CN109861310A - Supercritical thermal power unit primary frequency control system recognizes Variable Selection - Google Patents
Supercritical thermal power unit primary frequency control system recognizes Variable Selection Download PDFInfo
- Publication number
- CN109861310A CN109861310A CN201910152052.3A CN201910152052A CN109861310A CN 109861310 A CN109861310 A CN 109861310A CN 201910152052 A CN201910152052 A CN 201910152052A CN 109861310 A CN109861310 A CN 109861310A
- Authority
- CN
- China
- Prior art keywords
- variable
- data
- primary frequency
- frequency control
- control system
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Abstract
The present invention relates to a kind of supercritical thermal power unit primary frequency control systems to recognize Variable Selection, including chooses suitable variable and data segment by preliminary screening；Variable data is acquired by primary frequency control system upset test, and obtains sample data after pre-processing to data；Using the correlation matrix of data after PCA normalized, characteristic value, the feature vector of correlation matrix are calculated；Data pivot number is judged and chooses, by calculating T2Statistic is monitored multivariable process, finds out contribution of the variable to pivot；Using primary frequency control system output power as reference sequences, dependent variable is associated with angle value as comparing sequence, by what GM method calculated each comparison variable and reference variable；According to each variable to the contribution margin of pivot be associated with angle value variable screened.Compared with prior art, the mentioned method of the present invention reduces the data dimension of mode input variable, reduces System Discrimination scale, improves the precision of identification model, has more objectivity.
Description
Technical field
The present invention relates to supercritical thermal power unit control technology fields, primary more particularly, to a kind of supercritical thermal power unit
Frequency modulation system recognizes Variable Selection.
Background technique
The primary frequency control system of fired power generating unit is first stable of important leverage of system frequency, is referred in the active hair of system
When electricity and supply load mismatch, keep Power Systems control system inclined according to system frequency using manual or automatic mode
The regular active power output for increasing, subtracting corresponding generating set of difference adjusts unit processing according to certain regulations speed in real time
It is whole, with maintain whole system frequency of supply stabilization or region interconnection exchange power in defined variable range, make
Unit load meets continually changing user power utilization demand.The size of Primary frequency control ability and the speed-regulating system of generating set and
The control mode of generating set has extremely important relationship.But people are when carrying out Power System Stability Analysis at present, due to phase
Pass data and data do not carry out real-time update, generally use classical steam turbine control system model, and partial parameters will be adopted
With empirical value or default value, the characteristics of quickly adjusting with the operation and monoblock of large and super-critical unit, to thermal motor
Group participate in peak regulation, it is valley-fill propose higher actual requirement, the system spy that originally simple primary frequency control system model reaction goes out
Property with practical often differ larger, the especially dynamic response characteristic of primary frequency control system is not necessarily complete with practical operation situation
It is complete consistent, keep corresponding power grid dynamic analog and stability analysis calculated result not accurate enough, thus may will affect to power grid
The validity of the directive function of operation itself.The ginsengs such as element, the system in fired power generating unit calculating are established according to measured data thus
Several and model provides reliable basis to optimize the authenticity of governor parameter, raising dynamic power system simulations.
Supercritical thermal power unit primary frequency function it is more to be that control strategy based on boiler follow (BF) comes real
Existing, i.e., the main combination by DEH (steam turbine electro-hydraulic governing system) and CCS (Coordinated Control Systems) corresponding function is come real
It is existing.The variable for influencing primary frequency control system output not only includes the rotational speed difference of steam turbine side, pitch aperture, valve characteristic, high plus pumping
Vapour amount, condensing water flow etc., the variables such as main steam pressure, temperature, reheat steam temperature and boiler calorific requirement of boiler side are also to primary
Adjustment amplitude and the system output of frequency modulation have a significant impact.Therefore, all relatively difficult to the identification difficulty and precision of its model.
Summary of the invention
It is an object of the present invention to overcome the above-mentioned drawbacks of the prior art and provide a kind of supercritical thermal power machines
Group primary frequency control system recognizes Variable Selection.
The purpose of the present invention can be achieved through the following technical solutions:
A kind of supercritical thermal power unit primary frequency control system identification Variable Selection, comprising the following steps:
S1, variable is tentatively chosen；
S2, variable data is acquired by primary frequency control system upset test, and obtains sample after pre-processing to data
Data；
S3, the correlation matrix that sample data is calculated using PCA method, calculate the characteristic value for obtaining correlation matrix
And feature vector, and calculate the contribution rate and contribution rate of accumulative total for obtaining each pivot；
S4, data pivot number is judged and chosen according to contribution rate of accumulative total, by calculating T2Statistic is to multivariable process
It is monitored, obtains each variable to the contribution margin of pivot；
S5, using primary frequency control system output power as reference sequences, dependent variable passes through GM method meter as sequence is compared
That calculates each comparison variable and reference variable is associated with angle value；
S6, according to each variable to the contribution margin of pivot be associated with angle value variable screened, obtain primary frequency modulation system
The leading variable of system changed power, as identification variable.
Further, the pretreatment in the step S2 is to carry out Wavelet Denoising Method, the processing of coarse value, mark change and normalization
Processing.
Further, Wavelet Denoising Method processing the following steps are included:
One S21, selection wavelet function, while initial data is subjected to multilevel wavelet decomposition；
S22, the decomposition coefficient after wavelet decomposition carry out threshold value quantizing processing；
S23, data are reconstructed by wavelet inverse transformation, the data sequence after being denoised.
Further, the formula of data normalization processing are as follows:
Wherein, xminMinimum value before indicating to normalize in data, xmaxMaximum value before indicating to normalize in data,
xgiI-th of data after indicating normalization, x indicate the individual data point to be normalized.
Further, the correlation matrix R in the step S3 are as follows:
Wherein, rijFor the element in correlation matrix, specific formula for calculation are as follows:
Wherein, xiAnd xjFor sample data.
Further, the GM method the following steps are included:
S51, it determines reference sequence and compares ordered series of numbers；
S52, nondimensionalization processing is carried out to data；
S53, the grey incidence coefficient for calculating reference sequence ordered series of numbers compared with；
S54, the degree of association for calculating each relatively ordered series of numbers and reference sequence, wherein the degree of association is bigger to be illustrated to be associated between factor
Degree is bigger.
Further, in step S53, grey incidence coefficient ξi(k) calculation formula are as follows:
Wherein, x0={ x0 (1),x0 (2),…,x0 (n)Be variable reference sequences, ρ is resolution ratio.
Further, in step S54, degree of association biCalculation formula are as follows:
Compared with prior art, the present invention is using PCA (pivot analysis) method and GM (grey correlation analysis) method to overcritical
Unit primary frequency control system Model Distinguish parameter carries out dimensionality reduction, reduces the data dimension of mode input variable, reduces system
Identification scale improves the precision of identification model, has certain directive significance to using scene operation big data modeling, with classics
Analysis on Mechanism modeling compare, it is more objective using this method preferred variable, result is clear.
Detailed description of the invention
Fig. 1 is the process schematic that fired power generating unit primary frequency modulation initializaing variable is chosen；
Fig. 2 is the structural schematic diagram of fired power generating unit primary frequency control system；
Fig. 3 is the flow chart schematic diagram of Wavelet Denoising Method；
Fig. 4 is fired power generating unit primary frequency modulation output power change curve；
Fig. 5 is the fired power generating unit primary frequency modulation output power change curve after Wavelet Denoising Method；
Fig. 6 is PCA schematic illustration；
Fig. 7 is PCA algorithm flow chart；
Fig. 8 is primary frequency modulation raw-data map；
Fig. 9 is grey correlation analysis algorithm flow chart；
Figure 10 is the resulting PCA histogram of embodiment；
Figure 11 is the resulting T of embodiment2Statistical chart；
Figure 12 is each variable obtained by embodiment to the contribution margin histogram of the first pivot；
Figure 13 is the association angle value curve graph that GM analyzes between lower variable when not distinguishing positive and negative sequence in embodiment；
Figure 14 is the association angle value curve graph that GM analyzes between lower variable when distinguishing positive and negative sequence in embodiment.
Specific embodiment
The present invention is described in detail with specific embodiment below in conjunction with the accompanying drawings.The present embodiment is with technical solution of the present invention
Premised on implemented, the detailed implementation method and specific operation process are given, but protection scope of the present invention is not limited to
Following embodiments.
The more control plan being achieved in that based on boiler follow of supercritical thermal power unit primary frequency function
Slightly, i.e., the main combination by DEH (steam turbine electro-hydraulic governing system) and CCS (Coordinated Control Systems) corresponding function is come real
It is existing.As shown in Figure 1, a kind of supercritical thermal power unit primary frequency control system identification Variable Selection is present embodiments provided, it is main
Wanting process is that preliminary primary frequency modulation variable is obtained by Analysis on Mechanism, then by PCA and GM method to the number of preliminary variable
Factually show dimensionality reduction and screening, the specific steps are as follows:
Step 1: carrying out Analysis on Mechanism to fired power generating unit primary frequency control system, the variable of primary frequency control system output is influenced
The not only rotational speed difference including steam turbine side, pitch aperture, valve characteristic, high plus steam extraction amount, the variables such as condensing water flow, boiler side
Main steam pressure, temperature, reheat steam temperature and boiler need the variables such as amount of stored heat also to export to the adjustment amplitude and system of primary frequency modulation
Have a significant impact.As shown in Fig. 2, being fired power generating unit primary frequency control system complete diagram, fired power generating unit one can be found out from figure
The variable that secondary frequency modulation system involves is various, and system structure is complicated, and ω is generator speed, P in figureEFor unit load, PMFor machine
Group mechanical output, PTFor unit main steam pressure, BD is the information interchange of coordination system and boiler, and TD is coordination system and steamer
The information interchange of machine, CV refer to governor and steam turbine information interchange.
Step 2: being pre-processed by primary frequency control system upset test collection site variable data to historical data
After obtain sample data, specifically include:
A1, acquisition primary frequency modulation historical data, using noise present in smoothing method removal data, in the present embodiment,
Wavelet threshold denoising is specifically used, the full details of data can be caught and carry out the frequency analysis of different resolution, performance ratio
Fourier (Fourier) analysis method is more excellent.
Wherein the process of wavelet threshold denoising is as shown in figure 3, step are as follows:
(1) wavelet function is first selected, carries out N layers of wavelet decomposition, decomposition layer such as sym5 function, and by initial data
Number is determined by specific data noise situation, and 3 layers are selected in the present embodiment；
(2) threshold value quantizing processing is carried out to the decomposition coefficient in upper step after wavelet decomposition；
(3) data are reconstructed in wavelet inverse transformation, the data sequence after being denoised.
By taking fired power generating unit primary frequency control system output power as an example, Fig. 4 and Fig. 5 are respectively primary before and after passing through Wavelet Denoising Method
Frequency modulation output power change curve.
A2, zero initialization and normalized, zero initialization process formula are carried out to historical data are as follows:
Wherein, u*(k)、y*(k), u (k), y (k) respectively indicate the value after input data zero initializes, at the beginning of output data zero
The value before value, the initialization of input data zero, the value after the initialization of output data zero after beginningization, N is constant, indicates access evidence
Initial value of the mean value of top N as data；
A3, normalized map the data into [0,1] using maximin, to the formula of data normalized
Are as follows:
Wherein, xminMinimum value before indicating to normalize in data, xmaxMaximum value before indicating to normalize in data,
xgiI-th of data after indicating normalization, x indicate the individual data point to be normalized.
Step 3: calculating the spy of correlation matrix using the correlation matrix of data after PCA method normalized
Value indicative, feature vector, and calculate the contribution rate and contribution rate of accumulative total of each pivot；
Wherein the analysis diagram of PCA is as shown in fig. 6, in Fig. 6, and the left side can indicate original two variables x1, x2, and there are one
Fixed correlativity, and the right then indicates to obtain two orthogonal variables z1, z2 after mapping transformation, and the information of data is main
All on z1, this variations per hour number is just effectively reduced.
Pivot analysis specific flow chart is as shown in fig. 7, the step of PCA are as follows:
B1, the square correlation matrix for calculating the sample after data prediction are as follows:
Wherein, rijCalculation formula are as follows:
X is the data element values after standardization, indicates that each sample point numerical value, p are positive integer.
B2, the characteristic value for calculating R and feature vector and pivot contribution rate；
The calculation formula of characteristic value R is as follows:
The eigenvalue λ of RkFor equation | λ I-R |=0 solution, I are unit matrix；
Wherein, if λkFor k-th of characteristic value of the covariance matrix of X, then the contribution rate of k-th of pivot is defined as:
Step 4: data pivot number is judged and chosen according to contribution rate of accumulative total, by calculating T2Statistic is to multivariable
Process is monitored, and obtains each variable to the contribution margin of pivot；
C1, T is used2The form of statistic is monitored multivariable process, ifMatrix after being standardized for X, for X
In process variableT2The definition of statistic are as follows:
The loss of c2, in order to prevent important information, the contribution rate of accumulative total of k pivot before generally makingIt is greater than
85, select the pivot contribution rate of accumulative total to be 90% in the present invention.
Step 5: using primary frequency control system output power as reference sequences, dependent variable passes through GM as sequence is compared
(grey correlation analysis) method calculates be associated with angle value of each comparison variable with reference variable, grey correlation analysis algorithm flow such as Fig. 8
It is shown.Specific steps are as follows:
D1, it determines reference sequence and compares ordered series of numbers；
D2, nondimensionalization processing is carried out to data；
D3, the grey incidence coefficient for seeking reference sequence ordered series of numbers compared with；
D4, the degree of association b for asking each relatively ordered series of numbers and reference sequencei；
D5, degree of association biIt is bigger illustrate factor between correlation degree it is bigger.
Grey correlation analysis is suitable for judging dynamic variable trend consistency, incidence coefficient ξi(k) calculation formula are as follows:
In formula, x0={ x0 (1),x0 (2),…,x0 (n)Be data reference sequences, analyzed to primary frequency modulation unit
When, reference variable selects primary frequency modulation output power, cries and compares sequence, indicates the factor for influencing system change, selects herein not
With variable as sequence is compared, ρ indicates resolution ratio, belong to be one (0,1) real number, occurrence depends on the circumstances, generally
When less than 0.5463, the resolving effect of algorithm is best, ρ=0.5 herein.
The degree of association of each factor and reference sequences is calculated separately out according to following formula:
Step 6: screening by PCA and GM to primary frequency control system upset test parameter, primary frequency control system is found out
The leading factor of changed power realizes the overcritical selection for crossing motor group primary frequency control system identification variable.Through the above steps
Obtain PCA pivot histogram, variable to the contribution curve and data of pivot each variable sequence compared with after grey correlation analysis
Fired power generating unit primary frequency control system identification leading factor can be obtained in degree of association curve between column.
In order to which the fired power generating unit primary frequency control system Variable Selection for verifying proposed is effective, Jiangxi power plant is acquired
The primary frequency control system data of 660MW unit select total valve instruction, frequency difference, high pitch aperture, first stage pressure, hot repressing
Power, middle row pressure power, main steam pressure, main steam temperature, unit function power are as initial auxiliary variable, and all initial data are as schemed
Shown in 8.
PCA algorithm flow chart obtains the variance contribution ratio of each pivot and adds up contribution as shown in figure 9, after PCA is analyzed
Rate such as table 1, PCA pivot distribution histogram is as shown in Figure 10, by table 1 and Figure 10 it is found that comparing in R matrix there are two characteristic root
Greatly, respectively 6.5678 and 1.1803, when pivot number is 2, contribution rate of accumulative total has reached 98%, wherein first pivot can
To explain the variation of given total data 82.1%, the first two pivot contains all information of initial data substantially, certain
Required precision under (generally greater than 85%), the first, second pivot can be used to replace the original input data matrix of system,
To reduce the dimension of input data, the complexity of system model is reduced.
The variance contribution ratio and contribution rate of accumulative total of each pivot of table 1
Figure 11 and Figure 12 is the T drawn when primary frequency modulation unit output power changes greatly respectively2Count spirogram and every
A auxiliary variable changes the case where size of made contribution margin to the power of the assembling unit.Pass through pivot analysis histogram, each pivot variance
Contribution rate and each auxiliary variable are combined to the contribution plot of the first pivot, then with unit primary frequency modulation operating experience, final choice
5 variables: total valve aperture, first stage pressure, reheat pressure, main steam pressure, middle row pressure power are primary as unit is influenced
The leading factor of frequency modulation power output.
Using primary frequency modulation unit output power data sequence as reference sequences, total valve instruction, frequency difference, high pitch are opened
Degree, valve value of feedback, first stage pressure, reheat pressure, middle row pressure power, main steam pressure Variables Sequence are respectively to compare sequence,
It is handled using grey correlation analysis described in the present invention (GM), the degree of association curve between obtained comparative quantity and reference variable
As shown in Figure 13 and Figure 14.
From Analysis on Mechanism it is known that main steam pressure variation and reference sequences variation tendency are on the contrary, so if it is desired to directly
The influence degree size to reference sequences is directly found out from degree of association curve, and positive and negative sequence differentiation, figure need to be carried out to data sequence
13 indicate is the size for not carrying out the association angle value of positive and negative sequence differentiation to variable, Figure 14 reflection be by main steam pressure and
Its dependent variable did the association angle value size of each variable after positive and negative sequence is distinguished.Comparing two kinds of situations can be seen that, although different
Processing mode is different, the association angle value between each variable also have it is biggish not, when not considering positive and negative sequence, main steam pressure and machine
The association angle value of group power be it is the smallest, the variation tendency that can also regard variable 8 and reference sequences as is least consistent (may be at
Negative correlation), therefore it is defeated to fired power generating unit primary frequency modulation to combine the pivot analysis of positive and negative sequence and front that can clearly obtain
The factor that power is affected out is main are as follows: total valve aperture, first stage pressure, reheat pressure, main steam pressure, middle row pressure
Power, this is consistent with pivot analysis result.
Under two kinds of processing obtained comparison variables and reference variable to be associated with angle value as shown in table 2:
Angle value is associated between each variable of 2 association analysis of table and reference variable
In conclusion using the method for the present invention can relatively directly objectively to heat power engineering system image parameter data carry out dimensionality reduction and
Screening has more preferable more direct effect than Analysis on Mechanism, in smart grid big data era, to making full use of data information to have
There is highly important realistic meaning.
The preferred embodiment of the present invention has been described in detail above.It should be appreciated that those skilled in the art without
It needs creative work according to the present invention can conceive and makes many modifications and variations.Therefore, all technologies in the art
Personnel are available by logical analysis, reasoning, or a limited experiment on the basis of existing technology under this invention's idea
Technical solution, all should be within the scope of protection determined by the claims.
Claims (8)
1. a kind of supercritical thermal power unit primary frequency control system recognizes Variable Selection, which comprises the following steps:
S1, variable is tentatively chosen；
S2, variable data is acquired by primary frequency control system upset test, and obtains sample data after pre-processing to data；
S3, the correlation matrix that sample data is calculated using PCA method, calculate the characteristic value for obtaining correlation matrix and spy
Vector is levied, and calculates the contribution rate and contribution rate of accumulative total for obtaining each pivot；
S4, data pivot number is judged and chosen according to contribution rate of accumulative total, by calculating T2Statistic supervises multivariable process
It surveys, obtains each variable to the contribution margin of pivot；
S5, using primary frequency control system output power as reference sequences, dependent variable is calculated each as sequence is compared by GM method
Comparison variable is associated with angle value with reference variable；
S6, according to each variable to the contribution margin of pivot be associated with angle value variable screened, obtain primary frequency control system function
The leading variable of rate variation, as identification variable.
2. supercritical thermal power unit primary frequency control system according to claim 1 recognizes Variable Selection, feature exists
In the pretreatment in the step S2 is to carry out Wavelet Denoising Method, the processing of coarse value, mark change and normalized.
3. supercritical thermal power unit primary frequency control system according to claim 2 recognizes Variable Selection, feature exists
In the processing of, Wavelet Denoising Method the following steps are included:
One S21, selection wavelet function, while initial data is subjected to multilevel wavelet decomposition；
S22, the decomposition coefficient after wavelet decomposition carry out threshold value quantizing processing；
S23, data are reconstructed by wavelet inverse transformation, the data sequence after being denoised.
4. supercritical thermal power unit primary frequency control system according to claim 2 recognizes Variable Selection, feature exists
In the formula of data normalization processing are as follows:
Wherein, xminMinimum value before indicating to normalize in data, xmaxMaximum value before indicating to normalize in data, xgiTable
I-th of data after showing normalization, x indicate the individual data point to be normalized.
5. supercritical thermal power unit primary frequency control system according to claim 1 recognizes Variable Selection, feature exists
In correlation matrix R in the step S3 are as follows:
Wherein, rijFor the element in correlation matrix, specific formula for calculation are as follows:
Wherein, xiAnd xjFor sample data.
6. supercritical thermal power unit primary frequency control system according to claim 1 recognizes Variable Selection, feature exists
In, the GM method the following steps are included:
S51, it determines reference sequence and compares ordered series of numbers；
S52, nondimensionalization processing is carried out to data；
S53, the grey incidence coefficient for calculating reference sequence ordered series of numbers compared with；
S54, the degree of association for calculating each relatively ordered series of numbers and reference sequence, wherein the bigger correlation degree between illustrating factor of the degree of association
It is bigger.
7. supercritical thermal power unit primary frequency control system according to claim 6 recognizes Variable Selection, feature exists
In, in step S53, grey incidence coefficient ξi(k) calculation formula are as follows:
Wherein, x0={ x0 (1),x0 (2),…,x0 (n)Be variable reference sequences, ρ is resolution ratio.
8. supercritical thermal power unit primary frequency control system according to claim 7 recognizes Variable Selection, feature exists
In, in step S54, degree of association biCalculation formula are as follows:
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910152052.3A CN109861310B (en) | 2019-02-28 | 2019-02-28 | Identification variable selection method for primary frequency modulation system of supercritical thermal power generating unit |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910152052.3A CN109861310B (en) | 2019-02-28 | 2019-02-28 | Identification variable selection method for primary frequency modulation system of supercritical thermal power generating unit |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109861310A true CN109861310A (en) | 2019-06-07 |
| CN109861310B CN109861310B (en) | 2020-12-22 |
Family
ID=66899345
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910152052.3A Active CN109861310B (en) | 2019-02-28 | 2019-02-28 | Identification variable selection method for primary frequency modulation system of supercritical thermal power generating unit |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109861310B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110397479A (en) * | 2019-07-18 | 2019-11-01 | 西安西热节能技术有限公司 | A kind of multivariable collaboration frequency modulation peak regulation system and method |
| CN112149903A (en) * | 2020-09-22 | 2020-12-29 | 内蒙古电力(集团)有限责任公司内蒙古电力科学研究院分公司 | Primary frequency modulation analysis and optimization method of thermal power generating unit based on BP neural network algorithm |
| CN112329271A (en) * | 2020-12-04 | 2021-02-05 | 国网山东省电力公司电力科学研究院 | Thermal power generating unit peak regulation key index identification method and device based on multiple PCAs |
| CN113391575A (en) * | 2020-03-13 | 2021-09-14 | 新疆金风科技股份有限公司 | Method and device for identifying online state of variable pitch actuating mechanism of wind generating set |
| CN114999581A (en) * | 2022-06-13 | 2022-09-02 | 华东交通大学 | Time lag identification method and system for rare earth extraction separation process |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103324139A (en) * | 2013-06-07 | 2013-09-25 | 华中科技大学 | Numerical control machine tool milling tool breakage monitoring method |
| CN103544399A (en) * | 2013-11-04 | 2014-01-29 | 牛丽仙 | Building electricity energy-saving management method based on grey correlation |
| CN104807039A (en) * | 2015-04-23 | 2015-07-29 | 华北电力大学 | Variable dimensionality reduction modeling method for boiler combustion optimization |
| CN104951851A (en) * | 2015-07-08 | 2015-09-30 | 华侨大学 | Wind turbine state prediction model establishing method based on grey relation-regression SVM (support vector machine) |
| CN106372755A (en) * | 2016-08-31 | 2017-02-01 | 国家电网公司 | BP neural network intelligent industrial park energy consumption model establishment method based on principal component analysis |
| CN106597311A (en) * | 2016-12-21 | 2017-04-26 | 哈尔滨工业大学 | Fused HI equivalence lithium ion battery degradation prediction method based on principal component analysis |
-
2019
- 2019-02-28 CN CN201910152052.3A patent/CN109861310B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103324139A (en) * | 2013-06-07 | 2013-09-25 | 华中科技大学 | Numerical control machine tool milling tool breakage monitoring method |
| CN103544399A (en) * | 2013-11-04 | 2014-01-29 | 牛丽仙 | Building electricity energy-saving management method based on grey correlation |
| CN104807039A (en) * | 2015-04-23 | 2015-07-29 | 华北电力大学 | Variable dimensionality reduction modeling method for boiler combustion optimization |
| CN104951851A (en) * | 2015-07-08 | 2015-09-30 | 华侨大学 | Wind turbine state prediction model establishing method based on grey relation-regression SVM (support vector machine) |
| CN106372755A (en) * | 2016-08-31 | 2017-02-01 | 国家电网公司 | BP neural network intelligent industrial park energy consumption model establishment method based on principal component analysis |
| CN106597311A (en) * | 2016-12-21 | 2017-04-26 | 哈尔滨工业大学 | Fused HI equivalence lithium ion battery degradation prediction method based on principal component analysis |
Non-Patent Citations (3)
| Title |
|---|
| 徐晗: "陕西省干旱时空演变特征分析及脆弱性评价", 《中国博士学位论文全文数据库 基础科学辑》 * |
| 袁世通: "1000MW超超临界机组建模理论与方法的研究", 《中国博士学位论文全文数据库 工程科技II辑》 * |
| 贺化: "《中国知识产权区域布局理论与政策机制》", 31 July 2017, 知识产权出版社 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110397479A (en) * | 2019-07-18 | 2019-11-01 | 西安西热节能技术有限公司 | A kind of multivariable collaboration frequency modulation peak regulation system and method |
| CN113391575A (en) * | 2020-03-13 | 2021-09-14 | 新疆金风科技股份有限公司 | Method and device for identifying online state of variable pitch actuating mechanism of wind generating set |
| CN112149903A (en) * | 2020-09-22 | 2020-12-29 | 内蒙古电力(集团)有限责任公司内蒙古电力科学研究院分公司 | Primary frequency modulation analysis and optimization method of thermal power generating unit based on BP neural network algorithm |
| CN112329271A (en) * | 2020-12-04 | 2021-02-05 | 国网山东省电力公司电力科学研究院 | Thermal power generating unit peak regulation key index identification method and device based on multiple PCAs |
| CN114999581A (en) * | 2022-06-13 | 2022-09-02 | 华东交通大学 | Time lag identification method and system for rare earth extraction separation process |
| CN114999581B (en) * | 2022-06-13 | 2023-11-10 | 华东交通大学 | Time lag identification method and system for rare earth extraction and separation process |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109861310B (en) | 2020-12-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109861310A (en) | Supercritical thermal power unit primary frequency control system recognizes Variable Selection | |
| CN109491358B (en) | Control performance monitoring method for boiler dynamic information of million-kilowatt ultra-supercritical unit | |
| CN110285403A (en) | Main Steam Temperature Control method based on controlled parameter prediction | |
| CN108647809A (en) | A kind of exhaust enthalpy of turbine real-time computing technique based on least square method supporting vector machine | |
| CN106773666A (en) | A kind of model parameter acquisition methods for presurized water reactor primary Ioops system | |
| CN110532674A (en) | A kind of coal-fired power station boiler fire box temperature measurement method | |
| CN112377985B (en) | Heat supply unit peak regulation lower limit evaluation method and system based on feedback neural network | |
| CN111555363B (en) | AGC real-time control strategy based on deep learning under big data environment | |
| US20230152786A1 (en) | Industrial equipment operation, maintenance and optimization method and system based on complex network model | |
| CN109992921A (en) | A kind of online soft sensor method and system of the coal-fired plant boiler thermal efficiency | |
| CN112186761A (en) | Wind power scene generation method and system based on probability distribution | |
| CN107831652A (en) | A kind of unit load intelligent optimized control method based on cold end system energy storage | |
| Baghsheikhi et al. | Real-time exergoeconomic optimization of a steam power plant using a soft computing-fuzzy inference system | |
| CN110515304A (en) | Overheating steam temperature PID forecast Control Algorithm based on ARX-Laguerre function model | |
| Wang et al. | Big data analytics for price forecasting in smart grids | |
| CN112288157A (en) | Wind power plant power prediction method based on fuzzy clustering and deep reinforcement learning | |
| CN112821420B (en) | XGboost-based prediction method and system for dynamic damping factor and multidimensional frequency index in ASFR model | |
| Shi et al. | Modelling and diagnosis of feedback-controlled processes using dynamic PCA and neural networks | |
| CN104865944B (en) | Gas separation unit control system performance estimating method based on PCA LSSVM | |
| Zhou et al. | A very short-term probabilistic prediction method of wind speed based on ALASSO-nonlinear quantile regression and integrated criterion | |
| CN102967393B (en) | Supercritical unit boiler separator outlet temperature on-line calibration method | |
| CN115014454A (en) | Soft measurement method, system, equipment and medium for main steam flow of thermal power generating unit | |
| CN116029433A (en) | Energy efficiency reference value judging method, system, equipment and medium based on grey prediction | |
| CN114565209A (en) | Process industry energy consumption state evaluation method based on clustering | |
| CN112531725A (en) | Method and system for identifying parameters of static var generator |
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 |