Comprehensive management and optimization method and system for low-voltage distribution area

Abstract

The invention discloses a comprehensive management and optimization method and system for low-voltage distribution areas, and the method comprises the steps: collecting communication data of each distribution area, and constructing a first model through inter-network coordination calculation; based on the first model and the communication data, constructing a second model through feature importance analysis; positioning an electricity stealing point by comparing the deviation between the second model and the real-time electricity consumption data; based on the first model and the second model, performing comparative analysis on the line loss data of different time and regions, and positioning a line loss abnormal region; and processing results of electricity stealing point positioning and line loss abnormal area positioning are integrated and displayed. According to the invention, comprehensive management of the low-voltage distribution area from a topological structure to electric energy loss to electricity larceny prevention and the like is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of low - voltage power distribution, and particularly to a comprehensive management and optimization method and system for low - voltage power distribution substations. Background Art

[0002] In modern power systems, the effective management of low - voltage power distribution substations is crucial for ensuring power supply reliability, reducing line losses, and preventing electricity theft. Traditional low - voltage power distribution substation management faces many challenges. For example, the electrical topology structure is not clear, relying on manual investigation and recording, which is not only inefficient but also error - prone, leading to difficulties in maintenance, fault diagnosis, etc. At the same time, the phenomenon of electricity theft in the power distribution network persists, and the problem of line losses is also prominent. Traditional anti - electricity - theft and line - loss analysis methods lack accuracy and real - time performance, making it difficult to accurately locate the electricity - theft positions and effectively reduce line losses.

[0003] Therefore, there is an urgent need for an innovative technology to integrate electrical topology management, anti - electricity - theft, and line - loss analysis management to improve the overall management level of low - voltage power distribution substations.

[0004] I. Electrical Topology Identification and Association Verification Technology for Low - Voltage Power Distribution Substations

[0005] Problems of communication interference and signal attenuation: Some topology identification technologies based on power - line carrier communication are greatly affected by the load of the power distribution substation and have transmission dead zones. When the load of the substation changes significantly or there are high - interference sources, the carrier signal may be severely interfered, resulting in communication interruption or signal attenuation, affecting the accuracy and real - time performance of topology identification.

[0006] High equipment installation and maintenance costs: Some technologies require a large number of intelligent sensors or monitoring devices to be installed to achieve topology identification and association verification. For example, monitoring units are installed at each branch node and user intelligent meter box. This not only increases the equipment cost but also raises the workload and difficulty of installation and maintenance. Moreover, once a device fails, it may affect the topology identification function of the entire system.

[0007] Limited adaptability to complex networks: In the face of complex low - voltage power distribution substation networks, such as those containing multiple distributed power sources, various types of loads, and complex branch lines, some topology identification algorithms may have problems such as large computational amounts, low efficiency, or even inability to accurately identify. For example, when there are a large number of distributed photovoltaic power generation devices in the substation, their access and withdrawal may cause frequent changes in the topology structure, and existing technologies may be difficult to identify and update in a timely and accurate manner.

[0008] II. Precision Location Technology for Anti - Electricity - Theft in Power Distribution Networks

[0009] Insufficient detection ability for new electricity theft means: With the development of technology, electricity theft means have become increasingly diversified and concealed, such as using high-tech electricity theft equipment and attacking smart meters through network. Some traditional anti-electricity-theft technologies may be difficult to effectively detect these new electricity theft behaviors, resulting in the failure to discover electricity theft behaviors in a timely manner.

[0010] Difficulties in data fusion and analysis: Precise anti-electricity-theft positioning requires comprehensive analysis of multiple data sources, such as power metering data, user electricity consumption behavior data, and distribution network topology data. However, there may be differences in data formats, collection frequencies, accuracies, etc. of different data sources, and problems such as data mismatch and information loss may occur during the data fusion process, affecting the accuracy and reliability of anti-electricity-theft analysis.

[0011] False alarm and missed alarm problems: In the actual distribution network environment, there are various interference factors, such as power system fluctuations and sudden changes in load, which may cause false alarms or missed alarms in the anti-electricity-theft system. For example, when a large device in the distribution network suddenly starts or stops, it may cause instantaneous fluctuations in current and voltage, which are misjudged as electricity theft behaviors.

[0012] III. Lean analysis and management technology for distribution network line loss

[0013] Problems with the quality of basic data: Lean analysis and management of line loss rely on accurate basic data, including equipment parameters, line impedance, user electricity consumption, etc. However, in actual applications, these data may have errors, be incomplete, or be updated in a timely manner, resulting in inaccurate line loss calculation results and affecting the analysis and judgment of the reasons for abnormal line loss in the distribution network area.

[0014] Model limitations: Some current line loss analysis models are usually established based on certain assumptions and simplified conditions, and there may be deviations from the actual operation of the distribution network. For example, when considering factors such as load distribution, power factor change, and three-phase imbalance, it is not comprehensive enough, resulting in a large difference between the line loss analysis result and the actual value and being unable to provide accurate decision-making basis for lean management of line loss.

[0015] Lack of real-time performance: The operating state of the distribution network is real-time and dynamically changing, while some line loss analysis systems are unable to obtain and process data in real time, resulting in lag in line loss analysis results. This makes it difficult for management personnel to discover line loss anomalies in a timely manner and take corresponding measures, which is not conducive to the lean management of line loss. Summary of the Invention

[0016] The purpose of this part is to outline some aspects of the embodiments of the present invention and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this part, the abstract of the specification, and the title of the invention to avoid obscuring the purpose of this part, the abstract of the specification, and the title of the invention, and such simplifications or omissions shall not be used to limit the scope of the present invention.

[0017] In view of the above existing problems, the present invention is proposed.

[0018] Therefore, the present invention provides a comprehensive management and optimization method and system for low-voltage distribution transformer areas to solve the problem that existing management technologies often target single problems and lack the integration of electrical topology, anti-theft electricity, and line loss management.

[0019] To solve the above technical problems, the present invention provides the following technical solutions:

[0020] In a first aspect, the present invention provides a comprehensive management and optimization method for low-voltage distribution transformer areas, including:

[0021] Collect communication data of each distribution transformer area, and construct a first model through inter-network coordination calculation;

[0022] Based on the first model and communication data, construct a second model through feature importance analysis;

[0023] Locate the electricity theft points by comparing the deviation between the second model and the real-time electricity consumption data;

[0024] Based on the first model and the second model, conduct a comparative analysis of line loss data at different times and regions, and locate the abnormal line loss regions;

[0025] Integrate and display the processing results of the electricity theft point location and the abnormal line loss region location.

[0026] As a preferred solution of the comprehensive management and optimization method for low-voltage distribution transformer areas according to the present invention, wherein:

[0027] The construction of the first model includes the following steps:

[0028] Collect communication data of the transformer area nodes;

[0029] Based on the collected communication data, calculate the first target parameter;

[0030] According to the first target parameter and combined with inter-network coordination calculation, obtain the connection relationship between the user electricity meter and the transformer;

[0031] Obtain the first model according to the connection relationship.

[0032] As a preferred solution of the comprehensive management and optimization method for low-voltage distribution transformer areas according to the present invention, wherein:

[0033] The inter-network coordination calculation includes the following steps:

[0034] Construct a first data set through the first target parameter and combined with communication data;

[0035] Optimize the communication parameters between nodes using the first algorithm to obtain the optimal communication parameters.

[0036] As a preferred solution of the comprehensive management and optimization method for low-voltage distribution transformer areas described in the present invention, wherein:

[0037] The construction of the second model includes the following steps:

[0038] Perform first preprocessing on the user electricity consumption data to obtain a second data set

[0039] The first preprocessing includes performing first analysis on the electricity consumption data and screening out abnormal data;

[0040] Construct a second model according to the second data set in combination with the second algorithm.

[0041] As a preferred solution of the comprehensive management and optimization method for low-voltage distribution transformer areas described in the present invention, wherein:

[0042] The second algorithm includes the following steps:

[0043] Extract samples of normal electricity consumption and electricity theft behaviors from historical data to construct multiple decision trees;

[0044] Perform feature selection and node splitting based on the Gini coefficient;

[0045] Calculate the feature importance through the splitting times and information gain of the feature in all decision trees;

[0046] For new input data, each decision tree outputs a prediction result;

[0047] Combine the prediction results of the first model and the second algorithm, determine the substation area range of the abnormal data source according to the power grid topology structure, and screen out electricity theft users in the abnormal substation area.

[0048] As a preferred solution of the comprehensive management and optimization method for low-voltage distribution transformer areas described in the present invention, wherein:

[0049] The positioning of the electricity theft point by comparing the deviation between the second model and the real-time monitored electricity consumption data includes the following steps:

[0050] Compare the real-time monitored electricity consumption data with the second model to identify the first abnormal electricity consumption behaviors; the first abnormal electricity consumption behaviors include at least one of the following: voltage anomaly, current anomaly, power anomaly, and electric energy anomaly;

[0051] Combine the prediction results of the first model and the second model to lock and confirm the electricity theft users.

[0052] As a preferred solution of the comprehensive management and optimization method for low-voltage distribution transformer areas in the present invention, wherein:

[0053] Comparatively analyze the line loss data of different times and regions, and locate the regions with abnormal line losses, including the following steps:

[0054] Collect data of sectional metering devices;

[0055] Use the third algorithm in combination with the first model to calculate the line losses of each level of branches;

[0056] Statistically analyze the daily, hourly, and 15-minute line loss rates of the transformer area to identify abnormal line losses;

[0057] Locate the specific positions of abnormal line losses by comparing the actual line losses with the theoretical line losses and combining with the second model.

[0058] In a second aspect, the present invention provides a comprehensive management and optimization system for low-voltage distribution transformer areas, including:

[0059] An electrical topology modeling module, configured to collect communication data of each distribution transformer area, and construct a first model through inter-network coordination calculation;

[0060] An anti-stealing electricity analysis module, configured to construct a second model through feature importance analysis based on the first model and communication data;

[0061] A stealing electricity point positioning module, configured to locate the stealing electricity points by comparing the deviation between the second model and real-time electricity consumption data;

[0062] A line loss anomaly analysis module, configured to comparatively analyze the line loss data of different times and regions based on the first model and the second model, and locate the regions with abnormal line losses;

[0063] A comprehensive result display module, configured to integrally display the processing results of stealing electricity point positioning and line loss anomaly region positioning.

[0064] In a third aspect, the present invention provides a computing device, including:

[0065] A memory, configured to store programs;

[0066] A processor, configured to execute the computer-executable instructions, and when the computer-executable instructions are executed by the processor, the steps of the comprehensive management and optimization method for low-voltage distribution transformer areas are implemented.

[0067] In a fourth aspect, the present invention provides a computer-readable storage medium, including: when the program is executed by the processor, the steps of the comprehensive management and optimization method for low-voltage distribution transformer areas are implemented.

[0068] Advantages of the present invention: The present invention integrates multiple key technologies such as electrical topology identification and correlation verification, precise anti-theft power positioning in the distribution network, and lean analysis and management of line losses, achieving comprehensive management of low-voltage distribution substations in multiple aspects, from topological structure to power loss to anti-theft prevention. It is no longer a technical application with a single function, but forms an organic whole that can comprehensively address various operation and management problems of distribution substations. Traditional technologies often can only solve problems in one aspect and lack systematicness and comprehensiveness. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] In order to more clearly illustrate the technical solutions of 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 based on these drawings. Among them:

[0070] Figure 1 It is a schematic diagram of the basic process of a comprehensive management and optimization method for a low-voltage distribution substation provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0071] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will provide a detailed description of the specific embodiments of the present invention in conjunction with the drawings of the specification. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention.

[0072] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention can also be implemented in other ways different from those described herein. Those skilled in the art can make similar extensions without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0073] Secondly, the so-called "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that can be included in at least one implementation manner of the present invention. The "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it a separate or selectively exclusive embodiment from other embodiments.

[0074] The present invention is described in detail in conjunction with the schematic diagrams. When detailing the embodiments of the present invention, for the convenience of explanation, the cross-sectional views showing the device structure will be enlarged locally out of the general proportion, and the schematic diagrams are only examples and should not limit the scope of protection of the present invention herein. In addition, in actual production, three-dimensional spatial dimensions including length, width, and depth should be included.

[0075] In the description of the present invention, it should be noted that the orientation or positional relationship indicated by terms such as "upper, lower, inner, and outer" is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be construed as a limitation to the present invention. In addition, the terms "first, second, or third" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance.

[0076] Unless otherwise clearly specified and defined in the present invention, the terms "installation, connection, and coupling" shall be understood in a broad sense. For example, it may be a fixed connection, a detachable connection, or an integral connection; it may also be a mechanical connection, an electrical connection, or a direct connection, or may be indirectly connected through an intermediate medium, or may be the communication inside two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0077] Embodiment 1

[0078] Referring to Figure 1 , an embodiment of the present invention provides a comprehensive management and optimization method for low-voltage distribution substations, including:

[0079] S1: Collect communication data of each distribution substation, and construct a first model through inter-network coordination calculation;

[0080] S2: Based on the first model and communication data, construct a second model through feature importance analysis;

[0081] S3: Locate the power theft points by comparing the deviation between the second model and the real-time electricity consumption data;

[0082] S4: Based on the first model and the second model, conduct a comparative analysis of the line loss data at different times and regions, and locate the abnormal line loss regions;

[0083] S5: Integrate and display the processing results of power theft point location and abnormal line loss region location.

[0084] It should be noted that this method integrates multiple key technologies such as electrical topology identification and correlation verification, precise power theft prevention location in the distribution network, and lean line loss analysis management, realizing comprehensive management of low-voltage distribution substations in multiple aspects from topological structure to power loss to power theft prevention. It is no longer a single-functional technology application, but forms an organic whole that can comprehensively handle various operation and management problems of distribution substations, while traditional technologies often can only solve problems in one aspect and lack systematicness and comprehensiveness.

[0085] Embodiment 2

[0086] This is an embodiment of the present invention. Based on the previous embodiment, a comprehensive management and optimization method for low-voltage distribution transformer areas is provided, including:

[0087] In the embodiment of the present application, the communication data collected in step S1 includes collecting the communication data of the area nodes through the HPLC communication module and the power frequency distortion communication technology. Among them, the HPLC communication module is mainly responsible for collecting high-frequency and stable data information, and the power frequency distortion communication technology focuses on obtaining specific topology identification-related data. The two cooperate with each other to provide comprehensive data support for subsequent analysis.

[0088] In the embodiment of the present application, constructing the first model in step S1 includes the following steps:

[0089] Collect the communication data of the area nodes;

[0090] Based on the collected communication data, calculate the first target parameter;

[0091] According to the first target parameter combined with the inter-network coordination calculation, obtain the connection relationship between the user electric meter and the transformer;

[0092] Obtain the first model according to the connection relationship.

[0093] In the embodiment of the present application, the first target parameter includes but is not limited to key parameters such as calculated signal-to-noise ratio, gain ratio, and subcarrier delay ratio.

[0094] In the embodiment of the present application, the signal-to-noise ratio (SNR) is the ratio of the signal power to the noise power, and the calculation formula is:

[0095]

[0096] Where P signal is the signal power; P noise is the noise power.

[0097] In the embodiment of the present application, in HPLC communication, the signal power can be calculated from the signal strength received by the receiving end, and the noise power is obtained by analyzing the signal fluctuations over a period of time to obtain the average power of the background noise. The higher the signal-to-noise ratio, the clearer the signal, the less affected by noise interference, and the better the communication quality; conversely, the lower the signal-to-noise ratio, the easier the signal is to be submerged by noise, resulting in data transmission errors or losses.

[0098] In the embodiment of the present application, the gain ratio is the ratio of the power of the amplified signal to the power of the original signal, and the calculation formula is:

[0099]

[0100] Where Pamplified is the amplified signal power; P original is the original signal power.

[0101] It should be noted that a higher gain ratio means that the signal is effectively amplified during transmission and can be transmitted over a longer distance or in a more complex environment. However, too high a gain may also introduce additional noise and distortion. Therefore, in practical applications, it is necessary to reasonably set the gain ratio according to the specific communication environment and requirements to achieve the best communication effect.

[0102] In the embodiments of the present application, the subcarrier delay ratio is the ratio of the transmission times of different subcarrier signals, and the calculation formula is:

[0103]

[0104] In the formula, T a and T b are the transmission times of different subcarriers respectively.

[0105] In the embodiments of the present application, by analyzing the subcarrier delay ratio, the characteristics of the signal transmission path can be understood, such as the length of the line, the electromagnetic interference situation, etc. In the identification of the household-transformer relationship, the subcarrier delay ratios between different user electricity meters and concentrators will vary due to different physical distances and line conditions. Using this characteristic, the household-transformer relationship can be identified and judged.

[0106] In the embodiments of the present application, these key parameters reflect the characteristics of the signal during transmission. Different line connections and electrical equipment will have different effects on the signal. By analyzing and calculating these effects, the actual connection situation between households and transformers can be inferred.

[0107] In the embodiments of the present application, the inter-network coordination calculation in step S1 includes the following steps:

[0108] Construct a first data set by combining the first target parameter with communication data;

[0109] Optimize the communication parameters between nodes using the first algorithm to obtain the optimal communication parameters.

[0110] In the embodiments of the present application, constructing a first data set by combining the first target parameter with communication data includes forming a multi-dimensional feature vector by constituting a data set with the calculated signal-to-noise ratio (SNR), gain ratio, and subcarrier delay ratio of different nodes and node data, such as X i = [i, SNR i , Gain_Ratio i , Delay_Ratio i . Set a calculation program in the code to calculate the required node feature vector data to form a data set.

[0111] In an optional embodiment, the first algorithm may be a particle swarm optimization algorithm, a genetic algorithm, or a simulated annealing algorithm;

[0112] In an optional embodiment, the particle swarm optimization algorithm includes initializing a group of particles to represent communication parameters, evaluating the fitness of each particle, and iteratively updating the position of the particle based on the best experience of the individual and the group to minimize the communication error and optimize the communication performance.

[0113] In an optional embodiment, the genetic algorithm includes initializing a set of communication parameters as a population, continuously iteratively updating the population through fitness evaluation, selection, crossover and mutation operations, and ultimately finding the optimal communication parameter combination that minimizes the communication error.

[0114] In an optional embodiment, the gradient descent algorithm includes a gradient descent algorithm that initializes a set of communication parameters, calculates the gradient of the communication error, and iteratively updates the parameters in the opposite direction of the gradient to minimize the communication error and optimize the communication performance.

[0115] It should be noted that although a variety of optimization algorithms can be used, the particle swarm optimization algorithm is an ideal choice in the precise positioning technology of anti-electricity theft in distribution networks because of its simplicity, ease of implementation, strong global search capability and fast convergence speed.

[0116] In an embodiment of the present application, the first algorithm selects a particle swarm optimization algorithm, which is used to adjust the communication parameters between nodes based on the previously formed data set, with minimizing the communication error as the objective function, and continuously iteratively updates the communication parameters of the nodes to achieve optimization of communication performance.

[0117] In an embodiment of the present application, the connection relationship between the user's electric meter and the transformer includes drawing a signal transmission path diagram based on the analysis results of the signal-to-noise ratio (SNR), gain ratio and subcarrier delay ratio. By analyzing these key parameters, the strength changes of the signal during transmission, the amplification situation and the delay differences of different subcarriers can be clearly understood, thereby intuitively displaying the signal transmission path. The optimal communication parameters obtained by inter-network coordination calculation and optimization will affect the transmission characteristics of the signal, and then affect the judgment of the topological relationship. On this basis, the connection relationship between the node and the transformer is mapped to an electrical logic topology diagram of the substation area, and the electrical connection relationship between the various devices in the substation area is presented in a visual manner. Finally, a substation topology relationship table is generated to form a complete first model, namely the electrical topology model. The complex topology information is presented in a tabular form, which is convenient for management personnel to consult and refer to, and provides strong support for the operation and maintenance management of the substation area.

[0118] In the embodiment of the present application, the first model further includes, in the automatic maintenance work of the ledger, first updating the ledger, automatically updating the substation area equipment ledger according to the identified household-transformer relationship, and recording the communication parameters of each node, such as signal-to-noise ratio (SNR), gain ratio, and subcarrier delay ratio. At the same time, continuously carry out anomaly detection. By continuously monitoring the changes in SNR, gain ratio, and subcarrier delay ratio, once anomalies are found, such as a sudden drop in SNR, abnormal gain ratio, etc., promptly alarm and make records. Finally, generate reports, generating a substation area topology relationship report and an updated substation area equipment ledger report, providing comprehensive and accurate information for the operation and maintenance management of the substation area.

[0119] In the embodiment of the present application, constructing the second model in step S2 includes the following steps:

[0120] Perform a first preprocessing on the user power consumption data to obtain a second data set;

[0121] The first preprocessing includes performing a first analysis on the power consumption data and screening out abnormal data;

[0122] Construct a second model according to the second data set in combination with a second algorithm.

[0123] In the embodiment of the present application, performing a first preprocessing on the user power consumption data to obtain a second data set includes, in the data fusion analysis stage, the system collects multi-parameter data such as voltage, current, and power, and conducts a comparative analysis of real-time and historical data to identify abnormal changes. First, perform data preprocessing, perform cleaning, normalization, and feature extraction operations on the collected raw data to eliminate noise and redundant information; by performing multi-parameter fusion analysis on data such as voltage, current, and power, form a data set, such as Y i =[i, U i , I i , P i , …]; Finally, perform anomaly detection, and initially screen out abnormal data with the help of statistical analysis (such as standard deviation, mean) and a rule engine (such as voltage imbalance, current anomaly).

[0124] In an alternative embodiment, the second algorithm can be a random forest algorithm, can be a support vector machine (SVM) algorithm, or can also be a K-nearest neighbor (KNN) algorithm;

[0125] In an alternative embodiment, the random forest algorithm includes constructing multiple decision trees, randomly sampling and feature selecting the training data, splitting nodes based on information gain or Gini index, and finally classifying new input data through a voting mechanism to identify and locate electricity theft behavior.

[0126] In an alternative embodiment, the Support Vector Machine (SVM) algorithm includes mapping the training data to a high-dimensional space by selecting an appropriate kernel function and hyperparameters to find the optimal classification hyperplane, and classifying new input data to identify and locate electricity theft behavior.

[0127] In an alternative embodiment, the K-Nearest Neighbor (KNN) algorithm includes calculating the distance between new input data and training samples, selecting the nearest K neighbors, and classifying the new data according to the majority class of these neighbors to identify and locate electricity theft behavior.

[0128] It should be noted that although there are various alternative algorithms, in the anti-electricity-theft warning model, the random forest has high accuracy. By integrating multiple decision trees, it can effectively reduce the risk of overfitting and thus improve the prediction accuracy. At the same time, it has strong robustness and has a good tolerance for noise data and missing values. In terms of interpretability, the random forest can identify the features that have the greatest impact on electricity theft behavior through feature importance analysis. In addition, the random forest also supports parallel computing, that is, distributed computing, which is very suitable for processing large-scale data.

[0129] In the embodiment of the present application, the random forest algorithm includes the following steps:

[0130] (1) First, extract samples of normal electricity consumption and electricity theft behavior from historical data, and divide the data set into a 70% training set and a 30% test set.

[0131] (2) The random forest consists of n decision trees. Each decision tree is trained by randomly sampling training data and features, randomly drawn from the training set, and selecting the optimal feature for node splitting based on information gain or Gini index.

[0132] Decision tree node splitting criterion (Gini coefficient):

[0133]

[0134] In the formula: D is the data set, p k is the proportion of samples in the k-th class.

[0135] For the sorting of the feature parameters of the training set, it is necessary to calculate the feature importance, calculated by the number of splits of the feature in all decision trees and the information gain:

[0136]

[0137] In the formula: ΔGini(t,j) is the reduction in the Gini coefficient after node t is split using feature j.

[0138] (3) For the new input data, each decision tree outputs a prediction result, and the final prediction result is determined by voting.

[0139] (4) In the precise positioning phase, by combining the topology relationship of the power distribution area and the prediction results of the random forest, first determine the power distribution area range where the abnormal data source is located according to the power grid topology structure, then screen out suspected electricity theft users in the abnormal power distribution area, and finally conduct on-site inspections according to the system prompts to confirm and handle electricity theft behaviors.

[0140] In the embodiment of the present application, the second algorithm selects the random forest algorithm. By constructing multiple decision trees and conducting voting, a second model, namely the anti-electricity-theft early warning model, is created to achieve precise identification of electricity theft behaviors.

[0141] In the embodiment of the present application, the specific technical implementation method, common abnormal patterns and their characteristics of the anti-electricity-theft early warning model include:

[0142] In terms of data collection, it is necessary to monitor the voltage data at the user end in real time, including phase voltage and line voltage; current data, covering phase current and neutral line current; power data, namely active power, reactive power and power factor; and electric energy data, that is, the electricity consumption (kWh) of the user. The data monitoring method mainly collects the user's electricity consumption data in real time through smart meters, installs a power distribution area intelligent diagnosis device on the side of the power distribution area transformer to monitor the electricity consumption data of the entire power distribution area, and at the same time performs high-speed sampling on the voltage and current signals to capture instantaneous abnormalities.

[0143] In terms of abnormal behavior detection, it is divided into voltage abnormal detection, current abnormal detection, power abnormal detection, and electric energy abnormal detection. Voltage abnormal detection includes monitoring the balance of three-phase voltages. If a certain phase voltage is significantly lower than the other two phases, or an abnormal voltage drop is detected, there may be electricity theft behaviors; current abnormal detection monitors the balance of three-phase currents. If a certain phase current is significantly lower than the other two phases, or the neutral line current increases abnormally, and analyzes whether the current waveform is distorted, such as clipping, truncation, etc., it may be an electricity theft behavior; power abnormal detection focuses on the power factor of the user. If the power factor is abnormally low, or the active power is significantly lower than the normal value, or the reactive power is abnormally high, it may be that the electricity theft user steals electricity through a specific load or device; electric energy abnormal detection monitors the user's electricity consumption. If the electricity consumption suddenly drops significantly, or the electricity consumption fluctuates abnormally, such as a sudden increase in electricity consumption at night, there may be electricity theft behaviors.

[0144] The common abnormal patterns mainly include the following electricity theft methods:

[0145] (1) Electricity theft by wrong wiring, whose characteristic manifestations are voltage imbalance, that is, a certain phase voltage is significantly lower than the other two phases; current imbalance, a certain phase current is significantly lower than the other two phases; abnormal neutral line current, the neutral line current increases significantly. When detecting, it can be judged by monitoring the balance of three-phase voltages and currents and whether the neutral line current is abnormal.

[0146] (2) Power theft by voltage method, characterized by voltage drop, with the voltage of a certain phase significantly lower than the normal value, and at the same time, the current waveform is distorted, showing clipping or truncation. The detection method is to monitor whether the voltage shows abnormal drop and analyze whether the current waveform is distorted.

[0147] (3) Bypassing method of power theft, characterized by a sudden drop in electricity consumption, with the user's electricity consumption suddenly dropping significantly, and the power factor being abnormally low, significantly lower than the normal value. It can be detected by monitoring the change in the user's electricity consumption and whether the power factor is abnormal.

[0148] (4) Power theft by non-linear load, characterized by distorted current waveform, showing obvious clipping and truncation, and abnormal reactive power, significantly higher than the normal value. When detecting, it is only necessary to analyze whether the current waveform is distorted and monitor whether the reactive power is abnormal.

[0149] (5) Power theft through the neutral line, characterized by abnormal neutral line current, with the neutral line current significantly increasing, and voltage imbalance, with the voltage of a certain phase significantly lower than the other two phases. When detecting, monitor whether the neutral line current is abnormal and the balance of the three-phase voltages.

[0150] In the embodiment of the present application, in step S4, the line loss data of different times and regions are compared and analyzed to locate the line loss abnormal regions, including collecting data from sectional metering devices, combining the electrical topology information of the substation area, calculating the branch and hierarchical line losses, and reporting line loss abnormal events. The power flow calculation method is one of the core methods for line loss analysis, used to calculate the voltage, current, and power distribution in the power system, and then deduce the line loss. The power flow calculation is based on the node voltage equation of the power system and is usually represented by the node admittance matrix. For each node in the power system, there is the following power balance equation:

[0151]

[0152] In the formula, P i is the active power injection of node i; Q i is the reactive power injection of node i; V i is the voltage of node i; V ij is the admittance between node i and node j; n is the total number of system nodes.

[0153] In an optional embodiment, the third algorithm can be the Newton-Raphson method, can be the Gauss-Seidel Method, or can also be the Fast Decoupled Method;

[0154] In an alternative embodiment, the Newton-Raphson method includes calculating the power imbalance by initializing the node voltage values, and iteratively updating the voltage values using the Jacobian matrix until the convergence condition is met, thereby calculating the voltage, current, and power distributions at each node, and further deriving the line loss.

[0155] In an alternative embodiment, the Gauss-Seidel method includes initializing the node voltage values, updating the voltage of each node one by one, calculating the power imbalance based on the current voltage values, and repeating the iteration until the convergence condition is met, thereby calculating the voltage, current, and power distributions at each node, and further deriving the line loss.

[0156] In an alternative embodiment, the fast decoupled method includes initializing the node voltage values, decomposing the Jacobian matrix into independent sub-matrices of active and reactive power, and iteratively updating the voltage phase angle and amplitude respectively until the convergence condition is met, thereby calculating the voltage, current, and power distributions at each node, and further deriving the line loss.

[0157] It should be noted that in the lean analysis and management of distribution network line loss, although multiple algorithms can be used for power flow calculation, the Newton-Raphson method is characterized by its high convergence speed, applicability to complex power systems, strong robustness, and high accuracy.

[0158] In the embodiment of the present application, the power flow calculation is iteratively solved using the Newton-Raphson method. The specific steps are as follows:

[0159] (1) Initialization: Given the initial voltage values of each node (usually set to the nominal voltage).

[0160] (2) Calculate the power imbalance:

[0161]

[0162] Where P i spec and are the given active and reactive powers of node i; and are the active and reactive powers calculated based on the current voltage values.

[0163] (3) Update the voltage values

[0164]

[0165] Where Δθ and Δ|V| are the correction amounts of the voltage phase angle and amplitude respectively; J is the Jacobian matrix, representing the partial derivative of power with respect to voltage.

[0166] (4) Iterative convergence: Repeat the above steps until the power imbalance meets the convergence condition.

[0167] Line power loss:

[0168]

[0169] Where ΔP ij and ΔQ ij are the active power and reactive power losses of the line between node i and node j respectively, P i , Q i and U i are the active power, reactive power and voltage on the side of node i respectively; R ij is the line resistance between node i and node j; X ij is the line reactance between node i and node j.

[0170] It should be noted that using the power flow calculation method to calculate line losses for lean analysis and management has obvious advantages compared with the integration by parts method, average current method, and maximum current method. It calculates more accurately and comprehensively, fully considering the complex electromagnetic relationships and component characteristics in the power system, rather than simply relying on data statistics to estimate line losses, avoiding errors caused by ignoring electrical parameters and actual operating conditions, and can more accurately reflect the actual line loss situation of the power system; it has stronger adaptability and is applicable to power systems of various scales and complexities. Whether it is a simple small-scale distribution network or a large-scale transmission network with a complex structure, accurate line loss calculations can be carried out through reasonable parameter settings and model construction. It is not limited by the scale of the power system or the complexity of the network topology structure, and has strong generality and adaptability; power flow calculation can comprehensively consider factors such as the network topology structure, power source distribution, and load characteristics of the power system, and conduct a detailed analysis of the operating state of the entire power system. When calculating line losses, not only the total line loss value can be obtained, but also the line loss distribution of each line and each node can be analyzed, which provides rich and detailed information for lean analysis and management of line losses, facilitating maintenance personnel to deeply understand the specific locations and causes of line losses, and thus formulating targeted loss reduction measures.

[0171] It should be noted that the anti-stealing electricity precise positioning technology has important applications in scenarios such as distribution transformer area line loss analysis, reporting of line loss abnormal events, and distribution network optimization. In terms of distribution transformer area line loss analysis, the power flow calculation method is used to accurately calculate the line losses of each level of branches in the distribution transformer area, and at the same time, the daily, hourly, and 15-minute line loss rates of the distribution transformer area are statistically analyzed to identify abnormal line losses. In the link of reporting line loss abnormal events, abnormal events are identified by comparing the actual line loss with the theoretical line loss, and the specific location of the line loss abnormality, such as a certain line or a certain transformer, is located in combination with the electrical topology information of the distribution transformer area. In terms of distribution network optimization, network reconstruction is carried out by means of power flow calculation analysis to optimize the operation mode of the distribution network, reduce line losses, and the best installation location and capacity of reactive power compensation devices are determined through power flow calculation to improve the power factor and further reduce line losses, so as to comprehensively improve the operation efficiency and stability of the power system and facilitate the efficient development of anti-stealing electricity work.

[0172] It should be noted that the present invention uses the multi-source data fusion technology to integrate various electrical parameters such as voltage, current, and power, as well as information such as smart meter readings and switch states. Compared with the traditional method that only relies on partial limited data or manual inspection, it can construct an electrical topology model more comprehensively and accurately. For example, by analyzing the complex relationship between voltage drops and current flow directions between different nodes and combining the influence of switch opening and closing states on current paths, the connection relationships between various devices can be accurately depicted, effectively avoiding topological structure errors caused by one-sided data or manual judgment errors. By comparing in detail with historical topology data and equipment ledger information, inconsistencies or errors in the topology model can be discovered and corrected in a timely manner. For example, after equipment updates or line renovations, the real-time accuracy of topology information can be quickly ensured, providing a reliable topological basis for power system operation and maintenance, while traditional methods are difficult to achieve such efficient and accurate verification and update. Continuously collect data and transmit it to the topology recognition and verification module, automatically process and analyze the data to construct the topology, greatly shortening the topology recognition cycle, and being able to respond in a timely manner to changes in the distribution transformer area, such as the addition of equipment or line adjustments, etc. In the face of complex distribution transformer areas, traditional manual methods often take a lot of time and are difficult to keep up with the change rhythm. The normal electricity consumption pattern baseline model established based on data mining and machine learning algorithms can accurately capture the subtle differences in electricity consumption characteristics of different users at different times through in-depth analysis of a large amount of historical electricity consumption data and real-time collected data. For example, by analyzing multi-dimensional electricity consumption characteristics such as the shape of the power curve and the change of the current-voltage phase relationship, compared with the traditional anti-stealing electricity methods based on simple threshold judgment or empirical judgment, abnormal electricity consumption behaviors can be identified more accurately, effectively reducing the misjudgment rate.

[0173] It should be noted that the present invention locates electricity theft in combination with accurate electrical topology information, and starts from the transformer end to gradually analyze the power change of each branch node along the topology structure. Once a region with abnormal power loss and no reasonable electricity consumption explanation is found, the scope of suspicion can be quickly narrowed down to specific branch lines and their downstream regions, and then the possible electricity theft users or electricity theft points can be accurately located. However, traditional anti-electricity-theft technologies are often relatively vague when locating electricity theft positions and it is difficult to pinpoint specific users or line nodes.

[0174] It should be noted that the present invention divides the low-voltage distribution power station area into multiple fine calculation units according to the accurate electrical topology structure, such as main line segments, branch line segments, and units containing transformers, etc., and accurately calculates the specific line parameters (length, wire material, cross-sectional area, etc.) and operation data (current, load rate, etc.) of each unit. For example, when calculating the line resistance loss, the actual situation of different line segments is fully considered. Compared with the traditional method of roughly estimating or simply averaging the line loss calculation, a more accurate line loss value can be obtained, providing a more reliable data basis for line loss management. By adopting the combination of real-time calculation and segmented analysis, the change of line loss can be tracked in real time, and abnormal fluctuations of line loss can be found in time. Through the detailed comparison and analysis of line loss data in different time periods (peak-valley periods, different seasons, etc.) and different regions, the change law of line loss can be deeply understood, while traditional line loss analysis methods are difficult to achieve such refined real-time monitoring and analysis. When the line loss is found to increase abnormally, the reason can be accurately judged in combination with the results of the anti-electricity-theft analysis module. If the electricity theft factor is excluded, it is possible to further deeply analyze whether there are other factors such as increased resistance due to line aging and three-phase imbalance, and formulate targeted loss reduction measures, such as arranging line maintenance and adjusting load distribution. This lean management mode from cause analysis to measure formulation can more effectively reduce line loss, improve the operation efficiency and economic benefits of the power system compared with the traditional relatively extensive line loss management method.

[0175] It should be noted that the present invention integrates a number of key technologies such as electrical topology recognition and correlation verification, precise positioning of electricity theft in the distribution network, and lean analysis and management of line loss, realizing comprehensive management of the low-voltage distribution power station area from topology structure to power loss to electricity theft prevention. It is no longer a single-function technology application, but forms an organic whole, which can comprehensively handle various operation management problems of the distribution power station area, while traditional technologies can often only solve problems in one aspect and lack systematicness and comprehensiveness.

[0176] This embodiment also provides a comprehensive management and optimization system for a low-voltage distribution power station area, including:

[0177] An electrical topology modeling module, which is used to collect communication data of each distribution power station area and construct a first model through inter-network coordination calculation;

[0178] An anti-stealing electricity analysis module, configured to construct a second model based on a first model and communication data through feature importance analysis;

[0179] A stealing electricity point positioning module, configured to locate a stealing electricity point by comparing the deviation between the second model and real-time electricity consumption data;

[0180] A line loss anomaly analysis module, configured to perform a comparative analysis on line loss data of different times and regions based on the first model and the second model, and locate a line loss anomaly region;

[0181] A comprehensive result display module, configured to integrally display the processing results of stealing electricity point positioning and line loss anomaly region positioning.

[0182] Furthermore, it further includes:

[0183] A memory, configured to store a program;

[0184] A processor, configured to load the program to execute the comprehensive management and optimization method for a low-voltage distribution network area.

[0185] This embodiment further provides a computer-readable storage medium, which stores a program, and when the program is executed by a processor, the comprehensive management and optimization method for a low-voltage distribution network area is implemented.

[0186] The storage medium proposed in this embodiment and the comprehensive management and optimization method for a low-voltage distribution network area proposed in the above embodiment belong to the same inventive concept. Technical details not described in detail in this embodiment can be referred to in the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0187] Through the above description of the embodiments, those skilled in the art can clearly understand that the present invention can be implemented by means of software and necessary general-purpose hardware. Of course, it can also be implemented by hardware, but in many cases, the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, in essence, or the part that makes a contribution to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk, or optical disc of a computer, etc., including several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of various embodiments of the present invention.

[0188] Embodiment 3

[0189] This is an embodiment of the present invention, which provides a comprehensive management and optimization system for low-voltage distribution substations, including an electrical topology modeling module, an anti-stealing electricity analysis module, a stealing electricity point positioning module, a line loss anomaly analysis module, and a comprehensive result display module;

[0190] In the embodiment of the present application, the electrical topology modeling module includes collecting communication data of each distribution substation and constructing a first model through inter-network coordination calculation;

[0191] In the embodiment of the present application, the electrical topology modeling module is constructed based on the edge physical components, branch line monitoring terminals, and meter box monitoring units of the intelligent substation situation awareness complete set of equipment, and realizes the electrical physical topology based on the characteristic current algorithm.

[0192] In the embodiment of the present application, the electrical topology modeling module further includes an edge physical component as the core control unit, which is communicatively connected to the branch line monitoring terminal and the meter box monitoring unit. The branch line monitoring terminal detects the characteristic current in the line and transmits the information back to the edge physical component. The meter box monitoring unit collects and monitors the meter information in the meter box, and jointly realizes the construction of the electrical physical topology of the substation area.

[0193] In the embodiment of the present application, the anti-stealing electricity analysis module includes constructing a second model through feature importance analysis based on the first model and communication data;

[0194] In the embodiment of the present application, the anti-stealing electricity analysis module further includes using big data analysis technology to process and analyze the collected communication data. Through feature importance analysis, the key features affecting anti-stealing electricity are identified, such as voltage, current, power, etc. An anti-stealing electricity early warning model is constructed, and machine learning algorithms (such as random forest, support vector machine, etc.) are used to train according to historical data and feature importance to generate a second model.

[0195] In the embodiment of the present application, the stealing electricity point positioning module includes positioning the stealing electricity point by comparing the deviation between the second model and the real-time electricity consumption data;

[0196] In the embodiment of the present application, the stealing electricity point positioning module further includes real-time collecting the electricity consumption data of the substation area, including parameters such as voltage, current, and power. The real-time electricity consumption data is input into the second model to calculate the deviation between the predicted value and the actual value. If the deviation exceeds the set threshold, it is judged as abnormal data, and the possible stealing electricity point is located.

[0197] In the embodiment of the present application, the line loss anomaly analysis module includes comparing and analyzing the line loss data at different times and regions based on the first model and the second model, and positioning the line loss anomaly region;

[0198] In the embodiment of the present application, the line loss anomaly analysis module further includes collecting and analyzing the power consumption data of the master meter and each level of branch meters according to the identified electrical topology. Using calculation methods such as the shortest path algorithm in graph theory, the branch-level line loss is quickly calculated. By comparing the line loss data at different times and regions, the line loss anomaly regions are identified, and the reasons are further analyzed.

[0199] In the embodiment of the present application, the comprehensive result display module includes integrating and displaying the processing results of the electricity theft point positioning and the line loss anomaly region positioning.

[0200] In the embodiment of the present application, the comprehensive result display module further includes building a terminal system and an intelligent display platform, simulating various operating conditions of the power distribution area, testing functional indicators such as the topology recognition accuracy rate of the complete set of equipment, the fault judgment accuracy, and the line loss calculation accuracy. The test results are intuitively displayed through the intelligent display platform to evaluate and optimize the performance of the equipment. The results of the electricity theft point positioning and the line loss anomaly region positioning are summarized and displayed on the platform for the reference and decision-making of the operation and maintenance personnel.

[0201] Through the data sharing and collaborative working mechanism among the modules, efficient information interaction and linkage are achieved. For example, the electrical topology modeling module provides an accurate topology structure basis for the anti-electricity-theft analysis module and the line loss calculation and management module. The results of the anti-electricity-theft analysis module can provide reference for judging the reasons for line loss anomalies. The comprehensive result display module integrates and displays the processing results of each module and realizes human-computer interaction and remote control. This collaborative working mode can quickly respond to various changes in the operation conditions of the power distribution area, take corresponding measures in a timely manner, improve the management efficiency and operation reliability of the entire low-voltage power distribution area. However, due to the lack of integration and coordination in traditional technologies, they are often inefficient and prone to information island phenomena when dealing with complex problems.

[0202] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A comprehensive management and optimization method for low-voltage distribution transformer areas, characterized in that, Including: Collect communication data of each distribution transformer area, and construct a first model through inter-network coordination calculation; Based on the first model and communication data, construct a second model through feature importance analysis; Locate the electricity theft point by comparing the deviation between the second model and real-time electricity consumption data; Based on the first model and the second model, conduct a comparative analysis of the line loss data at different times and regions, and locate the line loss abnormal regions; Integrate and display the processing results of the electricity theft point location and the line loss abnormal region location.

2. The integrated management and optimization method for low-voltage distribution transformer areas according to claim 1, characterized in that: The constructing of the first model includes the following steps: Collect communication data of the transformer area nodes; Based on the collected communication data, calculate the first target parameter; According to the first target parameter and combined with inter-network coordination calculation, obtain the connection relationship between the user electricity meter and the transformer; Obtain the first model according to the connection relationship.

3. The integrated management and optimization method for low-voltage distribution transformer areas as claimed in claim 1 or 2, characterized in that: The inter-network coordination calculation includes the following steps: Construct a first data set through the first target parameter and combined with communication data; Adopt the first algorithm to optimize the communication parameters between nodes to obtain the optimal communication parameters.

4. The integrated management and optimization method for low-voltage distribution substation areas according to claim 3, characterized in that: The constructing of the second model includes the following steps: Conduct a first preprocessing on the user electricity consumption data to obtain a second data set; The first preprocessing includes conducting a first analysis on the electricity consumption data and screening out abnormal data; Construct a second model according to the second data set and combined with the second algorithm.

5. The integrated management and optimization method for low-voltage distribution substation areas according to claim 4, characterized in that: The second algorithm includes the following steps: Extract samples of normal electricity consumption and electricity theft behaviors from historical data and construct multiple decision trees; Conduct feature selection and node splitting based on the Gini coefficient; Conduct feature importance calculation through the splitting times and information gain of this feature in all decision trees; For the new input data, each decision tree outputs a prediction result; Combined with the prediction results of the first model and the second algorithm, determine the range of the transformer area where the abnormal data source is located according to the power grid topology structure, and screen out the electricity theft users in the abnormal transformer area.

6. The integrated management and optimization method for low-voltage distribution transformer areas as described in claim 5, characterized in that: The locating of the electricity theft point by comparing the deviation between the second model and real-time monitored electricity consumption data includes the following steps: Compare the real-time monitored electricity consumption data with the second model to identify the first abnormal electricity consumption behavior; the first abnormal electricity consumption behavior includes at least one of the following: voltage abnormality, current abnormality, power abnormality, and electric energy abnormality; Combined with the prediction results of the first model and the second model, lock and confirm the electricity theft users.

7. The integrated management and optimization method for low-voltage distribution substation area according to claim 6, characterized in that: The comparative analysis of the line loss data at different times and regions and the location of the line loss abnormal regions include the following steps: Collect data of the sectional metering device; Use the third algorithm and combined with the first model to calculate the line loss of each level of branch; Statistically analyze the daily, hourly, and 15-minute line loss rates of the transformer area to identify line loss abnormalities; Locate the specific location of the line loss abnormality by comparing the actual line loss with the theoretical line loss and combined with the second model.

8. A system for the comprehensive management and optimization method of a low-voltage distribution transformer area according to claim 1, characterized in that: An electrical topology modeling module, configured to collect communication data of each distribution transformer area, and construct a first model through inter-network coordination calculation; An anti-electricity-theft analysis module, configured to construct a second model through feature importance analysis based on the first model and communication data; The electricity theft point positioning module is used to locate the electricity theft point by comparing the deviation between the second model and the real-time electricity consumption data; The line loss anomaly analysis module is used to compare and analyze the line loss data in different times and regions based on the first model and the second model, and locate the line loss anomaly regions; The comprehensive result display module is used to integrally display the processing results of the electricity theft point positioning and the line loss anomaly region positioning.

9. A computing device, characterized in that, It includes: A memory for storing programs; A processor for loading the program to execute the steps of the comprehensive management and optimization method for low-voltage distribution transformer areas according to any one of claims 1-7.

10. A computer-readable storage medium storing a program, characterized in that, When the program is executed by the processor, the steps of the comprehensive management and optimization method for low-voltage distribution transformer areas according to any one of claims 1-7 are implemented.

Images