Method, system, and device for identifying sectionalized phase states of a substation based on measurement data

Abstract

The present invention belongs to the field of power equipment, and specifically relates to a method, system, and device for identifying sub - layers and phases of a transformer substation based on measurement data. First, the type of each metering device is determined according to its file information. Then, the time - series features of each metering device are extracted based on the measurement data uploaded each time, and the corresponding statistical features are extracted based on all the measurement data uploaded multiple times throughout the day. Then, the feature recognition is completed by using the depth - first search algorithm and the locality - sensitive hashing algorithm according to the time - series features and statistical features respectively, and a label relationship table that can represent the phase sequence and subordination relationship of each measurement switch and electric energy meter is generated. Finally, the true attributes of each measurement switch and electric energy meter are determined by combining the two label relationship tables determined by different methods, thereby completing the task of identifying the topological relationship of the metering devices in the transformer substation. The present invention solves the problem that the existing sub - layer and phase identification of metering devices in the transformer substation relies on manual work, resulting in low efficiency and insufficient reliability.

Description

Technical Field

[0001] The present invention belongs to the field of power equipment, and particularly relates to a method, system, and device for identifying the hierarchical and phase separation of a substation based on measurement data. Background Art

[0002] As a key component of the low-voltage distribution network, the metering box shoulders the important responsibility of supplying power to users, and its stable operation has a crucial impact on the stability of the entire power grid. However, with the growth of power demand, numerous users have accessed the power grid disorderly, resulting in an increasingly complex structure of the low-voltage distribution network. The complexity of the distribution network makes it difficult for operation and maintenance personnel to accurately master the topological structure of substation users and their power consumption phase sequence, leading to an increasingly prominent problem of three-phase imbalance in substation power consumption and frequent occurrence of low-voltage events. In addition, the complexity of the distribution network also makes it difficult to quickly locate other faults when they occur, which seriously threatens the safety and stability of user power consumption. Therefore, quickly and accurately mastering the topological structure of the substation and the power consumption phase of users is of great significance to power supply companies and users.

[0003] Traditionally, the identification of the topological structure and phase of a distribution substation mainly relies on two methods: manual mapping and statistics by operation and maintenance personnel, and on-site detection using a substation identifier. The manual mapping and statistics method requires operation and maintenance personnel to determine the corresponding relationship between the transformer and each power consumption user one by one along the low-voltage distribution line and record it in detail in the substation file. The on-site detection method based on the substation identifier requires installing a terminal identifier on the side of the substation transformer, performing phase detection operations on the electric energy meters of each power user, and then recording the relevant information in the substation file. Both of these methods have problems of huge workload, low verification efficiency, and untimely information update. With the continuous increase in the number of users, these problems become more prominent. Summary of the Invention

[0004] In order to solve the problem that it is difficult to accurately distinguish the phase sequence and hierarchy of each power node in the distribution network of a complex substation in the prior art, the present invention provides a method, system, and device for identifying the hierarchical and phase separation of a substation based on measurement data.

[0005] The solution provided by the present invention includes:

[0006] A method for identifying the hierarchical and phase separation of a substation based on measurement data, which is used to analyze and determine the phase sequence and subordination relationship of each meter in the substation according to the measurement data generated by each meter in the substation. The hierarchical and phase separation method includes the following steps:

[0007] S1: Obtain the file information of all meters in the substation from the power consumption information collection system and collect the load data of the current day in real time.

[0008] Among them, the load data includes phase voltage, phase current, and phase power. The file information includes the device identification code, the electricity consumption address, or the management unit code.

[0009] S2: Divide each meter into a gateway meter, a measurement switch, or a watt-hour meter according to the file information, and generate a data sequence for each gateway meter, measurement switch, and watt-hour meter in combination with the corresponding load data.

[0010] S3: Calculate the change in the phase current of each meter based on the data sequences of the meters collected at adjacent sampling times, and use it as the time series feature on each phase of the meter.

[0011] S4: Use the depth-first search algorithm to perform phase sequence division between the gateway meter and the measurement switch, and perform subordinate relationship classification and phase sequence division between the measurement switch and the watt-hour meter according to the time series features on each phase of each meter; and then obtain the corresponding first label relationship table Z.

[0012] S5: Calculate the mean, variance, and extreme values of the phase voltage, phase current, and phase power of each meter based on the data sequences of the three types of meters collected throughout the day; and perform normalization processing on them as the statistical features of each phase.

[0013] S6: Use the locality-sensitive hashing algorithm to perform phase sequence division between the gateway meter and the measurement switch, and perform subordinate relationship classification and phase sequence division between the measurement switch and the watt-hour meter according to the statistical features on each phase of each meter; and then obtain the corresponding second label relationship table Q.

[0014] S7: Take the relevant attributes of the meters with the same phase sequence and subordinate relationship in the first label relationship table Z and the second label relationship table Q as the true attributes, and classify the meters without complete attributes into the pending set.

[0015] S8: Continue to collect the load data of each meter in the substation area on the next day, use the meters in the pending set as the classification objects, and repeat steps S2 - S7 until the pending set is an empty set, and then determine the phase sequence and subordinate relationship of all meters.

[0016] As a further improvement of the present invention, in step S2, each meter in the substation area uploads the measurement data every 15 minutes, and a total of 96 groups of data are generated on the same day; then the data sequence D n (t) of the gateway meter, measurement switch, or watt-hour meter is expressed as follows:

[0017]

[0018] Among them, id represents the device identification code; U, P, and I respectively represent the phase - separated voltage, phase - separated power, and phase - separated current, and the subscripts A, B, and C are the corresponding phase marks; t represents the sampling time mark, where t = 1, 2, …, 96; n represents the type mark of the metering table, and when n = 1, 2, 3, it corresponds to the gateway meter, measurement switch, and electricity meter respectively.

[0019] As a further improvement of the present invention, in step S3, the calculation formula for the change amount of the phase - separated current is as follows:

[0020]

[0021] In the above formula, represents the current change amount of the j - th phase of the i - th metering table at time t; represents the phase - separated current of the j - th phase of the i - th metering table at time t; represents the phase - separated current of the j - th phase of the i - th metering table at time t - 1.

[0022] As a further improvement of the present invention, in step S4, the process of generating the first label relationship table Z using the depth - first search algorithm is as follows:

[0023] S41: Represent the connection relationship between the upper - level metering table and the lower - level metering table as a directed graph, where the nodes of the graph represent each metering table, and the edges represent the subordinate relationship and phase - sequence relationship between the metering tables; then the construction method of the first label relationship table Z is as follows:

[0024] S42: Initialize an empty label relationship table Z to record the relationship between each pair of metering tables;

[0025] S43: Start traversing from a certain measurement switch node corresponding to the root node, and record the current change amount of this measurement switch;

[0026] S44: Recursively traverse the adjacent electricity meter nodes under this measurement switch node. If the sum of the power change amounts of several lower - level metering table nodes is equal to the current change amount of a certain upper - level metering table, it means that the lower - level metering table is subordinate to the corresponding upper - level metering table, record the corresponding edge weight, and generate the first label relationship table Z.

[0027] As a further improvement of the present invention, in step S5, the mean value μ x and variance σ x of the phase - separated voltage, phase - separated current, or phase - separated power are calculated as follows:

[0028]

[0029] In the above formula, x(t) represents the original value of the phase - separated voltage, phase - separated current, or phase - separated power on each phase at time t.

[0030] As a further improvement of the present invention, in step S5, the expression for the normalization process of the statistical features is as follows:

[0031]

[0032] Among them, x represents the data of the original features, including three categories: voltage U, power P, and current I; D' n (x) represents the normalized feature vector; σ x represents the variance of the original features, and μ x represents the mean of the original features; α x and β x are two custom parameters used to adjust the normalization effect of different features.

[0033] As a further improvement of the present invention, in step S6, the process of generating the second label relationship table Q according to the statistical features on each phase of each meter is as follows:

[0034] S61: Based on the locality-sensitive hashing algorithm, perform hash mapping on the feature vectors between the measurement switch and the electricity meter, and between the gateway meter and the measurement switch, to obtain different hash buckets.

[0035] S62: Calculate the Euclidean distance value between any two data points belonging to different types of meters corresponding to the same hash bucket.

[0036] S63: Take the two data points with the Euclidean distance value less than the threshold as a similarity pair. In each similarity pair, the two meters belong to the upper and lower levels and the phase sequences correspond. Then, construct the second label relationship table Q according to all the similarity pairs in each hash bucket.

[0037] As a further improvement of the present invention, in step S61, the hash mapping adopted by the locality-sensitive hashing algorithm is as follows:

[0038]

[0039] In the above formula, a k and b k respectively represent the feature vectors established according to the data of two different types of meters; w represents the width of the hash bucket; x' represents the feature data including the normalized phase voltage, phase current, phase power, as well as the mean, variance, and extreme values of the phase voltage, the mean, variance, and extreme values of the phase current, and the mean, variance, and extreme values of the phase power.

[0040] As a further improvement of the present invention, in step S62, the calculation method of the Euclidean distance value between any two data points is as follows:

[0041] Suppose there are two data points V and Q: V=(v 1, v 2 , …, v n ), Q = (q 1 , q 2 , …, q n ), and the calculation formula for their Euclidean distance is as follows:

[0042]

[0043] Where n is the vector length of the data point.

[0044] As a further improvement of the present invention, in step S8, it is assumed that after a period of the specified duration, there are still meters in the pending set whose complete attributes have not been determined. Then, the management personnel of the power distribution area go to the site for on-site detection to determine the phase sequence and subordination relationship of the corresponding meters.

[0045] The present invention also includes a hierarchical and phase-separated system for meters in a power distribution area based on measurement data, which is used to determine the phase sequence and subordination relationship of each meter in the power distribution area by adopting the aforementioned method for identifying hierarchical and phase separation of power distribution areas based on measurement data. The hierarchical and phase-separated system includes: an object library, a data acquisition unit, a timing feature generation unit, a statistical feature generation unit, a first classification module, a second classification module, and an attribute update unit.

[0046] The object library stores the file information of all meters whose phase sequence and subordination relationship have not been confirmed. The data acquisition unit is used to obtain the file information of all meters and collect the load data of the current day in real time; then, according to the file information, each meter is divided into a reference meter, a measurement switch, or an electric energy meter, and data sequences of each reference meter, measurement switch, and electric energy meter are generated in combination with the corresponding load data.

[0047] The timing feature generation unit is used to calculate the change amount of the phase-separated current of the meter after each update of the data sequence of each meter, and use it as the timing feature of each phase. The statistical feature generation unit is used to calculate the mean, variance, and extreme values of the phase-separated voltage, phase-separated current, and phase-separated power in the data sequence of the meter on the current day after the daily update of the data sequence of each meter; and perform normalization processing on them and use them as the statistical features of each phase.

[0048] The first classification module uses the depth-first search algorithm to perform phase sequence division between the reference meter and the measurement switch according to the timing features of each phase of each meter in the object library, and perform subordination relationship classification and phase sequence division between the measurement switch and the electric energy meter; and then obtain the corresponding first label relationship table.

[0049] The second classification module is used to perform phase sequence division between the gateway meter and the measurement switch, and perform subordination relationship classification and phase sequence division between the measurement switch and the electricity meter according to the statistical features of each phase of each meter in the object library by using the locality-sensitive hashing algorithm; and then obtain the corresponding second label relationship table.

[0050] The attribute update unit is used to obtain the first label relationship table and the second label relationship table generated by the first classification module and the second classification module; then, the overlapping phase relationship and subordination relationship in the first label relationship table and the second label relationship table are used as the true attributes of the corresponding meter, and the meter is deleted from the object library.

[0051] The present invention also includes a hierarchical and phase-separated device for meters in a power distribution area based on measurement data, which includes a memory, a processor, and a computer program stored in the memory and running in the processor. When the processor executes the computer program, the above-mentioned method for identifying hierarchical and phase separation of power distribution areas based on measurement data is implemented, and then the phase sequence and subordination relationship of each meter in the power distribution area are determined.

[0052] The technical solution provided by the present invention has the following beneficial effects:

[0053] The present invention utilizes real-time acquisition of load data such as current, voltage, and power, extracts the time-series features and statistical features of each meter, and then combines the depth-first search algorithm and the locality-sensitive hashing algorithm to realize the hierarchical and phase separation of meters according to the two features. Then, the classification results of the two methods are compared and mutually verified, thereby significantly improving the working efficiency and accuracy of the hierarchical and phase separation tasks. Compared with the current method that relies on manual mapping and statistics by operation and maintenance personnel and uses a power distribution area identifier for household detection, the present invention realizes online automatic identification of the topological structure and phase of the power distribution area, effectively reduces the overall cost, and improves the working efficiency and timeliness;

[0054] Through the hierarchical and phase separation identification of the power distribution area, the present invention can efficiently carry out the "three-point" line loss calculation in the power distribution area, quickly lock the three-phase unbalanced users, and meet the refined management of the smart grid. In addition, the method provided by the present invention is not only applicable to existing gateway meters, measurement switches, and electricity meters, but also applicable to future newly added devices of the same type, and has strong practicability. Description of the Drawings

[0055] Figure 1 It is a schematic diagram of the topological structure of each meter in a typical power distribution area.

[0056] Figure 2 It is a flowchart of the steps of the method for identifying hierarchical and phase separation of power distribution areas based on measurement data provided in Embodiment 1 of the present invention.

[0057] Figure 3This is the system architecture diagram of the hierarchical and phase-splitting system of the metering devices in the power distribution area based on measurement data provided in Embodiment 1 of the present invention. Detailed implementation manners

[0058] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0059] Embodiment 1

[0060] In a typical power distribution area, the metering devices of the master station system that regularly generate and upload measurement data include three types, namely the electric energy meters (or called: household electric energy meters) of the power distribution area, measurement switches, and gateway meters. Among them, the electric energy meters are installed at the power consumption nodes corresponding to each power user and measure the power load generated by the user in real time. Multiple electric energy meters located in a certain area are usually installed in a digital metering box, and the measurement switch in the metering box measures the total load of the power users corresponding to all the electric energy meters therein. The gateway meter is responsible for measuring the load data of the entire power distribution area. Therefore, the topological structure of the power distribution area should be in the Figure 1 pyramid structure as shown. However, in actual power distribution scenarios, new power demands brought by electric vehicle chargers, charging stations, and sporadic businesses often occur in disorderly access situations, and for power users, household electric energy meters are also divided into different types such as single-phase meters and three-phase meters. Therefore, the actual power topological structure of the power distribution area is often more complex. This will bring great difficulties to the traditional manual metering device phase-splitting and hierarchical tasks.

[0061] To address the above problems, this embodiment provides a method for identifying hierarchical and phase-splitting of power distribution areas based on measurement data, which is used to analyze and determine the phase sequence and subordination relationship of each metering device according to the measurement data generated by the metering devices in the power distribution area. The overall idea of the technical solution provided in this embodiment is to first determine the type of the meter as an electric energy meter, a measurement switch, or a gateway meter according to the archive information of each metering device and add a type label to it. Then, use the measurement data uploaded by each metering device every day to perform phase sequence classification tasks on the gateway meter and multiple measurement switches respectively, and perform phase sequence classification and subordination relationship discrimination tasks between multiple measurement switches and multiple electric energy meters.

[0062] In the phase sequence classification and subordination relationship differentiation tasks of this embodiment, first, the time series characteristics of each meter were generated based on the measurement data uploaded each time, and the statistical characteristics of each meter were generated based on all the measurement data uploaded multiple times throughout the day. Then, the feature recognition was completed using the depth-first search algorithm and the locality-sensitive hashing (LSH) algorithm according to the time series characteristics and statistical characteristics respectively, and a label relationship table that can represent the phase sequence and subordination relationship of each measurement switch and electric energy meter was generated. Finally, the true attributes of each measurement switch and electric energy meter were determined by combining the two label relationship tables determined by different methods, thereby realizing the identification of the topological relationship of the meters in the substation area.

[0063] Specifically, as Figure 2 shown, the substation area hierarchical and phase identification method based on measurement data provided in this embodiment includes the following steps:

[0064] S1: Obtain the file information of all meters in the substation area from the power consumption information acquisition system, denoted as D info , and collect the load data D 0 of the current day in real time.

[0065] Among them, the load data includes phase voltage, phase current, and phase power. The phase voltage includes U A , U B , U C ; the phase current includes I A , I B , I C ; the phase power includes P A , P B , P C . The file information includes the device identification code id, the power consumption address, or the management unit code.

[0066] S2: Divide each meter into a reference meter, a measurement switch, or an electric energy meter according to the file information, and generate a data sequence for each reference meter, measurement switch, and electric energy meter in combination with the corresponding load data.

[0067] In the existing power system, the automatic collection of power data has basically been realized. In such a system, the substation area management personnel do not need to go to the site to copy the meters. Instead, the meters in the substation area automatically report their own measurement data to the concentrator at regular time intervals according to the preset time interval, and the concentrator sends it to the master station. Currently, each meter in the substation area under the existing system uploads the measurement data every 15 minutes, and a total of 96 groups of data are generated on the current day; then the expression of the data sequence D n (t) of the reference meter, measurement switch, or electric energy meter is as follows:

[0068]

[0069] Among them, id represents the device identification code; U, P, and I respectively represent the phase voltage, phase power, and phase current, and the subscripts A, B, and C are the corresponding phase marks; t represents the sampling time mark, where t = 1, 2, …, 96; n represents the type mark of the metering table, and when n = 1, 2, 3, it corresponds to the gateway meter, measurement switch, and electricity meter respectively.

[0070] For example, it can represent generating a data sequence based on the 25th group of load data reported by the measurement switch numbered id 1 at 06:15 on the same day. Correspondingly, D 1 represents the data sequence of the substation gateway meter, and D 3 represents the data sequence of the electricity meter.

[0071] S3: Calculate the change in the phase current of each metering table based on the data sequences of the metering tables collected at adjacent sampling times, and use it as the time-series feature on each phase of the metering table.

[0072] In this embodiment, the calculation formula for the change in the phase current of any metering table (including the gateway meter, measurement switch, and electricity meter) is as follows:

[0073]

[0074] In the above formula, represents the current change of the j-th phase of the i-th metering table at time t; represents the phase current of the j-th phase of the i-th metering table at time t; represents the phase current of the j-th phase of the i-th metering table at time t - 1.

[0075] S4: Use the depth-first search algorithm to perform phase sequence division between the gateway meter and the measurement switch according to the time-series features on each phase of each metering table, and perform subordinate relationship classification and phase sequence division between the measurement switch and the electricity meter; and then obtain the corresponding first label relationship table Z.

[0076] In this embodiment, the gateway meter is located at the top of the entire substation area, and there is only one gateway meter in a substation area. All measurement switches are subordinate devices of the gateway meter. Therefore, only phase sequence division is required between the gateway meter and the measurement switch. The number of measurement switches and electricity meters in the substation area is more than one. Any two electricity meters may belong to the subordinate devices of the same measurement switch or may belong to the subordinate devices of two different measurement switches. Therefore, for any measurement switch and electricity meter, not only the phase sequence relationship between them needs to be distinguished, but also the subordinate relationship between the two needs to be divided.

[0077] For this special hierarchical and phase - splitting task, this embodiment proposes a label relationship table to store the corresponding information. In the label relationship table, a phase - sequence label for distinguishing the phase sequence is added to each gateway table, and a phase - sequence label for distinguishing the phase sequence and a subordinate relationship label for characterizing the corresponding upper - level device are added to each measurement switch and watt - hour meter simultaneously.

[0078] In the solution of this embodiment, technicians extract the timing features, which reflect the increment of the measurement results of each meter. Generally speaking, the increment of the measurement results of the same upper - level device should be after the increments of the measurement results of all subordinate devices. Based on this characteristic, this embodiment first uses the depth - first search algorithm to generate the first label relationship table according to the timing features on each phase of each meter.

[0079] The depth - first algorithm is an algorithm for traversing or searching a tree or a graph. Its basic idea is to start from the root node, traverse the nodes of the tree along the depth of the tree, and search the branches of the tree as deep as possible. When all the edges of node v have been explored, or it is impossible to move forward, the search will backtrack to the starting node of the edge where node v was discovered and continue traversing. This process is repeated until all nodes have been visited.

[0080] In this embodiment, the process of generating the first label relationship table Z using the depth - first search algorithm is as follows:

[0081] S41: Represent the connection relationship between the upper - level meter and the lower - level meter as a directed graph. The nodes of the graph represent each meter, and the edges represent the subordinate relationship and the phase - sequence relationship between the meters. Then, the construction method of the first label relationship table Z is as follows:

[0082] S42: Initialize an empty label relationship table Z to record the relationship between each pair of meters;

[0083] S43: Start traversing from a certain measurement - switch node corresponding to the root node, and record the current - change amount of this measurement switch;

[0084] S44: Recursively traverse the adjacent watt - hour - meter nodes under this measurement - switch node. If the sum of the power - change amounts of several lower - level meter nodes is equal to the current - change amount of a certain upper - level meter, it means that the lower - level meter is subordinate to the corresponding upper - level meter. Record the corresponding edge weight and generate the first label relationship table Z.

[0085] It can be seen that the deep learning algorithm can perform a search each time new measurement data is generated by all the meters, and then assign multiple child nodes whose incremental sum is equal to the total increment of the parent node to the corresponding parent node. Every day, the deep learning algorithm can execute 96 times according to the 96 pieces of uploaded data, respectively correct and update the previous search results, and in this way, a more accurate first label relationship table Z can be obtained.

[0086] S5: According to the data sequences of the three types of meters collected throughout the day, calculate the mean, variance, and extreme values of the phase voltage, phase current, and phase power of each meter; and perform normalization processing on them as the statistical features of each phase.

[0087] In this embodiment, each meter uploads 96 pieces of measurement data throughout the day. According to the corresponding data sequences of these measurement data, the mean, variance, and extreme values of the phase current, voltage, and power data of each meter can be generated, and then using these data, the statistical features of the meter's all-day data can be obtained. Specifically,

[0088] The mean μ of the phase voltage, phase current, or phase power x and the variance σ x are calculated as follows:

[0089]

[0090] In the above formula, x(t) represents the original value of the phase voltage, phase current, or phase power on each phase at time t.

[0091] The expression for the normalization processing of the statistical features is as follows:

[0092]

[0093] Among them, x represents the data of the original features, including three categories: voltage U, power P, and current I; D' n (x) represents the normalized feature vector; σ x represents the variance of the original features, μ x represents the mean of the original features; α x and β x are two custom parameters used to adjust the normalization effect of different features; during the normalization process, the tanh and weighted smoothing methods are adopted to ensure that the data has a consistent scale after normalization and eliminate noise.

[0094] S6: According to the statistical features of each phase of each meter, use the locality-sensitive hashing algorithm to perform phase sequence division between the gateway meter and the measurement switch, and perform subordinate relationship classification and phase sequence division between the measurement switch and the electricity meter; and then obtain the corresponding second label relationship table Q.

[0095] In this embodiment, the process of generating the second label relationship table Q according to the statistical characteristics on each phase of each meter is as follows:

[0096] S61: Based on the locality-sensitive hashing algorithm, perform hashing mapping on the feature vectors between the measurement switch and the electricity meter, and between the gateway meter and the measurement switch, to obtain different hash buckets; among them, the hashing mapping adopted by the locality-sensitive hashing algorithm is as follows:

[0097]

[0098] In the above formula, a k and b k respectively represent the feature vectors established according to the data of two different types of meters; w represents the width of the hash bucket; x' represents the feature data including the normalized phase voltage, phase current, phase power, as well as the mean, variance and extreme values of the phase voltage, the mean, variance and extreme values of the phase current, and the mean, variance and extreme values of the phase power.

[0099] S62: Calculate the Euclidean distance value between any two data points belonging to different types of meters corresponding to the same hash bucket; the calculation method is as follows:

[0100] Suppose there are two data points V and Q: V=(v 1 ,v 2 ,…,v n ), Q=(q 1 ,q 2 ,…,q n ), the calculation formula of their Euclidean distance is as follows:

[0101]

[0102] Among them, n is the vector length of the data point.

[0103] S63: Take the two data points with the Euclidean distance value less than the threshold as a similarity pair. In each similarity pair, the two meters belong to the upper and lower levels and the phase sequences correspond. Then, construct the second label relationship table Q according to all the similarity pairs in each hash bucket.

[0104] S7: Take the relevant attributes of the meters with the same phase sequence and subordination relationship in the first label relationship table Z and the second label relationship table Q as the true attributes, and classify the meters without complete attributes into the pending set.

[0105] Based on the 96 - point measurement data reported daily by each meter, in this embodiment, the temporal characteristics of the data reported each time are extracted, and the depth - first search algorithm is used to iteratively search 96 times to generate and update the first label relationship table Z that can represent the phase relationship and subordination relationship between each meter. Also, the statistical characteristics of all the reported data of each meter are extracted from the measurement data of the whole day, and the locality - sensitive hashing algorithm is used to generate and update the second label relationship table Q that can represent the phase relationship and subordination relationship between each meter.

[0106] On this basis, in order to make the confidence level of the final hierarchical and phase - splitting results higher, in this embodiment, the relevant attributes of the meters with the same phase sequence and subordination relationship in the two label relationship tables are used as the true attributes.

[0107] For example, for a certain electricity meter, if it has complete attributes (both phase - sequence label and subordination label) in the first label relationship table Z, but the attribute content of this electricity meter is incomplete in the second label relationship table Q, then it is considered that the attribute information of this electricity meter in the first label relationship table Z is not credible. For a certain electricity meter, if the attribute information in the first label relationship table Z and the second label relationship table Q is complete, but one of the phase - sequence label and subordination label does not match in the two label relationship tables, then it is considered that the attribute information of this electricity meter in the first label relationship table Z and the second label relationship table Q is not credible. For a certain electricity meter, if the attribute information in the first label relationship table Z and the second label relationship table Q is complete, and the phase - sequence label and subordination label are exactly the same in the two label relationship tables, then it is considered that the attribute information of this electricity meter in the first label relationship table Z and the second label relationship table Q is credible. This embodiment takes it as the true attribute of this meter and completes the hierarchical and phase - splitting tasks for this electricity meter.

[0108] S8: Continue to collect the load - type data of each meter in the sub - station area on the next day. Taking the meters in the undetermined set as the classification objects, repeat steps S2 - S7 until the undetermined set is an empty set, and then determine the phase sequence and subordination relationship of all meters.

[0109] In the solution of this embodiment, generally, combining the measurement data of a certain day can complete the hierarchical and phase - splitting of the meters in a sub - station area with fewer nodes and simple topological relationships. However, for a complex large - scale sub - station area, the measurement data of a single day may not be sufficient to achieve the hierarchical and phase - splitting of all meters. In response to this situation, in this embodiment, the meters that have not completed the hierarchical and phase - splitting are separately put into an undetermined set, and then continue to collect the load - type data uploaded by all meters, and repeat the above classification process to complete the phase - sequence division and subordination - relationship classification tasks of the remaining measurement switches and electricity meters.

[0110] Generally speaking, by using data of consecutive days, the hierarchical and phase-separated tasks of more than 90% of the meters can be basically completed. On this basis, assuming that there are some special nodes, the electrical loads generated by them are difficult to distinguish through time series characteristics and statistical characteristics. The solution provided in this embodiment also provides a relief solution, where relevant technical personnel go to the site to check the meters on the spot to complete the corresponding hierarchical and phase-separated tasks. That is: assuming that after a period of a specified duration, there are still meters with undetermined complete attributes in the to-be-determined set, the management personnel of the transformer area go to the site for on-site detection to determine the phase sequence and subordination relationship of the corresponding meters.

[0111] Combined with the above content, it can be known that: the technical solution provided in this embodiment uses the load data such as current, voltage, and power obtained in real time, and can judge the topology structure and phase of the distribution transformer area in real time, with advantages such as high precision, low cost, and high efficiency. It effectively overcomes the problems of large workload, low efficiency, and lagging information update of traditional methods, and provides a solid technical support for the construction and progress of the smart grid.

[0112] Embodiment 2

[0113] On the basis of the solution of Embodiment 1, this embodiment further provides a hierarchical and phase-separated system for meters in a transformer area based on measurement data, which is used to determine the phase sequence and subordination relationship of each meter in the transformer area by using the hierarchical and phase-separated identification method for transformer area based on measurement data as in Embodiment 1. As Figure 3 shown, the hierarchical and phase-separated system includes: an object library, a data acquisition unit, a time series feature generation unit, a statistical feature generation unit, a first classification module, a second classification module, and an attribute update unit.

[0114] The object library stores the file information of all meters whose phase sequence and subordination relationship have not been confirmed. The data acquisition unit is used to obtain the file information of all meters and collect the load data of the current day in real time; then, according to the file information, each meter is divided into a gateway meter, a measurement switch, or an electric energy meter, and data sequences of each gateway meter, measurement switch, and electric energy meter are generated in combination with the corresponding load data.

[0115] The time series feature generation unit is used to calculate the change amount of the phase-separated current of the meter after each update of the data sequence of each meter, and use it as the time series feature of each phase. The statistical feature generation unit is used to calculate the mean value, variance, and extreme value of the phase-separated voltage, phase-separated current, and phase-separated power in the data sequence of the meter on the current day after the daily update of the data sequence of each meter; and perform normalization processing on them and use them as the statistical features of each phase.

[0116] The first classification module uses the depth - first search algorithm to perform phase - sequence division between the gateway meter and the measurement switch, as well as subordinate - relationship classification and phase - sequence division between the measurement switch and the electricity meter according to the timing characteristics of each phase of each meter in the object library; and then obtains the corresponding first label - relationship table.

[0117] The second classification module is used to perform phase - sequence division between the gateway meter and the measurement switch, as well as subordinate - relationship classification and phase - sequence division between the measurement switch and the electricity meter according to the statistical characteristics of each phase of each meter in the object library by using the locality - sensitive hashing algorithm; and then obtains the corresponding second label - relationship table.

[0118] The attribute update unit is used to obtain the first label - relationship table and the second label - relationship table generated by the first classification module and the second classification module; then, the overlapping phase - relationship and subordinate - relationship in the first label - relationship table and the second label - relationship table are used as the true attributes of the corresponding meter, and the meter is deleted from the object library.

[0119] Embodiment 3

[0120] Based on the solution of Embodiment 1, this embodiment further provides a hierarchical and phase - separated device for meters in a power distribution area based on measurement data, which includes a memory, a processor, and a computer program stored in the memory and running in the processor. When the processor executes the computer program, it implements the method for hierarchical and phase - separated identification of power distribution areas based on measurement data in Embodiment 1, and then determines the phase - sequence and subordinate - relationship of each meter in the power distribution area.

[0121] The hierarchical and phase - separated device for meters in a power distribution area based on measurement data provided in this embodiment is essentially a computer device for implementing the solution in Embodiment 1. This computer device can be an intelligent terminal capable of executing programs, a tablet computer, a laptop computer, a desktop computer, a rack - mounted server, a blade - server, a tower - server, or a cabinet - type server (including an independent server or a server cluster composed of multiple servers), etc.

[0122] The computer device pointed out in this embodiment includes at least, but is not limited to, a memory and a processor that can be communicatively connected to each other through a system bus. Among them, the memory (i.e., the readable storage medium) includes flash memory, hard disk, multimedia card, card-type memory (such as SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory may be an internal storage unit of the computer device, such as the hard disk or memory of the computer device. In other embodiments, the memory may also be an external storage device of the computer device, such as a plug-in hard disk equipped on the computer device, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, etc. Of course, the memory may also include both the internal storage unit and the external storage device of the computer device. In this embodiment, the memory is generally used to store the operating system and various application software installed on the computer device. In addition, the memory can also be used to temporarily store various data that have been output or will be output.

[0123] In some embodiments, the processor may be a central processing unit (CPU), a graphics processing unit (GPU), a controller, a microcontroller, a microprocessor, or other data processing chips. The processor is generally used to control the overall operation of the computer device. In this embodiment, the processor is used to run the program code stored in the memory or process data.

[0124] The technical features of the above-described embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0125] The above-described embodiments merely represent several implementation manners of the present invention. The description is relatively specific and detailed, but it should not be construed as a limitation on the scope of the invention patent. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the present invention patent shall be subject to the appended claims.

Claims

1. A method for identifying the stratification and phase separation of a substation based on measurement data, which is used to analyze and determine the phase sequence and subordinate relationship of each meter according to the measurement data generated by each meter in the substation; characterized in that: The stratified phase separation method comprises the following steps: S1: Obtain the archive information of all meters in the substation from the electricity consumption information collection system, and collect the load data of the day in real time; The load data includes phase voltage, phase current and phase power; the file information includes equipment identification code, electricity address or management unit code; S2: Divide each meter into a gateway meter, a measuring switch or an electric energy meter according to the archive information, and generate a data sequence for each gateway meter, measuring switch and electric energy meter in combination with the corresponding load type data; S3: Calculate the change of the phase current of each meter according to the data sequence of the meter collected at adjacent sampling moments, and use it as the time series feature of each phase of the meter; S4: using a depth-first search algorithm to perform phase sequence division between the gateway meter and the measuring switch according to the timing characteristics of each phase of each meter, and to perform subordinate relationship classification and phase sequence division between the measuring switch and the electric energy meter; thereby obtaining a corresponding first label relationship table Z; S5: Based on the data sequences of the three meters collected throughout the day, the mean, variance, and extreme value of the phase voltage, phase current, and phase power of each meter are calculated; and the normalized data are used as the statistical features of each phase; S6: Based on the statistical characteristics of each phase of each meter, a local sensitive hash algorithm is used to perform phase sequence division between the gateway meter and the measuring switch, and to perform subordinate relationship classification and phase sequence division between the measuring switch and the electric energy meter; thereby obtaining a corresponding second label relationship table Q; S7: taking the relevant attributes of the meters with the same phase sequence and subordinate relationship in the first label relationship table Z and the second label relationship table Q as the real attributes, and classifying the meters that do not contain complete attributes into the pending set; S8: Continue to collect the load data of each meter in the substation for the next day, take the meters in the pending set as the classification object, repeat steps S2-S7 until the pending set is an empty set, and then determine the phase sequence and subordinate relationship of all meters.

2. The method for identifying the stratification and phase separation of a station based on measurement data according to claim 1, characterized in that: In step S2, each meter in the area uploads metering data every 15 minutes, and a total of 96 sets of data are generated on the day; then the data sequence D of the gateway meter, measuring switch or energy meter n The expression of (t) is as follows: Wherein, id represents the device identification code; U, P, I represent the phase voltage, phase power and phase current respectively, and the subscripts A, B, C are the corresponding phase marks; t represents the sampling time mark, t = 1, 2, ..., 96; n represents the meter type mark, and n = 1, 2, 3 correspond to the gateway meter, measuring switch and electric energy meter respectively.

3. The method for identifying the stratification and phase separation of a station based on measurement data according to claim 2, characterized in that: In step S3, the calculation formula of the change amount of the phase current is as follows: In the above formula, represents the current change of the jth phase of the ith meter at time t; represents the phase current of the jth phase of the ith meter at time t; Represents the phase-splitting current of the j-th phase of the ith meter at time t-1.

4. The method for identifying the stratification and phase separation of a station based on measurement data according to claim 1, characterized in that: In step S4, the process of generating the first label relationship table Z using the depth-first search algorithm is as follows: S41: The connection relationship between the upper meter and the lower meter is represented as a directed graph, the nodes of the graph represent the meters, and the edges represent the subordinate relationship and phase sequence relationship between the meters; then the method for constructing the first label relationship table Z is as follows: S42: Initialize an empty label relationship table Z to record the relationship between each pair of meters; S43: traverse from a certain measurement switch node corresponding to the root node, and record the current change of the measurement switch; S44: recursively traverse the adjacent electric energy meter nodes under the measurement switch node. If the sum of the power change values ​​of several lower-level meter nodes is equal to the current change of a certain upper-level meter, it means that the lower-level meter is subordinate to the corresponding upper-level meter. The corresponding edge weight is recorded and a first label relationship table Z is generated.

5. The method for identifying the stratification and phase separation of a station based on measurement data according to claim 1, characterized in that: In step S5, the mean value μ of the phase voltage, phase current or phase power is x and variance σ x The calculation formula is as follows: In the above formula, x(t) represents the original value of the phase voltage, phase current or phase power on each phase at time t.

6. The method for identifying the stratification and phase separation of a station based on measurement data according to claim 2, characterized in that: In step S5, the expression for normalization of statistical features is as follows: Among them, x represents the original feature data, including voltage U, power P and current I; D' n (x) represents the normalized eigenvector; σ x Represents the variance of the original feature, μ x Represents the mean of the original features; α x and β x are two custom parameters used to adjust the normalization effect of different features.

7. The method for identifying the stratification and phase separation of a station based on measurement data according to claim 6, characterized in that: In step S6, the process of generating the second label relationship table Q according to the statistical features of each phase of each meter is as follows: S61: Based on the local sensitive hashing algorithm, hash mapping is performed on the feature vectors between the measurement switch and the electric energy meter, and between the gateway meter and the measurement switch to obtain different hash buckets; wherein the hash mapping adopted by the local sensitive hashing algorithm is as follows: In the above formula, a k and b k They represent feature vectors established based on data of two different types of meters; w represents the width of the hash bucket; x′ represents feature data including normalized split-phase voltage, split-phase current, split-phase power, and the mean, variance and extreme value of the split-phase voltage, the mean, variance and extreme value of the split-phase current, and the mean, variance and extreme value of the split-phase power; S62: Calculate the Euclidean distance between any two data points belonging to different types of meters corresponding to the same hash bucket; the calculation method is as follows: Suppose there are two data points V and Q: V = (v1, v2, ..., v n ), Q=(q1,q2,…,q n ), their Euclidean distance is calculated as follows: Where n is the length of the vector of data points; S63: Set the Euclidean distance value to be less than the threshold The two data points are regarded as a similarity pair, and the two meters in each similarity pair are upper and lower and correspond in sequence, and then the second label relationship table Q is constructed according to all similarity pairs in each hash bucket.

8. The method for identifying the stratification and phase separation of a station based on measurement data according to claim 1, characterized in that: In step S8, assuming that after a specified period of time, there are still meters with incomplete attributes in the pending set, the substation management personnel will go to the site for field inspection to determine the subordinate relationship and phase relationship of the corresponding meters.

9. A layered and phased system for meters in a substation based on measurement data, characterized in that: It is used to determine the phase sequence and subordinate relationship of each meter in the substation by using the substation layered phase identification method based on measurement data as described in any one of claims 1 to 8; The stratified phase separation system comprises: The object library stores the archive information of all meters with unconfirmed phase sequence and subordinate relationship; A data acquisition unit is used to obtain the archive information of all meters and collect the load data of the day in real time; then, each meter is divided into a gateway meter, a measuring switch or an electric energy meter according to the archive information, and a data sequence of each gateway meter, measuring switch and electric energy meter is generated in combination with the corresponding load data; A time series feature generating unit, which is used to calculate the change of the phase current of each meter after the data sequence of each meter is updated, and use it as the time series feature of each phase; A statistical feature generation unit is used to calculate the mean, variance, and extreme value of the phase voltage, phase current, and phase power in the data sequence of each meter on that day after the data sequence of each meter is updated daily; and perform normalization processing on the calculated values ​​as the statistical features of each phase; A first classification module, which uses a depth-first search algorithm to perform phase sequence division between the gateway meter and the measuring switch, and performs subordinate relationship classification and phase sequence division between the measuring switch and the electric energy meter according to the timing characteristics of each phase of each meter in the object library; thereby obtaining a corresponding first label relationship table; A second classification module, which is used to perform phase sequence division between the gateway meter and the measuring switch, and perform subordinate relationship classification and phase sequence division between the measuring switch and the electric energy meter according to the statistical characteristics of each phase of each meter in the object library by using a local sensitive hashing algorithm; thereby obtaining a corresponding second label relationship table; and An attribute updating unit is used to obtain the first label relationship table and the second label relationship table generated by the first classification module and the second classification module; then the overlapping phase relationship and subordinate relationship in the first label relationship table and the second label relationship table are used as the real attributes of the corresponding meter, and the meter is deleted from the object library.

10. A stratified and phased device for meters in a substation based on measured data, characterized in that: It includes a memory, a processor, and a computer program stored in the memory and running in the processor, and is characterized in that: when the processor executes the computer program, it implements the substation stratification and phase identification method based on measurement data as described in any one of claims 1-8, and then determines the phase sequence and subordinate relationship of each meter in the substation.

Images