An online monitoring system for electric energy meter status based on correlation information

Abstract

The present invention relates to an online monitoring system for the status of an electric energy meter based on associated information, comprising a topology construction subsystem and an abnormality analysis subsystem, and also comprising a plurality of virtual load units, wherein the virtual load units are used to generate virtual loads, each virtual load unit has a corresponding electric energy meter, and the virtual load unit and the corresponding electric energy meter have associated data; through the setting of the virtual load units, on the one hand, the topological relationship can be accurately verified and determined, thereby avoiding abnormalities in collected data caused by changes in the topological relationship and reducing dependence on input topological information; on the other hand, the data relationship between the electric energy meters can be accurately determined through the associated information, so that when analyzing the abnormalities of the electric energy meters, the data is more accurate, and the action can be completed automatically, and the associated information can be dynamically adjusted. In combination with the existing abnormality analysis strategy, more accurate abnormality analysis results can be obtained.

Description

Technical Field

[0001] The present invention relates to the technical field of power station data acquisition equipment, and more particularly to an online monitoring system for electric energy meter status based on correlation information. Background Art

[0002] As an important electricity data collection unit for electricity metering and electricity price trading, the electricity meter has played an increasingly important role with the development and popularization of smart grids. At the same time, more requirements have been put forward for the functions of the electricity meter. Among them, the more critical requirement is the discovery and correction of abnormal conditions of the electricity meter. Whether it is the original standard meter comparison method or the online electricity meter error evaluation method proposed in publication number CN114460529A, by substituting the initial error value into the energy conservation electricity meter error solution model, it is judged whether the electricity meter is an out-of-tolerance electricity meter based on the Pearson correlation coefficient between the electricity meter's electricity and the line loss in the substation area. , or a method for locating an abnormal electric energy meter proposed in publication number CN114942402A, which locates abnormal electric energy meters through the Fisher discriminant method based on Tellegen's theorem, thereby realizing remote electric energy meter fault detection and positioning. Regardless of which method is used, it is impossible to avoid dynamic data errors, such as line aging, increased internal resistance, etc., and if data analysis is performed only based on the premise that hardware parameters and topological relationships remain unchanged, the conclusion obtained in this way will still be in the normal change of parameters or topological relationships of the power grid system, resulting in the system being unable to obtain the change situation or change amount in the first time, and unable to make accurate judgments on abnormal situations. Summary of the Invention

[0003] In view of this, an object of the present invention is to provide an online monitoring system for electric energy meter status based on correlation information.

[0004] In order to solve the above technical problems, the technical solution of the present invention is: an online monitoring system for electric energy meter status based on correlation information, including a topology construction subsystem and an abnormality analysis subsystem, wherein the topology construction subsystem is used to construct the topology relationship of the electric energy meters, and the abnormality analysis subsystem is used to analyze the abnormal conditions of the electric energy meters based on the topology relationship of the electric energy meters and the feedback information of each electric energy meter. It is characterized in that: it also includes a plurality of virtual load units, wherein the virtual load units are used to generate virtual loads, each virtual load unit has a corresponding electric energy meter, and the virtual load unit and the corresponding electric energy meter have associated data;

[0005] The topology construction subsystem is configured with a topology update strategy, which includes

[0006] Step A1: generating a topology update instruction set according to a preset topology information table, and sending the topology update instructions in the topology update instruction set to the corresponding virtual load unit;

[0007] Step A2: receiving collected data fed back by each electric energy meter;

[0008] Step A3: extracting collection features from the collected data, and matching topology feature data in the topology information table according to the collection features;

[0009] Step A4: Associating electric energy meters with the same topological characteristic data, and determining their topological relationship based on the collected data;

[0010] Step A5: determining association information between the electric energy meters based on the collected data corresponding to the electric energy meters having a topological relationship;

[0011] Step A6: Marking the topological relationship lines between the electric energy meters using the association information to generate a corresponding electric energy topology model;

[0012] The abnormality analysis subsystem is configured with an abnormality correction strategy and an abnormality analysis strategy. The abnormality analysis strategy is used to analyze abnormal conditions of the electric energy meter. The abnormality correction strategy includes

[0013] Step B1: obtaining collected data fed back by the electric energy meter;

[0014] Step B2: Retrieve corresponding association information based on the topological relationship between the electric energy meters;

[0015] Step B3: updating the collected data according to the associated information to obtain new abnormality analysis data;

[0016] Step B4: Bring the new anomaly analysis data into the corresponding anomaly analysis strategy.

[0017] Furthermore, the abnormality analysis subsystem is also configured with an abnormality verification strategy. After determining the location of the abnormal energy meter, the abnormality analysis strategy is executed. The abnormality verification strategy includes

[0018] Step C1: generating an abnormal verification instruction set according to the position of the abnormal electric energy meter, and sending the abnormal verification instruction in the abnormal verification instruction set to the corresponding virtual load unit;

[0019] Step C2: obtaining collected data fed back by the electric energy meter;

[0020] Step C3: The acquired collected data is again brought into the abnormality analysis strategy;

[0021] Step C4, repeating step C2 until the abnormality analysis value of the preset abnormality verification number is obtained;

[0022] Step C5: Match the abnormality analysis value with the preset abnormality type table to determine the abnormality type.

[0023] Furthermore, the virtual load unit is configured as a programmable AC load.

[0024] Furthermore, the topology information table records the topology level, topology area, and topology association relationship of each electric energy meter. In step A1, a characteristic abundance value of each electric energy meter and a characteristic difference benchmark between electric energy meters are calculated based on the topology information table. The characteristic abundance value is calculated based on the topology level and the topology association relationship, and the characteristic difference benchmark is calculated based on the topology area and the topology association relationship. A topology instruction update set is generated based on the characteristic abundance value of each electric energy meter, and topology feature data is determined in each topology instruction update set based on the feature difference benchmark. The topology feature data reflects the correspondence between the topology update instruction and the virtual load unit.

[0025] When the virtual load unit receives a topology update instruction, different virtual loads are generated according to different topology update instructions. When the power branch where the electric energy meter is located is connected to the virtual load, virtual power consumption data with virtual power consumption characteristics is generated, and the virtual power consumption characteristics correspond to the collection characteristics.

[0026] Furthermore, in step A5, the correlation function is calculated by the first correlation algorithm, and there is Where F(x) reflects the number The parent node corresponds to the association function between the child nodes numbered βn, where n is the number of The number of child nodes corresponding to the parent node, For the number The acquisition function of the electric energy meter corresponding to the parent node, is the acquisition function of the electric energy meter corresponding to the child node numbered βn, a n is the inheritance weight corresponding to the nth child node,

[0027] have [x 2n-1 ,x 2n ] is the value range of the acquisition feature corresponding to the child node numbered βn and the acquisition function, and the association information is generated according to the association function.

[0028] Furthermore, in step B3, a correction algorithm is configured to calculate abnormal analysis data, and the correction algorithm includes Among them, χ w is a corresponding abnormal analysis value in the new abnormal analysis data, χ s is the corresponding collection value in the collected data, α d is the baseline trust value, α sis a type correction parameter, which is determined by querying the correction type table according to the acquisition type corresponding to the acquisition value. F(x) is the association function in the association information. [S2, S1] is the value range of the acquisition type and the association function. The value condition is determined by querying the correction type table according to the corresponding acquisition type, and the value range is determined according to the value condition.

[0029] The correction type table pre-stores value conditions and corresponding type correction parameters, and uses the acquisition type as an index.

[0030] Furthermore, the anomaly verification strategy includes a learning correction sub-strategy, wherein the learning correction sub-strategy obtains the difference between the anomaly analysis value to be used for verification and the initially obtained anomaly analysis value to obtain a trust anomaly difference value, and the learning correction sub-strategy includes configuring the learning correction strategy with a baseline trust range, and when the trust anomaly difference value is higher than the baseline trust range, the baseline trust value is corrected by a preset first correction algorithm, and there is α d =δ a α d When the trust abnormality difference is lower than the reference trust range, the reference trust value is corrected by the preset second correction algorithm, and α d =δ e α d , where δ a is the preset trust gain value, δ e is the preset trust decay value, and δ a >1>δ e .

[0031] Furthermore, the abnormality analysis strategy is configured as Pearson correlation coefficient analysis and / or interquartile range analysis.

[0032] Furthermore, step C1 also includes calculating the verification complexity and querying the preset instruction classification table to determine the abnormal verification instruction set through the verification complexity. The verification complexity is r q To verify the complexity, q f is the topological level corresponding to the abnormal electric energy meter, q z1 is the number of parent nodes of the abnormal electric energy meter, q z2 The number of child nodes it has; the instruction classification table stores a number of abnormal verification instruction sets, each abnormal verification instruction set corresponds to an abnormal verification number, and the abnormal verification instruction set is indexed by verification complexity.

[0033] Furthermore, step C5 also includes vectorizing the exception analysis value according to different exception verification instructions, and calculating the vector sum of all obtained exception analysis values to obtain an exception analysis vector, and determining the exception type through a preset exception vector analysis table, the exception vector index table stores several exception types, and the exception type is indexed by a vector condition. When the obtained exception analysis vector meets the corresponding vector condition, the corresponding exception type is output.

[0034] The technical effects of the present invention are mainly reflected in the following aspects: through the setting of the virtual load unit, on the one hand, the topological relationship can be accurately verified and determined, avoiding the abnormal collection data caused by the change of the topological relationship, and reducing the dependence on the input topological information; on the other hand, the data relationship between the electric energy meters can be accurately determined through the associated information, so that when analyzing the abnormalities of the electric energy meters, the data is more accurate, and the action can be completed automatically, and the associated information can be dynamically adjusted. In combination with the existing abnormality analysis strategy, more accurate abnormality analysis results can be obtained. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 : Schematic diagram of the online monitoring system for electric energy meter status based on correlation information of the present invention;

[0036] Figure 2 : Flowchart of the topology update strategy of the online monitoring system for electric energy meter status based on correlation information of the present invention;

[0037] Figure 3 : Flowchart of abnormal correction strategy of online monitoring system of electric energy meter status based on correlation information of the present invention;

[0038] Figure 4 : Flowchart of abnormality verification strategy of online monitoring system of electric energy meter status based on correlation information of the present invention.

[0039] Reference numerals: 100, topology construction subsystem; 200, abnormality analysis subsystem; 1, electric energy meter; 20, virtual load unit. DETAILED DESCRIPTION

[0040] The specific embodiments of the present invention are further described below in conjunction with the accompanying drawings to make the technical solutions of the present invention easier to understand and grasp.

[0041] An online monitoring system for the status of an electric energy meter 1 based on associated information includes a topology construction subsystem 100 and an abnormality analysis subsystem 200. Generally, online monitoring systems for the status of electric energy meters 1 are established based on the topological relationship between the electric energy meters 1, so the core part of the present invention is also divided into two subsystems. The topology construction subsystem 100 is used to construct the topological relationship of the electric energy meters, and the abnormality analysis subsystem 200 is used to analyze the abnormal conditions of the electric energy meters according to the topological relationship of the electric energy meters and the feedback information of each electric energy meter 1 to achieve the purpose of online monitoring. It also includes a number of virtual load units 20, and the virtual load units 20 are used to generate virtual loads. Each virtual load unit 20 has a corresponding electric energy meter 1, and the virtual load unit 20 and the corresponding electric energy meter 1 have associated data; the virtual load unit 20 can be understood as a load that is actually within the statistical range of the corresponding electric energy meter 1, but the virtual load unit 20 is a load that is actually within the statistical range of the corresponding electric energy meter 1. The load unit 20 is different from an ordinary load. The virtual load has higher stability, the output load value is deterministic, and the impact on the entire power network can also be inferred. This design is completely different from the original electricity meter 1 monitoring system, because the original electricity meter 1 monitoring system performs calculations, analysis and judgments based on the uncertainty of the actual load value, and thus cannot provide high accuracy. The virtual load has the following forms: 1. The virtual load unit 20 can be directly configured in the power equipment of the corresponding power grid, and different loads can be achieved by controlling the power equipment of the power grid to work at different powers. The advantage is that the cost is low and no additional wiring networking is required. The disadvantage is that the power of the power equipment itself is not stable, and the load range of the virtual load it can generate is limited, which cannot meet various needs. The type of power equipment, working mode, and internal hardware composition are not unified, and additional peripherals are required. 2. The virtual load unit 20 is configured as a programmable AC load. Programmable AC loads can adjust loads with different resistance values, inductive and capacitive loads. The advantage is that various loads can be adjusted through instructions, and the load can be changed in one instruction with high accuracy. The impact of the load on the electricity meter 1 can be predicted more accurately. The disadvantage is that the cost is high and additional wiring and networking are required. 3. The load can also be set to a combination of multiple charging circuits and discharge circuits. Through the combination of multiple different charging circuits, different load conditions can be simulated, and the electrical energy stored during the simulated load is used to compensate for the working waveform and other purposes when the power grid system needs it through the discharge circuit. The advantage is that this method is more energy-efficient, but its accuracy is between options 2 and 1, because the accuracy of the charging circuit's working load is not high. In addition, the range of selectable load compositions and the efficiency of its response to instructions are also between options 1 and 2.

[0042] As the first core function of the present invention, the topology construction subsystem 100 is configured with a topology update strategy, which includes

[0043] Step A1, generate a topology update instruction set according to a preset topology information table, and send the topology update instruction in the topology update instruction set to the corresponding virtual load unit 20; first, the topology information table records the topological level, topological area and topological association relationship of each electric energy meter 1, the topological level refers to the topological level in the power grid, the topological level of the end electric energy meter 1 is 1, the topological level of the parent node corresponding to the end electric energy meter 1 is 2, and so on, and the topological area reflects the mapping of the topological physical position corresponding to the electric energy meter 1 node in the model, and generally speaking, the electric energy meters 1 that are closer are more likely to be replaced or interlaced, and the topological association relationship reflects the actual connection relationship of the electric energy meter 1, and the topology information table The generation method is to obtain it by entering, scanning, etc., and the present invention aims to generate a corresponding topology update instruction set based on a known topology information table. Preferably, the topology update is performed when the power station is idle, or when an abnormality occurs, the topology update is performed, and the corresponding topology information table is updated again after each update. In this way, the topology change can be checked more accurately. In the step A1, the characteristic abundance value of each electric energy meter 1 and the characteristic difference benchmark between the electric energy meters 1 are calculated according to the topology information table. The characteristic abundance value is calculated according to the topological level and the topological association relationship. The characteristic abundance value reflects the characteristic richness of the virtual load, such as the virtual load requires more frequent changes, larger amplitude changes, and changes in the load in inductive and capacitive quantities. The richness of the virtual load, under a single instruction, the more changes, the more changes involved, and the more types of changes, the more complex the data changes of the electric energy meter 1 caused by the load, and vice versa. The reasons for calculating the feature abundance value of the topological level and the topological association relationship are as follows: the higher the topological level, the higher the complexity of the power grid statistics involved, the more complex the corresponding topological association relationships, and the greater the number. The changes caused by the load changes are more difficult to be identified, so a higher feature abundance value is required. On the other hand, the feature difference benchmark is calculated based on the topological area and the topological association relationship. If two meters belong to the same parent node, then if at this time, the virtual loads generated by the virtual load units 20 corresponding to the two meters are similar, then for the parent node It is difficult to distinguish which child node generates the load, so the concept of feature difference benchmark is introduced, that is, when the virtual load is generated, if the distance between the two meters is close or they belong to the same parent node, or the node relationship is close, then the corresponding feature difference benchmark is large, that is, two similar virtual loads cannot be configured in the corresponding two meters. The specific method is as follows: first, a topology instruction update set is generated according to the feature abundance value of each electric energy meter 1, and an instruction feature database is configured. The instruction feature database stores several instruction features, each instruction feature has an instruction value, and a combination of several instruction features is selected according to the feature abundance value. To meet the requirement that the sum of the instruction values is greater than the feature abundance value, the topology update instruction corresponding to the instruction feature combination of the current meter is determined.Each instruction feature is also configured with an instruction similarity value, so that a topology update instruction set can be obtained, and in each topology instruction update set, topology feature data is determined based on the feature difference benchmark. The topology update instruction and the electricity meter matching relationship are exchanged under the condition that the feature richness value is satisfied until the sum of the instruction similarity values is less than the feature difference benchmark. If any match cannot be satisfied, it is necessary to re-search the instruction feature based on the feature richness value and generate a new topology update instruction. The topology feature data reflects the correspondence between the topology update instruction and the virtual load unit 20, that is, the topology update instruction and the corresponding executed virtual load unit 20 are associated; when the virtual load unit 20 receives the topology update instruction, different virtual loads are generated according to different topology update instructions. When the power branch where the energy meter 1 is located is connected to the virtual load, virtual power consumption data with virtual power consumption characteristics is generated, and the virtual power consumption characteristics correspond to the collection characteristics.

[0044] Step A2: receiving the collected data fed back by each electric energy meter 1; the method for obtaining the collected data has not been particularly improved, so it will not be described in detail.

[0045] Step A3: Extract the acquisition features from the acquired data, and match the topology feature data in the topology information table according to the acquisition features. Since the virtual load is known, the acquisition features that the virtual load can generate are also known. Therefore, the corresponding topology feature data can be analyzed by extracting the acquisition features. Since if the child node is connected to the virtual load, the acquisition data of the parent node will also have a virtual load, and they have the same topology feature data. The acquisition features correspond one-to-one to the instruction features. For example, by judging whether the acquired data contains a certain waveform, the acquisition features can be determined, thereby determining the topology feature data.

[0046] Step A4: Associate the electric energy meters 1 with the same topological characteristic data, and determine their topological relationship based on the collected data; the method of determining the topological relationship based on the collected data is as follows: since the collected data of the parent node is the collected data and superposition of the child node, the corresponding topological relationship can be analyzed through the collected data.

[0047] Step A5: Determine the association information between the electric energy meters based on the collected data corresponding to the electric energy meters with topological relationships; in step A5, calculate the association function using the first association algorithm, Where F(x) reflects the number The parent node corresponds to the association function between the child nodes numbered βn, where n is the number of The number of child nodes corresponding to the parent node, For the number The acquisition function of the electric energy meter corresponding to the parent node, g βn(x) is the acquisition function of the electric energy meter corresponding to the child node numbered βn, a n is the inheritance weight corresponding to the nth child node,

[0048] have [x 2n-1 ,x 2n ] is the value range of the acquisition function of the acquisition feature corresponding to the child node numbered βn, and the association information is generated according to the association function. The first association algorithm can calculate the corresponding association function. The purpose is that if each association information is determined, then if the data of each node is to be calculated according to the relationship between the corresponding nodes, the association function can be directly called to eliminate the error. The two association functions use the above formula. First, since the acquisition function in the acquisition data is known, the acquisition function of its parent node is also known, so the deviation between the sum of the acquisition function of the child node and the acquisition function of the parent node is regarded as an error, and the deviation of each child node to the error is the inheritance weight, and the inheritance weight is calculated by the wallpaper of the integral of the range corresponding to the acquisition feature corresponding to the child node in the total range.

[0049] Step A6: Mark the topological relationship lines between the electric energy meters using the association information to generate a corresponding electric energy topology model. Then, the corresponding topological relationship lines are marked using the association information, and the marked model is defined as the electric energy topology model.

[0050] The abnormal analysis subsystem is configured with an abnormal correction strategy and an abnormal analysis strategy. The abnormal analysis strategy is used to analyze abnormal conditions of the electric energy meter. The abnormal analysis strategy is configured as Pearson correlation coefficient analysis and / or interquartile range analysis. The abnormal analysis strategy can also be analyzed by other analysis methods. It only needs to use the collected data of the electric energy meter as the analysis basis, and the analysis results are represented in numerical form. Such analysis methods can be adapted to this system. The purpose of the abnormal correction strategy is to make the collected values more accurate during analysis and to accurately eliminate errors caused by losses. The abnormal correction strategy includes

[0051] Step B1: obtaining collected data fed back by the electric energy meter;

[0052] Step B2: Retrieve corresponding association information based on the topological relationship between the electric energy meters;

[0053] Step B3: Update the collected data according to the associated information to obtain new abnormal analysis data; In step B3, a correction algorithm is configured to calculate the abnormal analysis data, and the correction algorithm includes Among them, χ w is a corresponding abnormal analysis value in the new abnormal analysis data, χ s is the corresponding collection value in the collected data, α d is the baseline trust value, αs is a type correction parameter, which is determined by querying the correction type table according to the collection type corresponding to the collection value. F(x) is the correlation function in the correlation information, [S2, S1] is the value range of the collection type and the correlation function. The value condition is determined by querying the correction type table according to the corresponding collection type, and the value range is determined according to the value condition. Through the above correction method, the collected data can be corrected to obtain more accurate abnormality analysis data. The pre-configured correction type table determines the corresponding value conditions based on the acquisition type, because the correction parameters corresponding to different acquisition types will also be different. For example, when collecting voltage data, the corresponding correction parameters are different from those corresponding to collecting current data. At the same time, the value conditions reflect the different acquisition parameters within a certain acquisition feature range, specifically for different acquisition types. Because the interval where an acquisition waveform is located is a combination of several instruction features, each instruction feature has independent stability in addition to its function of being identified. For example, a certain instruction feature is more stable in characterizing current data, while another instruction feature is more stable in characterizing voltage data. The corresponding value ranges in the function are different, so the corresponding value conditions are determined according to the type. For the stability requirements of a specific type, the corresponding value range can be determined according to the range of the corresponding instruction feature. In this way, the abnormal analysis data can be determined as accurately as possible. Specifically, the correction type table pre-stores the value conditions and the corresponding type correction parameters, and uses the acquisition type as an index.

[0054] Step B4: Submit the new anomaly analysis data into the corresponding anomaly analysis strategy. The anomaly analysis strategy can obtain more accurate results.

[0055] As another functional point of the present invention, the abnormal analysis subsystem is also configured with an abnormal verification strategy. After determining the location of the abnormal electric energy meter, the abnormal analysis strategy is executed. The abnormal verification strategy includes

[0056] Step C1: Generate an abnormal verification instruction set according to the position of the abnormal electric energy meter, and send the abnormal verification instructions in the abnormal verification instruction set to the corresponding virtual load unit; since the position of the abnormal electric energy meter is known after preliminary analysis, step C1 also includes calculating the verification complexity, and querying the preset instruction classification table through the verification complexity to determine the abnormal verification instruction set, and the verification complexity is r q To verify the complexity, q f is the topological level corresponding to the abnormal electric energy meter, q z1 is the number of parent nodes of the abnormal electric energy meter, q z2The number of child nodes it has; the instruction classification table stores several abnormal verification instruction sets, each of which corresponds to an abnormal verification number, and the abnormal verification instruction set is indexed by verification complexity. By calculating the verification complexity, the complexity of the instructions required for the abnormal electric energy meter to be verified can be determined, and the corresponding abnormal verification instruction set is determined by calculating the verification complexity, and then the corresponding abnormal verification instruction is sent to the virtual load to wait for the next acquisition. It should be noted that at this time, each abnormal verification instruction will be resent multiple times under different logics until an abnormal analysis value with an abnormal verification number is obtained. The purpose of this is to determine the abnormal situation through different virtual loads. For example, the electric energy meter loses its metering effect under capacitive load or inductive load. If this conclusion needs to be obtained, the load needs to be tested for changes to obtain the corresponding abnormal type conclusion.

[0057] Step C2: obtaining collected data fed back by the electric energy meter;

[0058] Step C3: The acquired collected data is brought back into the anomaly analysis strategy; the anomaly verification strategy includes a learning correction sub-strategy, the learning correction sub-strategy obtains the difference between the anomaly analysis value to be used for verification and the initially acquired anomaly analysis value to obtain a trust anomaly difference value, the learning correction sub-strategy includes configuring the learning correction strategy with a baseline trust range, and when the trust anomaly difference value is higher than the baseline trust range, the baseline trust value is corrected by a preset first correction algorithm, and there is α d =δ a α d When the trust abnormality difference is lower than the reference trust range, the reference trust value is corrected by the preset second correction algorithm, and α d =δ e α d , where δ a is the preset trust gain value, δ e is the preset trust decay value, and δ a >1>δ e By continuously introducing the abnormal analysis sub-strategy, the obtained results can be corrected. If the abnormal analysis value corresponding to the Pearson correlation coefficient analysis is the correlation value P, and if the abnormal analysis value corresponding to the interquartile range rule is the voltage judgment score G, it can be extended to other abnormal analysis strategies. Then, by learning the correction strategy to judge the abnormal situation, the baseline trust value can be adjusted to adjust the corresponding influence. Preferably, δ a Set to 1.01, δ e Set it to 0.96, so that the whole system will continue to approach the truth under verification.

[0059] Step C4, repeating step C2 until the abnormality analysis value of the preset abnormality verification number is obtained;

[0060] Step C5: Match the exception analysis value with a preset exception type table to determine the exception type. Step C5 also includes vectorizing the exception analysis value according to different exception verification instructions, and obtaining the vector sum of all obtained exception analysis values to obtain an exception analysis vector. The exception type is determined using a preset exception vector analysis table. The exception vector index table stores a number of exception types, each indexed by a vector condition. When the obtained exception analysis vector meets the corresponding vector condition, the corresponding exception type is output. Because different exception types may exist, and multiple exception analysis values may appear during the verification process, a vectorized design is constructed to divide different areas of the coordinate system into corresponding exception types. The corresponding exception type can then be determined based on the vectorized results. This facilitates maintenance or replacement.

[0061] Of course, the above are only typical examples of the present invention. In addition, the present invention may also have many other specific implementation methods. Any technical solutions formed by equivalent replacement or equivalent transformation fall within the scope of protection required by the present invention.

Claims

1. An online monitoring system for electric energy meter status based on correlation information, comprising a topology construction subsystem and an anomaly analysis subsystem. The topology construction subsystem is used to construct the topology relationship of electric energy meters, and the anomaly analysis subsystem is used to analyze the abnormal conditions of electric energy meters based on the topology relationship of electric energy meters and feedback information from each electric energy meter. The system is characterized by: It also includes a plurality of virtual load units, wherein the virtual load units are used to generate virtual loads, each virtual load unit has a corresponding electric energy meter, and the virtual load unit and the corresponding electric energy meter have associated data; The topology construction subsystem is configured with a topology update strategy, which includes Step A1: generating a topology update instruction set according to a preset topology information table, and sending the topology update instructions in the topology update instruction set to the corresponding virtual load unit; Step A2: receiving collected data fed back by each electric energy meter; Step A3: extracting collection features from the collected data, and matching topology feature data in the topology information table according to the collection features; Step A4: Associating electric energy meters with the same topological characteristic data, and determining their topological relationship based on the collected data; Step A5: determining association information between the electric energy meters based on the collected data corresponding to the electric energy meters having a topological relationship; Step A6: Marking the topological relationship lines between the electric energy meters using the association information to generate a corresponding electric energy topology model; The abnormality analysis subsystem is configured with an abnormality correction strategy and an abnormality analysis strategy. The abnormality analysis strategy is used to analyze abnormal conditions of the electric energy meter. The abnormality correction strategy includes Step B1: obtaining collected data fed back by the electric energy meter; Step B2: Retrieve corresponding association information based on the topological relationship between the electric energy meters; Step B3: updating the collected data according to the associated information to obtain new abnormality analysis data; Step B4: Bring the new anomaly analysis data into the corresponding anomaly analysis strategy; The topology information table records the topology level, topology area, and topology association relationship of each electric energy meter. In step A1, a characteristic abundance value of each electric energy meter and a characteristic difference benchmark between electric energy meters are calculated according to the topology information table. The characteristic abundance value is calculated according to the topology level and the topology association relationship. The characteristic abundance value reflects the characteristic richness of the virtual load. The characteristic difference benchmark is calculated according to the topology area and the topology association relationship. A topology instruction update set is generated according to the characteristic abundance value of each electric energy meter, and topology feature data is determined in each topology instruction update set according to the feature difference benchmark. The topology feature data reflects the correspondence between the topology update instruction and the virtual load unit. An instruction similarity value is also configured between each instruction feature, and topology feature data is determined in each topology instruction update set according to the feature difference benchmark. The topology update instruction and the electric meter matching relationship are interchanged under the condition that the characteristic abundance value is satisfied until the sum of the instruction similarity values is less than the feature difference benchmark. When the virtual load unit receives a topology update instruction, different virtual loads are generated according to different topology update instructions. When the power branch where the electric energy meter is located is connected to the virtual load, virtual power consumption data with virtual power consumption characteristics is generated, and the virtual power consumption characteristics correspond to the collection characteristics.

2. The online monitoring system for electric energy meter status based on correlation information according to claim 1, characterized in that: The abnormality analysis subsystem is also configured with an abnormality verification strategy. After determining the location of the abnormal electric energy meter, the abnormality analysis strategy is executed. The abnormality verification strategy includes Step C1: generating an abnormal verification instruction set according to the position of the abnormal electric energy meter, and sending the abnormal verification instruction in the abnormal verification instruction set to the corresponding virtual load unit; Step C2: obtaining collected data fed back by the electric energy meter; Step C3: The acquired collected data is again brought into the abnormality analysis strategy; Step C4, repeating step C2 until the abnormality analysis value of the preset abnormality verification number is obtained; Step C5: Match the abnormality analysis value with the preset abnormality type table to determine the abnormality type.

3. The online monitoring system for electric energy meter status based on correlation information according to claim 1, characterized in that: The virtual load unit is configured as a programmable AC load.

4. The online monitoring system for electric energy meter status based on correlation information according to claim 2, characterized in that: In step A5, the correlation function is calculated by the first correlation algorithm, ,in The reflection number is The corresponding number of the parent node is The association function between the child nodes of For the number The number of child nodes corresponding to the parent node, For the number The acquisition function of the electric energy meter corresponding to the parent node, For the number The acquisition function of the electric energy meter corresponding to the child node of is the inheritance weight corresponding to the nth child node, , For the number The collection characteristics corresponding to the child nodes of the collection function are within the value range of the collection function, and the association information is generated according to the association function.

5. The online monitoring system for electric energy meter status based on correlation information according to claim 4, characterized in that: In step B3, a correction algorithm is configured to calculate abnormal analysis data, and the correction algorithm includes ,in, is a corresponding abnormal analysis value in the new abnormal analysis data, For a certain collection value in the collected data, is the baseline trust value, is a type correction parameter, which is determined by querying the correction type table according to the acquisition type corresponding to the acquisition value. is the correlation function in the correlation information, To determine the value range of the acquisition type and the associated function, the value condition is determined by querying the correction type table through the corresponding acquisition type, and the value range is determined according to the value condition; the correction type table pre-stores the value condition and the corresponding type correction parameter, and uses the acquisition type as the index.

6. The online monitoring system for electric energy meter status based on correlation information according to claim 5, characterized in that: The anomaly verification strategy includes a learning correction sub-strategy, wherein the learning correction sub-strategy obtains the difference between the anomaly analysis value to be used for verification and the initially obtained anomaly analysis value to obtain a trust anomaly difference value, and the learning correction sub-strategy includes configuring the learning correction strategy with a baseline trust range, and when the trust anomaly difference value is higher than the baseline trust range, correcting the baseline trust value by a preset first correction algorithm, When the trust abnormality difference is lower than the reference trust range, the reference trust value is corrected by the preset second correction algorithm. ,in is the preset trust gain value, is the preset trust decay value, and .

7. The online monitoring system for electric energy meter status based on correlation information according to claim 2, characterized in that: The abnormal analysis strategy is configured as Pearson correlation coefficient analysis and / or interquartile range analysis.

8. The online monitoring system for electric energy meter status based on correlation information according to claim 7, characterized in that: In step C1, the verification complexity is calculated, and the abnormal verification instruction set is determined by querying the preset instruction classification table through the verification complexity. The verification complexity is , To verify the complexity, is the topological level corresponding to the abnormal electric energy meter, is the number of parent nodes of the abnormal electric energy meter, The number of child nodes it has; the instruction classification table stores a number of abnormal verification instruction sets, each abnormal verification instruction set corresponds to an abnormal verification number, and the abnormal verification instruction set is indexed by verification complexity.

9. The online monitoring system for electric energy meter status based on correlation information according to claim 2, characterized in that: The step C5 also includes vectorizing the exception analysis value according to different exception verification instructions, and calculating the vector sum of all obtained exception analysis values to obtain an exception analysis vector, and determining the exception type through a preset exception vector analysis table. The exception vector index table stores a plurality of exception types, and the exception type is indexed by a vector condition. When the obtained exception analysis vector meets the corresponding vector condition, the corresponding exception type is output.

Images