CN106290263B - A kind of LIBS calibration and quantitative analysis methods based on genetic algorithm - Google Patents
A kind of LIBS calibration and quantitative analysis methods based on genetic algorithm Download PDFInfo
- Publication number
- CN106290263B CN106290263B CN201510259910.6A CN201510259910A CN106290263B CN 106290263 B CN106290263 B CN 106290263B CN 201510259910 A CN201510259910 A CN 201510259910A CN 106290263 B CN106290263 B CN 106290263B
- Authority
- CN
- China
- Prior art keywords
- spectral line
- libs
- quantitative analysis
- line
- genetic algorithm
- 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.)
- Active
Links
Landscapes
- Investigating, Analyzing Materials By Fluorescence Or Luminescence (AREA)
Abstract
The LIBS calibration and quantitative analysis methods based on genetic algorithm that the present invention relates to a kind of, the specific steps are：1) LIBS spectroscopic datas are obtained；2) elemental characteristic spectrum to be measured is obtained；3) parameter coding forms genetic algorithm initial population；4) each individual adaptation degree in population is calculated；5) selection, intersection and mutation probability are pressed and forms new population；6) repeat 4), 5) to termination condition is met, export optimal spectral line (spectral line to) position；7) (internal calibration) quantitative analysis is calibrated according to optimal spectral line (spectral line to).The optimal spectral line (spectral line to) obtained using this method may be implemented as analytical line (analytical line and reference line) to the accurate quantitative analysis of concentration of element to be measured.It the advantage is that and be not necessarily to artificial selection analytical line (reference line), can accurately find high coefficient of determination (R2), the element spectral line (spectral line to) of low detection limits (LOD) and low relative standard deviation (RSD) are used as analytical line (analytical line and reference line).
Description
Technical field
It is specifically that one kind is calculated based on heredity the invention belongs to spectrum analysis and material component analysis field
The LIBS calibration and quantitative analysis methods of method.
Background technology
Laser induced breakdown spectroscopy (LIBS) analytical technology is a kind of emission spectrum using pulse laser as energy source
The qualitative and quantitative analysis of substance chemical element may be implemented in analytical technology.It has without sample preparation, directly quick, sample damages
The features such as vector is small becomes the research hotspot in the fields such as metallurgical analysis, historical relic's protection, geologic chemistry, environmental project in recent years.
Calibration (internal calibration) tracing analysis method is that it is strong to draw spectral line by the LIBS measurements to known concentration standard sample
Degree --- the relation curve (calibration curve) of concentration of element directly passes through relationship song after measuring analysis sample spectrum intensity
Line obtains the quantitative analysis method of the concentration of element.
As most basic quantitative analysis method, either basic calibration analysis method or base based on single the intensity of spectral line
In the internal calibration analytic approach of analytical line and reference line pair, what is relied primarily on is the selection of analytical line (reference line).Selection is clear
Accurately, it is the key that scaling method to interfere small spectral line.Traditional analytical line (reference line) be by analysis personnel by observe spectral line,
Selection is carried out in conjunction with spectra database and experience.With the increase of LIBS measurement data amounts, this artificial selection analysis
The method inefficiency of line (reference line) can not find the spectral line of global optimum substantially, and resulting calibration curve is to measuring
Sample carries out quantitative analysis and is extremely difficult to effect that is good and stablizing.
Invention content
Place aiming at the above shortcomings existing in the prior art, the technical problem to be solved in the present invention is to provide one kind can
Automatic optimal spectral line search, and using obtained optimal spectral line realization calibration curve method concentration of element quantitative analysis based on heredity
The LIBS calibration and quantitative analysis methods of algorithm.
Present invention technical solution used for the above purpose is：A kind of LIBS calibrations based on genetic algorithm are quantitative
Analysis method includes the following steps：
Step 1：The LIBS data for obtaining standard sample, determine wave-length coverage；
Step 2：It is loaded into characteristic spectral line database, it is all special in the wave-length coverage that step 1 determines to read element to be measured
Spectral position information is levied, the peak-seeking near the characteristic spectral line position of characteristic spectral line database corresponding element determines the LIBS measured
Corresponding characteristic spectral line specific location in data；
Step 3：Number of encoding bits are determined according to LIBS data lengths, selection spectral line quantity, form genetic algorithm initial population；
Step 4：With coefficient of determination R2, detection limit LOD and relative standard deviation RSD weighted sum as fitness function, seek
Look for the corresponding individual of optimal spectral line in population；
Step 5：Initial population is selected, is intersected and mutation operation, obtained new individual is inserted into original seed group's shape again
At new population；
Step 6：Step 4, step 5 are repeated, until genetic algorithm meets termination condition, terminates entire algorithmic procedure, is exported
Finally obtained optimal position of spectral line；
Step 7：Calibration and quantitative analysis is carried out to concentration of element to be measured using optimal spectral line.
The characteristic spectral line database is Given information, is measured by high purity substance or atomic emission spectrum database obtains
.
Further include：It is attached in the corresponding position of sample to be tested LIBS data to the characteristic spectrum location information obtained in step 2
Closely do peak-seeking processing.
It is described that number of encoding bits are determined according to LIBS data lengths, selection spectral line quantity, specially：For being in length
N spectral line is selected in the data of length, number of encoding bits are by determiningIt determines.
The fitness function is：
Wherein R2For coefficient of determination,For coefficient of determination penalty threshold, LOD is detection limit, LOD0Threshold is punished for detection limit
Value, RSD is relative standard deviation, RSD0For relative standard deviation's penalty threshold, a, b, c be respectively the corresponding punishment of three parameters because
The index of son.
The step 5 is formed according to LIBS data length selected populations size, crossover probability, mutation probability, select probability
New population.
The crossover probability between 0.7~0.9, mutation probability between 0.05~0.15, select probability between
Between 0.7~0.9.
The termination condition is：Adaptive optimal control degree individual does not change or reaches the maximum of setting and evolves after several generations evolve
Algebraically.
The present invention has the following advantages and beneficial effects：
1. it is not required to by observing LIBS measure spectrum artificial selection analytical lines (reference line), it only need to be rational suitable by being arranged
Response function can be obtained corresponding optimal spectral line, and the spectral line obtained in this way has more Global Optimality, passes through determining for its foundation
Mark song line can more accurately describe its relationship between concentration of element.
2. by changing genetic algorithm chromosome coding structure and length, fitness function is adjusted accordingly, it can be flexible
The different elements of selection, different evaluation index a plurality of spectral line, you can be used for the quantitative analysis method of unitary variant, and can be more
The quantitative analysis method of variable provides primary data.
3. the method applied in the present invention is to apply genetic algorithm, in conjunction with characteristic spectral line information in spectra database, setting
Rational fitness function achievees the purpose that search for the optimal spectral line of element to be measured in spectral region automatically by Evolution of Population,
The final optimal spectral line provided in LIBS measurement ranges, it is dense to element to establish calibration analysis curve as analytical line (reference line)
Degree carries out quantitative analysis.
Description of the drawings
Fig. 1 is the method for the present invention implementation flow chart；
Fig. 2 is the optimal spectral line of low-alloy steel sample to calibration curve quantitative analysis result the method for the present invention flow chart.
Specific implementation mode
The present invention is described in further detail with reference to the accompanying drawings and embodiments.
The present invention is directed to different elemental characteristic position of spectral line differences, the same element characteristic spectrum of various concentration in LIBS signals
The different feature of line intensity value finds optimal characteristics position of spectral line, corresponding element concentration is determined according to its intensity of spectral line.
As shown in Figure 1, method reads the experimental spectrum data obtained by LIBS experiment porch as input after starting, find
The corresponding each spectral line of element to be measured selects optimal position of spectral line by genetic algorithm optimization, and passes through it to establish calibration bent
Line analyzes concentration of element to be measured.Steps are as follows for specific implementation：
Step 1：The LIBS data for obtaining sample, determine wave-length coverage.The sample being directed to is standard sample, to be measured
Member is known as actual concentrations；LIBS data are obtained by testing to measure.
Step 2：It is loaded into characteristic spectral line database, it is all special in the wave-length coverage that step 1 determines to read element to be measured
Levy spectral position information.Characteristic spectral line database is measured by high purity substance or atomic emission spectrum database obtains, and is standard
Characteristic spectral line database is suitable for all samples.The peak-seeking near the characteristic spectral line position of characteristic spectral line database corresponding element,
Determine corresponding characteristic spectral line specific location in the LIBS data measured.
Due to experimental situation and parameter, the difference of operation, the spectrum of the LIBS data character pair position of spectral line measured
There is offset in summit, operated by the least displacement peak-seeking of the feature spectral position environs in java standard library, find specific LIBS
Characteristic spectral line position in data.
Step 3：Number of encoding bits are determined according to the spectral line quantity finally obtained, form genetic algorithm initial population.
Step 4：With coefficient of determination (R2), the weighted sum of detection limit (LOD) and relative standard deviation (RSD) is as fitness letter
Number, finds the corresponding individual of optimal spectral line.
By adding, deleting the parameter for participating in evaluating and adjusting parameter weights, the emphasis of optimal value concern can be adjusted, is obtained
To the fitness function for more adapting to sample to be tested characteristic and experimental situation.
Step 5：(duplication), intersection and mutation operation are selected to initial population, original seed group is inserted into again and forms new population.
Forming one by the high fitness individual for retaining the high fitness individual of parent population and being inserted into progeny population has higher
Fitness and different from parent population new population.According to LIBS data length selected populations size (generally 100), intersect generally
Rate (0.7~0.9), mutation probability (0.05~0.15), select probability (0.7~0.9) form new population.
Step 6：Step 4, step 5 are repeated, until genetic algorithm meets termination condition, terminates entire algorithmic procedure, is exported
Finally obtained optimal spectral line (spectral line to) position.
Algorithm termination condition is traditionally arranged to be the maximum that N does not change or reach setting for adaptive optimal control degree individual after evolving
Evolutionary generation, using the optimal value ensured as far as possible as global optimum.
Step 8：(internal calibration) quantitative analysis is calibrated to concentration of element to be measured using optimal spectral line (spectral line to).
Entire method, can also be by changing code length and knot other than calibration curve quantitative analysis is realized in route selection
Structure adjusts fitness function, selects a plurality of optimal spectral line, primary data is provided for other multivariate quantitative analysis methods.
Analyze the dense of tetra- kinds of elements of Cr, Ni, Mn, Si in 10 pieces of low-alloy steel standard samples respectively by approach described above
Degree, using wherein 8 pieces of samples as training sample, remaining 2 pieces of samples are as verification sample, the final effect of test method.
It is as follows to set fitness function：
Wherein R2For coefficient of determination, LOD is detection limit, and RSD is relative standard deviation, and a, b, c are respectively three parameters pair
The index for the penalty factor answered, setting is as 0.95 ＜ R2When≤1, a=0；R2When≤0.95, a=1.Equally, 0 LOD≤1000 ＜
When, b=0；When LOD ＞ 1000, b=1.And when 0≤RSD≤0.1, c=0；When RSD ＞ 0.1, c=1.I.e. when individual judgement system
When number, detection limit or relative standard deviation exceed ideal range, corresponding fitness function part one is given greatly
Penalty factor, so that it is eliminated in the next generation evolves, ensure in new population that contain is all coefficient of determination, detection limit and phase
The individual that standard deviation is met the requirements.
According to the corresponding Analysis of Genetic Algorithms of the above fitness function, the corresponding analytical line of four kinds of elements and its corresponding is obtained
Reference line (Fe characteristic spectral lines) it is as shown in the table.
According to obtained analytical line, reference line pair, establishes internal calibration curve and four kinds of concentration of element in sample are determined
Amount analysis, the results are shown in Figure 2, then respectively with formulaAnd formula
Calculate the root-mean-square error and verification collection root-mean-square error (C in formula of each elemental analysis resultiWithIn respectively sample i
Concentration of element actual value and measured value, t and v are respectively training set and verification collection sample number), it finally obtains and is related to using this patent
The LIBS internal calibrations quantitative analysis results such as following table based on genetic algorithm arrived, it is seen that this method can effectively reach quantitative analysis sample
Condition closes the requirement of concentration of element.
Claims (7)
1. a kind of LIBS calibration and quantitative analysis methods based on genetic algorithm, which is characterized in that include the following steps：
Step 1：The LIBS data for obtaining standard sample, determine wave-length coverage；
Step 2：It is loaded into characteristic spectral line database, reads whole characteristic lights of the element to be measured in the wave-length coverage that step 1 determines
Spectral position information, the peak-seeking near the characteristic spectral line position of characteristic spectral line database corresponding element determine the LIBS data measured
In corresponding characteristic spectral line specific location；
Step 3：Number of encoding bits are determined according to LIBS data lengths, selection spectral line quantity, form genetic algorithm initial population；
Step 4：With coefficient of determination R2, detection limit LOD and relative standard deviation RSD weighted sum as fitness function, find population
The corresponding individual of interior optimal spectral line；The fitness function is：
Wherein R2For coefficient of determination,For coefficient of determination penalty threshold, LOD is detection limit, LOD0For detection limit penalty threshold, RSD
For relative standard deviation, RSD0For relative standard deviation's penalty threshold, a, b, c are respectively the finger of the corresponding penalty factor of three parameters
Number；
Step 5：Initial population is selected, is intersected and mutation operation, obtained new individual is inserted into original seed group again and is formed newly
Population；
Step 6：Step 4, step 5 are repeated, until genetic algorithm meets termination condition, terminates entire algorithmic procedure, output is final
Obtained optimal position of spectral line；
Step 7：Calibration and quantitative analysis is carried out to concentration of element to be measured using optimal spectral line.
2. a kind of LIBS calibration and quantitative analysis methods based on genetic algorithm according to claim 1, which is characterized in that institute
It is Given information to state characteristic spectral line database, is measured by high purity substance or atomic emission spectrum database obtains.
3. a kind of LIBS calibration and quantitative analysis methods based on genetic algorithm according to claim 1, which is characterized in that also
Including：To the characteristic spectrum location information obtained in step 2, done at peak-seeking near the corresponding position of sample to be tested LIBS data
Reason.
4. a kind of LIBS calibration and quantitative analysis methods based on genetic algorithm according to claim 1, which is characterized in that institute
It states and number of encoding bits is determined according to LIBS data lengths, selection spectral line quantity, specially：For in the data that length is length
N spectral line is selected, number of encoding bits are by determiningIt determines.
5. a kind of LIBS calibration and quantitative analysis methods based on genetic algorithm according to claim 1, which is characterized in that institute
Step 5 is stated according to LIBS data length selected populations size, crossover probability, mutation probability, select probability, forms new population.
6. a kind of LIBS calibration and quantitative analysis methods based on genetic algorithm according to claim 5, which is characterized in that institute
Crossover probability is stated between 0.7~0.9, mutation probability between 0.05~0.15, select probability between 0.7~0.9 it
Between.
7. a kind of LIBS calibration and quantitative analysis methods based on genetic algorithm according to claim 1, which is characterized in that institute
Stating termination condition is：Adaptive optimal control degree individual does not change or reaches the maximum evolutionary generation of setting after several generations evolve.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510259910.6A CN106290263B (en) | 2015-05-19 | 2015-05-19 | A kind of LIBS calibration and quantitative analysis methods based on genetic algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510259910.6A CN106290263B (en) | 2015-05-19 | 2015-05-19 | A kind of LIBS calibration and quantitative analysis methods based on genetic algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106290263A CN106290263A (en) | 2017-01-04 |
| CN106290263B true CN106290263B (en) | 2018-08-24 |
Family
ID=57633843
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510259910.6A Active CN106290263B (en) | 2015-05-19 | 2015-05-19 | A kind of LIBS calibration and quantitative analysis methods based on genetic algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106290263B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109668862B (en) * | 2017-10-17 | 2021-02-05 | 中国科学院沈阳自动化研究所 | Aluminum electrolyte molecular ratio detection method based on laser-induced breakdown spectroscopy |
| CN107887289B (en) * | 2017-11-13 | 2021-03-09 | 北京半导体专用设备研究所（中国电子科技集团公司第四十五研究所） | Method and device for obtaining parameter value of film to be measured |
| CN108414475B (en) * | 2018-01-30 | 2020-06-26 | 中国科学院上海技术物理研究所 | LIBS analysis method based on optical chromatography simultaneous iterative reconstruction |
| CN110780021A (en) * | 2019-10-30 | 2020-02-11 | 广船国际有限公司 | Method and device for determining standard substance, terminal and storage medium |
| CN112414996B (en) * | 2020-07-24 | 2022-06-17 | 北京工商大学 | Finite difference and difference evolution algorithm-based ICP-AES spectral line overlapping interference correction method |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101915753A (en) * | 2010-07-30 | 2010-12-15 | 浙江师范大学 | Genetic Neural NetworkQuantitative analysis method for laser induced breakdown spectroscopy based on gGenetic Neural Networkgenetic neural network |
| CN103778469A (en) * | 2013-01-23 | 2014-05-07 | 辽宁工程技术大学 | Blasting scheme selection method based on neural network optimization genetic algorithm |
-
2015
- 2015-05-19 CN CN201510259910.6A patent/CN106290263B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101915753A (en) * | 2010-07-30 | 2010-12-15 | 浙江师范大学 | Genetic Neural NetworkQuantitative analysis method for laser induced breakdown spectroscopy based on gGenetic Neural Networkgenetic neural network |
| CN103778469A (en) * | 2013-01-23 | 2014-05-07 | 辽宁工程技术大学 | Blasting scheme selection method based on neural network optimization genetic algorithm |
Non-Patent Citations (3)
| Title |
|---|
| "基于遗传神经网络的激光诱导击穿光谱元素定量分析技术";沈沁梅;《中国优秀硕士学位论文全文数据库 基础科学辑》;20120715(第07期);论文全文 * |
| "基于遗传算法和偏最小二乘法的土壤激光诱导击穿光谱定量分析研究";邹孝恒 等;《分析化学》;20150228;第43卷(第2期);第182-184页 * |
| "多光谱峰值分离技术在LIBS中的应用研究";刘金桐;《中国优秀硕士学位论文全文数据库 信息科技辑》;20131215(第S2期);论文全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106290263A (en) | 2017-01-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106290263B (en) | A kind of LIBS calibration and quantitative analysis methods based on genetic algorithm | |
| CN105181678B (en) | Rice varieties discrimination method based on LIBS | |
| CN106680238B (en) | Method based on infrared spectrum analysis material component content | |
| CN106918567B (en) | A kind of method and apparatus measuring trace metal ion concentration | |
| CN105717066B (en) | A kind of near infrared spectrum identification model based on weighted correlation coefficient | |
| CN112231621B (en) | Method for reducing element detection limit based on BP-adaboost | |
| CN111965140B (en) | Wavelength point recombination method based on characteristic peak | |
| CN104730042A (en) | Method for improving free calibration analysis precision by combining genetic algorithm with laser induced breakdown spectroscopy | |
| CN107860743A (en) | Utilize the method and its application of the model of reflective near infrared fibre-optical probe structure fast prediction oil property | |
| CN106442474B (en) | A kind of cement slurry three ratio measurement method based on Partial Least Squares | |
| CN106990056A (en) | A kind of total soil nitrogen spectrum appraising model calibration samples collection construction method | |
| CN102128805A (en) | Method and device for near infrared spectrum wavelength selection and quick quantitative analysis of fruit | |
| CN100370242C (en) | Process for simultaneous determination of stained element content in gray glass utilizing plasma emission spectrometer | |
| CN103134770B (en) | Eliminate moisture detects total nitrogen content of soil impact method near infrared spectrum | |
| CN110567941B (en) | Rice seed moisture content grading detection method based on main element spectral intensity | |
| CN108693139A (en) | The near infrared prediction model method for building up of electronics tobacco tar physical and chemical index and application | |
| CN115436407A (en) | Element content quantitative analysis method combining random forest regression with principal component analysis | |
| CN114486821B (en) | Metallurgical spectral feature regression method, device, electronic equipment and storage medium | |
| CN104350378B (en) | Method and apparatus for the performance of measure spectrum system | |
| CN106198433A (en) | Infrared spectrum method for qualitative analysis based on LM GA algorithm | |
| CN106295667A (en) | A kind of method and application thereof selecting optimum spectrum based on genetic algorithm | |
| CN101666746B (en) | Laser induced spectrum data processing method based on wavelet analysis | |
| CN115165770A (en) | Water body COD (chemical oxygen demand) and turbidity simultaneous detection method based on broad spectrum and BPNN (BPNN) | |
| CN114354666A (en) | Method for extracting and optimizing spectral characteristics of soil heavy metal based on wavelength frequency selection | |
| CN109871356A (en) | A kind of data processing method and device for Soil K+adsorption |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |