CN108519347B - Infrared spectrum wavelength selection method based on binary dragonfly algorithm - Google Patents
Infrared spectrum wavelength selection method based on binary dragonfly algorithm Download PDFInfo
- Publication number
- CN108519347B CN108519347B CN201810314240.7A CN201810314240A CN108519347B CN 108519347 B CN108519347 B CN 108519347B CN 201810314240 A CN201810314240 A CN 201810314240A CN 108519347 B CN108519347 B CN 108519347B
- Authority
- CN
- China
- Prior art keywords
- dragonfly
- individual
- infrared spectrum
- binary
- wavelength selection
- 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
- 241000238633 Odonata Species 0.000 title claims abstract description 90
- 238000002329 infrared spectrum Methods 0.000 title claims abstract description 44
- 238000004422 calculation algorithm Methods 0.000 title claims abstract description 28
- 238000010187 selection method Methods 0.000 title claims abstract description 15
- 238000000034 method Methods 0.000 claims abstract description 21
- 238000005457 optimization Methods 0.000 claims abstract description 21
- 238000012216 screening Methods 0.000 claims abstract description 9
- 238000011156 evaluation Methods 0.000 claims description 16
- 239000011159 matrix material Substances 0.000 claims description 12
- 230000008569 process Effects 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000000926 separation method Methods 0.000 claims description 5
- 230000002776 aggregation Effects 0.000 claims description 4
- 238000004220 aggregation Methods 0.000 claims description 4
- 238000013507 mapping Methods 0.000 claims description 3
- 230000004931 aggregating effect Effects 0.000 claims description 2
- 238000004458 analytical method Methods 0.000 claims description 2
- 125000004122 cyclic group Chemical group 0.000 claims description 2
- 238000004445 quantitative analysis Methods 0.000 claims description 2
- 239000007789 gas Substances 0.000 abstract description 2
- 239000007791 liquid phase Substances 0.000 abstract description 2
- 239000012071 phase Substances 0.000 abstract description 2
- 239000007790 solid phase Substances 0.000 abstract description 2
- 230000006870 function Effects 0.000 description 19
- 230000003595 spectral effect Effects 0.000 description 5
- 238000010521 absorption reaction Methods 0.000 description 4
- 239000000126 substance Substances 0.000 description 4
- 230000006399 behavior Effects 0.000 description 2
- TVMXDCGIABBOFY-UHFFFAOYSA-N octane Chemical compound CCCCCCCC TVMXDCGIABBOFY-UHFFFAOYSA-N 0.000 description 2
- 238000005295 random walk Methods 0.000 description 2
- 238000001228 spectrum Methods 0.000 description 2
- 241001465754 Metazoa Species 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012844 infrared spectroscopy analysis Methods 0.000 description 1
- 238000012417 linear regression Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 238000012628 principal component regression Methods 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Images
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
Abstract
The invention belongs to the technical field of infrared spectrum wavelength selection methods, and particularly relates to an infrared spectrum wavelength selection method based on a binary dragonfly algorithm. The method converts the infrared spectrum wavelength screening problem into a binary optimization problem, innovatively utilizes a dragonfly algorithm to perform optimization solution, and simultaneously provides a series of strategies to ensure the overall search and convergence speed of optimization, thereby screening the optimal wavelength point combination. Compared with the prior art, the method has the advantages of avoiding falling into local extreme points, high convergence speed and the like, and can be widely applied to the field of infrared spectrum wavelength selection of solid phase, liquid phase and gas phase.
Description
Technical Field
The invention belongs to the technical field of infrared spectrum wavelength selection methods, and particularly relates to an infrared spectrum wavelength selection method based on a binary dragonfly algorithm.
Background
In recent years, infrared spectroscopy analysis technology has been widely applied in the fields of agriculture, chemical industry, environmental monitoring and the like due to the advantages of rapidness, no damage, no pollution and the like. However, in practical applications, three problems often need to be faced: (1) the number of wavelength points in the infrared spectrum waveband range is usually far larger than the number of scanned samples; (2) different components often have identical absorption peaks, i.e. the absorption peaks overlap; (3) there is often a correlation between adjacent wavelength points, i.e. a problem of collinearity. The above problems create great difficulties for qualitative and quantitative modeling analysis of infrared spectra. In order to solve the problems, spectra in the whole wave band range are screened before modeling, and the most representative wavelength point combination is selected, so that the method has very important theoretical significance and practical value.
For the infrared spectrum wavelength selection problem, many researchers have proposed a series of methods, such as: a forward selection method, a group optimization algorithm, an interval partial least square method, an information-free variable elimination method and the like. The method comprises the steps of simulating living habits of organisms (including animals and plants) in nature by using a swarm optimization algorithm represented by a genetic algorithm, a particle swarm algorithm and a bat algorithm, abstracting a series of local search and global search strategies, and realizing the screening of infrared spectrum wavelengths by optimizing an objective function.
In 2016, Seyedali Mirjalii creatively proposes a dragonfly algorithm (Neural comprehensive & Applic,2016,27: 1053-. Therefore, there is a need for an improvement to the basic dragonfly algorithm to enable the selection of wavelengths in the infrared spectrum.
Disclosure of Invention
Aiming at the technical problem, the invention provides an infrared spectrum wavelength selection method based on a binary dragonfly algorithm.
In order to solve the technical problems, the invention adopts the technical scheme that:
an infrared spectrum wavelength selection method based on a binary dragonfly algorithm abstracts infrared spectrum wavelength selection into a binary optimization problem, utilizes the dragonfly algorithm to solve, introduces strategies to ensure the optimization global property and convergence speed, and screens out the optimal wavelength point combination.
Abstracting infrared spectral wavelength selection as a binary optimization problem includes:
selecting wavelength, namely dividing the whole infrared spectrum band range into k subintervals, wherein k is a positive integer;
constructing a binary sequence M ═ M with length k1,m2,...,mk]Wherein m isi∈{0,1},i=1,2,...,k,miThe value 1 or 0 represents that the corresponding subinterval is selected or not selected.
The interval division of the subintervals is uniform or non-uniform, and when each subinterval only contains 1 wavelength point, the screening of the single wavelength point can be realized.
The dragonfly algorithm comprises the following steps: initializing dragonfly population, randomly generating L binary sequence dragonfly individuals to form initial population S0={M1,M2,...,ML};
Evaluating the fitness function value of dragonfly individuals, and aiming at each individual M in the initial populationi(i ═ 1, 2.. gtoren., L), establishing a corresponding infrared spectrum quantitative analysis model, evaluating the generalization performance of the model, and taking the generalization performance as the evaluation of each individual M in the populationiA fitness function of;
updating the positions of the food source and the natural enemy, wherein the position of the optimal individual is used as the food source, and the worst individual is used as the natural enemy;
updating the position of each dragonfly individual in the population;
binarization, performing binarization on each individual in the population, and checking and ensuring that a new individual never appears in the old population;
judging whether a shutdown condition is met, if so, ending and outputting an infrared spectrum wavelength selection optimization result; and if the shutdown condition is not met, returning the evaluation of the individual fitness function value of the dragonfly, and performing cyclic calculation until the shutdown condition is met.
In the evaluation of the dragonfly individual fitness function value, the infrared spectrum scanning matrix X of the whole wave band is extracted according to the position of the contained non-zero element by columns to obtain a new matrix Xi(i ═ 1, 2.., L), i.e., XiIs a subset of X; for each Xi(i ═ 1, 2.. times, L), a mapping model is established between it and the component content matrix Y to be analyzed.
In the updating of the food source and the natural enemy position, all dragonfly individuals in the population are ranked according to the fitness function value, and the position of the dragonfly individual with the optimal fitness function value is used as the position of the food source; and taking the position of the dragonfly individual with the worst fitness function value as the position of the natural enemy.
The updating of the position of each dragonfly individual in the population means that if the dragonfly individual MiThere are other dragonflies individuals in the neighborhood of (A), each dragonfly is individualThe location update of the body consists of the following 5 processes: separating, aligning, aggregating, predating, and deterrent, as follows:
(1) separation:wherein M isiRepresenting the current position of the ith dragonfly individual;represents and MiThe position of the adjacent jth dragonfly individual; n represents and MiThe number of adjacent dragonfly individuals;
(2) alignment:wherein the content of the first and second substances,represents and MiThe speed of the adjacent jth dragonfly individual;
(4) predation: fi=M+-MiWherein M is+Indicating the location of the food source;
(5) and (3) avoiding the enemy: ei=M--MiWherein M-represents the location of a natural enemy;
otherwise, if dragonfly individual MiThere are no other dragonfly individuals in the neighborhood of (A), dragonfly individual MiThe position updating of (a) is realized by random walk meeting the Levy flight strategy, and the specific calculation formula is as follows: mi,t+1=Mi,t+Lévy(d)×Mi,tWhere d is the dimension of the position vector; l (y) (x) 0.01 × r1×σ/|r2|1/βWherein r is1And r2Is a random number between 0 and 1; β is a constant, taken here to be 1.5; the formula for calculating σ is as follows:
in the evaluation of the dragonfly individual fitness function value, the evaluation indexes of the model generalization performance comprise absolute error, relative error, error square sum, root mean square error and decision coefficient.
The shutdown conditions include: maximum iteration times and the error tolerance of the fitness function value of two adjacent iterations.
Compared with the prior art, the invention has the following beneficial effects:
the method converts the infrared spectrum wavelength screening problem into a binary optimization problem, innovatively utilizes a dragonfly algorithm to perform optimization solution, and simultaneously provides a series of strategies to ensure the overall search and convergence speed of optimization, thereby screening the optimal wavelength point combination.
Compared with the prior art, the method has the advantages of avoiding falling into local extreme points, high convergence speed and the like, and can be widely applied to the field of infrared spectrum wavelength selection of solid phase, liquid phase and gas phase.
Drawings
FIG. 1 is a near infrared spectrum of 60 gasoline samples;
FIG. 2 is a flow chart of the present invention;
FIG. 3 is a graph showing the results of selecting wavelengths in the infrared spectrum of the present invention;
Detailed Description
The technical solutions in the embodiments of the present invention are clearly and completely described below, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Assuming that N samples are provided, the infrared spectrum signal scanned by the spectrometer isCorresponding to a content of the component to be analyzed ofWherein, P is the number of wavelength points of the infrared spectrum, and N & lt P is generally used.
The content prediction model of the component to be analyzed can be expressed as
Y=XΦ+ (1)
Wherein the content of the first and second substances,is the regression coefficient to be fitted;is a noise error.
The infrared spectral wavelength selection is abstracted to a binary optimization problem:
to achieve wavelength selection, the entire infrared spectral band range is first divided into k subintervals, k being a positive integer. It should be noted that the division of the interval may be uniform or non-uniform (in most cases, non-uniform division is more reasonable, since the absorption peak range of the component to be analyzed is usually non-uniform). In particular, when k is equal to P, i.e. each subinterval contains only 1 wavelength point, it is possible to screen for a single wavelength point.
Secondly, the infrared spectrum wavelength selection problem needs to be abstracted to a binary optimization problem. Constructing a binary sequence M ═ M with length k1,m2,...,mk]Wherein m isi∈{0,1},i=1,2,...,k。
Initializing dragonfly population:
then, L binary sequences are randomly generated to form an initial population S0={M1,M2,...,ML}。
Dragonfly individual fitness function value evaluation:
for each individual M in the initial populationi(i 1, 2.. times.l), and extracting the infrared spectrum scanning matrix X of the whole wave band according to the positions of the non-zero elements contained in the infrared spectrum scanning matrix X by columnsGetting to obtain a new matrix Xi(i ═ 1, 2.., L). Namely XiA subset of X.
Then, for each Xi(i ═ 1, 2.. times, L), a mapping model is established between it and the component content matrix Y to be analyzed. Note that the model can be built by linear methods (such as multiple linear regression, principal component regression, partial least squares, etc.) or by non-linear methods (such as artificial neural networks, support vector machines, deep learning, etc.). After the model is built, the generalization performance of the model needs to be evaluated, and the evaluation indexes generally include: absolute error, relative error, sum of squared errors, root mean square error, coefficient of determination, etc., as the evaluation of each individual M in the populationiA fitness function of. The bigger the fitness function value is, the larger the fitness function value is, the individual M is showniThe better; otherwise, the individual M is indicatediThe worse.
And (3) updating the positions of the food sources and the natural enemies:
in each iteration process, sequencing all dragonfly individuals in the population according to the fitness function value, wherein the position of the dragonfly individual with the optimal fitness function value is used as the position of a food source; and conversely, the position of the dragonfly individual with the worst fitness function value is taken as the position of the natural enemy.
Updating the position of each dragonfly individual in the population:
on the basis, each individual M is subjected to dragonfly algorithmiThe position is updated, and each individual is assumed to correspond to a dragonfly, and the daily behaviors of the dragonfly generally comprise the following 5 types: separation (Separation), Alignment (Alignment), aggregation (Cohesion), predation (Food) and deterrence (entity). Dragonfly individual MiThe updating of the location is realized through the above 5 actions, and the specific process is as follows:
(1) separation:wherein M isiRepresenting the current position of the ith dragonfly individual;represents and MiThe position of the adjacent jth dragonfly individual; n represents and MiThe number of adjacent dragonfly individuals.
(2) Alignment:wherein the content of the first and second substances,represents and MiThe speed of the adjacent jth dragonfly individual.
(4) predation: fi=M+-Mi. Wherein M is+Indicating the location of the food source.
(5) And (3) avoiding the enemy: ei=M--Mi. Wherein M is-Indicating the location of the natural enemy.
Dragonfly individual MiThe next generation of location updates are as follows: mi,t+1=Mi,t+ΔMi,t+1. Wherein, Δ Mi,t+1Is MiThe calculation formula of the position update step length of (2) is as follows: Δ Mi,t+1=(sSi+aAi+cCi+fFi+eEi)+wΔMi,t. Wherein the subscripts t and t +1 denote the t and t +1 th iterations, respectively; parameters s, a, c, f and e respectively represent weight coefficients of the 5 dragonfly behaviors; the parameter w represents the inertial weight.
From the above process, it can be easily found that the individual M of dragonflyiThe premise for realizing the position updating is that other dragonfly individuals exist in the neighborhood. Otherwise, it will not be possible to update its position, in this case the dragonfly individual MiThe position updating of (a) is realized by random walk meeting the Levy flight strategy, and the specific calculation formula is as follows: mi,t+1=Mi,t+Lévy(d)×Mi,t. Where d is the dimension of the position vector; l (y) (x) 0.01 × r1×σ/|r2|1/βWherein r is1And r2To be connected toA random number between 0 and 1; β is a constant, taken here to be 1.5; the formula for calculating σ is as follows:
binarization:
unlike the continuous optimization problem, the individual dragonfly M is here, as described aboveiIs a binary sequence, i.e. the value of an element in the sequence is not 0, i.e. 1. Therefore, each individual after location update needs to be binarized, and the specific process is as follows:
wherein r is a random number between 0 and 1.
It should be noted that each dragonfly individual in the initial population will get a new population after being subjected to location update through the above process. Since each individual dragonfly is a binary sequence, there may be a case where: the newly generated individual dragonfly is the same as the individual dragonfly in the previous iteration. In order to avoid the situation, the newly generated dragonfly individuals need to be compared with the previous population, if the dragonfly individuals are the same, the new individuals are deleted, and new dragonfly individuals are generated in a way of randomly walking by L é vy flight until the new individuals never appear in the previous population.
Judging whether the shutdown condition is met:
and if so, ending and outputting the infrared spectrum wavelength selection optimization result. Otherwise, if the shutdown condition is not met, returning to the dragonfly individual fitness function value evaluation step, and circularly calculating until the shutdown condition is met.
In summary, a detailed flow of wavelength screening of infrared spectra using the binary dragonfly algorithm is shown in fig. 2. The related model generalization performance evaluation indexes include but are not limited to: absolute error, relative error, error sum of squares, root mean square error, decision coefficient, etc.; the shutdown conditions involved include, but are not limited to: maximum iteration times, error tolerance of fitness function values of two adjacent iterations and the like.
The following will analyze the present invention in accordance with the embodiments shown in fig. 1, 2 and 3.
FIG. 1 is a near infrared spectrum of 60 gasoline samples, wherein the spectrum scanning range is 900-1700nm, and a wavelength point is scanned every 2 nm. Thus, the scanned spectral information matrix isThe corresponding component to be analyzed is the octane number contained in the gasoline sample, i.e. the octane number
Firstly, according to the result characteristics of a near-infrared spectrogram of gasoline, a non-uniform division method is adopted to divide the whole spectral range into 98 blocks, wherein 63 blocks only comprise two wavelength points, 1 block comprises 3 wavelength points, and the rest 34 blocks comprise 8 wavelength points.
Next, according to the steps in fig. 2, the binary dragonfly algorithm is used to perform the optimization solution. The parameter settings involved are as follows: the maximum number of iterations is 200; the population size is 50; the quantitative modeling adopts an Extreme Learning Machine (ELM); the training set contained 40 samples; the test set contained the remaining 20 samples; the evaluation index of the model generalization performance is a determination coefficient R2。
The calculated wavelength selection results are shown in fig. 3. It can be seen that 32 wavelength points are screened out from 64 blocks with larger contribution, and are uniformly distributed near strong absorption peaks (1150nm, 1190nm, 1400nm and 1700 nm); of the 34 blocks with less contribution, 18 blocks are screened out, and 144 wavelength points are screened out.
Although only the preferred embodiments of the present invention have been described in detail, the present invention is not limited to the above embodiments, and various changes can be made without departing from the spirit of the present invention within the knowledge of those skilled in the art, and all changes are encompassed in the scope of the present invention.
Claims (8)
1. An infrared spectrum wavelength selection method based on a binary dragonfly algorithm is characterized by comprising the following steps: selecting and abstracting infrared spectrum wavelength into a binary optimization problem, solving by using a dragonfly algorithm, introducing a strategy to ensure the global property and convergence speed of optimization, and screening out an optimal wavelength point combination;
the dragonfly algorithm comprises the following steps:
initializing dragonfly population, randomly generating L binary sequence dragonfly individuals to form initial population S0={M1,M2,...,ML};
Evaluating the fitness function value of dragonfly individuals, and aiming at each individual M in the initial populationiAnd L, establishing a corresponding infrared spectrum quantitative analysis model, evaluating the generalization performance of the model, and taking the generalized analysis model as an evaluation result of each individual M in the populationiA fitness function of;
updating the positions of the food source and the natural enemy, wherein the position of the optimal individual is used as the food source, and the worst individual is used as the natural enemy;
updating the position of each dragonfly individual in the population;
binarization, performing binarization on each individual in the population, and checking and ensuring that a new individual never appears in the old population;
judging whether a shutdown condition is met, if so, ending and outputting an infrared spectrum wavelength selection optimization result; and if the shutdown condition is not met, returning the evaluation of the individual fitness function value of the dragonfly, and performing cyclic calculation until the shutdown condition is met.
2. The infrared spectrum wavelength selection method based on the binary dragonfly algorithm as claimed in claim 1, wherein abstracting infrared spectrum wavelength selection to a binary optimization problem comprises:
selecting wavelength, namely dividing the whole infrared spectrum band range into k subintervals, wherein k is a positive integer;
construct a lengthBinary sequence M ═ M for k1,m2,...,mk]Wherein m isi∈{0,1},i=1,2,...,k,miThe value 1 or 0 represents that the corresponding subinterval is selected or not selected.
3. The infrared spectrum wavelength selection method based on the binary dragonfly algorithm as claimed in claim 2, wherein the interval division of the subintervals is uniform or non-uniform, and when each subinterval only contains 1 wavelength point, the screening of the single wavelength point can be realized.
4. The infrared spectrum wavelength selection method based on the binary dragonfly algorithm as claimed in claim 1, wherein in the evaluation of the dragonfly individual fitness function value, the infrared spectrum scanning matrix X of the whole waveband is extracted in columns according to the position of the contained non-zero element, so as to obtain a new matrix Xi1, 2, L, i.e. XiIs a subset of X; for each XiAnd i 1, 2, L, establishing a mapping model between the content matrix and the component content matrix Y to be analyzed.
5. The infrared spectrum wavelength selection method based on the binary dragonfly algorithm as claimed in claim 1, wherein in the updating of the food source and the natural enemy position, all dragonfly individuals in the population are ranked according to the fitness function value, and the position of the dragonfly individual with the optimal fitness function value is taken as the position of the food source; and taking the position of the dragonfly individual with the worst fitness function value as the position of the natural enemy.
6. The method as claimed in claim 1, wherein the updating of the individual location of each dragonfly in the population is based on the M number of individual dragonfliesiOther individual dragonflies exist in the neighborhood of (2), and the position update of each individual dragonflies consists of the following 5 processes: separating, aligning, aggregating, predating, and deterrent, as follows:
(1) separation:wherein M isiRepresenting the current position of the ith dragonfly individual;represents and MiThe position of the adjacent jth dragonfly individual; n represents and MiThe number of adjacent dragonfly individuals;
(4) predation: fi=M+-MiWherein M is+Indicating the location of the food source;
(5) and (3) avoiding the enemy: ei=M--MiWherein M is-Representing the location of a natural enemy;
otherwise, if dragonfly individual MiThere are no other dragonfly individuals in the neighborhood of (A), dragonfly individual MiBy satisfying the location update ofThe random walking of the flight strategy is realized by the following specific calculation formula: where d is the dimension of the position vector;wherein r is1And r2Is a random number between 0 and 1; β is a constant, taken here to be 1.5; the formula for calculating σ is as follows:
7. the infrared spectrum wavelength selection method based on the binary dragonfly algorithm as claimed in claim 1, wherein in the evaluation of the dragonfly individual fitness function value, the evaluation indexes of the model generalization performance comprise absolute error, relative error, error sum of squares, root mean square error and decision coefficient.
8. The method of claim 1, wherein the shutdown condition comprises: maximum iteration times and the error tolerance of the fitness function value of two adjacent iterations.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810314240.7A CN108519347B (en) | 2018-04-10 | 2018-04-10 | Infrared spectrum wavelength selection method based on binary dragonfly algorithm |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810314240.7A CN108519347B (en) | 2018-04-10 | 2018-04-10 | Infrared spectrum wavelength selection method based on binary dragonfly algorithm |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108519347A CN108519347A (en) | 2018-09-11 |
CN108519347B true CN108519347B (en) | 2020-11-06 |
Family
ID=63432307
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810314240.7A Active CN108519347B (en) | 2018-04-10 | 2018-04-10 | Infrared spectrum wavelength selection method based on binary dragonfly algorithm |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108519347B (en) |
Families Citing this family (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110517326B (en) * | 2019-07-25 | 2023-06-13 | 江苏大学 | Colorimetric sensor array optimization method based on weight dragonfly algorithm |
CN111080031A (en) * | 2019-12-27 | 2020-04-28 | 圆通速递有限公司 | Vehicle path optimization method and system based on improved dragonfly algorithm |
CN113570289B (en) * | 2021-09-26 | 2022-01-25 | 广东电网有限责任公司中山供电局 | Account card object consistency management method and system based on improved dragonfly optimization algorithm |
CN115952402B (en) * | 2022-09-29 | 2023-06-27 | 南京林业大学 | Method for selecting near-infrared model transfer standard sample set based on binary dragonfly algorithm |
CN115983434A (en) * | 2022-12-07 | 2023-04-18 | 广州番禺职业技术学院 | Short-term wind field wind speed prediction method and system based on neural network |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107703751A (en) * | 2017-10-13 | 2018-02-16 | 河南工程学院 | PID controller optimization method based on dragonfly algorithm |
-
2018
- 2018-04-10 CN CN201810314240.7A patent/CN108519347B/en active Active
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107703751A (en) * | 2017-10-13 | 2018-02-16 | 河南工程学院 | PID controller optimization method based on dragonfly algorithm |
Non-Patent Citations (3)
Title |
---|
Dragonfly algorithm: a new meta-heuristic optimization technique for solving single-objective, discrete, and multi-objective problems;Mirjalil Seyedali;《Neural Comput & Applic》;20160531;第27卷(第4期);第1053-1073页 * |
一种改进的多目标蜻蜓优化算法;韩鹏 等;《微型机与应用》;20171031;第36卷(第20期);第27-33页 * |
基于改进蝙蝠算法的红外光谱特征选择;陈媛媛 等;《红外与激光工程》;20140831;第43卷(第8期);第2715-2721页 * |
Also Published As
Publication number | Publication date |
---|---|
CN108519347A (en) | 2018-09-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108519347B (en) | Infrared spectrum wavelength selection method based on binary dragonfly algorithm | |
Dray et al. | Principal component analysis with missing values: a comparative survey of methods | |
CN105279397B (en) | A kind of method of key protein matter in identification of protein interactive network | |
Swenson | Phylogenetic imputation of plant functional trait databases | |
Folguera et al. | Self-organizing maps for imputation of missing data in incomplete data matrices | |
Clark et al. | Testing the link between functional diversity and ecosystem functioning in a Minnesota grassland experiment | |
CN109886464B (en) | Low-information-loss short-term wind speed prediction method based on optimized singular value decomposition generated feature set | |
CN111785329A (en) | Single-cell RNA sequencing clustering method based on confrontation automatic encoder | |
CN109214503B (en) | Power transmission and transformation project cost prediction method based on KPCA-LA-RBM | |
CN110674947B (en) | Spectral feature variable selection and optimization method based on Stacking integrated framework | |
CN107885971B (en) | Method for identifying key protein by adopting improved flower pollination algorithm | |
Mota et al. | Evaluating the performance of machine learning methods and variable selection methods for predicting difficult-to-measure traits in Holstein dairy cattle using milk infrared spectral data | |
WO2023103130A1 (en) | Quantum walk-based time-series multiscale analysis method | |
CN106650926A (en) | Robust boosting extreme learning machine integrated modeling method | |
CN112668611A (en) | Short-term photovoltaic power generation power prediction method based on Kmeans and CEEMD-PE-LSTM | |
Yu et al. | Prediction of soil properties based on characteristic wavelengths with optimal spectral resolution by using Vis-NIR spectroscopy | |
Lai et al. | Lightcts: A lightweight framework for correlated time series forecasting | |
Zhao et al. | Comparison of two cluster analysis methods using single particle mass spectra | |
Hemmateenejad et al. | Clustering of variables in regression analysis: a comparative study between different algorithms | |
CN111795944B (en) | Near infrared spectrum wavelength screening method for improving team progress algorithm | |
CN106529680B (en) | A kind of multiple dimensioned extreme learning machine integrated modelling approach based on empirical mode decomposition | |
CN116501925B (en) | Method, system, equipment and medium for predicting leaf area index under multi-source remote sensing data | |
WO2021091883A1 (en) | Deep imitation learning for molecular inverse problems | |
Wei et al. | MOO-DNAS: Efficient neural network design via differentiable architecture search based on multi-objective optimization | |
Tan et al. | Calibration transfer between two near-infrared spectrometers based on a wavelet packet transform |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |