CN102680427A - Method for identifying cigarette surface aroma quality by applying near infrared spectrum analysis technology - Google Patents

Abstract

The invention discloses a method for identifying the quality of cigarette surface aroma by applying near-infrared spectrum analysis technology. Firstly, the spectral data of qualified table fragrances are stored in the computer, and the similarity matching model of the brand table fragrances is established by using the qualitative analysis method of similarity matching in TQ analyst7.1 software, and the similarity matching criticality of the brand table fragrances is obtained through calculation. value. When identifying the quality of the surface aroma, apply the same method to collect the spectral data of the surface aroma to be identified, and calculate the spectral similarity matching value of the surface aroma to be identified, when the similarity matching value of the surface aroma to be identified ≥ the similarity It is qualified when the degree matches the critical value, otherwise it is unqualified. The identification method requires less samples, and the identification process does not require pretreatment and sample preparation procedures. The identification process is simple, and it is very suitable for promotion in the tobacco industry, and has a good application prospect in the quality stability control of cigarette flavor.

Description

A method for identifying the quality of cigarette flavor by using near-infrared spectral analysis technology

technical field

The invention belongs to the technical field of cigarette quality control, and in particular relates to a method for identifying the quality of cigarette surface aroma by applying near-infrared (NIR) spectral analysis technology. the

Background technique

The purpose of cigarette flavoring is mainly to set off the aroma of cigarettes and cover up miscellaneous odors without damaging the original aroma of tobacco leaves. Due to the difference in the quality of the leaves and the application of coke reduction measures, the aroma of cigarette products will be lacking or sluggish. Flavoring can supplement the beautiful aroma; flavoring can also add highly volatile characteristic spices to give cigarettes a unique aroma , to increase the appeal to consumers; in addition, flavoring can also increase the sweetness, improve the taste, cover up or weaken some miscellaneous, irritating, dry and rough; the more important role of flavoring is to make different types of The aromas of different grades of tobacco leaves produced in different varieties, different origins, and different years are organically combined and coordinated, and at the same time, it can cover up or dilute the miscellaneous odors and improve the quality. Therefore, the quality stability of cigarette top flavor is directly related to the quality stability of cigarette products. The surface flavors for tobacco are mostly made of natural spices, which are affected by various factors such as raw materials and processing. The complexity and diversity of its chemical components have always been the focus and difficulty of quality control. the

The existing quality control methods for cigarette surface aroma are mainly based on physical indicators such as acidity, miscibility, refractive index, total volatile matter, optical rotation, etc., and manual smelling by perfumers. The method is relatively single and one-sided. The use of chromatographic fingerprints to control the quality of cigarette flavors and fragrances is a research hotspot at present, but the pretreatment process of chromatographic analysis technology is complicated, which limits the application of chromatographic fingerprints. Therefore, in order to ensure the consistency of the quality of cigarette flavors from different raw material origins and different batches, it is urgent to develop a more optimized method for identifying the quality of cigarette flavors. Near-infrared spectral analysis technology is a convenient, efficient and low-cost green analysis technology developed rapidly in recent years. It has been widely used in industries such as agriculture, forestry, papermaking, tea, food, and Chinese herbal medicine. The application of the quality identification of surface aroma has not been reported yet. the

Contents of the invention

The purpose of the present invention is to provide a method for identifying the quality of cigarette surface aroma by applying near-infrared spectrum analysis technology to address the existing deficiencies. Using FT-NIR spectral data, the qualitative analysis method of similarity matching is used to directly establish the quality control model of cigarette flavor, so as to realize the rapid and accurate identification of the quality of cigarette flavor. the

The purpose of the present invention is achieved through the following technical solutions. the

Unless otherwise stated, the percentages used in the present invention are all percentages by weight. the

A method for identifying the quality of cigarette surface aroma by using near-infrared spectral analysis technology, comprising the following steps:

(1) Spectral data collection: Put the qualified surface aroma samples of a certain brand of cigarettes into small test tubes, place them on the transmission sample rack to collect the near-infrared transmission spectrum data of the surface aroma samples, and repeat the data collection 10 times for each sample, and Store the collected spectral data into the computer; the spectral collection method is: transmission spectrum collection mode; scanning range: 4000-10000 cm ^-1 ; resolution: 8 cm ^-1 ; scanning times: 16 times; A background spectrum must be collected;

(2) Model construction: Apply the qualitative analysis method of similarity matching in TQ analyst7.1 software to establish the similarity matching model of the cigarette brand's surface flavor;

(3) Calculation of similarity matching threshold value: Based on the similarity matching model, use TQ analyst7.1 software to calculate the spectral similarity matching value (Similarity Match Value, SMV) of the brand's cigarette surface flavor, and the brand's cigarette table flavor Statistically analyze the SMV value of cigarettes, analyze the range of SMV data distribution, and determine the minimum value Min of SMV; similarly, collect the spectral data of non-brand cigarettes produced by the same enterprise in batches, and analyze the non-brand cigarettes. Calculate the SMV value of the cigarette, conduct statistical analysis on the obtained SMV value, analyze the distribution range of SMV data, and determine the upper limit value Max of SMV; the similarity matching critical value CV of the brand's cigarette surface fragrance is obtained through calculation:

CV＝(Min+Max)/2

(4) Identification: Repeat step (1) to collect the spectral data of the surface flavor of the cigarette to be identified, and calculate the SMV of the surface flavor of the cigarette to be identified by TQ analyst7.1 software based on the similarity matching model of the brand’s cigarette surface flavor. When the SMV ≥ CV of the surface flavor of the cigarette to be identified is identified as qualified, otherwise it is unqualified. the

The qualified surface fragrance samples mentioned in step (1) are qualified samples produced by the brand at different times and in different batches. the

Compared with prior art, the present invention has following advantage:

1. Using the existing near-infrared spectral analysis instrument, combined with the qualitative analysis function of spectral data similarity matching in the TQ analyst spectral analysis software, it is very simple to establish an identification model. the

2. The identification process is simple. After the model is built, the identification process can be completed without professional background personnel, which is easy to promote.

3. The test of model identification results shows that the accuracy of model identification is high, and it can be used as an effective evaluation and detection tool for quality stability of cigarette flavor. the

Detailed ways

The present invention will be further described in detail through application examples below, but the examples do not limit the technical solution of the present invention. the

Application Example 1

The quality of surface fragrance of Yunyan (soft treasure) is non-destructively identified by near-infrared spectral analysis technology. the

1. Sample preparation: prepare 20 batches of qualified surface aroma samples of Yunyan (soft treasure) produced by the ingredient center of Kunming Cigarette Factory of Hongyun Honghe Tobacco (Group) Co., Ltd., and take 100 grams from each batch for the construction of the model. the

2. Spectral data acquisition: The Nicolet Antaris FT-near-infrared spectrometer of Thermoelectric Company was used to collect the spectrum of the sample. Before spectrum collection, turn on the FT-NIR spectrometer to warm up for 2 hours. Put different surface fragrance samples into small test tubes, and place them on the transmission sample holder to collect the near-infrared transmission spectrum data of the surface fragrance samples. Data collection was repeated 10 times for each sample, and the collected spectral data was stored in the computer; the spectral collection method was: transmission spectrum collection mode; scanning range (Data Range): 4000～10000cm ^-1 ; resolution (Resolution) : 8 cm ^-1 ; number of scans: 16 times; the background spectrum must be collected before each sample spectrum is collected.

3. Model construction:

In order to reduce the multiple deviations caused by factors such as sample uniformity, the original FT-NIR spectrum was processed with the additional scatter correction technology (Multiplicative Scatter Correction, MSC) in TQ analyst7.1 software. At the same time, adjusting and optimizing relevant parameters when building a model is one of the main means to improve the predictive ability and effect of the model. Smoothing and baseline correction of the spectrum can effectively remove various interferences and noises such as baseline drift and light scattering in the near-infrared spectrum signal, fully extract the effective feature information contained in the FT-NIR spectrum, and improve the prediction accuracy of the model. In addition, choosing an appropriate spectral range is also one of the important links in the modeling process. Through multiple optimization calculations on the model, appropriate parameters were selected to establish a Similarity Match model for the surface aroma of Yunyan (soft treasure) cigarettes. Table 1 shows the modeling parameter settings of Yunyan (soft treasure) cigarette flavor. the

Table 1 The modeling optimization parameters of the similarity matching model of Yunyan (soft treasures)

4. Calculate the similarity matching threshold:

Based on the similarity matching model of Yunyan (soft treasure) surface fragrance, use TQ analyst7.1 software to calculate the SMV of Yunyan (soft treasure) surface fragrance, and conduct statistical analysis on the obtained SMV value, and analyze the distribution range of SMV data , determine the minimum value Min of SMV as 94.12; similarly, collect the spectral data of non-yunyan (soft treasures) surface fragrance in batches, calculate the SMV of non-yunyan (soft treasures) surface fragrance, and calculate the obtained SMV Statistical analysis was carried out to analyze the distribution range of SMV data, and the upper limit value Max of SMV was determined to be 78.17; the similarity matching critical value of Yunyan (soft treasure) surface fragrance was obtained by calculation CV=(94.12+78.17)/2=86.15. the

5. Identification: Repeat step 2 to collect the spectral data of the surface fragrance to be identified. Based on the similarity matching model of Yunyan (soft treasure) surface fragrance, calculate the SMV of the surface fragrance to be identified by TQ analyst7.1 software. When the SMV of the surface incense is ≥ 86.15, it is identified as a qualified surface incense of Yunyan (soft treasure), otherwise it is unqualified. the

Application Example 2

Non-destructive identification of the quality of Yunyan (purple) table fragrance by using near-infrared spectral analysis technology: 

1. Sample preparation: use Yunyan (purple) surface fragrance for the sample, and the others are the same as in Example 1. the

2. Spectral data collection: same as in Example 1. the

3. Model construction: the selected modeling parameters are shown in Table 2, and others are the same as in Example 1. the

Table 2 The modeling optimization parameters of the similarity matching model of Yunyan (purple) fragrance

4. Calculate the similarity matching critical value: Based on the similarity matching model of Yunyan (purple) superficial incense, calculate the similarity matching critical value of Yunyan (purple) superficial incense CV=(97.07+77.94)/2=87.51. the

5. Identification: Repeat step 2 to collect the spectral data of the table fragrance to be identified. Based on the similarity matching model of Yunyan (purple) table fragrance, calculate the SMV of the table fragrance to be identified by TQ analyst7.1 software. When the table fragrance to be identified When the SMV of the incense is ≥ 87.51, it is identified as a qualified surface incense of Yunyan (purple), otherwise it is unqualified. the

Application Example 3

Non-destructive identification of the authenticity of Yunyan (red) cigarettes by using near-infrared spectroscopy analysis technology:

1. Sample preparation: Yunyan (red) is used as the sample, and the others are the same as in Example 1. the

2. Spectral data collection: same as in Example 1. the

3. Model construction: the selected modeling parameters are shown in Table 3, and others are the same as in Example 1. the

Table 3 The modeling optimization parameters of the similarity matching model of Yunyan (red) table fragrance

4. Calculation of similarity matching critical value: Based on the similarity matching model of Yunyan (red) surface incense, calculate the similarity matching critical value of Yunyan (red) surface incense CV=(96.08+79.60)/2=87.84. the

5. Identification: Repeat step 2 to collect the spectral data of the table fragrance to be identified, and use the TQ analyst7.1 software to calculate the SMV of the table fragrance to be identified based on the similarity matching model of Yunyan (red) table fragrance. When the SMV of the incense is ≥ 87.84, it is identified as a qualified table incense of Yunyan (red), otherwise it is unqualified. the

Application Example 4

The surface aroma of Yunyan (soft treasure), Yunyan (purple) and Yunyan (red) cigarettes were identified, and the identification accuracy of the method was investigated. Take 30 surface fragrance samples (20 of which are brand surface fragrances in the model calibration set, and 10 are non-model calibration set brand surface fragrances) for identification, and obtain the SMV of each sample and the sample in the calibration set. The results are shown in Tables 4, 5 and 6. The results showed that their ability to identify and judge the quality of cigarette surface flavors all reached 100%. the

Table 4 The prediction results of Yunyan (soft treasure) surface aroma similarity matching model for different types of surface aroma

Table 5 Prediction results of Yunyan (purple) superficial fragrance similarity matching model for different types of superficial fragrance

Table 6 The prediction results of Yunyan (red) surface aroma similarity matching model for different types of surface aroma

Claims (1)

1. use the fragrant method for quality of near-infrared spectral analysis technology discriminating cigarette table for one kind, may further comprise the steps:

(1) spectrum data gathering: the fragrant sample of the table that certain brand cigarette is qualified is packed in the small test tube; Place the near-infrared transmission spectroscopic data of the fragrant sample of acquisition tables on the transmission sample frame; Each sample carries out 10 repeating data collections, and deposits the spectroscopic data of being gathered in computing machine; The spectra collection mode is: the transmitted spectrum drainage pattern; Sweep limit: 4000～10000cm ^-1Resolution: 8 cm ^-1Scanning times: 16 times; Must gather background spectrum before each collected specimens spectrum;

(2) model construction: the method for qualitative analysis of using the similarity coupling in the TQ analyst7.1 software is set up the fragrant similarity Matching Model of this cigarette brand table;

(3) similarity coupling critical value is calculated: be the basis with the similarity Matching Model; The fragrant spectrum similarity matching value SMV of utilization TQ analyst7.1 this brand cigarette table of computed in software; And the fragrant SMV value of this brand cigarette table carried out statistical study; Analyze the scope of SMV DATA DISTRIBUTION, confirm the lower limit numerical value Min of SMV; Likewise; Gather the fragrant spectroscopic data of non-this brand cigarette table of same enterprise production in batches, and the fragrant SMV value of non-this brand cigarette table is calculated, resulting SMV value is carried out statistical study; Analyze the scope of SMV DATA DISTRIBUTION, confirm the upper limit numerical value Max of SMV; Calculate the fragrant similarity coupling critical value CV of this brand cigarette table:

CV＝（Min+?Max）/2

(4) differentiate: repeating step (1); Gather the fragrant spectroscopic data of cigarette table to be identified; The similarity Matching Model fragrant with this brand cigarette table is the basis; Through the fragrant SMV of TQ analyst7.1 computed in software cigarette table to be identified, it is up-to-standard when the fragrant SMV of cigarette table to be identified >=CV, to be accredited as this table perfume, otherwise is defective.