CN111784537A

CN111784537A - Power distribution network state parameter monitoring method and device and electronic equipment

Info

Publication number: CN111784537A
Application number: CN202010622269.9A
Authority: CN
Inventors: 张维; 刘玉民; 李温静; 刘柱; 张骁; 孟洪民; 李艺欣
Original assignee: State Grid Information and Telecommunication Co Ltd
Current assignee: State Grid Information and Telecommunication Co Ltd
Priority date: 2020-06-30
Filing date: 2020-06-30
Publication date: 2020-10-16
Anticipated expiration: 2040-06-30
Also published as: CN111784537B

Abstract

The application discloses a method and a device for monitoring state parameters of a power distribution network and electronic equipment, wherein power distribution network monitoring data of the power distribution network at a target moment are obtained, and the power distribution network monitoring data comprise a plurality of state parameters; acquiring a target parameter group corresponding to the monitoring data of the power distribution network from a set comprising a plurality of preset parameter groups, wherein the target parameter group comprises a plurality of state parameters; the preset parameter group in the set is generated based on historical data of the power distribution network corresponding to the power distribution network at a plurality of historical moments respectively, the historical data of the power distribution network corresponding to each historical moment comprises a plurality of state parameters, and the historical moments are moments selected in the fault-free moment section of the power distribution network; acquiring a parameter similarity value between a state parameter in the target parameter group and a state parameter in the power distribution network monitoring data; and outputting an alarm result corresponding to the power distribution network monitoring data under the condition that the parameter similarity value meets a preset alarm condition.

Description

Power distribution network state parameter monitoring method and device and electronic equipment

Technical Field

The application relates to the technical field of electric power, in particular to a method and a device for monitoring state parameters of a power distribution network and electronic equipment.

Background

The distribution network is located at the end of the power system and directly interacts with most electricity consumers, and the power supply reliability and the electricity consumption experience of the consumers are directly influenced by the quality of the operation condition of the distribution network. Statistical analysis of average customer blackout time data currently shows that 90% of blackout time is caused by power distribution network faults. Therefore, it is necessary to monitor whether a fault condition of the power distribution network occurs in real time.

At present, when monitoring a fault state of a power distribution network, a fixed threshold is usually set for the collected current and voltage data of the power distribution network to determine whether state parameters such as current or voltage are abnormal.

However, in an actual environment, the distribution network may have different characteristics due to changes of power consumers or changes of user states of the power consumers, for example, power consumption power, and the like may be different due to changes of the user clients, and therefore, there is a case where monitoring is inaccurate when a single fixed threshold is used to monitor the state parameters of the distribution network.

Disclosure of Invention

In view of this, the present application provides a method and an apparatus for monitoring power distribution network state parameters, and an electronic device, including:

a power distribution network state parameter monitoring method comprises the following steps:

acquiring power distribution network monitoring data of a power distribution network at a target moment, wherein the power distribution network monitoring data comprises a plurality of state parameters;

acquiring a target parameter group corresponding to the power distribution network monitoring data in a set comprising a plurality of preset parameter groups, wherein the target parameter group comprises a plurality of state parameters;

the preset parameter group in the set is generated based on historical data of the power distribution network corresponding to the power distribution network at a plurality of historical moments respectively, the historical data of the power distribution network corresponding to each historical moment comprises a plurality of state parameters, and the historical moment is a moment selected in the fault-free moment section of the power distribution network;

obtaining a parameter similarity value between the state parameters in the target parameter group and the state parameters in the power distribution network monitoring data;

and outputting an alarm result corresponding to the power distribution network monitoring data under the condition that the parameter similarity value meets a preset alarm condition.

The above method, preferably, further comprises:

obtaining a set corresponding to the power distribution network in advance, specifically including:

acquiring first historical state data, wherein the first historical state data comprises a group of historical data of the power distribution network corresponding to a plurality of historical moments respectively, the historical data of the power distribution network corresponding to each historical moment comprises a plurality of state parameters, and the historical moments are moments selected in the fault-free moment section of the power distribution network;

fuzzification processing is carried out on the power distribution network historical data corresponding to each historical moment in the first historical state data, so as to obtain fuzzy data corresponding to each historical moment, wherein the fuzzy data corresponding to each historical moment comprises a fuzzy set corresponding to each state parameter in the power distribution network historical data and a membership value corresponding to the fuzzy set;

generating an initial fuzzy rule statement corresponding to each historical moment according to a fuzzy set and a membership value in fuzzy data corresponding to each historical moment;

screening at least the initial fuzzy rule statement according to the minimum confidence value and the minimum support value of the fuzzy rule statement to obtain a target fuzzy rule statement of which the minimum confidence value is greater than or equal to a confidence threshold value and the minimum support value is greater than or equal to a support threshold value;

and performing defuzzification processing on the target fuzzy rule sentences respectively to obtain preset parameter groups corresponding to each target fuzzy rule sentence, wherein the preset parameter groups corresponding to each historical moment form a set corresponding to the power distribution network.

In the above method, preferably, before performing the defuzzification processing on the target fuzzy rule statement, the method further includes:

screening at least two fuzzy rule sentences meeting the logic contradiction condition in the target fuzzy rule sentences according to the interest degree values of the fuzzy rule sentences so as to delete the fuzzy rule sentences meeting the logic contradiction condition, wherein the interest degree values of the fuzzy rule sentences are smaller than or equal to the interest degree threshold value;

wherein the logical contradiction conditions include: the fuzzy sets corresponding to at least one state parameter are different among the fuzzy rule sentences.

Preferably, before the initial fuzzy rule statement is screened according to the minimum confidence value and the minimum support value of the fuzzy rule statement, the method further includes:

merging the same fuzzy rule sentences in the initial fuzzy rule sentences, wherein the same fuzzy rule sentences are as follows: the fuzzy set corresponding to each state parameter and the corresponding membership value of the fuzzy set are the same between the fuzzy rule sentences.

Preferably, the method for obtaining a target parameter group corresponding to the power distribution network monitoring data in a set including a plurality of preset parameter groups includes:

obtaining at least one first state parameter corresponding to the power distribution network monitoring data from a plurality of state parameters in the power distribution network monitoring data;

and according to the first state parameter, obtaining a target parameter group in a set containing a plurality of preset parameter groups, wherein the target parameter group contains at least one second state parameter, and the second state parameter is matched with the first state parameter.

Preferably, in the method, the parameter similarity value satisfies a preset alarm condition, and the method includes:

the parameter similarity value is greater than or equal to a similarity threshold value;

wherein the similarity threshold is obtained by:

obtaining second historical state data, wherein the second historical state data comprises a group of historical data of the power distribution network corresponding to a plurality of historical moments respectively, the historical data of the power distribution network corresponding to each historical moment comprises a plurality of state parameters, and the historical moments are moments selected in the fault-free moment section of the power distribution network;

fuzzifying power distribution network historical data corresponding to each historical moment in the second historical state data to obtain fuzzy data corresponding to each historical moment, wherein the fuzzy data corresponding to each historical moment comprises a fuzzy set corresponding to each state parameter in the power distribution network historical data and a membership value corresponding to the fuzzy set;

respectively carrying out defuzzification processing on the target fuzzy rule sentences to obtain preset parameter groups corresponding to the target fuzzy rule sentences;

obtaining a minimum similarity value between the preset parameter group and the power distribution network historical data in the second historical state data;

and obtaining a similarity threshold according to the minimum similarity value.

Preferably, the method for outputting the alarm result corresponding to the power distribution network monitoring data includes:

outputting a first alarm result, wherein the first alarm result is used for prompting that the state parameters of the power distribution network are abnormal;

and/or the presence of a gas in the gas,

outputting a second alarm result, wherein the second alarm result is used for prompting that the target state parameter in the power distribution network is abnormal; wherein the target state parameter is determined by:

acquiring parameter error data between each state parameter in the power distribution network monitoring data and the corresponding state parameter in the target parameter group;

and determining the state parameter corresponding to the parameter error data meeting the error abnormal condition as the target state parameter.

In the method, preferably, the error abnormal condition includes: the relative error value in the parameter error data is the largest in the parameter error data corresponding to all the state parameters;

wherein the relative error value is: a ratio between a parameter difference and a parameter value of a corresponding state parameter in the target parameter set, the parameter difference being: and the difference value of the parameter value between the state parameter in the power distribution network monitoring data and the corresponding state parameter in the target parameter group.

A distribution network state parameter monitoring device, the device includes:

the first acquisition unit is used for acquiring power distribution network monitoring data of a power distribution network at a target moment, wherein the power distribution network monitoring data comprise a plurality of state parameters;

a second obtaining unit, configured to obtain, in a set including multiple preset parameter groups, a target parameter group corresponding to the power distribution network monitoring data, where the target parameter group includes multiple state parameters, where the preset parameter group in the set is generated based on power distribution network historical data corresponding to the power distribution network at multiple historical times, and the power distribution network historical data corresponding to each historical time includes the multiple state parameters, and the historical time is a time selected within a fault-free time period of the power distribution network;

a third obtaining unit, configured to obtain a parameter similarity value between a state parameter in the target parameter set and a state parameter in the power distribution network monitoring data;

and the output unit is used for outputting an alarm result corresponding to the power distribution network monitoring data under the condition that the parameter similarity value meets a preset alarm condition.

An electronic device, the electronic device comprising:

a memory for storing an application program and data generated by the application program running;

a processor for executing the application to implement: acquiring power distribution network monitoring data of a power distribution network at a target moment, wherein the power distribution network monitoring data comprises a plurality of state parameters;

From the above technical solutions, in a power distribution network state parameter monitoring method, a power distribution network state parameter monitoring device and an electronic device disclosed in the present application, after power distribution network monitoring data including a plurality of state parameters of a power distribution network at a target time is obtained, a target parameter group corresponding to the power distribution network monitoring data and also including a plurality of state parameters is obtained in a set including a plurality of preset parameter groups, where the preset parameter group in the set is generated based on power distribution network historical data corresponding to the power distribution network at a plurality of historical times respectively, and the historical time is a time selected within a fault-free time period of the power distribution network, based on this, by obtaining a parameter similarity value between the state parameters in the target parameter group and the state parameters in the power distribution network monitoring data, and further, under the condition that the parameter similarity value satisfies a preset alarm condition, and outputting an alarm result corresponding to the power distribution network monitoring data. It can be seen that this application is through the parameter similarity between the historical state parameter of obtaining the state parameter in the distribution network monitoring data and distribution network in the no fault period, and then monitors whether the distribution network appears unusually according to the parameter similarity, avoids single and fixed threshold value to monitor and can have the unsafe condition in time measuring, can realize the distribution network monitoring through the similarity comparison between the state parameter in this application to improve the monitoring accuracy.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings without creative efforts.

Fig. 1 is a flowchart of a power distribution network state parameter monitoring method according to an embodiment of the present disclosure;

FIG. 2 is a partial flowchart of a first embodiment of the present application;

FIG. 3 is a diagram showing the function of the triangular membership degree in the present embodiment;

FIG. 4 is a partial flowchart of a first embodiment of the present application;

fig. 5 is a schematic structural diagram of a power distribution network state parameter monitoring device according to a second embodiment of the present application;

fig. 6 is a schematic structural diagram of another power distribution network state parameter monitoring device according to the second embodiment of the present application;

fig. 7 is a schematic structural diagram of another power distribution network state parameter monitoring device according to the second embodiment of the present application;

fig. 8 is a schematic structural diagram of an electronic device according to a third embodiment of the present application;

fig. 9 is a schematic block diagram of a power distribution network state parameter monitoring scheme provided in an embodiment of the present application.

Detailed Description

At present, when monitoring the fault state of the power distribution network, a fixed threshold is usually set for the collected current and voltage data of the power distribution network to determine whether the state parameters such as current or voltage are abnormal, and the monitoring efficiency and the fault judgment standard of the method are rough.

In order to solve the problems, the inventor of the application provides a power distribution network state parameter monitoring method through further research, so that the monitoring efficiency can be improved, and the fault state of a power distribution network can be monitored more accurately. The method comprises the following specific steps:

firstly, acquiring power distribution network monitoring data of a power distribution network at a target moment, wherein the power distribution network monitoring data comprises a plurality of state parameters; then, obtaining a target parameter group corresponding to the power distribution network monitoring data in a set comprising a plurality of preset parameter groups, wherein the target parameter group comprises a plurality of state parameters, the preset parameter groups in the set are generated based on power distribution network historical data corresponding to the power distribution network at a plurality of historical moments respectively, the power distribution network historical data corresponding to each historical moment comprises a plurality of state parameters, and the historical moments are moments selected in the fault-free moment section of the power distribution network; and finally, acquiring a parameter similarity value between the state parameter in the target parameter group and the state parameter in the power distribution network monitoring data, and outputting an alarm result corresponding to the power distribution network monitoring data under the condition that the parameter similarity value meets a preset alarm condition.

Therefore, the parameter similarity of the power distribution monitoring data and the target parameter group is obtained by obtaining the state parameters in the power distribution network monitoring data and the state parameters in the set of the target parameter group, and the state is judged according to the parameter similarity, so that normal state parameter reasoning and state judgment monitoring early warning of the power distribution network are realized.

The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and 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 application.

As shown in fig. 1, a method for monitoring a power distribution network state parameter provided in an embodiment of the present invention may be applied to an electronic device capable of performing data processing, such as a computer or a server. The electronic device in the application can be an edge computing device in a power distribution network area, such as a transformer area intelligent distribution device. The technical scheme in the embodiment is mainly used for judging the state according to the parameter similarity based on the acquired power distribution monitoring data and the parameter similarity of the state parameters of the power distribution network in the fault-free state when the fault state of the power distribution network is monitored, so that the monitoring and early warning of the state of the power distribution network are realized.

In a specific implementation, the method in this embodiment may include the following steps:

step 101: and acquiring the power distribution network monitoring data of the power distribution network at the target moment.

The power distribution network monitoring data comprise a plurality of state parameters. The state parameters refer to parameters corresponding to the distribution network on corresponding distribution project, such as three-phase voltage, current, active power, reactive power, frequency, voltage-current unbalance, frequency deviation, transformer area load rate and other state parameters of the distribution network on the low-voltage side.

The target time may be the current real-time or a certain time in the past historical time period.

In specific implementation, in this embodiment, power distribution network monitoring data of the power distribution network at a target time may be acquired by using the power distribution network device, for example, the current transformer and the voltage transformer are used to acquire state parameters of three-phase voltage, current, active power, reactive power, frequency, voltage-current imbalance, frequency deviation, transformer area load rate, and the like of the power distribution network at the current real-time.

Step 102: and obtaining a target parameter group corresponding to the power distribution network monitoring data in a set containing a plurality of preset parameter groups.

The target parameter group comprises a plurality of state parameters.

The preset parameter group in the set is generated based on historical data of the power distribution network corresponding to the power distribution network at a plurality of historical moments respectively, the historical data of the power distribution network corresponding to each historical moment comprises a plurality of state parameters, and the historical moments are moments selected in the fault-free moment section of the power distribution network;

in a specific implementation, in this embodiment, a plurality of historical moments may be selected in a power distribution network failure-free time period, and then, historical data of the power distribution network corresponding to each historical moment is collected in a historical database, where the historical data of the power distribution network corresponding to each historical moment includes a plurality of state parameters, a plurality of preset parameter groups are obtained based on the state parameters in the historical data of the power distribution network, and each obtained preset parameter group includes a plurality of state parameters. Based on this, in step 102, a target parameter set matched with the power distribution network monitoring data is obtained from the preset parameter sets. For example, if no fault occurs in the distribution network in 2018, historical data of the distribution network corresponding to two historical moments of month 1 and month 2 in 2018 may be selected and a plurality of preset parameter sets may be generated based on the historical data, where the historical data of the distribution network corresponding to month 1 in 2018 includes state parameters such as three-phase voltage and current of the low-voltage side of the distribution transformer, and the historical data of the distribution network corresponding to month 2 in 2018 includes state parameters such as active power and reactive power. In the application, the preset parameter groups corresponding to each historical moment are obtained respectively based on the historical data of the power distribution network, the preset parameter groups at the moment are data groups consisting of state parameters in a normal state, namely a fault-free state of the power distribution network in 2018, and based on the preset parameter groups, the target parameter groups corresponding to the monitoring data of the power distribution network are obtained, and the target parameter groups also comprise a plurality of state parameters. It should be noted that the state parameters in the target parameter group are matched with the state parameters in the power distribution network monitoring data in category.

Step 103: and obtaining a parameter similarity value between the state parameters in the target parameter group and the state parameters in the power distribution network monitoring data.

In a specific implementation, in this embodiment, a similarity function may be constructed to obtain a parameter similarity value between a state parameter in the target parameter set and a state parameter in the power distribution network monitoring data.

It should be noted that the parameter similarity value between the state parameter in the target parameter group and the state parameter in the power distribution network monitoring data may be understood as a similarity value between the target parameter group and the power distribution network monitoring data with respect to the state parameter included in each of the target parameter group and the power distribution network monitoring data, and the parameter similarity value represents a degree of similarity between the target parameter group and the power distribution network monitoring data.

Step 104: and outputting an alarm result corresponding to the power distribution network monitoring data under the condition that the parameter similarity value meets a preset alarm condition.

The parameter similarity value represents the similarity between the target parameter group and the power distribution network monitoring data, the alarm condition represents a condition capable of triggering an alarm mechanism, and correspondingly, the alarm mechanism is triggered under the condition that the parameter similarity value meets the alarm condition, and the information of the alarm result output at the moment can include the information that the power distribution network monitoring data fails and the information that one or more state parameters in the power distribution network monitoring data fail.

In specific implementation, the alarm condition in this embodiment may be: the parameter similarity value is greater than or equal to a similarity threshold. Correspondingly, in this embodiment, when the parameter similarity value is greater than or equal to the similarity threshold value, an alarm result corresponding to the power distribution network monitoring data is output. For example, if the preset alarm condition is that the parameter similarity value is greater than or equal to 0.8 and the obtained parameter similarity value is 0.9, the obtained parameter similarity value meets the preset alarm condition, that is, the similarity between the power distribution network monitoring data and the target parameter set generated according to the state parameters of the power distribution network in the normal state is high, a power distribution network fault exists, an alarm result is output at this time, and the alarm result is at least that the voltage state parameters in the power distribution network detection data have faults.

According to the above technical scheme, the method for monitoring the state parameter of the power distribution network provided by the embodiment of the application, after obtaining the power distribution network monitoring data including a plurality of state parameters of the power distribution network at the target moment, obtaining a target parameter group which also comprises a plurality of state parameters and corresponds to the power distribution network monitoring data from a set comprising a plurality of preset parameter groups, the preset parameter group in the set is generated based on historical data of the power distribution network corresponding to a plurality of historical moments of the power distribution network respectively, the historical moments are moments selected in the fault-free moment section of the power distribution network, based on the moment, the parameter similarity value between the state parameters in the target parameter group and the state parameters in the monitoring data of the power distribution network is obtained, and further outputting an alarm result corresponding to the power distribution network monitoring data under the condition that the parameter similarity value meets a preset alarm condition. It can be seen that this application is through the parameter similarity between the historical state parameter of obtaining the state parameter in the distribution network monitoring data and distribution network in the no fault period, and then monitors whether the distribution network appears unusually according to the parameter similarity, avoids single and fixed threshold value to monitor and can have the unsafe condition in time measuring, can realize the distribution network monitoring through the similarity comparison between the state parameter in this application to improve the monitoring accuracy.

Based on the method disclosed in fig. 1 in the embodiment of the present application, if the parameter similarity is to be known, a set corresponding to the power distribution network needs to be obtained in advance, and a specific implementation process is shown in fig. 2:

step 201: first historical state data is obtained.

The first historical state data comprises a group of historical data of the power distribution network corresponding to a plurality of historical moments respectively, the historical data of the power distribution network corresponding to each historical moment comprises a plurality of state parameters, and the historical moments are moments selected in the fault-free moment section of the power distribution network.

In specific implementation, in this embodiment, a plurality of historical moments may be selected in the power distribution network failure-free time period, a front part of the historical moments in the failure-free time period may be selected, a rear part of the historical moments in the failure-free time period may be selected, a middle part of the historical moments in the failure-free time period is optimally selected, and a plurality of state parameters corresponding to the historical moments are acquired as first historical state data, where the first historical state data includes a plurality of power distribution network historical state parameters corresponding to the plurality of historical moments respectively. For example, the power distribution network in 2017 has no fault, 6 months and 7 months in 2017 are selected as a first historical time period, current state parameters and voltage state parameters at multiple moments in 6 months in 2017 are obtained, multiple state parameters corresponding to each moment in the obtained state parameters form a group of power distribution network historical data corresponding to the moment, frequency state parameters and active power state parameters at multiple moments in 7 months in 2017 are obtained, and multiple state parameters corresponding to each moment in the obtained state parameters form a group of power distribution network historical data corresponding to the moment. And forming first historical state data by the multiple groups of power distribution network historical data.

Step 202: and fuzzifying the historical data of the power distribution network corresponding to each historical moment in the first historical state data to obtain fuzzy data corresponding to each historical moment.

The fuzzy data corresponding to each historical moment comprises a fuzzy set corresponding to each state parameter in the historical data of the power distribution network and a membership value corresponding to the fuzzy set.

Specifically, in this embodiment, the fuzzy controller may be used to perform fuzzification processing on historical data of the power distribution network, so as to obtain fuzzy data corresponding to each historical time.

Wherein, the fuzzification refers to a process that the fuzzy controller converts the determined value of the input quantity of the fuzzy controller into a corresponding fuzzy language variable value. The fuzzification processing in this embodiment is to convert the historical data of the power distribution network into output of fuzzy language variable values of different state parameters in the normal state parameter domain.

For example, in the present embodiment, fuzzy domains [ -6, 6] are preset and divided into 7 stages, each stage determines a fuzzy set, each fuzzy set corresponds to linguistic variables sequentially representing NB (negative large), NM (negative medium), NS (negative small), Z (zero), PS (positive small), PM (positive medium) and PB (positive large), and is represented by a triangular membership function, as shown in fig. 3, in the figure, the triangular membership function indicates NB, NM, NS, Z, PS, PM and PB, the first fuzzy set refers to a set of values from 0 to c2, the second fuzzy set refers to a set of values from c1 to c3, the third fuzzy set refers to a set of values from c2 to c4, the fourth fuzzy set refers to a set of values from c3 to c5, the fifth fuzzy set refers to a set of values from c4 to c6, the sixth fuzzy set refers to a set of values from c5 to c7, the seventh fuzzy set refers to the value set from c6 to a certain value later, wherein c1-c7 represents the center point of each fuzzy set of the fuzzy domain.

In a specific implementation, in this embodiment, the fuzzy processing of the hierarchical fuzzy set method may be performed on the power distribution network historical data corresponding to each historical time in the first historical state data to obtain fuzzy data corresponding to each historical time, where the fuzzy data corresponding to each historical time includes a fuzzy set corresponding to each state parameter in the power distribution network historical data and a membership value corresponding to the fuzzy set. For example, in this embodiment, the voltage state parameters and the current state parameters of the power distribution network corresponding to a certain historical time in the 5 th month in 2018 are respectively subjected to fuzzification processing by a stepped fuzzy set method, so as to obtain fuzzy data corresponding to the 5 th month in 2018, where the fuzzy data include fuzzy sets corresponding to the voltage state parameters and the current state parameters, and membership values corresponding to the fuzzy sets where the voltage state parameters and the current state parameters are located.

Step 203: and generating an initial fuzzy rule statement corresponding to each historical moment according to the fuzzy set and the membership value in the fuzzy data corresponding to each historical moment.

In a specific implementation, in this embodiment, according to the fuzzy set and the membership value in the fuzzy data corresponding to each historical time, an initial fuzzy rule statement corresponding to each power distribution and transformation state parameter at each historical time is obtained through mining and reasoning. For example, according to a fuzzy set and a membership value in fuzzy data corresponding to a certain historical moment in the month 1 of 2018, an initial fuzzy rule statement corresponding to the current state parameter and the voltage state parameter corresponding to the historical moment in the month 1 of 2018 is obtained through mining and reasoning.

The initial fuzzy rule statement can be understood as the association relation among various state parameters in the historical data of the power distribution network or the description statement data of the association rule. For example, the association rule may employ If-Then statements, described as follows:

If X₁＝X₁ ⁱ，

Then X₂＝X₂ ⁱ,X₃＝X₃ ⁱ,…,X_n＝X_n ⁱ

wherein: x₁-X_nRepresenting the state parameters, X, of the distribution substation of n items at the same time_2～nRepresents by X₁The external distribution substation state parameters, n represents the number of the distribution substation state parameters at the same moment; i represents the sequence number of the association rule.

For example, the initial fuzzy rule statement corresponding to the current state parameter and the voltage state parameter at a certain historical time a in month 1 of 2018 may be:

If X₁＝I₁ ^A，

Then X₂＝U₂ ^A；

wherein, I₁ ^AIs a fuzzy set, U, corresponding to the current state parameter corresponding to the historical time A₂ ^AAnd the fuzzy sets correspond to the voltage state parameters corresponding to the historical time A.

Step 204: and at least screening the initial fuzzy rule statement according to the minimum confidence value and the minimum support value of the fuzzy rule statement to obtain a target fuzzy rule statement of which the minimum confidence value is greater than or equal to the confidence threshold and the minimum support value is greater than or equal to the support threshold.

Wherein, the confidence degree refers to the probability of B occurring under the condition that A occurs in all transactions, and the support degree refers to the probability of A and B occurring simultaneously in all transactions. In the present embodiment, the association rule uses the minimum support s and the minimum confidence c as selection criteria, and the expressions of the minimum support s and the minimum confidence c are as shown in formula (1) and formula (2):

in the above formula, μ (X) represents X (X)₁To X_n) D represents the total number of transactions in the X dataset. In the application, the minimum support degree and the minimum confidence degree of each initial fuzzy rule statement can be obtained through the above manner, and then the initial fuzzy rule statements are screened according to the preset confidence degree threshold value and the support degree threshold value, and the screened initial fuzzy rule statements have to reach the minimum support degree and the minimum confidence degree, as shown in formula (3):

s(X₁∪X_2～n)≥s_min,c(X₁∪X_2～n)≥c_minformula (3)

Wherein s is_minTo support threshold, c_minIs a confidence threshold.

In a specific implementation, in this embodiment, at least the initial fuzzy rule statement is screened according to the minimum confidence value and the minimum support value of the fuzzy rule statement, and a target fuzzy rule statement with the minimum confidence value greater than or equal to the confidence threshold and the minimum support value greater than or equal to the support threshold is screened out. For example, if the confidence threshold and the support threshold are 0.3 and 0.6, respectively, then the minimum support and the minimum confidence in each initial fuzzy rule statement are compared with the corresponding support threshold and confidence threshold, respectively, to implement the screening of the initial fuzzy rule statements, and then the minimum support s in the initial fuzzy rule statement can be obtained_minGreater than or equal toAt a support threshold and a minimum confidence level c_minTarget fuzzy rule statements greater than or equal to the confidence threshold.

Step 205: and respectively carrying out defuzzification processing on the target fuzzy rule sentences to obtain a preset parameter group corresponding to each target fuzzy rule sentence.

And the preset parameter group corresponding to each historical moment forms a set corresponding to the power distribution network. Defuzzification refers to the conversion of fuzzy values inferred into numerical values of explicit state parameters.

In specific implementation, in this embodiment, the fuzzy rule statements obtained by the preceding fuzzification processing are subjected to defuzzification processing, and numerical data of different state parameters in a normal state of each target fuzzy rule statement are obtained through conversion and stored in an xml configuration file, that is, a set corresponding to the power distribution network.

In one implementation, before performing the defuzzification processing on the target fuzzy rule statement in the fuzzy rule statement set, step 205 further includes the following steps:

and screening at least two fuzzy rule sentences which meet logic contradiction conditions in the target fuzzy rule sentences in the fuzzy rule sentence set according to the interest values of the fuzzy rule sentences so as to delete the target fuzzy rule sentences which meet the logic contradiction conditions and obtain the target fuzzy rule sentences of which the interest values are less than or equal to the interest threshold, and finally obtaining the target fuzzy rule sentences which are reserved after screening.

Wherein the logical contradiction conditions include: the fuzzy sets corresponding to at least one state parameter are different among the fuzzy rule sentences. For example, in the target fuzzy rule statement a and the target fuzzy rule statement B, the fuzzy sets corresponding to the current state parameters are matched, if the fuzzy sets are both in the positive small fuzzy set, but the fuzzy sets corresponding to the voltage state parameters are respectively in the positive small fuzzy set and the positive middle fuzzy set, so that the fuzzy sets corresponding to the target fuzzy rule statement a and the target fuzzy rule statement B in the voltage state parameters are different, which is not logical, and therefore, one of the target fuzzy rule statement a and the target fuzzy rule statement B needs to be deleted. Specifically, one of the target fuzzy rule statement a and the target fuzzy rule statement B having an interest value smaller than or equal to the interest threshold may be deleted.

Specifically, the interestingness I is a measure for characterizing the attention degree of the rule, the larger the interestingness value is, the better the parameter relevance is under the guidance of the rule, and the expression is as in formula (4):

based on this, in this embodiment, for two or more target fuzzy rule statements that satisfy the logical contradiction condition, the interest value of each target fuzzy rule statement is obtained, and the fuzzy rule statements whose interest values are smaller than or equal to the interest threshold value in the target fuzzy rule statements are deleted, for example, there are three target fuzzy rule statements with different fuzzy sets corresponding to at least one state parameter, and the three target fuzzy rule statements are screened according to the interest values of the fuzzy rule statements, where the interest threshold value of the fuzzy rule statement is 0.5, and the target fuzzy rule statements whose interest values are smaller than or equal to 0.5 are deleted.

In one implementation, step 204 further includes the following method before the initial fuzzy rule statement is filtered according to the minimum confidence value and the minimum support value of the fuzzy rule statement:

merging the same fuzzy rule sentences in the initial fuzzy rule sentences, wherein the same fuzzy rule sentences are as follows: the corresponding fuzzy sets on each state parameter and the corresponding membership values on the fuzzy sets are the same among the fuzzy rule sentences.

In a specific implementation, in the initial fuzzy rule statement, the fuzzy sets corresponding to the fuzzy rule statements in the initial fuzzy rule statement on each state parameter and the initial fuzzy rule statements with the same membership value on the fuzzy sets are merged, for example, the fuzzy sets corresponding to the voltage state parameters in the two initial fuzzy rule statements and the membership values corresponding to the fuzzy sets are NB, and then the two initial fuzzy rule statements are merged, that is, one of the initial fuzzy rule statements is deleted.

In one implementation, when the target parameter group is obtained in step 102, the target parameter group matched with the power distribution network monitoring data may be determined in a plan according to the state parameters in the power distribution network monitoring data. The method comprises the following specific steps:

firstly, obtaining at least one first state parameter corresponding to the power distribution network monitoring data from a plurality of state parameters in the power distribution network monitoring data; the first state parameter herein refers to a state parameter having a flag in the power distribution network monitoring data, for example, among state parameters such as low-voltage side three-phase voltage, current, active power, load factor, reactive power, and frequency in the power distribution network monitoring data, the voltage, the current, the frequency, and the load factor are identification variables, and at this time, the first state parameter corresponding to these identification variables is obtained, that is, the voltage, the current, the frequency, and the load factor are obtained as the first state parameter.

Then, the target parameter group may be obtained in a set including a plurality of preset parameter groups according to the first state parameter.

The target parameter group comprises at least one second state parameter, wherein the second state parameter refers to a state parameter with a sign in the parameter group, and the second state parameter is matched with the first state parameter. In other words, the selected target parameter group has a second state parameter group consistent with the first state parameter in the power distribution network monitoring data. It should be noted that, the matching between the first state parameter and the second state parameter means that the types of the state parameters corresponding to the first state parameter and the second state parameter are the same, but the parameter values of the state parameters are not the same. It should be noted that the first status parameter and the second status parameter are not sequentially divided, but only represent two status parameters.

In a specific implementation, in this embodiment, at least two of the plurality of state parameters in the power distribution network monitoring data are obtained as first state parameters of the power distribution network monitoring data, then, in a set including a plurality of preset parameter groups, a target parameter group is obtained, then, at least two of the target parameter groups are obtained as second state data of the target parameter group, and the first state parameters are matched with the second state parameters, for example, the plurality of state parameters in the power distribution network monitoring data include state parameters of low-voltage side three-phase voltage, current, active power, reactive power, and frequency, identification variables such as low-voltage side three-phase voltage, current, and active power are obtained from the plurality of state parameters in the power distribution network monitoring data as first state parameters of the power distribution network monitoring data, in the set including the plurality of preset parameter groups, the target parameter group is obtained, each preset parameter group set in the set contains state parameters such as low-voltage side three-phase voltage, current, active power, frequency deviation and transformer area load rate (the state parameter types in each preset parameter group can be the same or different), then the preset parameter group containing the state parameters such as the low-voltage side three-phase voltage, the current and the active power is obtained and is a target parameter group, the state parameters in the target parameter group are second state parameters, and at the moment, the first state parameters are matched with the second state parameters.

In one implementation, step 104 may obtain whether the parameter similarity value satisfies a preset alarm condition by comparing the parameter similarity value with a similarity threshold value, for example, in a case that the parameter similarity value is less than or equal to the similarity threshold value, it may be determined that the alarm condition is satisfied, and in a case that the alarm condition is satisfied, a corresponding alarm result is output.

Wherein, the similarity threshold may be obtained by the following manner, as shown in fig. 4:

step 401: second historical state data is obtained.

The second historical state data comprises a group of historical data of the power distribution network corresponding to a plurality of historical moments respectively, the historical data of the power distribution network corresponding to each historical moment comprises a plurality of state parameters, and the historical moments are moments selected in the fault-free moment section of the power distribution network.

In a specific implementation, in this embodiment, a plurality of historical moments may be selected in the power distribution network failure-free time period, a front part of the historical moments in the failure-free time period may be selected, a rear part of the historical moments in the failure-free time period may be selected, a middle part of the historical moments in the failure-free time period is optimally selected, and a plurality of state parameters corresponding to the historical moments are acquired as second historical state data, where the second historical state data includes a plurality of power distribution network historical state parameters corresponding to the plurality of historical moments respectively. For example, the power distribution network in 2019 has no fault, 6 months and 7 months in 2019 are selected as a second historical time period, voltage and current unbalance degree and frequency deviation state parameters at multiple moments in 6 months in 2019 are obtained, the obtained state parameters form a group of power distribution network historical data, current and active power state parameters at multiple moments in 7 months in 2019 are obtained, and the obtained state parameters form a group of power distribution network historical data. And forming second historical state data by the multiple groups of power distribution network historical data.

Step 402: and fuzzifying the power distribution network historical data corresponding to each historical moment in the second historical state data to obtain fuzzy data corresponding to each historical moment.

The fuzzy data corresponding to each historical moment comprises a fuzzy set corresponding to each state parameter in the historical data of the power distribution network and a membership value corresponding to the fuzzy set;

specifically, the fuzzy processing of the hierarchical fuzzy set method may be performed on the power distribution network historical data corresponding to each historical time in the second historical state data to obtain fuzzy data corresponding to each historical time, where the fuzzy data corresponding to each historical time includes a fuzzy set corresponding to each state parameter in the power distribution network historical data and a membership value corresponding to the fuzzy set.

The implementation manner of the step 402 in implementing the fuzzification processing may refer to the implementation manner of the step 202 in the foregoing, and is not described in detail here.

Step 403: and generating an initial fuzzy rule statement corresponding to each historical moment according to the fuzzy set and the membership value in the fuzzy data corresponding to each historical moment.

Specifically, according to the fuzzy set and the membership value in the fuzzy data corresponding to each historical moment, the initial fuzzy rule statement corresponding to each power distribution and transformation state parameter in each historical moment is obtained through mining and reasoning.

The implementation manner in step 403 when generating the initial fuzzy statement may refer to the implementation manner in step 203, and is not described in detail here.

Step 404: and at least screening the initial fuzzy rule statement according to the minimum confidence value and the minimum support value of the fuzzy rule statement to obtain a target fuzzy rule statement of which the minimum confidence value is greater than or equal to the confidence threshold and the minimum support value is greater than or equal to the support threshold.

Specifically, at least the initial fuzzy rule sentences are screened according to the minimum confidence value and the minimum support value of the fuzzy rule sentences, and target fuzzy rule sentences with the minimum confidence value larger than or equal to the confidence threshold value and the minimum support value larger than or equal to the support threshold value are screened out.

The implementation manner of the step 404 in the filtering of the minimum confidence value and the minimum support value may refer to the implementation manner of the step 204, and is not described in detail here.

Step 405, performing defuzzification processing on the target fuzzy rule statements respectively to obtain a preset parameter group corresponding to each target fuzzy rule statement.

The implementation manner of step 405 in implementing the defuzzification processing may refer to the implementation manner of step 205 in the foregoing, and is not described in detail here.

Step 406: and obtaining the minimum similarity value between the preset parameter group and the historical data of the power distribution network in the second historical state data.

Wherein, similarity refers to a measure of similarity between two sets of state parameters. In this embodiment, a euclidean distance calculation method in a vector space distance calculation algorithm may be used to obtain a similarity value between each preset parameter group and corresponding power distribution network historical data, and a minimum similarity value is obtained based on the similarity value. Wherein, the Euclidean distance formula is as the formula (5):

in the above formula, d_ijParameter set x representing real-time monitoring state_i(i.e. power distribution network historical data) and the corresponding normal state parameter set x in the xml file (set)_j(i.e., the preset parameter set), n represents the dimension of the state parameter set, k represents the parameter number in the state parameter set, x_ikAnd x_jkRespectively representing the normal state parameter set x_iAnd monitoring state parameter group x_jThe corresponding kth parameter element. In engineering applications, S_ij＝1/(1+d_ij) Is often used in the standardized Euclidean distance range to [0,1 ]]. Thus, for evaluating the normal state parameter group x_iAnd monitoring state parameter group x_jThe similarity function between them is set as formula (6):

in a specific implementation, the preset parameter group obtained in step 305 is used as a minimum similarity value calculation basis, and a minimum similarity value between the preset parameter group and the power distribution network historical data in the second historical state data is obtained as S_m. Alternatively, in this embodiment, the preset parameter group (i.e., the normal state parameter data) of the power distribution network in the xml file obtained in step 205 may be used as a minimum similarity value calculation basis, and the minimum similarity value between the preset parameter group and the power distribution network historical data in the first historical state data is obtained as S_m。

For example, in this embodiment, a value (380 v) and a value (50 hz) of a frequency of a distribution network preset parameter group normal state parameter distribution transformation low-voltage side three-phase voltage in an xml configuration file are obtained as judgment criteria, and when the state parameters in the distribution network history data are obtained: after the three-phase voltage 450v and the frequency 45hz of the low-voltage side of the distribution network are changed, the state parameters in the historical data of the distribution network are as follows: and performing similarity calculation on the distribution transformer low-voltage side three-phase voltage 450v and the frequency 45hz and the distribution transformer low-voltage side three-phase voltage 380v and the frequency 50hz in the preset parameter group obtained by fuzzy inference to obtain a minimum similarity value of 0.8 between the preset parameter group and the power distribution network historical data in the second historical state data.

Step 407: and obtaining a similarity threshold according to the minimum similarity value.

In specific implementation, the minimum similarity S is set in this embodiment_mAnd a threshold coefficient k_tIs set as a similarity threshold S_t(also referred to as a similarity warning threshold), as in equation (7):

S_t＝k_t·S_mformula (7)

In the above formula, the threshold coefficient k_tThe value can be set to a suitable value according to the monitoring sensitivity requirement in practical application, for example, the minimum similarity S_mIs 0.4, the threshold coefficient k_tkt is 2, then a similarity threshold value of 0.8 is obtained, based on which the similarity value of the parameter is found to be less than or equal to the similarity threshold value S in this embodiment_tAnd outputting an alarm result.

In one implementation, when outputting the alarm result corresponding to the power distribution network monitoring data, step 104 may be implemented by:

outputting a first alarm result, wherein the first alarm result is used for prompting that the state parameters of the power distribution network are abnormal; and/or outputting a second alarm result, wherein the second alarm result is used for prompting that the target state parameter in the power distribution network is abnormal.

When the first alarm result and the second alarm result are output simultaneously, the output of the first alarm result and the output of the second alarm result do not have the priority, or the first alarm result is output first and then the second alarm result is output.

In specific implementation, in this embodiment, there are three situations of outputting the alarm result, and only the first alarm result may be output to prompt that the state parameter of the power distribution network is abnormal, such as the power distribution network is abnormal; or only outputting a second alarm result to prompt that the target parameter in the power distribution network is abnormal, for example, the voltage state parameter is abnormal; and a first alarm result and a second alarm result can be output, wherein the first alarm result is used for prompting that the state parameter of the power distribution network is abnormal, and the second alarm result is used for prompting that the target parameter in the power distribution network is abnormal, such as the power distribution network is abnormal and the abnormality occurs on the voltage state parameter.

The first embodiment is as follows: and outputting the first alarm result to the distribution and transformation equipment platform and the database, and prompting that the state parameters of the power distribution network are abnormal at the target moment by the first alarm result output to the distribution and transformation equipment platform and the database.

Example two: and outputting the second alarm result to the distribution and transformation equipment platform and the database, and prompting the abnormity of the voltage state parameter and the current state parameter in the state parameters of the power distribution network by the second alarm result output to the distribution and transformation equipment platform and the database.

Example three: and outputting the first alarm result and the second alarm result to a distribution and transformation equipment platform and a database, wherein the first alarm result output to the distribution and transformation equipment platform and the database prompts that the state parameters of the power distribution network are abnormal at a target moment, and meanwhile, the second alarm result prompts that the voltage state parameters and the current state parameters in the state parameters of the power distribution network are abnormal.

Wherein the target state parameter is determined by:

and acquiring parameter error data between each state parameter in the power distribution network monitoring data and the corresponding state parameter in the target parameter group, and determining the state parameter corresponding to the parameter error data meeting the error abnormal condition as the target state parameter.

Wherein the error exception condition comprises:

the relative error value in the parameter error data is the largest in the parameter error data corresponding to each state parameter, wherein the relative error value is as follows: the ratio of the parameter difference to the parameter value of the corresponding state parameter in the target parameter group is as follows: and the difference value of the parameter value between the state parameter in the power distribution network monitoring data and the corresponding state parameter in the target parameter group.

In the specific implementation, in this embodiment, each state parameter in the power distribution network monitoring data is compared with the corresponding normal state parameter in the target parameter set, a difference between the two state parameters is obtained first, and then the difference is divided by the corresponding normal state parameter in the target parameter set, so as to obtain the relative error, after obtaining the relative errors, the multiple relative errors in the group of state parameters at the same time are compared, the state parameter with the maximum relative error is determined as the target state parameter, for example, the current value in the obtained power distribution network monitoring data is 500A, the voltage value is 360V, the current value of the corresponding normal state parameter in the target parameter set is 600A, the voltage value is 380V, the current value in the normal state parameter 600A is subtracted by the current value 500A in the power distribution network monitoring data, and then the difference 100A between the two state parameters is divided by the current value 600A in the normal state parameter, so as to obtain the relative error of 0.167, and subtracting the voltage value 360V in the monitoring data of the power distribution network from the voltage value 380V of the normal state parameter, dividing the difference value of the two by the voltage value 380V of the normal state parameter to obtain a relative error of 0.053, and finally determining the current state parameter with the maximum relative error as the target state parameter.

Referring to fig. 5, a schematic structural diagram of a power distribution network state parameter monitoring device provided in the second embodiment of the present application is shown, the device can be configured in an electronic device capable of performing data processing, the technical scheme in the present application is mainly used for monitoring faults and early warning of power distribution, an intelligent device and a general software platform with an edge computing function are created, a power distribution network fault monitoring system is deployed in a power distribution network device software platform in a form of software APP, so as to implement power distribution network running state monitoring and fault real-time early warning functions, thereby improving monitoring efficiency, and making fault judgment standards more accurate.

Specifically, the apparatus may include the following units:

the first obtaining unit 501 is configured to obtain power distribution network monitoring data of a power distribution network at a target moment, where the power distribution network monitoring data includes multiple state parameters;

a second obtaining unit 502, configured to obtain a target parameter group corresponding to the power distribution network monitoring data in a set including multiple preset parameter groups, where the target parameter group includes multiple state parameters, where the preset parameter group in the set is generated based on power distribution network historical data corresponding to the power distribution network at multiple historical times, and the power distribution network historical data corresponding to each historical time includes the multiple state parameters, and the historical time is a time selected within a fault-free time period of the power distribution network;

the second obtaining unit 402 is specifically configured to: obtaining at least one first state parameter corresponding to the power distribution network monitoring data from a plurality of state parameters in the power distribution network monitoring data; and according to the first state parameter, obtaining a target parameter group from a set containing a plurality of preset parameter groups, wherein the target parameter group contains at least one second state parameter, and the second state parameter is matched with the first state parameter.

A third obtaining unit 503, configured to obtain a parameter similarity value between a state parameter in the target parameter group and a state parameter in the power distribution network monitoring data;

the output unit 504 is configured to output an alarm result corresponding to the power distribution network monitoring data when the parameter similarity value meets a preset alarm condition.

The condition that the parameter similarity value in the output unit 504 meets the preset alarm condition is specifically used for: the parameter similarity value is greater than or equal to a similarity threshold value; wherein the similarity threshold is obtained by: acquiring second historical state data, wherein the second historical state data comprises a group of historical data of the power distribution network corresponding to a plurality of historical moments respectively, the historical data of the power distribution network corresponding to each historical moment comprises a plurality of state parameters, and the historical moments are moments selected in the fault-free moment section of the power distribution network; fuzzifying the power distribution network historical data corresponding to each historical moment in the second historical state data to obtain fuzzy data corresponding to each historical moment, wherein the fuzzy data corresponding to each historical moment comprises a fuzzy set corresponding to each state parameter in the power distribution network historical data and a membership value corresponding to the fuzzy set; generating an initial fuzzy rule statement corresponding to each historical moment according to the fuzzy set and the membership value in the fuzzy data corresponding to each historical moment; screening at least the initial fuzzy rule statement according to the minimum confidence value and the minimum support value of the fuzzy rule statement to obtain a target fuzzy rule statement of which the minimum confidence value is greater than or equal to a confidence threshold value and the minimum support value is greater than or equal to a support threshold value; respectively carrying out defuzzification processing on the target fuzzy rule sentences to obtain preset parameter groups corresponding to each target fuzzy rule sentence; obtaining a minimum similarity value between the preset parameter group and the historical data of the power distribution network in the second historical state data; and obtaining a similarity threshold according to the minimum similarity value.

When the output unit 504 outputs the alarm result corresponding to the power distribution network monitoring data, the following implementation manners may be adopted:

outputting a first alarm result, wherein the first alarm result is used for prompting that the state parameters of the power distribution network are abnormal; and/or outputting a second alarm result, wherein the second alarm result is used for prompting that the target state parameter in the power distribution network is abnormal; wherein the target state parameter is determined by: acquiring parameter error data between each state parameter in the power distribution network monitoring data and the corresponding state parameter in the target parameter group; determining the state parameters corresponding to the parameter error data meeting the error abnormal conditions as target state parameters, wherein the error abnormal conditions comprise: the relative error value in the parameter error data is the largest in the parameter error data corresponding to all the state parameters; wherein, the relative error value is: the ratio of the parameter difference to the parameter value of the corresponding state parameter in the target parameter group is as follows: and the difference value of the parameter value between the state parameter in the power distribution network monitoring data and the corresponding state parameter in the target parameter group.

According to the scheme, the second embodiment of the present application provides a power distribution network state parameter monitoring device, after obtaining the power distribution network monitoring data including a plurality of state parameters of the power distribution network at the target moment, by obtaining a target parameter group which corresponds to the monitoring data of the power distribution network and also comprises a plurality of state parameters from a set comprising a plurality of preset parameter groups, the preset parameter group in the set is generated based on historical data of the power distribution network corresponding to a plurality of historical moments of the power distribution network respectively, the historical moments are moments selected in the fault-free moment section of the power distribution network, based on the moment, the parameter similarity value between the state parameters in the target parameter group and the state parameters in the monitoring data of the power distribution network is obtained, and further outputting an alarm result corresponding to the power distribution network monitoring data under the condition that the parameter similarity value meets a preset alarm condition. It can be seen that this application is through the parameter similarity between the historical state parameter of obtaining the state parameter in the distribution network monitoring data and distribution network in the no fault period, and then monitors whether the distribution network appears unusually according to the parameter similarity, avoids single and fixed threshold value to monitor and can have the unsafe condition in time measuring, can realize the distribution network monitoring through the similarity comparison between the state parameter in this application to improve the monitoring accuracy.

Referring to fig. 6, the apparatus in the second embodiment of the present application further includes a fourth obtaining unit 505, configured to obtain a set corresponding to the power distribution network in advance, where the fourth obtaining unit 505 specifically includes the following structure, as shown in fig. 7:

the historical data obtaining module 701 is configured to obtain first historical state data, where the first historical state data includes a group of power distribution network historical data corresponding to a plurality of historical moments, the power distribution network historical data corresponding to each historical moment includes a plurality of state parameters, and the historical moment is a moment selected within a fault-free moment segment of the power distribution network.

The fuzzy data obtaining module 702 is configured to perform fuzzification processing on the power distribution network historical data corresponding to each historical time in the first historical state data to obtain fuzzy data corresponding to each historical time, where the fuzzy data corresponding to each historical time includes a fuzzy set corresponding to each state parameter in the power distribution network historical data and a membership value corresponding to the fuzzy set.

The statement obtaining module 703 is configured to generate an initial fuzzy rule statement corresponding to each historical time according to the fuzzy set and the membership value in the fuzzy data corresponding to each historical time.

The initial screening module 704 is configured to screen at least the initial fuzzy rule statement according to the minimum confidence value and the minimum support value of the fuzzy rule statement to obtain a target fuzzy rule statement of which the minimum confidence value is greater than or equal to the confidence threshold and the minimum support value is greater than or equal to the support threshold.

A merging module 705, configured to merge identical fuzzy rule statements in the initial fuzzy rule statement before the initial filtering module 704 filters the initial fuzzy rule statement, where the identical fuzzy rule statements are: the corresponding fuzzy sets on each state parameter and the corresponding membership values on the fuzzy sets are the same among the fuzzy rule sentences.

The defuzzification module 706 is configured to perform defuzzification processing on the target fuzzy rule statements respectively to obtain preset parameter groups corresponding to each target fuzzy rule statement, and the preset parameter groups corresponding to each historical moment form a set corresponding to the power distribution network.

An interest degree screening module 707, configured to, before the defuzzification processing is performed by the defuzzification module 706, screen at least two fuzzy rule statements, which satisfy the logical contradiction condition, in the target fuzzy rule statement according to the interest degree values of the fuzzy rule statements, so as to delete the fuzzy rule statements, which satisfy the logical contradiction condition and whose interest degree values are less than or equal to the interest degree threshold value, in the fuzzy rule statements; wherein the logical contradiction conditions include: the fuzzy sets corresponding to at least one state parameter are different among the fuzzy rule sentences.

It should be noted that, for the specific implementation of each unit in the present embodiment, reference may be made to the corresponding content in the foregoing, and details are not described here.

Referring to fig. 8, a schematic structural diagram of an electronic device according to a third embodiment of the present disclosure is provided, where the electronic device may be an electronic device capable of performing data processing, such as a computer or a server. The electronic device in this embodiment mainly establishes a prosody recognition model.

Specifically, the electronic device in this embodiment may include the following structure:

a memory 801 for storing applications and data generated by the application operations;

a processor 802 for executing an application to implement: acquiring power distribution network monitoring data of a power distribution network at a target moment, wherein the power distribution network monitoring data comprises a plurality of state parameters; acquiring a target parameter group corresponding to the monitoring data of the power distribution network from a set comprising a plurality of preset parameter groups, wherein the target parameter group comprises a plurality of state parameters; the preset parameter group in the set is generated based on historical data of the power distribution network corresponding to the power distribution network at a plurality of historical moments respectively, the historical data of the power distribution network corresponding to each historical moment comprises a plurality of state parameters, and the historical moments are moments selected in the fault-free moment section of the power distribution network; acquiring a parameter similarity value between a state parameter in the target parameter group and a state parameter in the power distribution network monitoring data; and outputting an alarm result corresponding to the power distribution network monitoring data under the condition that the parameter similarity value meets a preset alarm condition.

According to the above scheme, after the power distribution network monitoring data including a plurality of state parameters of the power distribution network at the target time are obtained, by obtaining a target parameter group which corresponds to the monitoring data of the power distribution network and also comprises a plurality of state parameters from a set comprising a plurality of preset parameter groups, the preset parameter group in the set is generated based on historical data of the power distribution network corresponding to a plurality of historical moments of the power distribution network respectively, the historical moments are moments selected in the fault-free moment section of the power distribution network, based on the moment, the parameter similarity value between the state parameters in the target parameter group and the state parameters in the monitoring data of the power distribution network is obtained, and further outputting an alarm result corresponding to the power distribution network monitoring data under the condition that the parameter similarity value meets a preset alarm condition. It can be seen that this application is through the parameter similarity between the historical state parameter of obtaining the state parameter in the distribution network monitoring data and distribution network in the no fault period, and then monitors whether the distribution network appears unusually according to the parameter similarity, avoids single and fixed threshold value to monitor and can have the unsafe condition in time measuring, can realize the distribution network monitoring through the similarity comparison between the state parameter in this application to improve the monitoring accuracy.

It should be noted that, the specific implementation of the processor in the present embodiment may refer to the corresponding content in the foregoing, and is not described in detail here.

Taking monitoring of a power distribution network in a certain country or region as an example, fig. 9 is a schematic diagram of a module of a monitoring scheme of state parameters of the power distribution network provided in the embodiment of the present application, where the following modules are respectively used for early-stage state parameter data acquisition and power distribution network fault monitoring and early warning, and the schematic diagram includes the following modules: module 1: a distribution and transformation equipment platform and a database; and (3) module 2: a data sampling, selecting and preprocessing module; and a module 3: a fuzzy association rule mining and expert knowledge base module; and (4) module: and a state parameter similarity judgment module.

As shown in fig. 9, the distribution and transformation equipment platform and database module is connected to the data selection and preprocessing module and the state parameter similarity determination module, the data selection and preprocessing module is connected to the fuzzy association rule mining and expert knowledge base module and the state parameter similarity determination module, and the fuzzy association rule mining and expert knowledge base module is connected to the state parameter similarity determination module.

Firstly, the distribution and transformation equipment platform and database module transmits data in a distribution and transformation area to the data sampling, selecting and preprocessing module, and simultaneously receives a fault early warning signal which is sent by the state parameter similarity judging module and is finally judged by the state parameter similarity, so that the platform function of data information interaction is achieved;

the data sampling, selecting and preprocessing module comprises two parts: one part is parameter sampling, selection and preprocessing in an off-line state, and the other part is a modeling basis of a fuzzy association rule mining inference machine; the other part is sampling, selecting and preprocessing of real-time monitoring state parameters in an online state, and the part is a working basis for judging the similarity of the state parameters;

the fuzzy association rule mining and expert knowledge base module is used for mining fuzzy rule sentences among variables from state parameters after offline state downsampling processing, describing the correlation among the state parameters by a rule language, and obtaining the fuzzy rule sentences through reasoning, namely the expert knowledge base. Regular data in the expert knowledge base needs to be converted into power distribution network normal state parameter numerical data, and the power distribution network normal state parameter numerical data needs to be constructed into a configuration file form which can be called by power distribution network equipment so as to be used for judging subsequent state parameters;

the state parameter similarity judging module is used for comparing similarity difference between the real-time monitoring state parameters and the inferred standard normal state parameters and sending out early warning when the similarity difference exceeds a similarity threshold value. The main work task of the state parameter similarity judgment module is embodied in development and compilation of a functional APP, the APP needs to read the normal state parameter value of the power distribution network in a configuration file to serve as a judgment standard, receive real-time monitoring state parameter data and output a state judgment result in real time after state judgment.

Specifically, combine above module, the technical scheme of this application realizes the monitoring to distribution network state parameter through following flow:

the method comprises the following steps: and searching normal historical data from the distribution and transformation equipment platform and the database of the module 1, and sampling, selecting and preprocessing the historical data related to power distribution network monitoring in an off-line state. According to the sampling time of 50 mus, the local time of the equipment is taken as a reference, the power distribution network fault related data group at the same time is sampled to be used as a group of power distribution network state parameters, and the specific power distribution network state parameters comprise: and three-phase voltage, current, active power, reactive power, frequency, voltage and current unbalance, frequency deviation, transformer area load rate and other state parameters of the low-voltage side of the distribution transformer. In order to ensure the accuracy and the representativeness of data, data samples need to be cleaned and preprocessed, and outliers caused by metering errors, sensor faults and environmental changes are eliminated;

step two: and inputting the normal distribution network state parameters processed in the off-line state of the module 2 into a fuzzy association rule mining inference engine model. The core work of the step is to utilize historical data of actual power distribution network state parameters to mine and infer association relations among the power distribution network state parameters, namely fuzzy rule statements. The fuzzy rule statement is a fuzzy project set implication among all monitoring state parameters, the two fuzzy project sets do not have intersection, fuzzy association rules obtained by the inference engine are stored in an expert knowledge base in a rule group mode, and each rule qualitatively represents the range grade of the parameters in the domain of discourse by fuzzy data. The association rule uses If-Then statement, described as follows:

wherein: x₁To X_nRepresenting different distribution and transformation state parameters, X, at the same time_2～nRepresents by X₁The state parameters of the external power distribution network, n represents the number of the state parameters of different power distribution networks at the same time; and i represents the sequence numbers of the rules at different moments in the expert knowledge base.

For fuzzy association rule mining, the most crucial is how to select the most valuable association rule from the association rules corresponding to a large amount of data. For the traditional fuzzy association mining method, the association rule takes the minimum support degree s and the minimum confidence degree c as selection criteria, and the expressions are respectively shown as formula (1) and formula (2).

Wherein: μ (X) represents X (X)₁To X_n) D represents the total number of transactions in the X dataset. The purpose of data mining is to find out credible and representative rules, firstly, the minimum support degree and the minimum confidence degree of each initial fuzzy rule statement are obtained, then, the initial fuzzy rule statements are screened according to a preset confidence degree threshold value and a preset support degree threshold value, and the minimum support degree s_minAnd minimum confidence c_minSupport and confidence thresholds are specified which respectively specify the minimum support and confidence that the association rule must meet and the minimum support and confidence that the initial fuzzy rule statement that is screened must meet, as shown in equation (3).

However, if only the minimum support and confidence conditions are used as rule selection criteria, many inference rules all satisfy the minimum support and confidence conditions, and these inference rules contradict each other, which may cause inference decision failure. Thus, minimum support and minimum confidence often do not ensure that the associated rule is both valuable, and even rule bias or misleading sometimes occurs.

When the fuzzy association rule mining is carried out, based on the traditional minimum support degree and confidence coefficient, the interestingness concept is introduced to screen the most valuable rule. The interestingness I is a measure for representing the attention degree of the rule, the larger the interestingness value is, the better the parameter relevance is under the guidance of the rule, and the expression is shown in formula (4).

The detailed modeling steps of the fuzzy association rule mining inference engine model in the module 3 are expressed as follows:

(1) determining that X1-Xn represent the state parameters of n distribution substations at the same moment, and performing fuzzification processing of a grading fuzzy set method on the sampled and preprocessed distribution network state parameter historical data: dividing the precise quantity on the state parameter domain into several grades, defining a fuzzy set for each grade, presetting the fuzzy domain as [ -6,6]The number of fuzzy sets is 7, and the corresponding linguistic variables sequentially represent NB (negative large), NM (negative medium), NS (negative small), Z (zero), PS (positive small), PM (positive medium), and PB (positive large), and are represented by a trigonometric membership function, as shown in fig. 3. Wherein c is₁To c₇And the central points of all the data fuzzy sets in the EKB are represented, and the central points are cluster centers obtained by clustering analysis of all the state parameters through a K-means clustering algorithm.

(2) Establishing a power distribution network state parameter X by fuzzy data obtained by fuzzifying a large amount of historical data in Step1₁To X_nThe initial fuzzy Rule base Rule 1.

(3) Merging X in Rule1₁To X_nThe same fuzzy Rule statements are obtained to obtain a fuzzy Rule base Rule2, and the minimum support degree s corresponding to each fuzzy Rule in the fuzzy Rule base Rule2 is calculated according to the formulas (1) and (2)_minAnd minimum confidence c_min。

(4) Selecting standard according to conventional rule, and setting minimum support degree s_minAnd confidence c_minSet to 0.3 and 0.6, respectively, according to equation (3), the less than minimum support s in the fuzzy Rule base Rule2 is removed_minAnd confidence c_minThe fuzzy Rule statements are sorted to obtain a fuzzy Rule base Rule 3.

(5) Checking whether contradictory redundant rules exist in the fuzzy Rule base Rule3, if yes, calculating the interestingness of different fuzzy Rule sentences by using a formula (4), and selecting the fuzzy Rule sentences with the maximum interestingness calculation result from the Rule3 to form the fuzzy Rule base Rule4 according to the principle that the interestingness is maximum, wherein the fuzzy Rule base Rule4 is the expert knowledge base in the module 3.

(6) And (3) taking fuzzy association rules in an expert knowledge base as guidance, and obtaining the output of different state parameter normalization values under normal state parameter domains by using a fuzzy controller in MATLAB software, namely the synthesis of input vectors and fuzzy relations.

(7) Finally, the fuzzy value output is utilized to perform defuzzification processing, and the numerical data of different state parameters under the normal state are obtained and stored in the xml configuration file.

Step three: the module 4 parameter similarity judging module is used for comparing the similarity difference between the real-time monitoring state parameter and the inferred normal state parameter, so that a similarity function needs to be constructed. The similarity function reads the value of the normal state parameter of the power distribution network in the xml configuration file in the module 3 as a judgment standard, simultaneously receives the value of the real-time monitoring state parameter in the module 2 as an input, and outputs the state judgment result to the power distribution and transformation equipment platform and the database of the module 1 in real time after the state judgment. The similarity function is a measure describing the similarity between two sets of state parameter samples, and is constructed based on the most common euclidean distance in the vector space distance, and the euclidean distance formula is shown in formula (5).

In the formula (d)_ijParameter set x representing real-time monitoring state_iNormal state parameter set x corresponding to xml file_jEuclidean distance between them, n representing the state parameter set dimension, k representing the shapeParameter number, x, in the set of state parameters_ikAnd x_jkRespectively representing the normal state parameter set x_iAnd monitoring state parameter group x_jThe corresponding kth parameter element.

In engineering applications, S_ij＝1/(1+d_ij) Is often used in the standardized Euclidean distance range to [0,1 ]]. Therefore, the target parameter group x for evaluating the normal state_iAnd monitoring state parameter group x_jThe similarity function setting between them is shown in equation (6).

Collecting historical data of the power distribution network state parameters in the normal state from the power distribution network equipment again, and defining the minimum similarity between the power distribution network normal operation state parameter group and the actual normal state parameter group obtained by all fuzzy reasoning as S_mThe minimum similarity S_mAnd a threshold coefficient k_tThe product of (A) is set as a similarity early warning threshold S_tAs shown in equation (7).

When the similarity of the monitoring state is lower than the similarity early warning threshold S_tThe time system issues an alarm, where the threshold coefficient k_tAppropriate values may be set according to monitoring sensitivity requirements. In addition, after fault early warning, the parameter serial number k with the largest relative error in each state parameter group is subjected to abnormal marking, so that fault state parameters are determined, monitoring and maintenance personnel can analyze the fault state parameter characteristics and coordinate equipment inspection, and finally the fault type is determined.

According to the algorithm logic, embedded programming development is carried out by using C + + language, a similarity function is constructed to capture the difference between the state parameters, and the comparison function between the real-time monitoring state parameters and the inference normal state parameters is realized. After the linux platform is compiled, the obtained function APP is matched with the xml configuration file in the step 2 and is transmitted to the power distribution network equipment through an equipment debugging tool, finally, the on-line monitoring of the state parameters of the power distribution network is achieved, and when fault early warning occurs, an early warning signal is reported to the equipment platform through an interactive interface.

The embodiments in the present description are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other. The device disclosed by the embodiment corresponds to the method disclosed by the embodiment, so that the description is simple, and the relevant points can be referred to the method part for description.

Those of skill would further appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that the various illustrative components and steps have been described above generally in terms of their functionality in order to clearly illustrate this interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.

The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. A software module may reside in Random Access Memory (RAM), memory, Read Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A power distribution network state parameter monitoring method is characterized by comprising the following steps:

2. The method of claim 1, further comprising:

3. The method of claim 2, wherein prior to the defuzzifying the target fuzzy rule statement, the method further comprises:

4. The method of claim 2, wherein prior to filtering the initial fuzzy rule statement according to the minimum confidence value and the minimum support value of the fuzzy rule statement, the method further comprises:

5. The method of claim 1, wherein obtaining a target parameter set corresponding to the distribution network monitoring data in a set comprising a plurality of preset parameter sets comprises:

6. The method of claim 1, wherein the parameter similarity value satisfies a preset alarm condition, comprising:

the parameter similarity value is less than or equal to a similarity threshold;

wherein the similarity threshold is obtained by:

and obtaining a similarity threshold according to the minimum similarity value.

7. The method of claim 1, wherein outputting an alarm result corresponding to the power distribution network monitoring data comprises:

and/or the presence of a gas in the gas,

8. The method of claim 7, wherein the error exception condition comprises: the relative error value in the parameter error data is the largest in the parameter error data corresponding to all the state parameters;

9. A distribution network state parameter monitoring device, characterized in that the device includes:

10. An electronic device, comprising:

a processor for executing the application to implement: acquiring power distribution network monitoring data of a power distribution network at a target moment, wherein the power distribution network monitoring data comprises a plurality of state parameters; acquiring a target parameter group corresponding to the power distribution network monitoring data in a set comprising a plurality of preset parameter groups, wherein the target parameter group comprises a plurality of state parameters; the preset parameter group in the set is generated based on historical data of the power distribution network corresponding to the power distribution network at a plurality of historical moments respectively, the historical data of the power distribution network corresponding to each historical moment comprises a plurality of state parameters, and the historical moment is a moment selected in the fault-free moment section of the power distribution network; obtaining a parameter similarity value between the state parameters in the target parameter group and the state parameters in the power distribution network monitoring data; and outputting an alarm result corresponding to the power distribution network monitoring data under the condition that the parameter similarity value meets a preset alarm condition.