Summary of the invention
The present invention has overcome the deficiencies in the prior art, and the detection method of the integral body formation that solves the reflection flavouring essence for tobacco is provided.
For solving above-mentioned technical matters, the present invention by the following technical solutions:
A kind of detection method of flavouring essence for tobacco is characterized in that according to following steps:
(1) sets up the ultraviolet fingerprint principal component analysis model of essence sample for tobacco to be measured
Step 1 is gathered the uv absorption spectra of essence sample for tobacco to be measured;
The ultraviolet spectrogram of the essence sample for tobacco to be measured that step 2 will collect adopts S-G smoothly to be reconstructed ultraviolet fingerprint in conjunction with first order derivative;
(2) prepare qualified data
Step 3 selects to be no less than 7 qualified samples;
Step 4 is set up the ultraviolet fingerprint principal component analysis model of qualified essence sample with the Principal Component Analysis Method in the Chemical Measurement multivariable analysis, obtains qualified samples SIMCA disaggregated model;
Step 5 in level of significance be calculate under 5% the condition each qualified samples to qualified samples SIMCA disaggregated model apart from Si and each qualified samples distance H i to qualified samples SIMCA disaggregated model center, and calculate Si upper control limit and Hi upper control limit, lower control limit is 0;
(3) judge relatively
The ultraviolet fingerprint of the reconstruct that step 6 obtains in step 2 essence sample for tobacco to be measured with the qualified samples SIMCA disaggregated model that is obtained in the step 4 is 5% to carry out the SIMCA discriminant classification with level of significance, calculate essence sample for tobacco to be measured to qualified samples SIMCA disaggregated model apart from Si and essence sample for tobacco to be measured distance H i to qualified samples SIMCA disaggregated model center, have only when the essence sample to qualified samples SIMCA disaggregated model apart from Si less than Si upper control limit and essence sample for tobacco to be measured to the distance H i at qualified samples SIMCA disaggregated model center during less than the Hi upper control limit this essence sample for tobacco to be measured qualified, otherwise this essence sample for tobacco to be measured is defective.
Further technical scheme is the wavelength coverage of the UV scanning when gathering essence sample for tobacco uv absorption spectra to be measured: 190~450nm, wavelength interval 1nm; Quartz colorimetric utensil light path: 0.1cm; Each sample scanning 7 times.
Further technical scheme be in the step 5 the Si upper control limit be each qualified samples to qualified samples SIMCA disaggregated model apart from peaked three times of Si, the Hi upper control limit is each qualified samples to peaked three times of the distance at qualified samples SIMCA disaggregated model center.
Further technical scheme is each qualified essence sample directly obtaining from database apart from Si and the distance H i of each qualified essence sample to qualified samples SIMCA disaggregated model center to qualified samples SIMCA disaggregated model in the step 5.
Further technical scheme is after the step 6, with the essence sample for tobacco to be measured of test passes joining in the database apart from the essence sample for tobacco to be measured of Si and the test passes distance H i to qualified samples SIMCA disaggregated model center to qualified samples SIMCA disaggregated model.
Further technical scheme is that the primary data in the database is to select at least seven batches of qualified essence sample for tobacco for use, carries out according to following steps:
Step 7 is gathered the uv absorption spectrogram of qualified essence sample for tobacco;
The spectrogram that step 8 will collect qualified essence sample adopts S-G smoothly to be reconstructed ultraviolet fingerprint in conjunction with first order derivative;
Step 9 is set up the ultraviolet fingerprint principal component analysis model of essence sample for tobacco to be measured with the Principal Component Analysis Method in the Chemical Measurement multivariable analysis, obtains qualified samples SIMCA disaggregated model;
Step 10 in level of significance be calculate under 5% the condition each qualified samples to qualified samples SIMCA disaggregated model apart from Si and qualified samples distance H i to qualified samples SIMCA disaggregated model center;
Step 11 deposits each qualified samples in database apart from Si and each qualified samples to the distance H i at qualified samples SIMCA disaggregated model center to qualified samples SIMCA disaggregated model.
Further technical scheme is that Si upper control limit and Hi upper control limit obtain according to following steps in the step 5:
Step 12 is gathered the uv absorption spectra of at least 7 batches of qualified essence samples;
The spectrogram that step 13 will collect qualified essence sample adopts S-G smoothly to be reconstructed ultraviolet fingerprint in conjunction with first order derivative;
Step 14 is set up the ultraviolet fingerprint principal component analysis model of qualified essence sample with the Principal Component Analysis Method in the Chemical Measurement multivariable analysis, obtains qualified samples SIMCA disaggregated model;
Step 15 in level of significance be calculate under 5% the condition each qualified samples to qualified samples SIMCA disaggregated model apart from Si and qualified samples distance H i to qualified samples SIMCA disaggregated model center.
Compared with prior art, the invention has the beneficial effects as follows and detect the global feature that has reflected flavouring essence for tobacco that the detecting instrument of use is few, method is simple, can field quick detection.
Embodiment
Below the present invention is further elaborated.
Select for use 7 different batches to set up standard database through the essence sample for tobacco of the sample that is up to the standards, calculate according to step 7 to step 11, in level of significance be obtain under 5% the condition all qualified essence sample for tobacco to qualified samples SIMCA disaggregated model apart from Si and each qualified essence sample for tobacco distance H i to qualified samples SIMCA disaggregated model center, result of calculation sees Table 1:
The sample sequence number |
S1 |
S2 |
S3 |
S4 |
S5 |
S6 |
S7 |
Si |
0.00029 |
0.00004 |
0.00017 |
0.00027 |
0.00039 |
0.00048 |
0.00047 |
Hi |
0.0436 |
0.0053 |
0.0266 |
0.0405 |
0.0623 |
0.0766 |
0.0756 |
Table 1
With qualified essence sample for tobacco to qualified samples SIMCA disaggregated model apart from Si peaked three times 0.00144 as the Si upper control limit, qualified essence sample for tobacco to the distance H i at qualified samples SIMCA disaggregated model center peaked three times 0.2299 as Hi upper control limit UCL.
10 batches of sample 1#~10# with to be measured measure according to step 1 to step 3, obtain ultraviolet fingerprint, and according to the Si and the Hi value of the SIMCA disaggregated model of all samples in step 6 calculating and the database, result of calculation sees Table 2 again:
The sample sequence number |
#1 |
#2 |
#3 |
#4 |
#5 |
Si |
0.00081 |
0.00092 |
0.00084 |
0.00065 |
0.00039 |
Hi |
0.0172 |
0.0192 |
0.0176 |
0.0141 |
0.0116 |
The sample sequence number |
#6 |
#7 |
#8 |
#8 |
#10 |
Si |
0.00033 |
0.02723 |
0.02767 |
0.02832 |
0.02804 |
Hi |
0.091 |
4.642 |
4.753 |
5.351 |
5.390 |
Table 2
Adjudicate according to step 6, the Si of 1#~6# sample is all less than Si upper control limit 0.00144, and Hi is all less than Hi upper control limit 0.2299, and 1#~6# sample is qualified.The Si of 7#~10# sample is all greater than Si upper control limit 0.00144, and Hi all beats in Hi upper control limit 0.2299, and 7#~10# sample is defective.
1#~6# sample is deposited into database to the Si and the Hi value of SIMCA disaggregated model, finishes new database, use when being used for detecting from now on, new database sees Table 3:
The sample sequence number |
S1 |
S2 |
S3 |
S4 |
S5 |
S6 |
S7 |
Si |
0.00029 |
0.00004 |
0.00017 |
0.00027 |
0.00039 |
0.00048 |
0.00047 |
Hi |
0.0436 |
0.0053 |
0.0266 |
0.0405 |
0.0623 |
0.0766 |
0.0756 |
The sample sequence number |
S8 |
S9 |
S10 |
S11 |
S12 |
S13 |
|
Si |
0.00081 |
0.00092 |
0.00084 |
0.00065 |
0.00039 |
0.00033 |
|
Hi |
0.0172 |
0.0192 |
0.0176 |
0.0141 |
0.0116 |
0.091 |
|
Table 3