CN115640950A - Method for diagnosing abnormal line loss of distribution network line in active area based on factor analysis - Google Patents
Method for diagnosing abnormal line loss of distribution network line in active area based on factor analysis Download PDFInfo
- Publication number
- CN115640950A CN115640950A CN202210710883.XA CN202210710883A CN115640950A CN 115640950 A CN115640950 A CN 115640950A CN 202210710883 A CN202210710883 A CN 202210710883A CN 115640950 A CN115640950 A CN 115640950A
- Authority
- CN
- China
- Prior art keywords
- factor
- line loss
- line
- calculating
- matrix
- 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
Images
Landscapes
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention relates to an active station area line loss abnormity diagnosis method based on factor analysis, which comprises the following steps: step 1, analyzing factor indexes influencing line loss; step 2, establishing an active transformer area line loss abnormity diagnosis model; and 3, solving the line loss abnormity diagnosis model of the active area established in the step 2, and calculating the influence degree of the factor indexes influencing the line loss in the step 1 on the line loss. The method can accurately position the main reason of the abnormal line loss.
Description
Technical Field
The invention belongs to the technical field of abnormal line loss diagnosis, relates to a method for diagnosing abnormal line loss of distribution network lines in an active area, and particularly relates to a method for diagnosing abnormal line loss of distribution network lines in the active area based on factor analysis.
Background
Line loss is an abbreviation for grid power loss. The method refers to power loss and loss generated in the process of transmitting power, transforming power, distributing power and selling electric energy from a power plant to a power consumer. Specifically, the loss of the functional quantity and the loss of the reactive power generated when the current flows through various power devices of the power grid within a certain time are referred to as the loss of the functional quantity. The line loss rate of the power grid is a comprehensive economic and technical index, and can be used for measuring technical management, operation management, metering management, power utilization management and service management of power supply enterprises. Through data analysis of a metering device, a line load condition and a gate electricity utilization statistical report of a certain provincial line, the line loss rate of a plurality of distribution network lines still has abnormal line loss or even negative line loss at present.
At present, few researches on the mechanism of abnormal line loss caused by distribution networks are carried out. A power distribution network theoretical line loss calculation method based on system clustering and a drosophila optimization support vector machine provides a scientific auxiliary tool for power supply enterprises to quickly evaluate the line loss of a power distribution network. However, the mechanism of abnormal line loss has not been analyzed, and they cannot effectively guide the analysis and processing of abnormal line loss such as negative line loss and high line loss. The gray correlation based on the entropy weight method is used for screening out the 10kV line with high loss, the emphasis is on theoretical analysis, practical application analysis is lacked, abnormal line loss analysis based on multi-system data information interaction can be used for quickly positioning abnormal line loss, but all systems need to be tightly combined, and the implementation difficulty is high.
Through searching, no prior art publication which is the same as or similar to the present invention is found.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provides an active station area line loss abnormity diagnosis method based on factor analysis, which can accurately position the main reason of abnormal line loss.
The invention solves the practical problem by adopting the following technical scheme:
a method for diagnosing line loss abnormity of an active transformer area based on factor analysis comprises the following steps:
step 1, analyzing factor indexes influencing line loss;
furthermore, the factors affecting the line loss in step 1 are: line power factor index, line load factor index, three-phase imbalance rate index, and distribution transformer load rate index
Further, the specific steps of step 2 include:
n samples in the raw data matrix and p variables to be analyzed are X = (X) ij ) n×p The raw data is then normalized as follows:
wherein i =1, \8230;, n, j =1,2, \8230;, p, the normalized matrix is recorded as
The expression of the correlation coefficient matrix R is as follows:
from the eigenvalues λ of the correlation coefficient matrix R i The factor load α can be obtained i ：
Thus, the score F of each factor can be obtained i Wherein l is i Feature vector for R:
load matrix A (alpha) according to factors ij ) And a data matrix X (X) ij ) A specific factor B (. Beta.) can be obtained ij )：
B＝X-A·F (6)
Scoring F according to each factor i And a specific factor beta j The composite score F can be calculated General (1) ：
F General assembly ＝β 1 F 1 +β 2 F 2 +…+β k F k (7)
Further, the specific steps of step 3 include:
(1) Performing an index KMO test, analyzing the relative sizes of simple correlation coefficients and partial correlation coefficients among indexes, and judging whether the data is suitable for factor analysis;
(2) X = (X) for n samples and p variables in the raw data matrix according to equation (2) ij ) n×p Carrying out standardization treatment;
(3) Calculating a correlation coefficient matrix between the variables according to the formula (3);
(4) Searching characteristic value lambda of correlation coefficient matrix i (the feature root is greater than 1) and a feature vector l i (i =1, \8230;, n), and determining the number of factors;
(5) Calculating the factor load alpha according to equation (4) i ；
(6) Calculating the score F of each factor according to the formula (5) i ；
(7) Calculating a special factor B according to the formula (6);
(8) And (4) calculating and sequencing the comprehensive score F according to the formula (7).
Moreover, the specific method in the step (1) of the step 2 comprises the following steps:
calculating a value of a factor index KMO affecting the line loss, the value ranging from 0 to 1; if the value of KMO is less than 0.5, it indicates that there is a difference between the variables. The weaker the correlation between the indices, the less suitable the factor analysis.
The invention has the advantages and beneficial effects that:
1. aiming at the problems in the abnormal line loss processing of the power distribution network, the invention provides a method for diagnosing the line loss abnormality of an active station area based on factor analysis, which comprises a factor analysis model and an abnormal line loss diagnosis model which are suitable for the active station area, analyzes the influence degree of the abnormal factors on the line loss according to historical data, and processes high-loss factors in time.
2. According to the basic operation attribute of the active station area, the abnormal line loss analysis is carried out by calculating the common factor and the special factor, and the method has a remarkable effect of improving the processing capacity of the abnormal line loss of the active station area.
Drawings
FIG. 1 is a table area topology diagram of the present invention;
fig. 2 is a diagram illustrating characteristic values of different influence factors according to the present invention.
Detailed Description
The invention is described in further detail below with reference to the accompanying drawings:
a method for diagnosing line loss abnormity of an active transformer area based on factor analysis comprises the following steps:
step 1, analyzing factor indexes influencing line loss;
the factors influencing the line loss in the step 1 are as follows: line power factor index, line load factor index, three-phase imbalance index and distribution transformer load rate index
In the present embodiment, the line loss is mainly divided into two parts, i.e., a constant line loss and a variable line loss. The constant line loss is only influenced by the voltage of the power system, and the loss change is small. The constant line loss mainly comes from power loss generated by an iron core inside electric equipment such as a motor, a transformer and the like. The variable line loss mainly considers the power grid operation factors influencing the actual line loss of the line, the loss caused by the line and the load borne by the line. Secondly, the compensation effect of the reactive power compensation device, the balance of the three-phase load of the user terminal, the balance of the three-phase load, and the voltage problem caused by the power supply radius all influence the variable loss of the line. Therefore, the influence factors to be considered include a line power factor index, a line load factor index, a three-phase unbalance rate index, a distribution transformer load rate index, and the like.
the specific steps of the step 2 comprise:
n samples and p variables in the raw data matrix are X = (X) ij ) n×p The raw data is then normalized as follows:
wherein i =1, \8230;, n, j =1,2, \8230;, p, the normalized matrix is recorded as
The expression of the correlation coefficient matrix R is as follows:
according to the eigenvalue of the correlation coefficient matrix R, the factor load alpha can be obtained i ：
Thus, the score F of each factor can be obtained i ：
Load matrix A (alpha) according to factors ij ) And a data matrix X (X) ij ) Specific factor B (. Beta.) can be obtained ij )：
B＝X-A·F (6)
Scoring F according to each factor i And a specific factor beta j The composite score F can be calculated General assembly ：
F General assembly ＝β 1 F 1 +β 2 F 2 +…+β k F k (7)
the specific steps of the step 2 comprise:
(1) Performing an index KMO test, analyzing the relative sizes of simple correlation coefficients and partial correlation coefficients among indexes, and judging whether the data is suitable for factor analysis;
the specific method of the step (1) in the step 2 comprises the following steps:
calculating a value of the index KMO, the value ranging from 0 to 1; if the value of KMO is less than 0.5, it indicates that there is a difference between the variables. The weaker the correlation between the indexes, the less suitable the factor analysis;
(2) X = (X) for n samples and p variables in the raw data matrix according to equation (2) ij ) n×p Carrying out standardization treatment;
(3) Calculating a correlation coefficient matrix between the variables according to the formula (3);
(4) Finding eigenvalues lambda of the correlation coefficient matrix i (the feature root is greater than 1) and a feature vector l i (i =1, \8230;, n), and determining the number of factors;
(5) Calculating the factor load alpha according to equation (4) i ；
(6) Calculating the score F of each factor according to the formula (5) i ；
(7) Calculating a special factor B according to the formula (6);
(8) And (4) calculating a comprehensive score F according to the formula (7) and sequencing.
The principle of the factor analysis model of the present invention is as follows:
the core of the factor analysis method is to perform factor analysis on a plurality of comprehensive indexes, extract common factors and then construct a scoring function by using the variance contribution rate of each factor as the sum of weights and the scoring multiplier of the factor. Each variable is composed of common factors and special factors, and the special factors and the common factors should not have relevance.
The mathematical expression for the factorial analysis is a matrix:
vector X (X) where k ≦ p j ) The method is an observable random variable, namely an original observed variable, which is expressed as the measured abnormal line loss in the model, and p is the number of samples. F (F) j ) Is a common factor for X, i.e. the factors appearing in each original observed variable expression are non-observable theoretical variables that are independent of each other. The specific meaning of the common factor must be defined in connection with the actual research problem, and in this model, the common factor representsThe power distribution network in the transformer area generates factors of abnormal line loss, such as power supply voltage, power supply radius, overhead line length and the like. A (alpha) ij ) Is a coefficient of a common factor, named factor load matrix. Factor load alpha ij Is the load of the ith original variable on the jth factor. Alpha is alpha ij Is x j And f j Of (2), i.e. x j And f j Is a correlation coefficient of (a) represents x j And f j The degree of dependence or the correlation coefficient. Alpha (alpha) ("alpha") ij The larger the absolute value of (A), the larger x is represented j Upper common factor f j The greater the load of (a). B (. Beta.) of j ) Is a special factor for X, i.e. no observable variable and no correlation between variables.
The invention is further illustrated by the following specific examples:
in order to verify the effectiveness of the method, 8 factors influencing the line loss are selected, including power supply voltage, common transformer power, special transformer power, power supply radius, power transformer capacity, overhead line length, cable line length and the proportion of power occupied by a common transformer, incomplete data and data with obvious errors are removed by combining line loss theoretical calculation results of 2018, 7, month and 26 days, factor analysis data are formed, 1315 are formed, the assessment index referring to the line loss rate of 0-10 is qualified data, the line loss rate of filtering is smaller than 0 or larger than 10 data, and 783 items are diagnosed. The cell topology is shown in fig. 1.
In analyzing the line loss influence factor, a KMO test and a Bartlett ball test were first performed. The confidence level of the data was selected for factor analysis when the KMO test coefficient >0.5 and the P value of the Bartlett ball test was < 0.05. The KMO value calculated by selecting data is 0.635 and is larger than a threshold value of 0.5, which indicates that correlation exists between variables and meets requirements, and the significance value of the Bartlett ball test result is 0.000 and is smaller than 0.05. This indicates that the data can be used to perform factor analysis.
To calculate the variance of the commonality, each variable may be represented by a commonality. The larger the extracted value, the better the variable expressed in common factors. A requirement of greater than 0.7 is sufficient to indicate that the variable can be expressed in a common factor. In this calculation example, the extracted values are all greater than 0.7, so the variables can be expressed reasonably.
The calculated total variance is shown in table 1, including initial eigenvalues and variance contribution rates for the eight contributing factors. The initial eigenvalues of the first to third components are 2.731, 2.368 and 1.99, respectively, which are all greater than 1, and starting with the fourth component, which are all less than 1. Therefore, the cumulative contribution rate of 88.627% can be obtained by selecting three common factors, namely the three common factors can explain about 89% of the total variance, and the result is ideal. Meanwhile, as can be seen by combining the line chart in fig. 2, the first, second and third common factors have the largest variation, the cumulative contribution rate is 88.627%, and it can be seen that enough original information can be expressed by selecting three common factors from eight variables.
TABLE 1 Total variance
And obtaining a factor score coefficient matrix by adopting a Kaiser normalized maximum variance method. Weighting and summarizing the proportion of the variance contribution rate of each factor in the total variance contribution rate of the three factors as weight to obtain comprehensive score F General assembly ：
And analyzing the influence degree of the abnormal factors on the line loss through comprehensive scoring. The higher the score, the greater the degree of influence. Through the scoring of all factors, the main influence factors of the abnormal wiring can be analyzed, and the key points of the line loss management can be determined in an auxiliary mode.
The following 8 wires were selected for emphasis analysis and the results are shown in table 2.
TABLE 2 results of line analysis
The analysis results are compared with the actual conditions through the table:
the line A, the line B and the line C are affected most by a factor of 1, namely: power supply radius, overhead line length, cable line length. Verification shows that the lengths of the 3 distribution lines have problems, seriously exceed the normal reasonable range and are consistent with the analysis result.
Line D, line E, and line F are most affected by a factor of 2, namely: utility transformer power, transformer capacity, utility transformer power. The proportion of the common transformer in the three lines is proved to be large. The load operation condition is consistent with the analysis result.
The factor 3 has the greatest effect on the G-line and the H-line, i.e. the power supply and the dedicated transformer power supply. Verification shows that the special transformer connected with the two lines has the rheological saturation problem and is consistent with the analysis result.
It should be emphasized that the described embodiments of the present invention are illustrative rather than restrictive, and thus, the present invention is not limited to the described embodiments, but other embodiments, which are derived from the technical solutions of the present invention by those skilled in the art, also belong to the protection scope of the present invention.
Claims (5)
1. A factor analysis-based active transformer area line loss abnormity diagnosis method is characterized by comprising the following steps: the method comprises the following steps:
step 1, analyzing factor indexes influencing line loss;
step 2, establishing an active transformer area line loss abnormity diagnosis model;
and 3, solving the line loss abnormity diagnosis model of the active area established in the step 2, and calculating the influence degree of the factor indexes influencing the line loss in the step 1 on the line loss.
2. The method for diagnosing the line loss abnormality of the active platform area based on the factor analysis according to claim 1, wherein the method comprises the following steps: the factors influencing the line loss in the step 1 are as follows: the system comprises a line power factor index, a line load coefficient index, a three-phase unbalance rate index and a distribution transformer load rate index.
3. The method for diagnosing the line loss abnormality of the active platform area based on the factor analysis as claimed in claim 1, wherein: the specific steps of the step 2 comprise:
n samples in the raw data matrix and p variables to be analyzed are X = (X) ij ) n×p The raw data is then normalized as follows:
wherein i =1, \8230;, n, j =1,2, \8230;, p, the normalized matrix is recorded as
The expression of the correlation coefficient matrix R is as follows:
from the eigenvalues λ of the correlation coefficient matrix R i The factor load α can be obtained i ：
Thus, the score F of each factor can be obtained i In which I i Feature vector for R:
F i ＝α i l i (5)
load matrix A (alpha) according to factors ij ) And a data matrix X (X) ij ) Specific factor B (. Beta.) can be obtained ij )：
B＝X-A·F (6)
Scoring F according to each factor i And a specific factor beta j The composite score F can be calculated General assembly ：
F General assembly ＝β 1 F 1 +β 2 F 2 +…+β k F k (7)
4. The method for diagnosing the line loss abnormality of the active platform area based on the factor analysis according to claim 1, wherein the method comprises the following steps: the specific steps of the step 3 comprise:
(1) Performing an index KMO test, analyzing the relative sizes of simple correlation coefficients and partial correlation coefficients among indexes, and judging whether the data is suitable for factor analysis;
(2) X = (X) for n samples in the raw data matrix and p variables to be analyzed according to equation (2) ij ) n×p Carrying out standardization treatment;
(3) Calculating a correlation coefficient matrix between the variables according to the formula (3);
(4) Searching characteristic value lambda of correlation coefficient matrix i (feature root greater than 1) and feature vector I i (i =1, \8230;, n), and determining the number of factors;
(5) Calculating the factor load alpha according to equation (4) i ；
(6) Calculating the score F of each factor according to the formula (5) i ；
(7) Calculating a special factor B according to the formula (6);
(8) And (4) calculating and sequencing the comprehensive score F according to the formula (7).
5. The method for diagnosing the line loss abnormality of the active platform area based on the factor analysis as claimed in claim 4, wherein: the specific method of the step 3 and the step (1) comprises the following steps:
calculating a value of a factor index KMO affecting the line loss, the value ranging from 0 to 1; if the value of KMO is less than 0.5, it indicates that there is a difference between the variables. The weaker the correlation between the indices, the less suitable the factor analysis.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210710883.XA CN115640950A (en) | 2022-06-22 | 2022-06-22 | Method for diagnosing abnormal line loss of distribution network line in active area based on factor analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210710883.XA CN115640950A (en) | 2022-06-22 | 2022-06-22 | Method for diagnosing abnormal line loss of distribution network line in active area based on factor analysis |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115640950A true CN115640950A (en) | 2023-01-24 |
Family
ID=84939654
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210710883.XA Pending CN115640950A (en) | 2022-06-22 | 2022-06-22 | Method for diagnosing abnormal line loss of distribution network line in active area based on factor analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115640950A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116596199A (en) * | 2023-07-19 | 2023-08-15 | 广东电网有限责任公司 | Superposition analysis method and system for line loss influence factors of power distribution network |
| CN116882766A (en) * | 2023-09-07 | 2023-10-13 | 国网湖北省电力有限公司超高压公司 | Power consumption abnormal distribution risk analysis method and system |
-
2022
- 2022-06-22 CN CN202210710883.XA patent/CN115640950A/en active Pending
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116596199A (en) * | 2023-07-19 | 2023-08-15 | 广东电网有限责任公司 | Superposition analysis method and system for line loss influence factors of power distribution network |
| CN116596199B (en) * | 2023-07-19 | 2024-02-06 | 广东电网有限责任公司 | Superposition analysis method and system for line loss influence factors of power distribution network |
| CN116882766A (en) * | 2023-09-07 | 2023-10-13 | 国网湖北省电力有限公司超高压公司 | Power consumption abnormal distribution risk analysis method and system |
| CN116882766B (en) * | 2023-09-07 | 2023-11-24 | 国网湖北省电力有限公司超高压公司 | Power consumption abnormal distribution risk analysis method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115640950A (en) | Method for diagnosing abnormal line loss of distribution network line in active area based on factor analysis | |
| CN108020752B (en) | Distribution line loss diagnosis method and system based on multi-source through correlation | |
| CN109003000A (en) | A kind of energy quality comprehensive assessment method of active distribution network | |
| CN106208043B (en) | A kind of low and medium voltage distribution network key energy consumption link discrimination method | |
| CN114862229A (en) | Power quality evaluation method and device, computer equipment and storage medium | |
| CN110231503B (en) | High-loss platform area electricity stealing user identification and positioning method based on Glandum causal test | |
| CN111832174A (en) | Wire loss rate processing method and device based on multiple regression | |
| CN112508254B (en) | Method for determining investment prediction data of transformer substation engineering project | |
| CN107274025B (en) | System and method for realizing intelligent identification and management of power consumption mode | |
| CN115879799A (en) | Transformer substation electric energy quality analysis method | |
| CN111626559A (en) | Main factor analysis method-based medium-voltage distribution network line loss key characteristic index extraction method and system | |
| CN111080089A (en) | Method and device for determining critical factors of line loss rate based on random matrix theory | |
| CN116629632A (en) | Comprehensive evaluation method and device for power quality of power distribution network and storage medium | |
| CN115564170A (en) | Reliability assessment method for intelligent power distribution terminal | |
| CN110717725B (en) | Power grid project selection method based on big data analysis | |
| CN114465276A (en) | Power distribution network wind storage combined optimization configuration method and related device thereof | |
| CN113393099A (en) | Power distribution network project group value index evaluation method and device and terminal equipment | |
| CN106886851A (en) | Electric energy meter standby redundancy Forecasting Methodology based on electric energy meter repair based on condition of component | |
| CN112016838A (en) | Method and system for calculating contribution rate of power distribution network energy efficiency index system and terminal equipment | |
| CN111598420A (en) | Regional power quality assessment method considering sensitivity index of main equipment of user | |
| CN110929219A (en) | Distribution network equipment group state evaluation method, device and system and computer storage medium | |
| Sonwane et al. | Algorithms and methodology for development of OCP-PSO tool for optimal capacitor placement to enhance system reliability | |
| CN117350446A (en) | Complex power grid key line identification method based on brittleness theory and combined weighting | |
| CN114677031B (en) | Harmonic wave treatment demand evaluation method based on data envelope analysis | |
| CN115511326A (en) | Grid-connected power supply adjustable performance evaluation method and system |
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 |