CN102928382B - Near-infrared spectral characteristic wavelength selecting method based on improved simulated annealing algorithm - Google Patents
Near-infrared spectral characteristic wavelength selecting method based on improved simulated annealing algorithm Download PDFInfo
- Publication number
- CN102928382B CN102928382B CN201210451289.XA CN201210451289A CN102928382B CN 102928382 B CN102928382 B CN 102928382B CN 201210451289 A CN201210451289 A CN 201210451289A CN 102928382 B CN102928382 B CN 102928382B
- Authority
- CN
- China
- Prior art keywords
- new explanation
- current solution
- parameter
- value
- solution
- 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 Or Analysing Materials By Optical Means (AREA)
- Spectrometry And Color Measurement (AREA)
Abstract
The invention provides a near-infrared spectral characteristic wavelength selecting method based on an improved simulated annealing algorithm. The near-infrared spectral characteristic wavelength selecting method comprises the following steps of: firstly, randomly initializing an existing solution S for a near-infrared spectral data set; then, generating a novel solution S' based on the current solution S under control of a temperature parameter t and a markoff chain length parameter L, and judging importance of the existing solution S and the novel solution S' according to an improved Metropoli acceptance criterion; and finally, returning to effects of the optimal characteristic wavelength and a model corresponding to the optimal characteristic wavelength when the temperature control parameter t reaches a set ending temperature. According to the near-infrared spectral characteristic wavelength selecting method, an Metropolis acceptance criterion in a traditional simulated annealing algorithm is improved, so that the improved simulated annealing algorithm not only improves an effect of selecting a near-infrared spectral characteristic wavelength and simplifies parameter setting of the simulated annealing algorithm, but also keeps the characteristics that the traditional simulated annealing algorithm jumps out of a local optimal solution and rapid convergence is achieved.
Description
Technical field
The present invention relates to near-infrared spectrum technique field, particularly relate to a kind of characteristic wavelength of near-infrared spectrum system of selection based on improved Simulated Annealing Algorithm.
Background technology
In recent years, near infrared spectrum is quantitative and method for qualitative analysis as one, obtains Successful utilization in many fields such as agricultural, environment, pharmacy.First near-infrared spectral analytical method obtains the spectral information that sample to be tested (is generally 700 ~ 2500nm) under specific near-infrared band, then utilizes spectral information to carry out quantitatively or qualitative analysis sample.The ultimate principle of near-infrared spectrum analysis is frequency multiplication and the sum of fundamental frequencies information that near infrared spectrum can record the vibrations of chemical bond fundamental frequency, and main hydric group X-H(X is mainly C, N, O) frequency multiplication that shakes and sum of fundamental frequencies absorb.Because the composition of sample is made up of these groups mostly, therefore near infrared light spectrum information can carry out qualitative and quantitative analysis to sample.By modern near infrared spectroscopy instrument, a large amount of near infrared light spectrum informations (spectral information of sample under hundreds and thousands of even up to ten thousand wavelength) can be obtained within the short time (being generally less than 1 minute).Consider stability and the complicacy of model, near infrared quantitatively, in qualitative analysis process only by a few features variable (being generally less than 10) for Modling model.Therefore from a large amount of spectral informations, how to select the maximally related characteristic wavelength with sample to be tested is one of committed step in NIR technology.Domestic and international expert, scholar do a lot of work in characteristic wavelength of near-infrared spectrum selection, propose the characteristic wavelength systems of selection such as principal component analysis (PCA), interval partial least square, genetic algorithm, result of study shows compared with full spectral model, and characteristic wavelength system of selection effectively can improve model accuracy and reduce model complexity.At patent of invention " system of selection based on the characteristic wavelength of near-infrared spectrum of simulated annealing " (application number: 201010123934.6) disclose a kind of characteristic wavelength of near-infrared spectrum system of selection based on traditional analog annealing algorithm.The method, when selecting characteristic wavelength, judge the importance of old solution and new explanation, but effect is unsatisfactory with Metropolis criterion.
In view of this, be necessary to propose a kind of characteristic wavelength of near-infrared spectrum system of selection based on improved Simulated Annealing Algorithm.
Summary of the invention
The object of the present invention is to provide a kind of characteristic wavelength of near-infrared spectrum system of selection based on improved Simulated Annealing Algorithm, improve the effect that traditional analog annealing algorithm selects characteristic wavelength of near-infrared spectrum.
A kind of characteristic wavelength of near-infrared spectrum system of selection based on improved Simulated Annealing Algorithm of the present invention, said method comprising the steps of:
S1, input near infrared spectrum data collection, the value of random initializtion current solution S, initialization global variable S_global is the value of current solution S, initialization temperature parameter t is initial temperature t
0;
S2, judge whether temperature parameter t is greater than end temp t
f, if so, then perform step S3; If not, then step S9 is performed;
The value of S3, initialization Markov chain length parameter L is 1;
S4, judge whether Markov chain length parameter L is less than maximum Markov chain length L_max, if so, then perform step S5; If not, then step S8 is performed;
S5, on the basis of current solution S, produce new explanation S ', if f (S ') < f (S_global), S_global is updated to new explanation S ';
S6, according to modified Metropolis acceptance criterion judge current solution S and new explanation S ' importance;
The value of S7, Markov chain length parameter L increases progressively 1, and returns step S4;
The value of S8, temperature parameter t successively decreases 1, and returns step S2;
S9, return S_global, set up PLS model, provide the validation-cross root-mean-square error of model, calibration set related coefficient, forecast set root-mean-square error and forecast set related coefficient.
As a further improvement on the present invention, in described step S6, " modified Metropolis acceptance criterion " is specially:
Calculate the current solution objective function f (S) of S and new explanation S ' and f (S '), wherein f (S) is the validation-cross root-mean-square error of PLS calibration model choosing wavelength corresponding in current solution S, the validation-cross root-mean-square error that f (S ') is the PLS calibration model of choosing wavelength corresponding in new explanation S ';
Judge
whether set up, wherein r, α is a constant in [0,1] scope if being a random number in [α, 1] scope;
If
set up and then judge that new explanation S ' is as important solution, and current solution S is updated to the value of new explanation S ',
If
be false, judge that current solution S is as important solution, and keep the value of current solution S constant.
As a further improvement on the present invention, described modified Metropolis acceptance criterion judges that the process of the importance of current solution S and new explanation S ' comprising:
If with current solution S-phase ratio, new explanation S ' is optimization solution, and meet f (S)-f (S ') > 0,
then
certain establishment, i.e. the acceptance probability P of new explanation (S ') be 100%;
If with current solution S-phase ratio, new explanation S ' separates for worsening, and meets
then
there is certain probability to set up, i.e. the acceptance probability P of new explanation (S ') be
If with current solution S-phase ratio, new explanation S ' separates for worsening, and meets
then
set up scarcely, i.e. the acceptance probability P of new explanation (S ') be 0.
As a further improvement on the present invention, need before described step S1 to determine following parameter:
Initial temperature t
0, corresponding to the initial value of temperature parameter t, for being greater than the constant of 0;
End temp t
f, corresponding to the end value of temperature parameter t, be initial temperature t
0little constant;
Parameter alpha and β, parameter alpha is the constant in [0,1] scope, β be greater than 0 constant;
Parameter h, corresponds to and produce the columns changed in new explanation S ' process, for being greater than the constant of 1 on the basis of current solution S;
Maximum Markov chain length L_max, for being greater than the integer of 1.
As a further improvement on the present invention, described near infrared spectrum data collection comprises calibration set near infrared spectrum Xcal, calibration set chemical score Ycal, forecast set near infrared spectrum Xpre and forecast set chemical score Ypre.
As a further improvement on the present invention, the spectrum of described calibration set near infrared spectrum Xcal and calibration set chemical score Ycal, forecast set near infrared spectrum Xpre and forecast set chemical score Ypre and chemical score one_to_one corresponding respectively.
The invention has the beneficial effects as follows: by the improvement to Metropolis acceptance criterion in traditional analog annealing algorithm, make improved Simulated Annealing Algorithm not only increase the effect selecting characteristic wavelength of near-infrared spectrum, the optimum configurations simplifying simulated annealing, and remain the feature that traditional analog annealing algorithm jumps out locally optimal solution and Fast Convergent.
Accompanying drawing explanation
Fig. 1 is the schematic flow sheet of the characteristic wavelength of near-infrared spectrum system of selection that the present invention is based on improved Simulated Annealing Algorithm;
Fig. 2 is the schematic flow sheet of modified Metropolis acceptance criterion of the present invention;
The graph of a relation that Fig. 3 (a) and 3 (b) are parameter alpha, β and deterioration solution acceptance probability in the present invention.
Embodiment
Describe the present invention below with reference to each embodiment shown in the drawings.But these embodiments do not limit the present invention, the structure that those of ordinary skill in the art makes according to these embodiments, method or conversion functionally are all included in protection scope of the present invention.
Ginseng Figure 1 shows that the schematic flow sheet of a kind of characteristic wavelength of near-infrared spectrum system of selection based on improved Simulated Annealing Algorithm of the present invention, comprises the following steps:
S1, input near infrared spectrum data collection, the value of random initializtion current solution S, initialization global variable S_global is the value of current solution S, initialization temperature parameter t is initial temperature t
0;
S2, judge whether temperature parameter t is greater than end temp t
f, if so, then perform step S3; If not, then step S9 is performed;
The value of S3, initialization Markov chain length parameter L is 1;
S4, judge whether Markov chain length parameter L is less than maximum Markov chain length L_max, if so, then perform step S5; If not, then step S8 is performed;
S5, on the basis of current solution S, produce new explanation S ', if f (S ') <f (S_global), S_global is updated to new explanation S ';
S6, according to modified Metropolis acceptance criterion judge current solution S and new explanation S ' importance;
The value of S7, Markov chain length parameter L increases progressively 1, and returns step S4;
The value of S8, temperature parameter t successively decreases 1, and returns step S2;
S9, return S_global, set up PLS model, provide the validation-cross root-mean-square error (RMSECV) of model, calibration set related coefficient (Rcal), forecast set root-mean-square error (RMSEP) and forecast set related coefficient (Rpre).
Near infrared spectrum data collection in the present invention comprises calibration set near infrared spectrum Xcal (m
1× n), calibration set chemical score Ycal (m
1× 1), forecast set near infrared spectrum Xpre (m
2× n) and forecast set chemical score Ypre (m
2× 1); Wherein m1 is calibration set sample number, m
2for forecast set sample number, n is the number of wavelengths of near infrared spectrum; The spectrum of the spectrum in calibration set and chemical score, forecast set and chemical score one_to_one corresponding respectively.
Current solution S is the one-dimension array of a 1 × n, and represent a kind of combination of selected characteristic wavelength, the length of this array is consistent with the number of wavelengths of near infrared spectrum, the i-th row S of current solution S
i(i ∈ [1, n]) is corresponding with the wavelength of i-th near infrared spectrum.Random initializtion current solution S refers to each row random assignment 0 or 1 of current solution S, if S
ithe value of (i ∈ [1, n]) is 0, represents that i-th wavelength is not selected; If S
ithe value of (i ∈ [1, n]) is 1, then represent that i-th wavelength is selected.Current solution S is only initialised when algorithm just brings into operation once, and current solution S upgrades according to modified Metropolis acceptance criterion subsequently.
Temperature parameter t is the simulation to temperature in solid annealing process, the beginning of control algolithm and end.When algorithm is incipient, t is assigned initial temperature, and cool 1 degree subsequently, when t is cooled to end temp, algorithm terminates at every turn.
The iterations of algorithm when Markov chain length parameter L refers to parametric t.
New explanation S ' is the one group of solution produced on the basis of current solution S, represents a kind of combination of selected characteristic wavelength.First-selected is current solution S by new explanation S ' assignment; Then h (h ∈ [1, the n]) row of Stochastic choice new explanation S '; The last value changing above-mentioned h row in new explanation S ' one by one, if the i-th row S in the h row of current solution S
ithe value of (i ∈ [1, h]) is 0, then the i-th row S ' in being arranged by the h of new explanation S '
ithe value of (i ∈ [1, h]) is set to 1, if the i-th row S in the h row of current solution S
ithe value of (i ∈ [1, h]) is 1, then the i-th row S ' in being arranged by the h of new explanation S '
ithe value of (i ∈ [1, h]) is set to 0.Current solution S and the difference of new explanation S ' are that the value that selected above-mentioned h arranges is different.
Modified Metropolis acceptance criterion removes parametric t on the basis of traditional Metropolis acceptance criterion, introduces parameter alpha and β.
Shown in ginseng Fig. 2, " modified Metropolis acceptance criterion " is specially:
Calculate the current solution objective function f (S) of S and new explanation S ' and f (S '), wherein f (S) is current validation-cross root-mean-square error (RMSECV) of separating the PLS calibration model choosing wavelength corresponding in S, computing method are the wavelength according to choosing in S, extract the spectral information choosing wavelength corresponding in calibration set spectrum Xcal, set up PLS model with the chemical score Ycal of calibration set, the validation-cross root-mean-square error (RMSECV) that this PLS model is corresponding is f (S); The validation-cross root-mean-square error (RMSECV) that f (S ') is the PLS calibration model of choosing wavelength corresponding in new explanation S ', computing method are according to the wavelength chosen in S ', extract the spectral information choosing wavelength corresponding in calibration set spectrum Xcal, set up PLS model with the chemical score Ycal of calibration set, validation-cross root-mean-square error (RMSECV) corresponding to this PLS model is f (S ');
Judge
whether set up, wherein r, α is a constant in [0,1] scope if being a random number in [α, 1] scope;
If
set up and then judge that new explanation S ' is as important solution, and current solution S is updated to the value of new explanation S ',
If
be false, judge that current solution S is as important solution, and keep the value of current solution S constant.
Modified Metropolis acceptance criterion judges that the process of the importance of current solution S and new explanation S ' comprising:
If with current solution S-phase ratio, new explanation S ' is optimization solution, and meet f (S)-f (S ') > 0,
then
certain establishment, i.e. the acceptance probability P of new explanation (S ') be 100%;
If with current solution S-phase ratio, new explanation S ' separates for worsening, and meets
then
there is certain probability to set up, i.e. the acceptance probability P of new explanation (S ') be
If with current solution S-phase ratio, new explanation S ' separates for worsening, and meets
then
set up scarcely, i.e. the acceptance probability P of new explanation (S ') be 0.
The present invention the effect returning the corresponding model of optimal characteristics wavelength and optimal characteristics wavelength refer to just start at algorithm time, set up a global variable S_global for recording the optimal characteristics wavelength combinations obtained in algorithm operational process, the initial value of this global variable S_global equals the initial value of current solution S, whenever algorithm obtains (f that satisfies condition (S ') <f (S_global) when new explanation S ' is better than S_global), S_global is updated to new explanation S '.When temperature control parameter t reaches the end temp of setting, return characteristic wavelength combination that S_global chooses and PLS model is set up to the characteristic wavelength combination chosen, and providing validation-cross root-mean-square error (RMSECV) corresponding to PLS model, calibration set related coefficient (Rcal), forecast set root-mean-square error (RMSEP) and forecast set related coefficient (Rpre).
The present invention needs to determine following parameter before adopting improved Simulated Annealing Algorithm to select characteristic wavelength:
(1) initial temperature t
0: corresponding to the initial value of temperature parameter t, for being greater than the constant of 0;
(2) end temp t
f: corresponding to the end value of temperature parameter t, be initial temperature t
0little constant;
(3) parameter alpha and β: parameter alpha is the constant in [0,1] scope, β be greater than 0 constant; Due to the important parameter that parameter alpha and β are modified Metropolis acceptance criterions, point 2 kinds of situations are discussed the size of α and β and worsen the relation of separating acceptance probability:
Situation (1), keeps the value of parameter beta constant, allows parameter alpha increase progressively in [0,1] scope, as shown in Fig. 3 (a), suppose to get β=5, and 3 of α value α
1< α
2< α
3increase progressively gradually, middle horizontal ordinate X=f (the S)-f of Fig. 3 (a) (S '), X ∈ [-5,5], ordinate P (S ')=exp (X/ β), the less new explanation S ' that represents is poorer than current solution S for the value of X, and when X is less than 0, new explanation S ' separates for worsening; As can be seen from 3 value α of Fig. 3 (a), α
1< α
2< α
3when increasing progressively gradually, the value X of corresponding X
1< X
2< X
3increase progressively gradually, show increasing progressively along with α, worsen the acceptance probability separated and to taper off trend;
Situation (2), keeps the value of parameter alpha constant, allows parameter beta increase progressively, as shown in Fig. 3 (b), suppose to get α=0.5, and 3 of β value β
1< β
2< β
3increase progressively gradually, middle horizontal ordinate X=f (the S)-f of Fig. 3 (b) (S '), and X ∈ [-5,1], ordinate P (S ')=exp (X/ β), as can be seen from 3 value β of Fig. 3 (b), β
1< β
2< β
3increase progressively gradually, the value X of corresponding X
1> X
2> X
3successively decrease gradually, show increasing progressively along with β, worsening the acceptance probability separated is increasing trend;
(4) parameter h: correspond to and produce the columns changed in new explanation S ' process, for being greater than the constant of 1 on the basis of current solution S;
(5) maximum Markov chain length L_max: for being greater than the integer of 1.
To near infrared spectrum data collection in the present invention, first random initializtion current solution S; Then, under the control of temperature parameter t and Markov chain length parameter L, so that the basis of current solution S to produce new explanation S ', and the importance of current solution S and new explanation S ' is judged with modified Metropolis acceptance criterion; Finally when temperature control parameter t reaches the end temp of setting, return the effect of optimal characteristics wavelength and the corresponding model of optimal characteristics wavelength.
Beneficial effect is: by the improvement to Metropolis acceptance criterion in traditional analog annealing algorithm, make improved Simulated Annealing Algorithm not only increase the effect selecting characteristic wavelength of near-infrared spectrum, the optimum configurations simplifying simulated annealing, and remain the feature that traditional analog annealing algorithm jumps out locally optimal solution and Fast Convergent.
Be to be understood that, although this instructions is described according to embodiment, but not each embodiment only comprises an independently technical scheme, this narrating mode of instructions is only for clarity sake, those skilled in the art should by instructions integrally, technical scheme in each embodiment also through appropriately combined, can form other embodiments that it will be appreciated by those skilled in the art that.
A series of detailed description listed is above only illustrating for feasibility embodiment of the present invention; they are also not used to limit the scope of the invention, all do not depart from the skill of the present invention equivalent implementations done of spirit or change all should be included within protection scope of the present invention.
Claims (5)
1., based on a characteristic wavelength of near-infrared spectrum system of selection for improved Simulated Annealing Algorithm, it is characterized in that, said method comprising the steps of:
S1, input near infrared spectrum data collection, the value of random initializtion current solution S, initialization global variable S_global is the value of current solution S, initialization temperature parameter t is initial temperature t
0;
S2, judge whether temperature parameter t is greater than end temp t
f, if so, then perform step S3; If not, then step S9 is performed;
The value of S3, initialization Markov chain length parameter L is 1;
S4, judge whether Markov chain length parameter L is less than maximum Markov chain length L_max, if so, then perform step S5; If not, then step S8 is performed;
S5, on the basis of current solution S, produce new explanation S ', if f (S ') <f (S_global), S_global is updated to new explanation S '; The objective function that f (S ') is new explanation S ', the objective function that f (S_global) is global variable S_global;
S6, according to modified Metropolis acceptance criterion judge current solution S and new explanation S ' importance; Described " modified Metropolis acceptance criterion " is specially:
Calculate the current solution objective function f (S) of S and new explanation S ' and f (S '), wherein f (S) is the validation-cross root-mean-square error of PLS calibration model choosing wavelength corresponding in current solution S, the validation-cross root-mean-square error that f (S ') is the PLS calibration model of choosing wavelength corresponding in new explanation S ';
Judge
whether set up, wherein r, α is a constant in [0,1] scope if being a random number in [α, 1] scope, β be greater than 0 constant;
If
set up and then judge that new explanation S ' is as important solution, and current solution S is updated to the value of new explanation S ',
If
be false, judge that current solution S is as important solution, and keep the value of current solution S constant;
The value of S7, Markov chain length parameter L increases progressively 1, and returns step S4;
The value of S8, temperature parameter t successively decreases 1, and returns step S2;
S9, return S_global, set up PLS model, provide the validation-cross root-mean-square error of model, calibration set related coefficient, forecast set root-mean-square error and forecast set related coefficient.
2. method according to claim 1, is characterized in that, described modified Metropolis acceptance criterion judges that the process of the importance of current solution S and new explanation S ' comprising:
If with current solution S-phase ratio, new explanation S ' is optimization solution, and meet f (S)-f (S ') >0,
then
certain establishment, i.e. the acceptance probability P of new explanation (S ') be 100%;
If with current solution S-phase ratio, new explanation S ' separates for worsening, and meets
then
there is certain probability to set up, i.e. the acceptance probability P of new explanation (S ') be
If with current solution S-phase ratio, new explanation S ' separates for worsening, and meets
then
set up scarcely, i.e. the acceptance probability P of new explanation (S ') be 0.
3. method according to claim 1, is characterized in that, needs to determine following parameter before described step S1:
Initial temperature t
0, corresponding to the initial value of temperature parameter t, for being greater than the constant of 0;
End temp t
f, corresponding to the end value of temperature parameter t, be initial temperature t
0little constant;
Parameter alpha and β, parameter alpha is the constant in [0,1] scope, β be greater than 0 constant;
Parameter h, corresponds to and produce the columns changed in new explanation S ' process, for being greater than the constant of 1 on the basis of current solution S;
Maximum Markov chain length L_max, for being greater than the integer of 1.
4. method according to claim 1, is characterized in that, described near infrared spectrum data collection comprises calibration set near infrared spectrum Xcal, calibration set chemical score Ycal, forecast set near infrared spectrum Xpre and forecast set chemical score Ypre.
5. method according to claim 4, is characterized in that, the spectrum of described calibration set near infrared spectrum Xcal and calibration set chemical score Ycal, forecast set near infrared spectrum Xpre and forecast set chemical score Ypre and chemical score one_to_one corresponding respectively.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210451289.XA CN102928382B (en) | 2012-11-12 | 2012-11-12 | Near-infrared spectral characteristic wavelength selecting method based on improved simulated annealing algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210451289.XA CN102928382B (en) | 2012-11-12 | 2012-11-12 | Near-infrared spectral characteristic wavelength selecting method based on improved simulated annealing algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102928382A CN102928382A (en) | 2013-02-13 |
| CN102928382B true CN102928382B (en) | 2015-04-22 |
Family
ID=47643234
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210451289.XA Active CN102928382B (en) | 2012-11-12 | 2012-11-12 | Near-infrared spectral characteristic wavelength selecting method based on improved simulated annealing algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102928382B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107219189A (en) * | 2017-07-10 | 2017-09-29 | 天津工业大学 | A kind of Wavelength selecting method for complex sample spectrum |
| CN111125629B (en) * | 2019-12-25 | 2023-04-07 | 温州大学 | Domain-adaptive PLS regression model modeling method |
| CN111553338A (en) * | 2020-04-27 | 2020-08-18 | 南通智能感知研究院 | Hyperspectral feature selection method based on simulated annealing algorithm |
| CN113112031A (en) * | 2021-03-03 | 2021-07-13 | 河北科技大学 | Unmanned aerial vehicle task allocation method based on simulated annealing algorithm |
| CN115097723B (en) * | 2022-08-26 | 2022-11-22 | 南通启锦智能科技有限公司 | PID control parameter adjusting method |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6186397B1 (en) * | 1996-08-26 | 2001-02-13 | Hollandse Signaalapparaten B.V. | Method for operating a fire-control system based on a heuristic algorithm |
| CN101739508A (en) * | 2009-12-23 | 2010-06-16 | 北京邮电大学 | Quantum derivative simulated annealing algorithm adopting arbitrary initial rotation angle step length |
| CN101807228A (en) * | 2010-03-12 | 2010-08-18 | 江苏大学 | Method for selecting subinterval of near-infrared spectrum wavelength based on simulated annealing algorithm |
| CN101806728A (en) * | 2010-03-12 | 2010-08-18 | 江苏大学 | Method for selecting characteristic wavelength of near-infrared spectrum based on simulated annealing algorithm |
| CN101832909A (en) * | 2010-03-12 | 2010-09-15 | 江苏大学 | Selection method for subintervals of near infrared spectral characteristics based on simulated annealing-genetic algorithm |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004030413A (en) * | 2002-06-27 | 2004-01-29 | Fujitsu Ltd | Optimization processing apparatus |
-
2012
- 2012-11-12 CN CN201210451289.XA patent/CN102928382B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6186397B1 (en) * | 1996-08-26 | 2001-02-13 | Hollandse Signaalapparaten B.V. | Method for operating a fire-control system based on a heuristic algorithm |
| CN101739508A (en) * | 2009-12-23 | 2010-06-16 | 北京邮电大学 | Quantum derivative simulated annealing algorithm adopting arbitrary initial rotation angle step length |
| CN101807228A (en) * | 2010-03-12 | 2010-08-18 | 江苏大学 | Method for selecting subinterval of near-infrared spectrum wavelength based on simulated annealing algorithm |
| CN101806728A (en) * | 2010-03-12 | 2010-08-18 | 江苏大学 | Method for selecting characteristic wavelength of near-infrared spectrum based on simulated annealing algorithm |
| CN101832909A (en) * | 2010-03-12 | 2010-09-15 | 江苏大学 | Selection method for subintervals of near infrared spectral characteristics based on simulated annealing-genetic algorithm |
Non-Patent Citations (2)
| Title |
|---|
| 一种用于可见-近红外光谱特征波长选择的新方法;陈孝敬等;《光学学报》;20081130;第28卷(第11期);全文 * |
| 一种近红外光谱特征子区间选择新算法;石吉勇;《光谱学与光谱分析》;20101231;第30卷(第12期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102928382A (en) | 2013-02-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102928382B (en) | Near-infrared spectral characteristic wavelength selecting method based on improved simulated annealing algorithm | |
| Gonzalez-Perez et al. | The host dark matter haloes of [O II] emitters at 0.5< z< 1.5 | |
| Di Luca et al. | Challenges in the quest for added value of regional climate dynamical downscaling | |
| Collard | Selection of IASI channels for use in numerical weather prediction | |
| Ho et al. | Methods for probabilistic modeling of concentrating solar power plants | |
| CN109243172A (en) | Traffic flow forecasting method based on genetic algorithm optimization LSTM neural network | |
| CN103344600B (en) | Characteristic wavelength selecting method for near infrared spectrum in ant colony optimization algorithm | |
| Hertog et al. | Where systems biology meets postharvest | |
| Saha et al. | Deep learning for predicting the monsoon over the homogeneous regions of India | |
| CN103308463A (en) | Characteristic spectrum area selection method for near infrared spectrum | |
| Kunz et al. | Updating a synchronous fluorescence spectroscopic virgin olive oil adulteration calibration to a new geographical region | |
| US20170206468A1 (en) | Value addition dependent data mining techniques for assembly lines | |
| CN105738317A (en) | Textile near-infrared model transfer method | |
| Nandakumar et al. | Effects of the selection function on metallicity trends in spectroscopic surveys of the Milky Way | |
| CN106709588A (en) | Prediction model construction method and equipment and real-time prediction method and equipment | |
| CN103278467A (en) | Rapid nondestructive high-accuracy method with for identifying abundance degree of nitrogen element in plant leaf | |
| CN105158200A (en) | Modeling method capable of improving accuracy of qualitative near-infrared spectroscopic analysis | |
| Ollinaho et al. | Parameter variations in prediction skill optimization at ECMWF | |
| Hernandez et al. | Chemical abundances of two extragalactic young massive clusters | |
| CN107290305B (en) | A kind of near infrared spectrum quantitative modeling method based on integrated study | |
| Laur et al. | TOPz: Photometric redshifts for J-PAS | |
| Kügler et al. | Determining spectroscopic redshifts by using k nearest neighbor regression-I. Description of method and analysis | |
| CN102663439B (en) | Hyperspectral image sequential waveband selection method based on distance calculation | |
| CN113361194A (en) | Sensor drift calibration method based on deep learning, electronic equipment and storage medium | |
| Zibetti et al. | Stellar mass as the ‘glocal’driver of galaxies’ stellar population properties |
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 |