CN107860739A - A kind of tealeaves mid-infrared light profile classification method of fuzzy K mediations network clustering - Google Patents
A kind of tealeaves mid-infrared light profile classification method of fuzzy K mediations network clustering Download PDFInfo
- Publication number
- CN107860739A CN107860739A CN201711205752.1A CN201711205752A CN107860739A CN 107860739 A CN107860739 A CN 107860739A CN 201711205752 A CN201711205752 A CN 201711205752A CN 107860739 A CN107860739 A CN 107860739A
- Authority
- CN
- China
- Prior art keywords
- tealeaves
- fuzzy
- sample
- mediations
- mid
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 32
- 238000012360 testing method Methods 0.000 claims abstract description 31
- 238000002329 infrared spectrum Methods 0.000 claims abstract description 24
- 238000000513 principal component analysis Methods 0.000 claims abstract description 14
- 238000004458 analytical method Methods 0.000 claims abstract description 8
- 238000012937 correction Methods 0.000 claims abstract description 8
- 238000001514 detection method Methods 0.000 claims abstract description 6
- 241001122767 Theaceae Species 0.000 claims abstract 2
- 238000001228 spectrum Methods 0.000 claims description 11
- 238000012549 training Methods 0.000 claims description 10
- 238000004611 spectroscopical analysis Methods 0.000 claims description 8
- 230000003595 spectral effect Effects 0.000 claims description 6
- 238000007906 compression Methods 0.000 claims description 4
- 230000006835 compression Effects 0.000 claims description 4
- 239000000284 extract Substances 0.000 claims description 4
- 244000269722 Thea sinensis Species 0.000 description 16
- 235000013616 tea Nutrition 0.000 description 14
- 235000017166 Bambusa arundinacea Nutrition 0.000 description 5
- 235000017491 Bambusa tulda Nutrition 0.000 description 5
- 241001330002 Bambuseae Species 0.000 description 5
- 235000015334 Phyllostachys viridis Nutrition 0.000 description 5
- 241000270295 Serpentes Species 0.000 description 5
- 239000011425 bamboo Substances 0.000 description 5
- 238000012545 processing Methods 0.000 description 4
- 150000001875 compounds Chemical class 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- XEEYBQQBJWHFJM-UHFFFAOYSA-N Iron Chemical compound [Fe] XEEYBQQBJWHFJM-UHFFFAOYSA-N 0.000 description 2
- RYYVLZVUVIJVGH-UHFFFAOYSA-N caffeine Chemical compound CN1C(=O)N(C)C(=O)C2=C1N=CN2C RYYVLZVUVIJVGH-UHFFFAOYSA-N 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 235000009569 green tea Nutrition 0.000 description 2
- 241001269238 Data Species 0.000 description 1
- 238000001157 Fourier transform infrared spectrum Methods 0.000 description 1
- LPHGQDQBBGAPDZ-UHFFFAOYSA-N Isocaffeine Natural products CN1C(=O)N(C)C(=O)C2=C1N(C)C=N2 LPHGQDQBBGAPDZ-UHFFFAOYSA-N 0.000 description 1
- 206010028980 Neoplasm Diseases 0.000 description 1
- ISWSIDIOOBJBQZ-UHFFFAOYSA-N Phenol Chemical compound OC1=CC=CC=C1 ISWSIDIOOBJBQZ-UHFFFAOYSA-N 0.000 description 1
- 238000000862 absorption spectrum Methods 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 229940024606 amino acid Drugs 0.000 description 1
- 150000001413 amino acids Chemical class 0.000 description 1
- 230000001093 anti-cancer Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000008280 blood Substances 0.000 description 1
- 210000004369 blood Anatomy 0.000 description 1
- 229960001948 caffeine Drugs 0.000 description 1
- VJEONQKOZGKCAK-UHFFFAOYSA-N caffeine Natural products CN1C(=O)N(C)C(=O)C2=C1C=CN2C VJEONQKOZGKCAK-UHFFFAOYSA-N 0.000 description 1
- 201000011510 cancer Diseases 0.000 description 1
- 238000013144 data compression Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000012850 discrimination method Methods 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 235000013305 food Nutrition 0.000 description 1
- 125000000524 functional group Chemical group 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 229910052742 iron Inorganic materials 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 150000002894 organic compounds Chemical class 0.000 description 1
- 230000010355 oscillation Effects 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 230000000144 pharmacologic effect Effects 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 239000013598 vector Substances 0.000 description 1
- 229940088594 vitamin Drugs 0.000 description 1
- 229930003231 vitamin Natural products 0.000 description 1
- 235000013343 vitamin Nutrition 0.000 description 1
- 239000011782 vitamin Substances 0.000 description 1
- 150000003722 vitamin derivatives Chemical class 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
- G01N21/35—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light
- G01N21/3563—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light for analysing solids; Preparation of samples therefor
Landscapes
- Physics & Mathematics (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention discloses a kind of middle infrared spectrum of the tealeaves mid-infrared light profile classification method, first collection different cultivars tealeaves of fuzzy K mediations network clustering;Secondly, the middle infrared spectrum of tealeaves sample is pre-processed using multiplicative scatter correction, principal component analysis and linear discriminant analysis;Finally to the test sample comprising authentication information, the local tea variety in a kind of fuzzy K mediations network clustering method differential test sample is used.Fuzzy K mediation network clusterings are incorporated into the learning rate and more new strategy of Kohonen clustering networks by the present invention, are solved the problems, such as sensitive to initial classes center with Fuzzy Kohonen Clustering Network method and are caused cluster result unstable.The present invention has the advantages that detection speed is fast, and Detection accuracy is high, green non-pollution, and testing result is stable.
Description
Technical field
The invention belongs to field of artificial intelligence, especially a kind of tealeaves mid-infrared light of fuzzy K mediations network clustering
Profile classification method.
Background technology
China is the native place of tealeaves, and our people just has the custom drunk tea since ancient times.The polyatomic phenol that contains in tealeaves,
Caffeine, amino acid, vitamin etc. can play health care and pharmacological action, such as:Green tea helps anti-cancer and cancer-preventing, reduces blood fat etc.
Effect.With the continuous improvement of living standards of the people, the requirement more and more higher to tea leaf quality, and some low qualities, with secondary
The tealeaves substituted the bad for the good compromises consumer's interests.Therefore, the important topic that constant value must be studied that discerns between right and wrong of local tea variety, and design
A kind of simple and quick local tea variety discrimination method is very important.
Middle infrared spectrum detection technique as a kind of Fast nondestructive evaluation technology, in recent years applied to food, agricultural product and
In the Nondestructive Detections such as medicine.The wave-number range of middle infrared spectrum is in 4000cm-1~400cm-1Between, it is most inorganic
The fundamental frequency of the chemical bond oscillations of compound and organic compound is in this region.Functional group, the class of compound in different molecules
Other and compound stereochemical structure, its mid infrared absorption spectrum are not quite similar.The tealeaves of different cultivars, its component and content are often
Difference be present, then their middle infrared spectrum has differences.Around this principle, tea can be realized with mid-infrared light spectral technology
The classification of leaf kind.
Fuzzy Kohonen Clustering Network (Tsao E C, Bezdek J C, Pal N R.Fuzzy Kohonen
clustering networks.Pattern Recognition,1994,27(5):757-764.) it is a kind of unsupervised
Learning method.Fuzzy Kohonen Clustering Network is the study that Fuzzy C-Means Clustering (FCM) is incorporated into Kohonen clustering networks
In speed and more new strategy.But due to FCM, there is cause cluster result unstable initial classes center tender subject.
Introduce FCM Fuzzy Kohonen Clustering Network there is also it is identical with FCM the problem of.
The content of the invention
The present invention be directed to existing Fuzzy Kohonen Clustering Network method, in cluster data, there is in initial classes
Heart tender subject and the shortcomings that cause cluster result unstable, propose that a kind of fuzzy K reconciles the tealeaves mid-infrared light of network clustering
Profile classification method.Compared to original Fuzzy Kohonen Clustering Network method, a kind of fuzzy K mediations network clustering of the invention
Tealeaves mid-infrared light profile classification method calculates learning rate using the fuzzy membership of fuzzy K mediations cluster.The present invention has inspection
The advantages that degree of testing the speed is fast, and Detection accuracy is high, green non-pollution, and testing result is stable.
The present invention is achieved through the following technical solutions above-mentioned technical purpose.
A kind of tealeaves mid-infrared light profile classification method of fuzzy K mediations network clustering, comprises the following steps:
S1, gather tealeaves sample middle infrared spectrum:For the tealeaves sample of different cultivars, with infrared spectrometer to tealeaves sample
This is detected, and obtains the infrared spectrum information that diffuses in tealeaves sample, spectral information is stored in computer;By tealeaves sample
Originally it is divided into training sample and test sample;
S2, the successively mid-infrared light using multiplicative scatter correction, principal component analysis and linear discriminant analysis to tealeaves sample
Spectrum is pre-processed;
S3, the test sample to including authentication information in step 2, differentiated using a kind of fuzzy K mediations network clustering method
Local tea variety in test sample.
Further, in the S1 the infrared spectrum information that diffuses refer to the wave-number range of spectrum for 4001.569~
401.1211cm-1, the spectrum for collecting each tealeaves sample is the data of 1868 dimensions.
Further, the classification number for tealeaves sample being set in the S1 is k, number of training nr, test sample number is n.
Further, the S2 is specially:First tealeaves sample middle infrared spectrum is handled with multiplicative scatter correction, then
Spectroscopic data compression is carried out with principal component analysis, the authentication information that spectroscopic data is finally carried out with linear discriminant analysis extracts.
Further, the S3 is specially:
S3.1, initialization:Determine classification number k, test sample number n and weighted index m0Value, n>k>1 ,+∞>m0>1;If
Put iterations initial value r, maximum iteration rmaxAnd worst error parameter ε;It is determined that the initial classes center c of clusterj,0;
S3.2, calculate fuzzy membership angle value u during the r times iterationij,r;
S3.3, calculate learning rate α during the r times iterationij,r;
S3.4, calculate class center c during the r times iterationj,r;
S3.5, when | | cj,r-cj, r-1 | | < ε or r>Rmax then calculates termination, is otherwise recalculated since S3.2,
Wherein cj,rThe class center of -1 jth class when being the r-1 times iteration.
Further, fuzzy membership the angle value uij, r during the r times iteration are calculated:Wherein
Weighted index when mr is the r times iteration, mr=m0- r Δ m, Δ m=(m0-1)/rmax;uij, j-th of sample when r is the r times iteration
Originally it is under the jurisdiction of the fuzzy membership angle value of the i-th class, wherein dij=| | xi-cj, r-1 | |, xiFor i-th of sample data, dit=| | xi-
ct,r-1| |, ct,r-1For the r-1 times iteration when t classes class center.
Further, learning rate α during the r times iteration is calculatedij,r:
Further, class center c during the r times iteration is calculatedj,r:Wherein dil
=| | xi-cl,r-1| |, cl,r-1For the r-1 times iteration when l classes class center;αil,rFor the r times iteration when learning rateuil,rFor the r times iteration when l-th of sample be under the jurisdiction of the fuzzy membership angle value of the i-th class..
Beneficial effects of the present invention are:
1st, a kind of tealeaves mid-infrared light profile classification method of fuzzy K mediations network clustering of the invention is due to calculating the r times
Class center c during iterationj,r, so the stability and cluster accuracy rate in terms of tealeaves mid-infrared light modal data is clustered are better than
Fuzzy Kohonen Clustering Network method, there is the advantages of cluster accuracy rate is high, and cluster speed is fast, and cluster result is stable.
2nd, the present invention is by calculating fuzzy membership angle value and being classified with it, so the present invention includes noise number in cluster
According to mid-infrared light modal data in terms of be better than Kohonen clustering networks, can quickly realize local tea variety with reference to China and foreign countries' spectral technique
Quick and accurate discriminating.
Brief description of the drawings
Fig. 1 is the flow chart of the present invention;
Fig. 2 is the mid-infrared light spectrogram of tealeaves;
Fig. 3 is the tealeaves mid-infrared light spectrogram after MSC processing;
Fig. 4 is the test sample datagram that the middle infrared spectrum of tealeaves obtains after LDA Extraction and discrimination information;
Fig. 5 is initial fuzzy membership angle value figure;
Fig. 6 is fuzzy membership angle value figure caused by a kind of fuzzy K mediations network clustering method;
Fig. 7 is learning rate figure caused by a kind of fuzzy K mediations network clustering method.
Embodiment
Below in conjunction with accompanying drawing, the present invention is described further, but protection scope of the present invention is not limited to this.
Step 1: the collection of tealeaves sample middle infrared spectrum:The tealeaves sample of different cultivars is gathered, with infrared spectrometer pair
Tealeaves sample is detected, and obtains the infrared spectrum information that diffuses in tealeaves sample, spectral information is stored in computer.
Indoor temperature and humidity is kept to be basically unchanged as far as possible in experimentation;In the wave-number range of infrared diffusing reflection spectrum be
4001.569~401.1211cm-1, the spectrum for collecting each tealeaves sample is the data of 1868 dimensions.Tealeaves sample is divided into
Training sample and test sample.Classification number k, number of training n are setrIt is n with test sample number.
Step 2: the middle infrared spectrum of tealeaves sample is pre-processed:First spectrum is entered with multiplicative scatter correction (MSC)
Row processing, then carries out spectroscopic data compression with principal component analysis (PCA), finally carries out spectroscopic data with linear discriminant analysis
Authentication information extracts.
Step 3: to included in step 2 the test sample of authentication information with a kind of fuzzy K mediation network clustering method with
Local tea variety in differential test sample.
It is described in detail below:
1. initialization
(1) classification number k, test sample number n and weighted index m are determined0Value, and meet n>k>1 ,+∞>m0>1;Set
Iterations initial value r, maximum iteration rmaxAnd iteration worst error parameter ε;
(2) the initial classes center c of cluster is determinedj,0。
2. calculate fuzzy membership angle value u during the r times iterationij,r
Wherein:mrFor the r times iteration when weighted index, mr=m0- r Δ m, Δ m=(m0-1)/rmax;uij,rFor the r times
J-th of sample is under the jurisdiction of the fuzzy membership angle value of the i-th class, wherein d during iterationij=| | xi-cj,r-1| |, xiFor i-th of sample number
According to cj,r-1For the r-1 times iteration when jth class class center, dit=| | xi-ct,r-1| |, ct,r-1For the r-1 times iteration when t classes
Class center.
3. calculate learning rate α during the r times iterationij,r:
4. calculate class center c during the r times iterationj,r:
Wherein dil=| | xi-cl,r-1| |, cl,rThe class center of -1 l classes when being the r-1 times iteration;αil,rFor the r times iteration
When learning rate, anduil,L-th of sample is under the jurisdiction of the fuzzy membership of the i-th class when r is the r times iteration
Angle value.
(r is assigned to variable r) to 5.r=r+1 after increasing by 1, r+1
When | | cj,r-cj,r-1| | < ε or r>rmaxThen calculate and terminate, otherwise from " 2. calculate fuzzy person in servitude during the r times iteration
Belong to angle value uij,r" start to recalculate.
A kind of tealeaves mid-infrared light profile classification method of fuzzy K mediations network clustering of the present invention is applied to local tea variety
Discriminating, such as:The discriminating of the local tea varieties such as high-quality green tea, green bamboo snake, Dragon Well tea, Iron Guanyin.Because different cultivars tealeaves, its internal group
Point difference, therefore its middle infrared spectrum is there is also difference, to realize that the discriminating of local tea variety provides condition.The implementation of the present invention
Flow chart is as shown in Figure 1.For convenience of narration, Mount Emei's tealeaves, the high-quality green bamboo snake in Leshan and green bamboo snake inferior are chosen as experiment
Object.
Embodiment:
Step 1:Tealeaves sample middle infrared spectrum gathers:The start of FTIR-7600 type FTIR spectrums analyzer is pre-
Hot 1 hour, scanning times 32, wave number 4001.569cm-1~401.1211cm-1 of spectral scan, sweep spacing are
1.928cm-1, resolution ratio 4cm-1;Tealeaves sample:Mount Emei's tealeaves, the high-quality green bamboo snake in Leshan and green bamboo snake inferior;Tealeaves
Ground crushing, then after being filtered with 40 mesh sieves, respectively take 0.5g respectively with KBr 1:100 uniformly mixing;Each sample takes
Mixture 1g carries out press mold, is then scanned 3 times with spectrometer, takes the average value of 3 times as sample spectrum data;When gathering spectrum
Environment temperature and relative humidity keep relative stability;Every kind of tealeaves gathers 32 samples, obtains 96 samples altogether.Each sample is
The data of one 1868 dimension;It is test set that the tealeaves sample of each kind, which chooses 22, then test sample number n is 66;Residue 10
Individual sample is training set, then number of training nr is 30;Test set is tealeaves sample to be identified, and training set is known kind
Tealeaves sample, setting classification number k=3, the middle infrared spectrum of tealeaves sample are as shown in Figure 2.
Step 2:The middle infrared spectrum of tealeaves sample is pre-processed:First spectrum is entered with multiplicative scatter correction (MSC)
Row processing, then carries out spectroscopic data compression with principal component analysis (PCA), finally carries out spectroscopic data with linear discriminant analysis
Authentication information extracts.
Multiplicative scatter correction (MSC) is carried out to the tealeaves middle infrared spectrum of 96 samples, its result is as shown in Figure 3;Then
Data compression after MSC is handled with principal component analysis (PCA) is as follows to 14 dimensions, PCA 14 characteristic value concrete numerical values:λ1=
293.91, λ2=129.02, λ3=19.00, λ4=14.88, λ5=6.43, λ6=3.82, λ7=2.00, λ8=1.43, λ9=
1.07 λ10=0.63, λ11=0.40, λ12=0.32, λ13=0.27, λ14=0.23.
Tealeaves sample middle infrared spectrum is projected in PCA 14 characteristic vectors and obtains the data of 14 dimensions, i.e., from 1868
Dimension is compressed to 14 dimensions.
It is 2 to set discriminant vectorses number, the mirror of 14 dimension datas obtained using linear discriminant analysis (LDA) extraction PCA processing
Obtain including the training sample and test sample data of authentication information after other information, wherein test sample data are as shown in Figure 4.
Step 3: to included in step 2 the test sample of authentication information with a kind of fuzzy K mediation network clustering method with
Local tea variety in differential test sample.
It is described in detail below:
1. initialization
(1) classification number k, test sample number n and weighted index m are determined0Value, and meet n>k>1 ,+∞>m0>1;Setting changes
Generation number initial value r=0 and maximum iteration are rmax;Iteration worst error parameter ε is set;
The numerical value of initialization is set:From step 1:Classification number k=3 (i.e. three classifications), test sample number n=66,
Weighted index m is set0=2, iterations initial value r=0 and greatest iteration number rmax=100, error higher limit ε=0.005.
(2) the initial classes center c of cluster is determinedj,0;
To the test data operation fuzzy C-means clustering (FCM) in step 2 after LDA is handled, after FCM iteration ends
Cluster centre reconcile the initial cluster center of network clustering method as a kind of fuzzy K, then initial cluster center cj,0For:c1,0
=(- 0.1548,0.0388), c2,0=(0.0034,0.0022), c3,0=(0.1204, -0.0056);Initial fuzzy membership
Angle value uij,0As shown in Figure 5.
2. fuzzy membership angle value u during the r times iteration is calculated using formula (1)ij,r, experimental result:(this during iteration ends
When r=3) when fuzzy membership angle value uij,3As shown in Figure 6.
3. learning rate during the r times iteration, experimental result are calculated using formula (2):During iteration ends (now r=3)
When learning rate αij,3As shown in Figure 7.
4. class center c during the r times iteration is calculated using formula (3)j,r, experimental result:R=3 during iteration ends, cj,3
For:
c1,3=[- 0.16240.0344];c2,3=[0.01010.0037];c3,3=[0.11760.0011].
5.r=r+1
When | | cj,r-cj,r-1| | < ε or r>rmaxThen calculate and terminate, otherwise from " fuzzy during 2. the r times iteration of calculating
It is subordinate to angle value uij,r" start to recalculate.
According to fuzzy membership angle value uij,3Realize that local tea variety differentiates, differentiate rate of accuracy reached to 93.94%.
It is described above that the present invention is briefly described, not by above-mentioned working range limit value, as long as taking the present invention
Thinking and method of work carry out simple modification and apply to other equipment, or make and changing in the case where not changing central scope principle of the present invention
Enter and retouch etc. behavior, within protection scope of the present invention.
Claims (8)
1. a kind of tealeaves mid-infrared light profile classification method of fuzzy K mediations network clustering, it is characterised in that comprise the following steps:
S1, gather tealeaves sample middle infrared spectrum:For the tealeaves sample of different cultivars, tealeaves sample is entered with infrared spectrometer
Row detection, obtains the infrared spectrum information that diffuses in tealeaves sample, spectral information is stored in computer;By tealeaves sample point
For training sample and test sample;
S2, the middle infrared spectrum of tealeaves sample is entered using multiplicative scatter correction, principal component analysis and linear discriminant analysis successively
Row pretreatment;
S3, the test sample to including authentication information in step 2, use a kind of fuzzy K mediations network clustering method differential test
Local tea variety in sample.
2. a kind of tealeaves mid-infrared light profile classification method of fuzzy K mediations network clustering as claimed in claim 1, its feature exist
In the infrared spectrum information that diffuses refers to that the wave-number range of spectrum is 4001.569~401.1211cm in the S1-1, collect
The spectrum of each tealeaves sample is the data of 1868 dimensions.
3. a kind of tealeaves mid-infrared light profile classification method of fuzzy K mediations network clustering as claimed in claim 1, its feature exist
In the classification number that tealeaves sample is set in the S1 is k, number of training nr, test sample number is n.
4. a kind of tealeaves mid-infrared light profile classification method of fuzzy K mediations network clustering as claimed in claim 1, its feature exist
In the S2 is specially:First tealeaves sample middle infrared spectrum is handled with multiplicative scatter correction, then uses principal component analysis
Spectroscopic data compression is carried out, the authentication information that spectroscopic data is finally carried out with linear discriminant analysis extracts.
5. a kind of tealeaves mid-infrared light profile classification method of fuzzy K mediations network clustering as claimed in claim 1, its feature exist
In the S3 is specially:
S3.1, initialization:Determine classification number k, test sample number n and weighted index m0Value, n>k>1 ,+∞>m0>1;Setting changes
Generation number initial value r, maximum iteration rmaxAnd worst error parameter ε;It is determined that the initial classes center c of clusterj,0;
S3.2, calculate fuzzy membership angle value u during the r times iterationij,r;
S3.3, calculate learning rate α during the r times iterationij,r;
S3.4, calculate class center c during the r times iterationj,r;
S3.5, r=r+1, when | | cj,r-cj,r-1| | < ε or r>rmaxThen calculate and terminate, otherwise recalculated since S3.2,
Wherein cj,r-1For the r-1 times iteration when jth class class center.
6. a kind of tealeaves mid-infrared light profile classification method of fuzzy K mediations network clustering as claimed in claim 5, its feature exist
In, calculate the r times iteration when fuzzy membership angle value uij,r:Wherein mrFor the r times iteration when
Weighted index, mr=m0- r Δ m, Δ m=(m0-1)/rmax;uij,rFor the r times iteration when j-th of sample be under the jurisdiction of the mould of the i-th class
Paste is subordinate to angle value, wherein dij=| | xi-cj,r-1| |, xiFor i-th of sample data, dit=| | xi-ct,r-1| |, ct,r-1For r-1
The class center of t classes during secondary iteration.
7. a kind of tealeaves mid-infrared light profile classification method of fuzzy K mediations network clustering as described in claim 5 or 6, it is special
Sign is, calculates learning rate α during the r times iterationij,r:
8. a kind of tealeaves mid-infrared light profile classification method of fuzzy K mediations network clustering as described in claim 5 or 6, it is special
Sign is, calculates class center c during the r times iterationj,r:Wherein dil=| | xi-
cl,r-1| |, cl,rThe class center of -1 l classes when being the r-1 times iteration;αil,rFor the r times iteration when learning rate, anduil,L-th of sample is under the jurisdiction of the fuzzy membership angle value of the i-th class when r is the r times iteration.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711205752.1A CN107860739A (en) | 2017-11-27 | 2017-11-27 | A kind of tealeaves mid-infrared light profile classification method of fuzzy K mediations network clustering |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711205752.1A CN107860739A (en) | 2017-11-27 | 2017-11-27 | A kind of tealeaves mid-infrared light profile classification method of fuzzy K mediations network clustering |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107860739A true CN107860739A (en) | 2018-03-30 |
Family
ID=61702890
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711205752.1A Pending CN107860739A (en) | 2017-11-27 | 2017-11-27 | A kind of tealeaves mid-infrared light profile classification method of fuzzy K mediations network clustering |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107860739A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111881738A (en) * | 2020-06-19 | 2020-11-03 | 江苏大学 | Tea near infrared spectrum classification method based on nuclear fuzzy orthogonal discriminant analysis |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103954582A (en) * | 2014-04-11 | 2014-07-30 | 江苏大学 | Apple cultivar near-infrared-spectrum sorting method based on hybrid K-harmonic means clustering |
CN106408012A (en) * | 2016-09-09 | 2017-02-15 | 江苏大学 | Tea infrared spectrum classification method of fuzzy discrimination clustering |
CN106570520A (en) * | 2016-10-21 | 2017-04-19 | 江苏大学 | Infrared spectroscopy tea quality identification method mixed with GK clustering |
CN107192686A (en) * | 2017-04-11 | 2017-09-22 | 江苏大学 | A kind of Possibility Fuzzy Clustering local tea variety discrimination method of fuzzy covariance matrix |
CN107271394A (en) * | 2017-05-16 | 2017-10-20 | 江苏大学 | A kind of fuzzy Kohonen differentiates the tealeaves infrared spectrum sorting technique of clustering network |
-
2017
- 2017-11-27 CN CN201711205752.1A patent/CN107860739A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103954582A (en) * | 2014-04-11 | 2014-07-30 | 江苏大学 | Apple cultivar near-infrared-spectrum sorting method based on hybrid K-harmonic means clustering |
CN106408012A (en) * | 2016-09-09 | 2017-02-15 | 江苏大学 | Tea infrared spectrum classification method of fuzzy discrimination clustering |
CN106570520A (en) * | 2016-10-21 | 2017-04-19 | 江苏大学 | Infrared spectroscopy tea quality identification method mixed with GK clustering |
CN107192686A (en) * | 2017-04-11 | 2017-09-22 | 江苏大学 | A kind of Possibility Fuzzy Clustering local tea variety discrimination method of fuzzy covariance matrix |
CN107271394A (en) * | 2017-05-16 | 2017-10-20 | 江苏大学 | A kind of fuzzy Kohonen differentiates the tealeaves infrared spectrum sorting technique of clustering network |
Non-Patent Citations (1)
Title |
---|
赵恒 等: "《模糊K-HarmonicMeans聚类算法》", 《西安电子科技大学学报(自然科学版)》 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111881738A (en) * | 2020-06-19 | 2020-11-03 | 江苏大学 | Tea near infrared spectrum classification method based on nuclear fuzzy orthogonal discriminant analysis |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11656176B2 (en) | Near-infrared spectroscopy-based method for chemical pattern recognition of authenticity of traditional Chinese medicine Gleditsiae spina | |
Cai et al. | Using FTIR spectra and pattern recognition for discrimination of tea varieties | |
CN101881726B (en) | Nondestructive detection method for comprehensive character living bodies of plant seedlings | |
CN104374738B (en) | A kind of method for qualitative analysis improving identification result based on near-infrared | |
CN105181650B (en) | A method of quickly differentiating local tea variety using near-infrared spectrum technique | |
CN110082298B (en) | Hyperspectral image-based wheat variety gibberellic disease comprehensive resistance identification method | |
CN104990892B (en) | The spectrum picture Undamaged determination method for establishing model and seeds idenmtification method of seed | |
CN106408012A (en) | Tea infrared spectrum classification method of fuzzy discrimination clustering | |
CN108872132A (en) | A method of fresh tea leaves kind is differentiated using near infrared spectrum | |
CN108734205A (en) | A kind of simple grain for different cultivars wheat seed pinpoints identification technology | |
CN102012365A (en) | Tea fermentation degree identification method based on infrared spectrum | |
Dong et al. | Deep learning for geographical discrimination of Panax notoginseng with directly near-infrared spectra image | |
Chen et al. | Fast detection of cumin and fennel using NIR spectroscopy combined with deep learning algorithms | |
CN110378374A (en) | A kind of tealeaves near infrared light profile classification method that fuzzy authentication information extracts | |
CN110363125A (en) | Using the method for Model Transfer identification different cultivars Citrus Huanglongbing pathogen | |
CN105138834A (en) | Tobacco chemical value quantifying method based on near-infrared spectrum wave number K-means clustering | |
CN110376202A (en) | Tea tree anthracnose scab recognition methods based on imaging hyperspectral technique | |
CN109685098A (en) | The local tea variety classification method of cluster is separated between a kind of Fuzzy Cluster | |
CN107271394A (en) | A kind of fuzzy Kohonen differentiates the tealeaves infrared spectrum sorting technique of clustering network | |
CN107192686A (en) | A kind of Possibility Fuzzy Clustering local tea variety discrimination method of fuzzy covariance matrix | |
CN107860739A (en) | A kind of tealeaves mid-infrared light profile classification method of fuzzy K mediations network clustering | |
CN106570520A (en) | Infrared spectroscopy tea quality identification method mixed with GK clustering | |
CN110186871A (en) | A kind of method of discrimination in the fresh tea leaves place of production | |
CN107886115A (en) | A kind of tealeaves mid-infrared light profile classification method of adaptively possible C mean clusters | |
CN109358022A (en) | A kind of method of the quick-fried pearl type of quick discrimination cigarette |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180330 |