CN107527114A - A kind of circuit taiwan area exception analysis method based on big data - Google Patents
A kind of circuit taiwan area exception analysis method based on big data Download PDFInfo
- Publication number
- CN107527114A CN107527114A CN201710675256.6A CN201710675256A CN107527114A CN 107527114 A CN107527114 A CN 107527114A CN 201710675256 A CN201710675256 A CN 201710675256A CN 107527114 A CN107527114 A CN 107527114A
- Authority
- CN
- China
- Prior art keywords
- line
- area
- transformer
- data
- loss
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
- G06Q50/06—Electricity, gas or water supply
Abstract
A kind of circuit taiwan area exception analysis method based on big data, it is characterised in that including：Table bottom and file data are extracted from magnanimity platform and line loss database；Calculate low-voltage customer, specially become user and taiwan area electricity sales amount；Result of calculation and high pressure are uploaded in Guo Wang general headquarters unstructured database by the unstructured transmission channel of data center；By in table bottom, archives and computation model deposit general headquarters distributed file system；Calculate circuit, the delivery and line loss per unit of taiwan area；Carry out line loss analyzing；Generate diagnosis report.The present invention can differentiate exactly occurs the reason for abnormal, improves the robustness and fault-tolerance of power network.
Description
Technical Field
The invention relates to the field of electric power automation, in particular to a line transformer area abnormity analysis method based on big data.
Background
Under the background of constructing a national grid company integrated electric quantity and line loss management platform, a power grid line loss abnormity analysis and decision system is provided as a high-level application module. The main function of the application module is to analyze causes of various abnormal line loss by qualitatively and quantitatively analyzing power grid production data, find out key points of problems and provide a preliminary processing suggestion, so that the whole data analysis process has practical significance.
The analysis of the line loss abnormity of the whole network is a very complex system engineering, and the analysis point can be divided into a macroscopic aspect and a microscopic aspect. The macro analysis is particularly implemented by examining line loss unit data, and mainly concerns the problems of line and plane. The measurement points with problems are positioned by observing and analyzing the change conditions of data of a transformer substation, a line, a transformer area and the like, and the process is from macro to micro; the micro analysis is particularly implemented by examining the data of the metering points, and is concerned about the problem of points. And (4) finding out the gates, special transformer users and low-voltage users with problems by analyzing the data of the metering points.
In the face of massive power data, the traditional computing framework cannot be competent for such complicated computing work. Therefore, the use of big data technology becomes an effective solution.
Disclosure of Invention
In order to solve the above problems, the present invention provides a method for analyzing an anomaly of a line block area of big data, which is characterized by comprising:
step 100: extracting table bottom and file data from the mass platform and the line loss database, and storing the table bottom and the file data into the database;
step 200: calculating the electricity sale quantity of low-voltage users, special transformer users and distribution areas according to the data of the meter bottoms and the archives in the database;
step 300: after the national network province company calculates and finishes the low-voltage users, the special transformer users and the distribution area electricity sales, the calculation result and the high voltage are uploaded to an unstructured database of the national network headquarters through an unstructured transmission channel of a data center;
step 400: extracting data of the unstructured database of the national network headquarters, and storing a table bottom, a file and a calculation model into a headquarter distributed file system;
step 500: calculating the power supply quantity and the line loss rate of the line and the transformer area according to the table bottom, the file and the calculation model obtained in the step 400;
step 600, selecting at least one line loss analysis mode to perform line loss analysis; and generating a diagnosis report according to the results of the one or more line loss analysis modes.
The method can solve the problem of transverse fusion of various professional data related to line loss, simultaneously responds to the concurrent calculation requirements of ten million levels of data quantity, accurately analyzes the circuit and transformer area abnormity of the national power grid on the basis, assists a leader in making a decision, improves the working efficiency, and creates great economic and social values.
Drawings
FIG. 1 is a flow chart of a big data based line platform anomaly analysis method of the present invention.
Fig. 2 is a diagram of the cumulative electric power ratio of the line for 5 users as an example;
fig. 3 is a table area accumulated electric quantity ratio chart taking 6 users as an example;
FIG. 4 is a schematic diagram of a method for analyzing anomalies in a line area based on a random forest according to the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below, and it is obvious that the described embodiments are a part of the embodiments of the present invention, but not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In the description of the present invention, it is to be understood that the terms "longitudinal", "lateral", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc., indicate orientations or positional relationships based on those shown in the drawings, and are used merely for convenience in describing the present invention, and do not indicate or imply that the device or element so referred to must have a particular orientation, be constructed and operated in a particular orientation, and thus, are not to be construed as limiting the present invention.
In the present invention, the following basic concepts are involved:
the amount of power supplied: the power supply quantity refers to all input power quantities of production activities of power supply enterprises, and comprises the following power quantities: (1) power plant online electricity quantity: the method refers to the recorded online electric quantity of the local area overall power plant (independent power generation company, directly-owned power plant and local power plant). (2) outsourcing electric quantity: refers to the amount of power purchased by each power supply company from a power grid outside the company's power supply area. (3) inputting and outputting electric quantity by the adjacent network: refers to the amount of power supplied to each other between regional grids.
The power supply quantity calculation formula is as follows:
the power supply quantity = power plant online power quantity + purchased power quantity + adjacent network input power quantity-adjacent network output power quantity.
Selling electric quantity: the electric quantity sold to users by the electric power enterprise and the electric quantity supplied by the electric power enterprise for non-electric power production are indicated.
Line loss: line loss or grid loss refers to energy loss dissipated in the form of heat energy, namely active power consumed by resistance and conductance. The line loss of the power system mainly comprises two parts, namely management line loss and technical line loss. The management line loss is also called controllable line loss, and is mainly caused by poor power grid structure, power grid operation and power grid management. The invention mainly discusses the technical line loss, which is also called theoretical line loss, the size of the line loss is mainly determined by the load condition of a power grid, power supply equipment and parameters of a power supply circuit, theoretically, the line loss cannot be avoided, and only measures can be taken to reduce the loss, so that the loss is also called uncontrollable loss, and is normal and reasonable electric energy consumption.
Line loss rate: the line loss rate of the power network refers to the percentage of line loss electricity quantity in power supply quantity, and is called the line loss rate for short, and the calculation formula is
Line loss rate = (line loss electric quantity ÷ power supply quantity) × 100%
= (power supply-electricity selling amount) ÷ power supply amount × 100%
In the qualitative analysis business activity of the line loss abnormity, the line loss abnormity mainly comprises the following components: the electric quantity is unusual, and the line loss is unusual, and the archives are unusual, and the operation is unusual, gathers unusually.
The abnormal electric quantity is mainly characterized in that the change condition of the electric quantity of the metering point is analyzed by continuously examining the meter counting value of the metering point so as to determine whether the sudden increase, sudden decrease or sudden proportion change condition of the electric quantity exists.
The line loss abnormity is mainly characterized in that the possible cause of the line loss abnormity is analyzed by calculating the line loss indexes of the examination objects (lines, transformer areas and the like) and combining some associated information and reference data, and then qualitative judgment is made.
The abnormal files mainly expose the problems in management, which are embodied in that the file information of some assessment objects is inconsistent with the real information of the production environment.
The abnormal operation is mainly realized by comparing and analyzing data information fed back by the acquisition system, and an assessment object with an operation state deviating from a normal level is found, such as: and the gate is in voltage loss and phase loss, and the current of the gate meter is overloaded.
The abnormal collection is mainly to find the problem of the collection communication system through the analysis of the success rate of data collection when the examination object runs.
Correlation coefficient:
the correlation coefficient is a statistical index for reflecting the degree of closeness of the correlation between the variables and the correlation direction thereof, and the correlation degree and the correlation direction of the line loss fluctuation and the power fluctuation of the distribution area can be found out by using the property of the correlation coefficient, so that the loss reduction efficiency is improved. Correlation coefficient properties:
correlation coefficient ρ XY The value is between-1 and 1;
when rho XY When =0, X and Y are said to be irrelevant;
when | ρ XY When | =1, X and Y are called to be completely related, and at this time, a linear functional relationship exists between X and Y;
when | ρ XY |&1, the partial variation of Y caused by the variation of X, rho XY The larger the absolute value of (c), the larger the fluctuation of Y caused by the fluctuation of X, | ρ XY |>, 0.8 is called highly correlated, when 0.5<|ρ XY |&0.8 is called significant correlation, when 0.3<| ρ XY |&0.5, the correlation becomes low, and when | ρ |, the correlation becomes low XY |&And lt, 0.3, is called no correlation.
As shown in fig. 1, an embodiment of the present invention discloses a line loss rate calculation method for a line and a station area based on big data, which is characterized by including:
step 100: extracting table bottom and file data from the mass platform and the line loss database, and storing the table bottom and the file data into the database;
step 200: calculating the electricity selling quantity of low-voltage users, special transformer users and distribution areas according to the table bottom and the archive data in the database;
step 300: after the national network province company calculates and finishes the low-voltage users, the special transformer users and the distribution area electricity sales, the calculation result and the high voltage are uploaded to an unstructured database of the national network headquarters through an unstructured transmission channel of a data center;
step 400: extracting data of the unstructured database of the national network headquarters, and storing a table bottom, a file and a calculation model into a headquarter distributed file system;
step 500: calculating the power supply quantity and the line loss rate of the line and the distribution room according to the table bottom, the file and the calculation model obtained in the step 400;
step 600, selecting at least one line loss analysis mode to perform line loss analysis, wherein the line loss analysis mode comprises a first line loss analysis mode, a second line loss analysis mode and a third line loss analysis mode; and generating a diagnosis report according to the results of the one or more line loss analysis modes. If a line loss analysis mode is selected, directly taking a line loss analysis result as a diagnosis report; if multiple line loss analysis modes are selected, then the diagnostic report is regenerated according to the various analysis mode weights.
Further, the first line loss analysis mode is a line and platform area diagnosis mode with zero power supply, comprising
Step A100, obtaining all lines and transformer area data with zero power supply quantity;
step A200, detecting whether the acquisition of the line and station power supply side meter fails;
specifically, the values of the power supply side meters of the line and the station area are detected, and if one or more of the following conditions occur: and if the upper surface bottom is empty, the lower surface bottom is empty, the upper surface bottom and the lower surface bottom are empty, the lower surface bottom is smaller than the upper surface bottom, and no power exists in the surface bottom, the power supply side meter is considered to fail to collect.
Step A300, detecting whether a power supply side meter is powered off or not;
specifically, the power supply side meters of the line and the platform area are detected to count the values of the upper and lower bottoms, and if the values of the upper and lower bottoms are the same, no power is discharged.
Step A400, detecting the configuration condition of the circuit and distribution area archive model;
specifically, the method for detecting the configuration condition of the line archive model comprises the following steps: diagnosing the configuration of the line profile model of the line if one or more of the following occurs: the method comprises the following steps that if the relation between a switch and a measuring point is lost, the relation between a measuring point and an electric energy meter is lost, the multiplying power of a power supply measuring point is empty, the reading date of the power supply measuring point is empty, the data source of the power supply measuring point is empty, and the effective date of the electric energy meter is empty, the configuration of a circuit file model is considered to be wrong;
the method for detecting the configuration condition of the platform area archive model comprises the following steps: and diagnosing the configuration condition of the zone file model, and if the information of the zone file metering points is inconsistent with the information of the zone model metering points, determining that the configuration of the zone file model is wrong.
And step A500, generating a zero power supply amount diagnosis report of the line and the station area.
Further, the second line loss analysis mode is a line and station diagnosis mode of retail electric quantity, and comprises
Step B100, acquiring all lines and distribution room data with zero electricity sales;
step B200, detecting the line-to-line transformation relation of the line and the operation and distribution transformation relation of the transformer area;
specifically, the method for detecting the line variation relationship of the line includes: if there is no public-private change under the line, the line is empty;
the method for detecting the operation and distribution transformation relation of the transformer area comprises the following steps: if no low-voltage user exists under the platform area, the platform area is empty;
step B300, detecting whether the public and private changes of the line and the user meter of the transformer area fail to collect;
specifically, the method for detecting whether the public-private transformer of the line fails to be collected is as follows: detecting the value of the common variable table meter if one or more of the following conditions occur: if the station area has no master table, the high-voltage users have no master table, zero degree users and the meters have no numerical values (including that the upper surface bottom is empty, the lower surface bottom is empty, the upper surface bottom and the lower surface bottom are empty, the lower surface bottom is smaller than the upper surface bottom and the meter bottom has no electric quantity), the public and private transformer acquisition of the line is considered to be failed;
the method for detecting whether the acquisition of the user meter of the distribution room fails comprises the following steps: detecting the value of the user meter, if one or more of the following conditions occur: if the user does not have a summary table, the user has a table and meters without numerical values (including that the upper table bottom is empty, the lower table bottom is empty, the upper table bottom and the lower table bottom are empty, the lower table bottom is smaller than the upper table bottom, and the table bottom has no electric quantity), and the user is in zero degree, the user table and meters in the station area are considered to be failed in acquisition;
step B400, detecting whether the public and private transformer of the line and the user meter of the transformer area are out of power;
the method for detecting whether the public and private transformer of the line is not powered off comprises the following steps: detecting the numerical values of the upper and lower bottoms of the public and special transformer meters, wherein if the numerical values of the upper and lower bottoms are the same, no power is lost;
the method for detecting whether the user meter in the distribution area is powered off or not comprises the following steps: and detecting the numerical values of the upper and lower table bottoms of the user meter, wherein if the numerical values of the upper and lower table bottoms are the same, the user does not run power.
And step B500, generating a retail electricity quantity diagnosis report of the line and the distribution area.
Further, the third line loss analysis mode is a high-loss and negative-loss line and platform area diagnosis mode, and comprises
Step C100, acquiring data of all negative-loss and high-loss lines and transformer areas;
step C200, detecting whether metering on the power supply side of the line and the transformer area is wrong;
specifically, if the line has high loss, the bus where the line is located has negative loss, or the line has negative loss, and the bus where the line is located has high loss, the metering at the power supply side of the line has errors;
the distribution area is high loss, the line where the distribution area is located is negative loss, or the distribution area is negative loss, the line where the distribution area is located is high loss, and metering errors occur on the power supply side of the distribution area;
the high loss and negative loss of the bus and the line are defined as follows: the unbalance rate of the bus or the line is more than 2 percent, namely high loss, and the unbalance rate of the bus or the line is less than-2 percent, namely negative loss;
step C300, detecting whether the abnormal user with the electric quantity hangs down under the line and the platform area;
the method for detecting whether the line hangs down the user with abnormal electric quantity comprises the following steps:
and (4) balancing a bus where the line is located, sorting the electric quantity of the lower-hanging public and special transformer total meter from large to small, and taking the accumulated electric quantity and the users just larger than or equal to the line power supply quantity as suspicious users.
Assuming that n users are arranged under a certain line in the descending order of the electric quantity, calculating the accumulated electric quantity of all the users, if the calculation result is greater than or equal to (greater than or equal to) the power supply quantity of the line, intercepting the users under the line, and respectively setting the users as: x is a radical of a fluorine atom 1 ,x 2 ,x 3 ……x n The corresponding electric quantity of each user is respectively as follows: y is 1 ,y 2 ,y 3 ……y n 。
The accumulated electric quantity is
The electric quantity of each user accounts for the ratio that: y is 1 /Y、(y 1 +y 2 )/Y、(y 1 +y 2 +y 3 )/Y、……(y 1 +y 2 +y 3 +…+ y n )/Y。
Taking 5 users under the line as an example, the accumulated power ratio graph is shown in fig. 2.
And (3) when the line loss rate of the line in which the transformer area is located is normal, sorting the total electric quantity of the hung users from large to small, determining the accumulated proportion of the suspicious users by using the accumulated electric quantity and the users just more than or equal to the power supply quantity of the transformer area as the suspicious users, and drawing a proportion broken line graph.
Assuming that n users are arranged under a certain area in the descending order of the electric quantity, calculating the accumulated electric quantity of all the users, if the calculation result is greater than or equal to (greater than or equal to) the power supply quantity of the area, intercepting the users under the area, and respectively setting the users as: x is the number of 1 ,x 2 ,x 3 ……x n The corresponding electric quantity of each user is respectively as follows: y is 1 ,y 2 ,y 3 ……y n 。
The accumulated electric quantity is
The electric quantity of each user accounts for the ratio that: y is 1 /Y、(y 1 +y 2 )/Y、(y 1 +y 2 +y 3 )/Y、……(y 1 +y 2 +y 3 +…+ y n )/Y。
Taking 6 users under the line as an example, the accumulated electricity consumption ratio graph is shown in fig. 3.
Step C400, detecting the relationship between the line change and the distribution change of the transformer area;
the method for detecting the line variation relation comprises the following steps:
and calculating a correlation coefficient of the line loss rate fluctuation of the line and the fluctuation of each underhung public and private variable electric quantity, and if the correlation is established, determining that the public and private variable line variable relation is wrong.
The process of checking the line-variable relation is as follows:
and (3) solving a correlation coefficient: and calculating a correlation coefficient of the line loss rate variation of the line and the variation of each lower-hanging public and private variable, wherein if the correlation exists, the public and private variable line variation relationship is wrong.
1) Acquiring the electricity quantity variation of each distribution transformer in the line in the last 5 months, and setting the electricity quantity variation as A = { A = { (A) 1 、A 2 、A 3 、A 4 、A 5 }；
2) Obtaining the change of the line loss rate of the line in the last 5 months, and setting the change as B = { B = { (B) } 1 、B 2 、B 3 、B 4 、B 5 }；
3) Calculating the correlation coefficient rho of A and B AB
4) If | ρ AB |>, 0.8, strong correlation; l ρ AB |<, 0.3 is low correlation, others are moderate.
The detection method of the distribution transformation relation of the transformer area comprises the following steps:
and calculating a correlation coefficient of the line loss rate fluctuation of the transformer area and the electric quantity of each off-hook user, and if the correlation coefficient is correlated, determining that the operation and distribution relation of the user is wrong.
The checking process of the operation and distribution transformation relation is as follows:
and (3) solving a correlation coefficient: and calculating a correlation coefficient of the line loss rate variation of the transformer area and the electric quantity of each off-hook user, and if the correlation is obtained, determining that the operation and distribution relation of the user is wrong.
(1) Acquiring the latest 5 months of electric quantity of each user in the distribution area, and setting the electric quantity as C = { C = { (C) 1 、C 2 、C 3 、C 4 、C 5 }；
(2) Obtaining the change of the line loss rate of the station zone in the last 5 months, and setting the change as D = { D = { (D) 1 、D 2 、D 3 、D 4 、D 5 }
(3) Calculating the correlation coefficient rho of C and D CD
Wherein
(4) If | ρ CD |>, 0.8, then strong correlation; | ρ CD |&low correlation is found at 0.3; others are moderately correlated. | ρ CD |>, 0.3.
Step C500, detecting whether the public and private transformer and the distribution area summary sheet fail to collect;
specifically, if there is one or more of the following: if a public-private transformer is hung under a line, no meter is arranged, no data is arranged in the meter (including that the upper surface bottom is empty, the lower surface bottom is empty, the upper surface bottom and the lower surface bottom are empty, the lower surface bottom is smaller than the upper surface bottom, and no electric quantity is arranged in the meter bottom), and zero degree family exists, the line acquisition fails;
specifically, if there is one or more of the following: if the table area is hung with users without meters and meters without data (including the upper meter bottom is empty, the lower meter bottom is empty, the upper meter bottom and the lower meter bottom are empty, the lower meter bottom is smaller than the upper meter bottom and the meter bottom has no electric quantity) and zero degree users exist, the table area general meter collection fails;
step C600, detecting whether the electric quantity of the user suddenly decreases in the line and the distribution room;
the method for detecting whether the line has the sudden reduction of the user electric quantity comprises the following steps: the current month public and private variable electric quantity of the line is reduced by more than 20% compared with the last month electric quantity, namely the electric quantity of the line user is suddenly reduced;
the method for detecting whether the power quantity of the user suddenly decreases in the distribution area comprises the following steps: the electric quantity of the off-hook users in the current month of the transformer area is reduced by more than 20% compared with the electric quantity in the previous month, namely the electric quantity of the users in the transformer area is suddenly reduced;
and step C700, generating a negative loss and high loss diagnosis report of the line and the transformer area.
Before the method is used for analyzing the line loss abnormity of the lines and the transformer areas, the method can only judge by manual checking, and each line and transformer area generally takes about two to twenty minutes according to the scale of the line and transformer area.
An improved method for step 600 is: determining line loss analysis models of the lines and the transformer areas by using a random forest algorithm, and generating a diagnosis report;
as shown in fig. 4, further comprising:
step S606: the line loss data of the line and the transformer area are arranged into a training set, line loss analysis feature selection is carried out, and features used by the training set are determined;
step S610: setting relevant parameters of a random forest for a given training set S of the line loss data of the processed line and the line loss data of the transformer area and a characteristic dimension F of the training set S: the number g of used classification trees, the maximum depth d of each tree, and the number f of features used by each node; and establishing a termination condition: the minimum sample number s on the node and the minimum information gain m on the node;
step S620: randomly extracting a training set S (i) with the same size as S from S in a putting-back manner, taking the training set S (i) as a sample of a root node, and starting training from the root node;
step S630: if the current node meets the termination condition, the current node is set as a leaf node, the prediction output of the leaf node is the type c (j) with the largest number in the current node sample set, and the probability p is defined as the proportion of the c (j) in the current sample set. And then continue training other nodes. And if the current node does not reach the termination condition, randomly selecting the F-dimensional features from the F-dimensional features without putting back. And searching the one-dimensional feature k with the best classification effect and a threshold th thereof by using the f-dimensional feature, wherein the sample with the kth-dimensional feature smaller than th on the current node is divided into a left node, and the rest are divided into a right node. Continuing to train other nodes;
step S640: if all the nodes are trained repeatedly or marked as leaf nodes, then go to step S650; otherwise, go to step S620;
step S650: if all the classification trees are trained, ending; otherwise, the process proceeds to step S620.
The specific implementation method of the random forest for constructing the line loss analysis of the line and the platform area comprises the following steps:
assuming that the characteristics determined after the line loss analysis characteristic selection are power supply side meter, power supply quantity meter and high loss or negative loss of the line, the training set is shown as the following table:
for a given processed electricity consumption information data training set S as the table above, a first column represents a line or platform power supply side meter, the normality is 1, and the abnormality is 0; the second column represents the power supply amount metering of the line or the platform area, the normal is 1, and the abnormal is 0; the third column indicates that the line or block is high loss or negative loss, the high loss is 1, and the negative loss is 0; the fourth column shows the reasons of the circuit abnormality, and the specific reasons are replaced by 1,2,3; the characteristic dimension F =3, and the related parameters of the random forest are set as follows: the number of used classification trees g =3, the maximum depth of each tree d =4, and the number of used features per node f =1; and establishing a termination condition: the minimum number of samples on a node s =1, and the minimum information gain on the node m =0.001;
randomly extracting a training set S (i) with the same size as S from S in a putting-back manner, taking the training set S (i) as a sample of a root node, and starting training from the root node; <xnotran> { (1,1,1,1), (1,1,0,2), (1,1,1,1) }, { (1,1,0,2), (1,1,0,2), (1,1,1,1) }, { (0,1,0,3), (1,1,0,2), (0,1,0,3) }, 630 ～ 650 , 4 , . </xnotran>
Step S600: inputting a line loss condition table of a line or a transformer area to be predicted as a test set;
step S700: preprocessing a test set;
step S800: selecting the characteristics of the test set by using an information entropy algorithm;
step S900: the method for analyzing the line loss abnormal reasons of the line and the transformer area comprises the following specific steps: acquiring the predicted values of all g trees; taking the line loss abnormal reasons with the largest quantity in g trees as final line loss abnormal reasons, and generating a diagnosis report; or the line loss abnormal reasons are arranged from large to small according to the number of the line loss abnormal reasons, and the number of the line loss abnormal reasons presented in the diagnosis report is selected by a user to generate the diagnosis report; or a user sets a threshold for the percentage of causes of abnormalities that the percentage exceeds, to be presented in a diagnostic report.
| Data volume | Step 600 original method | Step 600 improvement |
| 20 | 0.85 | 0.95 |
| 50 | 0.88 | 0.92 |
| 100 | 0.83 | 0.87 |
| 500 | 0.78 | 0.85 |
| 1000 | 0.79 | 0.91 |
| 10000 | 0.77 | 0.88 |
| 100000 | 0.74 | 0.89 |
Comparing the original method of the step 600 with the improved method, and investigating the accuracy of the original method and the improved method under the same data quantity, as can be seen from the table above, the original method of the step 600 is approximately about 0.8 and has a slow descending trend along with the increase of data; the step 600 improved method has an accuracy of approximately 0.9 and remains substantially constant as data increases. The data show that the improved method of step 600 is more accurate, and therefore achieves a good result compared to the prior art, with significant improvements.
By using the original method of step 600, although the data processing speed is improved, the data can be processed in series, and the processing time is increased linearly under the condition of increasing the data volume; the improved method of the step 600 is used for constructing a random forest, so that data can be processed in batches, the processing efficiency is greatly improved, and manpower and material resources are saved.
The embodiment of the invention can accurately judge the reason of the abnormity under the condition that the circuit and the transformer area are abnormal, quickly take corresponding measures to correct and improve, improve the robustness and fault tolerance of the power grid, ensure the electricity consumption of the society and people, improve the frontal image of the power grid and enlarge the influence.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present 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 solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (7)
1. A line transformer area abnormity analysis method based on big data is characterized by comprising the following steps:
step 100: extracting table bottom and file data from the mass platform and the line loss database, and storing the table bottom and the file data into the database;
step 200: calculating the electricity selling quantity of low-voltage users, special transformer users and distribution areas according to the table bottom and the archive data in the database;
step 300: after the national network province company calculates and finishes the low-voltage users, the special transformer users and the distribution area electricity sales, the calculation result and the high voltage are uploaded to an unstructured database of the national network headquarters through an unstructured transmission channel of a data center;
step 400: extracting data of the unstructured database of the national network headquarters, and storing a table bottom, a file and a calculation model into a headquarter distributed file system;
step 500: calculating the power supply quantity and the line loss rate of the line and the distribution room according to the table bottom, the file and the calculation model obtained in the step 400;
step 600, selecting at least one line loss analysis mode to perform line loss analysis; and generating a diagnosis report according to the results of the one or more line loss analysis modes.
2. The big-data-based line transformer area anomaly analysis method according to claim 1, wherein the line loss analysis modes in step 600 include a first line loss analysis mode, a second line loss analysis mode and a third line loss analysis mode.
3. The big-data-based line and platform area abnormity analysis method according to claim 2, wherein the first line loss analysis mode is a line and platform area diagnosis mode with zero power supply, and comprises the following steps:
step A100, obtaining all lines and transformer area data with zero power supply quantity;
step A200, detecting whether the acquisition of the line and station power supply side meter fails;
step A300, detecting whether a power supply side meter is powered off or not;
step A400, detecting the configuration condition of the circuit and distribution area archive model;
and step A500, generating a zero power supply amount diagnosis report of the line and the station area.
4. The method for analyzing the abnormity of the line and platform area based on the big data as claimed in claim 2, wherein the second line loss analysis mode is a line and platform area diagnosis mode of retail electric quantity, and comprises:
step B100, acquiring all lines and distribution room data with zero electricity sales;
step B200, detecting the line-to-line transformation relation of the line and the operation and distribution transformation relation of the transformer area;
step B300, detecting whether the public and private change of the line and the user meter of the transformer area fail to collect;
step B400, detecting whether the public and private transformer of the line and the user meter of the transformer area are out of power;
and step B500, generating a diagnosis report of the retail electric quantity of the line and the distribution area.
5. The big data-based line area anomaly analysis method according to claim 2, characterized by comprising:
step C100, acquiring data of all negative-loss and high-loss lines and transformer areas;
step C200, detecting whether metering on the power supply side of the line and the transformer area is wrong;
step C300, detecting whether users with abnormal electric quantity hang down under the line and the distribution room;
step C400, detecting the relationship between the line variation and the distribution variation of the transformer area;
step C500, detecting whether the public and private transformer and the distribution area summary sheet fail to collect;
step C600, detecting whether the electric quantity of the user suddenly decreases in the line and the distribution room;
and step C700, generating a negative loss and high loss diagnosis report of the line and the transformer area.
6. The method for analyzing the abnormal line distribution room based on the big data as claimed in claim 5, wherein in step C300, the method for detecting whether the user with the abnormal power is hung down on the line includes:
if the bus where the line is located is balanced, the electric quantity of the public-specific transformer master tables hung under the line is sorted from large to small, and the users with the accumulated electric quantity and the power supply quantity just larger than or equal to the line are suspicious users;
if n users are in total under a certain line, arranging the users in the order of the electric quantity from large to small, and arranging x i For the ith subscriber under the line, y i The electricity quantity of the ith user under the line is accumulated
7. The method for analyzing the abnormal line distribution room based on the big data as claimed in claim 5, wherein in step C400, the method for detecting the line variation relationship comprises:
calculating a correlation coefficient of line loss rate fluctuation and each lower-hanging public and private variable electric quantity fluctuation, and if the correlation coefficient is correlated, determining that the public and private variable line-variable relation is wrong;
the method for solving the correlation coefficient comprises the following steps:
1) Acquiring the electricity quantity variation of each distribution transformer in the line in the last 5 months, and setting the electricity quantity variation as A = { A = { (A) 1 、A 2 、A 3 、A 4 、A 5 }；
2) Obtaining the change of the line loss rate of the line in the last 5 months, and setting the change as B = { B = { (B) } 1 、B 2 、B 3 、B 4 、B 5 }；
3) Calculating the correlation coefficient rho of A and B AB :
4) If | ρ AB |>, 0.8, then strong correlation; | ρ AB |<, 0.3 is low correlation, others are moderate.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710675256.6A CN107527114B (en) | 2017-08-09 | 2017-08-09 | A kind of route platform area exception analysis method based on big data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710675256.6A CN107527114B (en) | 2017-08-09 | 2017-08-09 | A kind of route platform area exception analysis method based on big data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107527114A true CN107527114A (en) | 2017-12-29 |
| CN107527114B CN107527114B (en) | 2019-04-23 |
Family
ID=60681022
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710675256.6A Active CN107527114B (en) | 2017-08-09 | 2017-08-09 | A kind of route platform area exception analysis method based on big data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107527114B (en) |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109086950A (en) * | 2018-10-11 | 2018-12-25 | 国网上海市电力公司 | A kind of user's coulometric analysis prediction technique based on big data analysis |
| CN109767349A (en) * | 2018-10-23 | 2019-05-17 | 中国电力科学研究院有限公司 | A kind of Line Loss of Distribution Network System data check and modification method and system |
| CN109919425A (en) * | 2019-01-23 | 2019-06-21 | 国网浙江省电力有限公司 | A kind of platform area customer relationship error correction method based on correlation analysis |
| CN110134708A (en) * | 2019-03-03 | 2019-08-16 | 云南电网有限责任公司信息中心 | Electric net platform region line loss abnormal cause diagnostic method, device, computer equipment and storage medium |
| CN110276511A (en) * | 2019-04-16 | 2019-09-24 | 国网浙江海盐县供电有限公司 | A kind of line change relationship anomalous discrimination method based on electricity and line loss relevance |
| CN110472871A (en) * | 2019-08-16 | 2019-11-19 | 广东电网有限责任公司 | A kind of investigation method for management line loss abnormal cause |
| CN110749784A (en) * | 2019-08-05 | 2020-02-04 | 上海大学 | Line electricity stealing detection method based on electric power data wavelet analysis |
| CN111461577A (en) * | 2020-04-28 | 2020-07-28 | 广东电网有限责任公司惠州供电局 | Line loss calculation method, device, equipment and storage medium |
| CN111579928A (en) * | 2020-05-25 | 2020-08-25 | 国网江苏省电力有限公司无锡供电分公司 | Distribution line loss analysis method and analysis device |
| CN111598380A (en) * | 2020-04-01 | 2020-08-28 | 国网江西省电力有限公司电力科学研究院 | Distribution line negative loss reason distinguishing method |
| CN111695600A (en) * | 2020-05-13 | 2020-09-22 | 国网湖北省电力有限公司电力科学研究院 | Multi-threshold and KNN-based distribution room classification method |
| CN112035544A (en) * | 2020-08-29 | 2020-12-04 | 国网信通亿力科技有限责任公司 | Power distribution network data anomaly monitoring and diagnosis method |
| CN113655308A (en) * | 2021-07-30 | 2021-11-16 | 国网天津市电力公司 | Synchronous line loss monitoring and management system based on intelligent sensing |
| CN114236283A (en) * | 2021-12-15 | 2022-03-25 | 广东电网有限责任公司 | Method and device for determining line loss reason of power supply network |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102545213A (en) * | 2012-01-11 | 2012-07-04 | 藁城市供电公司 | System and method for managing line loss of power grid in real time |
| CN102637288A (en) * | 2012-03-31 | 2012-08-15 | 上海市电力公司 | Method for analysis of line loss in distribution-room area of power supply enterprise |
| US20140148963A1 (en) * | 2009-01-14 | 2014-05-29 | Integral Analytics, Inc. | Optimization of microgrid energy use and distribution |
| US20150088583A1 (en) * | 2010-01-29 | 2015-03-26 | Enginuity Worldwide, LLC | Sustainability campus of co-located facilities |
| CN106570778A (en) * | 2016-10-14 | 2017-04-19 | 国网信通亿力科技有限责任公司 | Big data-based data integration and line loss analysis and calculation method |
| CN106932639A (en) * | 2017-03-22 | 2017-07-07 | 深圳市均方根科技有限公司 | A kind of platform area line loss analyzing management system |
| CN106950440A (en) * | 2017-02-14 | 2017-07-14 | 国网新疆电力公司伊犁供电公司 | Transformer station's metering critical point quickly judges alarm method extremely |
| CN107024641A (en) * | 2017-03-29 | 2017-08-08 | 国网上海市电力公司 | A kind of platform area line loss per unit exception monitoring system |
-
2017
- 2017-08-09 CN CN201710675256.6A patent/CN107527114B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140148963A1 (en) * | 2009-01-14 | 2014-05-29 | Integral Analytics, Inc. | Optimization of microgrid energy use and distribution |
| US20150088583A1 (en) * | 2010-01-29 | 2015-03-26 | Enginuity Worldwide, LLC | Sustainability campus of co-located facilities |
| CN102545213A (en) * | 2012-01-11 | 2012-07-04 | 藁城市供电公司 | System and method for managing line loss of power grid in real time |
| CN102637288A (en) * | 2012-03-31 | 2012-08-15 | 上海市电力公司 | Method for analysis of line loss in distribution-room area of power supply enterprise |
| CN106570778A (en) * | 2016-10-14 | 2017-04-19 | 国网信通亿力科技有限责任公司 | Big data-based data integration and line loss analysis and calculation method |
| CN106950440A (en) * | 2017-02-14 | 2017-07-14 | 国网新疆电力公司伊犁供电公司 | Transformer station's metering critical point quickly judges alarm method extremely |
| CN106932639A (en) * | 2017-03-22 | 2017-07-07 | 深圳市均方根科技有限公司 | A kind of platform area line loss analyzing management system |
| CN107024641A (en) * | 2017-03-29 | 2017-08-08 | 国网上海市电力公司 | A kind of platform area line loss per unit exception monitoring system |
Non-Patent Citations (3)
| Title |
|---|
| 施文: "台区线损异常成因分析及解决方法", 《电力需求侧管理》 * |
| 肖建红: "低压台区线损异常的监测与分析", 《电力需求侧管理》 * |
| 蔡晓燕: "基于智能用电大数据分析的台区线损管理初探", 《科技创新与应用》 * |
Cited By (21)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109086950A (en) * | 2018-10-11 | 2018-12-25 | 国网上海市电力公司 | A kind of user's coulometric analysis prediction technique based on big data analysis |
| CN109767349A (en) * | 2018-10-23 | 2019-05-17 | 中国电力科学研究院有限公司 | A kind of Line Loss of Distribution Network System data check and modification method and system |
| CN109919425A (en) * | 2019-01-23 | 2019-06-21 | 国网浙江省电力有限公司 | A kind of platform area customer relationship error correction method based on correlation analysis |
| CN110134708A (en) * | 2019-03-03 | 2019-08-16 | 云南电网有限责任公司信息中心 | Electric net platform region line loss abnormal cause diagnostic method, device, computer equipment and storage medium |
| CN110276511A (en) * | 2019-04-16 | 2019-09-24 | 国网浙江海盐县供电有限公司 | A kind of line change relationship anomalous discrimination method based on electricity and line loss relevance |
| CN110749784B (en) * | 2019-08-05 | 2022-07-08 | 上海大学 | Line electricity stealing detection method based on electric power data wavelet analysis |
| CN110749784A (en) * | 2019-08-05 | 2020-02-04 | 上海大学 | Line electricity stealing detection method based on electric power data wavelet analysis |
| CN110472871A (en) * | 2019-08-16 | 2019-11-19 | 广东电网有限责任公司 | A kind of investigation method for management line loss abnormal cause |
| CN110472871B (en) * | 2019-08-16 | 2023-07-07 | 广东电网有限责任公司 | Investigation method for managing line loss abnormal reasons |
| CN111598380A (en) * | 2020-04-01 | 2020-08-28 | 国网江西省电力有限公司电力科学研究院 | Distribution line negative loss reason distinguishing method |
| CN111598380B (en) * | 2020-04-01 | 2022-08-16 | 国网江西省电力有限公司电力科学研究院 | Distribution line negative loss reason distinguishing method |
| CN111461577A (en) * | 2020-04-28 | 2020-07-28 | 广东电网有限责任公司惠州供电局 | Line loss calculation method, device, equipment and storage medium |
| CN111695600A (en) * | 2020-05-13 | 2020-09-22 | 国网湖北省电力有限公司电力科学研究院 | Multi-threshold and KNN-based distribution room classification method |
| CN111695600B (en) * | 2020-05-13 | 2023-08-18 | 国网湖北省电力有限公司电力科学研究院 | Multi-threshold and KNN-based station classification method |
| CN111579928B (en) * | 2020-05-25 | 2022-03-15 | 国网江苏省电力有限公司无锡供电分公司 | Distribution line loss analysis method and analysis device |
| CN111579928A (en) * | 2020-05-25 | 2020-08-25 | 国网江苏省电力有限公司无锡供电分公司 | Distribution line loss analysis method and analysis device |
| CN112035544B (en) * | 2020-08-29 | 2022-05-17 | 国网信通亿力科技有限责任公司 | Power distribution network data anomaly monitoring and diagnosis method |
| CN112035544A (en) * | 2020-08-29 | 2020-12-04 | 国网信通亿力科技有限责任公司 | Power distribution network data anomaly monitoring and diagnosis method |
| CN113655308A (en) * | 2021-07-30 | 2021-11-16 | 国网天津市电力公司 | Synchronous line loss monitoring and management system based on intelligent sensing |
| CN114236283A (en) * | 2021-12-15 | 2022-03-25 | 广东电网有限责任公司 | Method and device for determining line loss reason of power supply network |
| CN114236283B (en) * | 2021-12-15 | 2024-02-13 | 广东电网有限责任公司 | Method and device for determining line loss reason of power supply network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107527114B (en) | 2019-04-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107527114A (en) | A kind of circuit taiwan area exception analysis method based on big data | |
| CN108520357B (en) | Method and device for judging line loss abnormality reason and server | |
| CN105046402B (en) | A kind of state evaluating method applied to secondary equipment of intelligent converting station | |
| CN109409628B (en) | Acquisition terminal manufacturer evaluation method based on metering big data clustering model | |
| CN109308571B (en) | Distribution line variable relation detection method | |
| CN106651169A (en) | Fuzzy comprehensive evaluation-based distribution automation terminal state evaluation method and system | |
| CN110097297A (en) | A kind of various dimensions stealing situation Intellisense method, system, equipment and medium | |
| CN110909983B (en) | Multidimensional assessment method for electric energy quality of active power distribution network | |
| CN106548357B (en) | Client satisfaction evaluation method and system | |
| CN109389145B (en) | Electric energy meter manufacturer evaluation method based on metering big data clustering model | |
| CN108320043A (en) | A kind of distribution network equipment state diagnosis prediction method based on electric power big data | |
| CN110222991B (en) | Metering device fault diagnosis method based on RF-GBDT | |
| CN104504508B (en) | The taiwan area closed-loop data analysis method returned based on step analysis and small echo | |
| CN107451402A (en) | A kind of equipment health degree appraisal procedure and device based on alarm data analysis | |
| CN108898311A (en) | A kind of data quality checking method towards intelligent distribution network repairing dispatching platform | |
| CN111260198A (en) | Method and system for judging degree of rationality of line loss in transformer area synchronization and terminal equipment | |
| CN113189418B (en) | Topological relation identification method based on voltage data | |
| CN109636110A (en) | A kind of method and device obtaining protective relaying device operating status | |
| CN114519514B (en) | Low-voltage transformer area reasonable line loss value measuring and calculating method, system and computer equipment | |
| CN104331532A (en) | Power transformer state evaluation method based on rough set-cloud model | |
| CN109389280B (en) | Electric energy meter manufacturer evaluation method based on multi-attribute decision model | |
| Hu et al. | A general model, estimation, and procedure for modeling recurrent failure process of high-voltage circuit breakers considering multivariate impacts | |
| CN110879377A (en) | Metering device fault tracing method based on deep belief network | |
| CN114354783A (en) | Health degree evaluation method of extra-high voltage oil chromatography monitoring device based on-operation data | |
| CN113554361A (en) | Comprehensive energy system data processing and calculating method and processing 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 | ||
| CB02 | Change of applicant information |
Address after: No. 8 Huanglong Road, Hangzhou, Zhejiang Province Applicant after: State Grid Zhejiang Electric Power Company Applicant after: State Grid Inforamtion & Telecommunication Great Power Science and Technology Co., Ltd. Address before: 361009 402, room 36, two day watching day, Xiamen Software Park, Siming District, Xiamen, Fujian. Applicant before: State Grid Inforamtion & Telecommunication Great Power Science and Technology Co., Ltd. Applicant before: State Grid Zhejiang Electric Power Company |
|
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: Room 402, No. 36 Guanri Road, Phase II Software Park, Siming District, Xiamen City, Fujian Province Applicant after: State Grid Inforamtion & Telecommunication Great Power Science and Technology Co., Ltd. Applicant after: State Grid Zhejiang Electric Power Company Address before: No. 8 Huanglong Road, Hangzhou, Zhejiang Province Applicant before: State Grid Zhejiang Electric Power Company Applicant before: State Grid Inforamtion & Telecommunication Great Power Science and Technology Co., Ltd. |
|
| CB02 | Change of applicant information | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CP01 | Change in the name or title of a patent holder |
Address after: Room 402, No. 36 Guanri Road, Phase II Software Park, Siming District, Xiamen City, Fujian Province Patentee after: STATE GRID INFO-TELECOM GREAT POWER SCIENCE AND TECHNOLOGY Co.,Ltd. Patentee after: STATE GRID ZHEJIANG ELECTRIC POWER Co.,Ltd. Address before: Room 402, No. 36 Guanri Road, Phase II Software Park, Siming District, Xiamen City, Fujian Province Patentee before: STATE GRID INFO-TELECOM GREAT POWER SCIENCE AND TECHNOLOGY Co.,Ltd. Patentee before: STATE GRID ZHEJIANG ELECTRIC POWER Co. |
|
| CP01 | Change in the name or title of a patent holder |