CN115684918A

CN115684918A - Switch state identification method and device

Info

Publication number: CN115684918A
Application number: CN202310005708.5A
Authority: CN
Inventors: 李问溪; 李思源
Original assignee: Beijing Zhixiang Technology Co Ltd
Current assignee: Beijing Zhixiang Technology Co Ltd
Priority date: 2023-01-04
Filing date: 2023-01-04
Publication date: 2023-02-03
Anticipated expiration: 2043-01-04
Also published as: CN115684918B

Abstract

The invention relates to the technical field of electric energy metering, and provides a method and a device for identifying a switch state. The method comprises the following steps: determining a plurality of metering points corresponding to a switch to be detected and electric energy data of the metering points in a preset time interval; determining a plurality of target time periods from the preset time interval according to the change condition and the change trend of the electric energy data; clustering the plurality of metering points according to the electric energy data in the target time period aiming at each target time period to obtain a metering point communication set corresponding to the target time period; and determining the switch state of the switch to be detected based on the metering point communication sets corresponding to the target time periods. The method and the device solve the defect that the switch state information is not easy to obtain in the prior art, and realize accurate inference of the switch state in the transformer substation.

Description

Switch state identification method and device

Technical Field

The invention relates to the technical field of electric energy metering, in particular to a method and a device for identifying a switch state.

Background

System data is stored in multiple levels of security zones depending on the importance of each service in the dispatch system, the data flow, the current situation, and the security requirements. The power supply load adjustment in the jurisdiction area of the transformer substation can be realized through the combination of a plurality of switch states in the transformer substation. However, the switch state information is located in a high-level security zone of the scheduling system, and data flow across the security zone has certain implementation difficulty, so that the difficulty in acquiring the switch state information is high.

In the existing method, extra information is obtained by adding electric physical quantity measuring equipment on two sides of a switch or adding a visual sensor, and the state of the switch is judged. However, in the existing method, additional equipment is required, the system complexity of the transformer substation is increased, and additional fault hidden dangers are introduced. In addition, the existing method needs to integrate original information from multiple sources, and has complex flow and difficult implementation.

Disclosure of Invention

The invention provides a switch state identification method and device, which are used for solving the defect that switch state information is difficult to obtain in the prior art and realizing accurate inference of the switch state in a transformer substation.

The invention provides a switch state identification method, which comprises the following steps:

determining a plurality of metering points corresponding to a switch to be detected and electric energy data of the metering points in a preset time interval respectively;

determining a plurality of target time periods from the preset time interval according to the change condition and the change trend of the electric energy data;

clustering the plurality of metering points according to the electric energy data in the target time period aiming at each target time period to obtain a metering point communication set corresponding to the target time period;

and determining the switch state of the switch to be detected based on the metering point communication sets corresponding to the target time periods.

According to the method for identifying the switch state provided by the invention, the step of determining a plurality of target time periods from the preset time interval according to the change condition and the change trend of the electric energy data comprises the following steps:

determining a plurality of first time points at which the electric energy data change from the preset time interval based on the electric energy data;

removing second time points meeting preset conditions from the first time points to obtain a plurality of third time points; wherein the preset conditions include: the variation trend of the electric energy data and the electric energy derivative data in the time period between the second time point and the previous first time point and the variation trend of the electric energy data and the electric energy derivative data corresponding to the electric energy data and the electric energy derivative data in the time period between the second time point and the next first time point are smaller than a first threshold value;

and dividing the preset time interval based on the third time points to obtain the target time periods.

According to the method for identifying the switch state provided by the invention, the determining a plurality of first time points at which the electric energy data change from the preset time interval based on the electric energy data comprises the following steps:

based on the electric energy data of each metering point in a preset time interval, identifying the time point of the change of the electric energy data of each metering point from the preset time interval through a time sequence analysis method, and determining the time point as a plurality of switching points;

combining the switching points with the time difference smaller than a second threshold value in the two adjacent switching points, and determining the multiple combined switching points as the multiple first time points.

According to the switch state identification method provided by the invention, the first time point further comprises a credible weight, and the credible weight is determined by the occurrence frequency of each metering point at the time point when the electric energy data changes in the preset time interval.

According to the method for identifying the switch state provided by the invention, the step of removing the second time points meeting the preset condition from the plurality of first time points to obtain a plurality of third time points comprises the following steps:

respectively calculating the bus balance rate of the left-side period electric energy data and the bus balance rate of the right-side period electric energy data based on the left-side period electric energy data of each first time point and the previous first time point and the right-side period electric energy data of each first time point and the next first time point;

and according to the sequence of the credible weights of the first time points from low to high, sequentially removing second time points meeting preset conditions from the first time points through bus balance rate analysis to obtain a plurality of third time points.

According to the method for identifying the on-off state provided by the invention, for each target time period, clustering the plurality of metering points according to the electric energy data in the target time period to obtain a metering point communication set corresponding to the target time period comprises the following steps:

acquiring codes of the plurality of metering points based on the voltage data of the plurality of metering points of each target time period;

determining the distance between the metering points in the metering point set of each target time period based on the codes of the plurality of metering points;

and determining a metering point communicating set of each target time period by a clustering method based on the distance between the metering points of the metering point set of each target time period.

According to the switch state identification method provided by the invention, the determining of the switch state of the switch to be detected based on the metering point communication sets corresponding to the target time periods comprises the following steps:

determining a switch switching metering point of the latter target time period in the two adjacent target time periods based on the difference of the metering point connected sets of the two adjacent target time periods in the plurality of target time periods;

and determining the switch state of the switch to be detected in the next target time period based on the switch switching metering point in the next target time period.

According to the switch state identification method provided by the invention, the determining of the switch change metering point of the latter target time period in the two adjacent target time periods based on the difference of the metering point connected sets of the two adjacent target time periods in the plurality of target time periods comprises the following steps:

judging based on the number of the metering point connected sets of the two adjacent target time periods;

if the number of the metering point connected sets of the two adjacent target time periods is consistent, determining a switch switching metering point of the next target time period based on the difference of the metering point connected set of the next target time period relative to the metering point connected set of the previous time period;

and if the number of the metering point connected sets in the two adjacent target time periods is inconsistent, determining the switch switching metering point in the next target time period based on the metering point connected sets meeting the addition relation in the two adjacent target time periods.

According to the switch state identification method provided by the invention, the determining the switch state of the switch to be detected in the next target time period based on the switch change metering point in the next target time period comprises the following steps:

and determining the switch state of the switch to be detected in the next target time period based on the switch switching metering point in the next target time period and the information of the direct influence metering point and the indirect influence metering point of the switch to be detected.

The present invention also provides a switch state recognition apparatus, comprising:

the device comprises a preparation module, a detection module and a control module, wherein the preparation module is used for determining a plurality of metering points corresponding to a switch to be detected and electric energy data of the metering points in a preset time interval;

the segmentation module is used for determining a plurality of target time periods from the preset time interval according to the change condition and the change trend of the electric energy data;

the clustering module is used for clustering the plurality of metering points according to the electric energy data in the target time period aiming at each target time period to obtain a metering point communication set corresponding to the target time period;

and the identification module is used for determining the switch state of the switch to be detected based on the metering point communication sets corresponding to the target time periods.

The invention also provides an electronic device, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor executes the program to realize the switch state identification method.

The present invention also provides a non-transitory computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the switch state identification method as described in any of the above.

The invention also provides a computer program product comprising a computer program which, when executed by a processor, implements a method of identifying a switch state as described in any one of the above.

According to the method and the device for identifying the switch state, provided by the invention, the target time period can be segmented from the preset time interval through the change condition and the change trend of the electric energy data. And determining the switch state of the switch to be detected in each target time period based on the electric energy data of each target time period. The method can fully excavate the potential of the electric energy data of the metering point, and can realize accurate inference of the switch state in the transformer substation by a data analysis means without adding an additional monitoring device.

Drawings

In order to more clearly illustrate the technical solutions of the present invention or the prior art, the drawings needed for the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.

FIG. 1 is a schematic flow chart of a switch state identification method according to the present invention;

FIG. 2 is a second schematic flowchart of the switch state identification method provided by the present invention;

FIG. 3 is a schematic structural diagram of a switch state recognition device provided in the present invention;

fig. 4 is a schematic structural diagram of an electronic device provided in the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is obvious that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

For the measurement of the same physical quantity in the transformer substation, different systems are redundant mutually but have emphasis, the scheduling side data is more emphasized in time efficiency and abnormal fault tolerance, and the metering side data is more emphasized in metering accuracy. According to the importance degree and data flow, current condition and safety requirements of each relevant business system, the whole power secondary system is divided into four safety areas: the system comprises a real-time control area I, a non-control production area II, a production management area III and an information management area IV, and meanwhile, data flow across safety areas has certain implementation difficulty.

Because the accurate on-station switch state is the basis of further line loss analysis, abnormity diagnosis, load regulation and prediction and the switch state information is located in a high-grade safety area of a dispatching system, the data acquisition difficulty is high. Therefore, the embodiment of the invention discloses a switch state identification method, which can realize switch state estimation in a transformer substation by utilizing the potential of the existing metering data without additionally adding equipment. The switch state identification method of the present invention is described below with reference to fig. 1-2, and as shown in fig. 1, the method at least includes the following steps:

step 101, determining a plurality of metering points corresponding to a switch to be detected and electric energy data of the metering points in a preset time interval;

step 102, determining a plurality of target time periods from a preset time interval according to the change condition and the change trend of the electric energy data;

103, clustering a plurality of metering points according to electric energy data in the target time periods aiming at each target time period to obtain a metering point communication set corresponding to the target time period;

and 104, determining the switch state of the switch to be detected based on the metering point communication sets corresponding to the target time periods.

In step 101, it should be noted that, in the embodiment of the present invention, the electric energy metering and charging system of the electric power system acquires archive account information of a certain transformer substation and the in-station connection topology as a data source, and the electric energy metering and charging system may be, for example, a national power grid TMR system, where the archive account is information that needs to be acquired and the in-station connection topology is auxiliary information. Statistics is carried out based on the information table of the archive ledger information and the in-station connection topology, and the measuring point range, the in-station switch number, the bus section voltage grade and the bus section data related to the transformer substation can be determined. In addition, since the relative position information and the type of the metering point switch of the metering point can be directly obtained from the in-station wiring topology, the information can be integrated to obtain a metering point set related to all switches of a certain substation.

Specifically, the measurement point set further includes a direct-influence measurement point and an indirect-influence measurement point, where the direct-influence measurement point is, for example, a data change of the measurement point a directly influenced by the opening and closing of the switch P, and then the measurement point a is the direct-influence measurement point of the switch P, and the indirect-influence measurement point is, for example, the opening and closing of the switch Q influences the opening and closing of the switch P, and the opening and closing of the switch P directly influences the data change of the measurement point a, and then the measurement point a is the indirect-influence measurement point of the switch Q.

According to the range of metering points defined by the set of metering points in the electric energy metering and charging system of the electric power system, curve data such as electric quantity, voltage, current, power factor and the like in an extraction target time interval are taken as electric energy data of the target time interval, wherein the electric quantity data comprise active and reactive power and forward and reverse information, and the data such as the voltage, the current, the power factor and the like comprise split-phase information.

According to the embodiment of the invention, when the data of the metering point set is calculated step by step, all the metering points in the set are not calculated simultaneously, but a plurality of calculation units are divided according to the voltage level of the bus in the transformer substation, and the bus with the same voltage level is used as one calculation unit for subsequent calculation, so that the simple implementation of a plurality of subsequent methods can be ensured.

In step 102, the variation condition of the electric energy data, that is, the data variation of each time point in the preset time interval, may be determined preliminarily to screen out the time point at which the switch change point may be possible, and the variation trend of the electric energy data is the variation trend within a period of the preset time interval, so that the precise time point at which the switch change point may be screened out may be further screened out by the variation trend and the preset time interval may be segmented according to the time point. The obtained data in a plurality of target time periods can ensure that the switch switching action exists.

As for step 103, it should be noted that, the electric energy data of the metering point set in each target time period is clustered, for example, in combination with a perceptual hash and a density clustering manner, and each category corresponds to one connected metering point subset after clustering, so that the embodiment of the present invention can identify connected metering points through clustering, and further analyze the switch action state.

For step 104, it should be noted that after the metering point connected set of each target time period is obtained, the difference that two adjacent targets are simply connected sets may be compared, the switch state is identified, and the switch action is inferred based on the minimum variation principle.

According to the switch state identification method, the suspected switch switching time point can be identified from the target time interval by analyzing the electric energy data based on the plurality of metering points, then the time point with small electric energy data change is eliminated, and the adjacent time periods are combined. And determining the switch state of each target time period based on the separated electric energy data of each target time period. The method can fully excavate the potential of the electric energy data of the metering point, and can realize accurate inference of the switch state in the transformer substation by means of data analysis without adding an additional monitoring device.

It can be understood that, according to the variation situation and the variation trend of the electric energy data, a plurality of target time periods are determined from the preset time interval, including:

determining a plurality of first time points at which the electric energy data change from a preset time interval based on the electric energy data;

removing second time points meeting preset conditions from the first time points to obtain a plurality of third time points; wherein the preset conditions include: the variation trend of the electric energy data and the electric energy derivative data in the time period between the second time point and the previous first time point and the corresponding electric energy data and the electric energy derivative data in the time period between the second time point and the next first time point is smaller than a first threshold value;

and dividing the preset time interval based on the third time points to obtain a plurality of target time periods.

It should be noted that determining a plurality of first time points at which the electrical energy data changes from the preset time interval actually means identifying suspected switch change time points in the target time interval based on time series analysis, and the first time point set means a set of suspected switch change time points. Because the extracted electric energy data comprises various types, a general method can be adopted when different electric energy data are screened, and a method more suitable for different data characteristics can also be adopted, wherein the general method is used for example, sub-mode analysis after variational mode decomposition is carried out, a suspected switching point can be identified by the sequential change of the same-phase voltage values aiming at the voltage data, and the suspected switching point can be identified by the main power flow direction change of the circuit aiming at the electric quantity data.

The first time point set comprises some switch switching points which are identified by mistake, so the first time point set needs to be screened according to the fact that whether the adjacent time periods have obvious difference, after screening, the time points which are irrelevant to the switch switching, namely the second time points, are removed, and then the left time period and the right time period which are adjacent to each other and have no obvious data change are combined. And the time points left after the elimination are third time points, the target time intervals are segmented according to the screened third time points, and the head time or the tail time of each target time period is the time when the switch state changes.

Meanwhile, the variation trend of the electric energy data considered when the preset condition is set not only refers to the variation trend of the electric energy data, but also includes derivative variables calculated based on the electric energy data, such as a bus balance rate and probability distribution thereof.

It is understood that determining the electrical energy data for the set of metering points and the target time interval for the set of metering points includes:

determining a metering point set based on direct and indirect influence metering points of all switches with the same bus voltage grade in the transformer substation;

and acquiring original electric energy data of each metering point in the metering point set in a target time interval, preprocessing the original electric energy data, and determining the electric energy data of the target time interval of the metering point set.

It should be noted that, by acquiring the in-station connection topology and the archive ledger information of a certain substation in the early stage, the relative position information and the type of the metering point switch can be further acquired, and the direct-influence metering point and the indirect-influence metering point of each switch can be obtained. The preprocessing of the embodiment of the invention comprises data missing statistics, time scale alignment and abnormal data cleaning processing, and a data space of construction metering points, data types and sampling time is obtained after preprocessing. Meanwhile, in the embodiment of the invention, the calculation units are divided according to the voltage grade of the bus in the transformer substation, and the subsequent calculation takes the bus with the same voltage grade as the calculation range of the metering point set.

Specifically, the sampling frequency should not be lower than 96 sampling points per day, otherwise, the ideal frequency data is obtained by adopting the up-sampling processing (such as a bilinear interpolation method).

It is understood that, based on the power data, a plurality of first time points at which the power data changes are determined from a preset time interval, including:

identifying the time point of the change of the electric energy data of each metering point from the target time interval through a time sequence analysis method based on the electric energy data of each metering point in the preset time interval, and determining the time point as a plurality of switching points;

and combining the switching points with the time difference smaller than a second threshold value in the two adjacent switching points, and determining the combined switching points as a plurality of first time points.

It should be noted that, for various electric energy data in a target time interval, for example, electric energy data such as electric quantity, voltage, current, power factor, etc., a suspected switching point may be identified through sub-modal analysis after the metamorphic modal decomposition. The switching point may also be identified by different methods for the data characteristics of the voltage data, the electric quantity data, and the like, respectively. In specific practice, suspected change points are identified by adopting a time series analysis method as many as possible, and the detection is ensured to be performed as much as possible.

Specifically, the sub-modal analysis and identification comprises the following steps:

processing each type of electric energy data of each metering point into time series data in sequence, and obtaining each time series data through Variational Modal Decomposition (VMD)

Sub-modal components of limited bandwidth

；

Calculating the center frequency of each sub-modal component, identifying the specific frequency interval at the center frequency

The sub-mode of (2) is an abnormal mode, and the specific frequency interval is set based on an empirical value;

and recording the time point corresponding to the maximum amplitude position in the abnormal mode as a suspected switching point.

For voltage data, suspected switching points can be identified through sequential changes of the same-phase voltage values. For any time, calculating the same-phase voltage data of the metering points in the unit, and concretely realizing the following steps:

the same-phase voltage data at each time of the measurement point can be sorted according to a specified rule (such as descending sorting) and the serial number is recorded. For example, comparing the results of two adjacent sorting times, identifying the scene with significant difference in the sorting times, and recording the occurrence time of the next sorting as the suspected cut-off point of the phase. Wherein the significant differences in the order include, over

The number measurement point sequence changes or the absolute value of the sequence change of the individual measurement points exceeds

。

And aiming at the electric quantity data of the metering points at all times, the suspected switching points can be identified through the main power flow direction change of the line. For example, the difference between the positive and negative electricity quantities of each outlet metering point is recorded as a net electricity quantity curve, and the zero-crossing point in the net electricity quantity curve is identified as a suspected change point.

In addition, it should be noted that the suspected change-over points of each kind of electric energy data are collected, and the suspected switch change-over points of which the time difference between the two suspected change-over points is smaller than a second threshold value are merged, where the second threshold value is generally set to twice the sampling interval, and time points obviously not belonging to the change-over points are screened out to obtain a first time point set.

It is to be understood that the first time point further comprises a confidence weight, the confidence weight being determined by the number of occurrences of each metering point at the time point at which the electrical energy data changes in the target time interval.

It should be noted that, specifically, in this embodiment, a confidence weight is determined based on the number of times that each metering point appears at the time point when the electric energy data changes in the target time interval, and the first time point set is determined based on the merged time switch point set and the confidence weight of each point in the merged time switch point set. Since the suspected change points of all the metering points are summarized in the merged time change point set, the counted frequency of each time point, namely the occurrence frequency of the changed time point, can be counted according to the time dimension, a credible weight can be set according to the frequency of each time point, and if a certain time point is detected for the most times, the highest weight is given, so that the first time point set construction is completed.

It can be understood that, the second time points meeting the preset condition are removed from the plurality of first time points, and a plurality of third time points are obtained, including:

respectively calculating the bus balance rate of the left-side period electric energy data and the bus balance rate of the right-side period electric energy data on the basis of the left-side period electric energy data of each first time point and the previous first time point and the right-side period electric energy data of each first time point and the next first time point;

according to the sequence of the credible weights of the first time points from low to high, second time points meeting preset conditions are sequentially removed from the first time points through bus balance rate analysis, and a plurality of third time points are obtained.

It should be noted that, in the embodiment of the present invention, the analysis basis of the bus balance ratio analysis method is a calculation unit.

Specifically, according to the sequence from low to high of the credible weights of the first time points, the second time points meeting the preset conditions are sequentially removed from the first time points through bus balance rate analysis, namely the bus balance rate change trend, so as to obtain a plurality of third time points, and the method comprises the following steps:

traversing from the time point with the lowest credible weight in the first time point set, recording as the time point to be measured, selecting two time points which are most similar to the time point to be measured in the set, and respectively forming a left time period to be measured and a right time period to be measured with the time point to be measured;

respectively calculating mother average rate curves of the left and right time periods to be measured;

wherein the bus balance rate at any time point

Is defined as: the ratio of the difference between the total electric quantity flowing into the bus and the total electric quantity flowing out of the bus to the total electric quantity flowing into the bus is shown as formula 1:

formula 1

For the total amount of power flowing into the bus bar,

the total amount of electricity flowing out of the bus.

Calculating KL divergence values of the probability distribution of the right time interval balance rate curve relative to the probability distribution of the left time interval unbalance curve

As shown in formula 2:

formula 2

Such as

Below threshold

If the variation trend results of the left and right time periods are not obviously different, the time point to be measured is removed from the suspected cutting and changing point set until all suspected cutting and changing points except the highest weight grade in the first time point set are completely traversed to obtain a third time point set;

based on the third set of time points, the original target time interval is segmented.

It can be understood that, for each target time period, clustering the plurality of metering points according to the electric energy data in the target time period to obtain a metering point connected set corresponding to the target time period, including:

acquiring codes of a plurality of metering points based on the voltage data of the plurality of metering points of each target time period;

determining the distance between the metering points of the metering point set of each target time period based on the codes of the plurality of metering points;

It should be noted that, in the embodiment of the present invention, for all metering points in a same voltage class in a computing unit, a perceptual hash and density clustering mode are combined to realize identification of connected metering points. And sequentially traversing the target time period and intercepting corresponding voltage data.

Taking any one of the target time periods as an example, the perceptual hash code identification process for calculating the voltage data of each metering point comprises the following steps:

the split-phase voltage data of a certain metering point n is staggered to construct a two-dimensional voltage according to the sequence of A phase, B phase and C phase

Matrix array

Wherein, in the step (A),

，

raw single phase voltage data.

Will two-dimensional voltage matrix

Performing Discrete Cosine Transform (DCT) to obtain orthogonal matrix

Orthogonal matrix

As shown in equation 3:

formula 3

Wherein the content of the first and second substances,

。

then, the upper left 8 x 8 matrix is taken to construct the perceptual hash code of the metering point

The hash code is shown in equation 4:

formula 4

In addition, it should be noted that the hamming distance matrix

The method is characterized in that the Hamming distance of the perceptual hash code between each metering point is calculated, and the formula is shown as 5:

formula 5

And taking the Hamming distance matrix as input, limiting the lower limit of the number of the subclasses samples to be 2, limiting the upper limit of the total number of the subclasses to be the number of bus segments, and applying a DBSCAN density clustering algorithm. Obtaining a set of K connected metering point subclasses for any one target time period t

，

。

It can be understood that, determining the switch state of the switch to be detected based on the metering point communication sets corresponding to the target time periods includes:

determining a switch switching metering point of a later target time period in the two adjacent target time periods based on the difference of the metering point connected sets of the two adjacent target time periods in the multiple target time periods;

It should be noted that the difference of the connected sets obtained by clustering the current time period and the previous time period is determined by the method, because the connected set quantity is influenced by the state change of the bus tie switch and the outgoing line switch. At the time of judgment, if

And if the state of the bus-coupled switch is not changed, firstly recognizing the switching action of the line switch and then recognizing the switching action of the bus-coupled switch based on the minimum change principle. If it is not

It shows that the state of the bus-tie switch is likely to be changedTherefore, at the moment, the bus-bar switch switching action can be recognized firstly based on the minimum change principle, and then the line switch switching action is recognized.

According to the difference of the connected sets of the metering points of two adjacent target time periods, which metering point is most probably changed can be accurately positioned, and the switch state can be accurately predicted and judged according to the determined switch change metering point.

It can be understood that, based on the difference between the connected sets of the metering points of two adjacent target time periods in the plurality of target time periods, determining the switch-off switching metering point of the latter target time period in the two adjacent target time periods comprises:

judging based on the number of the metering point connected sets of two adjacent target time periods;

It should be noted that when it is determined that the number of the metering point connected sets in two adjacent target time periods is consistent, the switching and changing actions of the line switch are firstly identified, and a subclass label mapping relationship between two sets of sets is firstly established by taking the jarard index as a standard. Then, the metering point of the difference part of the two sets is determined as a switch switching metering point, and the action of the line switch is identified according to the metering point.

Illustratively, as shown in Table 1, the set of connected metering points for time period 0

Is composed of

Measurement Point connected set of time period 1

Is composed of

Due to the fact that

There are 3 connected sets, and

there are 3 connected sets, so both are equal in number. The reason for establishing the label mapping relationship between the two through the jarard index at this time is to ensure that the two are not recognized as difference information when the sequence numbers are different but substantially the same. The mapping relationship obtained finally is

Correspond to

，

Correspond to

，

Correspond to

. On this basis, the metering point f is selected from

Move to

Is the difference part of the two periods, so f is determined as the switch change metering point, and on is identified based on thisThe action is changed.

TABLE 1

In addition, when the number of the metering point connected sets in two adjacent target time periods is determined to be inconsistent, the switching action of the bus tie switch is firstly identified, and the part meeting the addition relation by identifying the part of which the number of the subclass elements meets the addition relation is identified.

Illustratively, as shown in Table 2, the metering point connected set for the 0 th time period

Is composed of

Metric point connected set for period 1

Is composed of

Due to the fact that

There are 3 connected sets, and

there are 4 connected sets, the two numbers being different. Thus, find

Then, the state of the bus tie switch is recognized in the cdef area and then compared respectively

And

and

and

the metering point of the difference part is determined as a switch switching metering point, and the switching action of the line switch is identified based on the metering point.

TABLE 2

It can be understood that, determining the switch state of the switch to be detected in the next target time period based on the switch change metering point in the next target time period includes:

and determining the switch state of the switch to be detected in the next target time period based on the information of the direct influence measurement point and the indirect influence measurement point of the switch to be detected and the measurement point of the switch to be detected in the next target time period.

It should be noted that, since the switch state obtained directly from the data analysis may not be combined with the actual situation in the substation, it is necessary to correct the switch operation identification estimation result by combining the information of the direct influence measurement point and the indirect influence measurement point related to each switch in the archive, thereby improving the reliability and generating the final switch state result. Taking the target time period t as an example, the switching state result is:

。

it can be understood that, as shown in fig. 2, the embodiment of the present invention further discloses a method for identifying a switch state, which includes the following steps:

step 201, curve data such as electric quantity, voltage, current, power factor and the like are obtained by combining in-station connection topology and archive ledger information;

step 202, carrying out data missing statistics, time scale alignment and abnormal data cleaning on curve data to construct a metering point-data type-sampling time data space;

step 203, carrying out time series analysis on various data of each metering point in sequence to identify the abnormality in the curve data;

step 204, summarizing and counting the abnormal recognition results according to time dimension, and calculating a credible weight according to frequency to construct a suspected switch switching time point set;

step 205, traversing from the time point with the lowest credible weight in the set, and recording as the time point to be measured;

step 206, selecting two time points in the set which are most similar to the time points to be measured to respectively form a left time period to be measured and a right time period to be measured with the time points to be measured;

step 207, respectively calculating a bus balance rate curve of the left and right periods to be measured and probability distribution information of the current balance rate curve;

208, if the results of the left and right periods to be measured have no significant difference, removing the point to be measured from the set of suspected switch change time points;

step 209, segmenting the time interval based on the screened switch change point;

step 210, sequentially traversing each time period to intercept corresponding voltage data;

step 211, clustering and identifying the connected metering points according to the similarity of the voltage data;

it should be noted that, in step 211, each metering point corresponds to each electric meter, so the process of obtaining connectivity includes: firstly, obtaining original split-phase voltage, secondly, obtaining voltage matrixes of each table according to voltage data, then, obtaining codes of each table through Hash codes after discrete cosine transform, and finally, obtaining communicated metering points through clustering based on distances among the tables, such as Hamming distances.

Step 212, determining whether the cluster number of the current time interval and the cluster number of the previous time interval are consistent:

step 213, if the two signals are consistent, preferentially identifying the switching action of the bus tie switch based on the minimum change principle;

step 214, if the two are not consistent, preferentially identifying the outlet switch switching action based on the minimum change principle;

step 215, a switch action event set is obtained through aggregation, and the switch action is corrected in an auxiliary mode according to topology matching.

The switch state identification method provided by the embodiment of the invention can be used for firstly acquiring curve-type electric energy data and then deducing a suspected switch switching time point through a time sequence analysis method. And then, the identification of the communicated metering points is realized by combining a perceptual hash and a density clustering mode, so that the action state of the switch is analyzed. Compared with the prior art, the method and the device have the advantages that the switch state identification can be completed only based on the metering data of the physical quantity of the power core without adding additional equipment. Meanwhile, the potential of the metering data can be fully mined and seamlessly connected with the subsequent analysis.

The following describes the switch state recognition device provided by the present invention, and the switch state recognition device described below and the switch state recognition method described above may be referred to in correspondence with each other. As shown in fig. 3, the apparatus comprises:

the preparation module 301 is configured to determine a plurality of metering points corresponding to the switch to be detected and electric energy data of each of the plurality of metering points in a preset time interval;

the segmentation module 302 is configured to determine a plurality of target time periods from a preset time interval according to a change condition and a change trend of the electric energy data;

the clustering module 303 is configured to cluster the plurality of metering points according to the electric energy data in the target time period to obtain a metering point communication set corresponding to the target time period;

the identification module 304 is configured to determine a switch state of the switch to be detected based on the metering point communication sets corresponding to the multiple target time periods.

The switch state recognition device provided by the embodiment of the invention can fully explore the potential of electric energy data of the metering point, and can accurately infer the switch state in the transformer substation by means of data analysis without adding an additional monitoring device.

It can be understood that, according to the change situation and the change trend of the electric energy data, a plurality of target time periods are determined from the preset time interval, including:

and combining the switching points with the time difference smaller than a second threshold value in the two adjacent switching points, and determining the plurality of combined switching points as a plurality of first time points.

It can be understood that, the step of removing the second time points satisfying the preset condition from the plurality of first time points to obtain a plurality of third time points includes:

determining a switch change metering point of a latter target time period in the two adjacent target time periods based on the difference of the metering point connected sets of the two adjacent target time periods in the plurality of target time periods;

It is understood that, based on the difference between the connected sets of the metering points of two adjacent target time periods in the multiple target time periods, determining the switch-off switching metering point of the latter target time period in the two adjacent target time periods includes:

Fig. 4 illustrates a physical structure diagram of an electronic device, which may include, as shown in fig. 4: a processor (processor) 410, a communication Interface 420, a memory (memory) 430 and a communication bus 440, wherein the processor 410, the communication Interface 420 and the memory 430 are communicated with each other via the communication bus 440. The processor 410 may call logic instructions in the memory 430 to perform a switch state identification method comprising:

determining a plurality of metering points corresponding to a switch to be detected and electric energy data of the metering points in a preset time interval;

determining a plurality of target time periods from a preset time interval according to the change condition and the change trend of the electric energy data;

clustering a plurality of metering points according to the electric energy data in the target time period to obtain a metering point communication set corresponding to the target time period;

In addition, the logic instructions in the memory 430 may be implemented in the form of software functional units and stored in a computer readable storage medium when the software functional units are sold or used as independent products. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.

In another aspect, the present invention also provides a computer program product, the computer program product including a computer program, the computer program being storable on a non-transitory computer-readable storage medium, the computer program, when executed by a processor, being capable of executing the switch state identification method provided by the above methods, the method including:

In yet another aspect, the present invention also provides a non-transitory computer-readable storage medium having stored thereon a computer program, which when executed by a processor, implements a method for identifying a switch state provided by the above methods, the method comprising:

The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and the parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.

Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment may be implemented by software plus a necessary general hardware platform, and may also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods of the various embodiments or some parts of the embodiments.

Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims

1. A switch state identification method is characterized by comprising the following steps:

2. The method for recognizing the on-off state according to claim 1, wherein the determining a plurality of target time periods from the preset time interval according to the change condition and the change trend of the electric energy data comprises:

3. The method for recognizing the switch state according to claim 2, wherein the determining a plurality of first time points at which the power data is changed from the preset time interval based on the power data comprises:

4. The switch-state recognition method according to claim 3, wherein the first time point further includes a confidence weight determined by the number of occurrences of each metering point at the time point at which the power data is changed in the target time interval.

5. The method for recognizing the on/off state according to claim 4, wherein the step of removing the second time points satisfying the preset condition from the plurality of first time points to obtain a plurality of third time points comprises:

6. The method for identifying the on-off state according to any one of claims 1 to 5, wherein the clustering the plurality of metering points according to the electric energy data in the target time period to obtain a metering point connected set corresponding to the target time period for each target time period comprises:

7. The method according to any one of claims 1 to 5, wherein the determining the switch state of the switch to be detected based on the measurement point communication sets corresponding to the target time periods comprises:

8. The switch state identification method according to claim 7, wherein the determining the switch change metering point of the latter one of the two adjacent target time periods based on the difference of the metering point connected sets of the two adjacent target time periods comprises:

9. The method for recognizing the switch state according to claim 7, wherein the determining the switch state of the switch to be detected in the next target time period based on the switch change metering point in the next target time period comprises:

10. A switch state identifying device, comprising:

11. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the switch state identification method according to any one of claims 1 to 9 when executing the program.

12. A non-transitory computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the switch state identification method according to any one of claims 1 to 9.