CN111628494B - Low-voltage distribution network topology identification method and system based on logistic regression method - Google Patents

Low-voltage distribution network topology identification method and system based on logistic regression method Download PDF

Info

Publication number
CN111628494B
CN111628494B CN202010393380.5A CN202010393380A CN111628494B CN 111628494 B CN111628494 B CN 111628494B CN 202010393380 A CN202010393380 A CN 202010393380A CN 111628494 B CN111628494 B CN 111628494B
Authority
CN
China
Prior art keywords
voltage
historical
data
time series
low
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.)
Active
Application number
CN202010393380.5A
Other languages
Chinese (zh)
Other versions
CN111628494A (en
Inventor
张弛
周金辉
赵健
邵先军
童力
江航
陈蕾
谢琳
王凯
赵启承
王子凌
陈超
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
State Grid Zhejiang Electric Power Co Ltd
Electric Power Research Institute of State Grid Zhejiang Electric Power Co Ltd
Shanghai Electric Power University
Zhejiang Huayun Information Technology Co Ltd
Original Assignee
State Grid Zhejiang Electric Power Co Ltd
Electric Power Research Institute of State Grid Zhejiang Electric Power Co Ltd
Shanghai Electric Power University
Zhejiang Huayun Information Technology Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by State Grid Zhejiang Electric Power Co Ltd, Electric Power Research Institute of State Grid Zhejiang Electric Power Co Ltd, Shanghai Electric Power University, Zhejiang Huayun Information Technology Co Ltd filed Critical State Grid Zhejiang Electric Power Co Ltd
Priority to CN202010393380.5A priority Critical patent/CN111628494B/en
Publication of CN111628494A publication Critical patent/CN111628494A/en
Application granted granted Critical
Publication of CN111628494B publication Critical patent/CN111628494B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for ac mains or ac distribution networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J3/00—Circuit arrangements for ac mains or ac distribution networks
    • H02J3/38—Arrangements for parallely feeding a single network by two or more generators, converters or transformers
    • H02J3/381—Dispersed generators
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J2203/00—Indexing scheme relating to details of circuit arrangements for AC mains or AC distribution networks
    • H02J2203/20—Simulating, e g planning, reliability check, modelling or computer assisted design [CAD]
    • H—ELECTRICITY
    • H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
    • H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
    • H02J2300/00—Systems for supplying or distributing electric power characterised by decentralized, dispersed, or local generation
    • H02J2300/20—The dispersed energy generation being of renewable origin

Abstract

The invention discloses a low-voltage distribution network topology identification method and system based on a logistic regression method. With the penetration of distributed energy, the electric power energy system puts higher requirements on a rapid and accurate online data analysis tool. The technical scheme adopted by the low-voltage distribution network topology identification method is as follows: firstly, collecting time sequence voltage amplitude information of a user of a low-voltage distribution area, secondly, training historical data to obtain a topological classification model, then obtaining a group of most similar historical data through a nearest neighbor algorithm, and finally, putting the historical data into a trained regression model to predict and finish the topological identification of the low-voltage distribution area. The method can better cope with the change of the power industry mode from the traditional physical model-based monitoring system structure to the data-driven resource management mode, and breaks through the current situation that the existing monitoring architecture-based model needs large-scale complex modeling and consumes a large amount of time.

Description

Low-voltage distribution network topology identification method and system based on logistic regression method
Technical Field
The invention belongs to the field of power distribution network topology identification, and relates to a low-voltage power distribution network topology identification method and system based on a logistic regression method.
Background
The power network analysis and optimization problem is the basis for power system control, operation and planning. In recent years, high-proportion renewable energy sources are connected to the power grid on a large scale, the power generation output of the high-proportion renewable energy sources has strong randomness, and higher requirements are provided for uncertainty analysis and optimization of a power system. The permeation of distributed energy, energy storage, electric vehicles and other equipment in a terminal power distribution network is gradually increased, so that the analysis and control of the power distribution network are more complicated. These factors present challenges to the power network analysis and optimization problem.
On the other hand, data analysis and processing technology has rapidly developed in recent years, and various industries around the world have made a great deal of investment and have developed intense competition in a brand new dimension of data acquisition, analysis and processing. Due to the characteristic of real-time balance of electric power, the electric power industry has higher requirements on real-time performance and accuracy in daily scheduling operation, so that the measurement, transmission and storage of data have better accumulation, and a good foundation is laid for data-driven analysis. The problem is mainly independent of a physical model of a power system, and can be summarized as that the dependence on the physical model of the power system is reduced by a process of solving unknown data (data to data, D2D) according to known data by virtue of a certain incidence relation. The application of data-driven thinking and technologies to the traditional power network analysis and optimization problem can comprehensively utilize historical data and priori knowledge, so that the results of the power network analysis and optimization are more accurate and updated in real time, and the stability of a power system is improved.
Problems in the existing research: most researches on the topology of the power system are limited to identification of topology errors of a power transmission network and detection of changes of a topological structure, but at present, the power distribution network is frequently updated and complicated in wiring, and mass data of all links of the power distribution network are difficult to identify the topology of the power distribution network along with development of a smart power grid.
Disclosure of Invention
The invention aims to provide a low-voltage distribution network topology identification method and system based on a logistic regression method aiming at the defects of the prior art, which can effectively solve the problem that the difficulty of topology identification is obviously increased due to frequent changes of network topology caused by large-scale grid connection of high-proportion renewable energy sources, and further perform calculation and optimization on an electric power system, thereby improving the stability of the electric power system.
Therefore, the invention adopts the following technical scheme: a low-voltage distribution network topology identification method based on a logistic regression method comprises the following steps:
1) collecting time series voltage amplitude information of a low-voltage distribution station user as data for identifying low-voltage distribution station topology, and using a corresponding time series historical topological structure as a label;
2) training a regression method to generate a regression model;
3) obtaining a set of most similar historical data by a nearest neighbor algorithm using time series historical voltage measurements of the historical data
Figure BDA0002486758530000021
4) Using the historical data
Figure BDA0002486758530000022
And (4) putting the power distribution network into a trained regression model for prediction to obtain the optimal topological structure of the low-voltage power distribution network.
The invention aims to improve the accuracy of topology identification by utilizing the massive historical data so as to deal with the uncertainty in the smart grid.
As a further optimization scheme, the data collected in step 1) comprises time-series historical voltage measurement values of historical data and a time-series historical topology structure of the historical data;
11) time series voltage amplitude measurements for low voltage distribution station users: { z1,z2,...,zk,...,zQQ is the length of the time series, zkA vector representing voltage measurements over a time series;
12) time series historical topology of historical data: { s1,s2,...,sk,...,sQQ represents the length of the time series, skThe switching state vector represents a switching state vector of the power distribution network under a time sequence, the value is 1 when the switch is closed, and the value is 0 when the switch is disconnected.
As a further optimization scheme, the specific content of step 2) is as follows:
21) putting the obtained historical data into a machine learning algorithm SVM for training to obtain a topological classification model, and training a regression model, namely the probability distribution with the maximum value as follows:
Figure BDA0002486758530000023
wherein w is { w ═ w0,w1,...,wnParameter representing probability, zkAnd s(l)Time series historical voltage measurement vector and collection respectively representing historical dataTime series historical topology structure vectors of historical data;
22) log the first equation in 21) yields:
Figure BDA0002486758530000024
in the formula (I), the compound is shown in the specification,
Figure BDA0002486758530000025
representing the kth element in the switching state vector of the power distribution network under a time node k;
23)
Figure BDA0002486758530000026
is a binary variable, and is obtained by substituting into a formula:
Figure BDA0002486758530000027
for this formula, the gradient descent method is used to obtain the optimal parameters
Figure BDA0002486758530000031
Completing the training of the model; w is aiWhich represents the (i) th parameter,
Figure BDA0002486758530000032
representing the kth element in the distribution network switch state vector at time node i.
As a further optimization, the acquisition in step 3) is carried out
Figure BDA0002486758530000033
The formula of (1) is:
Figure BDA0002486758530000034
d (v) is a function of time and distance, representing the measured value z of the current voltagecurrentWith historical voltage measurements zkThe distance of (d);
Figure BDA0002486758530000035
is a vector in which the voltage measurement vector and the switch state vector are merged together; q is the number of all data points in the time data set; n is a set of temporal natural numbers; k is the index of the data point; k is the base of v, which represents the voltage.
The invention also provides a low-voltage distribution network topology identification system based on a logistic regression method, which comprises the following steps:
a data acquisition module: collecting time series voltage amplitude information of a low-voltage distribution station user as data for identifying low-voltage distribution station topology, and using a corresponding time series historical topological structure as a label;
a model generation module: training a regression method to generate a regression model;
a historical data acquisition module: obtaining a set of most similar historical data by a nearest neighbor algorithm using time series historical voltage measurements of the historical data
Figure BDA0002486758530000036
A topology prediction module: using the historical data
Figure BDA0002486758530000037
And (4) putting the power distribution network into a trained regression model for prediction to obtain the optimal topological structure of the low-voltage power distribution network.
Further, in the data acquisition module, the acquired data includes a time-series historical voltage measurement value of the historical data and a time-series historical topology structure of the historical data, and the specific contents are as follows:
11) time series voltage amplitude measurements for low voltage distribution station users: { z1,z2,...,zk,...,zQQ is the length of the time series, zkA vector representing voltage measurements over a time series;
12) time series historical topology of historical data: { s1,s2,...,sk,...,sQQ represents the length of the time series, skThe switching state vector represents a switching state vector of the power distribution network under a time sequence, the value is 1 when the switch is closed, and the value is 0 when the switch is disconnected.
Further, the specific content of the model generation module is as follows:
21) putting the obtained historical data into a machine learning algorithm SVM for training to obtain a topological classification model, and training a regression model, namely the probability distribution with the maximum value as follows:
Figure BDA0002486758530000038
wherein w is { w ═ w0,w1,...,wnParameter representing probability, zkAnd s(l)Respectively representing a time series historical voltage measured value vector of historical data and a time series historical topological structure vector of collected historical data;
22) log the first equation in 21) yields:
Figure BDA0002486758530000041
in the formula (I), the compound is shown in the specification,
Figure BDA0002486758530000042
representing the kth element in the switching state vector of the power distribution network under a time node k;
23)
Figure BDA0002486758530000043
is a binary variable, and is obtained by substituting into a formula:
Figure BDA0002486758530000044
for this formula, the gradient descent method is used to obtain the optimal parameters
Figure BDA0002486758530000045
Completing the training of the model; w is aiWhich represents the (i) th parameter,
Figure BDA0002486758530000046
representing the kth element in the distribution network switch state vector at time node i.
Further, the historical data acquisition module acquires
Figure BDA0002486758530000047
The formula of (1) is:
Figure BDA0002486758530000048
d (v) is a function of time and distance, representing the measured value z of the current voltagecurrentWith historical voltage measurements zkThe distance of (d);
Figure BDA0002486758530000049
is a vector in which the voltage measurement vector and the switch state vector are merged together; q is the number of all data points in the time data set; n is a set of temporal natural numbers; k is the index of the data point; k is the base of v, which represents the voltage.
The invention has the following beneficial effects: the method can effectively solve the problem that the difficulty of topology identification is obviously increased due to frequent changes of network topology caused by large-scale grid connection of high-proportion renewable energy sources, and further performs calculation and optimization on the electric power system, so that the stability of the electric power system is improved; the method has high accuracy of topology identification and short running time of topology identification.
Drawings
Fig. 1 is a flow chart of a low-voltage distribution network topology identification method of the present invention;
fig. 2 is a diagram of an IEEE30 node according to an embodiment of the present invention.
Detailed Description
The technical scheme of the invention is explained in detail in the following with the accompanying drawings of the specification.
Example 1
The embodiment provides a low-voltage distribution network topology identification method based on a logistic regression method, a flow chart of which is shown in fig. 1, and the method comprises the following steps:
step (1): the method comprises the steps of collecting time series voltage amplitude information of low-voltage power distribution station users as data for identifying low-voltage power distribution station topology, and using a corresponding time series historical topological structure as a label.
And measuring the voltage of each node at regular time by using the intelligent electric meter to obtain the historical time sequence value of the voltage of each node.
Time series voltage amplitude measurements for low voltage distribution station users: { z1,z2,...,zk,...,zQQ represents the length of the time series, zkRepresenting a vector of voltage measurements in a time series.
Collecting a time-series historical topology of historical data: { s1,s2,...,sk,...,sQQ represents the length of the time series, skThe switching state vector represents a switching state vector of the power distribution network under a time sequence, the value is 1 when the switch is closed, and the value is 0 when the switch is disconnected.
Step (2): and putting the obtained historical data into a machine learning algorithm SVM for training to obtain a topological classification model.
Training the regression model, i.e. the probability distribution at the maximum:
Figure BDA0002486758530000051
wherein w ═ { w ═ w0,w1,...,wn}。
Taking log of the equation yields:
Figure BDA0002486758530000052
Figure BDA0002486758530000053
the binary variable represents the kth element in the switching state vector of the power distribution network under the time node k.
Substituting into the formula can obtain:
Figure BDA0002486758530000054
the optimal parameters can be obtained by using a gradient descent method
Figure BDA0002486758530000055
And finishing the training of the model. w is aiWhich represents the (i) th parameter,
Figure BDA0002486758530000056
representing the kth element in the distribution network switch state vector at time node i.
Step (3) using the time series historical voltage measurement value of the historical data to obtain a group of most similar historical data through a nearest neighbor algorithm
Figure BDA0002486758530000057
And classifying all the nodes by adopting a K neighbor algorithm according to the historical time sequence value of the node voltage to obtain K neighbor points of each node.
And calculating the distance between any two nodes according to the voltage time sequence value, wherein k nodes with the minimum distance from the selected node are used as adjacent points of the selected node.
The k value is selected to meet the dual requirements of the accuracy rate and the running time of the topology identification, and the larger the value of k is, the higher the accuracy rate of the topology identification is, but the longer the running time is.
The value range of k is 2-n, and n represents the number of nodes of the power distribution network.
k is equal to the integer closest to n/4, n representing the number of nodes of the distribution network.
Obtaining
Figure BDA0002486758530000058
The formula of (a) is specifically:
Figure BDA0002486758530000061
and (4): obtaining a set of most similar historical data by a nearest neighbor algorithm
Figure BDA0002486758530000064
And putting the topological structure into a trained regression model for prediction to obtain the optimal topological structure of the low-voltage distribution area.
Fig. 2 is an IEEE43 node diagram.
In the IEEE43 node map, a conventional topology identification method and a topology identification method based on logistic regression are respectively employed.
The rapid algorithm provided by the invention is used for identifying the topology of the power distribution network, and has the following advantages:
1) the accuracy rate of topology identification is high
The definition of the accuracy of the topology identification algorithm is as follows:
Figure BDA0002486758530000062
wherein the content of the first and second substances,
Figure BDA0002486758530000063
for the set of node pairs for topology estimation, | ε | is the size of the exact set of topology node pairs ε.
2) The running time of topology identification is short.
Example 2
The embodiment provides a low voltage distribution network topology identification system based on a logistic regression method, which includes:
a data acquisition module: collecting time series voltage amplitude information of a low-voltage distribution station user as data for identifying low-voltage distribution station topology, and using a corresponding time series historical topological structure as a label;
a model generation module: training a regression method to generate a regression model;
a historical data acquisition module: obtaining a set of most similar historical data by a nearest neighbor algorithm using time series historical voltage measurements of the historical data
Figure BDA0002486758530000065
A topology prediction module: using the historical data
Figure BDA0002486758530000066
And (4) putting the power distribution network into a trained regression model for prediction to obtain the optimal topological structure of the low-voltage power distribution network.
In the data acquisition module, the acquired data comprises time-series historical voltage measurement values of historical data and a time-series historical topological structure of the historical data, and the specific contents are as follows:
11) time series voltage amplitude measurements for low voltage distribution station users: { z1,z2,...,zk,...,zQQ is the length of the time series, zkA vector representing voltage measurements over a time series;
12) time series historical topology of historical data: { s1,s2,...,sk,...,sQQ represents the length of the time series, skThe switching state vector represents a switching state vector of the power distribution network under a time sequence, the value is 1 when the switch is closed, and the value is 0 when the switch is disconnected.
The specific contents of the model generation module are as follows:
21) putting the obtained historical data into a machine learning algorithm SVM for training to obtain a topological classification model, and training a regression model, namely the probability distribution with the maximum value as follows:
Figure BDA0002486758530000071
wherein w is { w ═ w0,w1,...,wnRepresenting probabilitiesParameter, zkAnd s(l)Respectively representing a time series historical voltage measured value vector of historical data and a time series historical topological structure vector of collected historical data;
22) log the first equation in 21) yields:
Figure BDA0002486758530000072
in the formula (I), the compound is shown in the specification,
Figure BDA0002486758530000073
representing the kth element in the switching state vector of the power distribution network under a time node k;
23)
Figure BDA0002486758530000074
is a binary variable, and is obtained by substituting into a formula:
Figure BDA0002486758530000075
for this formula, the gradient descent method is used to obtain the optimal parameters
Figure BDA0002486758530000076
Completing the training of the model; w is aiWhich represents the (i) th parameter,
Figure BDA0002486758530000077
representing the kth element in the distribution network switch state vector at time node i.
The historical data acquisition module acquires
Figure BDA0002486758530000078
The formula of (1) is:
Figure BDA0002486758530000079
d (v) is the distance in timeFunction representing the measured value z of the present voltagecurrentWith historical voltage measurements zkThe distance of (d);
Figure BDA00024867585300000710
is a vector in which the voltage measurement vector and the switch state vector are merged together; q is the number of all data points in the time data set; n is a set of temporal natural numbers; k is the index of the data point; k is the base of v, which represents the voltage.

Claims (4)

1. A low-voltage distribution network topology identification method based on a logistic regression method is characterized by comprising the following steps:
1) collecting time series voltage amplitude information of a low-voltage distribution station user as data for identifying low-voltage distribution station topology, and using a corresponding time series historical topological structure as a label;
2) training a regression method to generate a regression model;
3) obtaining a set of most similar historical data by a nearest neighbor algorithm using time series historical voltage measurements of the historical data
Figure FDA0003242917310000011
4) Using the historical data
Figure FDA0003242917310000012
Putting the power distribution network into a trained regression model for prediction to obtain an optimal topological structure of the low-voltage power distribution area;
the data collected in the step 1) comprises time-series historical voltage measurement values of historical data and a time-series historical topological structure of the historical data;
11) time series voltage amplitude measurements for low voltage distribution station users: { z1,z2,...,zk,...,zQQ is the length of the time series, zkA vector representing voltage measurements over a time series;
12) time series historical topology of historical data: { s1,s2,...,sk,...,sQQ represents the length of the time series, skRepresenting a switch state vector of the power distribution network under a time sequence, wherein the value is 1 when the switch is closed and the value is 0 when the switch is disconnected;
the specific content of the step 2) is as follows:
21) putting the obtained historical data into a machine learning algorithm SVM for training to obtain a topological classification model, and training a regression model, namely the probability distribution with the maximum value as follows:
Figure FDA0003242917310000013
wherein w is { w ═ w0,w1,...,wnParameter representing probability, zkAnd s(l)Respectively representing a time series historical voltage measured value vector of historical data and a time series historical topological structure vector of collected historical data;
22) log the first equation in 21) yields:
Figure FDA0003242917310000014
in the formula (I), the compound is shown in the specification,
Figure FDA0003242917310000015
representing the kth element in the switching state vector of the power distribution network under a time node k;
23)
Figure FDA0003242917310000016
is a binary variable, and is obtained by substituting into a formula:
Figure FDA0003242917310000017
for this formula, the gradient descent method is used to obtain the optimal parameters
Figure FDA0003242917310000021
Completing the training of the model; w is aiWhich represents the (i) th parameter,
Figure FDA0003242917310000022
representing the kth element in the distribution network switch state vector at time node i.
2. The method for identifying the topology of the low-voltage distribution network based on the logistic regression method as claimed in claim 1, wherein the topology obtained in the step 3) is obtained
Figure FDA0003242917310000023
The formula of (1) is:
Figure FDA0003242917310000024
d (v) is a function of time and distance, representing the measured value z of the current voltagecurrentWith historical voltage measurements zkThe distance of (d);
Figure FDA0003242917310000025
is a vector in which the voltage measurement vector and the switch state vector are merged together; q is the number of all data points in the time data set; n is a set of temporal natural numbers; k is the index of the data point; k is the base of v, which represents the voltage.
3. A low voltage distribution network topology identification system based on a logistic regression method is characterized by comprising the following steps:
a data acquisition module: collecting time series voltage amplitude information of a low-voltage distribution station user as data for identifying low-voltage distribution station topology, and using a corresponding time series historical topological structure as a label;
a model generation module: training a regression method to generate a regression model;
a historical data acquisition module: make itObtaining a group of most similar historical data by a nearest neighbor algorithm by using time series historical voltage measurement values of the historical data
Figure FDA0003242917310000026
A topology prediction module: using the historical data
Figure FDA0003242917310000027
Putting the power distribution network into a trained regression model for prediction to obtain an optimal topological structure of the low-voltage power distribution area;
in the data acquisition module, the acquired data comprises time-series historical voltage measurement values of historical data and a time-series historical topological structure of the historical data, and the specific contents are as follows:
11) time series voltage amplitude measurements for low voltage distribution station users: { z1,z2,...,zk,...,zQQ is the length of the time series, zkA vector representing voltage measurements over a time series;
12) time series historical topology of historical data: { s1,s2,...,sk,...,sQQ represents the length of the time series, skRepresenting a switch state vector of the power distribution network under a time sequence, wherein the value is 1 when the switch is closed and the value is 0 when the switch is disconnected;
the specific contents of the model generation module are as follows:
21) putting the obtained historical data into a machine learning algorithm SVM for training to obtain a topological classification model, and training a regression model, namely the probability distribution with the maximum value as follows:
Figure FDA0003242917310000028
wherein w is { w ═ w0,w1,...,wnParameter representing probability, zkAnd s(l)Time series calendar respectively representing historical dataHistorical voltage measurement value vector and historical topological structure vector of time series of collected historical data;
22) log the first equation in 21) yields:
Figure FDA0003242917310000031
in the formula (I), the compound is shown in the specification,
Figure FDA0003242917310000032
representing the kth element in the switching state vector of the power distribution network under a time node k;
23)
Figure FDA0003242917310000033
is a binary variable, and is obtained by substituting into a formula:
Figure FDA0003242917310000034
for this formula, the gradient descent method is used to obtain the optimal parameters
Figure FDA0003242917310000035
Completing the training of the model; w is aiWhich represents the (i) th parameter,
Figure FDA0003242917310000036
representing the kth element in the distribution network switch state vector at time node i.
4. The system for identifying topology of low-voltage distribution network based on logistic regression method according to claim 3, wherein the historical data acquisition module acquires
Figure FDA0003242917310000037
The formula of (1) is:
Figure FDA0003242917310000038
d (v) is a function of time and distance, representing the measured value z of the current voltagecurrentWith historical voltage measurements zkThe distance of (d);
Figure FDA0003242917310000039
is a vector in which the voltage measurement vector and the switch state vector are merged together; q is the number of all data points in the time data set; n is a set of temporal natural numbers; k is the index of the data point; k is the base of v, which represents the voltage.
CN202010393380.5A 2020-05-11 2020-05-11 Low-voltage distribution network topology identification method and system based on logistic regression method Active CN111628494B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010393380.5A CN111628494B (en) 2020-05-11 2020-05-11 Low-voltage distribution network topology identification method and system based on logistic regression method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010393380.5A CN111628494B (en) 2020-05-11 2020-05-11 Low-voltage distribution network topology identification method and system based on logistic regression method

Publications (2)

Publication Number Publication Date
CN111628494A CN111628494A (en) 2020-09-04
CN111628494B true CN111628494B (en) 2021-12-10

Family

ID=72271821

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010393380.5A Active CN111628494B (en) 2020-05-11 2020-05-11 Low-voltage distribution network topology identification method and system based on logistic regression method

Country Status (1)

Country Link
CN (1) CN111628494B (en)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112241812B (en) * 2020-10-20 2023-10-03 江苏其厚智能电气设备有限公司 Topology identification method for low-voltage distribution network based on single-side optimization and genetic algorithm cooperation
CN112182499B (en) * 2020-10-23 2022-10-14 国网天津市电力公司 Low-voltage distribution network topological structure identification method based on time sequence electric quantity data
CN112688365B (en) * 2020-12-26 2023-07-04 四川大川云能科技有限公司 Mutual information-Bayesian network-based power distribution network topology robust identification method
CN112986658B (en) * 2021-01-25 2023-07-18 浙江华云信息科技有限公司 Topology identification method and device for low-voltage circuit
CN114814420B (en) * 2022-04-20 2024-02-02 北京飞利信信息安全技术有限公司 Low-voltage distribution network topology identification method and system based on frozen data

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108183475B (en) * 2017-12-27 2021-02-09 国网河南省电力公司电力科学研究院 Low-voltage distribution network topology reconstruction method and device
CN109359271B (en) * 2018-12-21 2020-06-30 浙江大学 Transformer winding deformation degree online detection method based on logistic regression

Also Published As

Publication number Publication date
CN111628494A (en) 2020-09-04

Similar Documents

Publication Publication Date Title
CN111628494B (en) Low-voltage distribution network topology identification method and system based on logistic regression method
CN110659693B (en) K-nearest neighbor classification-based power distribution network rapid topology identification method, system and medium
CN111900731B (en) PMU-based power system state estimation performance evaluation method
CN111368904B (en) Electrical equipment identification method based on electric power fingerprint
CN110363334B (en) Grid line loss prediction method of photovoltaic grid connection based on gray neural network model
CN113723844B (en) Low-voltage station theoretical line loss calculation method based on ensemble learning
CN104578053A (en) Power system transient stability prediction method based on disturbance voltage trajectory cluster features
CN116148753A (en) Intelligent electric energy meter operation error monitoring system
CN113937764A (en) Low-voltage distribution network high-frequency measurement data processing and topology identification method
CN111654392A (en) Low-voltage distribution network topology identification method and system based on mutual information
CN111091223A (en) Distribution transformer short-term load prediction method based on Internet of things intelligent sensing technology
CN110569888A (en) transformer fault diagnosis method and device based on directed acyclic graph support vector machine
CN111965442A (en) Energy internet fault diagnosis method and device under digital twin environment
CN115296295B (en) Low-voltage distribution transformer area topology identification method and system based on cluster analysis and graph theory
CN106816871B (en) State similarity analysis method for power system
CN109638892A (en) A kind of photovoltaic plant equivalent modeling method based on improvement fuzzy clustering algorithm
CN106684854B (en) A kind of active power distribution network voltage limit risk analysis method based on node equivalent
CN109378834A (en) Large scale electric network voltage stability margin assessment system based on information maximal correlation
CN115377999A (en) Subsynchronous oscillation identification method based on broadband measurement data
CN114838923A (en) Fault diagnosis model establishing method and fault diagnosis method for on-load tap-changer
Wei et al. Fault location method for active distribution network based on SVM and feature search algorithm
Zhang et al. Topology identification method of low voltage aea based on topological data analysis
Miao et al. A Transformer District Line Loss Calculation Method Based on Data Mining and Machine Learning
CN117371623B (en) Electric energy meter running state early warning method and system
Yang et al. Data-driven Method of Renewable Energy Based on Generative Adversarial Networks and EnergyPLAN

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant