CN104849323B - A kind of method of fining agent in quick detection fruit juice based on Electronic Nose - Google Patents
A kind of method of fining agent in quick detection fruit juice based on Electronic Nose Download PDFInfo
- Publication number
- CN104849323B CN104849323B CN201510226487.XA CN201510226487A CN104849323B CN 104849323 B CN104849323 B CN 104849323B CN 201510226487 A CN201510226487 A CN 201510226487A CN 104849323 B CN104849323 B CN 104849323B
- Authority
- CN
- China
- Prior art keywords
- electronic nose
- sample
- sensor
- value
- fruit juice
- 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.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims abstract description 46
- 235000015203 fruit juice Nutrition 0.000 title claims abstract description 36
- 238000001514 detection method Methods 0.000 title claims abstract description 30
- 239000006025 fining agent Substances 0.000 title claims abstract description 8
- 238000007637 random forest analysis Methods 0.000 claims abstract description 40
- 241000675108 Citrus tangerina Species 0.000 claims abstract description 19
- 238000004458 analytical method Methods 0.000 claims abstract description 18
- 238000007789 sealing Methods 0.000 claims abstract description 16
- 235000011389 fruit/vegetable juice Nutrition 0.000 claims abstract description 14
- 238000000513 principal component analysis Methods 0.000 claims abstract description 11
- 238000003066 decision tree Methods 0.000 claims description 31
- 235000015205 orange juice Nutrition 0.000 claims description 10
- 238000012952 Resampling Methods 0.000 claims description 9
- 239000003205 fragrance Substances 0.000 claims description 9
- 239000003795 chemical substances by application Substances 0.000 claims description 8
- 238000005352 clarification Methods 0.000 claims description 5
- 239000000706 filtrate Substances 0.000 claims description 5
- 238000001914 filtration Methods 0.000 claims description 5
- 239000011159 matrix material Substances 0.000 claims description 5
- 238000005457 optimization Methods 0.000 claims description 5
- 239000006228 supernatant Substances 0.000 claims description 5
- 238000004381 surface treatment Methods 0.000 claims description 5
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 5
- 108010059820 Polygalacturonase Proteins 0.000 claims description 4
- 108010093305 exopolygalacturonase Proteins 0.000 claims description 4
- 239000007788 liquid Substances 0.000 claims description 4
- 230000002000 scavenging effect Effects 0.000 claims description 4
- 239000013049 sediment Substances 0.000 claims description 4
- 230000006641 stabilisation Effects 0.000 claims description 4
- 238000011105 stabilization Methods 0.000 claims description 4
- 230000003068 static effect Effects 0.000 claims description 4
- 239000003381 stabilizer Substances 0.000 abstract description 7
- 238000004140 cleaning Methods 0.000 abstract description 2
- 238000004445 quantitative analysis Methods 0.000 abstract description 2
- 238000005070 sampling Methods 0.000 abstract 1
- 210000001331 nose Anatomy 0.000 description 41
- 239000007789 gas Substances 0.000 description 20
- 229920001661 Chitosan Polymers 0.000 description 8
- 239000000203 mixture Substances 0.000 description 5
- 229920002101 Chitin Polymers 0.000 description 4
- VYPSYNLAJGMNEJ-UHFFFAOYSA-N Silicium dioxide Chemical compound O=[Si]=O VYPSYNLAJGMNEJ-UHFFFAOYSA-N 0.000 description 4
- 239000004615 ingredient Substances 0.000 description 4
- 235000013399 edible fruits Nutrition 0.000 description 3
- MWUXSHHQAYIFBG-UHFFFAOYSA-N nitrogen oxide Inorganic materials O=[N] MWUXSHHQAYIFBG-UHFFFAOYSA-N 0.000 description 3
- 238000003756 stirring Methods 0.000 description 3
- 102100028260 Gamma-secretase subunit PEN-2 Human genes 0.000 description 2
- 108010010803 Gelatin Proteins 0.000 description 2
- 101000579663 Homo sapiens Gamma-secretase subunit PEN-2 Proteins 0.000 description 2
- 150000001335 aliphatic alkanes Chemical class 0.000 description 2
- 229910000278 bentonite Inorganic materials 0.000 description 2
- 239000000440 bentonite Substances 0.000 description 2
- SVPXDRXYRYOSEX-UHFFFAOYSA-N bentoquatam Chemical compound O.O=[Si]=O.O=[Al]O[Al]=O SVPXDRXYRYOSEX-UHFFFAOYSA-N 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 235000013305 food Nutrition 0.000 description 2
- 229920000159 gelatin Polymers 0.000 description 2
- 239000008273 gelatin Substances 0.000 description 2
- 235000019322 gelatine Nutrition 0.000 description 2
- 235000011852 gelatine desserts Nutrition 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- VNWKTOKETHGBQD-UHFFFAOYSA-N methane Chemical compound C VNWKTOKETHGBQD-UHFFFAOYSA-N 0.000 description 2
- 238000001556 precipitation Methods 0.000 description 2
- 239000003755 preservative agent Substances 0.000 description 2
- 230000002335 preservative effect Effects 0.000 description 2
- 230000001373 regressive effect Effects 0.000 description 2
- 230000035945 sensitivity Effects 0.000 description 2
- 239000000126 substance Substances 0.000 description 2
- 238000005303 weighing Methods 0.000 description 2
- 241000196324 Embryophyta Species 0.000 description 1
- 108090000790 Enzymes Proteins 0.000 description 1
- 102000004190 Enzymes Human genes 0.000 description 1
- LFQSCWFLJHTTHZ-UHFFFAOYSA-N Ethanol Chemical compound CCO LFQSCWFLJHTTHZ-UHFFFAOYSA-N 0.000 description 1
- UCKMPCXJQFINFW-UHFFFAOYSA-N Sulphide Chemical compound [S-2] UCKMPCXJQFINFW-UHFFFAOYSA-N 0.000 description 1
- 239000000654 additive Substances 0.000 description 1
- 230000000996 additive effect Effects 0.000 description 1
- 238000013019 agitation Methods 0.000 description 1
- 235000011114 ammonium hydroxide Nutrition 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 235000013361 beverage Nutrition 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 230000006866 deterioration Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011049 filling Methods 0.000 description 1
- 239000003292 glue Substances 0.000 description 1
- 210000003128 head Anatomy 0.000 description 1
- 239000001257 hydrogen Substances 0.000 description 1
- 229910052739 hydrogen Inorganic materials 0.000 description 1
- 125000004435 hydrogen atom Chemical class [H]* 0.000 description 1
- 230000003902 lesion Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 229910044991 metal oxide Inorganic materials 0.000 description 1
- 150000004706 metal oxides Chemical class 0.000 description 1
- 229930003811 natural phenol Natural products 0.000 description 1
- 150000002898 organic sulfur compounds Chemical class 0.000 description 1
- 230000003647 oxidation Effects 0.000 description 1
- 238000007254 oxidation reaction Methods 0.000 description 1
- 229920001277 pectin Polymers 0.000 description 1
- 235000010987 pectin Nutrition 0.000 description 1
- 239000001814 pectin Substances 0.000 description 1
- 238000005375 photometry Methods 0.000 description 1
- 150000008442 polyphenolic compounds Chemical class 0.000 description 1
- 235000013824 polyphenols Nutrition 0.000 description 1
- 239000000047 product Substances 0.000 description 1
- 108090000623 proteins and genes Proteins 0.000 description 1
- 102000004169 proteins and genes Human genes 0.000 description 1
- 238000004451 qualitative analysis Methods 0.000 description 1
- 230000001953 sensory effect Effects 0.000 description 1
- 238000002798 spectrophotometry method Methods 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
Landscapes
- Investigating Or Analyzing Materials By The Use Of Electric Means (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention discloses a kind of method of fining agent in quick detection fruit juice based on Electronic Nose.Oranges and tangerines surface is cleaned, is dried, pick out the basically identical oranges and tangerines of size color as detection object.Oranges and tangerines are removed the peel, squeeze the juice, filtered, it is stand-by.Pectase is made into the solution of various concentrations, a certain proportion of mixing is carried out with orange blossom.Take mixed 10 ~ 30ml of solution to be put in the sealing container of 500ml, the solution of each concentration scale respectively takes 24 repeat samples, to the head space of sample, other carry out detection by electronic nose.Electronic Nose sensor settings condition is：Under certain flow rate, the sampling time is 80s, and the air cleaning time is 60s；Using steady-state value as Electronic Nose characteristic value.Using principal component analysis, discriminant analysis, and random forest carry out quantitative analysis to stabilizer in fruit juice.The present invention substantially improves the quick detection of fruit juice internal stabilizer, convenient, objective, and with good discrimination, with popularization higher and value.
Description
Technical field
The invention belongs to food fruit juice clarifier detection technique field, it is related to a kind of based in Electronic Nose quick detection fruit juice
The method of fining agent.
Background technology
Fruit juice is susceptible to turbidity and precipitation in long-term storage process, and oxidation deterioration may occur.Muddy the reason for
A lot, it is mainly relevant with the material such as naturally occurring phenols.When protein and pectin substance and polyphenols in fruit juice
When coexisting for a long time, the phaneroplasm of muddiness will be produced, or even precipitated.So can usually add in the food industry various clear
Clear agent, the material of precipitation is easily caused to remove a part or major part, fruit juice is formed more stable solution.
Beverage industry is often added, such as bentonite, diatomite, gelatin, Ludox, pectase clarification aid or they
Be used in combination to reach the purpose of clarification.The conventional determining method of stabiliser content has chemical detection, light splitting in current fruit juice
Photometry, fluorescence spectrophotometry etc., but the shortcomings of these methods have complex operation, analysis time is long.Therefore, one is explored
It is necessary to plant fast and convenient stabilizer detection method.
Electronic Nose quickly provides the whole of sample also known as smell scanner with specific sensor and PRS
Body information, indicate sample hidden feature, with high sensitivity, reliability, repeatability, it sample can be quantified, together
When some component contents can be carried out fast qualitative quantify.At present, using fruit in olfactory sensor quantitative determination fruit juice
The correlative study of glue enzyme content is not yet reported.Present invention aim at stabilizer in Quantitative detection fruit juice, while filling up state
The inside and outside blank on the quick detection of stabiliser content in fruit drink.
The content of the invention
Regarding to the issue above, it is an object of the invention to provide a kind of side of fining agent in quick detection fruit juice based on Electronic Nose
Method.
The purpose of the present invention is achieved through the following technical solutions, and one kind is based on clear in Electronic Nose quick detection fruit juice
The method of clear agent, comprises the following steps：
(1) by oranges and tangerines surface treatment it is clean, remove the peel, squeeze the juice, filter, abandon filter residue, the oranges and tangerines filtrate of pure concentration is standby；Respectively
The pectase of 3g, 6g, 9g, 12g is weighed, under 40 DEG C of water bath condition, pectase is repeated to dissolve, filter residue is abandoned in filtering；In room
Under temperature, during pectinase solution constant volume put into 100ml volumetric flasks, with pure oranges and tangerines solution in mass ratio 1:11 ratio is mixed,
Stirring；After static 1 hour of blended fruit juice, it is centrifuged under the centrifugal speed not less than 4000r/min, is abandoned sediment, is taken
Clear liquid is stand-by；
(2) take the 10~30ml of fruit juice supernatant after treatment to be put into sealing container of the volume not less than 250ml, sealing,
30min is placed at room temperature, the fragrance level of orange juice is reached saturation in the beaker headspace of sealing；Each concentration
The solution of grade is no less than 3 repeat samples respectively, is the regression modeling collection Electronic Nose data in later stage；It is built-in by Electronic Nose
Pump by the sensor array passage of the headspace gas suction Electronic Nose in sealing container, with sample gas send out by Electronic Nose sensor
Raw reaction produces sensor signal；Electrical conductivity G when the sensor signal is sensor contacts sample gas exists with sensor
Electrical conductivity G during by calibration gas0Ratio, i.e. G/G0；The detection time of the Electronic Nose sensor is 80s, scavenging period
It is 60s, internal pump gas flow rate is 200ml/min；
(3) the sensor response of detection by electronic nose orange juice is a data matrix, by many responses of sensor
Value composition, from each sensor stabilization value as characteristic value, in SPSS softwares, is entered using principal component analysis, discriminant analysis
Row analysis；In Matlab softwares, according to characteristic value by bootstrap (boot-strap) resampling technique, training is continuously generated
Sample and test sample, some decision trees are generated by training sample, so as to set up Random Forest model；In Random Forest model
In, the number to decision tree in the variable number and random forest of the tree node of decision tree is optimized, and test result is by certainly
Depending on the average value of plan tree voting results；
(4) the unknown sample to be tested of agent content will be clarified its Electronic Nose sensor response is obtained by step 2, will sensed
The Random Forest model that device response steps for importing 3 is obtained, the clarification agent content in final prediction sample to be tested.
Further, in the step 3, by correlation coefficient r, root-mean-square error RMSE value to the random forest after optimization
Model is evaluated, specially：
Wherein, N represents the number of samples in modeling process；
XiI-th test value in for modeling process；
It is the average value of the sample responses value in modeling process；
YiIt is i-th predicted value of sample in modeling process；
It is the average value of the sample predictions value in modeling process.
The beneficial effects of the invention are as follows by principal component analysis, discriminant analysis, random forest scheduling algorithm is set up and good determined
Property Quantitative Prediction Model, evaluation and pre- is carried out to the pectase of different content in juice solution using intelligent smell sensory system
Survey, it is simple to operate, quick, quick indirect evaluation truly has been tested, it is that fruit drink industry additive is determined, carry
A kind of new method is supplied.
Brief description of the drawings
The Electronic Nose of Fig. 1 present examples 1 sensor response signal when detection is containing pectase fruit juice；
The principal component analysis result of the pectase samples of juice of the various concentrations of Fig. 2 present examples 1；
The discriminant analysis result of the pectase samples of juice of the various concentrations of Fig. 3 present examples 1；
The random forest regressive prediction model of the pectase samples of juice of the various concentrations of Fig. 4 present examples 1；
The Electronic Nose of Fig. 5 present examples 2 sensor response signal when detection is containing chitin fruit juice
The principal component analysis result of the chitin samples of juice of the various concentrations of Fig. 6 present examples 2；
The discriminant analysis result of the chitin samples of juice of the various concentrations of Fig. 7 present examples 2；
The random forest regressive prediction model of the chitin samples of juice of the various concentrations of Fig. 8 present examples 2.
Specific embodiment
The present invention contains suitable for the various fruit juice stabilizers such as pectase, shitosan, bentonite, diatomite, gelatin, Ludox
It is fixed to measure.Electronic Nose quickly provides the Global Information of sample with specific sensor and PRS, indicates sample
Hidden feature, with high sensitivity, reliability, repeatability, it sample can be quantified, while can be to some compositions
Content carries out fast qualitative and quantifies.Be embodied in case also added based on smell finger print information to fruit juice in shitosan
Qualitative and quantitative analysis, so as to further illustrate popularity of the invention.
The method of fining agent in a kind of quick detection fruit juice based on Electronic Nose of the present invention, using olfactory sensor to containing not
It is used for quickly detecting with the juice solution of concentration pectase, sets up effective Quantitative Prediction Model, is comprised the following steps that：
(1) by oranges and tangerines surface treatment it is clean, remove the peel, squeeze the juice, filter, abandon filter residue, the oranges and tangerines filtrate of pure concentration is standby；Respectively
The accurate pectase for weighing 3g, 6g, 9g, 12g, under 40 DEG C of water bath condition, pectase is repeated to dissolve, and filter residue is abandoned in filtering.
At room temperature, in pectinase solution constant volume being put into 100ml volumetric flasks.Pressed with 1 with pure oranges and tangerines solution:11 ratio is mixed,
Stirring.After static 1 hour of blended fruit juice, it is centrifuged under the centrifugal speed not less than 4000r/min, is abandoned sediment, is taken
Clear liquid is stand-by；
(2) take the 10~30ml of fruit juice supernatant after treatment to be put into sealing container of the volume not less than 250ml, sealing,
30min is placed at room temperature, the fragrance level of orange juice is reached saturation in the beaker headspace of sealing.Each concentration
The solution of grade is no less than 3 repeat samples respectively, is the regression modeling collection Electronic Nose data in later stage.It is built-in by Electronic Nose
Pump by the sensor array passage of the headspace gas suction Electronic Nose in sealing container, with sample gas send out by Electronic Nose sensor
Raw reaction produces sensor signal；Electrical conductivity G when the sensor signal is sensor contacts sample gas exists with sensor
The ratio of electrical conductivity G0 during by calibration gas, i.e. G/G0；The detection time of the Electronic Nose sensor is 80s, during cleaning
Between be 60s, internal pump gas flow rate be 200ml/min；
(3) the sensor response of detection by electronic nose orange juice is a data matrix, by many responses of sensor
Value composition, from each sensor stabilization value as characteristic value, in SPSS softwares, is entered using principal component analysis, discriminant analysis
Row analysis；In Matlab softwares, according to characteristic value by bootstrap (boot-strap) resampling technique, training is continuously generated
Sample and test sample, some decision trees are generated by training sample, so as to set up Random Forest model；In Random Forest model
In, the number to decision tree in the variable number and random forest of the tree node of decision tree is optimized, and test result is by certainly
Depending on the average value of plan tree voting results；
(4) the unknown sample to be tested of agent content will be clarified its Electronic Nose sensor response is obtained by step 2, will sensed
The Random Forest model that device response steps for importing 3 is obtained, the clarification agent content in final prediction sample to be tested.
In the step 3, the Random Forest model after optimization is carried out by correlation coefficient r, root-mean-square error RMSE value
Evaluate, specially：
Wherein, N represents the number of samples in modeling process；
XiI-th test value in for modeling process；
It is the average value of the sample responses value in modeling process；
YiIt is i-th predicted value of sample in modeling process；
It is the average value of the sample predictions value in modeling process.
Embodiment 1
The present embodiment using pectase as analysis object.The PEN2 types Electronic Nose of German AIRSENSE companies is detecting instrument
Elaborate, the Electronic Nose is made up of 10 metal oxide sensors, its model is as shown in the table with response characteristic：
| Sequence number | Title | Performance characteristics |
| 1 | S1 | It is sensitive to fragrance ingredient |
| 2 | S2 | It is very sensitive to nitrogen oxides |
| 3 | S3 | It is sensitive to ammoniacal liquor, fragrance ingredient |
| 4 | S4 | It is selective to hydrogen |
| 5 | S5 | It is sensitive to alkane, fragrance ingredient |
| 6 | S6 | It is sensitive to methane |
| 7 | S7 | It is sensitive to sulfide |
| 8 | S8 | To alcohol sensible |
| 9 | S9 | It is sensitive to fragrance ingredient, organic sulfur compound |
| 10 | S10 | It is sensitive to alkane |
Specific detecting step is as follows：
1st, by oranges and tangerines surface treatment it is clean, remove the peel, squeeze the juice, filter, abandon filter residue, the oranges and tangerines filtrate of pure concentration is standby；Respectively
The accurate pectase for weighing 3g, 6g, 9g, 12g, under 40 DEG C of water bath condition, pectase is repeated to dissolve, and filter residue is abandoned in filtering.
At room temperature, in pectinase solution constant volume being put into 100ml volumetric flasks.Pressed with 1 with pure oranges and tangerines solution:11 ratio is mixed,
Stirring.After static 1 hour of blended fruit juice, it is centrifuged under the centrifugal speed not less than 4000r/min, is abandoned sediment, is taken
Clear liquid is stand-by；
2nd, take the fruit juice supernatant 10ml after treatment to be put into the beaker of volume 500ml, sealed with double-deck preservative film, in room
Temperature is lower to place 30min, the fragrance level of orange juice is reached saturation in the beaker headspace of sealing；Each concentration scale
Fruit juice distinguish 24 repeat samples, be the later stage regression modeling collection Electronic Nose data；To be sealed by Electronic Nose internal pump
In the sensor array passage of the headspace gas suction Electronic Nose in container, Electronic Nose sensor and sample gas react product
Raw sensor signal；Electrical conductivity G when the sensor signal is sensor contacts sample gas is with sensor by calibrating
The ratio of electrical conductivity G0 during gas, i.e. G/G0；The detection time of the Electronic Nose sensor is 80s, and scavenging period is 60s,
Internal pump gas flow rate is 200ml/min；
3rd, the sensor response of detection by electronic nose orange juice is a data matrix, by many responses of sensor
Composition, from each sensor in 60s stationary values as characteristic value, in SPSS softwares, using principal component analysis, discriminant analysis
It is analyzed；It is respectively that, based on principal component analysis, linear discriminant analysis is sentenced to the preliminary of chitosan content in fruit juice that Fig. 2, Fig. 3 are
It is fixed；In Matlab softwares, according to characteristic value by bootstrap (boot-strap) resampling technique, training sample is continuously generated
And test sample, some decision trees are generated by training sample, so as to set up Random Forest model；It is right in Random Forest model
The number of decision tree is optimized in the variable number and random forest of the tree node of decision tree, and test result is thrown by decision tree
Depending on the average value of ticket result；The substantially step of random forest is as follows：
(1) bootstrap (boot-strap) resampling technique is utilized, T training set S is randomly generated1,S2,...,ST；
Bootstrap (boot-strap) resampling：If there is n different sample { x in set1,x2,...,xn, if every time
A sample is extracted from set S with putting back to, is extracted n times altogether, form new set S*, then set S*In include certain sample
This xi(i=1,2 ..., probability n) isAs n → ∞, have
Therefore, the total sample number of new set is identical with former set, but contains repeated sample (putting back to extraction), is only wrapped in new set
The sample of former set S collection and about 1-0.368*100%=63.2% is contained；
(2) each training set is utilized, corresponding decision tree C is generated1,C2,...CT；It is being based in each non-leaf nodes
Best divisional mode in the Split Attribute collection m of front nodal point enters line splitting to the node (in general, in this random forest
Growth course in, the value of m is to maintain constant)；
(3) each tree is completely grown up, without being pruned；
(4) for test set sample X, tested using each decision tree, obtained corresponding classification C1(X),C2
(X),...,CT(X)；
(5) by the way of ballot, the final result of Random Forest model is by the T average value of decision tree output valve
It is fixed.
4th, in Random Forest model, in the variable number and random forest of the tree node of decision tree decision tree
Number is optimized, and depending on average value of the test result by decision tree voting results, the result for finally optimizing obtains the change of tree node
It is 3 to measure, and the number of decision tree is 40.The Random Forest model after optimization is entered by correlation coefficient r, root-mean-square error RMSE value
Row is evaluated, specially：
Wherein, N represents the number of samples in modeling process；
XiI-th test value in for modeling process；
It is the average value of the sample responses value in modeling process；
YiIt is i-th predicted value of sample in modeling process；
It is the average value of the sample predictions value in modeling process.
Shown in Fig. 4, black squares point is the content of the pectase of the sample of the known pectase content model prediction
There is good correlation (r=0.9879, RMSE=0.0902) between actual pectase content.
5th, the unknown sample to be tested of pectase content is obtained into its Electronic Nose sensor response by step 2, will be sensed
The Random Forest model that device response steps for importing 4 is obtained, the pectase content in final prediction sample to be tested.Such as Fig. 4's is white
Triangulation point is the sample of unknown pectase content on color, and Random Forest model has prediction energy well to orange blossom pectase content
Power (r=0.9764, RMSE=0.1073).
Embodiment 2
Present case is using the shitosan commonly used in fruit juice as analysis object.The PEN2 type Electronic Noses of German AIRSENSE companies
For detecting instrument has elaborated in case 1.
1st, it is oranges and tangerines surface treatment is clean, lesion is rejected, get colors, the sample that size, shape are consistent.Oranges and tangerines go
Skin, squeezes the juice, filtering, abandons filter residue, and the oranges and tangerines filtrate of pure concentration is standby.
2nd, accurate accurate title 3g, 6g, 9g, 12g shitosan, pours into cold water, is soaked through 10 minutes, adds in 45 DEG C of hot bath methods
Thermal agitation is fully dissolved, at room temperature, during chitosan solution constant volume put into 100ml volumetric flasks.With the pure oranges and tangerines solution of step (1)
By 1：11 ratios repeat to mix, and treat and detection by electronic nose.
3rd, take the fruit juice 10ml after treatment to be put into the beaker of volume 500ml, sealed with double-deck preservative film, put at room temperature
30min is put, the fragrance level of orange juice is reached saturation in the beaker headspace of sealing；The fruit juice of each concentration scale
24 repeat samples, are the regression modeling collection Electronic Nose data in later stage respectively；By Electronic Nose internal pump by sealing container
Headspace gas suction Electronic Nose sensor array passage in, Electronic Nose sensor and sample gas react generation sensing
Device signal；Electrical conductivity G when the sensor signal is sensor contacts sample gas is with sensor when by calibration gas
Electrical conductivity G0 ratio, i.e. G/G0；The detection time of the Electronic Nose sensor is 80s, and scavenging period is 60s, internal pump
Gas flow rate is 200ml/min；
4th, the sensor response of detection by electronic nose orange juice is a data matrix, by many responses of sensor
Composition, from each sensor stabilization value as characteristic value, in SPSS softwares, is carried out using principal component analysis, discriminant analysis
Analysis；Fig. 6, Fig. 7 are to be respectively based on principal component analysis, preliminary judgement of the linear discriminant analysis to chitosan content in fruit juice；
In Matlab softwares, according to characteristic value by bootstrap (boot-strap) resampling technique, be continuously generated training sample and
Test sample, some decision trees are generated by training sample, so as to set up Random Forest model；In Random Forest model, fight to the finish
The number of decision tree is optimized in the variable number and random forest of the tree node of plan tree, and test result is voted by decision tree
Depending on the average value of result；The substantially step of random forest is as follows：
(1) bootstrap (boot-strap) resampling technique is utilized, T training set S is randomly generated1,S2,...,ST；
Bootstrap (boot-strap) resampling：If there is n different sample { x in set1,x2,...,xn, if every time
A sample is extracted from set S with putting back to, is extracted n times altogether, form new set S*, then set S*In include not individual sample
This xi(i=1,2 ..., probability n) isAs n → ∞, have
Therefore, the total sample number of new set is identical with former set, but contains repeated sample (putting back to extraction), is only wrapped in new set
The sample of former set S collection and about 1-0.368*100%=63.2% is contained；
(2) each training set is utilized, corresponding decision tree C is generated1,C2,...CT；It is being based in each non-leaf nodes
Best divisional mode in the Split Attribute collection m of front nodal point enters line splitting to the node (in general, in this random forest
Growth course in, the value of m is to maintain constant)；
(3) each tree is completely grown up, without being pruned；
(4) for test set sample X, tested using each decision tree, obtained corresponding classification C1(X),C2
(X),...,CT(X)；
(5) by the way of ballot, depending on the average value of the head piece result exported in T decision tree.
4th, in Random Forest model, in the variable number and random forest of the tree node of decision tree decision tree
Number is optimized, and depending on average value of the test result by decision tree voting results, the result for finally optimizing obtains the change of tree node
It is 3 to measure, and the number of decision tree is 50.The Random Forest model after optimization is entered by correlation coefficient r, root-mean-square error RMSE value
Row is evaluated, specially：
Wherein, N represents the number of samples in modeling process；
XiI-th test value in for modeling process；
It is the average value of the sample responses value in modeling process；
YiIt is i-th predicted value of sample in modeling process；
It is the average value of the sample predictions value in modeling process.
Shown in Fig. 8, black squares point is the content of the pectase of the sample of the known chitosan content model prediction
There is good correlation (r=0.9950, RMSE=0.0542) between actual chitosan content.
6th, the unknown sample to be tested of chitosan content is obtained into its Electronic Nose sensor response by step 3, will be sensed
The Random Forest model that device response steps for importing 5 is obtained, the chitosan content in final prediction sample to be tested.Such as Fig. 4's is white
Triangulation point is the sample of unknown chitosan content on color, and Random Forest model has prediction energy well to orange blossom pectase content
Power (r=0.9721, RMSE=0.1449).
Claims (2)
1. in a kind of quick detection fruit juice based on Electronic Nose fining agent method, it is characterised in that comprise the following steps：
(1) by oranges and tangerines surface treatment it is clean, remove the peel, squeeze the juice, filter, abandon filter residue, the oranges and tangerines filtrate of pure concentration is standby；Weigh respectively
The pectase of 3g, 6g, 9g, 12g, under 40 DEG C of water bath condition, pectase is repeated to dissolve, and filter residue is abandoned in filtering；In room temperature
Under, during pectinase solution constant volume put into 100ml volumetric flasks, with pure oranges and tangerines solution in mass ratio 1:11 ratio is mixed, and is stirred
Mix；After static 1 hour of blended fruit juice, it is centrifuged under the centrifugal speed not less than 4000r/min, is abandoned sediment, is taken supernatant
Liquid is stand-by；
(2) take the 10~30ml of fruit juice supernatant after treatment to be put into sealing container of the volume not less than 250ml, sealing, in room
Temperature is lower to place 30min, the fragrance level of orange juice is reached saturation in the beaker headspace of sealing；Each concentration scale
Solution respectively be no less than 3 repeat samples, be the later stage regression modeling collection Electronic Nose data；Will by Electronic Nose internal pump
In the sensor array passage of the headspace gas suction Electronic Nose in sealing container, Electronic Nose sensor occurs anti-with sample gas
Sensor signal should be produced；Electrical conductivity G when the sensor signal is sensor contacts sample gas is passing through with sensor
Electrical conductivity G during calibration gas0Ratio, i.e. G/G0；The detection time of the Electronic Nose sensor is 80s, and scavenging period is
60s, internal pump gas flow rate is 200ml/min；
(3) the sensor response of detection by electronic nose orange juice is a data matrix, by many response groups of sensor
Into from each sensor stabilization value as characteristic value, in SPSS softwares, being divided using principal component analysis, discriminant analysis
Analysis；In Matlab softwares, according to characteristic value by bootstrap (boot-strap) resampling technique, training sample is continuously generated
And test sample, some decision trees are generated by training sample, so as to set up Random Forest model；It is right in Random Forest model
The number of decision tree is optimized in the variable number and random forest of the tree node of decision tree, and test result is thrown by decision tree
Depending on the average value of ticket result；
(4) the unknown sample to be tested of agent content will be clarified its Electronic Nose sensor response is obtained by step (2), by sensor
The Random Forest model that response steps for importing (3) is obtained, the clarification agent content in final prediction sample to be tested.
2. in a kind of quick detection fruit juice based on Electronic Nose according to claim 1 fining agent method, it is characterised in that
In the step (3), the Random Forest model after optimization is evaluated by correlation coefficient r, root-mean-square error RMSE value, had
Body is：
Wherein, N represents the number of samples in modeling process；
XiI-th test value in for modeling process；
It is the average value of the sample responses value in modeling process；
YiIt is i-th predicted value of sample in modeling process；
It is the average value of the sample predictions value in modeling process.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510226487.XA CN104849323B (en) | 2015-05-06 | 2015-05-06 | A kind of method of fining agent in quick detection fruit juice based on Electronic Nose |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510226487.XA CN104849323B (en) | 2015-05-06 | 2015-05-06 | A kind of method of fining agent in quick detection fruit juice based on Electronic Nose |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104849323A CN104849323A (en) | 2015-08-19 |
| CN104849323B true CN104849323B (en) | 2017-06-30 |
Family
ID=53849126
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510226487.XA Expired - Fee Related CN104849323B (en) | 2015-05-06 | 2015-05-06 | A kind of method of fining agent in quick detection fruit juice based on Electronic Nose |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104849323B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105548268A (en) * | 2016-02-01 | 2016-05-04 | 浙江大学 | Method for fast predicting processing time of pecan based on electronic nose |
| CN105738503B (en) * | 2016-02-01 | 2017-10-20 | 浙江大学 | A kind of method based on electronic nose fast prediction hickory nut content of fatty acid |
| CN105699437B (en) * | 2016-02-01 | 2018-12-04 | 浙江大学 | A kind of quick nondestructive differentiating method of the different freshness hickory nuts based on electronic nose |
| CN107064438A (en) * | 2017-06-07 | 2017-08-18 | 南京财经大学 | A kind of method of pseudomonad rapid screening in fresh White mushroom |
| CN108651182B (en) * | 2018-03-27 | 2019-12-31 | 青岛农业大学 | Rapid screening method of high-pungency pepper breeding resources |
| CN111307973A (en) * | 2020-03-09 | 2020-06-19 | 西北农林科技大学 | Method for releasing combined-state aroma substances of kiwi fruit juice |
| CN112285296A (en) * | 2020-11-11 | 2021-01-29 | 重庆长安汽车股份有限公司 | Automobile interior part smell evaluation method based on electronic nose |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102353701A (en) * | 2011-07-20 | 2012-02-15 | 浙江大学 | Diagnostic method for insect attacks on crops by utilizing volatile matter |
| CN103376295A (en) * | 2013-07-24 | 2013-10-30 | 上海交通大学 | Rapid method for distinguishing oak sheet baking degree and ageing time of wine |
| CN103424433A (en) * | 2013-08-07 | 2013-12-04 | 浙江大学 | Remote wireless electronic nose system used for detecting quality of agricultural products |
| CN104569313A (en) * | 2015-01-15 | 2015-04-29 | 上海应用技术学院 | Method for rapidly analyzing fragrance of bayberry juice based on electronic nose smell fingerprint information |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| ES2268973B1 (en) * | 2005-05-23 | 2008-03-16 | Consejo Superior Investig. Cientificas | AUTOMATIC SYSTEM OF CONTINUOUS ANALYSIS OF THE EVOLUTION OF WINE. |
| DE102013111576A1 (en) * | 2013-10-21 | 2015-04-23 | Gea Mechanical Equipment Gmbh | Process for clarifying a flowable product with a centrifuge, in particular a separator |
-
2015
- 2015-05-06 CN CN201510226487.XA patent/CN104849323B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102353701A (en) * | 2011-07-20 | 2012-02-15 | 浙江大学 | Diagnostic method for insect attacks on crops by utilizing volatile matter |
| CN103376295A (en) * | 2013-07-24 | 2013-10-30 | 上海交通大学 | Rapid method for distinguishing oak sheet baking degree and ageing time of wine |
| CN103424433A (en) * | 2013-08-07 | 2013-12-04 | 浙江大学 | Remote wireless electronic nose system used for detecting quality of agricultural products |
| CN104569313A (en) * | 2015-01-15 | 2015-04-29 | 上海应用技术学院 | Method for rapidly analyzing fragrance of bayberry juice based on electronic nose smell fingerprint information |
Non-Patent Citations (2)
| Title |
|---|
| 基于电子鼻和电子舌的樱桃番茄汁品质检测方法研究;洪雪珍;《中国博士学位论文全文数据库 工程科技I辑》;20141215(第12期);第B024-46页 * |
| 电子鼻MOS传感器阵列优化及其在中药材快速鉴别中的应用;邹慧琴 等;《中国中药杂志》;20130131;第38卷(第2期);第161－166页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104849323A (en) | 2015-08-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104849323B (en) | A kind of method of fining agent in quick detection fruit juice based on Electronic Nose | |
| CN101382531B (en) | Method for detecting fresh degree of shrimp by electronic nose | |
| Gallardo et al. | Application of a potentiometric electronic tongue as a classification tool in food analysis | |
| Escuder-Gilabert et al. | Highlights in recent applications of electronic tongues in food analysis | |
| CN104914225B (en) | A kind of based on the method for fining agent content in sense of smell finger print information prediction fruit juice | |
| CN104849321B (en) | A kind of method based on smell finger-print quick detection Quality Parameters in Orange | |
| CN101493431A (en) | Method for detecting fresh degree of chicken meat by electronic nose | |
| CN106568907A (en) | Chinese mitten crab freshness damage-free detection method based on semi-supervised identification projection | |
| CN105954412B (en) | For the sensor array optimization method of hickory nut freshness detection | |
| Zhang et al. | Authentication of Zhongning wolfberry with geographical indication by mineral profile | |
| CN110954499B (en) | Mixed identification method and system for producing areas of imported salmon | |
| CN103822886A (en) | Analysis detection of biological humic acid and carbon ratio determination method | |
| Lou et al. | Stable Isotope Ratio and Elemental Profile Combined with Support Vector Machine for Provenance Discrimination of Oolong Tea (Wuyi‐Rock Tea) | |
| CN102914584A (en) | Rapid detection system and detection method for lactogenesis mixing | |
| CN104897738B (en) | A kind of method based on smell finger print information quick detection super-pressure fruit juice quality | |
| Griboff et al. | Differentiation between Argentine and Austrian red and white wines based on isotopic and multi-elemental composition | |
| Ulberth | Advances in testing for adulteration in honey | |
| CN104849328B (en) | The method that benzoic acid in fruit juice is quickly detected based on electronic tongues | |
| Yang et al. | A colorimetric sensor array based on peroxidase activity nanozyme for the highly efficient differential sensing of tea polyphenols and Tieguanyin adulteration | |
| Li et al. | Zooplankton size structure in relation to environmental factors in the Xiangxi Bay of Three Gorges Reservoir, China | |
| CN104849327B (en) | A kind of method that benzoic acid content in fruit juice is predicted based on sense of taste finger print information | |
| CN106770477A (en) | One kind optimization sensing data and pattern-recognition differentiate and adulterated fast detecting method to nectar source | |
| CN112268993A (en) | Method for rapidly monitoring river channel water quality parameters based on electronic nose technology | |
| CN209727908U (en) | A kind of electronic nose equipment based on food inspection | |
| Tsoi et al. | How do abiotic environmental variables shape benthic diatom assemblages in subtropical streams? |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170630 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |