CN103776797B - A kind of near infrared spectrum differentiates the method for flat interest Herb Gynostemmae Pentaphylli - Google Patents
A kind of near infrared spectrum differentiates the method for flat interest Herb Gynostemmae Pentaphylli Download PDFInfo
- Publication number
- CN103776797B CN103776797B CN201410065240.XA CN201410065240A CN103776797B CN 103776797 B CN103776797 B CN 103776797B CN 201410065240 A CN201410065240 A CN 201410065240A CN 103776797 B CN103776797 B CN 103776797B
- Authority
- CN
- China
- Prior art keywords
- sample
- near infrared
- neural network
- spectrum
- gynostemma pentaphylla
- 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 33
- 238000002329 infrared spectrum Methods 0.000 title claims abstract description 24
- 238000013528 artificial neural network Methods 0.000 claims abstract description 27
- 238000012549 training Methods 0.000 claims abstract description 14
- 230000003595 spectral effect Effects 0.000 claims abstract description 11
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 8
- 238000003062 neural network model Methods 0.000 claims abstract description 5
- 235000002956 Gynostemma pentaphyllum Nutrition 0.000 claims description 68
- 240000006509 Gynostemma pentaphyllum Species 0.000 claims description 60
- 238000001228 spectrum Methods 0.000 claims description 18
- 239000011159 matrix material Substances 0.000 claims description 12
- 238000007781 pre-processing Methods 0.000 claims description 10
- 240000008027 Akebia quinata Species 0.000 claims description 8
- 235000007756 Akebia quinata Nutrition 0.000 claims description 8
- 241001065361 Gynostemma Species 0.000 claims description 8
- 238000004458 analytical method Methods 0.000 claims description 6
- 238000012937 correction Methods 0.000 claims description 5
- 238000010606 normalization Methods 0.000 claims description 5
- 238000004497 NIR spectroscopy Methods 0.000 claims description 4
- 238000000149 argon plasma sintering Methods 0.000 claims description 3
- 238000001035 drying Methods 0.000 claims description 3
- 238000000862 absorption spectrum Methods 0.000 claims description 2
- 238000012847 principal component analysis method Methods 0.000 claims description 2
- 239000010453 quartz Substances 0.000 claims description 2
- VYPSYNLAJGMNEJ-UHFFFAOYSA-N silicon dioxide Inorganic materials O=[Si]=O VYPSYNLAJGMNEJ-UHFFFAOYSA-N 0.000 claims description 2
- 238000000605 extraction Methods 0.000 abstract description 2
- 239000000470 constituent Substances 0.000 abstract 2
- 210000005036 nerve Anatomy 0.000 abstract 2
- 230000007935 neutral effect Effects 0.000 abstract 2
- 238000004064 recycling Methods 0.000 abstract 1
- 238000005516 engineering process Methods 0.000 description 9
- 239000000126 substance Substances 0.000 description 7
- 230000006870 function Effects 0.000 description 6
- 238000003909 pattern recognition Methods 0.000 description 5
- 230000001186 cumulative effect Effects 0.000 description 4
- 238000011160 research Methods 0.000 description 4
- 238000012545 processing Methods 0.000 description 3
- -1 20W tungsten halogen Chemical class 0.000 description 2
- 240000007594 Oryza sativa Species 0.000 description 2
- 235000007164 Oryza sativa Nutrition 0.000 description 2
- 150000001413 amino acids Chemical class 0.000 description 2
- 239000008280 blood Substances 0.000 description 2
- 210000004369 blood Anatomy 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 230000005284 excitation Effects 0.000 description 2
- 235000013305 food Nutrition 0.000 description 2
- 150000004676 glycans Chemical class 0.000 description 2
- 238000012567 pattern recognition method Methods 0.000 description 2
- 229920001282 polysaccharide Polymers 0.000 description 2
- 239000005017 polysaccharide Substances 0.000 description 2
- 239000001397 quillaja saponaria molina bark Substances 0.000 description 2
- 238000012827 research and development Methods 0.000 description 2
- 235000009566 rice Nutrition 0.000 description 2
- 238000007873 sieving Methods 0.000 description 2
- 238000010183 spectrum analysis Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 238000012546 transfer Methods 0.000 description 2
- RBTBFTRPCNLSDE-UHFFFAOYSA-N 3,7-bis(dimethylamino)phenothiazin-5-ium Chemical compound C1=CC(N(C)C)=CC2=[S+]C3=CC(N(C)C)=CC=C3N=C21 RBTBFTRPCNLSDE-UHFFFAOYSA-N 0.000 description 1
- 241000219104 Cucurbitaceae Species 0.000 description 1
- 241000208251 Gymnema Species 0.000 description 1
- 240000007817 Olea europaea Species 0.000 description 1
- 241000203383 Schefflera Species 0.000 description 1
- GAMYVSCDDLXAQW-AOIWZFSPSA-N Thermopsosid Natural products O(C)c1c(O)ccc(C=2Oc3c(c(O)cc(O[C@H]4[C@H](O)[C@@H](O)[C@H](O)[C@H](CO)O4)c3)C(=O)C=2)c1 GAMYVSCDDLXAQW-AOIWZFSPSA-N 0.000 description 1
- 238000010521 absorption reaction Methods 0.000 description 1
- 239000003513 alkali Substances 0.000 description 1
- 230000036772 blood pressure Effects 0.000 description 1
- 229910052729 chemical element Inorganic materials 0.000 description 1
- 239000003153 chemical reaction reagent Substances 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000003814 drug Substances 0.000 description 1
- 238000003912 environmental pollution Methods 0.000 description 1
- 229930003944 flavone Natural products 0.000 description 1
- 150000002212 flavone derivatives Chemical class 0.000 description 1
- 235000011949 flavones Nutrition 0.000 description 1
- 229930003935 flavonoid Natural products 0.000 description 1
- 235000017173 flavonoids Nutrition 0.000 description 1
- 229910052736 halogen Inorganic materials 0.000 description 1
- 210000004185 liver Anatomy 0.000 description 1
- QSHDDOUJBYECFT-UHFFFAOYSA-N mercury Chemical compound [Hg] QSHDDOUJBYECFT-UHFFFAOYSA-N 0.000 description 1
- 229910052753 mercury Inorganic materials 0.000 description 1
- 229960000907 methylthioninium chloride Drugs 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 150000007524 organic acids Chemical class 0.000 description 1
- 238000000643 oven drying Methods 0.000 description 1
- 239000003208 petroleum Substances 0.000 description 1
- 230000000144 pharmacologic effect Effects 0.000 description 1
- 239000010465 pomace olive oil Substances 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000010298 pulverizing process Methods 0.000 description 1
- 238000012797 qualification Methods 0.000 description 1
- 238000003908 quality control method Methods 0.000 description 1
- 230000036632 reaction speed Effects 0.000 description 1
- 238000007670 refining Methods 0.000 description 1
- 238000000985 reflectance spectrum Methods 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 239000011347 resin Substances 0.000 description 1
- 229920005989 resin Polymers 0.000 description 1
- 229930182490 saponin Natural products 0.000 description 1
- 150000007949 saponins Chemical class 0.000 description 1
- 238000001179 sorption measurement Methods 0.000 description 1
- 239000011573 trace mineral Substances 0.000 description 1
- 235000013619 trace mineral Nutrition 0.000 description 1
- 229910052721 tungsten Inorganic materials 0.000 description 1
- 239000010937 tungsten Substances 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- 239000010463 virgin olive oil Substances 0.000 description 1
- 239000011782 vitamin Substances 0.000 description 1
- 229940088594 vitamin Drugs 0.000 description 1
- 229930003231 vitamin Natural products 0.000 description 1
- 235000013343 vitamin Nutrition 0.000 description 1
- VHBFFQKBGNRLFZ-UHFFFAOYSA-N vitamin p Natural products O1C2=CC=CC=C2C(=O)C=C1C1=CC=CC=C1 VHBFFQKBGNRLFZ-UHFFFAOYSA-N 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Landscapes
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The present invention provide a kind of near infrared spectrum differentiate flat interest Herb Gynostemmae Pentaphylli method, comprise the steps: A, set up flat interest Herb Gynostemmae Pentaphylli near infrared spectrum differentiate model: A 1, select spectral region 4000 12500cm‑1, scan flat interest Herb Gynostemmae Pentaphylli near infrared light spectrogram;A 2, to spectral region 4000 9500cm‑1Data carry out pretreatment;A 3, extraction main constituent;A 4, set up artificial nerve network model: take artificial neural network algorithm, determine that according to inputoutput data feature the structure of neutral net, recycling training data train this neutral net;MATLAB software is used to set up the BP artificial nerve network model of input layer 10 hidden layer node 5 output layer node 2;B, the discriminating of unknown sample: unknown sample scans near infrared light spectrogram under the same conditions, choose main constituent number, according to the neural network model trained to judge the true and false of unknown sample, output node represents with binary code respectively, 10 representatives are flat interest Herb Gynostemmae Pentaphylli, and 01 representative is non-flat interest Herb Gynostemmae Pentaphylli.
Description
Technical Field
The invention relates to a method for identifying Pingli gynostemma pentaphylla by near infrared spectrum, in particular to a method for identifying Pingli gynostemma pentaphylla by combining near infrared spectrum technology with artificial neural network algorithm, belonging to the field of near infrared spectrum detection and analysis.
Background
Gynostemma pentaphylla, also known as fiveleaf gynostemma herb and scandent schefflera root, has the functions of reducing blood pressure, blood fat and blood sugar, protecting heart and liver, regulating fat and losing weight, and is called as 'longevity herb'. Since the quality control bureau of 2004 performs regional protection of the origin of the Shaanxi Pingli gynostemma pentaphylla, the price of Pingli gynostemma pentaphylla is multiplied, the phenomena of inferior quality and counterfeit and shoddy occur occasionally in the market, and the technology for identifying the origin of the Shaanxi Pingli gynostemma pentaphylla is imperative in order to effectively identify the Pingli gynostemma pentaphylla of different origins and protect the rights and interests of consumers.
The near infrared spectrum has the advantages of high reaction speed, rich information content, less pretreatment, no environmental pollution and the like, is widely applied in many fields, and becomes one of the most popular spectral analysis technologies in current research. The near infrared spectrum contains a great deal of information of the sample, so that the near infrared analysis technology and the pattern recognition method are combined, and the grade and the category of the sample can be more effectively distinguished. The near-infrared pattern recognition technology is a technology for deducing the attribution of a substance from near-infrared data of the substance by applying a chemical pattern recognition method. All methods of chemical pattern recognition can be used for the study of near infrared pattern recognition. At present, the near-infrared-based pattern recognition technology is widely applied to the fields of agriculture, medicine, food, petroleum and the like, and plays an important role in the aspects of true and false discrimination, grade classification, origin and place identification and the like. However, the recognition models established by the pattern recognition are all specific to specific products and have strong specificity. The applicant has adopted the near infrared spectroscopy combined with the mahalanobis distance algorithm and the qualification test to effectively identify the rice with the rice water; and successfully identifying the virgin olive oil and the olive pomace oil by using a fi sher discrimination algorithm. The invention discloses a method for identifying Pingli gynostemma pentaphylla based on near infrared spectrum technology and artificial neural network algorithm. At present, the researches on gynostemma pentaphylla by scholars at home and abroad mainly focus on chemical components and pharmacological actions of gynostemma pentaphylla. It mainly contains saponin [1], polysaccharide [2], amino acid [4], flavone [3], organic acid and trace elements [4] and other chemical components. The reports prove that the components of the gynostemma pentaphylla in different producing areas are different, so the methods have certain reference value for identifying the authenticity of the gynostemma pentaphylla in different producing areas, but the report for identifying the authenticity of the gynostemma pentaphylla by utilizing the component difference is not found at present.
Disclosure of Invention
The invention aims to provide a method for quickly and accurately identifying the truth of the Pingli gynostemma pentaphylla by combining a near infrared spectrum technology and an artificial neural network algorithm.
The technical scheme of the invention is as follows: the method for identifying Pingli gynostemma pentaphylla by using the near infrared spectrum comprises the following steps:
A. establishing a near infrared spectrum identification model of Pingli gynostemma pentaphylla
A-1, selective spectral range 4000--1Scanning a near-infrared spectrogram of the Pingli gynostemma pentaphylla;
a-2, 12500cm in spectral range 4000--1Preprocessing the data;
a-3, extracting main components;
a-4, establishing an artificial neural network model: determining the structure of a neural network according to the characteristics of input and output data by adopting an artificial neural network algorithm, and training the neural network by utilizing training data to obtain an identification model of the fiveleaf gynostemma herb;
B. identification of unknown samples
Scanning a near-infrared spectrogram of an unknown sample under the same condition, selecting the number of main components, judging the authenticity of the unknown sample according to a trained neural network model, respectively representing output nodes by binary codes, wherein 10 represents Pingli gynostemma pentaphylla, and 01 represents non-Pingli gynostemma pentaphylla.
The method for identifying the Pingli gynostemma pentaphylla by the near infrared spectrum comprises the following steps of A-1, wherein the scanning Pingli gynostemma pentaphylla near infrared spectrogram comprises the following steps: drying and crushing an effective amount of a Gynostemma Pentaphyllum sample, uniformly placing the dried and crushed Gynostemma Pentaphyllum sample in a quartz sample cell, and scanning an absorption spectrum by using a Fourier near infrared spectrometer; the scanning mode is rotation diffuse reflection, and the resolution ratio is 8cm-1Scanning each sample for multiple times, and taking an average spectrum as a final analysis spectrum of the sample;
the method for identifying the Pingli gynostemma pentaphylla by the near infrared spectrum comprises the following steps of A-2, wherein the data preprocessing of the Pingli gynostemma pentaphylla near infrared spectrogram comprises the following steps: and preprocessing of multivariate scattering correction and proper normalization is carried out on the spectrum of the stranded blue sample, and the influence of interference factors such as sample nonuniformity, light scattering, instrument noise and the like is eliminated through the preprocessing, so that the prediction precision and the stability of the model are improved.
In the method for identifying Pingli gynostemma pentaphylla by using the near infrared spectrum, the main component extraction in the step A-3 is to reduce the dimension of spectrogram information by using a main component analysis method, the cumulative contribution rate of the first 10 main components is 99.99%, the limited amount of input is used for reducing the calculation complexity of the model, and the prediction precision of the model is improved.
The method for identifying the Pingli gynostemma pentaphylla by the near infrared spectrum comprises the following steps of A-4, establishing an artificial neural network model by the Pingli gynostemma pentaphylla, wherein the artificial neural network model comprises an input layer node 10, a hidden layer node 5 and an output layer node 2, and by using MATLAB software:
a-4-1, determining the number of nodes of an input layer: taking 10 principal component scores as parameters, and determining the input layer node of the network to be 10;
a-4-2, determining the number of hidden nodes: the following equation was used to determine:
m and n are respectively the number of input nodes and output nodes, the number of hidden nodes can obtain an initial value through a formula, and then the initial value is corrected by utilizing a step-by-step growth method to obtain an empirical value 5;
a-4-3, determining the number of nodes of an output layer: and determining 2 output nodes of the neural network according to two results of judging whether the gynostemma pentaphylla sample belongs to a peaceful producing area or a non-peaceful producing area.
The method for identifying the Pingli gynostemma pentaphylla by the near infrared spectrum comprises the following steps of:
let x1,x2,…,xnIs a sample taken from the population x, where xi=(xi1,xi2,…,xip)′(i=1,2,…n);
The sample observation matrix is recorded as:
each row of x corresponds to a sample and each column corresponds to a variable;
recording the sample covariance matrix and the sample correlation coefficient matrix as follows:
wherein,is the sample average;
taking S as an estimate of sigma,as an estimate of R, from S orThe principal components of the sample can be determined.
The method for identifying Pingli gynostemma pentaphylla by near infrared spectroscopy is characterized by comprising the following steps of: the principal component of the sample is composed of a matrix of correlation coefficients of the slave samplesStarting and solving:
is provided withIs composed ofThe number of p characteristic values of (a),for the corresponding orthonormal unit feature vector, the p principal components of the sample are
Sample xiNormalized observed valueSubstituting into the jth main component to obtain a sample xiJ-th principal component score of
The invention selects the spectral range of 4000--1According to the original spectral analysis, the characteristic peak of the interval containing the main near-infrared absorption of the gynostemma pentaphylla sample is subjected to preprocessing of multi-element scattering correction and vector normalization on the spectrum of the gynostemma pentaphylla sample, so that the influence of interference factors such as sample nonuniformity, light scattering and instrument noise is eliminated, and the prediction precision and stability of a neural network model are improved; according to the method, the principal component scores of the first 10 principal components are extracted and used as new variable input, the computation complexity of the model is reduced through the limited input, and the prediction accuracy of the model is improved; the method can quickly realize the identification of the trueness of the fiveleaf gynostemma herb, and the identification accuracy of the model training set and the prediction set is 100 percent. The established discrimination model has great significance for realizing the discrimination of the trueness of the fiveleaf gynostemma herb.
Drawings
FIG. 1 is a diagram of a neural network architecture
Each letter in the figure indicates:
xjan input representing the jth node of the input layer, j ═ 1, …, M;
wijrepresenting the weight from the ith node of the hidden layer to the jth node of the input layer;
θia threshold value representing the ith node of the hidden layer;
Φxan excitation function representing the hidden layer;
wkjrepresenting the weight value from the kth node of the output layer to the ith node of the hidden layer, wherein i is 1, … and q;
aka threshold value indicating the kth node of the output layer, k being 1, …, L;
Ψxan excitation function representing an output layer;
Okrepresenting the output of the kth node of the output layer.
FIG. 2 is a near infrared spectrum of Pingli gynostemma pentaphyllum
FIG. 3 cumulative contribution rate of the top ten principal component scores
FIG. 4 is a program diagram of a neural network computational implementation
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and specific embodiments.
1. Instruments and reagents: the sample spectrum was collected using MPA near infrared spectrometer and diffuse reflectance accessory from Bruker, Germany, with its own analytical software and MATLAB software.
1-1, instrument noise
Preparing a voltage-stabilized power supply, starting up the device, preheating until the instrument is sufficiently stable, and ensuring the proper test environment temperature to be 15-25 ℃;
1-2, wavelength accuracy and reproducibility
The accuracy of the wavelength was corrected with a low pressure mercury lamp and a methylene blue solution with a mass fraction of 0.005% to prevent drift.
2. Sample preparation and spectral scanning: the gynostemma pentaphylla samples were purchased at the place of origin.
TABLE 1 Gynostemma pentaphyllum sample information table
3. Establishing near infrared spectrum identification model of geographical sign gynostemma pentaphylla
3-1, scanning the gynostemma pentaphylla near infrared spectrogram
The method comprises the steps of drying a gynostemma pentaphyllum sample at 60 ℃ for 4 hours, crushing, sieving with a 60-mesh sieve, uniformly placing about 50g of the sample in a sample cell, and collecting a rotating diffuse reflection spectrum of the sample by using an MPA near infrared spectrometer and a diffuse reflection accessory of Bruker company, Germany, as shown in figure 2. The light source of the instrument is a 20W tungsten halogen lamp, and the spectral range is 4000-12500cm-1. The spectrum acquisition software used was OPUS 6.5 from Bruker, with a scan number of 64 and a resolution of 8cm-1The reference is the built-in background of the instrument. Each sample was scanned 3 times and the average spectrum of the 3 scans was taken as the sample spectrum.
3-2, preprocessing the spectral data
At the wave band 4000--1The discrimination model is established, MATLAB software is used for processing the spectrum of the sample, a multivariate scattering correction and vector normalization preprocessing method is used for processing the spectrum, and a principal component analysis method is used for reducing the dimension of the spectrum of the stranded blue sample.
3-3, extracting main components;
the method for identifying the Pingli gynostemma pentaphylla by the near infrared spectrum comprises the following steps of:
let x1,x2,…,xnIs a sample taken from the population x, where xi=(xi1,xi2,…,xip)′(i=1,2,…n);
The sample observation matrix is recorded as:
each row of x corresponds to a sample and each column corresponds to a variable;
recording the sample covariance matrix and the sample correlation coefficient matrix as follows:
wherein,is the sample average;
taking S as an estimate of sigma,as an estimate of R, from S orThe principal components of the sample can be determined.
The method for identifying Pingli gynostemma pentaphylla by near infrared spectroscopy is characterized by comprising the following steps of: the principal component of the sample is composed of a matrix of correlation coefficients of the slave samplesStarting and solving:
is provided withIs composed ofThe number of p characteristic values of (a),for the corresponding orthonormal unit feature vector, the p principal components of the sample are
Sample xiNormalized observed valueSubstituting into the jth main component to obtain a sample xiJ-th principal component score of
Using MATLAB software to take the top 10 principal component scores as the input level nodes of the network, the cumulative contribution rate of the top ten principal components in fig. 3 reaches 99.99%, as shown in table 2. Can represent most effective information of the gynostemma pentaphylla, so the scores of the first ten principal components are adopted as the parameters of the nodes of the input layer.
TABLE 2 cumulative contribution of the first ten principal components
The output point of the network adopts binary code to represent the output of the producing area, as shown in table 3, the samples of eight producing areas are only divided into two categories, one is the original producing area peaceful gynostemma pentaphylla, which is represented by binary code 10; the other is non-native gynostemma pentaphyllum, represented by binary code 01.
TABLE 3 output node origin code
3-4, establishing artificial neural network model
3-4-1, determining the number of nodes of an input layer: taking 10 principal component scores as parameters, determining the input layer nodes of the network to be 10, dividing the sample into a training set and a prediction set, wherein the training set is used for training the neural network, and establishing the artificial neural network model. The prediction set is used for verifying the accuracy of the network model, and if the accuracy of the method is found to be not up to the requirement, the model is trained again, namely the parameters are optimized until the accurate neural network model is established.
In this embodiment, a total of 405 samples are collected, 90 samples are collected in a fair producing area, 45 samples are collected in other producing areas, and the 405 samples are randomly divided into a training set and a prediction set, wherein the number of the training set samples is 270, and the number of the prediction set samples is 135.
3-4-2, determining the number of hidden nodes: the following equation was used to determine:
m and n are respectively the number of input nodes and output nodes, the number of hidden nodes can obtain an initial value through a formula, and then the initial value is corrected by utilizing a step-by-step growth method to obtain an empirical value 5;
3-4-3, determining the number of nodes of an output layer: and determining 2 output nodes of the neural network according to two results of judging whether the gynostemma pentaphylla sample belongs to a peaceful producing area or a non-peaceful producing area.
According to the calculation procedure shown in fig. 4, transfer functions tansig of the input layer and the hidden layer of the network are determined, the transfer function of the output layer is a thingdx function, and the training target is set to be 1x10-6The learning rate of the network is 0.05, the set training iteration times are 1000, and the number of hidden nodes is 5. MATLAB software is used for establishing a BP artificial neural network model of the input layer node 10, the hidden layer node 5 and the output layer node 2.
4. Unknown sample identification
4-1, sample treatment: oven drying about 50g unknown sample at 60 deg.C for 4 hr, pulverizing, sieving with 60 mesh sieve, and uniformly placing in sampleIn the cell, the rotational diffuse reflectance spectrum of the sample is collected. The spectral range is 4000-12500cm-1Scanning times 64, resolution 8cm-1The reference is the built-in background of the instrument. Each sample was scanned 3 times and the average spectrum of the 3 scans was taken as the sample spectrum.
4-2 neural network recognition model
Selecting the spectral range of the unknown sample to be 4000--1And (3) processing the internal near-infrared spectrogram by adopting a multivariate scattering correction and vector normalization preprocessing method, extracting scores of the first 10 main components by using MATLAB software, inputting the scores into a discrimination model, judging that the fiveleaf gynostemma herb is obtained if the output result of the model is 10, and judging that the fiveleaf gynostemma herb is not fiveleaf gynostemma herb if the output result is 01. The results show that the recognition accuracy of 270 training samples is 100%, the recognition accuracy of 135 prediction samples is 100%, and the specific verification results are shown in table 4.
TABLE 4 judgment result of recognition feasibility of the artificial neural network to Gymnema pentaphyllum
The above description is only provided as an implementable technical solution of the method for identifying the Pingli gynostemma pentaphylla by using the near infrared spectrum of the invention, and is not a single limitation condition for the technical solution.
Reference documents:
[1] "bamboo-Ben Chang Song", a research on the composition of Cucurbitaceae plants-saponin component of Gynostemma pentaphyllum (Ma) Makino.) Makino, volume 7, phase 5, 1985, month 4
[2] Research and development of water-soluble polysaccharide of Gynostemma pentaphyllum Makino by WANGSHUANGJING, Lord Shihui, alkali, food research and development, Vol.27, No. 5, 2006, 5 months
[3] Research on refining process of gynostemma pentaphylla flavonoid compound by using Wangqinghao, Zhangluo and macroporous adsorption resin [ J ] forest chemical communication, volume 39, 6 th year, 6 th 2005
[4] Analysis of amino acids, vitamins and various chemical elements in Gynostemma pentaphyllum, Deng Shilin, well-known Juan, J, volume 19, university of Hunan medical sciences, volume 6, month 6, 1994
Claims (2)
1. A method for identifying Pingli gynostemma pentaphylla by near infrared spectrum is characterized by comprising the following steps:
A. establishing a near infrared spectrum identification model of Pingli gynostemma pentaphylla
A-1, selective spectral range 4000--1Scanning a near-infrared spectrogram of the Pingli gynostemma pentaphylla: drying and crushing an effective amount of a Gynostemma Pentaphyllum sample, uniformly placing the dried and crushed Gynostemma Pentaphyllum sample in a quartz sample cell, and scanning an absorption spectrum by using a Fourier near infrared spectrometer; the scanning mode is rotation diffuse reflection, and the resolution ratio is 8cm-1Each sample is scanned multiple times and leveledThe average spectrum is the final analysis spectrum of the sample;
a-2, selecting the spectrum range of 4000--1The data of (2) are preprocessed: preprocessing of multivariate scattering correction and proper normalization is carried out on the spectrum of the stranded blue sample, the influence of factors such as sample nonuniformity, light scattering and instrument noise interference is eliminated through the preprocessing, and the prediction precision and stability of the model are improved;
a-3, extracting main components: dimensionality reduction is carried out on spectrogram information through a principal component analysis method, the accumulated contribution rate of the first 10 principal components is 99.99%, the computation complexity of the model is reduced through limited input, and the prediction accuracy of the model is improved;
the principal component of the sample is determined by the following method:
let x1,x2,…,xnIs a sample taken from the population x, where xi=(xi1,xi2,…,xip)′,i=1,2,…n;
The sample observation matrix is recorded as:
each row of x corresponds to a sample and each column corresponds to a variable;
recording the sample covariance matrix and the sample correlation coefficient matrix as follows:
wherein,is the sample average;
taking S as the estimate of ∑,as an estimate of R, from S orStarting to obtain the principal component of the sample;
a-4, establishing an artificial neural network model: determining the structure of a neural network according to the characteristics of input and output data by adopting an artificial neural network algorithm, and training the neural network by utilizing training data to obtain an identification model of the fiveleaf gynostemma herb;
B. identification of unknown samples
Scanning a near-infrared spectrogram of an unknown sample under the same condition, selecting the number of main components, judging the authenticity of the unknown sample according to a trained neural network model, wherein output nodes are respectively represented by binary numbers, 10 represents Pinglie gynostemma pentaphylla, and 01 represents non-Pinglie gynostemma pentaphylla;
the principal component of the sample is determined by a matrix of correlation coefficients of the slave samplesStarting and solving:
is provided withIs composed ofThe number of p characteristic values of (a),for the corresponding orthonormal unit feature vector, the p principal components of the sample are
Sample xiNormalized observed valueSubstituting into the jth main component to obtain a sample xiJ-th principal component score of
2. The method for identifying Gynostemma pentaphyllum by near infrared spectroscopy according to claim 1, wherein: the step A-4 of establishing the artificial neural network model comprises the steps of establishing a BP artificial neural network model of an input layer node 10, a hidden layer node 5 and an output layer node 2 by using MATLAB software:
a-4-1, determining the number of nodes of an input layer: taking 10 principal component scores as parameters, and determining the input layer node of the network to be 10;
a-4-2, determining the number of hidden nodes: the following equation was used to determine:
m and n are respectively the number of input nodes and the number of output nodes, the number of hidden nodes obtains an initial value by a formula, and then the initial value is corrected by a step-by-step growth method to obtain an empirical value 5;
a-4-3, determining the number of nodes of an output layer: and determining 2 output nodes of the neural network according to two results of judging whether the gynostemma pentaphylla sample belongs to a peaceful producing area or a non-peaceful producing area.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410065240.XA CN103776797B (en) | 2014-02-25 | 2014-02-25 | A kind of near infrared spectrum differentiates the method for flat interest Herb Gynostemmae Pentaphylli |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410065240.XA CN103776797B (en) | 2014-02-25 | 2014-02-25 | A kind of near infrared spectrum differentiates the method for flat interest Herb Gynostemmae Pentaphylli |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103776797A CN103776797A (en) | 2014-05-07 |
CN103776797B true CN103776797B (en) | 2016-09-21 |
Family
ID=50569301
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410065240.XA Expired - Fee Related CN103776797B (en) | 2014-02-25 | 2014-02-25 | A kind of near infrared spectrum differentiates the method for flat interest Herb Gynostemmae Pentaphylli |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103776797B (en) |
Families Citing this family (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106485094A (en) * | 2016-11-30 | 2017-03-08 | 华东理工大学 | A kind of PX oxidation reaction production process agent model modeling method |
CN108324286B (en) * | 2018-01-26 | 2020-11-10 | 重庆大学 | Infrared noninvasive blood glucose detection device based on PCA-NARX correction algorithm |
CN108444944A (en) * | 2018-02-24 | 2018-08-24 | 盐城工学院 | A kind of radix polygoni multiflori powder place of production discrimination method for the spectrometry that diffused based on near-infrared |
CN108509997A (en) | 2018-04-03 | 2018-09-07 | 深圳市药品检验研究院(深圳市医疗器械检测中心) | A method of Chemical Pattern Recognition is carried out to the true and false that Chinese medicine Chinese honey locust is pierced based on near-infrared spectrum technique |
WO2023279338A1 (en) * | 2021-07-08 | 2023-01-12 | Shanghaitech University | Neural spectral field reconstruction for spectrometer |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1603794A (en) * | 2004-11-02 | 2005-04-06 | 江苏大学 | Method and device for rapidly detecting tenderness of beef utilizing near infrared technology |
CN101532954A (en) * | 2008-03-13 | 2009-09-16 | 天津天士力现代中药资源有限公司 | Method for identifying traditional Chinese medicinal materials by combining infra-red spectra with cluster analysis |
CN101957316A (en) * | 2010-01-18 | 2011-01-26 | 河北大学 | Method for authenticating Xiangshui rice by near-infrared spectroscopy |
CN101968438A (en) * | 2010-09-25 | 2011-02-09 | 西北农林科技大学 | Method for distinguishing water injection of raw material muscles quickly |
CN101995392A (en) * | 2010-11-15 | 2011-03-30 | 中华人民共和国上海出入境检验检疫局 | Method for rapidly detecting adulteration of olive oil |
CN102636452A (en) * | 2012-05-03 | 2012-08-15 | 中国科学院长春光学精密机械与物理研究所 | NIR (Near Infrared Spectrum) undamaged identification authenticity method for wild ginseng |
-
2014
- 2014-02-25 CN CN201410065240.XA patent/CN103776797B/en not_active Expired - Fee Related
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1603794A (en) * | 2004-11-02 | 2005-04-06 | 江苏大学 | Method and device for rapidly detecting tenderness of beef utilizing near infrared technology |
CN101532954A (en) * | 2008-03-13 | 2009-09-16 | 天津天士力现代中药资源有限公司 | Method for identifying traditional Chinese medicinal materials by combining infra-red spectra with cluster analysis |
CN101957316A (en) * | 2010-01-18 | 2011-01-26 | 河北大学 | Method for authenticating Xiangshui rice by near-infrared spectroscopy |
CN101968438A (en) * | 2010-09-25 | 2011-02-09 | 西北农林科技大学 | Method for distinguishing water injection of raw material muscles quickly |
CN101995392A (en) * | 2010-11-15 | 2011-03-30 | 中华人民共和国上海出入境检验检疫局 | Method for rapidly detecting adulteration of olive oil |
CN102636452A (en) * | 2012-05-03 | 2012-08-15 | 中国科学院长春光学精密机械与物理研究所 | NIR (Near Infrared Spectrum) undamaged identification authenticity method for wild ginseng |
Non-Patent Citations (3)
Title |
---|
Fast discrimination of traditional Chinese medicine according to geographical origins with FTIR spectroscopy and advanced pattern recognition techniques;Ning Li et al.;《OPTICS EXPRESS》;20060821;第14卷(第17期);第2-3节 * |
中草药绞股蓝的傅里叶变换红外和拉曼光谱分析;郭萍 等;《光谱学与光谱分析》;20041031;第24卷(第10期);第1210-1212页 * |
基于主成分分析和人工神经网络的五味子质量鉴定方法研究;姜健 等;《红外》;20091231;第30卷(第12期);第2.1、2.3.1节,第3节 * |
Also Published As
Publication number | Publication date |
---|---|
CN103776797A (en) | 2014-05-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107677647B (en) | Method for identifying origin of traditional Chinese medicinal materials based on principal component analysis and BP neural network | |
CN103776797B (en) | A kind of near infrared spectrum differentiates the method for flat interest Herb Gynostemmae Pentaphylli | |
CN104677875B (en) | A kind of three-dimensional fluorescence spectrum combines the method that parallel factor differentiates different brands Chinese liquor | |
CN108875913B (en) | Tricholoma matsutake rapid nondestructive testing system and method based on convolutional neural network | |
Kunz et al. | Updating a synchronous fluorescence spectroscopic virgin olive oil adulteration calibration to a new geographical region | |
Xun et al. | Mapping cotton cultivated area combining remote sensing with a fused representation-based classification algorithm | |
CN106092990A (en) | A kind of three-dimensional fluorescence spectrum discrimination method of lycium barbarum | |
CN104374739A (en) | Identification method for authenticity of varieties of seeds on basis of near-infrared quantitative analysis | |
Sonobe et al. | Estimating leaf carotenoid contents of shade-grown tea using hyperspectral indices and PROSPECT–D inversion | |
CN103278467A (en) | Rapid nondestructive high-accuracy method with for identifying abundance degree of nitrogen element in plant leaf | |
CN104778349B (en) | One kind is used for rice table soil nitrogen application Classified Protection | |
Li et al. | Manufacturer identification and storage time determination of “Dong’e Ejiao” using near infrared spectroscopy and chemometrics | |
CN104568824A (en) | Method and device for detecting freshness grade of shrimps based on visible/near-infrared spectroscopy | |
CN109187443A (en) | Water body bacterial micro-organism based on multi-wavelength transmitted spectrum accurately identifies method | |
Wang et al. | Extraction and classification of origin characteristic peaks from rice Raman spectra by principal component analysis | |
CN110378373B (en) | Tea variety classification method for fuzzy non-relevant linear discriminant analysis | |
Mishra et al. | A generic workflow combining deep learning and chemometrics for processing close-range spectral images to detect drought stress in Arabidopsis thaliana to support digital phenotyping | |
Chen et al. | A rapid and effective method for species identification of edible boletes: FT-NIR spectroscopy combined with ResNet | |
CN114112983A (en) | Python data fusion-based Tibetan medicine all-leaf artemisia rupestris L producing area distinguishing method | |
CN112782148B (en) | Method for rapidly identifying Arabica and Robertia coffee beans | |
Fan et al. | Non-destructive detection of single-seed viability in maize using hyperspectral imaging technology and multi-scale 3D convolutional neural network | |
CN107101972A (en) | A kind of near infrared spectrum quick detection radix tetrastigme place of production method | |
CN116151454A (en) | Method and system for predicting yield of short-forest linalool essential oil by multispectral unmanned aerial vehicle | |
Liu et al. | A modified feature fusion method for distinguishing seed strains using hyperspectral data | |
CN111881738B (en) | Near infrared spectrum classification method for tea leaves through nuclear fuzzy orthogonal discriminant analysis |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20160921 Termination date: 20180225 |