CN105809304A - Method for analyzing correlation of production and operation parameters of power plant and pollution treatment facility - Google Patents
Method for analyzing correlation of production and operation parameters of power plant and pollution treatment facility Download PDFInfo
- Publication number
- CN105809304A CN105809304A CN201410840009.3A CN201410840009A CN105809304A CN 105809304 A CN105809304 A CN 105809304A CN 201410840009 A CN201410840009 A CN 201410840009A CN 105809304 A CN105809304 A CN 105809304A
- Authority
- CN
- China
- Prior art keywords
- data
- correlation
- parameter
- transfinites
- module
- 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
Landscapes
- Testing And Monitoring For Control Systems (AREA)
- Feedback Control In General (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the field of direct or distributed digital control systems for pollutants, and specifically relates to a method for analyzing the correlation of production and operation parameters of a power plant and a pollution treatment facility. The method for analyzing the correlation of production and operation parameters of a power plant and a pollution treatment facility adopts three modules, which are respectively a de-noising module, a long-cycle correlation calculation module and a fine correlation calculation module. The method has the characteristics of high precision, low calculation complexity, high calculation speed and the like, and can meet practical application requirements.
Description
Technical field
The present invention relates to and control system regions for the direct of pollutant or distributed digital loop, be specially a kind of power plant and the method for pollution treatment furnished equipments operational factor correlation analysis.
Background technology
At present, polluter automatically monitor be implementation environment supervision advanced means, there is the characteristics such as automatic, real-time, online, the sewage draining exit Monitoring Data of magnanimity can be provided, make environmental administration can grasp up-to-date discharge of pollutant sources and treatment facility ruuning situation in the very first time, have traditional environment supervision, the unrivaled advantage of monitoring means.Since China proposes the concept of polluter automatic monitored control system, environmental administrations at different levels just present quick growth in the input of association area.The on-line monitoring system that monitoring of environment uses, it is the result data (such as CEMS data and COD data) of monitoring pollution treatment facility mostly, namely by on-the-spot various acquisition means, set up transmission channel, environmental protection facilities of the enterprise operation result is transferred directly to environmental administration, utilizes traditional means to be preserved and analyze.Along with the demand of economic development and construction of environmental protection, the mode of " end monitoring " is difficult to pollution reducing facility carries out effectively monitoring and strong supervision.Online monitoring working conditions system can change the mode that the Pollutant emission concentration of enterprise, discharge capacity only carry out on-line monitoring, achieve from end monitoring to the transformation of whole process supervision, improve verity and the accuracy of data, improve the science of major polluting sources monitoring data and public letter property, strong technical support will be provided to promoting environment supervision level, to finding environmental illegality problem in time, the dynamics of environmental enhancement supervision law enforcement serves positive impetus.But, current on-line monitoring there is a problem in that: one is that management system is unsound.At present, the design of polluter automatic monitoring device and product standard wait to formulate further perfect.Two is that working service is lack of standardization.Though many enterprises have purchased on-line monitoring device, but corollary apparatus and auxiliary equipment do not catch up with, and some enterprises even put together mixed to polluter automatic monitoring device and chemical agent, cause that Online monitoring of pollution sources equipment is corroded.Three is that data reliability is not high.The blowdown of being in fear of to exceed standard of minority enterprise is punished, and just juggles things over the display, allows data fluctuate within the specific limits throughout the year, never exceed standard, differs greatly with the sampling observation Monitoring Data of environmental administration.Along with the automatic monitor and control facility market demand scope of polluter constantly expands, the how further strengthening automatic monitor and control facility operational management of polluter, it is ensured that polluter automatic monitored control system operation stability and data accuracy are more and more important.
Wherein data reliability is not high is the most direct reason restricting the application of polluter automatic monitored control system, it is embodied in shortage of data, data exception, data over run, steady state value, data jump, data validity exception, data dependence exception etc., and data dependence is the most obscure problem being most difficult to find extremely.The theory of existing correlation analysis and method are comprehensive all not, it is impossible to solve the production practical problem of field of Environment Protection.
Summary of the invention
In order to overcome the defect of prior art, it is provided that a kind of can determine that polluter monitors the method for dependency abnormal data in data automatically, the invention discloses a kind of power plant and the method for pollution treatment furnished equipments operational factor correlation analysis.
The present invention reaches goal of the invention by following technical solution:
A kind of method for power plant and pollution treatment furnished equipments operational factor correlation analysis, is characterized in that: be made up of three modules, respectively: noise reduction process module, long period correlation calculations module and the correlation calculations module that becomes more meticulous,
Noise reduction process module: the time series data got is carried out noise reduction process and includes deleting wherein disabled data, the data supplementing disappearance, the noise reduction process that makes time series data more smooth, by noise reduction process module, guarantee that production run data can be used for correlation analysis, and retain the original main characteristic of data；
Long period correlation calculations module: the data obtained are carried out the correlation analysis in the full time period, according to Pearson correlation coefficient formula, calculate the relative coefficient between the parameter index of time series data, and whether the correlation coefficient between the parameter index of review time sequence data is within preset correlation coefficient number scope, so that it is determined that correlation coefficient transfinites parameter in main trend；
The correlation calculations that becomes more meticulous module: after determining the parameter that transfinites, scans data relative fluctuation Amplitude Ratio larger data section, arranges and merges fluctuation data segment, and carries out multinomial relativity method and jointly determine that the dependency of the parameter that transfinites is whether abnormal.
The method of described power plant and pollution treatment furnished equipments operational factor correlation analysis, is characterized in that: correlation analysis relates to two class data: model parameter and time series data,
Model parameter includes unit load, Coal-fired capacity, the total air output of unit, inlet flue gas flow, exiting flue gas flow, booster fan electric current, air-introduced machine electric current；
Time series data is the real time data obtained running from power plant and pollution treatment furnished equipments, and each data point has time tag.
The method of described power plant and pollution treatment furnished equipments operational factor correlation analysis, is characterized in that: described noise reduction process module includes saltus step and checks module, purges inspection module and data noise reduction module:
Saltus step checks module: check the severe degree of the time series data saltus step read, if acutely saltus step frequently occurs the time series data within one period of cycle, in this time period, data cannot do correlation analysis, directly deletes the data in this time period；
Purging and check module: check the time series data read, for there is data jump once in a while in simply a period of time, being referred to as to purge phenomenon, carry out purging and filter, namely the data deletion that saltus step occurs is fallen, and do the data of linear interpolation arithmetic completion disappearance；
Noise reducing of data module: check the severe degree of the time series data saltus step read, if the time series data within one period of cycle there occurs frequently beating of lesser extent, adopting arithmetic rolling average algorithm that the data in this period are carried out noise reduction process, the data after noise reduction process are used for correlation analysis.
The method of described power plant and pollution treatment furnished equipments operational factor correlation analysis, is characterized in that: described long period correlation calculations module includes:
Pearson correlation coefficients matrix: formed Pearson correlation coefficients matrix by model parameter Pearson correlation coefficients between any two, and Pearson correlation coefficients matrix is a symmetrical matrix；
Preset correlation coefficient number scope: by the statistical analysis to normal data sample, obtains the bound of many times of standard deviations of Pearson correlation coefficients average as preset correlation coefficient number scope.
The method of described power plant and pollution treatment furnished equipments operational factor correlation analysis, is characterized in that: described in the correlation calculations module that becomes more meticulous include:
Integration correlation coefficient error rate computing module: calculate the integration correlation coefficient of the parameter that transfinites of current data, and the standard integral correlation coefficient calculated with integration correlation coefficient self learning model is compared, and calculates integration correlation coefficient error rate；
The Pearson correlation coefficients of the parameter that transfinites transfinites accounting computing module again: calculate Pearson correlation coefficients matrix between the model parameter of fluctuation section, and add up the quantity that the Pearson correlation coefficients of parameter transfinites again that transfinites, calculating the Pearson correlation coefficients of the parameter that transfinites transfinites accounting again.
Unusual determination module: determine abnormality decision conditions according to integration correlation coefficient error rate and the Pearson correlation coefficients of parameter of the transfiniting accounting that again transfinites, thus the data of the parameter that judges to transfinite whether to there is dependency abnormal.
The method of described power plant and pollution treatment furnished equipments operational factor correlation analysis, is characterized in that: the many times of standard deviations calculated in preset correlation coefficient number scope are chosen according to experimental result adjustment, generally choose 3 standard deviation scopes.
The method of described power plant and pollution treatment furnished equipments operational factor correlation analysis, is characterized in that: the calculating of integration correlation coefficient will choosing according to principal parameter, principal parameter is most important parameter during power plant and pollution treatment furnished equipments run, and the present invention chooses unit load.
The method of described power plant and pollution treatment furnished equipments operational factor correlation analysis, it is characterized in that: abnormality decision conditions includes two kinds of situations, meet one of them and namely judge the abnormal parameters that transfinites: one be integration correlation coefficient error rate more than 10%, and the Pearson correlation coefficients of the parameter that transfinites transfinites accounting more than 30%；Two be integration correlation coefficient error rate more than 5%, and the Pearson correlation coefficients of the parameter that transfinites transfinites accounting more than 70%.
The method of described power plant and pollution treatment furnished equipments operational factor correlation analysis, is characterized in that: implement successively as follows:
A. initialize system, Boot Model and real-time data base, read in model parameter and time series data；
B. the time series data got is carried out noise reduction process include deleting wherein disabled data, supplement the data of disappearance, smoothing and noise-reducing process, by noise reduction process module, guarantee that production run data can be used for correlation analysis, and retain the original main characteristic of data；
C. according to Pearson correlation coefficient formula, the relative coefficient between the parameter index of real time data is calculated；
D. by historical data statistical analysis, it is determined that parameter index long period strong correlation coefficient data preset range, check that the correlation coefficient between the parameter index of real time data is whether within preset correlation coefficient number scope；
E. capturing correlation coefficient fluctuation and exceed the data segment of preset range, carry out arranging merging, if correlation coefficient is within preset range, then model terminates；
F. calculating fluctuation segment data correlation matrix, by the correlation calculations that becomes more meticulous, the Pearson correlation coefficients of associative multiplication partial correlation coefficient error rate and the parameter that transfinites transfinites accounting again, it is judged that whether the data dependence of the parameter that transfinites is abnormal.
Described model refers to power plant and pollution treatment furnished equipments operational factor correlation analysis, and model parameter includes: the total air output of unit load, Coal-fired capacity, unit, inlet flue gas flow, exiting flue gas flow, booster fan electric current, air-introduced machine electric current.Between parameter index, relation is as follows:
1. load Coal-fired capacity: positive correlation (load becomes big, and coal-fired quantitative change is big)；
2. the total air output of load unit: positive correlation；
3. load inlet flue gas flow: positive correlation；
4. load booster fan electric current: positive correlation；
5. load air-introduced machine electric current: positive correlation；
6. the total air output of Coal-fired capacity unit: positive correlation；
7. Coal-fired capacity inlet flue gas flow: positive correlation；
8. Coal-fired capacity booster fan electric current: positive correlation；
9. Coal-fired capacity air-introduced machine electric current: positive correlation；
10. the total air output inlet flue gas flow of unit: positive correlation；
11. the total air output booster fan electric current of unit: positive correlation；
12. the total air output air-introduced machine electric current of unit: positive correlation；
13. inlet flue gas flowexit flue gas flow: positive correlation, and exiting flue gas flow > inlet flue gas flow；
14. inlet flue gas flow booster fan electric current: positive correlation；
15. inlet flue gas flow air-introduced machine electric current: positive correlation；
16. booster fan electric current blower fan electric current: positive correlation.
The method for power plant and pollution treatment furnished equipments operational factor correlation analysis that the present invention proposes.The method, under the correct premise ensureing data result, has the feature such as precision height, low, the fast operation of algorithm complex, can meet application request.
Accompanying drawing explanation
Fig. 1 is the flow chart in the present invention for power plant and pollution treatment furnished equipments operational factor correlation analysis；
Fig. 2 is integration Calculation of correlation factor method schematic diagram in the present invention.
Detailed description of the invention
The present invention is further illustrated below by way of specific embodiment.
Embodiment 1
A kind of method for power plant and pollution treatment furnished equipments operational factor correlation analysis, is made up of three modules, respectively: noise reduction process module, long period correlation calculations module and the correlation calculations module that becomes more meticulous,
Noise reduction process module: the time series data got is carried out noise reduction process and includes deleting wherein disabled data, the data supplementing disappearance, the noise reduction process that makes time series data more smooth, by noise reduction process module, guarantee that production run data can be used for correlation analysis, and retain the original main characteristic of data；
In the present embodiment, described noise reduction process module includes saltus step and checks module, purges inspection module and data noise reduction module:
Saltus step checks module: check the severe degree of the time series data saltus step read, if acutely saltus step frequently occurs the time series data within one period of cycle, in this time period, data cannot do correlation analysis, directly deletes the data in this time period；
Purging and check module: check the time series data read, for there is data jump once in a while in simply a period of time, being referred to as to purge phenomenon, carry out purging and filter, namely the data deletion that saltus step occurs is fallen, and do the data of linear interpolation arithmetic completion disappearance；
Noise reducing of data module: check the severe degree of the time series data saltus step read, if the time series data within one period of cycle there occurs frequently beating of lesser extent, adopting arithmetic rolling average algorithm that the data in this period are carried out noise reduction process, the data after noise reduction process are used for correlation analysis.
Long period correlation calculations module: the data obtained are carried out the correlation analysis in the full time period, according to Pearson correlation coefficient formula, calculate the relative coefficient between the parameter index of time series data, and whether the correlation coefficient between the parameter index of review time sequence data is within preset correlation coefficient number scope, so that it is determined that correlation coefficient transfinites parameter in main trend；
In the present embodiment, described long period correlation calculations module includes:
Pearson correlation coefficients matrix: formed Pearson correlation coefficients matrix by model parameter Pearson correlation coefficients between any two, and Pearson correlation coefficients matrix is a symmetrical matrix；
Preset correlation coefficient number scope: by the statistical analysis to normal data sample, obtains the bound of many times of standard deviations of Pearson correlation coefficients average as preset correlation coefficient number scope.
The correlation calculations that becomes more meticulous module: after determining the parameter that transfinites, scans data relative fluctuation Amplitude Ratio larger data section, arranges and merges fluctuation data segment, and carries out multinomial relativity method and jointly determine that the dependency of the parameter that transfinites is whether abnormal.
In the present embodiment, described in the correlation calculations module that becomes more meticulous include:
Integration correlation coefficient error rate computing module: calculate the integration correlation coefficient of the parameter that transfinites of current data, and the standard integral correlation coefficient calculated with integration correlation coefficient self learning model is compared, and calculates integration correlation coefficient error rate；
The Pearson correlation coefficients of the parameter that transfinites transfinites accounting computing module again: calculate Pearson correlation coefficients matrix between the model parameter of fluctuation section, and add up the quantity that the Pearson correlation coefficients of parameter transfinites again that transfinites, calculating the Pearson correlation coefficients of the parameter that transfinites transfinites accounting again；
Unusual determination module: determine abnormality decision conditions according to integration correlation coefficient error rate and the Pearson correlation coefficients of parameter of the transfiniting accounting that again transfinites, thus the data of the parameter that judges to transfinite whether to there is dependency abnormal.
The method of described power plant and pollution treatment furnished equipments operational factor correlation analysis, is characterized in that: correlation analysis relates to two class data: model parameter and time series data,
Model parameter includes unit load, Coal-fired capacity, the total air output of unit, inlet flue gas flow, exiting flue gas flow, booster fan electric current, air-introduced machine electric current；
Time series data is the real time data obtained running from power plant and pollution treatment furnished equipments, and each data point has time tag.
The many times of standard deviations calculated in preset correlation coefficient number scope are chosen according to experimental result adjustment, generally choose 3 standard deviation scopes.
The calculating of integration correlation coefficient will choosing according to principal parameter, principal parameter is most important parameter during power plant and pollution treatment furnished equipments run, and the present embodiment chooses unit load.
Abnormality decision conditions includes two kinds of situations, meets one of them and namely judges the abnormal parameters that transfinites: one be integration correlation coefficient error rate more than 10%, and the Pearson correlation coefficients of the parameter that transfinites transfinites accounting more than 30%；Two be integration correlation coefficient error rate more than 5%, and the Pearson correlation coefficients of the parameter that transfinites transfinites accounting more than 70%.
It is sequentially carried out as follows during enforcement:
A. initialize system, Boot Model and real-time data base, read in model parameter and time series data；
B. the time series data got is carried out noise reduction process include deleting wherein disabled data, supplement the data of disappearance, smoothing and noise-reducing process, by noise reduction process module, guarantee that production run data can be used for correlation analysis, and retain the original main characteristic of data；
C. according to Pearson correlation coefficient formula, the relative coefficient between the parameter index of real time data is calculated；
D. by historical data statistical analysis, it is determined that parameter index long period strong correlation coefficient data preset range, check that the correlation coefficient between the parameter index of real time data is whether within preset correlation coefficient number scope；
E. capturing correlation coefficient fluctuation and exceed the data segment of preset range, carry out arranging merging, if correlation coefficient is within preset range, then model terminates；
F. calculating fluctuation segment data correlation matrix, by the correlation calculations that becomes more meticulous, the Pearson correlation coefficients of associative multiplication partial correlation coefficient error rate and the parameter that transfinites transfinites accounting again, it is judged that whether the data dependence of the parameter that transfinites is abnormal.
Described model refers to power plant and pollution treatment furnished equipments operational factor correlation analysis, and model parameter includes: the total air output of unit load, Coal-fired capacity, unit, inlet flue gas flow, exiting flue gas flow, booster fan electric current, air-introduced machine electric current.Between parameter index, relation is as follows:
1. load Coal-fired capacity: positive correlation (load becomes big, and coal-fired quantitative change is big)；
2. the total air output of load unit: positive correlation；
3. load inlet flue gas flow: positive correlation；
4. load booster fan electric current: positive correlation；
5. load air-introduced machine electric current: positive correlation；
6. the total air output of Coal-fired capacity unit: positive correlation；
7. Coal-fired capacity inlet flue gas flow: positive correlation；
8. Coal-fired capacity booster fan electric current: positive correlation；
9. Coal-fired capacity air-introduced machine electric current: positive correlation；
10. the total air output inlet flue gas flow of unit: positive correlation；
11. the total air output booster fan electric current of unit: positive correlation；
12. the total air output air-introduced machine electric current of unit: positive correlation；
13. inlet flue gas flowexit flue gas flow: positive correlation, and exiting flue gas flow > inlet flue gas flow；
14. inlet flue gas flow booster fan electric current: positive correlation；
15. inlet flue gas flow air-introduced machine electric current: positive correlation；
16. booster fan electric current blower fan electric current: positive correlation.
Referring to Fig. 1 and Fig. 2, illustrate the method for power plant and pollution treatment furnished equipments operational factor correlation analysis that the present invention relates to:
Step S201, initial phase.Relating to two class master datas in the present invention, the real time data that a Lei Shi power plant and pollution treatment furnished equipments moving model parameter, an other Lei Shi power plant and pollution treatment furnished equipments produce in running, each data are attribute and unique parameter attribute if having time.At this step, Selection Model parameter { P1, P2, P3..., Pm, the historical data ended in current a period of time, such as the parameter P got according to a sample frequency per second1Time series data { x1, x2, x3..., xn}。
Step S202, sets saltus step round of visits according to time series data sample frequency, calculates the ratio of the degree of fluctuation index index in saltus step round of visits and average fluctuation and average ratio, and computing formula is as follows:
F=| xmax-xmin|,
In formula, xiIt it is data point；
N is the number of point；
F is the amplitude of point；
Being the mean amplitude of tide of point, during calculating, data being divided into length is m some sections put, and seeks its mean amplitude of tide；
Work as index > 0.05 and ratio > 20%, then enter step S203, otherwise, enter step S204.
Step S203, reports to the police to measuring point, and prompting data jump is serious, and deletes all time series datas in this saltus step round of visits from sample.
Step S204, carry out purging saltus step inspection, adopting 1hour to calculate once, fragment of every time fetching data is for this hour and front 1hour as data sample, and the variance carrying out 2hour data calculates, data minor cycle initial setting is 5min, every 1min carries out the mean value computation of front 5min, resets variance scope, if data point or data slot exceed the variance scope of setting, then think that data purge saltus step, in addition it is also necessary to further process.If data jump length is more than 1min, enter step S205, otherwise enter step S205, otherwise enter step S206.
Step S205, rejects saltus step length, and missing data carries out difference operation, completion time series data more than the data in the 1min time period.
Step S206, for the saltus step length time series data section less than 1min, then rejects in the data directly obtained.
Step S207, obtains current data, calculates the arithmetic moving average often organizing time series data, substitutes current data, thus obtaining one group of new data through noise reduction process.Compared with data before, new data is more smooth, and ensure that trend concordance.
Step S208, after noise reduction process, then carries out the calculating of Pearson (Pearson) correlation coefficient r, first carries out the Pearson correlation coefficients between two parameters and calculate, and each parameter does Calculation of correlation factor with other parameters.Then set up the Pearson correlation coefficients matrix between each parameter, as shown in table 1:
Table 1:
Correlation coefficient r computing formula is:
In formula, x, y represent model parameter, rxy∈ [-1,1] represents the Pearson correlation coefficients between parameter x and parameter y.
In table 1: r12It is the Pearson correlation coefficients on the 1st column heading hurdle " unit load " and the 2nd row headers hurdle " Coal-fired capacity ", rijIt is the i-th column heading hurdle and the Pearson correlation coefficients on jth row headers hurdle, by that analogy；
FGD (i.e. Fluegasdesulfurization) represents flue gas desulfurization device；
Strong positive correlation is all presented between selected model parameter.
Step S209, first additionally chooses normal data sample and adds up, and the bound of the average 3sigma of the Pearson correlation coefficients of the normal sample calculated, in this, as the preset range of Pearson correlation coefficients.All present strong positive correlation for the power plant selected by the present invention with between pollution treatment furnished equipments operational factor, be R ∈ (0.67,0.995) through calculating preset range.Whether the value of checking Pearson correlation coefficients matrix is in preset range, if within the scope of this, then judges that the dependence on parameter of all data is normal, carries out step S217；Otherwise, the parameter P that transfinites is recordedi(i > 0), carries out step S210.
Step S210, carries out fluctuation scanning, calculates the data segment that data relative fluctuation is bigger, intercept and arrange the data segment merging fluctuation, to make further correlation analysis time series data.
Step S211, calculates the Pearson correlation coefficients matrix between the parameter of fluctuation section.
Step S212, relative to the Pearson correlation coefficients preset range R of the data after noise reduction process, the preset range of fluctuation section is more loosely designated as R' ∈ (0.6,1), adds up the parameter P that transfinitesiExceed R' quantity k with other parameters at Pearson's coefficient of fluctuation section, calculate the parameter P that transfinitesiPearson correlation coefficients again transfinite accounting λ=k/ (m-1), λ ∈ [0,1] wherein m is the quantity of model parameter.
Step S213, first the present invention chooses unit load as principal parameter, and calculates the data of principal parameter in fluctuation section to time integral α；Secondly the parameter P that transfinites is obtainedi, calculate PiData in fluctuation section to time integral β.Finally calculating the ratio delta of integration α and integration β, δ is the parameter P that transfinitesiAnd the integration correlation coefficient between principal parameter.
Step S214, additionally chooses multiple normal data sample, calculates the parameter P that transfinitesiTo time integral and principal parameter to the fitting formula F between time integral, and by self learning model, constantly correct fitting formula.
Step S215, according to fitting formula F, calculating in principal parameter is integration correlation coefficient during α to time integral value, and result is designated as δ ', and transfinite parameter PiWith integration correlation coefficient error rate (ξ) computing formula is between principal parameter:
Step S216, by checking the Pearson correlation coefficients of parameter of transfiniting again to transfinite accounting λ scope and integration correlation coefficient error rate ξ scope, carries out unusual determination, it is determined that condition is as follows:
1. integration correlation coefficient error rate is more than the 10% of ξ, and the Pearson correlation coefficients of the parameter that transfinites transfinites accounting again more than the 30% of λ；
2. integration correlation coefficient error rate is more than the 5% of ξ, and the Pearson correlation coefficients of the parameter that transfinites transfinites accounting again more than the 70% of λ.
Meet any wherein one i.e. judgement in above-mentioned two: this parameter P that transfinitesiData dependence abnormal；Otherwise, transfinite parameter PiData dependence normal.
Step S217, terminates relation analysis of parameter.
Claims (9)
1., for a method for power plant and pollution treatment furnished equipments operational factor correlation analysis, it is characterized in that: be made up of three modules, respectively: noise reduction process module, long period correlation calculations module and the correlation calculations module that becomes more meticulous,
Noise reduction process module: the time series data got is carried out noise reduction process and includes deleting wherein disabled data, the data supplementing disappearance, the noise reduction process that makes time series data more smooth, by noise reduction process module, guarantee that production run data can be used for correlation analysis, and retain the original main characteristic of data；
Long period correlation calculations module: the data obtained are carried out the correlation analysis in the full time period, according to Pearson correlation coefficient formula, calculate the relative coefficient between the parameter index of time series data, and whether the correlation coefficient between the parameter index of review time sequence data is within preset correlation coefficient number scope, so that it is determined that correlation coefficient transfinites parameter in main trend；
The correlation calculations that becomes more meticulous module: after determining the parameter that transfinites, scans data relative fluctuation Amplitude Ratio larger data section, arranges and merges fluctuation data segment, and carries out multinomial relativity method and jointly determine that the dependency of the parameter that transfinites is whether abnormal.
2. the method for power plant as claimed in claim 1 and pollution treatment furnished equipments operational factor correlation analysis, is characterized in that: correlation analysis relates to two class data: model parameter and time series data,
Model parameter includes unit load, Coal-fired capacity, the total air output of unit, inlet flue gas flow, exiting flue gas flow, booster fan electric current, air-introduced machine electric current；
Time series data is the real time data obtained running from power plant and pollution treatment furnished equipments, and each data point has time tag.
3. the method for power plant as claimed in claim 1 and pollution treatment furnished equipments operational factor correlation analysis, is characterized in that: described noise reduction process module includes saltus step and checks module, purges inspection module and data noise reduction module:
Saltus step checks module: check the severe degree of the time series data saltus step read, if acutely saltus step frequently occurs the time series data within one period of cycle, in this time period, data cannot do correlation analysis, directly deletes the data in this time period；
Purging and check module: check the time series data read, for there is data jump once in a while in simply a period of time, being referred to as to purge phenomenon, carry out purging and filter, namely the data deletion that saltus step occurs is fallen, and do the data of linear interpolation arithmetic completion disappearance；
Noise reducing of data module: check the severe degree of the time series data saltus step read, if the time series data within one period of cycle there occurs frequently beating of lesser extent, adopting arithmetic rolling average algorithm that the data in this period are carried out noise reduction process, the data after noise reduction process are used for correlation analysis.
4. the method for power plant as claimed in claim 1 and pollution treatment furnished equipments operational factor correlation analysis, is characterized in that: described long period correlation calculations module includes:
Pearson correlation coefficients matrix: formed Pearson correlation coefficients matrix by model parameter Pearson correlation coefficients between any two, and Pearson correlation coefficients matrix is a symmetrical matrix；
Preset correlation coefficient number scope: by the statistical analysis to normal data sample, obtains the bound of many times of standard deviations of Pearson correlation coefficients average as preset correlation coefficient number scope.
5. the method for power plant as claimed in claim 1 and pollution treatment furnished equipments operational factor correlation analysis, is characterized in that: described in the correlation calculations module that becomes more meticulous include:
Integration correlation coefficient error rate computing module: calculate the integration correlation coefficient of the parameter that transfinites of current data, and the standard integral correlation coefficient calculated with integration correlation coefficient self learning model is compared, and calculates integration correlation coefficient error rate；
The Pearson correlation coefficients of the parameter that transfinites transfinites accounting computing module again: calculate Pearson correlation coefficients matrix between the model parameter of fluctuation section, and add up the quantity that the Pearson correlation coefficients of parameter transfinites again that transfinites, calculating the Pearson correlation coefficients of the parameter that transfinites transfinites accounting again.
Unusual determination module: determine abnormality decision conditions according to integration correlation coefficient error rate and the Pearson correlation coefficients of parameter of the transfiniting accounting that again transfinites, thus the data of the parameter that judges to transfinite whether to there is dependency abnormal.
6. the method for power plant as claimed in claim 4 and pollution treatment furnished equipments operational factor correlation analysis, is characterized in that: the many times of standard deviations calculated in preset correlation coefficient number scope choose 3 standard deviation scopes.
7. the method for power plant as claimed in claim 5 and pollution treatment furnished equipments operational factor correlation analysis, is characterized in that: the calculating of integration correlation coefficient will choosing according to principal parameter, principal parameter chooses unit load.
8. the method for power plant as claimed in claim 6 and pollution treatment furnished equipments operational factor correlation analysis, it is characterized in that: abnormality decision conditions includes two kinds of situations, meet one of them and namely judge the abnormal parameters that transfinites: one be integration correlation coefficient error rate more than 10%, and the Pearson correlation coefficients of the parameter that transfinites transfinites accounting more than 30%；Two be integration correlation coefficient error rate more than 5%, and the Pearson correlation coefficients of the parameter that transfinites transfinites accounting more than 70%.
9. the method for power plant as claimed in any of claims 1 to 8 in one of claims and pollution treatment furnished equipments operational factor correlation analysis, is characterized in that: implement successively as follows:
A. initialize system, Boot Model and real-time data base, read in model parameter and time series data；
B. the time series data got is carried out noise reduction process include deleting wherein disabled data, supplement the data of disappearance, smoothing and noise-reducing process, by noise reduction process module, guarantee that production run data can be used for correlation analysis, and retain the original main characteristic of data；
C. according to Pearson correlation coefficient formula, the relative coefficient between the parameter index of real time data is calculated；
D. by historical data statistical analysis, it is determined that parameter index long period strong correlation coefficient data preset range, check that the correlation coefficient between the parameter index of real time data is whether within preset correlation coefficient number scope；
E. capturing correlation coefficient fluctuation and exceed the data segment of preset range, carry out arranging merging, if correlation coefficient is within preset range, then model terminates；
F. calculating fluctuation segment data correlation matrix, by the correlation calculations that becomes more meticulous, the Pearson correlation coefficients of associative multiplication partial correlation coefficient error rate and the parameter that transfinites transfinites accounting again, it is judged that whether the data dependence of the parameter that transfinites is abnormal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410840009.3A CN105809304B (en) | 2014-12-27 | 2014-12-27 | Method for analyzing correlation of production operation parameters of power plant and pollution control facility |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410840009.3A CN105809304B (en) | 2014-12-27 | 2014-12-27 | Method for analyzing correlation of production operation parameters of power plant and pollution control facility |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105809304A true CN105809304A (en) | 2016-07-27 |
| CN105809304B CN105809304B (en) | 2021-04-09 |
Family
ID=56979982
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410840009.3A Active CN105809304B (en) | 2014-12-27 | 2014-12-27 | Method for analyzing correlation of production operation parameters of power plant and pollution control facility |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105809304B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110469522A (en) * | 2019-08-13 | 2019-11-19 | 浪潮通用软件有限公司 | A kind of method for detecting abnormality and device of drainage system |
| CN110895404A (en) * | 2018-09-12 | 2020-03-20 | 长鑫存储技术有限公司 | Method and system for automatically detecting correlation between integrated circuit parameters |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8301464B1 (en) * | 2008-07-18 | 2012-10-30 | Cave Consulting Group, Inc. | Method and system for producing statistical analysis of medical care information |
| CN103207944A (en) * | 2013-02-04 | 2013-07-17 | 国家电网公司 | Electric power statistical index relevance analysis method |
| CN103678869A (en) * | 2013-09-17 | 2014-03-26 | 中国人民解放军海军航空工程学院青岛校区 | Prediction and estimation method of flight parameter missing data |
| CN103813355A (en) * | 2014-02-21 | 2014-05-21 | 厦门大学 | Identification method for anomalous points of cooperative synchronization in distributed network |
-
2014
- 2014-12-27 CN CN201410840009.3A patent/CN105809304B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8301464B1 (en) * | 2008-07-18 | 2012-10-30 | Cave Consulting Group, Inc. | Method and system for producing statistical analysis of medical care information |
| CN103207944A (en) * | 2013-02-04 | 2013-07-17 | 国家电网公司 | Electric power statistical index relevance analysis method |
| CN103678869A (en) * | 2013-09-17 | 2014-03-26 | 中国人民解放军海军航空工程学院青岛校区 | Prediction and estimation method of flight parameter missing data |
| CN103813355A (en) * | 2014-02-21 | 2014-05-21 | 厦门大学 | Identification method for anomalous points of cooperative synchronization in distributed network |
Non-Patent Citations (2)
| Title |
|---|
| ABISH MALIK ETC.: "A Correlative Analysis Process in a Visual Analytics Environment", 《2012 IEEE CONFERENCE ON VISUAL ANALYTICS SCIENCE AND TECHNOLOGY (VAST)》 * |
| 王涓 等: "应用皮尔逊相关系数算法查找异常电能表用户", 《电力需求侧管理》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110895404A (en) * | 2018-09-12 | 2020-03-20 | 长鑫存储技术有限公司 | Method and system for automatically detecting correlation between integrated circuit parameters |
| CN110469522A (en) * | 2019-08-13 | 2019-11-19 | 浪潮通用软件有限公司 | A kind of method for detecting abnormality and device of drainage system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105809304B (en) | 2021-04-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106404403B (en) | Method and system for analysis of a turbomachine | |
| CN102063375B (en) | Software reliability assessment method and device based on hybrid testing | |
| CN108204341B (en) | method and device for identifying operating state of wind power plant | |
| CN114912079B (en) | Method, device and equipment for calculating carbon emission of sewage plant and storage medium | |
| CN110094251B (en) | SCR catalyst performance degradation analysis method based on time-interval multi-model modeling | |
| CN113240211B (en) | Method and device for predicting wastewater discharge, computer equipment and storage medium | |
| CN116399818A (en) | Sewage discharge supervision system for chemical enterprises | |
| CN112465239B (en) | Desulfurization system operation optimization method based on improved PSO-FCM algorithm | |
| CN114066262A (en) | Method, system and device for estimating cause-tracing reasoning of abnormal indexes after power grid dispatching and storage medium | |
| CN109783934A (en) | A kind of mean velocity in section fitting rating method based on H-ADCP | |
| CN105809304A (en) | Method for analyzing correlation of production and operation parameters of power plant and pollution treatment facility | |
| CN114858207A (en) | Soft measurement-based gridding source tracing investigation method for drain outlet of river channel | |
| CN109614570B (en) | Method and device for predicting section water quality parameter data | |
| CN111855219A (en) | Diesel engine lubricating oil entering security parameter prediction method based on grey theory | |
| CN105844083A (en) | Method for real-time calculating SCR denitration catalyst optimal exchange time | |
| CN104750144B (en) | Method of early warning operating condition of equipment | |
| CN116432439A (en) | Urban river sewage receiving capacity planning method and system based on numerical simulation | |
| CN116401501A (en) | Dredging operation leakage quantity prediction method and device, electronic equipment and medium | |
| CN114970311A (en) | Method for establishing remote module life prediction model and life prediction method | |
| CN114971164A (en) | Sludge treatment equipment abnormity detection method and system based on artificial intelligence | |
| CN113468813A (en) | Desulfurization system inlet SO2Concentration prediction method and device and electronic equipment | |
| CN117979667B (en) | Circulating cooling water control system, method, equipment and medium for large-scale machine room | |
| Lin et al. | The method for identifying the health state of aircraft hydraulic pump based on grey prediction | |
| Semenova et al. | Research of Statistic Distributions of Nonparametric Goodness-of-Fit Tests by Large Samples | |
| CN116104621B (en) | Nitrogen oxide emission control diagnosis method and device and electronic equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | 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 |