CN103913432A - Near infrared spectrum wavelength selecting method based on particle swarm optimization - Google Patents
Near infrared spectrum wavelength selecting method based on particle swarm optimization Download PDFInfo
- Publication number
- CN103913432A CN103913432A CN201410114669.3A CN201410114669A CN103913432A CN 103913432 A CN103913432 A CN 103913432A CN 201410114669 A CN201410114669 A CN 201410114669A CN 103913432 A CN103913432 A CN 103913432A
- Authority
- CN
- China
- Prior art keywords
- particle
- infrared spectrum
- wavelength
- selection
- spectrum wavelength
- 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
Landscapes
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention relates to a near infrared spectrum wavelength selecting method based on particle swarm optimization, wherein near infrared spectrum signals of a sample are collected to form a site historical database, the database comprises multiple wavelength variables, the particle swarm optimization is combined with partial linear squares (PIS) to select all wavelengths of the near infrared spectrum, so that an established calibration model has stronger predictive ability, accurate detection and analysis on concentration of material compositions are realized, good theoretical basis is provided for application of the near infrared spectrum technology in all industrial fields, and important practical significance is realized.
Description
Technical field
The present invention relates to the application of near infrared spectrum in material composition quantitative test, be specifically related to a kind of near-infrared spectrum wavelength system of selection based on particle cluster algorithm.
Background technology
Near infrared spectrum (near infrared spectroscopy, NIR) derives from the absorption of molecular vibration to light, can be used for the principal ingredient analysis of material.Infrared irradiation on material time, is accompanied by the difference of material composition, and the light absorbance log of different wave length is also different, and in infrared spectrogram, the position of absorption peak changes with the composition of material, absorption peak height with content of material number change.
The spectral instrument of near-infrared spectral analysis technology utilization precision and Chemical Measurement software obtain the absorption spectrum of material near infrared spectrum district, then obtained near infrared spectrum data is carried out to analyzing and processing, the final qualitative or quantitative analysis results that obtains this material composition, its performance is accurate, clean, there are the not comparable high efficiency of classic method and stability, be used widely in industries such as food, pharmacy, heavy industry, petrochemical compleies at present, on-the-spot on-line analysis, has brought considerable economic benefit.
Near-infrared spectral analysis technology mainly depends on calibration model, according to the difference of the composition of surveyed material and character, adopts different modeling methods, and constantly calibration model is expanded and safeguard.Partial least square method (partial linear squares, PIS) be one of the main method of near infrared spectrum data modeling, it is to adopt full wave spectroscopic data to carry out modeling, so not only makes arithmetic speed slack-off, and can reduce the precision of prediction of model.Can make set up model there is stronger predictive ability and better robustness by specific method screening characteristic wavelength or wavelength region may.
At present, wavelength select method mainly contain correlation coefficient process, without information variable null method, genetic algorithm etc.Correlation coefficient process is that absorbance vector corresponding to each wavelength of proofreading and correct in light harvesting spectrum matrix carried out to correlation calculations with the concentration of component vector to be measured in concentration matrix, obtain related coefficient or the coefficient of determination of wavelength, its wavelength information that corresponding related coefficient absolute value (or coefficient of determination) is larger should be more.In conjunction with the given threshold value of chemical knowledge, the wavelength of selecting related coefficient to be greater than this threshold value is set up model.Correlation coefficient process utilizes linear statistical method to set up, and when for nonlinear dependence system or calibration samples collection skewness, utilizes the method to carry out the model prediction ability that wavelength selection sets up poor.Without information variable removing method (elimination of uninformative variable, UVE) be a kind of Wavelength selecting method of setting up based on PLS regression coefficient, the basic thought of the method is the measurement index using regression coefficient as wavelength importance.Although the method, in the time choosing wavelength variable, has been considered the impact of noise and concentration information simultaneously, relatively more directly perceived practical, its effect relies on the generation of random initial mask to a great extent.Genetic algorithm (genetic algorithm, GA) be to utilize organic sphere natural selection and genetic mechanism, by the operation of the operators such as selection, exchange, variation and sudden change, along with continuous genetic iteration, make target function value preferably wavelength variable be retained, poor eliminates, and finally realizes the result that wavelength is selected.But the local search ability of genetic algorithm is poor, easily produce " precocity " phenomenon, and in modeling process, initial calibration collection sample choose and the computation process of whole algorithm is all to have very strong randomness.
Summary of the invention
The deficiency existing in order to overcome above-mentioned Wavelength selecting method, the object of the present invention is to provide a kind of near-infrared spectrum wavelength system of selection based on particle cluster algorithm, thereby make set up calibration model there is stronger predictive ability, realize accurate detection and analysis to material composition concentration, for near-infrared spectral analysis technology provides better theoretical foundation in the application of each industrial circle, have important practical significance.
In order to achieve the above object, the technical solution adopted in the present invention is:
A near-infrared spectrum wavelength system of selection based on particle cluster algorithm, step is as follows:
Step 1: the first near infrared light spectrum signal of collecting sample, form on-the-spot historical data base D, the measure spectrum of database D is near infrared spectrum; Database D includes N wavelength variable;
Step 2: near-infrared spectrum wavelength system of selection is used Monte Carlo (Monte-Carlo, MC) method is training set and checking collection according to preset ratio R:1 by database D random division;
Step 3: near-infrared spectrum wavelength system of selection initialization training set, choose at random Num particle, each particle represents a data object, and each particle is a N dimensional vector, and Num is population size; The flying speed of this Num particle is carried out to random initializtion;
Step 4: near-infrared spectrum wavelength system of selection adopts binary coding to carry out position encoded to each particle; Each particle length equals whole wavelength N, the corresponding binary code of each wavelength, and wherein numerical value ' 1 ' represents that corresponding wavelength is selected, numerical value ' 0 ' represents that corresponding wavelength is not selected;
Step 5: near-infrared spectrum wavelength system of selection adopts partial least square method (partial linear squares, PIS) set up analysis correction model, and choose cross validation root-mean-square error RMSECV as fitness function, calculate the fitness value of each particle, and recording individual optimum solution p
iwith globally optimal solution p
g; The computing formula of cross validation root-mean-square error RMSECV is:
In formula, Kp is the sample number of cross validation collection, y
kbe the actual measurement concentration of k sample,
it is the prediction concentrations of k sample;
Step 6: the near-infrared spectrum wavelength system of selection flying speed of new particle more according to the following formula,
In formula, p
cfor reference state, p
ijfor individual optimal solution p
iit is position encoded that j ties up, p
gjfor globally optimal solution p
git is position encoded that j ties up, r
1and r
2be the random number between [0,1], f
cfor selecting coefficient, k is current iteration number of times, and ω is inertial factor, and a is acceleration factor,
for the flying speed of i particle j dimension of current iteration,
position encoded for i particle j of current iteration dimension,
for the flying speed of i particle j dimension of next iteration;
Step 7: near-infrared spectrum wavelength system of selection new particle position encoded more according to the following formula,
In formula, ρ is the random number between [0,1], and k is current iteration number of times,
for the flying speed of i particle j dimension of next iteration,
position encoded for i particle j of next iteration dimension;
Step 8: near-infrared spectrum wavelength system of selection repeating step 4～step 7, until reach maximum iteration time Iter_n or fitness function RMSECV and reach the fitness value fitness of setting; Output globally optimal solution;
Step 9: near-infrared spectrum wavelength system of selection, according to the binary coding of the globally optimal solution of output, can obtain selected wavelength variable, and numerical value ' 1 ' represents that corresponding wavelength is selected, and numerical value ' 0 ' represents that corresponding wavelength is not selected.
Described near-infrared spectrum wavelength system of selection adopts the spectral signal acquisition system of USB2000+ fiber spectrometer and computing machine composition, and Related Component is measured to collection in the spectral absorption of each frequency range.
Embodiment
Methane composition in the flue gas discharging below in conjunction with combustion of natural gas is that the present invention will be described in more detail for example.
The present invention is the near-infrared spectrum wavelength system of selection based on particle cluster algorithm, and step is as follows:
Step 1: the first near infrared light spectrum signal of collecting sample, form on-the-spot historical data base D, the measure spectrum of database D is near infrared spectrum.The wavelength coverage of measure spectrum is 549.44nm～4193.28nm, and database D includes N wavelength variable, and N value is 473.
Step 2: near-infrared spectrum wavelength system of selection is used Monte Carlo (Monte-Carlo, MC) method, is training set and checking collection according to preset ratio R:1 by database D random division, and R value is set as 4.
Step 3: near-infrared spectrum wavelength system of selection initialization training set, choose at random Num particle, each particle represents a data object, be that each particle is a N dimensional vector, Num is population size, Num value is set as 30, and the flying speed of this Num particle is carried out to random initializtion, and initialization velocity amplitude is provided at random by system.
Step 4: near-infrared spectrum wavelength system of selection adopts binary coding to carry out position encoded to each particle.Each particle length equals whole wavelength N,, N value is 473, the corresponding binary code of each wavelength, wherein numerical value ' 1 ' represents that corresponding wavelength is selected, numerical value ' 0 ' represents that corresponding wavelength is not selected.
Step 5: near-infrared spectrum wavelength system of selection adopts partial least square method (partial linear squares, PIS) set up analysis correction model, and choose cross validation root-mean-square error RMSECV as fitness function, calculate the fitness value of each particle, and recording individual optimum solution p
iwith globally optimal solution p
g.The computing formula of cross validation root-mean-square error RMSECV is:
In formula, Kp is the sample number of cross validation collection, and Kp value is 151, y
kbe the actual measurement concentration of k sample,
it is the prediction concentrations of k sample.
Step 6: the near-infrared spectrum wavelength system of selection flying speed of new particle more according to the following formula,
In formula, p
cfor reference state, p
ijfor individual optimal solution p
iit is position encoded that j ties up, p
gjfor globally optimal solution p
git is position encoded that j ties up, r
1and r
2be the random number between [0,1], r
1and r
2value is provided at random by system, f
cfor selecting coefficient, f
cvalue is that 0.5, k is current iteration number of times, and ω is inertial factor, and ω value is that 0.7, a is acceleration factor, and a value is 2,
for the flying speed of i particle j dimension of current iteration,
position encoded for i particle j of current iteration dimension,
for the flying speed of i particle j dimension of next iteration.
Step 7: near-infrared spectrum wavelength system of selection new particle position encoded more according to the following formula,
In formula, ρ is the random number between [0,1], is provided at random by system, and k is current iteration number of times,
for the flying speed of i particle j dimension of next iteration,
position encoded for i particle j of next iteration dimension.
Step 8: near-infrared spectrum wavelength system of selection repeating step 4～step 7, until reach maximum iteration time Iter_n or fitness function RMSECV and reach the fitness value fitness of setting, Iter_n value is set as 100, fitness value and is set as 10
-6, output globally optimal solution.
Step 9: the binary coding of near-infrared spectrum wavelength system of selection output globally optimal solution: 00,000,000,000,001,000,100,000,100,000,000,001,000,100,000,010,000,000,000 00,000,100,000,001,100,000,000,000,100,010,100,010,000,001,100,000,000,000 00,000,000,101,000,000,000,000,010,000,000,000,010,000,110,000,000,000,010 00,101,100,000,010,000,000,000,001,100,000,000,100,000,000,000,010,001,001 10,100,010,000,001,000,000,001,000,100,000,010,000,000,000,000,000,000,010 10,001,010,000,000,000,000,000,010,000,001,100,000,000,000,110,110,000,010 11,001,000,000,001,100,011,000,000,100,000,101,001,010,000,100,001,000,001 11,000,001,000,101,000,001,011,000,001,001,000,100,000,001,100,000,100,000 0000000011000110000000000 numerical value ' 1 ' represent that corresponding wavelength is selected, numerical value ' 0 ' represents that corresponding wavelength is not selected.The wavelength of choosing according to result is respectively:
14,18,24,35,39,46,62,70,71,83,87,89,93,100,101,121,123，
137,149,154,155,167,171,173,174，
181,194,195,204,217,221,224,225,227，
231,238,247,251,258,279,281,285,287,305,312,313,325,326,328,329,
335,337,338,341,350,351,355,356,363，
369,371,374,376,381,386,392,393,394,400,404,406,412,414,415,421,
424,428,436,437,443,457,458,462,463 wavelength variablees.
Described near-infrared spectrum wavelength system of selection adopts the spectral signal acquisition system of USB2000+ fiber spectrometer and computing machine composition, and Related Component is measured to collection in the spectral absorption of each frequency range.
Claims (2)
1. the near-infrared spectrum wavelength system of selection based on particle cluster algorithm, is characterized in that: step is as follows:
Step 1: the first near infrared light spectrum signal of collecting sample, form on-the-spot historical data base D, the measure spectrum of database D is near infrared spectrum; Database D includes N wavelength variable;
Step 2: near-infrared spectrum wavelength system of selection is used Monte Carlo Monte-Carlo, MC method, is training set and checking collection according to preset ratio R:1 by database D random division;
Step 3: near-infrared spectrum wavelength system of selection initialization training set, choose at random Num particle, each particle represents a data object, and each particle is a N dimensional vector, and Num is population size; The flying speed of this Num particle is carried out to random initializtion;
Step 4: near-infrared spectrum wavelength system of selection adopts binary coding to carry out position encoded to each particle; Each particle length equals whole wavelength N, the corresponding binary code of each wavelength, and wherein numerical value ' 1 ' represents that corresponding wavelength is selected, numerical value ' 0 ' represents that corresponding wavelength is not selected;
Step 5: near-infrared spectrum wavelength system of selection adopts partial least square method partial linear squares, PIS sets up analysis correction model, and choose cross validation root-mean-square error RMSECV as fitness function, calculate the fitness value of each particle, and recording individual optimum solution p
iwith globally optimal solution p
g; The computing formula of cross validation root-mean-square error RMSECV is:
In formula, Kp is the sample number of cross validation collection, y
kbe the actual measurement concentration of k sample,
it is the prediction concentrations of k sample;
Step 6: the near-infrared spectrum wavelength system of selection flying speed of new particle more according to the following formula,
In formula, p
cfor reference state, p
ijfor individual optimal solution p
iit is position encoded that j ties up, p
gjfor globally optimal solution p
git is position encoded that j ties up, r
1and r
2be the random number between [0,1], f
cfor selecting coefficient, k is current iteration number of times, and ω is inertial factor, and a is acceleration factor,
for the flying speed of i particle j dimension of current iteration,
position encoded for i particle j of current iteration dimension,
for the flying speed of i particle j dimension of next iteration;
Step 7: near-infrared spectrum wavelength system of selection new particle position encoded more according to the following formula,
In formula, ρ is the random number between [0,1], and k is current iteration number of times,
for the flying speed of i particle j dimension of next iteration,
position encoded for i particle j of next iteration dimension;
Step 8: near-infrared spectrum wavelength system of selection repeating step 4～step 7, until reach maximum iteration time Iter_n or fitness function RMSECV and reach the fitness value fitness of setting; Output globally optimal solution;
Step 9: near-infrared spectrum wavelength system of selection, according to the binary coding of the globally optimal solution of output, obtains selected wavelength variable, and numerical value ' 1 ' represents that corresponding wavelength is selected, and numerical value ' 0 ' represents that corresponding wavelength is not selected.
2. a kind of near-infrared spectrum wavelength system of selection based on particle cluster algorithm according to claim 1, it is characterized in that: described near-infrared spectrum wavelength system of selection adopts the spectral signal acquisition system of USB2000+ fiber spectrometer and computing machine composition, and Related Component is measured to collection in the spectral absorption of each frequency range.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410114669.3A CN103913432B (en) | 2014-03-25 | 2014-03-25 | Based on the near-infrared spectrum wavelength system of selection of particle cluster algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410114669.3A CN103913432B (en) | 2014-03-25 | 2014-03-25 | Based on the near-infrared spectrum wavelength system of selection of particle cluster algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103913432A true CN103913432A (en) | 2014-07-09 |
| CN103913432B CN103913432B (en) | 2016-01-20 |
Family
ID=51039277
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410114669.3A Expired - Fee Related CN103913432B (en) | 2014-03-25 | 2014-03-25 | Based on the near-infrared spectrum wavelength system of selection of particle cluster algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103913432B (en) |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104155262A (en) * | 2014-08-20 | 2014-11-19 | 浙江中烟工业有限责任公司 | Method for selecting spectrum scope in tobacco water-soluble sugar near infrared quantification model |
| CN104502305A (en) * | 2014-12-09 | 2015-04-08 | 西北师范大学 | Near infrared spectrum useful information distinguishing method based on wavelet transform |
| CN105738317A (en) * | 2016-02-16 | 2016-07-06 | 广州纤维产品检测研究院 | Textile near-infrared model transfer method |
| CN108388965A (en) * | 2018-03-01 | 2018-08-10 | 武汉轻工大学 | Grease mixes pseudo- detection method of content, terminal device and computer readable storage medium |
| CN109255464A (en) * | 2018-07-17 | 2019-01-22 | 广东工业大学 | A kind of Multilayer Microwave Absorption Materials performance optimization method based on particle swarm algorithm |
| CN109507118A (en) * | 2018-11-14 | 2019-03-22 | 江南大学 | The detection method of dry green soy bean moisture content |
| CN110503156A (en) * | 2019-08-27 | 2019-11-26 | 黑龙江八一农垦大学 | A kind of multivariate calibration characteristic wavelength selection method based on least correlativing coefficient |
| CN111812041A (en) * | 2020-06-29 | 2020-10-23 | 重庆邮电大学 | Portable water body COD (chemical oxygen demand) measuring system and method |
| CN113222959A (en) * | 2021-05-26 | 2021-08-06 | 马翔 | Fresh jujube wormhole detection method based on hyperspectral image convolutional neural network |
| CN114018864A (en) * | 2021-11-10 | 2022-02-08 | 黑龙江八一农垦大学 | Method for rapidly detecting content change of alcohol-soluble protein in corn kernels in grouting period |
| CN114166793A (en) * | 2021-11-03 | 2022-03-11 | 杭州电子科技大学 | Leaf chlorophyll a and b content inversion method based on spectral band overlapping separation |
| CN114584252A (en) * | 2022-02-16 | 2022-06-03 | 暨南大学 | Micro-ring resonance wavelength searching method combined with particle swarm optimization |
| CN115114838A (en) * | 2022-07-22 | 2022-09-27 | 黑龙江八一农垦大学 | Spectral characteristic wavelength selection method based on particle swarm algorithm thought and simulated annealing strategy |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050196047A1 (en) * | 2004-02-03 | 2005-09-08 | Yuri Owechko | Object recognition system incorporating swarming domain classifiers |
| CN101975575A (en) * | 2010-10-15 | 2011-02-16 | 西安电子科技大学 | Multi-target tracking method for passive sensor based on particle filtering |
-
2014
- 2014-03-25 CN CN201410114669.3A patent/CN103913432B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050196047A1 (en) * | 2004-02-03 | 2005-09-08 | Yuri Owechko | Object recognition system incorporating swarming domain classifiers |
| CN101975575A (en) * | 2010-10-15 | 2011-02-16 | 西安电子科技大学 | Multi-target tracking method for passive sensor based on particle filtering |
Non-Patent Citations (4)
| Title |
|---|
| R. RAGHAVENDRA 等: "Particle swarm optimization based fusion of near infrared and visible images for improved face verification", 《PATTERN RECOGNITION》 * |
| 周延 等: "RMSECV曲线筛选光谱波段算法", 《光谱学与光谱分析》 * |
| 寇蔚 等: "多粒子群优化算法和RBF神经网络在缺陷参数红外智能识别中的应用", 《数据采集与处理》 * |
| 曹晖 等: "融合波长选择和异常光谱检测的天然气燃烧过程定量分析方法", 《光谱学与光谱分析》 * |
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104155262B (en) * | 2014-08-20 | 2017-01-11 | 浙江中烟工业有限责任公司 | Method for selecting spectrum scope in tobacco water-soluble sugar near infrared quantification model |
| CN104155262A (en) * | 2014-08-20 | 2014-11-19 | 浙江中烟工业有限责任公司 | Method for selecting spectrum scope in tobacco water-soluble sugar near infrared quantification model |
| CN104502305A (en) * | 2014-12-09 | 2015-04-08 | 西北师范大学 | Near infrared spectrum useful information distinguishing method based on wavelet transform |
| CN104502305B (en) * | 2014-12-09 | 2017-02-22 | 西北师范大学 | Near infrared spectrum useful information distinguishing method based on wavelet transform |
| CN105738317A (en) * | 2016-02-16 | 2016-07-06 | 广州纤维产品检测研究院 | Textile near-infrared model transfer method |
| CN105738317B (en) * | 2016-02-16 | 2018-06-01 | 广州纤维产品检测研究院 | Textile method for transferring near infrared model |
| CN108388965A (en) * | 2018-03-01 | 2018-08-10 | 武汉轻工大学 | Grease mixes pseudo- detection method of content, terminal device and computer readable storage medium |
| CN109255464A (en) * | 2018-07-17 | 2019-01-22 | 广东工业大学 | A kind of Multilayer Microwave Absorption Materials performance optimization method based on particle swarm algorithm |
| CN109507118A (en) * | 2018-11-14 | 2019-03-22 | 江南大学 | The detection method of dry green soy bean moisture content |
| CN109507118B (en) * | 2018-11-14 | 2020-11-24 | 江南大学 | Method for detecting moisture content of dried green soy beans |
| CN110503156B (en) * | 2019-08-27 | 2021-09-03 | 黑龙江八一农垦大学 | Multivariate correction characteristic wavelength selection method based on minimum correlation coefficient |
| CN110503156A (en) * | 2019-08-27 | 2019-11-26 | 黑龙江八一农垦大学 | A kind of multivariate calibration characteristic wavelength selection method based on least correlativing coefficient |
| CN111812041A (en) * | 2020-06-29 | 2020-10-23 | 重庆邮电大学 | Portable water body COD (chemical oxygen demand) measuring system and method |
| CN113222959A (en) * | 2021-05-26 | 2021-08-06 | 马翔 | Fresh jujube wormhole detection method based on hyperspectral image convolutional neural network |
| CN113222959B (en) * | 2021-05-26 | 2022-04-15 | 马翔 | Fresh jujube wormhole detection method based on hyperspectral image convolutional neural network |
| CN114166793A (en) * | 2021-11-03 | 2022-03-11 | 杭州电子科技大学 | Leaf chlorophyll a and b content inversion method based on spectral band overlapping separation |
| CN114166793B (en) * | 2021-11-03 | 2023-08-04 | 杭州电子科技大学 | Leaf chlorophyll a and b content inversion method based on spectrum band overlapping separation |
| CN114018864A (en) * | 2021-11-10 | 2022-02-08 | 黑龙江八一农垦大学 | Method for rapidly detecting content change of alcohol-soluble protein in corn kernels in grouting period |
| CN114018864B (en) * | 2021-11-10 | 2022-09-16 | 黑龙江八一农垦大学 | Method for rapidly detecting content change of alcohol-soluble protein in corn kernels in grouting period |
| CN114584252A (en) * | 2022-02-16 | 2022-06-03 | 暨南大学 | Micro-ring resonance wavelength searching method combined with particle swarm optimization |
| CN114584252B (en) * | 2022-02-16 | 2023-09-19 | 暨南大学 | Micro-ring resonance wavelength searching method combined with particle swarm algorithm |
| CN115114838A (en) * | 2022-07-22 | 2022-09-27 | 黑龙江八一农垦大学 | Spectral characteristic wavelength selection method based on particle swarm algorithm thought and simulated annealing strategy |
| CN115114838B (en) * | 2022-07-22 | 2023-02-07 | 黑龙江八一农垦大学 | Spectral characteristic wavelength selection method based on particle swarm algorithm thought and simulated annealing strategy |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103913432B (en) | 2016-01-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103913432B (en) | Based on the near-infrared spectrum wavelength system of selection of particle cluster algorithm | |
| CN104020127B (en) | A kind of near infrared spectrum is utilized quickly to measure the method for inorganic elements in Nicotiana tabacum L. | |
| CN103528990B (en) | A kind of multi-model Modeling Method of near infrared spectrum | |
| CN102750333B (en) | Method for extracting semiconductor nano-structure feature size | |
| CN101216426A (en) | Gas status quantitative analyzer based on extended kalman filter theory | |
| CN107632010B (en) | Method for quantifying steel sample by combining laser-induced breakdown spectroscopy | |
| CN105388123A (en) | Method for predicting crude oil characteristic through near infrared spectrum | |
| CN101949826B (en) | Positive model and inverse model-based quantitative spectrometric analysis and calibration method of multi-component gas | |
| CN103884670B (en) | Based on the smoke components quantitative analysis method of near infrared spectrum | |
| CN103487411A (en) | Method for recognizing steel grade by combining random forest algorithm with laser-induced breakdown spectroscopy | |
| CN114021436A (en) | Near-surface ozone inversion method based on near-surface ultraviolet radiation | |
| CN109521437B (en) | Multispectral laser radar wavelength selection method for vegetation biochemical parameter detection | |
| CN114460013B (en) | Coastal wetland vegetation overground biomass GAN model self-learning remote sensing inversion method | |
| CN104535528A (en) | Method for real time extraction of TDLAS gas absorption spectrum absorbance by BP neural network | |
| CN106990056A (en) | A kind of total soil nitrogen spectrum appraising model calibration samples collection construction method | |
| CN102680425B (en) | Multiple analysis model information fusion method for multicomponent gas Fourier transform spectral analysis | |
| Yu et al. | Prediction of soil properties based on characteristic wavelengths with optimal spectral resolution by using Vis-NIR spectroscopy | |
| CN105608296B (en) | A kind of blade potassium concn inversion method based on lichee canopy spectra | |
| Ju et al. | Rapid Identification of Atmospheric Gaseous Pollutants Using Fourier‐Transform Infrared Spectroscopy Combined with Independent Component Analysis | |
| Chen et al. | Hyperspectral detection of sugar content for sugar-sweetened apples based on sample grouping and SPA feature selecting methods | |
| CN105241823B (en) | Coal steam-electric plant smoke quantitative analysis method of spectrum based on rarefaction representation | |
| CN106529680B (en) | A kind of multiple dimensioned extreme learning machine integrated modelling approach based on empirical mode decomposition | |
| CN117057464A (en) | Soil organic matter spectrum prediction method and device based on nonlinear memory learning | |
| Wan et al. | A stacking-based ensemble learning method for available nitrogen soil prediction with a handheld micronear-infrared spectrometer | |
| de los Ángeles Sepúlveda et al. | Near-infrared spectroscopy: Alternative method for assessment of stable carbon isotopes in various soil profiles in Chile |
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: 20160120 Termination date: 20190325 |