CN116467557A - Accurate comparison method for load data of user power failure event - Google Patents
Accurate comparison method for load data of user power failure event Download PDFInfo
- Publication number
- CN116467557A CN116467557A CN202211595412.5A CN202211595412A CN116467557A CN 116467557 A CN116467557 A CN 116467557A CN 202211595412 A CN202211595412 A CN 202211595412A CN 116467557 A CN116467557 A CN 116467557A
- Authority
- CN
- China
- Prior art keywords
- power failure
- event load
- failure event
- model
- test
- 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
- 238000000034 method Methods 0.000 title claims abstract description 51
- 238000012360 testing method Methods 0.000 claims abstract description 41
- 238000012417 linear regression Methods 0.000 claims abstract description 31
- 238000010586 diagram Methods 0.000 claims abstract description 7
- 230000001419 dependent effect Effects 0.000 claims description 28
- 238000001134 F-test Methods 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 7
- 230000000750 progressive effect Effects 0.000 claims description 4
- 238000004458 analytical method Methods 0.000 abstract description 8
- 230000000694 effects Effects 0.000 abstract description 6
- 230000002411 adverse Effects 0.000 abstract description 3
- 238000013499 data model Methods 0.000 abstract description 3
- 238000012545 processing Methods 0.000 abstract description 3
- 230000008030 elimination Effects 0.000 abstract 1
- 238000003379 elimination reaction Methods 0.000 abstract 1
- 230000002159 abnormal effect Effects 0.000 description 5
- 238000003672 processing method Methods 0.000 description 3
- 238000012795 verification Methods 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000001556 precipitation Methods 0.000 description 2
- 238000000611 regression analysis Methods 0.000 description 2
- 238000007619 statistical method Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 238000003657 Likelihood-ratio test Methods 0.000 description 1
- 238000000540 analysis of variance Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001364 causal effect Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000010998 test method Methods 0.000 description 1
- 238000012956 testing procedure Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/17—Function evaluation by approximation methods, e.g. inter- or extrapolation, smoothing, least mean square method
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Mathematical Physics (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Operations Research (AREA)
- Probability & Statistics with Applications (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Complex Calculations (AREA)
Abstract
The invention discloses a precise comparison method of user power failure event load data, which comprises the following steps of establishing a power failure event load continuous model by using a multiple linear regression method; drawing a standardized residual error histogram; drawing a normalized residual normal P-P diagram; and analyzing the graph to verify the accuracy of the continuous power failure event load model. The invention adopts multiple linear regression, considers the influence of multiple factors on a user power failure event load data model, performs F test on the variables in the equation when introducing one independent variable, eliminates the variables meeting the elimination standard one by one, adds residual image auxiliary analysis and F and R test to ensure the accuracy of the processed user power failure event load data, eliminates adverse effects, adopts errors RES and RComparing the fitting goodness test by using the accurate methodIn the traditional method, various methods are used for ensuring the accuracy of data in the data processing process, and the reliability of data comparison is greatly improved.
Description
Technical Field
The invention relates to the technical field of power failure event load data, in particular to a method for accurately comparing user power failure event load data.
Background
In the process of continuously and rapidly developing social economy, the energy development mode of China is also undergoing transformation, and electric energy occupies an increasingly important position in the life of people. The new situation places higher demands on the power supply capacity of the power enterprises and the quality of the services provided. But in the process of production, construction and transformation of a power supply enterprise, faults and power failure can be inevitably generated, or power failure operation is carried out, so that the normal production and life of power users can be influenced. After a power failure, how to find out the reason of the power failure by accurately comparing the load data of the power failure event of a user becomes a great problem. The traditional user power failure event load data comparison method is simpler, abnormal data is not removed, and the reliability cannot be guaranteed by the data processing method.
Disclosure of Invention
The accurate comparison method of the user power failure event load data solves the problems that the traditional user power failure event load data comparison method provided in the background technology is simple, abnormal data is not removed, and the reliability of the data processing method cannot be guaranteed.
In order to achieve the above purpose, the present invention provides the following technical solutions: a precise comparison method of user power failure event load data comprises the following steps:
s1: establishing a power failure event load continuous model by using a multiple linear regression method;
s2: drawing a standardized residual error histogram;
s3: drawing a normalized residual normal P-P diagram;
s4: and analyzing the graphs obtained in the step S3 and the step S4 to verify the accuracy of the continuous model of the power failure event load in the step S1.
Preferably, the step S1 includes the steps of:
s101: selecting a meteorological factor as a dependent variable;
s102: analyzing factors which are easy to cause a power failure event, and determining two factors from a selected area as independent variables;
s103: analyzing the dependency relationship between the independent variable in the step S102 and the dependent variable in the step S101 by using a progressive method in multiple linear regression of SPSS;
s104: sequentially introducing equations according to the contribution rate of the independent variable to the dependent variable from large to small to obtain a continuous model of the power failure event load;
s105: and (5) checking the accuracy of the power failure event load continuous model.
Preferably, the using step S103 is preceded by determining criteria for selecting the argument.
Preferably, in the step S104, each time an independent variable is introduced, F-test is performed on the variables already in the equation, and the variables that meet the rejection criteria are rejected one by one.
Preferably, the step of F-test is as follows:
s01: firstly, according to the requirements of practical problem, a conclusion is put forward, called the original assumption, and is marked as H 0 ；
S02: based on the information about the sample, for H 0 Is true or false to judge and reject H 0 Or accept H 0 Is a decision of (a).
Preferably, the accuracy of the test equation in S105 is that R is applied 2 And the fitting goodness test is used for testing the fitting degree of the equation to the sample observation value so as to judge the accuracy of the power failure event load continuous model.
Preferably, the accuracy of the test equation in S105 is an error RES test, and the accuracy of the power outage event load duration model is determined by calculating the error RES.
Preferably, the power outage event load duration model in step S104 is as follows:
y＝β 0 +β 1 x 1 +…+β p x p +ε
wherein beta is 0 ,β 1 ,…,β p Is p+1 unknown parameters, ε is an undetectable random error, and usually assume ε N (0, σ) 2 ) Y is dependent variable, x i (i=1, 2, …, p) independent variables.
Preferably, said R 2 The expression for the goodness-of-fit test is as follows:
wherein yi is a regional power failure event load continuous model, y 0i A power outage event load duration model for another region.
Preferably, the load duration model in the step S104 builds a linear regression model as follows
Wherein yi is a regional power failure event load continuous model, y 0i A power outage event load duration model for another region.
Preferably, in the step S104, each time an independent variable is introduced, R test is performed on the variables already in the equation, and the variables that meet the rejection criteria are rejected one by one, where the R test includes the following steps:
s01: firstly, according to the requirements of practical problem, a conclusion is put forward, called the original assumption, and is marked as H 0 ；
S02: when |r| based on the information about the sample>r 1-α When rejecting H 0 The method comprises the steps of carrying out a first treatment on the surface of the Otherwise accept H 0 ，
Wherein r is a complex correlation coefficient, r 1-α The calculation formula of the complex correlation coefficient r is as follows:
wherein x is i 、For the ith argument and the argument average value, y i Is a dependent variable.
The negative correlation coefficient threshold is as follows:
wherein n is the degree of freedom of the sample, F 1-α (1, n-2) is a critical value.
The beneficial effects of the invention are as follows:
adopting multiple linear regression, considering the influence of multiple factors on a user power failure event load data model, performing F test on variables in an equation when introducing one independent variable, removing the variables meeting the removal standard one by one, adding residual image auxiliary analysis and F and R test to ensure the accuracy of the processed user power failure event load data, removing adverse effects, and adopting errors RES and R 2 Compared with the traditional method, the fitting goodness-of-fit testing method ensures the accuracy of data by using a plurality of methods in the data processing process, and greatly improves the reliability of data comparison.
Drawings
FIG. 1 is a flow chart of a method for accurately comparing load data of a user power outage event according to the present invention;
FIG. 2 is a normalized residual histogram for a region;
FIG. 3 is a normalized residual P-P diagram for a region;
FIG. 4 is a region two normalized residual histogram;
FIG. 5 is a normalized P-P diagram of a region two normalized residual.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are some, but not all embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The invention provides a precise comparison method of user power failure event load data, which is described by referring to fig. 1, and comprises the following steps:
s1: establishing a power failure event load continuous model by using a multiple linear regression method;
s2: drawing a standardized residual error histogram;
s3: drawing a normalized residual normal P-P diagram;
s4: and analyzing the graphs obtained in the step S3 and the step S4 to verify the accuracy of the continuous model of the power failure event load in the step S1.
Preferably, the step S1 includes the following steps:
s101: selecting a meteorological factor as a dependent variable;
s102: analyzing factors which are easy to cause a power failure event, and determining two factors from a selected area as independent variables;
s103: analyzing the dependency relationship between the independent variable in the step S102 and the dependent variable in the step S101 by using a progressive method in multiple linear regression of SPSS;
before this step again, criteria for the selected argument are determined.
S104: sequentially introducing equations according to the contribution rate of the independent variable to the dependent variable from large to small to obtain a continuous model of the power failure event load;
in this step, each time an independent variable is introduced, the F-test is performed on the variables already in the equation, and the variables that meet the rejection criteria are rejected one by one.
Wherein, the step of F test is as follows:
s01: firstly, according to the requirements of practical problem, a conclusion is put forward, called the original assumption, and is marked as H 0 ；
S02: based on the information about the sample, for H 0 Is true or false to judge and reject H 0 Or accept H 0 Is a decision of (a).
S105: and (5) checking the accuracy of the power failure event load continuous model.
Further, the accuracy of the test equation in S105 is to apply R 2 And the fitting goodness test is used for testing the fitting degree of the equation to the sample observation value so as to judge the accuracy of the power failure event load continuous model. It is also possible to use an error RES test,and judging the accuracy of the power failure event load continuous model through calculating the error RES.
The power outage event load duration model in step S104 is as follows:
y＝β 0 +β 1 x 1 +…+β p x p +ε
wherein beta is 0 ,β 1 ,…,β p Is p+1 unknown parameters, ε is an undetectable random error, and usually assume ε N (0, σ) 2 ) Y is dependent variable, x i (i=1, 2, …, p) independent variables.
The R is 2 The expression for the goodness-of-fit test is as follows:
wherein yi is a regional power failure event load continuous model, y 0i A power outage event load duration model for another region.
The load duration model in the step S104 builds a linear regression model as follows
Wherein yi is a regional power failure event load continuous model, y 0i A power outage event load duration model for another region.
In the step S104, each time an independent variable is introduced, R test is performed on the variables already in the equation, and the variables meeting the rejection criteria are rejected one by one, where the R test steps are as follows:
s01: firstly, according to the requirements of practical problem, a conclusion is put forward, called the original assumption, and is marked as H 0 ；
S02: when |r| based on the information about the sample>r 1-α When rejecting H 0 The method comprises the steps of carrying out a first treatment on the surface of the Otherwise accept H 0 ，
Wherein r is a complex correlation coefficient, r 1-α For the critical value of the complex correlation coefficient, the calculation of the complex correlation coefficient rThe formula is as follows:
wherein x is i 、For the ith argument and the argument average value, y i Is a dependent variable.
The negative correlation coefficient threshold is as follows:
wherein n is the degree of freedom of the sample, F 1-α (1, n-2) is a critical value.
The detailed principle and flow of the present embodiment are shown below.
First, multiple linear regression model and verification. Taking the distribution data of daily highest load, daily lowest load, daily peak valley difference and daily load rate of user outage events all year around in two areas as an example, the method of the embodiment establishes a load continuous model and curve drawing in the two areas by using a multiple linear regression method, and adds the accuracy of a standardized residual histogram and a standardized residual normal P-P graph auxiliary verification model.
Regression analysis is a widely applied statistical analysis method, which is used for analyzing statistical relationships among things, focusing on observing quantity change rules among variables, describing and reflecting the relationships in a form of a regression equation, and helping people to accurately grasp the influence degree of one or more other variables.
The unary linear regression analysis is a process of analyzing how one factor (independent variable) affects another (dependent variable) under the condition that other influencing factors are excluded or other influencing factors are assumed to be determined, and the analysis is more ideal. In fact, in real-world social life, any one thing (dependent variable) is always affected by a plurality of other things (independent variables).
The regression problem discussed in the unitary linear regression analysis involves only one independent variable, but in practical problems, there are often multiple factors affecting the dependent variable. Therefore, in many cases, it is not sufficient to consider only a single variable, and it is also necessary to examine the relationship of one dependent variable to a plurality of independent variables to obtain a more satisfactory result. This creates a problem in determining the correlation between multiple factors.
The quantitative change relation of two or more independent variables to one dependent variable under the linear correlation condition is studied, and is called multiple linear regression analysis, and a mathematical formula for representing the quantitative relation is called multiple linear regression model.
The multiple linear regression model is an extension of the single linear regression model, and the basic principle is similar to that of the single linear regression model, but is more complex in calculation and is generally completed by means of a computer.
Multiple linear regression model
Let y be an observable random variable that is subject to p nonrandom causal strands x 1 ,x 2 ,…,x p And the influence of a random factor ε, if y and x 1 ,x 2 ,…,x p There is a linear relationship as follows:
y＝β 0 +β 1 x 1 +…+β p x p +ε
wherein beta is 0 ,β 1 ,…,β p Is p+1 unknown parameters, ε is an undetectable random error, and usually assume ε N (0, σ) 2 ) The formula is called a multiple linear regression model, y is called an interpreted variable (dependent variable), x i (i=1, 2, …, p) is an explanatory variable (independent variable).
The present embodiment is based on the multiple linear regression model.
And carrying out multiple linear regression analysis in regression analysis SPSS on the relationship between the daily highest load of the user power failure event and each meteorological factor, and analyzing the dependency change relationship and linear correlation characteristic among a plurality of variables by a statistical method. Here, the dependence of daily maximum load, daily minimum load, daily average load and meteorological factors was analyzed using a stepwise method in multiple linear regression of SPSS. The standard for selecting the independent variable is determined in advance when using a progressive method, the equation only contains constant terms at the beginning, and the equation is selected in sequence from big to small according to the contribution rate of the independent variable to the dependent variable. And each time an independent variable is introduced, F-test is carried out on the variables in the equation, and the variables meeting the rejection standard are rejected one by one. The F-test is a likelihood ratio test reflecting the overall effect on dependent variables.
The principle and flow of the F test are described below:
f test is to verify the fitting degree of the fitting curve, F test is to verify the significance of the whole equation, i.e. to verify whether the parameters in the model are significantly different from zero, and is to be attributed to the assumption H 0 :β 1 ＝0；H 1 :β 1 Test not equal to 0, assume H 0 :β 1 If the value of the linear regression equation is rejected, the regression is obvious, and the relation between y and x is considered to exist, so that the solved linear regression equation is meaningful; otherwise the regression is not significant, the relationship of y to x cannot be described by a regression model, the resulting regression equation is also meaningless, and F is therefore also commonly used as a verification method to assist in verifying the fit equation.
The method used for the test is hypothesis testing in mathematical statistics, the basic task of which is to make a reasonable decision on the hypothesis of certain aspects of the unknown population distribution based on the information provided by the sample. The hypothesis testing procedure is that firstly, a judgment is put forward according to the requirements of the actual problem, which is called the original hypothesis and is marked as H 0 Then according to the information about the sample, to H 0 Is true or false to judge and reject H 0 Or accept H 0 Is a decision of (a). The inverse of the probability nature is based on the principle of small probability events, which is considered to be almost impossible for a small probability event to occur in one test.
We originally hypothesize H 0 An event is constructed next, this event being "assumption H 0 Is a small probability event under the correct condition. Randomly extracting a set of sample observations of capacity n to test the event, if a small probability event is observed in a sampling experiment, indicating that the original assumption is not authentic, because the small probability event that should not occur occurs, and therefore, shouldReject original hypothesis H 0 . Conversely, if the small probability event does not occur, there is no reason to reject the original hypothesis H 0 It should be accepted that the original hypothesis H 0 。
The F test principle is as follows
When H is 0 When the utility model is in the standing state,
in the above formula, F is an analysis of variance value, U is a regression square sum, Q e Is the sum of squares of the residuals and n is the sample degree of freedom.
Wherein the regression square sum U is obtained by:
in the above-mentioned method, the step of,y is the i-th dependent variable and the mean of the dependent variables.
F therefore>F 1-α (1, n-2), reject H 0 Otherwise accept H 0 Wherein F is 1-α (1, n-2) is a critical value.
When H is 0 When the utility model is in the standing state,
in the above formula, T is a normal value, L xx As a function of the variance of the values,t (n-2) is the normal degree of freedom,partial regression coefficient components;
in the above, x i X is the ith argument and the argument average.
The following is a power outage event load duration model for two regions obtained after F-test:
region 1: y is i ＝255.96T av -60.26T max +4567.395
Region 2: y is 0i ＝128.06T min +4.15P-9.89H+3009.524
Wherein the day maximum temperature T max Minimum temperature of day T min Average daily temperature T av Precipitation P, humidity H.
The regression equation can be used to find: for region one, there is a strong correlation between daily maximum load and maximum and average temperatures; that is, the variable that affects the highest daily load the most is the average temperature, followed by the highest temperature; the effect of precipitation, minimum temperature and relative humidity on the daily peak load is minimal. The highest temperature is inversely related to the daily peak load and the average temperature is positively related to the daily peak load.
For the second region, the daily minimum load and the minimum temperature, the relative humidity and the rainfall have strong correlation; that is, the variable that affects the lowest daily load the most is the lowest temperature, and then the relative humidity, the rainfall; the highest and average temperatures have minimal effect on the daily minimum load. The relative humidity is inversely related to the daily minimum load, and the minimum temperature and rainfall are positively related to the daily minimum load.
Fig. 2 is a normalized residual histogram of region one, and fig. 3 is a normalized P-P diagram of normalized residual of region one, and the goodness of fit of the linear regression equation is checked. By observing the normalized residual histogram of region one shown in fig. 2 and the normalized P-P map of the normalized residual of region one shown in fig. 3, it can be found that: since the residuals have a tendency to normally distribute, a regression model is appropriate.
Fig. 4 is a normalized residual histogram of region two, and fig. 5 is a normalized P-P plot of normalized residual of region two, and the linear regression equation goodness of fit is checked. By observing the normalized residual histogram of region two shown in fig. 4 and the normalized P-P map of standard residual of region two shown in fig. 5, it can be found that: since the residuals have a tendency to normally distribute, a regression model is appropriate.
The following table can be derived from MATLAB toolbox
In table 5, the sum of the remaining squares refers to the square of the difference between the fitted curve and the actual value, and the smaller the coefficient, the better the regression of the fitted curve is reflected and the smaller the difference from the actual value is. The correlation coefficient refers to the correlation degree of the fitting curve and the actual value, and the closer the correlation coefficient is to 1, the higher the correlation degree is, and when the correlation degree is more than 0.95, the fitting curve is considered to have correlation with the actual value. The adjustment of the correlation coefficient refers to the correlation coefficient after the exception of the abnormal data, and if the abnormal data does not exist, the adjustment of the correlation coefficient is equal to the value of the correlation coefficient. The residual standard deviation is the open square of the residual square sum, indirectly reflects the regression degree of the fitting curve, and the smaller the residual standard deviation is, the better the regression degree is.
We can clearly see the result, the correlation coefficient r 2 The value is equal to approximately 1, the R value is extremely large, the correlation degree is high, and the residual square sum P<The f value was very small, 0.05, indicating significant regression, and the linear regression equation was considered significant.
The method of the embodiment adopts the errors RES and R 2 And comparing the fitting goodness test method. The calculation formula of the error is as follows:
R 2 the calculation formula of the fitting goodness test is as follows:
the calculated error is 0.023, R 2 The fitting goodness test value is0.93。R 2 And when the regression equation is close to 1, the regression square sum accounts for most of the square sum of the total variation of the dependent variables, the variation of the dependent variables is mainly caused by different values of the independent variables, the regression equation fits sample data points well, the error of the curve is small, and the compared result shows that the load data of the power failure event of the user are similar.
In summary, in this embodiment, multiple linear regression is adopted, the influence of multiple factors on the user outage event load data model is considered, each time an independent variable is introduced, the F test is performed on the variables in the equation, the variables meeting the rejection criteria are removed one by one, the residual map auxiliary analysis and the F and R tests are added to ensure the accuracy of the processed user outage event load data, the adverse effect is eliminated, and the errors RES and R are adopted 2 Compared with the traditional method, the fitting goodness test precision method ensures the accuracy of data by using a plurality of methods in the data processing process, greatly improves the reliability of data comparison, and solves the problems that the traditional user power failure event load data comparison method is simpler, abnormal data is not removed and the data processing method cannot guarantee the reliability.
The above embodiments are only for illustrating the technical solution of the present invention, and not for limiting the same; although the invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some or all of the technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit of the invention, and are intended to be included within the scope of the appended claims and description.
Claims (10)
1. The accurate comparison method of the user power failure event load data is characterized by comprising the following steps of:
s1: establishing a power failure event load continuous model by using a multiple linear regression method;
s2: drawing a standardized residual error histogram;
s3: drawing a normalized residual normal P-P diagram;
s4: and analyzing the graphs obtained in the step S3 and the step S4 to verify the accuracy of the continuous model of the power failure event load in the step S1.
2. The accurate comparison method of user outage event load data according to claim 1, wherein said step S1 comprises the steps of:
s101: selecting a meteorological factor as a dependent variable;
s102: analyzing factors which are easy to cause a power failure event, and determining two factors from a selected area as independent variables;
s103: analyzing the dependency relationship between the independent variable in the step S102 and the dependent variable in the step S101 by using a progressive method in multiple linear regression of SPSS;
s104: sequentially introducing equations according to the contribution rate of the independent variable to the dependent variable from large to small to obtain a continuous model of the power failure event load;
s105: and (5) checking the accuracy of the power failure event load continuous model.
3. The method according to claim 2, wherein the step S103 is preceded by determining criteria for selecting an argument.
4. The method according to claim 2, wherein in step S104, each time an independent variable is introduced, F-test is performed on the variables already in the equation, and the variables that meet the rejection criteria are rejected one by one.
5. The method for accurate comparison of customer outage event load data according to claim 4, wherein said step of F-checking comprises the steps of:
s01: firstly, according to the requirements of practical problem, a conclusion is put forward, called the original assumption, and is marked as H 0 ；
S02: based on the information about the sample, for H 0 Is true or false to judge and reject H 0 Or accept H 0 Is a decision of (a).
6. The method for precisely comparing load data of a blackout event according to claim 2, wherein the accuracy of the model for continuously checking the blackout event in S105 is the application R 2 Fitting goodness test to test the fitting degree of the equation to the sample observation value, thereby judging the accuracy of the power failure event load continuous model, wherein R is as follows 2 The expression for the goodness-of-fit test is as follows:
wherein yi is a regional power failure event load continuous model, y 0i A power outage event load duration model for another region.
7. The method for precisely comparing the user outage event load data according to claim 2, wherein the method for checking the accuracy of the outage event load duration model in S105 is an error RES check, and the accuracy of the outage event load duration model is determined by calculating an error RES, and the calculation formula of the error RES check is as follows:
wherein yi is a regional power failure event load continuous model, y 0i A power outage event load duration model for another region.
8. The accurate comparison method of user outage event load data according to claim 2, wherein said outage event load duration model in step S104 is as follows:
y＝β 0 +β 1 x 1 +…+β p x p +ε
wherein beta is 0 ,β 1 ,…,β p Is p+1 unknown parameters, ε is an undetectable random error, and usually assume ε N (0, σ) 2 ) Y is dependent variable, x i (i=1, 2, …, p) independent variables.
9. The method according to claim 2, wherein the method for checking the accuracy of the check equation in S105 is a method for analyzing a residual map, which is a graph drawn by a result of a difference between an observed value of a dependent variable and a prediction obtained from a continuous model of the outage event load.
10. The accurate comparison method of user outage event load data according to claim 2, wherein in the step S104, each time an independent variable is introduced, R test is performed on the variables already in the equation, the variables meeting the rejection criteria are rejected one by one, and the step of R test is as follows:
s01: firstly, according to the requirements of practical problem, a conclusion is put forward, called the original assumption, and is marked as H 0 ；
S02: when |r| based on the information about the sample>r 1-α When rejecting H 0 The method comprises the steps of carrying out a first treatment on the surface of the Otherwise accept H 0 ，
Wherein r is a complex correlation coefficient, r 1-α The calculation formula of the complex correlation coefficient r is as follows:
wherein x is i 、For the ith argument and the argument average value, y i Is a dependent variable.
The negative correlation coefficient threshold is as follows:
wherein n is the degree of freedom of the sample, F 1-α (1, n-2) is a critical value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211595412.5A CN116467557A (en) | 2022-12-12 | 2022-12-12 | Accurate comparison method for load data of user power failure event |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211595412.5A CN116467557A (en) | 2022-12-12 | 2022-12-12 | Accurate comparison method for load data of user power failure event |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116467557A true CN116467557A (en) | 2023-07-21 |
Family
ID=87172324
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211595412.5A Pending CN116467557A (en) | 2022-12-12 | 2022-12-12 | Accurate comparison method for load data of user power failure event |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116467557A (en) |
-
2022
- 2022-12-12 CN CN202211595412.5A patent/CN116467557A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR20180041174A (en) | Risk Assessment Methods and Systems | |
| CN107168995B (en) | Data processing method and server | |
| CN108388751B (en) | Virtual-real comparison analysis method based on virtual test system and actual system | |
| CN111338972A (en) | Machine learning-based software defect and complexity incidence relation analysis method | |
| CN114839464A (en) | Rapid inspection method, system, equipment and terminal for index life type product | |
| CN107274025B (en) | System and method for realizing intelligent identification and management of power consumption mode | |
| CN113835947B (en) | Method and system for determining abnormality cause based on abnormality recognition result | |
| CN116467557A (en) | Accurate comparison method for load data of user power failure event | |
| CN116883153A (en) | Pedestrian credit investigation-based automobile finance pre-credit rating card development method and terminal | |
| CN107871213B (en) | Transaction behavior evaluation method, device, server and storage medium | |
| TW201135474A (en) | Method for sampling workpiece for inspection and computer program product performing the same | |
| CN113986970A (en) | Method for detecting measurement fee calculation result based on baseline database data | |
| Ahmed et al. | Amputated Life Testing for Weibull-Fréchet Percentiles: Single, Double and Multiple Group Sampling Inspection Plans with Applications | |
| CN114971223A (en) | Index life type sequential test method, system, equipment, medium and terminal | |
| Hanna | Some information measures for testing stochastic models | |
| CN112348685A (en) | Credit scoring method, device, equipment and storage medium | |
| CN112686423A (en) | Method and device for predicting yield of trusted product based on machine learning | |
| CN115587338B (en) | Complex system success rate assessment method based on sequential multi-layer fusion | |
| CN117851266B (en) | Safety key software reliability Bayesian verification method and device | |
| CN116500456B (en) | Ternary lithium battery capacity detection method based on machine learning | |
| Adeyeye et al. | Errors of Misclassification Associated with Edgworth Series Distribution (EDS) | |
| Seifi et al. | Designing different sampling plans based on process capability index | |
| CN114444601A (en) | Feature data item screening method and device and electronic equipment | |
| CN112884167B (en) | Multi-index anomaly detection method based on machine learning and application system thereof | |
| Liu et al. | Developing a new variables sampling plan for products with multiple quality characteristics based on process yield |
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 |