CN106442470A - Coal quality characteristic quantitative analysis method based on LIBS (laser induced breakdown spectrum) and genetic neural network - Google Patents
Coal quality characteristic quantitative analysis method based on LIBS (laser induced breakdown spectrum) and genetic neural network Download PDFInfo
- Publication number
- CN106442470A CN106442470A CN201610793585.6A CN201610793585A CN106442470A CN 106442470 A CN106442470 A CN 106442470A CN 201610793585 A CN201610793585 A CN 201610793585A CN 106442470 A CN106442470 A CN 106442470A
- Authority
- CN
- China
- Prior art keywords
- sample
- coal
- model
- quantitative analysis
- characteristic
- 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.)
- Pending
Links
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
- G01N21/62—Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light
- G01N21/71—Systems in which the material investigated is excited whereby it emits light or causes a change in wavelength of the incident light thermally excited
- G01N21/718—Laser microanalysis, i.e. with formation of sample plasma
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/086—Learning methods using evolutionary algorithms, e.g. genetic algorithms or genetic programming
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Theoretical Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Computing Systems (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Mathematical Physics (AREA)
- General Engineering & Computer Science (AREA)
- Physiology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Optics & Photonics (AREA)
- Plasma & Fusion (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention discloses a coal quality characteristic quantitative analysis method, which is used for monitoring coal quality data of a power plant in real time. The coal quality characteristic quantitative analysis method specifically comprises the following steps of adopting an LIBS (laser induced breakdown spectrum) analysis technique, combining with the nonlinear fitting capability of a BP-ANN (backward propagation-artificial neural network), and optimizing a network structure by an GA (genetic algorithm), so as to finally establish a coal quality characteristic quantitative analysis model. Compared with the traditional calibrating curve quantitative analysis method, the coal quality characteristic quantitative analysis method has the characteristics that the influence by matrix effect is little, the application range of coal variety is wide, the nonlinear suitability is good, and the analysis precision is high.
Description
Technical field
The invention belongs to the quantitative analysis method technical field of LIBS is and in particular to a kind of be based on laser
The coal characteristic quantitative analysis method of induced breakdown spectroscopy and genetic neural network.
Background technology
Coal resources in China rich reserves, distributional region is wide, and ature of coal differs greatly.In the production process of thermal power generation,
The composition of ature of coal and characteristic are run and are had extremely important impact on the safety of power generation, economy and environmental protection.At present, coal
Plant one of changeable main crux having become impact coal-fired power plant unit safety economical operation, therefore, how to realize as-fired coal
The real-time monitoring of matter characteristic becomes the emphasis of person skilled concern.LIBS (Laser-Induced
Breakdown Spectroscopy.LIBS it is) a kind of brand-new analysis means in spectrum analyses field, being attempted in recent years should
For ature of coal quantitative analyses field.Traditional quantitative analysis method has univariate model, partial least square method model, is based on and dominates
Factor combines partial least square method model, Quantitative Analysis Model based on spectrum standardization etc., but due to matrix effect and from
Quantitative analysis results are often brought certain non-linear effects by the presence of absorption effect, so that above-mentioned model is general
Change ability falls flat.
BP neural network (Back Propagation Artificial Neural Network, BP-ANN) be a kind of by
The Multi-layered Feedforward Networks of Back Propagation Algorithm training, can learn and store substantial amounts of input-output mode map relation.The heaviest
Want, BP-ANN has certain non-linear mapping capability, have been demonstrated the energy arbitrary nonlinear function of matching, and have
There is preferable generalization ability.Research shows BP-ANN to be applied in LIBS quantitative analyses and can weaken matrix effect to quantitative analyses
Impact.Publication No. disclosed in 15 days December in 2010 of State Intellectual Property Office is CN101915753A " based on genetic neural network
The LIBS quantitative analysis method of network ", a kind of LIBS element based on BP-ANN of this disclosure of the invention is quantitatively point
Analysis method, and using Genetic Algorithms, the weights of network and threshold value are carried out and optimized, but, this invention have ignored opens up to network
Flutter the optimization of structure.In addition, publication No. disclosed in 6 days April in 2011 of State Intellectual Property Office is " the one of CN 102004088A
Plant the coal characteristic On-line Measuring Method based on neutral net " there is also same problem, not to optimization network topology structure
Clear and definite solution is proposed.Therefore the present invention is directed to the determination usually rule of thumb formula of network topology structure, lacks reason
By the problem of foundation, it is proposed for different prediction objects, needs with genetic algorithm, network topology structure to be optimized.
Content of the invention
It is an object of the invention to, for the problem of existing coal data analytical technology required time length, a kind of base is proposed
Coal characteristic quantitative analysis method in LIBS and genetic neural network.The method utilizes LIBS analyze speed
Fast advantage, in conjunction with the nonlinear fitting ability of BP-ANN, and determines optimum network topological structure using GA global optimizing algorithm,
Set up coal characteristic Quantitative Analysis Model.The method and conventional coal analysis and LIBS tradition calibration curve quantitative analysis method phase
Than having the characteristics that to be affected that little, coal is applied widely, non-linear adaptive is good, analysis precision is high by matrix effect.
For reaching above-mentioned purpose, it is special that the present invention proposes a kind of ature of coal based on LIBS and genetic neural network
Property quantitative analysis method, technical scheme is as follows:
1. select correlated characteristic spectral line:Prediction coal dust calorific value, needs to consider from the following aspects:Pick out and calorific value is had
The element (being included in O, Si, Al, Fe, Ca, the Mg in ash and moisture) of the element (C, H, S) of front contribution and negative contribution
Characteristic spectral line, the also characteristic spectral line of element (K, Ti, Na) influential on matrix effect
2. the pretreatment of model signals, specifically includes following steps:
1) internal standard process is carried out for internal standard spectral line to above-mentioned input feature vector spectral line with C247.86;
2) output calorific value is normalized so as to be between [- 1,1];
3) arrange as input matrix E by following structureinWith output matrix Eout;
4) sample divides, and sample is divided into training set sample, monitoring collection sample and forecast set sample.Wherein train
Collect the structure for model;Monitoring collection, for controlling the absolute error of match value, adjusts the generalization ability of model;Forecast set is considered as
Unknown sample, for predicting ature of coal Characterization Data.
3. build Quantitative Analysis Model, specifically include following steps:
1) using the characteristic spectrum data matrix of training set sample as mode input, with the ature of coal to be measured of training set sample
Data exports as model;And it is optimized for above-mentioned BP-ANN model to be set up using GA, determine optimal net
Network topological structure and transmission function, specifically include following steps:
A. to network topology structure and transfer function by binary coding, be the vector of 12 bits, wherein before
Five represent the first hidden layer neuron number, and being decoded into decimal number is 0-31, and middle 5 represent the second hidden layer neuron
Number, latter two represents the transmission function of each hidden layer respectively, and 0 is logsig, and 1 is tansig, and randomly generates one initially
Population;
B. to monitor the root-mean-square error collecting sample as object function, and it is each individuality point by fitness assignment function
Join fitness;
C. judging whether each individual goal functional value of this generation meets preset value, if there being individual satisfaction, terminating calculating, and
It is decoded as network topology structure and transmission function;If being all unsatisfactory for, carrying out selecting, intersect, make a variation and weight update, and
Repeat step b and c, by that analogy, reach preset value until there being individuality to meet pre-conditioned or genetic algebra, optimum of decoding
Body, obtains network topology structure and transmission function.
2) refinement training GA-BP-ANN model, to monitor the match value of collection with the absolute error of reference value as termination instruction
The condition practiced, when absolute error is less than preset value, deconditioning, obtain the GA-BP-ANN Quantitative Analysis Model determining;
3. predict unknown sample coal characteristic:The coal analysis of the GA-BP-ANN model prediction unknown sample with training
Data, using forecast set characteristic spectrum matrix as mode input, calculates through GA-BP-ANN Quantitative Analysis Model, and will export
Data carry out renormalization, you can realize the prediction of the calorific value of unknown sample.
Brief description
Fig. 1 coal analysis method flow diagram
Fig. 2 genetic algorithm flow chart
Fig. 3 genetic algorithm optimum individual target value tracking figure
Fig. 4 GA-BP-ANN model topology figure
The Quantitative Analysis Model figure of Fig. 5 coal dust calorific value
Specific embodiment
For verifying the effect of the present invention, to predict the higher calorific value Q of coal dust under empty butt stategr,ad(MJ/kg), as a example, enter
Row experimental study, particular content is as follows:
This experiment has 29 samples altogether, gathers spectroscopic data, and is screened, and picks out and has front contribution to calorific value
The characteristic spectral line of element (C, H, S), has the element (O, Si, Al, Fe, Ca, Mg) of negative contribution (ash and moisture) to calorific value
Characteristic spectral line, also the characteristic spectral line of element (K, Ti, Na) influential on matrix effect is as input feature vector light spectrum matrix, coal
The higher calorific value of powder, as output vector, due to the principle of operation of neutral net, needs the form of unified input matrix, to arrange is
Each sample data, the spectroscopic data of behavior difference element.
For input feature vector light spectrum matrix, internal standard process is carried out for internal standard spectral line with C247.86;Number for uniform data
Magnitude, to output analysis of coal nature characteristics data high calorific power, is normalized between [- 1,1], to improve arithmetic speed.
29 samples are divided into training set sample, monitoring collection sample and forecast set sample.Wherein 20 samples of training set (bag
Calorific value containing minimum and maximum), remaining randomly selects monitoring 4 samples of collection, 5 samples of forecast set.
Using the characteristic spectrum data matrix of training set sample as mode input, with the ature of coal number to be measured of training set sample
Export according to as model;Using the GAs Toolbox of University of Sheffield's exploitation, by GA to above-mentioned BP- to be set up
ANN model is optimized (see Fig. 2), determines optimal network topology structure and transmission function, specifically includes following steps:
A. network topology structure and transfer function by binary coding, and each parameter of initial time genetic algorithm, heredity
Algebraically is 50, initial population number of individuals 60, and crossover probability is 0.4, and mutation probability is 0.1, and insertion probability is 0.5 again;
B. with monitor collection sample root-mean-square error RMSE as object function, and by fitness assignment function be every each and every one
Body distributes fitness, and RMSE is less, and individual fitness is bigger, and selected probability is higher;Tracing record when program is run
The desired value (see Fig. 3) of optimum individual;
C. judging whether each individual goal functional value of this generation meets preset value, if there being individual satisfaction, terminating calculating, and
It is decoded as network topology structure and transmission function;If being all unsatisfactory for, carrying out selecting, intersect, make a variation and weight update, and
Repeat step b and c, by that analogy, reach preset value until there being individuality to meet pre-conditioned or genetic algebra, optimum of decoding
Body, obtains network topology structure and transmission function, and the last vector obtaining of this test is [11,000 01,101 1 0], warp
The first hidden layer can be obtained after crossing decoding and comprise 24 neurons, transmission function is tansig, the second hidden layer comprises 13 nerves
Unit, transmission function is logsig, network topology structure such as Fig. 4,
Above-mentioned gained GA-BP-ANN is carried out with refinement training, with training set sample training network, and to monitor the matching of collection
The absolute error of result and reference value, as the standard weighing model generalization ability, arranges 0.5 (MJ/kg), here when all
When the absolute error of monitoring collection is all between [- 0.5,0.5], deconditioning, obtain the GA-BP-ANN quantitative analyses mould determining
Type, and preserve weights and the threshold value of network, result is as shown in Figure 5.According to fitting result, the RMSE=0.5271 of training set;Prison
Survey the RMSE=0.2781 of collection, monitoring collection maximum absolute error is 0.3168, meets pre-conditioned.It can be seen that use the method is set up
Model fitting precision higher.
Using the characteristic spectrum data matrix of forecast set sample as mode input, it is input to GA BP ANN quantitative analyses mould
In type, the data of output is carried out renormalization, you can obtain the coal analysis data of unknown sample, result see table.
By the visible forecast set absolute error of data in table within ± 1.25MJ/kg, in the case that experiment sample is less, use
The coal dust caloric value forecast model that the method is set up still can reach higher precision, therefore it is contemplated that, more there is generation collecting
After the sample of table, the precision of prediction of model can further improve, thus meeting the requirement of person skilled.
The above embodiment of the present invention is only intended to clearly illustrate example of the present invention, and is not to the present invention
Embodiment restriction.For those of ordinary skill in the field, can also make on the basis of the above description
The change of other multi-forms or variation.There is no need to be exhaustive to all of embodiment.All the present invention's
Any modification, equivalent and improvement made within spirit and principle etc., should be included in the protection of the claims in the present invention
Within the scope of.
Claims (7)
1. a kind of coal characteristic quantitative analysis method based on LIBS and genetic neural network it is characterised in that
Comprise the following steps:
1) select correlated characteristic spectral line:Prediction coal dust calorific value, gathers LIBS spectroscopic data, chooses coherent element characteristic spectral line conduct
Mode input;
2) pretreatment of model signals:Internal standard process is carried out to input feature vector spectral line, output calorific value is normalized, and
Sample after processing is divided and is originally divided into training set sample, monitoring collection sample and forecast set sample;
3) build Quantitative Analysis Model:Using the characteristic spectrum data matrix of training set sample as mode input, with training set sample
This calorific value exports as model;And it is optimized for above-mentioned neural network model to be set up using GA, determine the most suitable
The network topology structure closed and transmission function, then carry out refinement training to this model, obtain the GA-BP-ANN quantitation point determining
Analysis model;
4) predict sample coal characteristic:The coal analysis data of the GA-BP-ANN model prediction unknown sample with training, with pre-
The characteristic spectrum matrix surveying collection sample, as mode input, obtains the calorific value of sample.
2. according to claim 1 coal characteristic quantitative analysis method it is characterised in that described step 1) select characteristic spectral line
Need to consider from the following aspects:Prediction coal dust calorific value, picks out the element (C, H, S) having front contribution to calorific value and negatively
The characteristic spectral line of the element (being included in O, Si, Al, Fe, Ca, the Mg in ash and moisture) of contribution, also has shadow to matrix effect
The characteristic spectral line of the element (K, Ti, Na) ringing.
3. according to claim 1 coal characteristic quantitative analysis method it is characterised in that described step 2) model signals pre-
Process includes carrying out internal standard process for internal standard spectral line to above-mentioned input feature vector spectral line with C247.86, meanwhile to sample calorific value
It is normalized so as to be between [- 1,1], and arrange as input matrix E by following structureinWith output matrix Eout;
Sample is divided into training set sample, monitoring collection sample and forecast set sample.Wherein training set is used for the structure of model;Monitoring collection
For controlling the absolute error of match value, adjust the generalization ability of model;Forecast set is considered as unknown sample, special for predicting ature of coal
Property analytical data.
Wherein,Represent the intensity level after the characteristic spectral line internal standard process of j element of i-th sample,Represent i-th sample
Normalized after calorific value.
4. according to claim 1 coal characteristic quantitative analysis method it is characterised in that described step 3) build quantitative analyses
To the topological structure of BP neural network and transfer function by optimization with genetic algorithm during model, comprise the following steps:
A. to network topology structure and transfer function by binary coding, be the vector of 12 bits, wherein first five
Represent the first hidden layer neuron number, being decoded into decimal number is 0-31, and middle 5 represent the second hidden layer neuron number
Mesh, latter two represents the transmission function of each hidden layer respectively, and 0 is logsig, and 1 is tansig, and randomly generates an initial kind
Group;
B. to monitor root-mean-square error RMSE collecting sample as object function, and it is each individuality point by fitness assignment function
Join fitness;
Wherein x 'iFor the match value of i-th monitor sample, xiFor reference value.
C. judge whether every generation individual goal functional value meets preset value, if there being individual satisfaction, terminating calculating, and being decoded as
Network topology structure and transmission function;If being all unsatisfactory for, carrying out selecting, intersect, make a variation and weight update, and repeating to walk
Rapid b and c, by that analogy, reaches preset value until there being individuality to meet pre-conditioned or genetic algebra.
5. according to claim 1 coal characteristic quantitative analysis method it is characterised in that described step 3) model is carried out carefully
When changing training, to monitor the match value of collection and the absolute error of reference value as the condition terminating training, when absolute error is less than
During preset value, deconditioning.
6. according to claim 1 coal characteristic quantitative analysis method it is characterised in that described step 4) prediction sample ature of coal
During characteristic, using forecast set characteristic spectrum matrix as mode input, calculate through GA-BP-ANN Quantitative Analysis Model, and will be defeated
The data going out carries out renormalization, you can realize the prediction of the calorific value of unknown sample.
7. according to claim 1 coal characteristic quantitative analysis method it is characterised in that combustion not only can be set up with the method
Coal calorific value forecast model, can also set up the pre- of Industrial Analysis, elementary analysiss, ash fusion point, combustion characteristics and Slagging Characteristics respectively
Survey model.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610793585.6A CN106442470A (en) | 2016-08-31 | 2016-08-31 | Coal quality characteristic quantitative analysis method based on LIBS (laser induced breakdown spectrum) and genetic neural network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610793585.6A CN106442470A (en) | 2016-08-31 | 2016-08-31 | Coal quality characteristic quantitative analysis method based on LIBS (laser induced breakdown spectrum) and genetic neural network |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106442470A true CN106442470A (en) | 2017-02-22 |
Family
ID=58165157
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610793585.6A Pending CN106442470A (en) | 2016-08-31 | 2016-08-31 | Coal quality characteristic quantitative analysis method based on LIBS (laser induced breakdown spectrum) and genetic neural network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106442470A (en) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107037012A (en) * | 2017-04-05 | 2017-08-11 | 华中科技大学 | The echelle spectrometer dynamic correcting method gathered for LIBS |
CN108444981A (en) * | 2018-01-30 | 2018-08-24 | 中国科学院上海技术物理研究所 | The LIBS quantitative solving methods rebuild based on multiplying property of MART |
CN108627500A (en) * | 2018-04-16 | 2018-10-09 | 华南理工大学 | A kind of solid fuel measurement of caloric value method |
CN109521002A (en) * | 2018-11-29 | 2019-03-26 | 华南理工大学 | A kind of fuel characteristic measurement method of solid fuel particle stream |
CN109521001A (en) * | 2018-11-19 | 2019-03-26 | 华南理工大学 | A kind of flying marking measuring method based on PSO and ε-SVR |
CN109633885A (en) * | 2019-01-29 | 2019-04-16 | 清华大学 | Gradient-index lens micro imaging system and method based on focusing wavefronts |
CN110579467A (en) * | 2018-06-11 | 2019-12-17 | 中国农业机械化科学研究院 | Time-resolved laser-induced breakdown spectroscopy quantification method |
CN111754028A (en) * | 2020-06-08 | 2020-10-09 | 吉林大学 | Hyperspectrum-based coal ash content and moisture detection system and method |
CN113340831A (en) * | 2021-05-10 | 2021-09-03 | 哈尔滨理工大学 | Ultraviolet spectral characteristic analysis and quantitative detection method for yeast and escherichia coli in cow's milk |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6002985A (en) * | 1997-05-06 | 1999-12-14 | Halliburton Energy Services, Inc. | Method of controlling development of an oil or gas reservoir |
CN1912588A (en) * | 2006-08-23 | 2007-02-14 | 太原市海通自动化技术有限公司 | Coal on-line analyse equipment based on laser induced spectral and nerve network technology |
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 |
CN103745101A (en) * | 2013-12-30 | 2014-04-23 | 中冶南方(武汉)自动化有限公司 | Improved neural network algorithm based forecasting method of set value of rolling force of medium plate |
CN104867074A (en) * | 2015-05-15 | 2015-08-26 | 东北大学 | Student comprehensive quality evaluation method based on genetic algorithm optimization BP neural network |
CN104933245A (en) * | 2015-06-15 | 2015-09-23 | 南华大学 | Ship reactor shielding design optimization method based on neural network and genetic algorithm |
CN105631151A (en) * | 2016-01-06 | 2016-06-01 | 陈威宇 | Modeling method for pulverized coal fired boiler combustion optimization |
-
2016
- 2016-08-31 CN CN201610793585.6A patent/CN106442470A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6002985A (en) * | 1997-05-06 | 1999-12-14 | Halliburton Energy Services, Inc. | Method of controlling development of an oil or gas reservoir |
CN1912588A (en) * | 2006-08-23 | 2007-02-14 | 太原市海通自动化技术有限公司 | Coal on-line analyse equipment based on laser induced spectral and nerve network technology |
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 |
CN103745101A (en) * | 2013-12-30 | 2014-04-23 | 中冶南方(武汉)自动化有限公司 | Improved neural network algorithm based forecasting method of set value of rolling force of medium plate |
CN104867074A (en) * | 2015-05-15 | 2015-08-26 | 东北大学 | Student comprehensive quality evaluation method based on genetic algorithm optimization BP neural network |
CN104933245A (en) * | 2015-06-15 | 2015-09-23 | 南华大学 | Ship reactor shielding design optimization method based on neural network and genetic algorithm |
CN105631151A (en) * | 2016-01-06 | 2016-06-01 | 陈威宇 | Modeling method for pulverized coal fired boiler combustion optimization |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107037012A (en) * | 2017-04-05 | 2017-08-11 | 华中科技大学 | The echelle spectrometer dynamic correcting method gathered for LIBS |
CN108444981A (en) * | 2018-01-30 | 2018-08-24 | 中国科学院上海技术物理研究所 | The LIBS quantitative solving methods rebuild based on multiplying property of MART |
CN108444981B (en) * | 2018-01-30 | 2020-06-26 | 中国科学院上海技术物理研究所 | LIBS quantitative solving method based on MART multiplicative reconstruction |
CN108627500A (en) * | 2018-04-16 | 2018-10-09 | 华南理工大学 | A kind of solid fuel measurement of caloric value method |
CN110579467A (en) * | 2018-06-11 | 2019-12-17 | 中国农业机械化科学研究院 | Time-resolved laser-induced breakdown spectroscopy quantification method |
CN110579467B (en) * | 2018-06-11 | 2021-09-10 | 中国农业机械化科学研究院 | Time-resolved laser-induced breakdown spectroscopy quantification method |
CN109521001A (en) * | 2018-11-19 | 2019-03-26 | 华南理工大学 | A kind of flying marking measuring method based on PSO and ε-SVR |
CN109521002A (en) * | 2018-11-29 | 2019-03-26 | 华南理工大学 | A kind of fuel characteristic measurement method of solid fuel particle stream |
CN109521002B (en) * | 2018-11-29 | 2021-05-14 | 华南理工大学 | Fuel characteristic measuring method for solid fuel particle flow |
CN109633885A (en) * | 2019-01-29 | 2019-04-16 | 清华大学 | Gradient-index lens micro imaging system and method based on focusing wavefronts |
CN111754028A (en) * | 2020-06-08 | 2020-10-09 | 吉林大学 | Hyperspectrum-based coal ash content and moisture detection system and method |
CN113340831A (en) * | 2021-05-10 | 2021-09-03 | 哈尔滨理工大学 | Ultraviolet spectral characteristic analysis and quantitative detection method for yeast and escherichia coli in cow's milk |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106442470A (en) | Coal quality characteristic quantitative analysis method based on LIBS (laser induced breakdown spectrum) and genetic neural network | |
CN109828304B (en) | Deep learning-based method for predicting lithologic sequence model by using seismic data | |
CN112906982A (en) | GNN-LSTM combination-based network flow prediction method | |
CN112070356B (en) | Method for predicting carbonization resistance of concrete based on RF-LSSVM model | |
CN108921359A (en) | A kind of distribution gas density prediction technique and device | |
CN106656357B (en) | Power frequency communication channel state evaluation system and method | |
CN104408562A (en) | Photovoltaic system generating efficiency comprehensive evaluation method based on BP (back propagation) neural network | |
CN108595803A (en) | Shale gas well liquid loading pressure prediction method based on recurrent neural network | |
CN104539601B (en) | Dynamic network attack process analysis method for reliability and system | |
CN112765902A (en) | RBF neural network soft measurement modeling method based on TentFWA-GD and application thereof | |
Sakthivel et al. | A genetic algorithm-based artificial neural network model with TOPSIS approach to optimize the engine performance | |
CN108760592A (en) | A kind of unburned carbon in flue dust On-line Measuring Method based on BP neural network | |
CN114862035B (en) | Combined bay water temperature prediction method based on transfer learning | |
CN110245390A (en) | A kind of automobile engine oil consumption prediction technique based on RS-BP neural network | |
CN114266416A (en) | Photovoltaic power generation power short-term prediction method and device based on similar days and storage medium | |
CN112836902B (en) | Improved BP neural network-based coal combustion calorific value prediction method | |
CN111914487B (en) | Data-free regional hydrological parameter calibration method based on antagonistic neural network | |
CN112329269A (en) | Sintering ignition temperature modeling prediction method based on working condition identification | |
Chen et al. | Research on prediction methods of energy consumption data | |
CN111310974A (en) | Short-term water demand prediction method based on GA-ELM | |
CN106529816A (en) | Power line channel adjusting method and system | |
Xu et al. | Research on transmission line project cost forecast method based on BP neural network | |
CN113746813B (en) | Network attack detection system and method based on two-stage learning model | |
CN114065628B (en) | Auxiliary laser protective coating material selection and design method and system | |
CN116451582B (en) | Fire heat release rate measurement system and method based on machine learning fusion model |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
CB02 | Change of applicant information | ||
CB02 | Change of applicant information |
Address after: 2, 210A, 511458, A District, 20 South Ring Road, Nansha District, Guangdong, Guangzhou Applicant after: Guangzhou Bo Bo Energy Technology Co., Ltd. Address before: 510640 Guangdong city of Guangzhou province Tianhe District Wushan South China University of Technology No. 30 building 420 Applicant before: Guangzhou Bo Bo Energy Technology Co., Ltd. |
|
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170222 |