CN107462607A - A kind of detection method of the fermentation of black tea appropriateness based on electrical characteristic parameter - Google Patents
A kind of detection method of the fermentation of black tea appropriateness based on electrical characteristic parameter Download PDFInfo
- Publication number
- CN107462607A CN107462607A CN201710828757.3A CN201710828757A CN107462607A CN 107462607 A CN107462607 A CN 107462607A CN 201710828757 A CN201710828757 A CN 201710828757A CN 107462607 A CN107462607 A CN 107462607A
- Authority
- CN
- China
- Prior art keywords
- fermentation
- characteristic parameter
- electrical characteristic
- black tea
- appropriateness
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N27/00—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N27/00—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means
- G01N27/02—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating impedance
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N27/00—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means
- G01N27/02—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating impedance
- G01N27/04—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating impedance by investigating resistance
- G01N27/041—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating impedance by investigating resistance of a solid body
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N27/00—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means
- G01N27/02—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating impedance
- G01N27/22—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating impedance by investigating capacitance
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N30/00—Investigating or analysing materials by separation into components using adsorption, absorption or similar phenomena or using ion-exchange, e.g. chromatography or field flow fractionation
- G01N30/02—Column chromatography
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
Landscapes
- Chemical & Material Sciences (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Pathology (AREA)
- Biochemistry (AREA)
- Immunology (AREA)
- General Health & Medical Sciences (AREA)
- Analytical Chemistry (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical Kinetics & Catalysis (AREA)
- Electrochemistry (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Measuring Or Testing Involving Enzymes Or Micro-Organisms (AREA)
- Investigating Or Analyzing Materials By The Use Of Electric Means (AREA)
Abstract
The present invention relates to technical field of food detection, a kind of more particularly to detection method of the fermentation of black tea appropriateness based on electrical characteristic parameter, it is that the tealeaves of Various Fermenting Degree is respectively placed in the measure that electrical characteristic parameter is carried out in electrical characteristics test system, detect the index of quality simultaneously, carry out grade demarcation, pretreatment is standardized to electrical characteristic parameter, screen the preferred feature frequency that suitable fermentation quality differentiates, filter out preferred feature electrical characteristic parameter, carry out principal component analysis, so as to establish prediction black tea dynamic fermentation appropriateness model, for realizing that the prediction to appropriate grade of fermenting differentiates.The detection method of the present invention overcomes the deficiency of the methods of artificial sense evaluation and Physico-chemical tests, realize standardization, stabilisation, the intellectuality moderately detected to fermentation of black tea, rapid and precision, had important practical significance and application prospect in the detection of fermentation of black tea appropriateness.
Description
Technical field
The present invention relates to technical field of food detection, more particularly to a kind of fermentation of black tea appropriateness based on electrical characteristic parameter
Detection method.
Background technology
China is the area of origin of tea tree, is Chan Cha states the biggest in the world.Tea industry is in the Belt and Road national strategy
Environmental health industry in traditional advantage characteristic export-oriented industry, and the strategy of Health China 2030.Spy of the black tea as China
There is tea, important position is occupied in international tea market.Fermentation is the most key process in black tea processing, directly decision into
The quality and flavor characteristic sampled tea.During the fermentation, can occur after tea cell film is impaired based on polyphenol compound
Enzymatic oxidation reaction and association response, generate based on theaflavins (TFs), thearubigin class (TRs) and theabrownin class (TBs)
A series of organic pigment class materials, synthesis form the distinctive color and luster of black tea, taste and aroma quality.Only appropriate fermentation
The black tea of high-quality could be obtained, it is thus determined that fermentation appropriateness has vital effect.
The method of currently used control fermentation appropriateness has sense organ determining method, measurement temperture of leaves method, biochemical indicator method, electronic nose
Method etc., while some fermentation instruction equipments are have developed, but these equipment operations are complicated, expensive.Sent out in actual production
The regulation and control of ferment degree are main or processing staff changes to judge with the sense organ of oneself from the sign of fermentated leaves, such as see fermentated leaves
The degree of red change, change of fragrance etc. determine that fermentation is enough to reach appropriateness, and its accuracy is easily by experience and external environment
Influence, cause fermentation partially light or excessively, directly influence product quality, the stability of flavor and uniformity, quality safety risk
Factor is significantly greatly increased.Therefore, fermentation of black tea standardization is realized, stabilizes, is intelligent, quick, accurate judgement fermentation appropriateness has
Important meaning.
Response of the bound charge to extra electric field in biology interior molecule is referred to as dielectric property.Electrical characteristics detection technique is
The inner link with component content is established using the change of test substance electromagnetic property, and then realizes food quality or the inspection of material attribute
A kind of fast non-destructive detection method surveyed, its major parameter include：Its major parameter includes：Electric capacity, resistance, reactance, loss because
Son and impedance etc..Uneven electrolyte of the tealeaves between conductor and insulator, when black tea fermentation process sample is placed in electric field
When middle, its internal charged particle can form bio-electric field, there is the effect such as obstruction, receiving, loss, conduction to electric field charge, with
The change of fermented sample endoplasm composition, structure and physiological status, the electrically charged and quantity of electric charge of chemical substance in tissue can be caused
Changed in terms of spatial distribution and intensity, macroscopically show the change of electrical parameter.Electrical characteristic parameter extensive use
In quick, the lossless maturity for judging fruits and vegetables, cereal and meat products, nutrition condition etc., but its application in tealeaves research
Then report is few.And to having no report in the application of the detection of product property appropriateness during fermentation of black tea.
The content of the invention
In view of this, it is an object of the invention to provide a kind of detection side of the fermentation of black tea appropriateness based on electrical characteristic parameter
Method, it is intended to overcome the shortcomings of artificial sense evaluation and the methods of Physico-chemical tests, realize fermentation of black tea is moderately detected standardization,
Stabilize, be intelligent, rapid and precision, having important practical significance and answer in the on-line checking of fermentation of black tea appropriateness
Use prospect.
The present invention solves above-mentioned technical problem by following technological means：
A kind of detection method of the fermentation of black tea appropriateness based on electrical characteristic parameter, the tealeaves of Various Fermenting Degree is put respectively
The measure of electrical characteristic parameter is carried out in electrical characteristics test system, while detects the index of quality, grade demarcation is carried out, to electrical characteristics
Parameter carries out Screening Treatment, so as to establish prediction black tea dynamic fermentation appropriateness model, for realizing to the pre- of appropriate grade of fermenting
Survey and differentiate, comprise the following steps：
S1. the structure of electrical characteristics test system, the electrical characteristics test system include bridge test instrument, test electrode, taken
Sample box, operation main frame and acquisition software, are provided with disperser on the sampling box, the bridge test instrument by R232 serial ports with
Run main frame and carry out data exchange, and two test electrodes are connected by twisted-pair shielded wire；
S2. the measure of sample electrical characteristic parameter, the Tea Samples of different fermentations time are dissipated naturally by disperser respectively
Drop into sampling box, and Tea Samples is flooded test electrode, respectively at measuring electrical characteristic parameter under different frequency；
S3. fermentation quality detection and grade demarcation, the index of quality of different fermentations time Tea Samples is detected, and to fermentation
Quality carries out grade demarcation；
S4. data screening processing and Model checking, pretreatment are standardized to electrical characteristic parameter, using based on phase relation
The characteristic frequency composite index analysis method of number and principal component selects the preferred feature frequency that suitable fermentation quality differentiates, using nothing
Information variable null method-competitive adaptive weighting sampling method, which is combined, filters out preferred feature electrical characteristic parameter, to preferred feature
Each electrical characteristic parameter and preferred feature electrical characteristic parameter under frequency carries out PCA analyses respectively, using optimal principal component as input
Operating limit learning machine pattern discrimination is measured, so as to establish the dynamic fermentation appropriateness detection model of black tea.
S5. the quick detection of black tea dynamic fermentation appropriateness, computer client, real-time Transmission fermentation are connected to by data wire
The electrical characteristic parameter data of sample, the fermentation appropriateness detection model write with reference to acquisition software, are analyzed in real time at software interface end
The fermentation appropriateness grade of fermentation sample, realize the qualitative quick detection of black tea dynamic fermentation appropriateness.
Further, the test frequency scope of the bridge test instrument is 50Hz~200kHz, and the test electrode is copper
Parallel plate electrode plate, the area of the battery lead plate is 20cm2, the bottom plate side of the sampling box is hinged with side plate, is corresponding
Opposite side is detachably connected with corresponding side plate, and the disperser includes motor, rotating shaft and the moving tooth being fastened around in rotating shaft,
The rotating shaft is connected with motor, and the rotating speed of the rotating shaft is more than 1000r/min.
The electric parameters testing circuit of bridge test instrument is divided into series, parallel equivalent-circuit model, and usual low impedance element is used
Series connection, from numerically saying that impedance is less than 100 ohm with series connection, and middle impedance minimum value of the present invention is more than 200 ohm, therefore selects simultaneously
Join the electrical parameter under test circuit.
Further, the Tea Samples are tealeaves after kneading in 27 DEG C of temperature, humidity>Sent out in 90% artificial gas tank
Ferment 6h, in the sample that 0~6h different fermentations time samplings obtain.
Further, the electrical characteristic parameter in the S2 includes parallel equivalent electric capacity, complex impedance, resistance, reactance, fissipation factor
With phase loss angle.
Tea leaf fermentation sample can be considered the interior media of capacitor, when electric current is by black tea sample between battery lead plate, i.e. structure
Into capacitor equivalent circuit, by the algebraically processing to detection signal, terminal interface can receive complex impedance Z, parallel equivalent electricity
Hold the electrical characteristic parameter values such as Cp, reactance X, fissipation factor D and phase loss angle θ.
Further, the test condition of electrical characteristic parameter is test voltage 1V, 0.05~200kHz of test frequency in the S2.
Further, the index of quality in the S3 includes sensory evaluation scores, theaflavin, thearubigin and theabrownin, and the tea is yellow
Element, thearubigin and theabrownin are measured using high performance liquid chromatography.
Further, the demarcation of fermentation quality grade is to calculate to evaluate total score based on expert sensory's scoring comment in the S3, is examined
Sample of the total score more than 85 is commented to be assessed as proper fermentation, the sample before proper fermentation sample time node is slight fermentation, it
Sample afterwards is excessive fermentation.
Further, the standardization pretreatment is mainly using multiplicative scatter correction, deviation standardization, Smooth, S/G2st
Electrical characteristic parameter initial data is pre-processed with Zscore methods.
Further, the characteristic frequency composite index analysis method based on coefficient correlation and principal component in the S4 is selected preferably
Characteristic frequency, the correlation matrix of electrical parameter and each fermentation quality index is specially established, each group matrix data is led
Constituent analysis, correlation index is obtained, the array formed to the correlation index of each index of quality at different frequencies carries out PCA points
Analysis, the comprehensive index of correlation is calculated, frequency corresponding to the comprehensive index of correlation of maximum is preferred feature frequency.
Further, it is 0.2kHz that the preferred feature frequency moderately differentiated of suitably fermenting is filtered out in the S4.0.2kHz's
Comprehensive correlation coefficient value is up to 0.274.
Further, the extreme learning machine pattern discrimination is programmed in Matlab softwares, using ELM sorting algorithms, choosing
The activation primitive for being ELM with Sigmoid functions, using the number of principal components evidence of the electrical characteristic parameter extracted, three classes are fermented suitable
The black tea sample of degree carries out discriminant classification.
Beneficial effects of the present invention：
(1) present invention specify that the changing rule of electrical characteristic parameter during the fermentation under each detection frequency.As a result show,
Electrical impedance, reactance, impedance angle and fissipation factor gradually increase in fermentation, and electric capacity is gradually reduced with the extension of fermentation period, is said
Hinder the material of charge transfer more and more in bright fermentation process.
(2) present invention integrates correlation index method by principal component, specify that 0.2kHz can be used as electricity in fermentation of black tea special
The characteristic frequency of property parameter detecting.Variable optimization method based on MCUVE-CARS, it specify that the survey stable with each index of quality
Examination frequency is low-frequency range, and most related electrical parameters are D and X.
(3) present invention is shown the differentiation result of fermentation of black tea quality appropriateness using electrical characteristic parameter combination ELM, ELM moulds
Type estimated performance is more satisfactory, when its number of principal components is 3, when node in hidden layer is 20, and the differentiation of model calibration set and forecast set
Rate reaches 100%, shows to realize to the pre- of fermentation quality appropriateness grade using electrical characteristic parameter combination ELM discrimination models
Survey and differentiate.
(4) electrical characteristics detection technique of the invention is as a kind of mature technology, has that stable, quick, sensitive, cost is low
Advantage, had important practical significance and application prospect in the on-line checking of fermentation of black tea appropriateness.
Brief description of the drawings
Fig. 1 is a kind of flow chart of the detection method of the fermentation of black tea appropriateness based on electrical characteristic parameter of the present invention；
Fig. 2 is electrical characteristics test system in a kind of detection method of the fermentation of black tea appropriateness based on electrical characteristic parameter of the present invention
Structural representation；
Wherein, main frame 1, bridge test instrument 2, test electrode 3, sampling box 4, bottom plate 5, motor 61, rotating shaft 62, moving tooth are run
63；
Fig. 3 is influence curve figure of the test frequency in the range of 0.05kHz~200KHz to fermentation of black tea sample electrical parameter；
Fig. 4 is the procedure chart that MCUVE-CARS screens the feature electrical parameter related to sensory evaluation scores；
Fig. 5 is the three-dimensional shot chart of three PCA principal components under 0.2kHz and preferred feature electrical parameter；
Fig. 6 is differentiation result of the ELM models to fermentation of black tea quality.
Embodiment
Below with reference to the drawings and specific embodiments, the present invention is described in detail：
As shown in figure 1, a kind of detection method of fermentation of black tea appropriateness based on electrical characteristic parameter of the present invention, by different hairs
The tealeaves of ferment degree is respectively placed in the measure that electrical characteristic parameter is carried out in electrical characteristics test system, while detects the index of quality, enters
Row grade is demarcated, and Screening Treatment is carried out to electrical characteristic parameter, so as to establish prediction black tea dynamic fermentation appropriateness model, for realizing
Prediction to appropriate grade of fermenting differentiates, comprises the following steps：
S1. the structure of electrical characteristics test system, electrical characteristics test system include bridge test instrument 2, test electrode 3, sampling
Box 4, operation main frame 1 and acquisition software, acquisition software are mainly used to gathered data, in operation main frame 1.Bridge test instrument 2
Data exchange is carried out by R232 serial ports and operation main frame 1, and two test electrodes 3 are connected by twisted-pair shielded wire.
S2. the measure of sample electrical characteristic parameter, the Tea Samples of different fermentations time is separately added into by disperser and taken
In sample box 4, and test electrode 3 is flooded, respectively at measuring electrical characteristic parameter under different frequency.
S3. fermentation quality detection and grade demarcation, the index of quality of different fermentations time Tea Samples is detected, and to fermentation
Quality carries out grade demarcation.
S4. data screening processing and Model checking, pretreatment are standardized to electrical characteristic parameter, using based on phase relation
The characteristic frequency composite index analysis method of number and principal component selects the preferred feature frequency that suitable fermentation quality differentiates, using nothing
Information variable null method-competitive adaptive weighting sampling method, which is combined, filters out preferred feature electrical characteristic parameter, to preferred feature
Each electrical characteristic parameter and preferred feature electrical characteristic parameter under frequency carries out PCA analyses respectively, using optimal principal component as input
Operating limit learning machine pattern discrimination is measured, so as to establish the dynamic fermentation appropriateness detection model of black tea.
S5. the quick detection of black tea dynamic fermentation appropriateness, computer client, real-time Transmission fermentation are connected to by data wire
The electrical characteristic parameter data of sample, the fermentation appropriateness detection model write with reference to acquisition software, are analyzed in real time at software interface end
The fermentation appropriateness grade of fermentation sample, realize the qualitative quick detection of black tea dynamic fermentation appropriateness.
Detection of the specific embodiment to a kind of fermentation of black tea appropriateness based on electrical characteristic parameter of the present invention will be passed through below
Method is described in detail, and the detection method concrete operation step of fermentation of black tea appropriateness is as follows：
Step 1, the structure of electrical characteristics test system
As shown in Fig. 2 the electrical characteristics test system of the present embodiment include bridge test instrument 2, test electrode 3, sampling box 4,
Run main frame 1 and acquisition software.The test frequency scope of bridge test instrument 2 is 50Hz~200kHz, from test circuit in parallel
Under electrical parameter, bridge test instrument 2 carries out data exchange by R232 serial ports and operation main frame 1, and connected by twisted-pair shielded wire
Connect two test electrodes 3.Test electrode 3 is red copper parallel plate electrode plate, and the area of battery lead plate is 20cm2, it is grown and width
Respectively 5cm, 4cm.The side of bottom plate 5 of sampling box 4 is hinged with side plate, corresponding opposite side detachably connects with corresponding side plate
Connect, such structure makes bottom plate 5 to be rotated around side, and sample after testing can take out from sampling box bottom, be installed on sampling box 4
There is disperser, disperser includes motor 61, rotating shaft 62 and the moving tooth 63 being fastened around in rotating shaft 62, rotating shaft 62 and passed with motor 61
Dynamic connection, the rotational band turn axle 62 of motor 61 rotate.The present embodiment has customized terminal acquisition system and interface, realizes test data
Automatic record.
Step 2, the measure of sample electrical characteristic parameter
The leaf of a bud one of harvesting Fuding great Bai kinds is used for fermenting preparing black tea, and the tealeaves of harvesting is spread to water naturally in time
During point content about 60%, using 40 type kneading machines, rubbed according to sky, after gently pressure, middle pressure and each over-over mode of weight are kneaded,
Temperature is 27 DEG C, ferment 6h in growth cabinet of the fermentation humidity more than 90%.
It will be naturally scattered after disperser that Tea Samples pass through rotation, add in sampling box 4, flood test to Tea Samples
Electrode 3, determine 0.05~200kHz frequency ranges in 0.05kHz, 0.06kHz, 0.08kHz, 0.1kHz, 0.2kHz,
0.3kHz、0.4kHz、0.5kHz、0.6kHz、0.8kHz、1kHz、2kHz、3kHz、4kHz、6kHz、8kHz、10kHz、50kHz、
100kHz, 150kHz and 200kHz27 Frequency points, voltage are 1 volt of sine wave, and measure Tea Samples ferment 0~6h times
In the range of sample electrical characteristic parameter, including parallel equivalent electric capacity Cp, complex impedance Z, resistance R, reactance X, fissipation factor D and phase
Loss angle θ.It is placed in sampling box 4 and determines once every 0.5h samplings, each sample is repeated 6 times during measure, then by each parameter
Measured value of the average of value as each electrical parameter, the result of measure are as shown in Figure 3.
Wherein, Fig. 3-A are influence curve figure of the frequency to complex impedance, and Fig. 3-B are influence curve figure of the frequency to resistance, figure
3-C is influence curve figure of the frequency to impedance angle, and Fig. 3-D are influence curve figure of the frequency to reactance, and Fig. 3-E are frequencies to electric capacity
Influence curve figure, Fig. 3-F are the influence curve figures of the frequency versus path loss factor, and Fig. 3-G are shadow of the test frequency to resistance-reactance
Ring curve map.
Shown by Fig. 3 data results, detection frequency has obvious influence to fermentation of black tea sample electrical parameter, ferments
Time has considerable influence to the low frequency characteristic of electrical parameter.Electrical impedance, reactance, impedance angle and fissipation factor are gradual in fermentation
Increase, electric capacity are gradually reduced with the extension of fermentation period, illustrate to hinder the material of charge transfer more and more in fermentation process.
Step 3, fermentation quality detection and grade demarcation
While electrical characteristic parameter measure is carried out every 0.5h during fermentation of black tea, 100g is taken with fermentation time node
Fermented sample in liquid nitrogen, be repeated 3 times, after freeze-dried processing, be placed in -20 DEG C of refrigerator storage to detecting.According to
National Standard Method《Measure-high performance liquid chromatography of theaflavin in GB/T 30483-2013 tealeaves》Dark brown cellulose content is determined, i.e. tea is yellow
Element, thearubigin, the content of theabrownin and catechin.Take in addition the fermented sample of each same fermentation time node in 120 DEG C of primary drying for baking,
After processing is dried in 90 DEG C of final firings, according to GB/T 23776-2009《Tealeaves organoleptic evaluation method》And password evaluates form, by
Expert carries out sensory evaluation to fermented tea sample, evaluates the organoleptic quality of the produced black tea under each fermentation time node, and count
Total score is evaluated in calculation.
Expert sensory evaluation is the overall merit to tealeaves color and shape, and different fermentations journey can be embodied from flavour
Intrinsic stimuli of the sample biochemical composition to sense organ is spent, and can is levyd from the external performance upper table such as sample soup look, profile, tea residue in ferment
Externality of the biochemical products to quality.Calculated based on expert sensory's scoring comment and evaluate total score progress ranking, evaluated total
Sample point more than 85 is assessed as proper fermentation, and the sample before proper fermentation sample time node is slight fermentation, afterwards
Sample is excessive fermentation.The result of ranking is as shown in table 1.
The sensory evaluation of the fermentation of black tea process sample of table 1
Step 4, data screening processing and model are established
All data processings are completed under Matlab 2016a and Microsoft Windows7 platforms.Due to each test
The dimension of each electrical parameter, magnitude, susceptibility are different under frequency, therefore need first to make each electrical parameter data standardization pretreatment, go
Except the influence of dimension.
1. a pair electrical parameter initial data is standardized pretreatment
Multiplicative scatter correction MSC, deviation standardization Min-Max, Smooth, S/G2 is respectively adopted in the present embodimentstWith
Zscore methods pre-process to electrical characteristic parameter initial data, are then based on processing data and establish PLS models, each model
Can be as shown in table 2：
Performance impact of the different preprocess method of table 2 to PLS sense organ models
As a result show, Zscore is optimal electrical characteristic parameter preprocess method, and it carries Rp values from the 0.172 of initial data
Up to 0.842, and apparently higher than other preprocess methods；RMSEP is 2.194 and Bias is 0.232, also significantly lower than other pre-
Processing method.In addition, the RMSECV and RMSEP of Zscore models difference are 1.157, it is minimum in all preprocess methods,
Model is set to possess preferable Generalization Capability after showing Zscore standardizations.Therefore, in practical operation, it is possible to use only
Zscore methods are standardized pretreatment to electrical characteristic parameter initial data.
2. the preferred feature frequency that the suitable fermentation quality of screening differentiates
After being standardized pretreatment to electrical characteristic parameter initial data, using the feature based on coefficient correlation and principal component
Frequency synthesis index analysis method filters out the preferred feature frequency that suitable fermentation quality differentiates, specially establish electrical parameter with it is each
The correlation matrix of fermentation quality index, principal component analysis is carried out to each group matrix data, as a result as shown in table 3.
Electrical parameter and the principal component analysis of index of quality coefficient correlation under the different frequency of table 3
Using the characteristic value corresponding to each principal component as flexible strategy, with preceding 3 principal component scores (score-1, score-2,
Score-3 correlation synthesis Principal component, referred to as correlation index CI can be obtained by) being multiplied, and its calculation formula is：CI=(LATENT
(1,:)*score(:,1)+LATENT(2,:)*score(:,2)+LATENT(3,:)*score(:,3))/(LATENT(1,:)+
LATENT(2,:)+LATENT(3,:)), score-principal component scores matrix in formula, LATENT-eigenvectors matrix.Thus
The CI of each frequency calculated is shown in Table 4.Each index of quality and the synthesis correlation of 6 electrical parameters, each quality can be obtained by table 4
The CI values of index, the changing rule totally dropped afterwards in first increasing with frequency increase CI.
CI absolute value of the TFs contents in 0.2kHz is 0.474, is ranked the first；CI absolute value of the TRs activity in 0.1kHz be
0.429, rank the first；CI absolute values of the TBs in 0.5kHz is 0.432, is ranked the first；CI of the sensory evaluation scores in 0.2kHz is absolute
It is worth for 0.274, ranks the first；Although each index of quality is corresponding with the best Frequency point of comprehensive correlation, its optimal frequency is simultaneously
Do not unify.Therefore the array formed to the CI values of each index of quality in table 4 at different frequencies carries out secondary PCA analyses, and
Calculate it and integrate index of correlation CCI, the CCI of each frequency and overall ranking are as shown in most right two row of table 4.
Electrical parameter and index of quality synthesis index of correlation CCI under 4 different test frequencies of table
As can be seen from Table 4, CCI values are more than 0 Frequency point totally 7, wherein, 0.2kHz CCI values are up to 0.274,
Therefore the characteristic frequency that 0.2kHz can test as electrical characteristic parameter in fermentation of black tea.
3. filter out preferred feature electrical characteristic parameter
It is excellent using variable of the Monte Carlo without information variable null method MCUVE and competitive adaptive weighting sampling method CARS
Choosing method carries out preferably, rejecting useless characteristic variable to 162 (27 × 6) electrical characteristic parameters under 27 frequencies.MCUVE side
Method is by calculating the index of stability RI values of each variable come the importance of each variable in evaluation model, according to RI values size to becoming
Amount is ranked up, and establishes new variables collection, and then 1 variable that progressively adds up establishes PLS models respectively, with forecast set most
Evaluation index of the small RMSEP values as variable encumbrance, MCUVE feature electrical parameter selection results are as shown in table 5：
Table 5MCUVE feature electrical parameter selection results
Above-mentioned as shown by data, frequencies of the RI more than 4 of sensory evaluation scores selection are concentrated mainly on 0.06~0.4kHz of low-frequency range,
The high variable of stability is the fissipation factor D and reactance X of electrical parameter, and wherein the RI values of D variables are up under 0.4kHz frequencies
4.53.Frequencies of the RI more than 2 of TFs selections is concentrated mainly on middle 0.1~5kHz of low-frequency range, and the high variable of stability is electrical parameter
Fissipation factor D and reactance X, wherein under 1kHz frequencies the RI values of X variables be up to 3.17.The index of stability RI of TRs selections is big
Frequency in 2.5 is concentrated mainly on middle 0.1~5kHz of low-frequency range, and the high variable of stability is the fissipation factor D and electricity of electrical parameter
The RI values of X variables are up to 3.47 under anti-X, wherein 0.6kHz frequencies.The index of stability RI of TBs selections is more than 3.7 frequency
0.05~0.1kHz of low-frequency range is concentrated mainly on, the high variable of stability is the fissipation factor D and reactance X of electrical parameter, wherein
The RI values of X variables are up to 4.51 under 0.1kHz frequencies.Comprehensive analysis, stable test frequency is low frequency with each index of quality
Section, maximally related feature electrical parameter is D and X.
The electrical parameter selected using CARS to MCUVE is further optimized, and sets CARS algorithm continuous sampling numbers as 50
With 10 times of validation-cross, choose the variable subset corresponding to RMSECV minimum values and be characterized variable, obtain synteny minimum
Validity feature variable.As shown in figure 4, by taking the MCUVE-CARS feature electrical parameter screening processes of sensory evaluation scores forecast model as an example,
When sampling number is 8 times, its RMSECV reaches minimum value 1.235, and corresponding feature electrical parameter is 10.Similarly obtain
TFs, TRs, TBs feature electrical parameter are respectively 8,5,6, as a result as shown in table 6：
Table 6MCUVE-CARS feature electrical parameter selection results
4. principal component analysis
The preferred feature electricity ginseng that 6 electrical parameters under preferable 0.2kHz and MCUVE-CARS methods are screened from full rate
Number, carries out PCA analyses respectively, and the three-dimensional load diagram of preceding 3 principal components is as shown in Figure 5.Fig. 5-A are 6 under 0.2kHz frequencies
Electrical characteristic parameter principal component load diagram, Fig. 5-B are characterized electrical characteristic parameter principal component load diagram.As shown in Figure 5, each sample is pressed
Fermentation time difference is gathered in different area of space, certain intersection also be present between each area of space, it is still necessary to reference to mould
Formula method of discrimination makees further differentiation.
5. extreme learning machine pattern discrimination
Sent out optimal principal component as input quantity operating limit learning machine ELM pattern discriminations, the dynamic of fast prediction black tea
Ferment degree.Program in Matlab softwares, using ELM sorting algorithms, from the activation primitive that Sigmoid functions are ELM, utilize
The number of principal components evidence of the electrical characteristic parameter extracted, the black tea sample for appropriateness of being fermented to three classes carry out discriminant classification.Because of hidden layer
Number N and principal component factor number PCs are had a great influence to the estimated performance of discrimination model, therefore N and PCs is made in the range of selection
Further parameter collaboration optimizing processing.Choose respectively 10 Φ values (5~50, step-length 5) and 10 PCs numbers (1~10, walk
It is a length of 1), preferred parameter is determined with the differentiation accuracy of training pattern, Fig. 6 shows different principal component factor numbers to ELM moulds
The differentiation result of type calibration set.When it is 20 that PCs, which is 3, N, the differentiation rate of model calibration set and forecast set reaches 100%, mould
Type better performances.Table 7 shows differentiation result of the ELM models to different fermentations quality appropriateness grade black tea, it is known that electrical characteristics are joined
Number, which combines ELM discrimination models, can realize that the prediction to fermentation quality appropriateness grade differentiates.
Differentiation result of the table 7ELM models to different fermentations quality appropriateness grade black tea
Step 5, the quick detection of black tea dynamic fermentation appropriateness
Computer client, the electrical characteristic parameter data of real-time Transmission fermented sample, with reference to collection are connected to by data wire
The fermentation appropriateness detection model of software write-in, the fermentation appropriateness grade of fermentation sample is analyzed in real time at software interface end, is realized red
The qualitative quick detection of tea dynamic fermentation appropriateness.
As can be seen from the above embodiments, the present invention has using the method for electrical characteristics technology for detection fermentation of black tea quality appropriateness
There have the advantages that to be rapid, efficient, convenient, and probe is inserted into fermentation of black tea sample, so as to realize the reality of fermentation of black tea appropriateness
When online and Site Detection.
The above embodiments are merely illustrative of the technical solutions of the present invention and it is unrestricted, although with reference to preferred embodiment to this hair
It is bright to be described in detail, it will be understood by those within the art that, technical scheme can be modified
Or equivalent substitution, without departing from the objective and scope of technical solution of the present invention, it all should cover the claim in the present invention
Among scope.The present invention be not described in detail technology, shape, construction part be known technology.
Claims (9)
1. a kind of detection method of the fermentation of black tea appropriateness based on electrical characteristic parameter, it is characterised in that by Various Fermenting Degree
Tealeaves is respectively placed in the measure that electrical characteristic parameter is carried out in electrical characteristics test system, while detects the index of quality, carries out grade mark
It is fixed, Screening Treatment is carried out to electrical characteristic parameter, it is suitable to fermenting for realizing so as to establish prediction black tea dynamic fermentation appropriateness model
The prediction for spending grade differentiates, comprises the following steps：
S1. the structure of electrical characteristics test system, the electrical characteristics test system include bridge test instrument, test electrode, sampling box,
Main frame and acquisition software are run, disperser is installed on the sampling box, the bridge test instrument passes through R232 serial ports and operation
Main frame carries out data exchange, and connects two test electrodes by twisted-pair shielded wire；
S2. the measure of sample electrical characteristic parameter, the Tea Samples of different fermentations time are spilled into naturally by disperser respectively
Enter in sampling box, and Tea Samples is flooded test electrode, respectively at measuring electrical characteristic parameter under different frequency；
S3. fermentation quality detection and grade demarcation, the index of quality of different fermentations time Tea Samples is detected, and to fermentation quality
Carry out grade demarcation；
S4. data screening processing and Model checking, pretreatment is standardized to electrical characteristic parameter, using based on coefficient correlation with
The characteristic frequency composite index analysis method of principal component selects the preferred feature frequency that suitable fermentation quality differentiates, using without information
Variable elimination method-competitive adaptive weighting sampling method, which is combined, filters out preferred feature electrical characteristic parameter, to preferred feature frequency
Under each electrical characteristic parameter and preferred feature electrical characteristic parameter carry out PCA analyses respectively, optimal principal component is made as input quantity
With extreme learning machine pattern discrimination, so as to establish the dynamic fermentation appropriateness detection model of black tea.
S5. the quick detection of black tea dynamic fermentation appropriateness, computer client, real-time Transmission fermented sample are connected to by data wire
Electrical characteristic parameter data, with reference to acquisition software write fermentation appropriateness detection model, analyze fermentation in real time at software interface end
The fermentation appropriateness grade of sample, realize the qualitative quick detection of black tea dynamic fermentation appropriateness.
2. a kind of detection method of fermentation of black tea appropriateness based on electrical characteristic parameter according to claim 1, its feature exist
In the test frequency scope of the bridge test instrument is 50Hz~200kHz, and the test electrode is copper parallel plate electrode
Plate, the area of battery lead plate is 20cm2, the disperser includes motor, rotating shaft and the moving tooth being fastened around in rotating shaft, rotating shaft with
Motor is connected, and the rotating speed of the rotating shaft is more than 1000r/min.
3. a kind of detection method of fermentation of black tea appropriateness based on electrical characteristic parameter according to claim 1, its feature exist
In the electrical characteristic parameter in the S2 includes parallel equivalent electric capacity, complex impedance, resistance, reactance, fissipation factor and phase loss
Angle.
4. a kind of detection method of fermentation of black tea appropriateness based on electrical characteristic parameter according to claim 3, its feature exist
In the test condition of electrical characteristic parameter is test voltage 1V, 0.05~200kHz of test frequency in the S2.
5. a kind of detection method of fermentation of black tea appropriateness based on electrical characteristic parameter according to claim 1, its feature exist
In the index of quality in the S3 includes sensory evaluation scores, theaflavin, thearubigin and theabrownin, the theaflavin, thearubigin and tea
Brown element is measured using high performance liquid chromatography.
6. a kind of detection method of fermentation of black tea appropriateness based on electrical characteristic parameter according to claim 5, its feature exist
In the demarcation of fermentation quality grade is to calculate to evaluate total score based on expert sensory's scoring comment in the S3, evaluates total score more than 85
Sample be assessed as proper fermentation, the sample before proper fermentation sample time node is slight fermentation, and sample afterwards was
Degree fermentation.
7. a kind of detection method of fermentation of black tea appropriateness based on electrical characteristic parameter according to claim 1, its feature exist
In the characteristic frequency composite index analysis method based on coefficient correlation and principal component in the S4 selects preferred feature frequency, tool
Body is the correlation matrix for establishing electrical parameter and each fermentation quality index, carries out principal component analysis to each group matrix data, obtains
To correlation index, the array formed to the correlation index of each index of quality at different frequencies carries out PCA analyses, is calculated
The comprehensive index of correlation, frequency corresponding to the comprehensive index of correlation of maximum is preferred feature frequency.
8. a kind of detection method of fermentation of black tea appropriateness based on electrical characteristic parameter according to claim 7, its feature exist
In it is 0.2kHz that the preferred feature frequency moderately differentiated of suitably fermenting is filtered out in the S4.
9. a kind of detection method of fermentation of black tea appropriateness based on electrical characteristic parameter according to claim 8, its feature exist
In the extreme learning machine pattern discrimination is programmed in Matlab softwares, using ELM sorting algorithms, from Sigmoid functions
For ELM activation primitive, using the number of principal components evidence of the electrical characteristic parameter extracted, the black tea sample for appropriateness of being fermented to three classes enters
Row discriminant classification.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710828757.3A CN107462607B (en) | 2017-09-14 | 2017-09-14 | A kind of detection method of the fermentation of black tea appropriateness based on electrical characteristic parameter |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710828757.3A CN107462607B (en) | 2017-09-14 | 2017-09-14 | A kind of detection method of the fermentation of black tea appropriateness based on electrical characteristic parameter |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107462607A true CN107462607A (en) | 2017-12-12 |
| CN107462607B CN107462607B (en) | 2019-09-06 |
Family
ID=60552351
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710828757.3A Active CN107462607B (en) | 2017-09-14 | 2017-09-14 | A kind of detection method of the fermentation of black tea appropriateness based on electrical characteristic parameter |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107462607B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109329501A (en) * | 2018-10-29 | 2019-02-15 | 江苏大学 | A kind of intelligent preparation method of instant fermented tea powder |
| CN109497194A (en) * | 2018-10-29 | 2019-03-22 | 江苏大学 | A kind of fermented tea intelligence process for solid state fermentation |
| CN116777309A (en) * | 2023-08-23 | 2023-09-19 | 四川省眉山益稷农业科技有限公司 | Organic fertilizer fermentation quality evaluation method |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1763543A (en) * | 2005-09-27 | 2006-04-26 | 浙江大学 | Black tea quality identification method |
| CN103412006A (en) * | 2013-06-19 | 2013-11-27 | 中国农业科学院茶叶研究所 | Judgment method for appropriate rolling degree of red tea |
| CN103776717A (en) * | 2014-01-17 | 2014-05-07 | 中安高科检测科技（北京）有限公司 | Preparation method of gas sensor array for detecting Yunnan black tea |
| CN106538729A (en) * | 2016-10-24 | 2017-03-29 | 江苏丘陵地区镇江农业科学研究所 | A kind of black tea fresh leaf is processed and fermentation technology |
| CN107505358A (en) * | 2017-09-14 | 2017-12-22 | 中国农业科学院茶叶研究所 | A kind of detection method of the fermentation of black tea index of quality based on electrical characteristic parameter |
-
2017
- 2017-09-14 CN CN201710828757.3A patent/CN107462607B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1763543A (en) * | 2005-09-27 | 2006-04-26 | 浙江大学 | Black tea quality identification method |
| CN103412006A (en) * | 2013-06-19 | 2013-11-27 | 中国农业科学院茶叶研究所 | Judgment method for appropriate rolling degree of red tea |
| CN103776717A (en) * | 2014-01-17 | 2014-05-07 | 中安高科检测科技（北京）有限公司 | Preparation method of gas sensor array for detecting Yunnan black tea |
| CN106538729A (en) * | 2016-10-24 | 2017-03-29 | 江苏丘陵地区镇江农业科学研究所 | A kind of black tea fresh leaf is processed and fermentation technology |
| CN107505358A (en) * | 2017-09-14 | 2017-12-22 | 中国农业科学院茶叶研究所 | A kind of detection method of the fermentation of black tea index of quality based on electrical characteristic parameter |
Non-Patent Citations (1)
| Title |
|---|
| 龚琦: "LRC数字电桥检测茶叶品质的研究", 《TRANSACTIONS OF THE CSAE》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109329501A (en) * | 2018-10-29 | 2019-02-15 | 江苏大学 | A kind of intelligent preparation method of instant fermented tea powder |
| CN109497194A (en) * | 2018-10-29 | 2019-03-22 | 江苏大学 | A kind of fermented tea intelligence process for solid state fermentation |
| CN116777309A (en) * | 2023-08-23 | 2023-09-19 | 四川省眉山益稷农业科技有限公司 | Organic fertilizer fermentation quality evaluation method |
| CN116777309B (en) * | 2023-08-23 | 2023-10-27 | 四川省眉山益稷农业科技有限公司 | Organic fertilizer fermentation quality evaluation method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107462607B (en) | 2019-09-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107505358B (en) | A kind of detection method of the fermentation of black tea index of quality based on electrical characteristic parameter | |
| CN102818777B (en) | Fruit maturity degree evaluation method based on spectrum and color measurement | |
| CN107860740A (en) | A kind of evaluation method of the fermentation of black tea quality based on near-infrared spectrum technique | |
| CN107462607B (en) | A kind of detection method of the fermentation of black tea appropriateness based on electrical characteristic parameter | |
| CN108663339A (en) | Corn online test method of going mouldy based on spectrum and image information fusion | |
| CN108960315B (en) | Intelligent evaluation system and method for quality of cooked meat product | |
| Zhu et al. | Evaluation of green tea sensory quality via process characteristics and image information | |
| CN106841083A (en) | Sesame oil quality detecting method based on near-infrared spectrum technique | |
| CN104977258B (en) | The quality detecting methods such as tealeaves/cosmetics based on two-dimensional correlation spectra | |
| CN102735642B (en) | Method for quickly and losslessly identifying virgin olive oil and olive-residue oil | |
| CN106568907B (en) | A kind of steamed crab freshness lossless detection method based on semi-supervised identification projection | |
| CN106769970A (en) | A kind of method of the kind, pol and acidity of Fast nondestructive evaluation apple | |
| CN106483166A (en) | A kind of method based on dielectric spectra technology quick detection cow's milk fat content | |
| CN105973840B (en) | A kind of method of quick detection Maotai-flavor wine brewing fermented grain physical and chemical index | |
| CN108037256A (en) | The rapid assay methods of rice eating-quality | |
| CN107045013A (en) | Beef taste quality automatic classification detector and its detection method | |
| CN110487746A (en) | A method of baby cabbage quality is detected based near infrared spectrum | |
| CN110263969B (en) | Dynamic prediction system and prediction method for quality of apples with shelf life | |
| CN108613943A (en) | A kind of near-infrared single grain crop component detection method based on the transfer of spectrum form | |
| CN106770014A (en) | A kind of near infrared spectrum detection method of rice freshness | |
| CN109961179A (en) | A kind of aquatic products quality detecting method and portable Raman device | |
| Zhang et al. | Non-destructive detection of the fruit firmness of Korla fragrant pear based on electrical properties | |
| CN110231302A (en) | A kind of method of the odd sub- seed crude fat content of quick measurement | |
| Ren et al. | Description of tea quality using deep learning and multi-sensor feature fusion | |
| CN108444943A (en) | A kind of adulterated method for quick identification of coffee based on Fourier Transform Near Infrared |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |