CN106706558A - Method for eliminating abnormal sample in calibration set - Google Patents
Method for eliminating abnormal sample in calibration set Download PDFInfo
- Publication number
- CN106706558A CN106706558A CN201710017074.XA CN201710017074A CN106706558A CN 106706558 A CN106706558 A CN 106706558A CN 201710017074 A CN201710017074 A CN 201710017074A CN 106706558 A CN106706558 A CN 106706558A
- Authority
- CN
- China
- Prior art keywords
- sample
- calibration set
- exceptional
- suspicious
- exceptional sample
- 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
- 238000000034 method Methods 0.000 title claims abstract description 39
- 230000002159 abnormal effect Effects 0.000 title claims abstract description 8
- 238000010219 correlation analysis Methods 0.000 claims abstract description 8
- 238000002790 cross-validation Methods 0.000 claims abstract description 7
- 238000010586 diagram Methods 0.000 claims description 21
- 238000000513 principal component analysis Methods 0.000 claims description 17
- 238000004458 analytical method Methods 0.000 claims description 6
- 230000007812 deficiency Effects 0.000 claims description 5
- 238000001228 spectrum Methods 0.000 claims description 5
- 238000002329 infrared spectrum Methods 0.000 claims description 4
- 230000003595 spectral effect Effects 0.000 claims description 3
- 230000002411 adverse Effects 0.000 abstract description 2
- 230000000694 effects Effects 0.000 abstract 1
- 238000001514 detection method Methods 0.000 description 3
- 238000001704 evaporation Methods 0.000 description 3
- 239000012141 concentrate Substances 0.000 description 2
- 230000008030 elimination Effects 0.000 description 2
- 238000003379 elimination reaction Methods 0.000 description 2
- TVMXDCGIABBOFY-UHFFFAOYSA-N octane Chemical compound CCCCCCCC TVMXDCGIABBOFY-UHFFFAOYSA-N 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 238000012773 Laboratory assay Methods 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 230000005856 abnormality Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 229920006395 saturated elastomer Polymers 0.000 description 1
- 238000010025 steaming Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
- G01N21/35—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light
- G01N21/359—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light using near infrared light
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/25—Colour; Spectral properties, i.e. comparison of effect of material on the light at two or more different wavelengths or wavelength bands
- G01N21/31—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry
- G01N21/35—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light
- G01N21/3577—Investigating relative effect of material at wavelengths characteristic of specific elements or molecules, e.g. atomic absorption spectrometry using infrared light for analysing liquids, e.g. polluted water
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
Landscapes
- Physics & Mathematics (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Health & Medical Sciences (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Theoretical Computer Science (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention provides a method for eliminating an abnormal sample in a calibration set. Leave-one-out cross validation prediction is carried out on samples in the calibration set one by one by using a local modeling method, and the sample of which the predicted value exceeds the reproducibility is listed as a suspicious abnormal sample; and the abnormal sample is finally determined according to an analytical coordinate graph of main components and correlation analysis between properties. According to the method, the abnormal sample in the calibration set is relatively well eliminated, the adverse effects of the abnormal sample on other normal samples are eliminated, continuous enrichment and completion of the calibration set are facilitated, and the method has an important influence on improvement of the model prediction accuracy.
Description
Technical field
Present invention is mainly used for oil property field of fast detection, specially a kind of oil property based near infrared spectrum
The method that calibration set exceptional sample is rejected before detection.
Background technology
During gasoline property quick detection, the widely used modeling prediction techniques based near infrared spectrum of industry.Mesh
Before, there is the most frequently used modeling method of two classes, one is global modeling, and two is locally fine point.Though global modeling method has preferable
Versatility, but it is strict to calibration set sample distribution uniformity requirement, and the precision of prediction of model is not universal high;Locally fine point
Similar sample in method choice calibration set, predicts more accurate.It is local particularly in the case where computer speed is increasingly lifted
Modeling have developed rapidly in recent years.
However, it is possible that two class exceptional samples, a class is that laboratory values have conspicuousness with predicted value in calibration set
The calibration samples of difference, this is probably by laboratory values evaluated error is larger, spectral measurement error is larger or laboratory values typing mistake
Caused by etc. reason, this kind of sample either global modeling or locally fine point must be rejected before modeling;Another kind of is high bar
Bar value sample, compared with other samples in calibration set, containing extreme composition, away from the average value of model entirety sample, this kind of sample
This is obviously unhelpful to global modeling, because the uniformity of sample distribution is destroyed, but it is not only harmless to locally fine point, on the contrary favorably
In abundant calibration set, the precision of prediction of the follow-up similar sample to be tested of lifting.Therefore, accurately identify and rejecting abnormalities sample is to improving
Model prediction accuracy has great importance.
The content of the invention
In order to accurately identify and reject the exceptional sample in calibration set, improve the model inspection precision of oil property, this hair
It is bright to propose a kind of method for rejecting calibration set exceptional sample.The method will stay an intersection to test first with the method for locally fine point
Card predicted value is classified as suspicious exceptional sample beyond the sample of repeatability index, then using principal component analysis (PCA) coordinate diagram with
And the correlation analysis between property, finally determine exceptional sample;Wherein:Principal component analysis coordinate diagram is used to judge suspicious abnormal sample
Whether this is because modeling sample deficiency is caused:If suspicious exceptional sample is distributed in the edge of coordinate diagram, then it is assumed that be modeling sample
This deficiency is caused, and it temporarily is classified as into normal sample;If suspicious exceptional sample is distributed in coordinate diagram compact district, using between property
Correlation analysis determine whether it is exceptional sample.
The method specifically includes following steps:
(1) near infrared spectrum and property laboratory values of gasoline initial calibration collection sample are obtained;
(2) spectrum to calibration set sample carries out conventional pretreatment;
(3) using staying the cross-validation method to carry out offset minimum binary (PLS) modeling and forecasting one by one to sample in calibration set, point
The deviation between predicted value, and predicted value and laboratory values is not obtained;
(4) calibration samples of the prediction deviation beyond repeatability index are filtered out, suspicious exceptional sample is classified as;
(5) suspicious exceptional sample is filtered out from calibration set, remaining sample continues the calibration set modeled as next round;
(6) repeat step (3)~(5) are continued, until prediction deviation is all in corresponding repeatability indication range;
(7) suspicious exceptional sample is carried out into principal component analysis with residual correction collection sample one by one, draws PCA coordinate diagrams;
(8) PCA coordinate diagrams are observed, if suspicious exceptional sample is distributed in the edge of coordinate diagram, it is most likely that be modeling sample
What this deficiency was caused, the suspicious exceptional sample is considered as normal sample;If suspicious exceptional sample is distributed in coordinate diagram compact district,
Then exceptional sample is determined whether it is using the correlation analysis between property.
Related provision regulation is detected according to national oil property, for the repeatability index of gasoline property, organon octane
The repeatability for being worth (RON) is 0.6, and the repeatability of motor octane number is 0.8, and the repeatability of density (20 DEG C) is 5.0kg/m3,
The repeatability of saturated vapour pressure (RVP) is 5.0kPa, etc..
For specific gasoline, the related implementations between each property are generally:Density and RON, 50% evaporating temperature it
Between have good positive correlation, between density and RVP have negative correlation, this provides ginseng for the further analysis of exceptional sample
Examine foundation.
Beneficial effect:
The present invention proposes a kind of method for rejecting calibration set exceptional sample, first with the method for locally fine point, to school
The positive sample concentrated carries out staying a cross validation to predict one by one, and the sample by predicted value beyond repeatability index is classified as suspicious exception
Sample, then using the correlation analysis between principal component analysis coordinate diagram and property, finally determines exceptional sample.The method energy
The exceptional sample in calibration set is effectively rejected, model prediction accuracy is favorably improved.
Brief description of the drawings
The implementing procedure figure of Fig. 1 calibration set exceptional sample elimination methods
Fig. 2 numberings are 93#_26 and the spectrum comparison diagram that numbering is 93#_48 samples
Fig. 3 numberings are the PCA coordinate diagram of the suspicious exceptional sample with calibration set sample of 93#_132
Specific implementation process
The present invention is further illustrated with case study on implementation below in conjunction with the accompanying drawings.
The present invention introduces the elimination method of gasoline calibration set exceptional sample by taking certain 93# gasoline as an example.Present case is directed to vapour
The test of oily RON, selection in October, 2014 to 136 samples altogether of September in 2016, used as initial calibration collection, numbering was respectively
93#-1~93#-136.Using staying a cross validation predicted method to carry out locally fine point prediction one by one to the sample in calibration set, select
The modeling spectral coverage selected is 4000~4800cm-1, it is 3 using length, width and height ratio in three-dimensional principal component analysis coordinate diagram:2:1 length
Cube (is transverse axis 3 with first principal component, Second principal component, is the longitudinal axis 2, and the 3rd principal component is vertical pivot 1, drawing three-dimensional principal component point
Analysis figure) 50 ± 5 calibration samples are selected as similar Sample Establishing model.Initial calibration concentrates the modeling and forecasting result of sample such as
Shown in table 1.
The initial calibration of table 1 concentrates sample to stay a crossing prediction result
Because the repeatability of gasoline RON is 0.6, then sample of the prediction deviation absolute value more than 0.6 is filtered out, can by table 1
Know and have 9 prediction deviations of sample beyond repeatability index, respectively 93#_1,93#_26,93#_38,93#_47,93#_
55th, 93#_64,93#_84,93#_123,93#_132, are classified as suspicious exceptional sample, are continued to residual correction collection sample again
It is secondary to carry out staying a cross validation, 1 sample 93#_45 is filtered out again, put it into suspicious exceptional sample collection.Table 2 show suspicious
The sample predictions situation of exceptional sample collection.
The sample predictions situation of the suspicious exceptional sample collection of table 2
After filtering out suspicious exceptional sample, remaining 126 samples are normal sample in calibration set.Sat below with PCA
Mark on a map and each property between correlation, the suspicious exceptional sample to being given in table 2 is analyzed.With number be 93#_26 can
Doubt as a example by exceptional sample, PCA analyses are carried out to remaining 126 samples and 93#_26 samples in calibration set:
Firstly it is found that the sample that numbering is 93#_26 is in the sample distribution close quarters of PCA coordinate diagrams, in the absence of similar
The not enough problem of sample;The correlation between gasoline each property is next based on, finds it with the sample ratio that numbering is 93#_48
Compared with the two curve of spectrum is essentially coincided, as shown in Figure 2.And as shown in Table 1,93#_26 sample predictions value is 93.67, prediction
Deviation is 0.83;93#_48 sample predictions value is 93.81, and prediction deviation is -0.01, and prediction case is preferable.Table 3 is checked, according to
Positive correlation between 50% evaporating temperature and RON, 50% steaming of the 50% evaporating temperature value than 93#_48 sample of 93#_26 samples
Hair temperature value is small, then the RON values of 93#_26 samples ought to be smaller than the RON values of 93#_48 sample, actually but than 93#_48 sample
RON values it is bigger.It is therefore contemplated that the RON laboratory values of 93#_26 samples have overgauge, further laboratory assay is also confirmed that
Above-mentioned analysis, determines that it is exceptional sample.
The numbering of table 3 is respectively the sample properties table of 93#_26 and 93#_48
Using above-mentioned same procedure analyze, it is possible to determine that numbering be 93#_38,93#_45 and 93#_47 sample be it is different
Normal sample.
Again by taking the suspicious exceptional sample that numbering is 93#_132 as an example, to this 126 samples and the sample that numbering is 93#_132
Originally PCA analyses, and drawing three-dimensional coordinate diagram are carried out, as shown in Figure 3.In figure 3, numbering is the sample distribution of 93#_132 on side
Edge, and its prediction deviation is up to -1.92, far beyond 0.6 repeatability, it is most likely that there is a problem of that similar sample is not enough.Cause
This, is regarded as normal sample for the time being, waits supplying for subsequent samples.If however, combined using conventional principal component analysis
Mahalanobis distance method, the mahalanobis distance of the sample is then considered as exceptional sample outside prescribed limit, in causing sample set after rejecting
Sample it is fewer and feweri, extreme value range shorter is unfavorable for the perfect of Sample Storehouse, to the later stage modeling adversely affect.
It can be seen that, using method proposed by the present invention, calibration set exceptional sample can be effectively rejected, to set up oil property prediction
Model lays good basis.
Claims (5)
1. it is a kind of reject calibration set exceptional sample method, it is characterised in that the method first with locally fine point method, it is right
Sample in calibration set carries out staying a cross validation to predict one by one, and the sample by predicted value beyond repeatability is classified as suspicious abnormal sample
This, then using the correlation analysis between principal component analysis coordinate diagram and property, finally determines exceptional sample;Wherein:It is main into
Analysis coordinate diagram is used to judge whether suspicious exceptional sample is because modeling sample deficiency is caused:If suspicious exceptional sample distribution
At the edge of coordinate diagram, then it is assumed that be that modeling sample deficiency is caused, it temporarily is classified as into normal sample;If suspicious exceptional sample distribution
In coordinate diagram compact district, then exceptional sample is determined whether it is using the correlation analysis between property.
2. it is according to claim 1 it is a kind of reject calibration set exceptional sample method, it is characterised in that the method have with
Lower step:
(1) near infrared spectrum and property laboratory values of gasoline initial calibration collection sample are obtained;
(2) spectrum to calibration set sample carries out conventional pretreatment;
(3) using staying the cross-validation method to carry out PLS modeling and forecastings one by one to sample in calibration set, predicted value is respectively obtained, with
And the deviation between predicted value and laboratory values;
(4) calibration samples of the prediction deviation beyond repeatability index are filtered out, suspicious exceptional sample is classified as;
(5) suspicious exceptional sample is filtered out from calibration set, remaining sample continues the calibration set modeled as next round;
(6) repeat step (3)~(5) are continued, until prediction deviation is all in corresponding repeatability indication range;
(7) suspicious exceptional sample is carried out into principal component analysis with residual correction collection sample one by one, draws principal component analysis coordinate diagram;
(8) doubt exceptional sample and be considered as normal sample;If suspicious exceptional sample is distributed in coordinate diagram compact district, using property it
Between correlation analysis determine whether it is exceptional sample.
3. it is according to claim 2 it is a kind of reject calibration set exceptional sample method, it is characterised in that described in step (2)
Conventional preprocess method uses baseline correction and vector normalizing.
4. it is according to claim 2 it is a kind of reject calibration set exceptional sample method, it is characterised in that described in step (3)
PLS models are that, using the method for locally fine point, characteristic spectrum spectral coverage elects 4000~4800cm as-1, select 50 ± 5 calibration samples
As similar Sample Establishing model.
5. it is according to claim 2 it is a kind of reject calibration set exceptional sample method, it is characterised in that described in step (3)
It is 3 using length, width and height ratio in three-dimensional principal component analysis coordinate diagram that local regression method is:2:1 cuboid selects similar sample
This.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710017074.XA CN106706558B (en) | 2017-01-10 | 2017-01-10 | A method of rejecting calibration set exceptional sample |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710017074.XA CN106706558B (en) | 2017-01-10 | 2017-01-10 | A method of rejecting calibration set exceptional sample |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106706558A true CN106706558A (en) | 2017-05-24 |
CN106706558B CN106706558B (en) | 2019-03-22 |
Family
ID=58908178
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710017074.XA Active CN106706558B (en) | 2017-01-10 | 2017-01-10 | A method of rejecting calibration set exceptional sample |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106706558B (en) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107179293A (en) * | 2017-06-23 | 2017-09-19 | 南京富岛信息工程有限公司 | A kind of assessment method of oil property uncertainty |
CN107271400A (en) * | 2017-06-23 | 2017-10-20 | 南京富岛信息工程有限公司 | A kind of method of automatic addition calibration set sample |
CN107505282A (en) * | 2017-08-28 | 2017-12-22 | 南京富岛信息工程有限公司 | A kind of method for improving oil product near-infrared modeling robustness |
CN108267422A (en) * | 2017-12-29 | 2018-07-10 | 广州讯动网络科技有限公司 | Exceptional sample scalping method based on near-infrared spectrum analysis |
CN108416489A (en) * | 2018-01-10 | 2018-08-17 | 浙江中烟工业有限责任公司 | A kind of processing method and processing system of tobacco leaf product sensory evaluation data |
CN113536601A (en) * | 2021-08-17 | 2021-10-22 | 南京富岛信息工程有限公司 | Method for improving carbon content estimation precision of catalyst of continuous reforming device |
CN117093841A (en) * | 2023-10-18 | 2023-11-21 | 中国科学院合肥物质科学研究院 | Abnormal spectrum screening model determining method, device and medium for wheat transmission spectrum |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2006126100A (en) * | 2004-11-01 | 2006-05-18 | Kao Corp | Discrimination method using near-infrared spectrum |
CN104865222A (en) * | 2015-04-30 | 2015-08-26 | 北京林业大学 | Nondestructive testing method of content of fatty acid in peony seeds |
CN105928901A (en) * | 2016-07-11 | 2016-09-07 | 上海创和亿电子科技发展有限公司 | Qualitative and quantitative combined near infrared quantitative model construction method |
CN105954223A (en) * | 2016-04-28 | 2016-09-21 | 南京富岛信息工程有限公司 | Method for improving prediction accuracy of gasoline properties |
CN106018325A (en) * | 2016-04-29 | 2016-10-12 | 南京富岛信息工程有限公司 | Method for evaluating credibility of gasoline property modeling prediction result |
-
2017
- 2017-01-10 CN CN201710017074.XA patent/CN106706558B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2006126100A (en) * | 2004-11-01 | 2006-05-18 | Kao Corp | Discrimination method using near-infrared spectrum |
CN104865222A (en) * | 2015-04-30 | 2015-08-26 | 北京林业大学 | Nondestructive testing method of content of fatty acid in peony seeds |
CN105954223A (en) * | 2016-04-28 | 2016-09-21 | 南京富岛信息工程有限公司 | Method for improving prediction accuracy of gasoline properties |
CN106018325A (en) * | 2016-04-29 | 2016-10-12 | 南京富岛信息工程有限公司 | Method for evaluating credibility of gasoline property modeling prediction result |
CN105928901A (en) * | 2016-07-11 | 2016-09-07 | 上海创和亿电子科技发展有限公司 | Qualitative and quantitative combined near infrared quantitative model construction method |
Non-Patent Citations (4)
Title |
---|
MOHAMMADREZA KHANMOHAMMADI ET AL.: "Characterization of petroleum-based products by infrared spectroscopy and chemometrics", 《TRENDS IN ANALYTICAL CHEMISTRY》 * |
杨南林 等: "中药材三七中皂苷类成分的近红外光谱快速无损分析新方法", 《化学学报》 * |
王建义 等: "近红外光谱煤质分析模型中异常样品的剔除方法", 《工矿自动化》 * |
赵振英 等: "近红外光谱法分析油页岩含油率中异常样品识别和剔除方法的研究", 《光谱学与光谱分析》 * |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107179293A (en) * | 2017-06-23 | 2017-09-19 | 南京富岛信息工程有限公司 | A kind of assessment method of oil property uncertainty |
CN107271400A (en) * | 2017-06-23 | 2017-10-20 | 南京富岛信息工程有限公司 | A kind of method of automatic addition calibration set sample |
CN107505282A (en) * | 2017-08-28 | 2017-12-22 | 南京富岛信息工程有限公司 | A kind of method for improving oil product near-infrared modeling robustness |
CN108267422A (en) * | 2017-12-29 | 2018-07-10 | 广州讯动网络科技有限公司 | Exceptional sample scalping method based on near-infrared spectrum analysis |
CN108267422B (en) * | 2017-12-29 | 2021-01-12 | 广州讯动网络科技有限公司 | Abnormal sample removing method based on near infrared spectrum analysis |
CN108416489A (en) * | 2018-01-10 | 2018-08-17 | 浙江中烟工业有限责任公司 | A kind of processing method and processing system of tobacco leaf product sensory evaluation data |
CN113536601A (en) * | 2021-08-17 | 2021-10-22 | 南京富岛信息工程有限公司 | Method for improving carbon content estimation precision of catalyst of continuous reforming device |
CN113536601B (en) * | 2021-08-17 | 2023-09-01 | 南京富岛信息工程有限公司 | Method for improving estimation accuracy of carbon content of catalyst of continuous reforming device |
CN117093841A (en) * | 2023-10-18 | 2023-11-21 | 中国科学院合肥物质科学研究院 | Abnormal spectrum screening model determining method, device and medium for wheat transmission spectrum |
CN117093841B (en) * | 2023-10-18 | 2024-02-09 | 中国科学院合肥物质科学研究院 | Abnormal spectrum screening model determining method, device and medium for wheat transmission spectrum |
Also Published As
Publication number | Publication date |
---|---|
CN106706558B (en) | 2019-03-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106706558A (en) | Method for eliminating abnormal sample in calibration set | |
US11408875B2 (en) | Method for detecting and identifying toxic and harmful gases based on machine olfaction | |
US9401013B2 (en) | Method of design-based defect classification and system thereof | |
CN108829878B (en) | Method and device for detecting abnormal points of industrial experimental data | |
CN104713845A (en) | Mixture component identification method based on terahertz absorption spectrum processing | |
Nar et al. | The evaluation of analytical performance of immunoassay tests by using six-sigma method | |
CN111044638A (en) | Method for classifying and identifying different varieties of flue-cured tobacco leaves | |
CN115546211B (en) | Welding spot defect classification method, terminal and computer storage medium | |
CN106442122A (en) | Method for detecting ductile section percentage of fracture of steel material in drop weight tear test based on image segmentation and identification | |
CN107132311A (en) | A kind of fast gas recognizer extracted based on frequency domain character | |
CN113406249A (en) | Method for predicting adulterated oil types in tea oil | |
CN117370871B (en) | Quality analysis method and system for special steel | |
CN102565288A (en) | Method for analyzing component concentration of mixed volatile organic compounds (VOCs) | |
CN116842240B (en) | Data management and control system based on full-link management and control | |
CN106485049B (en) | A kind of detection method of the NIRS exceptional sample based on Monte Carlo cross validation | |
CN110793959A (en) | Metal brand identification method based on laser-induced breakdown spectroscopy | |
CN108920428B (en) | Fuzzy distance discrimination method based on joint fuzzy expansion principle | |
WO2020044165A1 (en) | Method and control system for detecting faults associated with gas chromotograph device in process plant | |
CN113740294B (en) | Near infrared modeling-based gasoline/diesel oil detection and analysis method and device | |
CN104181125A (en) | Method for rapidly determining Kol-bach value of beer malt | |
Xia et al. | A method for recognition of mixed gas composition based on PCA and KNN | |
CN113327204A (en) | Image calibration method and device, equipment and storage medium | |
CN107402191A (en) | A kind of oil product of gasoline quality determining method for evading disturbing factor | |
CN112257336A (en) | Mine water inrush source distinguishing method based on feature selection and support vector machine model | |
CN114756823B (en) | Method for improving prediction capability of pepper spectrum model |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |