An Adaptive Repair Method and System for Distribution Network Line Loss

Abstract

This application relates to the technical field of line loss repair, and provides a method and system for adaptive repair of distribution network line loss. The method includes: reading the line topology network to determine the ternary partition characteristics; building an abnormal positioning module; making a loss determination decision and probability calculation to determine the single column of line loss positioning; building a network reconstruction model, traversing the single column of line loss positioning, performing regional segmentation reconstruction and self-healing correction analysis to determine the reconstruction data; traversing the single column of line loss positioning, reading the non-self-healing line loss positioning data, and generating an abnormal repair instruction; combining the reconstruction data with the abnormal repair instruction to perform line loss repair management. This application solves the technical problems that the traditional repair method cannot accurately judge the line loss position and cannot adaptively adjust according to the real-time state of the distribution network, resulting in more ineffective repair operations and low repair efficiency, and realizes the improvement of the accuracy and repair efficiency of distribution network line loss positioning and the reduction of ineffective repair operations through network reconstruction and abnormal traceability repair.

Description

Technical Field

[0001] This application relates to the technical field of distribution network operation and maintenance, specifically to the technical field of line loss repair, and particularly to a method and system for adaptive repair of distribution network line loss. Background Art

[0002] With the rapid development of the power industry and the in-depth promotion of the construction of smart grids, the line loss management of distribution networks has become an increasingly important issue. Distribution network line loss not only directly affects the economic benefits of power enterprises, but also is related to the stable operation of the entire power grid and the energy utilization efficiency. However, with the continuous growth of power demand and the increasing complexity of the power grid structure, the line loss problem of distribution networks has become increasingly prominent. Due to the complexity of distribution networks, traditional line loss repair methods are difficult to cover all possible line loss points, resulting in some line loss problems being difficult to detect in a timely manner, and the operating state of the distribution network cannot be sensed in real time. Therefore, it is impossible to adaptively adjust the repair strategy according to real-time data, resulting in low repair efficiency and even possible ineffective repair operations. Summary of the Invention

[0003] This application provides a method and system for adaptive repair of distribution network line loss, aiming to solve the technical problems that traditional repair methods cannot accurately judge the line loss location, cannot adaptively adjust according to the real-time state of the distribution network, resulting in many ineffective repair operations and low repair efficiency.

[0004] In view of the above problems, this application provides a method and system for adaptive repair of distribution network line loss.

[0005] In the first aspect disclosed in this application, a method for adaptive repair of distribution network line loss is provided. The method includes: reading the line topology network of the target distribution network and determining the three-way partition features, where the three-way partition features include voltage-divided line loss, area-divided line loss, and component-divided line loss; based on the line topology network and the three-way partition features, building an anomaly location module, the anomaly location module includes parallel three-way decision branches, and the anomaly location module is built-in with a line loss probability calculation formula; transmitting back the distribution network data, combining with the anomaly location module, performing loss determination decision and probability calculation to determine the line loss location single column; building a network reconstruction model based on the network coupling characteristics, taking the regional network reconstruction as the line loss repair principle, traversing the line loss location single column, performing regional segmentation reconstruction and self-healing correction analysis to determine the reconstruction data, the reconstruction data includes network configuration; traversing the line loss location single column, reading the non-self-healing line loss location data, generating an anomaly repair instruction; combining the reconstruction data and the anomaly repair instruction to perform line loss repair management on the target distribution network.

[0006] Another aspect disclosed in this application provides a self - adaptive repair system for distribution network line loss. The system includes: a ternary partition feature unit for reading the line topology network of the target distribution network and determining ternary partition features, where the ternary partition features include voltage - divided line loss, area - divided line loss, and component - divided line loss; an abnormal location module building unit for building an abnormal location module based on the line topology network and the ternary partition features. The abnormal location module includes parallel ternary decision branches, and the abnormal location module has a built - in line loss probability calculation formula; a line loss location single - column determination unit for transmitting back the distribution network data, combining with the abnormal location module, making damage determination decisions and probability calculations to determine the line loss location single - column; a reconstructed data determination unit for building a network reconstruction model based on network coupling characteristics, taking regional network reconstruction as the line loss repair principle, traversing the line loss location single - column, performing regional segmentation reconstruction and self - healing correction analysis to determine the reconstructed data, where the reconstructed data includes network configuration; an abnormal repair instruction generation unit for traversing the line loss location single - column, reading the non - self - healing line loss location data, and generating abnormal repair instructions; a line loss repair management unit for combining the reconstructed data and the abnormal repair instructions to manage the line loss repair of the target distribution network.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] The above-mentioned method for self-adaptive repair of distribution network line loss first reads the line topology network of the distribution network, and constructs an abnormal location module according to this network structure and the ternary division characteristics of voltage division, area division, and component division. This module consists of three parallel decision branches, and each branch is built with a corresponding line loss probability calculation formula to quickly and accurately judge the possible line loss positions and degrees in the distribution network. Subsequently, the operation data of the distribution network are transmitted back in real time and combined with the abnormal location module to make a damage determination decision and probability calculation. Through this process, a specific line loss location list can be determined, that is, the specific positions or areas with line loss problems. After that, based on the network coupling characteristics of the distribution network, a network reconstruction model is built. This model traverses the determined line loss location list based on the principle of regional network reconstruction, and performs regional segmentation reconstruction and self-healing correction analysis. This process aims to reduce or eliminate line loss problems by optimizing the network configuration and improve the operation efficiency of the distribution network. During the traversal process, if it is found that some line loss problems cannot be self-healed through network reconstruction, the non-self-healing line loss location data will be read and corresponding abnormal repair instructions will be generated. Then, combined with the network reconstruction data and abnormal repair instructions, line loss repair management is carried out on the target distribution network. The whole process realizes the accurate location, self-adaptive repair and efficient management of the distribution network line loss problem, and improves the operation efficiency and stability of the distribution network.

[0009] The above description is only an overview of the technical solution of this application. In order to be able to understand the technical means of this application more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features and advantages of this application more obvious and understandable, the following specifically illustrates the specific implementation manners of this application. Brief Description of the Drawings

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained according to these drawings.

[0011] Figure 1 It is a schematic flow chart of a method for self-adaptive repair of distribution network line loss in an embodiment.

[0012] Figure 2 It is an architecture diagram of a self-adaptive repair system for distribution network line loss in an embodiment.

[0013] Description of the reference numerals: ternary division feature unit 1, abnormal location module building unit 2, line loss location list determination unit 3, reconstruction data determination unit 4, abnormal repair instruction generation unit 5, line loss repair management unit 6. Detailed Description of the Invention

[0014] By providing a method and system for adaptive repair of distribution network line loss in an embodiment of the present application, the technical problems that the traditional repair method cannot accurately judge the line loss position and cannot adaptively adjust according to the real-time state of the distribution network, resulting in a large number of ineffective repair operations and low repair efficiency are solved.

[0015] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present application without creative efforts shall fall within the protection scope of the present application.

[0016] It should be noted that the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or server that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or are inherent to these processes, methods, products, or devices.

[0017] In the first embodiment, as Figure 1 shown, the present application provides a method for adaptive repair of distribution network line loss, and the method includes:

[0018] Read the line topology network of the target distribution network and determine the three-way partition characteristics, where the three-way partition characteristics include voltage-divided line loss, area-divided line loss, and component-divided line loss.

[0019] With the rapid development of the power industry, electricity has become an indispensable energy source for the operation of modern society. As the link directly connected to users in the power system, the operation state of the distribution network directly affects the power consumption quality of users and the economic benefits of the power system. However, with the rapid development of industrial production and transportation, the power demand is continuously increasing, and the scale and complexity of the distribution network are also increasing continuously, which leads to the increasingly prominent problem of line loss in the distribution network. Line loss, that is, the energy loss caused by physical factors such as resistance, inductance, and conductance during the transmission and distribution of electricity, is an issue that cannot be ignored in the distribution network. It not only causes waste of energy, but also affects the stable operation of the power grid, reduces the power supply efficiency, and may even cause safety accidents.

[0020] In the embodiment of the present application, in the line loss management of the target distribution network, the system terminal reads the line topology network of the target distribution network to comprehensively understand the line structure of the target distribution network. The line topology network is a connection relationship diagram of each line and device in the target distribution network, which reflects the overall structure of the target distribution network. Subsequently, based on this topology network, the system terminal determines the ternary division characteristics of the target distribution network. The ternary division characteristics refer to voltage-divided line loss, area-divided line loss, and component-divided line loss. Among them, the voltage-divided line loss refers to the energy loss generated during the power transmission and distribution process in the target distribution network due to different voltage levels. This loss can help understand the energy efficiency under each voltage level, so as to optimize power transmission and distribution. The area-divided line loss refers to the energy loss generated when the target distribution network transmits electric energy in a specific area. By analyzing the area-divided line loss, the operation efficiency of the target distribution network in this area can be evaluated. The component-divided line loss refers to the energy loss generated by each component in the target distribution network, such as transformers, cables, etc., during the power transmission and distribution process. Understanding the line loss conditions of these components helps to accurately identify components with low energy efficiency.

[0021] Based on the line topology network and the ternary division characteristics, an anomaly location module is built. The anomaly location module includes parallel ternary decision branches, and the anomaly location module is built-in with a line loss probability calculation formula.

[0022] In one embodiment, based on the line topology network and the ternary partition characteristics of the target distribution network, the system terminal constructs an anomaly location module. The core of this module is three parallel decision branches, namely ternary decision branches, which independently analyze the line losses of voltage division, area division, and component division respectively. Inside each decision branch, a line loss probability calculation formula is integrated, which can estimate the probability of line loss occurring under specific conditions. Specifically, the system terminal collects the voltage division data of the sample distribution network and the sample voltage division line loss degree to construct a training set, a validation set, and a test set. Subsequently, according to the complexity of the problem and the characteristics of the data, the structures of the input layer, hidden layer, and output layer of the neural network are determined, including the number of neurons in each layer, activation functions, etc. Then, the hyperparameters of the neural network are configured, such as the learning rate, batch size, number of iterations, etc. After that, the training set is input into the neural network to train the neural network model. During the training process, the predicted voltage division line loss degree of the neural network is calculated through forward propagation and compared with the sample voltage division line loss degree, and the mean square error is used to calculate the value of the loss function. Then, the backpropagation algorithm is used to calculate the gradient of the loss function with respect to the neural network parameters, and the neural network parameters are updated to minimize the loss function. After each training cycle, the validation set is used to evaluate the performance of the neural network, and the hyperparameters are adjusted or the training is stopped as needed to prevent overfitting. After the training is completed, the system terminal uses the test set to evaluate the trained neural network model. Metrics such as the prediction accuracy, recall rate, and F1 value of the neural network are calculated to evaluate the performance of the neural network on unknown data. If the expected expectations are met, the current neural network is used as the voltage division line loss decision branch. Otherwise, the structure of the neural network and the hyperparameters are adjusted, etc. After constructing the voltage division line loss decision branch, the system terminal integrates the pre-constructed line loss probability calculation formula into the output layer of the voltage division line loss decision branch to calculate the prior probability and posterior probability of the corresponding node of the currently input distribution network data. After the integration of the line loss probability calculation formula is completed, the system terminal also integrates the matching algorithm into the output layer. This matching algorithm can label the calculated prior probability and posterior probability, as well as the voltage division line loss degree according to the node information to the line topology network respectively, forming a voltage division probability distribution network and a voltage division line loss degree distribution network. These two networks will be used as the final output of the voltage division line loss decision branch. Similarly, the system terminal uses the same method as above to train the area division line loss decision branch and the component division line loss decision branch, and connects these three branches in parallel to form a ternary decision branch. Further, the system terminal constructs an input layer and an output layer, and encapsulates them with this ternary decision branch to build an anomaly location module. Among them, the input layer has a built-in data division rule, which can divide the input distribution network data into distribution network voltage division data, distribution network area division data, and distribution network component division data according to historical experience for subsequent analysis by the ternary decision branch.The output layer stores the line topology network of a target distribution network, which is used to summarize the voltage division probability distribution network and the voltage division line loss degree distribution network generated by the ternary decision branch, the partition probability distribution network and the partition line loss degree distribution network, and the component probability distribution network and the component line loss degree distribution network, to form the probability distribution network and the line loss degree distribution network, as the final output of the anomaly location module. Through this built anomaly location module, the system terminal can more quickly and accurately identify potential problem areas or components in the target distribution network, so as to take corresponding measures to reduce or avoid energy losses and improve the operation efficiency and stability of the target distribution network.

[0023] Furthermore, the present application provides the line loss probability calculation formula, including:

[0024] The line loss probability calculation formula includes the prior probability and the posterior probability. Obtaining the line loss probability calculation formula includes: ; where is the prior probability, is the eigenvalue of the visual feature, is the weight of the visual feature, is the total amount of visual features; is the posterior probability, is the degradation probability matrix of the visual feature, is the transition probability matrix of the hidden feature; is the eigenvalue of the hidden feature, is the weight of the hidden feature, is the total amount of hidden features, is the degradation probability of the visual feature, is the transition probability of the hidden feature.

[0025] Optionally, the line loss probability calculation formula is divided into prior probability calculation and posterior probability calculation to calculate the line loss probability. Two types of features are used, namely visual features and hidden features, and different weights are assigned to them. The line loss probability calculation formula is specifically as follows: ; where is the prior probability, which is a calculated value. is the eigenvalue of the visual feature, which is a feature that can be directly observed or measured, such as current, voltage, etc. is the weight of the visual feature, which determines the influence degree of each visual feature on the prior probability and the posterior probability. Is the total amount of visual features. Is the posterior probability, which is a calculated value. Is the deterioration probability matrix of visual features. Deterioration refers to the possibility that the feature value changes from the normal state to the abnormal state. Each data inside represents the deterioration probability of a certain visual feature of the current node, which is pre-statistically obtained based on historical data. Is the transition probability matrix of hidden features, which is also pre-statistically obtained based on historical data. Is the Eigenvalue of the hidden feature, which is a factor that cannot be directly observed but may affect line loss. For example, parameters such as aging state and load fluctuation. Is the Weight of the hidden feature, which determines the influence degree of each hidden feature on the posterior probability. Is the total amount of hidden features. Is the Deterioration probability of the visual feature. Is the Transition probability of the hidden feature. The prior probability is obtained by multiplying the eigenvalue of each visual feature by its corresponding weight, then adding all the products and dividing by the total amount of visual features, providing a preliminary estimate based on the current observed data. The posterior probability not only considers the current observed data but also the potential influence of hidden factors. This enables the posterior probability to more comprehensively reflect the possibility of line loss.

[0026] Backhaul the distribution network data, combine with the abnormal location module, make a loss determination decision and probability calculation, and determine the single column of line loss location.

[0027] In one embodiment, a large amount of real-time data, i.e., distribution network data, will be generated during the operation of the target distribution network. These data are backhauled to the system terminal through data acquisition devices. Subsequently, the system terminal inputs these data into the abnormal location module for loss determination decision and probability calculation. The abnormal location module collaboratively analyzes through the internal input layer, ternary decision branch, and output layer to generate a probability distribution network and a line loss degree distribution network. Then, the probability distribution network and the line loss degree distribution network are combined and screened multiple times to generate a single column of line loss location. This single column will detail key information such as the description of abnormal events, location information, possible causes, expected losses, and calculated probabilities. This single column can help operators quickly locate the problem to reduce losses and restore the normal operation of the target distribution network.

[0028] Furthermore, the present application provides the determination of the single column of line loss location, including:

[0029] Based on the ternary decision branches, the distribution network data is analyzed, proofread, and fused in parallel to determine the probability distribution network and the line loss degree distribution network; based on the line loss degree distribution network, a first screening is performed in combination with the standard line loss coefficient to determine the first loss-determining data, and a second screening is performed based on the probability distribution network to determine the second loss-determining data; the first loss-determining data and the second loss-determining data are fitted to determine the line loss location single column.

[0030] Preferably, the system terminal inputs the distribution network data into the abnormal location module for loss determination decision-making and probability calculation. The abnormal location module divides the input distribution network data through the input layer to generate distribution network voltage-divided data, distribution network partitioned data, and distribution network component data. Subsequently, the ternary decision branches inside the abnormal location module calculate the corresponding data after division, and each decision branch will generate a corresponding probability distribution network and a line loss degree distribution network. Then, the output layer of the abnormal location module fuses the probability distribution network and the line loss degree distribution network generated by each decision branch to determine the final probability distribution network and the line loss degree distribution network. Then, the system terminal uses the standard line loss coefficient to perform a first screening on the line loss degree distribution network. During the first screening process, the system terminal compares the line loss degree of each node in the line loss degree distribution network with the standard line loss coefficient, extracts the nodes whose line loss degree is greater than or equal to the standard line loss coefficient, and adds them to the pre-constructed blank first loss-determining data. This standard line loss coefficient is determined based on historical data and safety criteria. This first loss-determining data represents the initially determined areas or nodes that may have problems. After completing the first screening, the system terminal performs a second screening on the probability distribution network based on the prior probability threshold and the posterior probability threshold to determine the second loss-determining data. Finally, the obtained first loss-determining data and the second loss-determining data are fitted, that is, the two sets of data are combined to form a more comprehensive and accurate line loss location single column. This line loss location single column contains a list of all areas or nodes identified as having anomalies, which is used to guide subsequent corrections.

[0031] Furthermore, the present application provides a second screening based on the probability distribution network, including:

[0032] Obtain the probability distribution network, where each network node includes a prior probability and a posterior probability; traverse the probability distribution network, and perform a threshold overrun determination based on the prior probability threshold and the posterior probability threshold; if the prior probability meets the prior probability threshold and the posterior probability meets the posterior probability threshold, add the network node to the second loss-determining data; if either the prior probability or the posterior probability does not meet the probability threshold, mark the network node as a pseudo-abnormal point.

[0033] Optionally, after obtaining the probability distribution network, the system terminal analyzes this probability distribution network to understand the prior probability and posterior probability stored in each node. Subsequently, the probability distribution network is traversed, and in each traversal, the system terminal makes a threshold overrun determination for each network node. This determination is based on the preset prior probability threshold and posterior probability threshold. These two thresholds are set according to historical data and safety criteria and are used to determine whether a node may have experienced a power loss. If the prior probability of a certain node meets the prior probability threshold and the posterior probability also meets the posterior probability threshold, then it is determined that this node may have experienced a power loss, and thus it is added to the pre-constructed blank second loss assessment data. This second loss assessment data can determine which areas or nodes may have power loss problems. However, if either the prior probability or the posterior probability of a certain node does not meet the corresponding threshold, then this node is determined to be a false anomaly point. These nodes may be caused by false alarms, noise, or other reasons, and they do not truly represent the occurrence of a power loss. By identifying these points as false anomaly points, unnecessary attention and processing can be avoided, thereby improving the accuracy and efficiency of the entire loss assessment process.

[0034] Build a network reconstruction model based on network coupling characteristics. Taking regional network reconstruction as the principle for line loss repair, traverse the single column of line loss location, conduct regional segmentation reconstruction and self-healing correction analysis, and determine the reconstruction data, where the reconstruction data includes network configuration.

[0035] In one embodiment, the system terminal first clarifies the goal of network reconstruction, that is, to reconfigure and optimize the network structure for a specific area with problems, so as to restore the normal operation state of the target distribution network and reduce or eliminate the line loss problem. Subsequently, the coupling relationship in the network is analyzed using the distribution network data. This includes identifying key nodes and lines, evaluating the degree of mutual influence between them, and determining their positions and roles in the network. Then, according to the analysis results of the network coupling characteristics, a network reconstruction model is set, such as the threshold of neurons, coupling strength, voltage level, current magnitude, etc. The setting of these parameters will directly affect the performance and accuracy of the model. After that, according to the electrical coupling method, the connection relationship between nodes and lines is implemented in the model. That is, nodes such as generators, transformers, and loads are connected by lines. Then, according to the mesh topology structure, the connection mode between nodes and lines is constructed in the model, and then the network reconstruction model is built. The mesh topology structure is often selected for power network reconstruction due to its high reliability and fault tolerance. After the network reconstruction model is built, for effective repair, the system terminal takes regional network reconstruction as the principle, that is, on the premise of maintaining the stable operation of the overall network, the network structure of a specific area with problems is reconfigured and optimized. The system terminal traverses the single column of line loss location, and divides and reconstructs these areas in the network reconstruction model according to the network coupling characteristics. Division means temporarily isolating the problem area from the network reconstruction model for separate analysis and processing. Reconstruction is to reconfigure the network structure of the problem area according to the division result to determine multiple transformer data. During the reconstruction process, self-healing correction analysis is also required. Self-healing correction means that when a fault or problem occurs in the target distribution network, through collision avoidance analysis, the transformer data is spliced to determine the network configuration to restore the normal operation state of the target distribution network. After obtaining the network configuration, the system terminal sorts out the network configuration to form reconstruction data, which contains specific information about the network configuration, such as node positions, line connections, equipment parameters, etc. These reconstruction data will be used to guide the actual network repair work to restore the normal operation state and reduce or eliminate the line loss problem.

[0036] Further, the present application provides the above-mentioned regional division and reconstruction and self-healing correction analysis, including:

[0037] Based on the network coupling characteristics, traverse the single column of line loss location to determine multiple damaged transformer regions, where the coupling degree within the transformer region is higher than the coupling threshold, and the coupling degree between transformer regions is lower than the coupling threshold; traverse the multiple damaged transformer regions to perform regional network reconstruction to determine multiple transformer data, where the regional network reconstruction analysis method is multi-objective optimization; perform collision avoidance analysis on the multiple transformer data and splice to determine the network configuration.

[0038] Preferably, the system terminal determines the high and low coupling degrees in the network by analyzing the network coupling characteristics. Subsequently, it traverses the single-column data of line loss positioning and identifies the areas with coupling degrees higher than the preset coupling threshold as distribution transformer areas. The couplings within these areas are tight, but the coupling degrees between the areas are low, thus forming relatively independent network areas. After obtaining multiple line loss-determined distribution transformer areas, the system terminal traverses all the line loss-determined distribution transformer areas, assigns a unique identifier to each area, and then initializes the network structure of each area, which includes the positions of nodes, the connections of lines, etc. After that, according to the objectives of network reconstruction, multiple objective functions are set, for example, improving network efficiency, reducing line loss, maintaining model complexity, etc. Then, corresponding weights are assigned according to the importance of each objective function. Then, the network configuration of each line loss-determined distribution transformer area is used as a chromosome, and the fitness of each chromosome is evaluated according to the weighted value of the objective function. Chromosomes with higher fitness are selected to enter the next generation. Two chromosomes are randomly selected for crossover operation to generate new offspring chromosomes, that is, new network configurations. This helps to explore new areas in the solution space. Random mutation operations are performed on the offspring chromosomes to introduce new genes, that is, new network configuration parameters. This helps to maintain the diversity of the population and avoid falling into local optima. Further, the system terminal repeats the selection, crossover, and mutation processes to generate a new generation of chromosome populations. Then, the fitness of the new generation of chromosomes is evaluated, and the optimal or near-optimal chromosomes are recorded. Subsequently, the above optimization process is iterated until the improvement of fitness is no longer significant. After that, the optimal or near-optimal chromosomes are selected from the final generation of chromosomes as the distribution transformer data for each area, and the key parameters of these chromosomes, such as node positions, line connections, equipment parameters, etc., are extracted to determine multiple distribution transformer data. Then, all the distribution transformer data are traversed to identify possible collision points between different areas. Collision points include node overlaps, line crossings, equipment conflicts, etc. Then, for the identified collision points, severity assessment is performed. The severity is determined according to factors such as the impact of the collision point on network performance, repair cost, time, etc. The higher the severity, the higher the processing priority of the corresponding collision point. After the priorities are determined, the system terminal re-performs multi-objective optimization on the multiple distribution transformer data according to the determined priorities to avoid conflicts and ensure that each area can reach the optimal state. Finally, the distribution transformer data of each area after collision avoidance analysis are spliced to form a complete network configuration. This configuration not only considers the influence of network coupling characteristics but also is optimized through multi-objective optimization and collision avoidance analysis, which can effectively reduce line loss and improve the stability and efficiency of the target distribution network.

[0039] Further, the present application provides for adjusting the area scale of the multiple line loss-determined distribution transformer areas, including:

[0040] Based on the correction requirements, the area scale of the multiple line loss-determined distribution transformer areas is adjusted, where the scale adjustment methods include area expansion and area contraction.

[0041] Optionally, over time, factors such as network status and load conditions may change. Therefore, these loss-determined and distribution-transformer regions also need to be adjusted or updated accordingly. When the network status or load condition near a loss-determined and distribution-transformer region changes, resulting in the need to expand the boundary of the region to better cover potential problem areas that may exist, the system terminal performs outwards expansion of the region. This involves incorporating certain devices or lines that originally did not belong to the region into the new loss-determined and distribution-transformer region. Conversely, if the network status or load condition within a loss-determined and distribution-transformer region improves, or if an area that was originally considered problematic is found to be normal after inspection, then the boundary of the region may need to be contracted inwards to reduce unnecessary resource waste and operation and maintenance costs.

[0042] Traverse the single column of line loss location, read the non-self-healing line loss location data, and generate an abnormal repair instruction.

[0043] In one embodiment, when performing the operation of traversing the single column of line loss location, the system terminal also traverses out the non-self-healing line loss location data. Non-self-healing line loss refers to the power loss in the target distribution network due to reasons such as equipment failure, line aging, and power theft, and this loss cannot be automatically restored through the self-adjustment or repair of the target distribution network itself. The traversal process involves checking each line loss location data and analyzing whether it indicates a persistent power loss problem that requires manual intervention. Once such non-self-healing line loss location data is identified, the system terminal performs line loss cause tracing and determines an abnormal repair instruction based on the repair time limit. This abnormal repair instruction includes information such as specific repair measures, repair time, required materials, and personnel allocation, which provide clear guidance and reference for the operator to help quickly and accurately find and repair the problem, thereby restoring the normal operation of the target distribution network.

[0044] Furthermore, the present application provides the generation of the abnormal repair instruction, including:

[0045] Perform line loss cause tracing on the non-self-healing line loss location data to determine a set of repair plans; based on the line loss level and network impact, set a repair time limit based on the repair priority, where the repair priority is positively correlated with the line loss level and network impact; map the set of repair plans and the repair time limit to generate an abnormal repair instruction.

[0046] Preferably, for the non-self-healing line loss location data, the system terminal traces the root causes of the line loss. The purpose of this step is to identify the fundamental reasons for the line loss. By deeply analyzing these causes, it is possible to better understand the essence of the problem and provide an important basis for formulating subsequent repair measures. Subsequently, based on the tracing results, the system terminal determines a set of repair plans. These plan sets contain multiple repair plans for different causes, and each plan details the repair steps, required materials, personnel allocation, etc. By preparing multiple plans, the most suitable repair plan can be flexibly selected according to the actual situation, improving the repair efficiency. After determining the set of repair plans, the system terminal sets the repair time limit according to the line loss level and network impact. The line loss level reflects the severity of the line loss, while the network impact measures the scope of the problem's impact on the entire target distribution network. The repair priority is positively correlated with these two factors, that is, the more severe the line loss and the wider the impact range, the higher the repair priority and the more urgent the repair time limit. After that, the system terminal maps the set of repair plans to the repair time limit to generate abnormal repair instructions. These instructions detail key information such as the repair measures to be taken, the operators, and the completion time. By sending these instructions to the operators, it is possible to ensure the rapid and accurate execution of the repair tasks, thereby restoring the normal operation of the target distribution network in the shortest time.

[0047] Combining the reconstructed data with the abnormal repair instructions, perform line loss repair management on the target distribution network.

[0048] In one embodiment, during the line loss repair management process of the target distribution network, the system terminal combines the reconstructed data with the generated abnormal repair instructions to perform targeted repair on the target distribution network. By implementing these instructions, the line loss problem in the target distribution network can be effectively solved, improving the operation efficiency and stability of the target distribution network. At the same time, it is also possible to continuously monitor the operation status of the target distribution network, promptly discover new problems and repair them, forming a dynamic and cyclic management process.

[0049] In summary, the embodiments of the present application have at least the following technical effects:

[0050] In an embodiment of the present application, the line topology network of the target distribution network is read to determine the ternary partition characteristics including voltage division, zoning, and component division. Based on these characteristics and the line topology network, an anomaly location module is constructed. This module includes parallel ternary decision branches for combining distribution network data to make loss determination decisions and probability calculations, thereby determining a single column of line loss location. Subsequently, a network reconstruction model is constructed using the network coupling characteristics to perform regional segmentation reconstruction and self-healing correction analysis on the single column of line loss location to determine the reconstruction data, including network configuration. During this process, the loss determination distribution transformer area is divided based on the coupling degree between regions, and regional network reconstruction is performed through multi-objective optimization. Then, the single column of line loss location is traversed to identify non-self-healing line loss location data, and the cause of the line loss is traced to determine the corresponding repair plan set. According to the line loss level and network impact, repair priorities and repair time limits are set for each repair plan. Then, combining the reconstruction data and the anomaly repair instructions, line loss repair management is performed on the target distribution network. During this process, the repair plan is dynamically adjusted and optimized by combining the real-time data of the distribution network to achieve efficient line loss repair. In addition, it includes steps to adjust the scale of the loss determination distribution transformer area to meet different correction requirements, and the area scale is adjusted by expanding the area or shrinking the area. These technical effects together solve the technical problems that traditional repair methods cannot accurately judge the line loss location and cannot adaptively adjust according to the real-time state of the distribution network, resulting in many ineffective repair operations and low repair efficiency, realizing network reconstruction and anomaly tracing repair, improving the accuracy and repair efficiency of distribution network line loss location, and reducing the effect of ineffective repair operations.

[0051] Embodiment 2, based on the same inventive concept as a method for adaptive repair of distribution network line loss in the foregoing embodiment, as Figure 2 shown, the present application provides a system for adaptive repair of distribution network line loss, and the system includes:

[0052] Three - element division feature unit 1: The three - element division feature unit 1 is used to read the line topology network of the target distribution network and determine the three - element division features, where the three - element division features include voltage - dividing line loss, area - dividing line loss, and component - dividing line loss; Abnormal location module building unit 2: The abnormal location module building unit 2 is used to build an abnormal location module based on the line topology network and the three - element division features. The abnormal location module includes parallel three - element decision branches, and the abnormal location module has a built - in line - loss probability calculation formula; Line - loss location single - column determination unit 3: The line - loss location single - column determination unit 3 is used to transmit the distribution network data back, combine with the abnormal location module, make a damage determination decision and probability calculation, and determine the line - loss location single - column; Reconstructed data determination unit 4: The reconstructed data determination unit 4 is used to build a network reconstruction model based on the network coupling characteristics, take the regional network reconstruction as the line - loss repair principle, traverse the line - loss location single - column, perform regional segmentation reconstruction and self - healing correction analysis, and determine the reconstructed data, where the reconstructed data includes network configuration; Abnormal repair instruction generation unit 5: The abnormal repair instruction generation unit 5 is used to traverse the line - loss location single - column, read the non - self - healing line - loss location data, and generate abnormal repair instructions; Line - loss repair management unit 6: The line - loss repair management unit 6 is used to combine the reconstructed data and the abnormal repair instructions to manage the line - loss repair of the target distribution network.

[0053] Further, the line - loss location single - column determination unit 3 is also used to execute the following method:

[0054] Based on the three - element decision branches, perform parallel analysis and proofreading fusion on the distribution network data to determine the probability distribution network and the line - loss degree distribution network; Based on the line - loss degree distribution network, combine with the standard line - loss coefficient for the first screening to determine the first damage - determination data, and perform the second screening based on the probability distribution network to determine the second damage - determination data; Fit the first damage - determination data and the second damage - determination data to determine the line - loss location single - column.

[0055] Further, the line - loss location single - column determination unit 3 is also used to execute the following method:

[0056] The line - loss probability calculation formula includes prior probability and posterior probability. Obtaining the line - loss probability calculation formula includes: ; where is the prior probability, is the eigenvalue of the visual feature, is the weight of the visual feature, is the total amount of visual features; is the posterior probability, is the deterioration probability matrix of the visual feature, is the transition probability matrix of hidden features; is the eigenvalue of the hidden feature, is the weight of the hidden feature, is the total amount of hidden features, is the deterioration probability of the visualization feature, is the transition probability of the hidden feature.

[0057] Furthermore, the single-column line loss location determination unit 3 is further configured to execute the following method:

[0058] Obtain a probability distribution network, where each network node includes a prior probability and a posterior probability; traverse the probability distribution network, and based on the prior probability threshold and the posterior probability threshold, perform a threshold exceeding determination; if the prior probability meets the prior probability threshold and the posterior probability meets the posterior probability threshold, add the network node to the second loss determination data; if either the prior probability or the posterior probability does not meet the probability threshold, mark the network node as a pseudo-abnormal point.

[0059] Furthermore, the reconstructed data determination unit 4 is further configured to execute the following method:

[0060] Based on the network coupling characteristics, traverse the single-column line loss location to determine multiple loss determination transformer regions, where the coupling degree within the transformer region is higher than the coupling threshold and the coupling degree between the transformer regions is lower than the coupling threshold; traverse the multiple loss determination transformer regions to perform regional network reconstruction to determine multiple transformer data, where the regional network reconstruction analysis method is multi-objective optimization; perform collision avoidance analysis on the multiple transformer data and splice to determine the network configuration.

[0061] Furthermore, the reconstructed data determination unit 4 is further configured to execute the following method:

[0062] Based on the correction requirements, perform regional scale adjustment on the multiple loss determination transformer regions, where the scale adjustment methods include regional expansion and regional contraction.

[0063] Furthermore, the abnormal repair instruction generation unit 5 is further configured to execute the following method:

[0064] For non-self-healing line loss location data, trace the source of the line loss cause to determine a set of repair plans; based on the line loss level and network impact, set the repair time limit based on the repair priority, where the repair priority is positively correlated with the line loss level and network impact; map the set of repair plans and the repair time limit to generate an abnormal repair instruction.

[0065] It should be noted that the above order of the embodiments of the present application is only for description and does not represent the superiority or inferiority of the embodiments. And the above specific embodiments of this specification have been described. The processes depicted in the drawings do not necessarily require the specific order and continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0066] The above are only the preferred embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included within the protection scope of the present application.

[0067] This specification and the drawings are only exemplary descriptions of the present application and are considered to have covered any and all modifications, variations, combinations, or equivalents within the scope of the present application. Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalent technologies, the present application is intended to include these changes and modifications.

Claims

1. An adaptive repair method for distribution network line loss, characterized in that The method includes: Reading the line topology network of the target distribution network and determining the ternary partition features, where the ternary partition features include voltage-dividing line loss, sectional line loss, and component line loss; Based on the line topology network and the ternary partition features, building an anomaly location module, which includes parallel ternary decision branches, and the anomaly location module has a built-in line loss probability calculation formula; Transmitting back the distribution network data, combining with the anomaly location module, making a damage assessment decision and probability calculation, and determining a single column of line loss location, which includes a list of all areas or nodes identified as having anomalies; Based on the network coupling characteristics, building a network reconstruction model, taking regional network reconstruction as the principle of line loss repair, traversing the single column of line loss location, and performing regional segmentation reconstruction and self-healing correction analysis to determine the reconstruction data, which includes network configuration; Traversing the single column of line loss location, reading the non-self-healing line loss location data, and generating an anomaly repair instruction; Combining the reconstruction data and the anomaly repair instruction, performing line loss repair management on the target distribution network; The determination of the single column of line loss location includes: Based on the ternary decision branches, performing parallel analysis and verification fusion on the distribution network data to determine the probability distribution network and the line loss degree distribution network; Based on the line loss degree distribution network, performing a first screening in combination with the standard line loss coefficient to determine the first damage assessment data, and performing a second screening based on the probability distribution network to determine the second damage assessment data; Fitting the first damage assessment data and the second damage assessment data to determine the single column of line loss location.

2. The self-adaptive repair method for distribution network line loss according to claim 1, characterized in that The line loss probability calculation formula includes prior probability and posterior probability. Obtaining the line loss probability calculation formula includes: Among them, is the prior probability, is the eigenvalue of the visual feature, is the weight of the visual feature, is the total amount of visual features; is the posterior probability, is the deterioration probability matrix of visual features, is the transition probability matrix of hidden features; is the eigenvalue of the hidden feature, is the weight of the hidden feature, is the total amount of hidden features, is the deterioration probability of visual features, is the transition probability of hidden features.

3. The self-adaptive repair method for distribution network line loss according to claim 2, wherein The second screening based on the probability distribution network includes: Obtaining the probability distribution network, where each network node includes a prior probability and a posterior probability; Traversing the probability distribution network, and performing a threshold overrun determination based on the prior probability threshold and the posterior probability threshold; If the prior probability meets the prior probability threshold and the posterior probability meets the posterior probability threshold, adding the network node to the second damage assessment data; If either the prior probability or the posterior probability does not meet the probability threshold, marking the network node as a pseudo-anomaly point.

4. The adaptive repair method for distribution network line loss according to claim 1, characterized in that, The performing of regional segmentation reconstruction and self-healing correction analysis includes: Based on the network coupling characteristics, traversing the single column of line loss location to determine multiple damaged distribution transformer areas, where the coupling degree within the distribution transformer area is higher than the coupling threshold, and the coupling degree between the distribution transformer areas is lower than the coupling threshold; Traversing the multiple damaged distribution transformer areas, performing regional network reconstruction to determine multiple distribution transformer data, where the regional network reconstruction analysis method is multi-objective optimization; Performing collision avoidance analysis on the multiple distribution transformer data and splicing to determine the network configuration.

5. The self-adaptive repair method for distribution network line loss according to claim 4, characterized in that, Based on the correction requirements, adjusting the regional scale of the multiple damaged distribution transformer areas, where the scale adjustment methods include regional expansion and regional contraction.

6. The self-adaptive repair method for distribution network line loss according to claim 1, characterized in that The generating of the anomaly repair instruction includes: For the non-self-healing line loss location data, tracing the cause of the line loss to determine a set of repair plans; Based on the line loss level and network impact, setting a repair time limit based on the repair priority, where the repair priority is positively correlated with the line loss level and network impact; Map the repair plan set and the repair time limit to generate an abnormal repair instruction.

7. An adaptive repair system for distribution network line loss, characterized in that The steps for implementing the method for adaptively repairing the line loss of a distribution network according to any one of claims 1 to 6 include: Three - element partition feature unit: Read the line topology network of the target distribution network and determine the three - element partition features, where the three - element partition features include voltage - divided line loss, area - divided line loss, and component - divided line loss; Abnormal location module building unit: Based on the line topology network and the three - element partition features, build an abnormal location module. The abnormal location module includes parallel three - element decision branches, and the abnormal location module has a built - in line loss probability calculation formula; Line loss location single - column determination unit: Transmit the distribution network data back, combine with the abnormal location module, make a damage determination decision and probability calculation, and determine the line loss location single - column; Reconstruction data determination unit: Build a network reconstruction model based on the network coupling characteristics, take the regional network reconstruction as the line loss repair principle, traverse the line loss location single - column, conduct regional segmentation reconstruction and self - healing correction analysis, and determine the reconstruction data, where the reconstruction data includes network configuration; Abnormal repair instruction generation unit: Traverse the line loss location single - column, read the non - self - healing line loss location data, and generate an abnormal repair instruction; Line loss repair management unit: Combine the reconstruction data and the abnormal repair instruction to manage the line loss repair of the target distribution network.

Images