A method for optimizing measurement data and measurement points in a hierarchical topology model of a distribution network in a substation.
By constructing a hierarchical topology model of the distribution network in the transformer substation, optimizing the processing of measurement data and the layout of measurement points, the problems of unreasonable layout of measurement points and poor data quality in the low-voltage transformer substation were solved, resulting in reduced equipment costs and improved reliability of fault location.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- STATE GRID SHANGHAI MUNICIPAL ELECTRIC POWER CO
- Filing Date
- 2026-01-21
- Publication Date
- 2026-06-02
AI Technical Summary
Existing technologies have unreasonable measurement point layouts in low-voltage distribution networks, leading to increased equipment costs, poor measurement data quality, insufficient adaptability, large line loss calculation errors, and numerous blind spots in fault location.
A hierarchical topology model of the distribution network in the transformer area is constructed. The measurement data processing and measurement point layout are optimized by combining the hierarchical electrical coupling law. Interference noise is removed by using the hierarchical adjacency matrix and blind source separation algorithm. The data is completed by using electrical correlation. The optimal measurement point location and number are determined by multi-objective optimization algorithm.
It reduces the investment cost of measuring equipment, improves the accuracy of measurement data and the reliability of fault location, reduces line loss calculation errors, and is suitable for different types of low-voltage distribution areas.
Smart Images
Figure CN122136915A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of power distribution network technology, and in particular to a method for optimizing measurement data and measurement points in a hierarchical topology model of a distribution network in a transformer substation. Background Technology
[0002] As the final link in the distribution network, the low-voltage distribution area exhibits a three-tiered topology of "main trunk-branch-user," specifically a hierarchical connection of "transformer → main line → branch box → user meter." It is the core carrier for ensuring power supply reliability and reducing line losses. Currently, the measurement system and measurement point layout of the distribution network mainly rely on the following technical solutions: First, at the level of measurement point layout, a "fixed location + uniform distribution" model is adopted. This involves installing smart meters, current transformers, and other measurement devices at the transformer outgoing terminals, the middle section of the main line, and some branch boxes. On the user side, basic electricity consumption data is collected only through the household meter. Second, at the level of measurement data processing, time-series data such as voltage, current, and power are collected through concentrators and directly used for line loss calculation, fault location, and other applications after simple filtering (such as moving average). Third, in terms of related technology dependence, existing measurement point optimization methods mostly focus on 10kV medium-voltage distribution networks and are designed based on the "full coverage" approach of global topology; measurement data processing mostly adopts general time-series data cleaning algorithms.
[0003] The aforementioned existing technical solutions have revealed many prominent problems in practical applications, mainly in the following three aspects: Unreasonable measurement point layout: The number and location of measurement points are mostly set based on experience without being optimized in combination with topology layer characteristics. Some layer nodes (such as branch boxes with low load density) have repeated measurement devices, resulting in an increase in equipment costs of 30%-40%. Measurement points are often not arranged at the junction of the main line and branch line and in high-load user cluster areas, resulting in line loss calculation errors of 15%-20% and fault location blind spots of more than 25%.
[0004] Poor measurement data quality: Data quality optimization was not performed for the electrical coupling characteristics of the hierarchical topology (such as branch load superposition and main line voltage drop accumulation). The coupling interference of branch line load fluctuations on the main line measurement data resulted in a noise rate of 8%-12% for voltage and current data. Traditional completion algorithms (such as linear interpolation) did not utilize the electrical correlation of the hierarchical topology, and the deviation between the completed data and the true value exceeded 10%.
[0005] The contradiction between adaptability and economy: Existing measurement point optimization methods do not consider the local correlation of the layered structure of low-voltage distribution areas. The measurement point optimization methods of medium-voltage distribution networks are directly transferred to low-voltage distribution areas, resulting in an increase of more than 50% in equipment investment costs, and they cannot adapt to the small-scale, layered operating characteristics of distribution areas. The measurement data processing does not take into account the hierarchical electrical laws of the distribution area's "main trunk-branch-user" structure, resulting in poor method adaptability.
[0006] It is understood that the above statements only provide background information related to the present invention and do not necessarily constitute prior art. Summary of the Invention
[0007] Based on the aforementioned technical problems, the purpose of this invention is to provide a method for optimizing measurement data and measurement points in a hierarchical topology model of a distribution network in a transformer substation. This method aims to solve problems such as unreasonable measurement point layout, poor measurement data quality, and insufficient adaptability. By constructing a hierarchical topology model and optimizing the measurement data processing and measurement point layout scheme in combination with the hierarchical electrical coupling law, the invention aims to reduce the investment cost of measurement point equipment, improve the accuracy of measurement data, and enhance the reliability of line loss calculation and fault location downstream applications, thereby providing support for the efficient operation and precise management of low-voltage transformer substations.
[0008] To achieve the above objectives, this invention proposes a method for optimizing measurement data and measurement points in a hierarchical topology model of a distribution network, comprising the following steps: S1. Construct a hierarchical topology model of the distribution network in the transformer area; S2. Based on the hierarchical topology model of the distribution network in the transformer area, the measurement data is processed hierarchically; S3. Based on the hierarchical topology model of the distribution network in the aforementioned area, the location and number of measuring points are optimized through multi-objective optimization to obtain an optimization scheme; S4. Verify whether the optimization scheme meets the objective. If it does not meet the objective, repeat steps S2-S4.
[0009] Furthermore, the method for constructing a hierarchical topology model of a distribution network in a transformer substation includes the following steps: S1.1 Construct a hierarchical adjacency matrix to represent the electrical connection relationships within and between levels; S1.2, Generate the hierarchical electrical feature matrix; S1.3 Verify the hierarchical topology model of the distribution network in the transformer substation to ensure the accuracy of the node connection relationship in the hierarchical topology model of the distribution network in the transformer substation.
[0010] Furthermore, the construction of the distribution network layered topology model for the transformer substation is as follows: the nodes and lines from the transformer outgoing end to the incoming end of each branch box are divided into the backbone layer; the nodes and lines from the outgoing end of the branch box to the incoming end of the user cluster are divided into the branch layer; and the nodes and lines from the incoming end of the user cluster to each user's meter are divided into the user layer.
[0011] Furthermore, the hierarchical processing of the measurement data includes the following steps: S2.1 Separate interference noise at different levels; S2.2, Use the electrical relationships between levels to complete missing data; S2.3. Verify the measurement data after hierarchical processing. If the deviation from the hierarchical electrical law exceeds the preset value, repeat S2.1-S2.3.
[0012] Furthermore, the method for separating interference noise at different levels includes the following steps: A blind source separation algorithm based on a hierarchical adjacency matrix is adopted, which takes synchronous measurement data of the backbone layer and the branch layer as input and outputs clean data after interference removal. The blind source separation algorithm uses the FastICA algorithm: In the formula, For measurement data after separating interference noise, It is a hierarchical adjacency matrix. The collected measurement data; Alternatively, the blind source separation algorithm may employ a wavelet transform algorithm to suit scenarios with high real-time requirements.
[0013] Furthermore, the method for completing missing data using electrical correlations between levels includes the following steps: A hierarchical electrical connectivity model is constructed, and gradient boosting trees are used to train and complete the electrical connectivity model: In the formula, For electrical correlation model, For training data, To complete the missing measurement data, This is a hierarchical electrical feature matrix; Constraints: Completing missing data must satisfy hierarchical power balance, i.e., the input power of the backbone layer = the sum of the output power of each branch layer + backbone loss.
[0014] Furthermore, the method for optimizing the location and number of measurement points for multiple objectives specifically includes the following steps: S3.1 Construct a multi-objective model of cost, accuracy, and reliability; S3.2. Use a non-dominated sorting genetic algorithm to find the optimal solution in terms of cost, accuracy, and reliability. S3.3. Based on the optimal solution, determine the location and number of measurement points at each level to obtain the optimized scheme.
[0015] Furthermore, the method for constructing a cost-accuracy-reliability multi-objective model includes the following steps: constructing the multi-objective model from three aspects: cost objective, accuracy objective, and reliability objective; wherein, Regarding the cost objectives: In the formula, For the first i The equipment and installation cost for each measuring point, where n is the number of measuring points; Regarding the accuracy targets: measurement coverage based on hierarchical topology is greater than or equal to 90%, and the data error rate is less than or equal to 5%; Regarding the accuracy targets: the downstream line loss calculation error is less than or equal to 8%, and the fault location accuracy rate is greater than or equal to 90%; The constraints of the cost-accuracy-reliability multi-objective model are: at least one measuring point should be arranged at each level, and the distance between measuring points should be greater than or equal to 50 meters to avoid redundancy.
[0016] Furthermore, the non-dominated sorting genetic algorithm employs the NSGA-III algorithm, and the method of the NSGA-III algorithm for finding the cost-accuracy-reliability optimal solution includes the following steps: S3.2.1 Generate an initial population containing the locations and numbers of measurement points; S3.2.2 Based on the hierarchical topology model of the distribution network in the aforementioned area, calculate the cost, accuracy, and reliability indicators of each individual component; S3.2.3. Retain non-dominated solutions and generate the next generation population through crossover mutation; S3.2.4 After the iteration converges, output the optimal solution set of cost-accuracy-reliability, and select the optimal measurement point scheme according to the actual engineering requirements; Alternatively, the non-dominated sorting genetic algorithm may employ a particle swarm optimization algorithm to suit scenarios with high real-time requirements.
[0017] Further, step S4 includes the following steps: S4.1 Build a hierarchical topology simulation model, input the optimization scheme, and verify whether the line loss calculation error and fault location accuracy meet the objectives; S4.2 Deploy the optimization plan in three typical transformer substations, collect on-site measurement data within one month, and compare the indicators before and after optimization; S4.2 If the on-site measurement data does not meet the target, return to S2 to adjust the data processing algorithm parameters, or return to the third step to optimize the measurement point position until the requirements are met; The three typical transformer substation areas are urban densely populated substation areas, rural scattered substation areas, and old and dilapidated substation renovation areas.
[0018] Compared with the prior art, the present invention has the following advantages: This invention constructs a hierarchical topology model and optimizes measurement data processing and measurement point layout schemes by combining hierarchical electrical coupling laws. This reduces the investment cost of measurement point equipment, enhances the reliability of downstream applications, and brings significant economic benefits. Furthermore, this invention is adaptable to different types of transformer substations, including densely populated urban areas, dispersed rural areas, and old substation renovation projects. The model is flexible and requires no additional adaptation costs, providing support for the efficient operation and precise management of low-voltage transformer substations. Attached Figure Description
[0019] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the description will be briefly introduced below. Obviously, the drawings in the following description are one embodiment of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort: Figure 1 This is a flowchart illustrating a method for optimizing measurement data and measurement points in a hierarchical topology model of a distribution network in an area, as provided in an embodiment of the present invention.
[0020] Figure 2 This is a schematic diagram of the hierarchical topology model of the distribution network in the transformer substation provided in an embodiment of the present invention. Detailed Implementation
[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0022] It should be noted that, in this document, the terms "comprising," "including," "having," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Unless otherwise specified, an element defined by the phrase "comprising..." or "including..." does not exclude the presence of additional elements in the process, method, article, or terminal device that includes said element.
[0023] It should be noted that the accompanying drawings are all in a very simplified form and use non-precise ratios, and are only used to facilitate and clearly illustrate the purpose of the embodiments of the present invention.
[0024] This invention proposes a method for optimizing measurement data and measurement points in a hierarchical topology model of a distribution network in a transformer substation. By constructing a hierarchical topology model of the distribution network in a transformer substation, optimizing the measurement data processing flow based on the hierarchical electrical coupling law, and finally determining the optimal measurement point layout scheme through a multi-objective optimization algorithm, the collaborative optimization of "topology-data-measurement points" is achieved.
[0025] like Figure 1 As shown in the figure, the method for optimizing measurement data and measurement points in a hierarchical topology model of a distribution network proposed in this invention includes the following steps: S1. Construct a hierarchical topology model of the distribution network in the transformer area.
[0026] In this embodiment, the hierarchical division standard of the distribution network topology model is as follows: the nodes and lines from the transformer outgoing end to the incoming end of each branch box are divided into the backbone layer, i.e., the first layer; the nodes and lines from the outgoing end of the branch box to the incoming end of the user cluster are divided into the branch layer, i.e., the second layer; and the nodes and lines from the incoming end of the user cluster to each user's meter are divided into the user layer, i.e., the third layer. The backbone layer includes transformer T, backbone line L1, and branch box incoming nodes B1-in / B2-in; the branch layer includes branch boxes B1 / B2 and branch lines L2 / L3; and the user layer includes user cluster nodes C1 / C2 and user meters U1-U6, as shown below. Figure 2 As shown.
[0027] In other embodiments, the distribution network hierarchical topology model of the distribution area is divided into two levels according to "transformer-main line-user", which simplifies the complexity of the distribution network hierarchical topology model and is suitable for small distribution areas with fewer than or equal to 30 users.
[0028] The method for constructing a hierarchical topology model of a distribution network in a transformer substation specifically includes the following steps: S1.1 Construct a hierarchical adjacency matrix.
[0029] The hierarchical adjacency matrix is divided into "main layer - branch layer - user layer", which can characterize the electrical connection relationship within and between each layer.
[0030] S1.2 Generate the hierarchical electrical feature matrix.
[0031] The electrical feature matrix records the electrical characteristics of each level of node, such as rated voltage, line impedance, and historical load fluctuation coefficient, providing a basis for data processing and measurement point optimization.
[0032] S1.3 Verify the hierarchical topology model of the distribution network in the transformer area.
[0033] By comparing on-site survey data with data from the Distribution Automation System (DAS), the accuracy of node connection relationships in the hierarchical topology model of the distribution network area is ensured to be ≥98%.
[0034] S2. Based on the hierarchical topology model of the distribution network in the transformer area, the measurement data is processed hierarchically.
[0035] Based on the electrical coupling law of the hierarchical topology model, the collected measurement data (voltage U, current I, power P) are processed by "interference separation-correlation completion-accuracy verification".
[0036] Specifically, the hierarchical processing of the measurement data includes the following steps: S2.1 Separate interference noise at different levels.
[0037] By utilizing the principle that the measured data of the backbone layer equals the sum of the load superposition data of each branch layer and the loss data of the backbone itself, interference noise at different levels is separated. In this embodiment, a blind source separation algorithm based on a hierarchical adjacency matrix is used. Synchronous measured data from the backbone layer and branch layers are input, and the clean data after interference removal is output. In this embodiment, the blind source separation algorithm uses the FastICA algorithm: In the formula, For measurement data after separating interference noise, It is a hierarchical adjacency matrix. The collected measurement data.
[0038] In other embodiments, the blind source separation algorithm employs a wavelet transform algorithm, which is suitable for scenarios with high real-time requirements.
[0039] By employing a blind source separation algorithm based on a hierarchical adjacency matrix to separate interference noise at different levels, the noise rate of measurement data can be reduced from 8%-12% to 3%-5%.
[0040] S2.2. Use the electrical relationships between levels to complete the missing data.
[0041] By leveraging the electrical correlations between layers, such as the fact that branch layer measurement data can be derived from trunk layer data and branch line impedance calculations, missing data can be supplemented. In this embodiment, the method for supplementing missing data using the electrical correlations between layers includes the following steps: A hierarchical electrical interconnection model is constructed, and gradient boosting trees are used to train and complete the electrical interconnection model. The GBRT algorithm is then used to train and complete the electrical interconnection model. In the formula, For electrical correlation model, For hierarchical electrical feature matrix, For training data, To complete the missing measurement data.
[0042] Constraints: Completing missing data must satisfy hierarchical power balance, i.e., the input power of the backbone layer = the sum of the output power of each branch layer + backbone loss.
[0043] S2.3. Verify the measurement data after hierarchical processing. If the deviation from the hierarchical electrical law exceeds the preset value, repeat S2.1-S2.3.
[0044] A hierarchical error verification mechanism is adopted. If the measurement data after hierarchical processing deviates from the hierarchical electrical law (such as the voltage drop formula) by more than 5%, the process returns to step S2.1 for reprocessing to ensure that the data accuracy is ≥95%.
[0045] S3. Based on the hierarchical topology model of the distribution network in the aforementioned area, the location and number of measuring points are optimized through multi-objective optimization to obtain an optimization scheme.
[0046] Based on the hierarchical topology model of the distribution network in the aforementioned area, the location and number of measuring points are optimized with the goal of "lowest equipment cost, highest measurement accuracy, and strongest reliability for downstream applications".
[0047] The method for optimizing the location and number of measurement points for multiple objectives specifically includes the following steps: S3.1 Construct a multi-objective model of cost, accuracy, and reliability; Specifically, a multi-objective model is constructed from three aspects: cost objective, accuracy objective, and reliability objective. Regarding the cost objectives: In the formula, For the first i The cost of equipment and installation for each measuring point, where n is the number of measuring points.
[0048] Regarding the accuracy targets: measurement coverage based on hierarchical topology is greater than or equal to 90%, and the data error rate is less than or equal to 5%; Regarding the accuracy targets: the downstream line loss calculation error is less than or equal to 8%, and the fault location accuracy rate is greater than or equal to 90%; The constraints of the cost-accuracy-reliability multi-objective model are: at least one measuring point should be arranged at each level, and the distance between measuring points should be greater than or equal to 50 meters to avoid redundancy.
[0049] S3.2. The non-dominated sorting genetic algorithm is used to solve for the optimal solution in terms of cost, accuracy, and reliability.
[0050] In this embodiment, the NSGA-III algorithm is used to solve for the Pareto Optimal Solution (Pareto), which specifically includes the following steps: S3.2.1 Generate an initial population containing the locations and numbers of measurement points; The measurement point locations are candidate sets of nodes at each level, and the number of measurement points in this embodiment is 100.
[0051] S3.2.2 Based on the hierarchical topology model of the distribution network in the aforementioned area, calculate the cost, accuracy, and reliability indicators of each individual component; S3.2.3. Retain non-dominated solutions and generate the next generation population through crossover mutation; S3.2.4 After the iteration converges, output the Pareto optimal solution set and select the optimal measurement point scheme according to the actual engineering requirements. In this implementation, the number of iterations of the non-dominated sorting genetic algorithm is selected to be 50.
[0052] In other embodiments, the particle swarm optimization algorithm is used to solve the Pareto optimal solution, which is suitable for scenarios with high real-time requirements.
[0053] S3.3. Based on the optimal solution, determine the location and number of measurement points at each level to obtain the optimized scheme.
[0054] Based on the optimal measurement point scheme, the location and number of measurement points at each level are determined, forming a visual layout diagram. It is worth noting that for high reliability requirements, such as in critical load areas, the weight of reliability targets can be increased, and the number of measurement points can be appropriately increased; for low-cost requirements, such as in rural areas, the weight of cost targets can be increased, and the number of measurement points at non-critical levels can be reduced.
[0055] S4. Verify whether the optimization scheme meets the objective. If it does not meet the objective, repeat steps S2-S4.
[0056] Step S4 specifically includes the following steps: S4.1 Build a hierarchical topology simulation model, input the optimization scheme, and verify whether the line loss calculation error and fault location accuracy meet the objectives; S4.2 Deploy the optimization plan in three typical transformer substations, collect on-site measurement data within one month, and compare the indicators before and after optimization; The three typical transformer substation areas are urban densely populated substation areas, rural scattered substation areas, and old and dilapidated substation renovation areas.
[0057] S4.2 If the on-site measurement data does not meet the target, return to S2 to adjust the data processing algorithm parameters, or return to step 3 to optimize the measurement point position until the requirements are met.
[0058] The measurement data and measurement point optimization method of the hierarchical topology model of the distribution network provided by this invention have been verified by simulation and piloted in three typical distribution areas, achieving the following technical effects: Taking a 100-unit transformer substation area as an example, before optimization, 5 measuring points were set up, costing approximately 12,000 yuan; after optimization, 3 measuring points were set up, costing approximately 7,000 yuan. The number of measuring points was reduced by 30%-40% compared to existing experience, and equipment and installation costs were reduced by 35%-45%; the noise rate of measurement data decreased from 8%-12% to 3%-5%, the deviation in data missing data completion decreased from over 10% to within 5%, and the data accuracy was ≥95%; the line loss calculation error decreased from 15%-20% to 6%-8%, the fault location accuracy increased from below 75% to over 90%, and the fault location blind zone was reduced to within 5%; the annual cost saving for line loss management in a single transformer substation area was approximately 8,000-12,000 yuan, and the cumulative cost saving in the three pilot substation areas within six months was 25,000 yuan, with a payback period of ≤1 year.
[0059] The measurement data and measurement point optimization method of the hierarchical topology model of the distribution network provided by this invention can reduce the investment cost of measurement point equipment, enhance the reliability of downstream applications, and bring significant economic benefits. Moreover, this invention can be adapted to different types of distribution areas, such as densely populated urban areas, dispersed rural areas, and old area renovation areas, and the model can be adjusted flexibly without additional adaptation costs.
[0060] Although the present invention has been described in detail through the preferred embodiments above, it should be understood that the above description should not be considered as a limitation of the present invention. Various modifications and substitutions to the present invention will be apparent to those skilled in the art after reading the above description. Therefore, the scope of protection of the present invention should be defined by the appended claims.
Claims
1. A method for optimizing measurement data and measurement points in a hierarchical topology model of a distribution network, characterized in that, Includes the following steps: S1. Construct a hierarchical topology model of the distribution network in the transformer area; S2. Based on the hierarchical topology model of the distribution network in the transformer area, the measurement data is processed hierarchically; S3. Based on the hierarchical topology model of the distribution network in the aforementioned area, the location and number of measuring points are optimized through multi-objective optimization to obtain an optimization scheme; S4. Verify whether the optimization scheme meets the objective. If it does not meet the objective, repeat steps S2-S4.
2. The method for optimizing measurement data and measurement points in a hierarchical topology model of a distribution network area as described in claim 1, characterized in that, The method for constructing a hierarchical topology model of a distribution network includes the following steps: S1.1 Construct a hierarchical adjacency matrix to represent the electrical connection relationships within and between levels; S1.2, Generate the hierarchical electrical feature matrix; S1.3 Verify the hierarchical topology model of the distribution network in the transformer substation to ensure the accuracy of the node connection relationship in the hierarchical topology model of the distribution network in the transformer substation.
3. The method for optimizing measurement data and measurement points in a hierarchical topology model of a distribution network area as described in claim 2, characterized in that, The proposed hierarchical topology model for the distribution network in the transformer substation is constructed by dividing the nodes and lines from the transformer outgoing end to the incoming end of each branch box into the backbone layer; dividing the nodes and lines from the outgoing end of the branch box to the incoming end of the user cluster into the branch layer; and dividing the nodes and lines from the incoming end of the user cluster to each user's meter into the user layer.
4. The method for optimizing measurement data and measurement points in a hierarchical topology model of a distribution network area as described in claim 3, characterized in that, The hierarchical processing of the measurement data includes the following steps: S2.1 Separate interference noise at different levels; S2.2, Use the electrical relationships between levels to complete missing data; S2.
3. Verify the measurement data after hierarchical processing. If the deviation from the hierarchical electrical law exceeds the preset value, repeat S2.1-S2.
3.
5. The method for optimizing measurement data and measurement points in a hierarchical topology model of a distribution network area as described in claim 4, characterized in that, The method for separating interference noise at different levels includes the following steps: A blind source separation algorithm based on a hierarchical adjacency matrix is adopted, which takes synchronous measurement data of the backbone layer and the branch layer as input and outputs clean data after interference removal. The blind source separation algorithm uses the FastICA algorithm: In the formula, For measurement data after separating interference noise, It is a hierarchical adjacency matrix. The collected measurement data; Alternatively, the blind source separation algorithm may employ a wavelet transform algorithm to suit scenarios with high real-time requirements.
6. The method for optimizing measurement data and measurement points in a hierarchical topology model of a distribution network area as described in claim 4, characterized in that, The method for completing missing data using electrical correlations between levels includes the following steps: A hierarchical electrical connectivity model is constructed, and gradient boosting trees are used to train and complete the electrical connectivity model: In the formula, For electrical correlation model, For training data, To complete the missing measurement data, It is a hierarchical adjacency matrix. This is a hierarchical electrical feature matrix; Constraints: Completing missing data must satisfy hierarchical power balance, i.e., the input power of the backbone layer = the sum of the output power of each branch layer + backbone loss.
7. The method for optimizing measurement data and measurement points in a hierarchical topology model of a distribution network area as described in claim 1, characterized in that, The method for optimizing the location and number of measurement points for multiple objectives specifically includes the following steps: S3.1 Construct a multi-objective model of cost, accuracy, and reliability; S3.
2. Use a non-dominated sorting genetic algorithm to find the optimal solution in terms of cost, accuracy, and reliability. S3.
3. Based on the optimal solution, determine the location and number of measurement points at each level to obtain the optimized scheme.
8. The method for optimizing measurement data and measurement points in a hierarchical topology model of a distribution network area as described in claim 7, characterized in that, The method for constructing a cost-accuracy-reliability multi-objective model includes the following steps: constructing the multi-objective model from three aspects: cost objective, accuracy objective, and reliability objective; wherein... Regarding the cost objectives: In the formula, For the first i The equipment and installation cost for each measuring point, where n is the number of measuring points; Regarding the accuracy targets: measurement coverage based on hierarchical topology is greater than or equal to 90%, and the data error rate is less than or equal to 5%; Regarding the accuracy targets: the downstream line loss calculation error is less than or equal to 8%, and the fault location accuracy rate is greater than or equal to 90%; The constraints of the cost-accuracy-reliability multi-objective model are: at least one measuring point should be arranged at each level, and the distance between measuring points should be greater than or equal to 50 meters to avoid redundancy.
9. The method for optimizing measurement data and measurement points in a hierarchical topology model of a distribution network area as described in claim 7, characterized in that, The non-dominated sorting genetic algorithm uses the NSGA-III algorithm. The method of the NSGA-III algorithm to find the optimal solution in terms of cost, accuracy, and reliability includes the following steps: S3.2.1 Generate an initial population containing the locations and numbers of measurement points; S3.2.2 Based on the hierarchical topology model of the distribution network in the aforementioned area, calculate the cost, accuracy, and reliability indicators of each individual component; S3.2.
3. Retain non-dominated solutions and generate the next generation population through crossover mutation; S3.2.4 After the iteration converges, output the optimal solution set of cost-accuracy-reliability, and select the optimal measurement point scheme according to the actual engineering requirements; Alternatively, the non-dominated sorting genetic algorithm may employ a particle swarm optimization algorithm to suit scenarios with high real-time requirements.
10. The method for optimizing measurement data and measurement points in a hierarchical topology model of a distribution network area as described in claim 1, characterized in that, Step S4 includes the following steps: S4.1 Build a hierarchical topology simulation model, input the optimization scheme, and verify whether the line loss calculation error and fault location accuracy meet the objectives; S4.2 Deploy the optimization plan in three typical transformer substations, collect on-site measurement data within one month, and compare the indicators before and after optimization; S4.2 If the on-site measurement data does not meet the target, return to S2 to adjust the data processing algorithm parameters, or return to the third step to optimize the measurement point position until the requirements are met; The three typical transformer substation areas are urban densely populated substation areas, rural scattered substation areas, and old and dilapidated substation renovation areas.