Low-voltage distribution network power supply line leakage fault positioning method, device and storage medium

Abstract

The application discloses a low-voltage distribution network power supply line electric leakage fault positioning method and device and a storage medium, and relates to the technical field of power grid power supply safety. The positioning method comprises the following steps: acquiring historical electrical quantity data of a target station area in a normal operation state, and then calculating equivalent mutual impedances between user nodes and determining an abnormal judgment threshold; acquiring measured electrical quantity data of the target station area when an electric leakage fault occurs, and then calculating current equivalent mutual impedances between the user nodes; comparing the current equivalent mutual impedances with corresponding abnormal judgment thresholds to identify abnormal mutual impedances; for each user node, counting the number of the current equivalent mutual impedances between the user node and all other user nodes that are identified as abnormal mutual impedances; and determining the user node with the largest number of abnormal mutual impedances as a priority investigation object of the electric leakage fault. The application solves the problem of large calculation error of the virtual impedance method depending on strong assumptions, and improves the fault investigation efficiency.

Description

Technical Field

[0001] This invention belongs to the field of power grid safety technology, and particularly relates to a method, equipment and storage medium for locating leakage faults in low-voltage power distribution lines. Background Technology

[0002] Low-voltage power distribution networks directly serve users' daily lives and production areas. Their electrical equipment and power lines are widely distributed and have numerous nodes. While providing power security for various users, they also present significant safety hazards due to issues such as improper equipment installation, complex operating environments, and insufficient electrical safety awareness among personnel in some locations. In particular, grounding faults (leakage faults) in power lines occur frequently and are characterized by their concealment and intermittent nature, making manual on-site troubleshooting extremely difficult. These faults not only significantly increase the risk of electric shock and electrical fires but also lead to increased line losses in distribution substations, directly impacting the economic efficiency and reliability of the power grid. Therefore, achieving rapid and accurate location of line leakage faults is a key technical requirement for improving user electrical safety and reducing line losses in low-voltage distribution substations.

[0003] Leakage faults alter the electrical connections between users in low-voltage distribution networks, thus affecting line parameters between them. Currently, some studies have attempted to locate faults by analyzing changes in electrical quantities on the user side. For example, based on the stability assumption of user circuit impedance, they calculate the virtual circuit impedance of users and capture its characteristic abrupt changes to achieve anomaly detection and location. However, such methods have significant limitations in practical applications:

[0004] On the one hand, it is difficult to accurately assess the actual distribution of trunk current in power supply lines, and usually requires strong assumptions for approximation, resulting in large errors in the calculation results; on the other hand, for persistent or intermittent, slowly changing leakage faults, traditional methods have limited ability to capture fault characteristics, are not adaptable enough, and are difficult to achieve long-term effective fault monitoring and location.

[0005] Therefore, how to fully utilize existing metering data from distribution areas and users (such as voltage and current information collected by smart meters) and break through the limitations of traditional methods to develop a technical solution that can accurately and quickly locate leakage faults in low-voltage distribution networks has become an important technical problem that urgently needs to be solved in the field of power system operation and maintenance. Summary of the Invention

[0006] To address the aforementioned deficiencies in the existing technology, the present invention aims to provide a method, device, and storage medium for locating leakage faults in low-voltage power distribution lines, thereby solving the problems of low efficiency in manual inspection, large calculation errors in virtual impedance methods that rely on strong assumptions, and insufficient detection capability for continuous and intermittent leakage faults in low-voltage power distribution network leakage fault location scenarios.

[0007] This invention solves the above-mentioned technical problems through the following technical solution: a method for locating leakage faults in low-voltage power distribution lines, comprising:

[0008] Acquire historical electrical quantity data of the target transformer area under normal operating conditions;

[0009] Based on the historical electrical quantity data, the equivalent mutual impedance between each user node is calculated, and based on the calculation results of multiple normal time periods, the abnormal judgment threshold of the equivalent mutual impedance between each user node is determined.

[0010] Acquire measured electrical quantity data of the target distribution area when a leakage fault occurs, and calculate the current equivalent mutual impedance between each user node based on the measured electrical quantity data;

[0011] Each of the current equivalent mutual impedances is compared with its corresponding anomaly detection threshold to identify abnormal mutual impedances;

[0012] For each user node, count the number of abnormal mutual impedances identified in the current equivalent mutual impedance between it and all other user nodes.

[0013] The user node with the highest number of abnormal mutual impedances is identified as the priority target for investigating leakage faults.

[0014] This invention directly calculates or fits the equivalent mutual impedance between nodes based on collected electrical quantity data (voltage and load current data of each user node). This method does not require prior knowledge or assumptions about the actual current distribution on the distribution trunk line, thus fundamentally avoiding systematic modeling errors introduced by inaccurate assumptions about current distribution, and improving the objectivity and reliability of parameter calculation.

[0015] The reliability of this invention does not depend on the accuracy of the absolute value of the equivalent mutual impedance, but rather on the relative change of the same mutual impedance under normal and fault conditions, and the relative comparison of the number of abnormal mutual impedances between different user nodes. This approach, based on relative analysis of actual operating data, reduces the dependence on the absolute accuracy of the calculation model and improves the robustness and adaptability of the method under complex actual operating conditions.

[0016] This invention analyzes historical electrical quantity data from multiple normal periods to statistically determine anomaly thresholds for each mutual impedance. These thresholds accurately reflect the natural fluctuation range of the system under fault-free operation. Therefore, it can effectively distinguish between significant parameter deviations caused by leakage faults and inherent random fluctuations during normal system operation. Any change in mutual impedance caused by a fault exceeding its historical normal fluctuation range can be identified as an anomaly. Thus, regardless of whether a leakage fault is continuous, intermittent, or slowly developing, as long as it has a sufficient impact on line parameters during its occurrence period, it can be effectively detected by the method of this invention. The detection capability of this invention is not limited by the duration or form of the fault and has good identification effects on concealed and intermittent faults.

[0017] This invention automatically processes meter data in a distribution area using an algorithm, outputting a clear "priority inspection target," namely the user node with the highest number of abnormal mutual impedances. This completely changes the outdated "blanket" inspection model that relies on manual experience, precisely narrowing the fault location scope from the entire distribution area covering dozens to hundreds of users to a specific user node and its related power supply lines. This significantly improves fault diagnosis efficiency, substantially reduces the cost and time consumption of manual inspections, and provides clear and operable technical guidance for on-site operation and maintenance.

[0018] Furthermore, both the historical electrical quantity data and the measured electrical quantity data include the voltage phasor and load current phasor of each user node.

[0019] This invention analyzes the voltage phasor and load current phasor of each user node. The data source is standardized and easy to obtain. There is no need to add a dedicated sensor. It directly uses the complete electrical phasor information collected by existing smart meters, which reduces the implementation cost and ensures the consistency and reliability of the data. This lays a real and unified data foundation for the subsequent accurate calculation of equivalent mutual impedance.

[0020] Furthermore, the equivalent mutual impedance between each user node is calculated, specifically including:

[0021] Based on Kirchhoff's laws, a system of multiple linear equations is established with the voltage difference between user nodes as the dependent variable and the load current of each user as the independent variable.

[0022] Based on the historical electrical quantity data or the measured electrical quantity data, the least squares method is used to solve the multivariate linear equation system to obtain the equivalent mutual impedance between each user node.

[0023] The method of this invention constructs a system of linear equations describing the network topology using Kirchhoff's laws and solves it using the least squares method. It can robustly fit the equivalent mutual impedance between user nodes based on noisy measurement data in actual operation, without requiring known precise network topology or complex high-order modeling. While ensuring computational accuracy, it significantly reduces the model's dependence on data quality and prior system knowledge, thereby improving the applicability and reliability of the method in actual power distribution networks.

[0024] Furthermore, when using the least squares method to solve the problem, the objective function is to minimize the root mean square error of the difference between the measured and calculated voltage difference.

[0025] This invention aims to minimize the root mean square error of the difference between the measured and calculated voltage difference, effectively balancing the influence of errors at each measurement point and avoiding excessive interference from individual abnormal data on the overall fitting results. This improves the overall accuracy and robustness of the equivalent mutual impedance solution, making the obtained parameters more accurately reflect the average state of the distribution network operation and laying a reliable foundation for the accurate extraction of subsequent fault characteristics.

[0026] Furthermore, the determination of the anomaly threshold for the equivalent mutual impedance between user nodes specifically includes:

[0027] The equivalent mutual impedance between each user node was calculated in several normal time periods.

[0028] For each specific pair of user nodes, the maximum value of their equivalent mutual impedance over all normal time periods is taken as the threshold for determining the abnormality of the equivalent mutual impedance between that specific pair of user nodes.

[0029] The method of this invention uses the historical maximum values ​​of each mutual impedance under multiple normal time periods as the anomaly judgment threshold, which fully considers the upper limit of natural fluctuations during normal system operation. It can effectively distinguish between significant parameter deviations caused by faults and normal random fluctuations, and minimizes the false alarm rate while ensuring fault sensitivity, thereby improving the adaptability and scenario adaptability of fault judgment criteria.

[0030] Furthermore, the calculation of the equivalent mutual impedance between each user node during multiple normal time periods specifically includes:

[0031] A sliding time window method is used to sequentially extract multiple normal time periods of equal length from the historical electrical quantity data; wherein adjacent normal time periods partially overlap or are connected end to end on the time axis;

[0032] Calculate the equivalent mutual impedance between each user node within the normal time period corresponding to each sliding time window.

[0033] The method of this invention continuously captures data during normal periods through a sliding time window, which can dynamically and densely cover historical operating states, thereby capturing the natural fluctuation characteristics of the system under fault-free conditions more comprehensively and meticulously. This provides a sufficient data basis for setting adaptive and highly reliable anomaly judgment thresholds, and enhances the adaptability of the fault diagnosis system to gradual changes in the operating environment and load pattern changes.

[0034] Furthermore, the identification of abnormal mutual impedance specifically means: if the current equivalent mutual impedance is greater than its corresponding abnormal determination threshold, then the current equivalent mutual impedance is determined to be an abnormal mutual impedance.

[0035] The judgment rules of this invention are simple and clear, and the computational efficiency is high. It can directly achieve rapid and batch identification of abnormal states based on preset thresholds, avoiding the computational overhead required for complex pattern recognition or dynamic threshold adjustment. It is conducive to deployment in embedded devices or real-time systems, ensuring the timeliness and feasibility of the fault location process.

[0036] Furthermore, the method of counting the number of abnormal mutual impedances identified among the current equivalent mutual impedances between the node and all other user nodes specifically includes:

[0037] For each user node, a set of mutual impedances is constructed, which contains the current equivalent mutual impedances between that user node and all other user nodes in the transformer area.

[0038] The number of abnormal mutual impedances identified in the set of mutual impedances is counted and taken as the number of abnormal mutual impedances of the user node.

[0039] The method of this invention transforms the complex network-wide topology correlation analysis into the calculation of quantitative indicators for a single node by constructing and counting the number of anomalies in the set of associated mutual impedances for each user node. This not only significantly reduces the analysis dimensions and computational complexity, but also makes the "clustering effect" of fault impacts more intuitive, providing a clear and comparable basis for accurately identifying the user nodes most significantly affected by faults.

[0040] Based on the same concept, the present invention also provides an electronic device, including a memory, a processor, and a computer program or instructions stored in the memory, wherein the processor executes the computer program or instructions to implement the low-voltage power distribution line leakage fault location method as described above.

[0041] Based on the same concept, the present invention also provides a computer-readable storage medium having a computer program or instructions stored thereon, which, when executed by a processor, implements the method for locating leakage faults in low-voltage power distribution lines as described above.

[0042] Compared with existing technologies, this invention, based on existing smart meter data, calculates and compares the relative changes in equivalent mutual impedance between user nodes under normal and fault conditions, achieving the following main beneficial effects:

[0043] It automatically outputs a unique priority target for troubleshooting, accurately narrowing the scope of the fault from the entire distribution area to the relevant lines of a single user node, which greatly improves troubleshooting efficiency and reduces labor costs.

[0044] It does not rely on trunk current assumptions and absolute impedance accuracy. It compares relative changes with historical fluctuation thresholds, is insensitive to data noise and model errors, and is suitable for various leakage faults such as continuous and intermittent ones.

[0045] It fully utilizes existing smart meter data, without the need to modify lines or add dedicated monitoring equipment, making it easy to deploy and promote on a large scale.

[0046] Thresholds are generated based on historical data statistics, avoiding human experience-based settings, making fault identification more objective, and can be adaptively updated according to the system's operating status, reducing false alarms and missed alarms. Attached Figure Description

[0047] To more clearly illustrate the technical solution of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only one embodiment of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 This is a flowchart of the method for locating leakage faults in low-voltage power distribution lines in an embodiment of the present invention;

[0049] Figure 2 This is a diagram showing the calculation results of the number of abnormal mutual impedances of each user node under leakage conditions in this embodiment of the invention. Detailed Implementation

[0050] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present invention, and 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.

[0051] The technical solution of the present invention will be described in detail below with reference to specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0052] Example 1

[0053] like Figure 1 As shown in the figure, the method for locating leakage faults in low-voltage power distribution lines provided by this invention includes the following steps:

[0054] Step S1: Obtain historical electrical quantity data of the target transformer area under normal operating conditions.

[0055] To establish a benchmark for locating leakage faults, it is first necessary to systematically collect and construct historical electrical quantity data for the target distribution substation under fault-free normal operation. Historical electrical quantity data refers to the time-series electrical quantity data automatically collected and recorded by the user-side metering device over a continuous period during which no leakage or other grounding faults have been confirmed to have occurred. The specific implementation method is as follows:

[0056] Historical electrical quantity data mainly comes from the existing metering and monitoring system in the target area, including but not limited to: collecting voltage phasors (usually phase voltage amplitude and phase) and load current phasors at each user access point.

[0057] To ensure the representativeness and validity of historical electrical quantity data, the data collection process must meet the following conditions:

[0058] During the selected historical period, the residual current monitoring value of the transformer area should be consistently below the safety threshold (e.g., less than 500mA) and there should be no leakage alarm records to ensure that the data represents a normal state. The data sampling interval should meet the needs of dynamic process capture, preferably at the second or minute level (e.g., 15 minutes / time) to obtain electrical quantity change sequences with sufficient time resolution. Data should be collected for multiple consecutive cycles (e.g., 7-30 consecutive days) to cover different date types (weekdays, weekends) and typical load change patterns to ensure the comprehensiveness of the dataset. Voltage and current data should be recorded in phasor form (i.e., including amplitude and phase information) to support subsequent accurate calculations based on circuit principles.

[0059] The collected raw data needs to be preprocessed to form high-quality data that can be directly used for calculation. The preprocessing in this embodiment includes data cleaning, that is, automatically or manually removing abnormal records that are missing due to communication interruption, malfunction of metering device, obvious exceedance or invalidity.

[0060] The preprocessed time-series data is organized into structured samples in chronological order. Each sample point corresponds to a sampling time t, containing the voltage phasor set of all N users at that sampling time. and current phasor set ,in, This represents the voltage phasor (measured value) of the i-th user at sampling time t. This represents the load current phasor (measured value) of the i-th user at sampling time t. Complete historical electrical quantity data consists of M time-series consecutive sample points.

[0061] To verify the effectiveness of the method of the present invention and optimize the parameters under controlled conditions, data can be obtained through simulation or experimental platforms:

[0062] In a power distribution network physical simulation experimental platform or high-fidelity digital simulation model, construct a transformer substation topology containing a typical number of users (e.g., 139 households). Set the substations to a fault-free, normal operating state in the simulation environment, and run them continuously for a period of time (e.g., 10 days) at the same sampling frequency as the actual system (e.g., 15 minutes / time), collecting time-series data on voltage and load current from all simulated user meters. This process generates clean, noise-free historical electrical quantity data (i.e., a baseline dataset) with known topology.

[0063] In actual engineering deployment, step S1 automatically extracts historical electrical quantity data of the target transformer area for a specified period through standard data interfaces (such as electricity information collection systems and distribution automation systems) and performs preprocessing.

[0064] Step S1 provides high-quality, time-consistent, and physically meaningful historical electrical quantity data for normal conditions, providing a solid data foundation for establishing an accurate "normal condition baseline" in subsequent steps.

[0065] Step S2: Based on historical electrical quantity data, calculate the equivalent mutual impedance between each user node, and determine the abnormal judgment threshold of the equivalent mutual impedance between each user node based on the calculation results of multiple normal time periods.

[0066] Step S2 is the core of establishing a leakage current fault diagnosis benchmark. It aims to extract parameters describing the electrical coupling relationship between users from historical electrical quantity data during normal operation and quantify their normal fluctuation range. The specific implementation method is as follows:

[0067] S2.1: Calculation principle and model establishment of equivalent mutual impedance.

[0068] Equivalent mutual impedance is an electrical parameter that describes the degree of mutual influence between two user nodes due to common paths or electromagnetic coupling. A network model is established based on Kirchhoff's Voltage Law (KVL) and the superposition principle.

[0069] For any two user nodes i and j (i≠j) within the transformer area, at a certain time t, their voltage difference This can be expressed as the sum of the voltage drops generated by all user load currents in this circuit. Based on this, the following system of multivariate linear equations can be established:

[0070] (1)

[0071] Where N represents the number of user nodes within the transformer area; This represents the equivalent mutual impedance to be solved, and its physical meaning is: the load current of user node p. The voltage difference between user node i and user node j The impedance component contributed by the above; This represents the modeling error and measurement noise. For p = i or j, This characterizes the main impact of user node p on its own voltage difference; for p ≠ i or j, It characterizes the mutual influence of other user load currents through network coupling.

[0072] Substituting all user voltage difference and load current data over M consecutive sampling times (e.g., a time period containing 160 sample points) into equation (1), an overdetermined set of equations can be constructed:

[0073] (2)

[0074] Therefore, we can conclude that: ,in, It is an M×(N(N-1) / 2) dimensional voltage difference matrix, where each column corresponds to a voltage difference sequence of a pair of user nodes i and j at M sampling times; It is an M×N dimensional current matrix, with each column corresponding to the load current sequence of a user node; It is the N×(N(N-1) / 2) dimensional equivalent mutual impedance matrix to be determined; This is the error matrix.

[0075] S2.2: Parameter solution based on least squares method.

[0076] To robustly estimate parameters from noisy measurement data The solution is obtained using the least squares method:

[0077] Solution objective: Find a set of The estimated value This minimizes the sum of squared deviations between the measured and calculated voltage differences at all sampling times and at all user nodes.

[0078] Objective function: Minimize the sum of squared residuals (or root mean square error RMSE):

[0079] (3)

[0080] in, This represents the Frobenius norm of the matrix (the square root of the sum of the squares of all its elements).

[0081] By using the normal equation or QR decomposition, the least squares problem in formula (3) is solved to obtain the estimated value. Estimated value Each element in This refers to the estimated equivalent mutual impedance value within the selected time period.

[0082] S2.3: Multi-time period calculation and determination of anomaly judgment threshold.

[0083] To obtain a robust baseline for normal operation, the above calculations need to be repeated over multiple normal periods, and the threshold for anomaly detection needs to be statistically determined.

[0084] A sliding time window method is used to sequentially extract Q (e.g., Q=6) equal-length normal time periods from the complete historical electrical quantity data. The length M of each normal time period should be sufficient to ensure the statistical significance of the parameter estimation (e.g., containing 160 sample points). Adjacent normal time periods can partially overlap or be joined end-to-end on the time axis.

[0085] For the q-th (q=1,2,...,Q) normal time period, perform steps S2.1 and S2.2 to calculate the estimated value of the equivalent mutual impedance matrix for that normal time period. .

[0086] For each specific pair of user node relationships (i.e., for each fixed combination of i, j, p indices), extract its equivalent mutual impedance calculation values ​​for all Q normal time periods. , , ..., The maximum value among the calculated equivalent mutual impedances in this set is taken as the specific mutual impedance. Anomaly detection threshold .Right now:

[0087] (4)

[0088] All Organized according to the same structure, forming a similar Dimensionally consistent anomaly detection threshold matrix Anomaly detection threshold This represents the specific mutual impedance under all historically observed normal operating conditions. The upper limit value that appears. Any new measured or calculated value exceeding this threshold indicates that the mutual impedance relationship has exceeded the historical normal fluctuation range, possibly caused by a fault.

[0089] Through step S2, the system not only learns the coupling relationship model between users under normal conditions ( The average trend), and more importantly, the range of natural random fluctuations of these relationships in a fault-free environment (anomaly detection threshold matrix). This provides an objective and adaptive judgment standard for accurately and reliably identifying significant anomalies caused by leakage faults in subsequent steps.

[0090] Step S3: Obtain the measured electrical quantity data of the target transformer area when the leakage fault occurs, and calculate the current equivalent mutual impedance between each user node based on the measured electrical quantity data.

[0091] Step S3 is the core of online diagnostics. Its goal is to collect data in real time and calculate the current system status parameters after detecting a possible leakage fault, so as to compare them with a normal baseline. The specific implementation is as follows:

[0092] S3.1: Leakage fault triggering and data acquisition.

[0093] When a leakage fault occurs, step S3 can be triggered and executed by at least one of the following methods:

[0094] Residual current operated protector alarm: Triggered when the total residual current monitoring value of the transformer area continuously exceeds the preset safety threshold.

[0095] Zero-sequence current / voltage over-limit: Triggered when a sustained abnormal increase in zero-sequence current or zero-sequence voltage is detected.

[0096] Planned diagnostic triggers: Diagnoses are performed periodically according to the operation and maintenance plan or manually triggered.

[0097] Once triggered, the system immediately acquires, in real-time or near real-time, the measured electrical quantity data of the target transformer area during the current leakage (or suspected leakage) period through the electricity information acquisition system or distribution automation system. The content and format of the measured electrical quantity data should be consistent with the historical electrical quantity data defined in step S1, specifically including:

[0098] Voltage phasors of each user node ( ), and the load current phasor of each user node. .

[0099] Collect synchronous data for a continuous time period. The length M' of this time period should be the same as or comparable to the length M of the historical normal time period used for calculation in step S2 (e.g., both containing 160 consecutive sampling points) to ensure consistency in the calculation. This time period should be close to the time of occurrence or cover the period when the fault characteristics are obvious.

[0100] S3.2: Calculation of the current equivalent mutual impedance.

[0101] The equivalent mutual impedance (i.e., the current equivalent mutual impedance) between each user node during the current fault period is calculated. The principle is exactly the same as the calculation model established in step S2.1, but the fault period data collected in real time is used.

[0102] Utilizing the fault time periods t1, t2, ..., t M′ The collected voltage and load current data are used to construct the same system of multivariate linear equations or matrix equations as in step S2.1:

[0103] (5)

[0104] The matrix form is:

[0105] (6)

[0106] in, and It consists of a voltage difference matrix and a load current matrix composed of data from the fault period. It is the current equivalent mutual impedance matrix to be determined, which reflects the coupling relationship between user nodes under the current (potentially faulty) system state.

[0107] The above equations are solved using the exact same parameter estimation algorithm as in step S2.2 (i.e., the least squares method, with the objective of minimizing the sum of squared residuals). That is:

[0108] (7)

[0109] The solution obtained This is the estimated value of the equivalent mutual impedance matrix calculated based on the data during the current fault period.

[0110] The output of step S3 is the current equivalent mutual impedance matrix. Each element of the matrix This represents the estimated mutual impedance value under the current operating conditions. Current equivalent mutual impedance matrix. This will be used as input and sent to the subsequent step S4, along with the anomaly detection threshold matrix determined in step S2. Perform element-by-element comparison.

[0111] In step S3, the system quantifies the current (potentially abnormal) system state into a set of comparable parameters. This provides a direct data basis for the next step of identifying which mutual impedance relationships have undergone abnormal changes beyond the normal range.

[0112] Step S4: Compare each current equivalent mutual impedance with the corresponding anomaly detection threshold to identify abnormal mutual impedances.

[0113] Step S4 is a crucial step in fault feature extraction. Its purpose is to map the numerous mutual impedance values ​​calculated under the current operating state into simple "normal" or "abnormal" binary state labels through element-by-element threshold comparison, thereby highlighting the electrical coupling paths most affected by leakage faults. The specific implementation method is as follows:

[0114] The current equivalent mutual impedance matrix calculated in step S3 Each element in (where i,j represent user node pairs (i≠j), and p represents the voltage difference) Contributing user load current index (p=1,2,...,N), and the anomaly detection threshold matrix determined in step S2. Threshold elements at exactly the same index position Compare them.

[0115] Each mutual impedance value Each has one and only one exclusive, predefined threshold. Correspondingly, this strict correspondence is based on historical normal data statistics, ensuring the relevance and fairness of the comparison.

[0116] For each pair of indices (i,j,p), perform the following judgment:

[0117] if > Then determine the current equivalent mutual impedance. It is an abnormal mutual impedance; otherwise (i.e.) ≤ Determine the current equivalent mutual impedance. This is normal mutual impedance.

[0118] This anomaly detection rule is based on the following assumption: Under normal operating conditions, each cross impedance value will reach its historical maximum value (i.e., the threshold). Fluctuations within a certain range. Leakage faults alter local network parameters, causing a significant increase in certain mutual impedance values, exceeding the upper limit of their historical normal fluctuation range. This "over-limit" phenomenon is marked as "abnormal".

[0119] To facilitate subsequent statistics, the system generates a [database name]. and A binary state matrix (or anomaly marker matrix) with identical dimensions. .matrix elements in Assign values ​​according to the following rules:

[0120] (8)

[0121] matrix It visually depicts the abnormal distribution of mutual impedance relationships among all user nodes in the transformer substation network under the current state.

[0122] In step S4, the system completes the transformation from complex continuous parameters to clear anomaly features. The output anomaly label matrix... It directly identifies which specific mutual impedance relationships in the network have undergone abnormal changes beyond the normal range, providing the most direct clues for finally locating the leakage fault point.

[0123] Step S5: For each user node, count the number of abnormal mutual impedances identified in the current equivalent mutual impedance between it and all other user nodes.

[0124] Step S5 aims to perform aggregate analysis on the anomaly distribution generated in step S4, summarizing and quantifying the anomaly clues scattered throughout the network by user node, thereby generating a quantitative indicator that directly reflects the degree of impact of leakage faults on each user. The specific implementation method is as follows:

[0125] For each user node k (k=1,2,...,N), count the total number of abnormal mutual impedances associated with user node k. Abnormal mutual impedances associated with user node k refer to the total number of abnormal mutual impedances in the mutual impedance matrix. (or anomaly marker matrix) In the diagram, the mutual impedance relationship represented by the element indices i, j, p that is at least equal to k. In a preferred embodiment of the invention, to focus on anomalies characterizing the "connectivity" of user node k itself, statistics typically focus on the following two more direct types of correlations:

[0126] Voltage difference loop with k as one end: i.e., all loops of the form or The mutual impedances (i≠k, j≠k) describe the coupling relationship on the voltage difference loop between user node k and other user nodes.

[0127] Other user circuits affected by the k load current: i.e., all circuits of the form The mutual impedances (i≠j, i≠k, j≠k) describe the effect of the load current at user node k on the voltage difference at other user nodes.

[0128] For each user node, an equivalent mutual impedance correlation set can be defined, which contains an anomaly marker matrix. The index of all elements related to this user node.

[0129] The current equivalent mutual impedance matrix and anomaly marker matrix Reorganize by row (or column) so that the k-th row (or column) directly corresponds to the equivalent mutual impedance vector of a user node k. This vector contains all mutual impedance values ​​(or anomaly markers) that satisfy the conditions i=k, j=k, or p=k. In other words, the vector is the data structure representation of the equivalent mutual impedance association set of user node k.

[0130] For each user node k, traverse its equivalent mutual impedance association set, and for each element in the association set, assign an anomaly marker. ,if =1 (indicating an anomaly), then the number of abnormal mutual impedances of user node k will be increased by 1 (the initial value is 0).

[0131] After completing the traversal, the number of abnormal mutual impedances of user node k can be obtained. This number directly quantifies the degree to which user node k is abnormal in the network mutual impedance relationship.

[0132] The statistics on abnormal mutual impedance are based on a core inference: a leakage fault point significantly alters the electrical parameters of nearby lines, making the mutual impedance along current paths flowing through or near the fault point more prone to abnormal changes. Therefore, the number of abnormal mutual impedances in the associated mutual impedances of user nodes closest to the fault point will be significantly higher than that of user nodes farther away from the fault point. A larger value for the number of abnormal mutual impedances indicates a more widespread and severe "ripple effect" or "impact" of the fault on the electrical relationships of that user node.

[0133] Through step S5, the system transforms complex network-level anomaly patterns into a single, comparable quantitative indicator for each user node, thus providing a direct ranking basis for the final step of determining priority investigation targets.

[0134] Step S6: Identify the user node with the highest number of abnormal mutual impedances as the priority target for investigating leakage faults.

[0135] Step S6 is the final decision-making stage of the location method of this invention. Its purpose is to automatically and clearly identify the users most likely to be near the leakage fault point based on the quantitative indicators generated in step S5, and to provide clear and actionable guidance for on-site maintenance personnel. The specific implementation method is as follows:

[0136] Based on the number of abnormal mutual impedances of each user node output in step S5, find the element with the largest value, and the corresponding user node is determined as the priority investigation target.

[0137] Under normal circumstances, there exists a single user node that results in the maximum number of abnormal cross impedances. If multiple user nodes have the same maximum number of abnormal cross impedances, they can be handled according to preset auxiliary rules. For example, maintenance personnel can conduct combined investigations according to network topology (e.g., from the transformer side to the user side) or geographical proximity, or analyze the changing trends of the number of abnormal cross impedances of these user nodes in a recent historical period, and prioritize the user node with the fastest rising trend.

[0138] Through step S6, the method of the present invention completes a closed loop from data acquisition, feature calculation, anomaly identification to final decision-making. The results directly serve on-site operations and maintenance, realizing the transformation from "data-driven analysis" to "action-driven guidance," effectively solving the core problems of low efficiency and wide scope of manual investigation in the background technology.

[0139] Example 2

[0140] To verify the effectiveness of the method of the present invention, a simulated transformer area containing 139 simulated users is used as an example. Based on steps S1 and S2, historical electrical quantity data of the simulated transformer area under normal conditions for 10 consecutive days are obtained. Using a sliding window method, the historical electrical quantity data is divided into 6 equal-length normal time period subsets, each containing 160 sample points. For each normal time period, the equivalent mutual impedance matrix between each user node is calculated by constructing a system of multivariate linear equations and applying the least squares method. For each equivalent mutual impedance parameter between each pair of user nodes, the maximum value among the calculated values ​​of all 6 normal time periods is taken as the anomaly judgment threshold for that equivalent mutual impedance parameter, thus forming a complete threshold benchmark.

[0141] When a leakage fault is simulated and injected into the branch power circuit to which user node 68 belongs, the system enters an abnormal leakage state. According to steps S3 to S5, the measured electrical quantity data under the leakage state (e.g., lasting for 2 days) is obtained, the current equivalent mutual impedance is calculated, and it is compared element by element with the aforementioned abnormal judgment threshold to identify abnormal mutual impedance. Finally, the number of abnormal mutual impedances for each user node is counted.

[0142] The number of abnormal cross impedances of user nodes obtained from statistics is visualized, and the results are as follows: Figure 2 As shown. From Figure 2 It can be clearly observed that the number of abnormal mutual impedances of user node 68 shows a very significant peak, which is much higher than that of all other user nodes in the area.

[0143] This phenomenon indicates that, for user node 68, the equivalent mutual impedance relationship between it and almost all other user nodes in the distribution area exceeds the historical normal fluctuation range and is therefore judged as abnormal. This directly verifies the core inference of the present invention: the leakage fault point will "contaminate" the electrical parameters of the line segment where it is located, causing a systematic and large-scale abnormal shift in the coupling relationship between the user node (user node 68 in this embodiment) located on that line segment and other user nodes in the entire network. Therefore, user node 68 is determined to be the location closest to the leakage fault point.

[0144] exist Figure 2 In the process, it can be observed that the number of abnormal mutual impedances of some users near user node 68 (such as user nodes 67, 69, etc.) also increased to a certain extent, but was significantly lower than that of user node 68. As the electrical or topological distance from user node 68 increases, the number of abnormalities of other user nodes rapidly decreases to near the baseline level. This intuitively reflects the physical law that the influence intensity of leakage faults decreases with increasing electrical distance, further confirming the rationality of the identification results of the method of this invention.

[0145] Based on the decision rules in step S6, the system automatically identifies the user node with the highest number of abnormal mutual impedances—user node 68—as the priority target for investigating leakage faults. This location result instantly and precisely narrows the scope of the maintenance personnel's investigation from a large distribution area containing 139 users to the single branch power circuit corresponding to user node 68 and its adjacent lines. On-site personnel can then directly conduct focused inspections of user node 68's access point, its branch box, and upstream power supply lines, thereby quickly identifying and isolating the leakage fault point.

[0146] Example 3

[0147] This invention also provides an electronic device, which includes a memory, a processor, and a computer program or instructions stored in the memory. The processor executes the computer program or instructions to implement the low-voltage power distribution line leakage fault location method in this invention.

[0148] Although not shown, the electronic device includes a processor that can perform various appropriate operations and processes based on programs and / or data stored in read-only memory (ROM) or loaded from a storage portion into random access memory (RAM). The processor can be a multi-core processor or may contain multiple processors. In some embodiments, the processor may include a general-purpose main processor and one or more specialized coprocessors, such as a central processing unit, graphics processing unit (GPU), neural network processor (NPU), digital signal processor (DSP), etc. Various programs and data required for device operation are also stored in RAM. The processor, ROM, and RAM are interconnected via a bus. Input / output (I / O) interfaces are also connected to the bus.

[0149] The processor and memory described above are used together to execute programs / instructions stored in the memory. When the program / instructions are executed by the computer, they can implement the methods, steps, or functions described in the above embodiments.

[0150] Although not shown, embodiments of the present invention also provide a computer-readable storage medium having a computer program or instructions stored thereon, which, when executed by a processor, implements the method for locating leakage faults in low-voltage power distribution lines according to embodiments of the present invention.

[0151] Readable storage media include both permanent and non-permanent, removable and non-removable media that can store information by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0152] The above description only discloses specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any changes or modifications that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for locating leakage faults in low-voltage power distribution lines, characterized in that, The positioning method includes: Acquire historical electrical quantity data of the target transformer area under normal operating conditions; Based on the historical electrical quantity data, the equivalent mutual impedance between each user node is calculated, and based on the calculation results of multiple normal time periods, the abnormal judgment threshold of the equivalent mutual impedance between each user node is determined. Acquire measured electrical quantity data of the target distribution area when a leakage fault occurs, and calculate the current equivalent mutual impedance between each user node based on the measured electrical quantity data; Each of the current equivalent mutual impedances is compared with its corresponding anomaly detection threshold to identify abnormal mutual impedances; For each user node, count the number of abnormal mutual impedances identified in the current equivalent mutual impedance between it and all other user nodes. The user node with the highest number of abnormal mutual impedances is identified as the priority target for investigating leakage faults. The calculation of the equivalent mutual impedance between user nodes specifically includes: Based on Kirchhoff's laws, a system of multiple linear equations is established with the voltage difference between user nodes as the dependent variable and the load current of each user as the independent variable. Based on the historical electrical quantity data or the measured electrical quantity data, the least squares method is used, with the root mean square error of the difference between the measured and calculated voltage difference as the objective function, to solve the multivariate linear equation system and obtain the equivalent mutual impedance between each user node.

2. The method for locating leakage faults in low-voltage power distribution lines according to claim 1, characterized in that, Both the historical electrical quantity data and the measured electrical quantity data include the voltage phasor and load current phasor of each user node.

3. The method for locating leakage faults in low-voltage power distribution lines according to claim 1, characterized in that, The determination of the anomaly threshold for the equivalent mutual impedance between user nodes specifically includes: The equivalent mutual impedance between each user node was calculated in several normal time periods. For each specific pair of user nodes, the maximum value of their equivalent mutual impedance over all normal time periods is taken as the threshold for determining the abnormality of the equivalent mutual impedance between that specific pair of user nodes.

4. The method for locating leakage faults in low-voltage power distribution lines according to claim 3, characterized in that, The calculation of the equivalent mutual impedance between each user node during multiple normal time periods specifically includes: A sliding time window method is used to sequentially extract multiple normal time periods of equal length from the historical electrical quantity data; wherein adjacent normal time periods partially overlap or are connected end to end on the time axis; Calculate the equivalent mutual impedance between each user node within the normal time period corresponding to each sliding time window.

5. The method for locating leakage faults in low-voltage power distribution lines according to claim 1, characterized in that, The identification of abnormal mutual impedance specifically involves: if the current equivalent mutual impedance is greater than its corresponding abnormal determination threshold, then the current equivalent mutual impedance is determined to be an abnormal mutual impedance.

6. The method for locating leakage faults in low-voltage power distribution lines according to any one of claims 1 to 5, characterized in that, The number of abnormal mutual impedances identified among the current equivalent mutual impedances between the node and all other user nodes specifically includes: For each user node, a set of mutual impedances is constructed, which contains the current equivalent mutual impedances between that user node and all other user nodes in the transformer area. The number of abnormal mutual impedances identified in the set of mutual impedances is counted and taken as the number of abnormal mutual impedances of the user node.

7. An electronic device comprising a memory, a processor, and a computer program or instructions stored in the memory, characterized in that, The processor executes the computer program or instructions to implement the method for locating leakage faults in low-voltage power distribution lines as described in any one of claims 1 to 6.

8. A computer-readable storage medium having a computer program or instructions stored thereon, characterized in that, When the computer program or instructions are executed by the processor, they implement the method for locating leakage faults in low-voltage power distribution lines as described in any one of claims 1 to 6.

Images