion. Measure the content of the tested additive X in the sample S. Therefore, the present invention utilizes the independence of the measured component and the background matrix, and for the overall matrix space, the variance of the series value of the angle between the measured object vector and the background space only changes with the relative content of the measured object in the whole system; As long as the other components of the background are determined relative to the overall content, they will not affect the quantitative results of the analyte. This is common in the formulation of agricultural products, food, and fine chemicals, and it is also the case in most cases for adulteration identification tasks. Therefore, the present invention does not require requantification of the added background.
2.工作量小，成本低2. Small workload and low cost
由于本发明不需要对新增背景进行重新定量，无需重新建模，其工作量小，成本低。Since the present invention does not need to re-quantify the newly added background, and does not need to re-model, the workload is small and the cost is low.
3. 适合于现场快速测定和常规采样后模型的快速调整3. Suitable for rapid on-site determination and rapid adjustment of models after routine sampling
本发明无需对新添加的样本重新定量，省却了实验室分析环节，非常适合于农产品、食品、精细化学品现场快速测定和常规采样后模型的快速调整，去除了目前分析方法在现场分析和模型更新中遇到的关键限制。The invention does not need to re-quantify the newly added samples, saves the laboratory analysis link, is very suitable for rapid on-site determination of agricultural products, food, and fine chemicals and rapid adjustment of models after conventional sampling, and eliminates the need for on-site analysis and model analysis in current analysis methods. Critical limitations encountered in updates.
下面，结合附图和实施例对本发明之混合体系中添加物直接光谱定量的背景库扩充方法的技术特征作进一步的说明。Below, the technical features of the background library expansion method for direct spectral quantification of additives in the mixed system of the present invention will be further described in conjunction with the accompanying drawings and examples.
附图说明Description of drawings
图1：案例一所述新添加的新背景组分（茶油new）与四种原背景(茶油1-4号)的红外光谱图；Figure 1: Infrared spectra of the newly added new background component (tea oil new) and four original backgrounds (tea oil No. 1-4) described in Case 1;
图2：案例一所述原标准曲线；Figure 2: The original standard curve described in Case 1;
图3：案例二所述添加的新背景组分（茶油new）与四种原背景物在1200-1500cm-1的红外光谱图；Figure 3: Infrared spectra of the new background component (tea oil new) added in Case 2 and four original background substances at 1200-1500cm-1 ;
图4：案例二所述采用校正样本集S0中被测添加物含量集C0={ c01，…，c05 }与新背景矩阵B1的方差值D0={d01，…，d05}建立的标准曲线。Figure 4: The variance valueD0 ={d01 ,..., d0 5} to establish the standard curve.
图1、图3中，横坐标表示波数（cm-1），纵坐标表示透过率（%）；图2、图4中，横坐标表示被测添加物含量，纵坐标表示方差值。In Figure 1 and Figure 3, the abscissa indicates the wave number (cm-1 ), and the ordinate indicates the transmittance (%); in Figure 2 and Figure 4, the abscissa indicates the content of the tested additive, and the ordinate indicates the variance value.
具体实施方式detailed description
一种混合体系中添加物直接光谱定量的背景库扩充方法，用于农产品、食品、精细化学品现场快速测定和常规采样后模型的快速调整，该方法包括以下步骤：A background library expansion method for direct spectral quantification of additives in a mixed system is used for rapid on-site determination of agricultural products, food, and fine chemicals and rapid adjustment of models after routine sampling. The method includes the following steps:
①通过光谱仪测定需要添加的新背景组分光谱bnew；① Measure the spectrum bnew of the new background component that needs to be added by a spectrometer;
②比较新背景组分光谱bnew与原背景矩阵B0的一致性，如果新背景组分光谱bnew与原背景矩阵B0高度一致或差异较大，则背景库不需扩充，操作结束；如果新背景组分光谱bnew与原背景矩阵B0存在一定差异时，进入步骤③；② Compare the consistency between the new background component spectrum bnew and the original background matrix B0. If the new background component spectrum bnew is highly consistent with the original background matrix B0 or if the difference is large, the background library does not need to be expanded, and the operation ends; if the new background component When there is a certain difference between the sub-spectrum bnew and the original background matrix B0, enter step ③;
③将新背景组分光谱 bnew加入到原背景矩阵B0中，构成新背景矩阵B1；③ Add the new background component spectrum bnew to the original background matrix B0 to form a new background matrix B1;
④计算校正样本集S0与新背景矩阵B1的空间夹角系列值的方差值集D0，将该方差值集D0与校正样本集S0中被测添加物含量集C0回归，得到调整后的新标准曲线；④ Calculate the variance value set D0 of the spatial angle series values between the calibration sample set S0 and the new background matrix B1, and regress the variance value set D0 with the measured additive content set C0 in the calibration sample set S0 to obtain the adjusted new standard curve;
⑤对于由新背景矩阵B1中组分与被测添加物X构成的被测样本S，测定其光谱，并计算被测样本S的光谱与新背景矩阵B1的空间夹角系列值的方差值dx，通过调整后的新标准曲线，即可预测背景扩充后的被测样本S中被测添加物X的含量。⑤ For the measured sample S composed of components in the new background matrix B1 and the measured additive X, measure its spectrum, and calculate the variance value of the space angle series between the spectrum of the measured sample S and the new background matrix B1 dx, through the adjusted new standard curve, the content of the tested additive X in the tested sample S after the background expansion can be predicted.
在步骤②中，所述的比较新背景组分光谱bnew与原背景矩阵B0的一致性采用归属系数ψ判断，该归属系数ψ=1-2D/π（参看本发明人公布号为CN105784637A的“标识光谱差异性的方法”，归属系数ψ=1-τ），如果归属系数ψ达到0.999以上（该值与信号噪声等因素相关，可视精度要求调整），则新背景组分光谱 bnew与原背景矩阵B0的一致性很高，直接沿用原背景矩阵B0及对应的标准曲线，背景不需扩充；如果归属系数ψ <0.9，则新背景组分光谱bnew与原背景矩阵B0差异很大，该组分与原有背景不属于同一类，不扩充为新的背景库，应作为新的分类；如果归属系数ψ在0.9至0.999之间，则新背景组分光谱 bnew与原背景矩阵B0存在一定差异，进入步骤③。In step ②, the consistency between the new background component spectrum bnew and the original background matrix B0 is judged by the belonging coefficient ψ, the belonging coefficient ψ=1-2D/π (referring to the inventor’s publication number CN105784637A " method for identifying spectral differences", the belonging coefficient ψ=1-τ), if the belonging coefficient ψ reaches 0.999 or more (this value is related to factors such as signal noise, and can be adjusted according to the requirement of visual accuracy), the new background component spectrum bnew and the original The consistency of the background matrix B0 is very high, the original background matrix B0 and the corresponding standard curve are directly used, and the background does not need to be expanded; if the belonging coefficient ψ <0.9, the new background component spectrum bnew is very different from the original background matrix B0, the The component does not belong to the same category as the original background, and should not be expanded into a new background library, but should be regarded as a new classification; if the belonging coefficient ψ is between 0.9 and 0.999, there is a certain relationship between the spectrum bnew of the new background component and the original background matrix B0. difference, go to step ③.
上述的原背景矩阵B0为原背景库针对同一类的n个样本b构成的背景矩阵，B0={b1，…，bn}；校正样本集S0为被测添加物X含量已知的p个原有校正样本s0的集合，S0={s01，…，s0p}，被测添加物含量集C0为被测添加物X的p个含量c0的集合，C0={c01，…，c0p}。The above-mentioned original background matrix B0 is the background matrix formed by the original background library for n samples b of the same class, B0={b1,...,bn}; the calibration sample set S0 is p original samples whose content of the tested additive X is known. There is a set of calibration samples s0 , S0={s0 1,...,s0 p}, the content set C0 of the tested additive is a set of p content c0 of the tested additive X, C0={c0 1 ,...,c0 p}.
在背景库扩充前，混合体系中添加物直接光谱定量的方法是：采集多种与被测混合体系为同一类物质的光谱作为背景库，构成背景库矩阵，选择背景库中所包含的一种或多种物质作为背景物质，按照不同含量分别往背景物质中加入被测添加物得到混合物，依次计算不同被测添加物含量的混合物的光谱与背景库矩阵的夹角系列值A、系列方差值D；绘制不同被测添加物含量值集C与系列方差值D的标准曲线，再计算被测混合体系光谱与背景库矩阵的方差值，将该方差值代入标准曲线中即可测出被测混合体系中被测添加物的实际含量值，具体包括以下步骤：Before the expansion of the background library, the method of direct spectral quantification of additives in the mixed system is: collect a variety of spectra of the same substance as the mixed system to be measured as the background library to form a background library matrix, and select one of the substances contained in the background library. One or more substances are used as background substances, and the measured additives are added to the background substances according to different contents to obtain a mixture, and the angle series value A and the series variance of the spectrum of the mixture with different measured additive contents and the background library matrix are calculated in turn value D; draw the standard curve of different measured additive content value sets C and series variance value D, then calculate the variance value of the measured mixed system spectrum and the background library matrix, and substitute the variance value into the standard curve Measure the actual content value of the tested additive in the tested mixed system, specifically including the following steps:
（1）选择与被测混合体系为同一类物质，分别采集其光谱构建背景库，每种光谱以列排列，构成背景库矩阵；背景库矩阵的每一行对应相同的光谱波长响应值，每一列对应每种物质的系列波长下的响应值；所采集的光谱是红外光谱或近红外光谱或拉曼光谱。(1) Select the same type of substance as the mixed system to be tested, and collect its spectra to construct a background library. Each spectrum is arranged in columns to form a background library matrix; each row of the background library matrix corresponds to the same spectral wavelength response value, and each column Responses at a range of wavelengths for each substance; spectra collected are either infrared or near-infrared or Raman.
（2）选择背景库中所包含的一种或多种物质作为背景物质，按照含量c序列，分别往背景物质中加入被测添加物得到系列混合物，测量系列混合物的光谱；混合物的光谱与背景库光谱的波长一一对应。(2) Select one or more substances contained in the background library as the background substance, add the measured additives to the background substance respectively according to the sequence of content c to obtain a series of mixtures, and measure the spectrum of the series of mixtures; the spectrum of the mixture and the background There is a one-to-one correspondence between the wavelengths of the library spectra.
（3）选择一条含被测添加物的光谱，计算该含被测添加物的光谱与背景库矩阵的移动窗口夹角系列值，并求取夹角系列值的方差值d；由p个不同被测添加物含量的系列混合物光谱可得到系列方差值D={d1，…，dp}；(3) Select a spectrum containing the additive to be tested, calculate the series of angles between the spectrum containing the additive to be tested and the background library matrix in the moving window, and calculate the variance value d of the series of angles; from p A series of mixture spectra with different contents of the tested additives can obtain a series of variance values D={d1,...,dp};
（4）绘制p个不同被测添加物含量值集C={c1，…，cp}和系列方差值D={d1，…，dp}的标准曲线；(4) Draw a standard curve of p different content value sets of tested additives C={c1,...,cp} and series variance value D={d1,...,dp};
（5）计算被测混合体系光谱与背景库矩阵的方差值，将该方差值代入标准曲线中测出被测混合体系中被测添加物的实际含量值。(5) Calculate the variance value of the spectrum of the tested mixed system and the background library matrix, and substitute the variance value into the standard curve to measure the actual content value of the tested additive in the tested mixed system.
步骤（3）包括以下具体内容：Step (3) includes the following specific contents:
（3）-1、选择含被测添加物的光谱全部波长点的一半，建立移动窗口；(3)-1. Select half of all wavelength points of the spectrum containing the additive to be tested, and establish a moving window;
（3）-2、移动窗口的起始位置位于背景库矩阵和作为向量的含被测添加物光谱的顶端，计算得移动窗口内背景库矩阵与向量的夹角a1；(3)-2. The starting position of the moving window is located at the top of the background library matrix and the spectrum containing the additive to be measured as a vector, and the angle a1 between the background library matrix and the vector in the moving window is calculated;
（3）-3、移动窗口下移，计算得移动窗口内背景库矩阵与向量的夹角a2；直至移动窗口移至背景库矩阵底部，得到夹角aend；(3)-3. The moving window moves down, and the angle a2 between the background library matrix and the vector in the moving window is calculated; until the moving window moves to the bottom of the background library matrix, the included angle aend is obtained;
（3）-4、将{a1，…，an，…，aend}构成夹角系列值A；(3)-4. Make {a1,...,an,...,aend} constitute the angle series value A;
（3）-5、计算夹角系列值A的方差值d；(3)-5. Calculate the variance value d of the included angle series value A;
（3）-6、由p个不同被测添加物含量的系列混合物光谱可得到系列方差值D={d1，…，dp}。(3)-6. A series of variance values D={d1,...,dp} can be obtained from the series of mixture spectra of p different measured additive contents.
以下是本发明的具体实施案例：The following are specific implementation cases of the present invention:
案例一Case number one
一种混合体系中添加物直接光谱定量的背景库扩充方法，用于不同产地茶油中大豆油添加含量分析，该方法包括以下步骤：A background library expansion method for direct spectral quantification of additives in a mixed system is used for the analysis of soybean oil additive content in camellia oil from different origins. The method includes the following steps:
①测定需要添加的新背景组分（茶油new）红外光谱bnew；图1为新添加的新背景组分（茶油new）红外光谱bnew与四种原背景物(茶油1-4号) 红外光谱b1～b4。① Determination of the infrared spectrum bnew of the new background component (tea oil new) that needs to be added; Figure 1 shows the newly added new background component (tea oil new) infrared spectrum bnew and four original background substances (tea oil No. 1-4) Infrared spectrum b1 ~ b4.
②比较新背景组分光谱bnew与原背景矩阵B0（b1~b4）的一致性：计算新添加背景光谱bnew与4种原背景物光谱库B0的归属系数ψ为0.9999，表明新加入的背景物质与原背景物质基本无差异，可直接将新背景物质加入背景库中，不需要重新建立标准曲线，背景不需扩充，操作结束。②Compare the consistency of the new background component spectrum bnew with the original background matrix B0 (b1~b4): calculate the belonging coefficient ψ between the newly added background spectrum bnew and the four original background object spectral libraries B0 is 0.9999, indicating that the newly added background material There is basically no difference with the original background substance, and the new background substance can be directly added to the background library without re-establishing the standard curve, the background does not need to be expanded, and the operation is over.
实验验证：本申请的发明人在新背景茶油new中分别加入25%、35%、45%、55%的被测添加物大豆油，得到4份被测样本N1-N4。测定被测样本光谱，计算其与原背景光谱矩阵B0的d值分别为：8.545×10-5、1.225×10-4、1.599×10-4、1.957×10-4。通过图2（图2中，横坐标为被测添加物含量，纵坐标为方差值）原标准曲线可得到4份样本的被测添加物含量值为：24.8%、34.9%、45.1%、54.9%，最大相对误差为0.8%。计算结果参见附表1。Experimental verification: the inventor of the present application added 25%, 35%, 45%, and 55% of the tested additive soybean oil to the new background camellia oil new, and obtained 4 tested samples N1-N4. Measure the spectrum of the measured sample, and calculate the d values of it and the original background spectrum matrix B0 as: 8.545×10-5 , 1.225×10-4 , 1.599×10-4 , 1.957×10-4 . Through the original standard curve in Figure 2 (in Figure 2, the abscissa is the content of the tested additive, and the ordinate is the variance value), the values of the tested additive content of the four samples can be obtained: 24.8%, 34.9%, 45.1%, 54.9%, and the maximum relative error is 0.8%. See attached table 1 for the calculation results.
从附表1可看出，当新添加背景组分与原背景物属于同一类时，背景库无需扩充，不需要重新建立标准曲线。It can be seen from the attached table 1 that when the newly added background component belongs to the same category as the original background object, the background library does not need to be expanded, and the standard curve does not need to be re-established.
案例二case two
一种混合体系中添加物直接光谱定量的背景库扩充方法，用于不同产地茶油中大豆油添加含量分析，该方法包括以下步骤：A background library expansion method for direct spectral quantification of additives in a mixed system is used for the analysis of soybean oil additive content in camellia oil from different origins. The method includes the following steps:
①测定需要添加的新背景组分（茶油new）红外光谱bnew；图3为新背景组分（茶油new）红外光谱bnew与4种原背景物(茶油1-4号)的红外光谱b1～b4；① Measure the infrared spectrum bnew of the new background component (tea oil new) that needs to be added; Figure 3 shows the infrared spectrum bnew of the new background component (tea oil new) and the infrared spectra of the four original background substances (tea oil No. 1-4) b1～b4;
②比较新背景组分光谱bnew与原背景矩阵B0的一致性，对新背景组分光谱bnew与原背景矩阵B0作一致性判断，计算得到归属系数ψ为0.9950，可判断新背景组分光谱bnew与原背景矩阵B0存在差异，但其性质上仍属于同一类；②Comparing the consistency between the new background component spectrum bnew and the original background matrix B0, making a consistency judgment on the new background component spectrum bnew and the original background matrix B0, the calculated belonging coefficient ψ is 0.9950, which can judge the new background component spectrum bnew There are differences with the original background matrix B0, but they still belong to the same category in nature;
③将新背景组分光谱 bnew加入到原背景矩阵B0中，构成新背景矩阵B1；③ Add the new background component spectrum bnew to the original background matrix B0 to form a new background matrix B1;
④直接计算已有的含有5个校正样本的校正样本集S0与新背景矩阵B1的空间夹角系列值的方差值集D0，将该方差值集D0与校正样本集S0中被测添加物含量集C0回归，得到调整后的新标准曲线，无需重新建立新样本；即是利用在背景扩充前的5个校正样本的光谱s1-s5，依次计算得到光谱s1-s5与新背景矩阵B1的系列夹角A={a1’，…，a5’}，并求取方差值集D0={d01，…，d05}，与被测添加物含量集C0={ c01，…，c05 }绘制标准曲线，相关系数为0.9999。④ directly calculate the variance value set D0 of the spatial angle series values between the existing calibration sample set S0 containing 5 calibration samples and the new background matrix B1, and add the variance value set D0 to the calibration sample set S0 Regression of substance content set C0 to obtain an adjusted new standard curve without re-establishing a new sample; that is, using the spectra s1-s5 of the 5 calibration samples before background expansion, and sequentially calculating the spectra s1-s5 and the new background matrix B1 The series angle A={a1',...,a5'}, and calculate the variance value set D0={d0 1,...,d0 5}, and the content set C0={ c0 1 ,...,c0 5 } draw a standard curve, and the correlation coefficient is 0.9999.
图4是采用校正样本集S0中被测添加物含量集C0={ c01，…，c05 }与新背景矩阵B1的方差值D0={d01，…，d05}建立的标准曲线，该图4中，横坐标为被测添加物含量，纵坐标为方差值）。Figure 4 is the variance value D0={d0 1,...,d0 5} of the measured additive content set C0={c0 1,...,c0 5} and the new background matrix B1 in the calibration sample set S0 The established standard curve, in Figure 4, the abscissa is the content of the tested additives, and the ordinate is the variance value).
⑤对于由新背景矩阵B1中组分与被测添加物X构成的被测样本S，测定其光谱，并计算被测样本S的光谱与与新背景矩阵B1的空间夹角系列值的方差值dx，通过调整后的新标准曲线，即可预测背景扩充后的被测样本S中被测添加物X的含量。⑤For the measured sample S composed of components in the new background matrix B1 and the measured additive X, measure its spectrum, and calculate the variance of the spectrum of the measured sample S and the space angle series value with the new background matrix B1 The value dx, through the adjusted new standard curve, can predict the content of the tested additive X in the tested sample S after background expansion.
实验验证：本申请的发明人对新增的新背景组分（茶油new），分别加入25%、35%、45%、55%的被测添加物大豆油得到四个被测样本Y1-Y4。计算被测样本Y1-Y4的红外光谱与新背景矩阵B1的方差值d分别为：8.613×10-5、1.235×10-4、1.599×10-4、1.966×10-4。由重新标定的标准曲线得到的含量值为：24.9%、35.1%、45.0%、55.1%，最大相对误差小于0.5%。计算结果见附表2。Experimental verification: the inventor of this application added 25%, 35%, 45%, and 55% of the tested additive soybean oil to the newly added new background component (tea oil new) to obtain four tested samples Y1- Y4. Calculate the variance values d of the infrared spectra of the tested samples Y1-Y4 and the new background matrix B1 as: 8.613×10-5 , 1.235×10-4 , 1.599×10-4 , 1.966×10-4 . The content values obtained from the re-calibrated standard curve are: 24.9%, 35.1%, 45.0%, 55.1%, and the maximum relative error is less than 0.5%. The calculation results are shown in Table 2.
从该附表2可看出，当新添加背景组分与原背景存在差异，需要扩充背景库。可将新添加背景组分纳入原背景矩阵，构成新的背景矩阵。利用原有的系列校正样本，计算这些样本与新背景库矩阵的空间夹角系列值的方差，与校正样本的含量回归，得到调整后的标准曲线。It can be seen from the attached table 2 that when the newly added background components are different from the original background, the background library needs to be expanded. The newly added background components can be incorporated into the original background matrix to form a new background matrix. Using the original series of calibration samples, calculate the variance of the series values of the space angle between these samples and the new background library matrix, and regress with the content of the calibration samples to obtain the adjusted standard curve.
因此，本发明无需对新添加的样本重新定量，省却了实验室分析环节，非常适合于农产品、食品、精细化学品等现场快速测定和常规采样后模型的快速调整，去除了目前分析方法在现场分析和模型更新中遇到的关键限制。Therefore, the present invention does not need to re-quantify newly added samples, saves the laboratory analysis link, and is very suitable for rapid on-site determination of agricultural products, food, fine chemicals, etc. and rapid adjustment of models after conventional sampling, eliminating the need for current analysis methods to Key limitations encountered in analysis and model updates.
。 .
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610710873.0ACN106323898B (en) | 2016-08-23 | 2016-08-23 | The quantitative context vault extending method of additive Spectrographic in mixed system |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610710873.0ACN106323898B (en) | 2016-08-23 | 2016-08-23 | The quantitative context vault extending method of additive Spectrographic in mixed system |
| Publication Number | Publication Date |
|---|---|
| CN106323898Atrue CN106323898A (en) | 2017-01-11 |
| CN106323898B CN106323898B (en) | 2019-01-15 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610710873.0AExpired - Fee RelatedCN106323898B (en) | 2016-08-23 | 2016-08-23 | The quantitative context vault extending method of additive Spectrographic in mixed system |
| Country | Link |
|---|---|
| CN (1) | CN106323898B (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109297918A (en)* | 2018-08-21 | 2019-02-01 | 广西科技大学 | A kind of method for detecting acid red 26 |
| CN114428067A (en)* | 2020-10-12 | 2022-05-03 | 中国石油化工股份有限公司 | Method for predicting gasoline octane number |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150016711A1 (en)* | 2013-07-10 | 2015-01-15 | Canon Kabushiki Kaisha | Devices, systems, and methods for classifying materials based on a bidirectional reflectance distribution function |
| CN105158189A (en)* | 2015-08-28 | 2015-12-16 | 广西科技大学 | Method for analyzing content of antioxidants in vegetable oil on basis of criterion of space angle |
| CN105784637A (en)* | 2016-03-30 | 2016-07-20 | 广西科技大学 | Method for identifying spectrum difference |
| CN105844646A (en)* | 2016-04-01 | 2016-08-10 | 广西科技大学 | Mixture spectrum consistency determination method |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150016711A1 (en)* | 2013-07-10 | 2015-01-15 | Canon Kabushiki Kaisha | Devices, systems, and methods for classifying materials based on a bidirectional reflectance distribution function |
| CN105158189A (en)* | 2015-08-28 | 2015-12-16 | 广西科技大学 | Method for analyzing content of antioxidants in vegetable oil on basis of criterion of space angle |
| CN105784637A (en)* | 2016-03-30 | 2016-07-20 | 广西科技大学 | Method for identifying spectrum difference |
| CN105844646A (en)* | 2016-04-01 | 2016-08-10 | 广西科技大学 | Mixture spectrum consistency determination method |
| Title |
|---|
| 刘柳 等: "基于向量-子空间夹角判据测定果冻中色素和防腐剂", 《食品研究与开发》* |
| 胡爱琴 等: "多组分复杂体系光谱多元定量分析方法研究", 《光谱学与光谱分析》* |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109297918A (en)* | 2018-08-21 | 2019-02-01 | 广西科技大学 | A kind of method for detecting acid red 26 |
| CN114428067A (en)* | 2020-10-12 | 2022-05-03 | 中国石油化工股份有限公司 | Method for predicting gasoline octane number |
| Publication number | Publication date |
|---|---|
| CN106323898B (en) | 2019-01-15 |
| Publication | Publication Date | Title |
|---|---|---|
| CN104062257B (en) | A kind of based on the method for general flavone content near infrared ray solution | |
| CN104062259B (en) | A kind of use the method for total saponin content near infrared spectrum quick test complex prescription glue mucilage | |
| CN105372195B (en) | A kind of micro ultraviolet specrophotometer quality determining method and detection kit | |
| CN103543123A (en) | Infrared spectrum recognition method for adulterated milk | |
| CN101216419A (en) | Method for quickly detecting yellow wine quality index | |
| CN105203498A (en) | Near infrared spectrum variable selection method based on LASSO | |
| Ouyang et al. | Measurement of non-sugar solids content in Chinese rice wine using near infrared spectroscopy combined with an efficient characteristic variables selection algorithm | |
| CN104020134A (en) | Rapid determination method and rapid determination system for adulterated starch in food based on near infrared spectrum | |
| CN108169168A (en) | Test and analyze rice grain protein content mathematical model and construction method and application | |
| CN101408501A (en) | Method for quantitatively detecting DNA base by using near-infrared spectrum-partial least squares method | |
| CN107505283A (en) | The method of nitrate ion content near infrared ray Secondary salinization soil | |
| CN104297201A (en) | Method for quickly, accurately and quantitatively detecting ratio of various oil components in blend oil | |
| CN106383091A (en) | A method of directly quantifying the content of an additive through spectrums | |
| CN106769927A (en) | A kind of quality determining method of Milkvetch Root | |
| CN106770005B (en) | A kind of division methods of calibration set for near-infrared spectrum analysis and verification collection | |
| CN106323898A (en) | Method for expanding background database for direct spectroscopy quantitative analysis of additives in mixing system | |
| WO2020248961A1 (en) | Method for selecting spectral wavenumber without reference value | |
| CN103592256A (en) | Mid-infrared spectroscopic method for distinguishing normal edible vegetable oil from refined hogwash oil based on Fourier transform | |
| CN101576486B (en) | Method for fast measuring essence mixture ratio in cigarette | |
| CN102485249A (en) | Detection method for quality evaluation of saffron | |
| CN107271396A (en) | The quick determination method of general flavone content in a kind of tealeaves | |
| CN104568827A (en) | Method for measuring sugar content of watermelon by near-infrared spectrometry | |
| CN101140225B (en) | Method for detecting lead in scenting agent with AOTF near-infrared spectrometer | |
| CN106092960A (en) | A kind of quick correction near infrared gear also detects the method for chemical composition in agricultural product | |
| Li et al. | Near infrared spectroscopy combined with multivariate analysis for monitoring the ethanol precipitation process of fraction I+ II+ III supernatant in human albumin separation |
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | Granted publication date:20190115 | |
| CF01 | Termination of patent right due to non-payment of annual fee |