Small current ground fault sectioning method and device, storage medium and electronic equipment

Abstract

This application discloses a method, device, storage medium, and electronic equipment for selecting fault sections with low current grounding faults, relating to the field of distribution network fault detection technology. The method includes: real-time acquisition of electrical data from the distribution network busbars; triggering a grounding fault assessment program and determining the type of low current grounding fault; dividing the distribution network into independent sections using distribution switches as boundaries; building equivalent circuit models for each section, adapted to the fault type, based on distribution network fundamental parameters; and simultaneously acquiring measured zero-sequence current data for each section; inputting distribution network fundamental parameters, fault type, and model characteristics into a trained artificial intelligence model; outputting the predicted distribution of zero-sequence current for each section; and determining the fault section based on similarity calculations with measured data; and generating switch operation commands based on the fault section to instruct the completion of physical isolation and automatic load transfer of the fault section. This application can achieve precise location of low current grounding faults at the section level.

Description

Technical Field

[0001] This application relates to the field of power distribution network fault detection technology, and in particular to a method, device, storage medium and electronic equipment for selecting sections of low-current grounding faults. Background Technology

[0002] The power distribution network is the core link in the power system for power distribution and supply. Low-current grounding systems with voltage levels of 10kV and below are widely used in power distribution network construction due to their high power supply reliability and small fault impact range. These systems are prone to single-phase grounding faults during operation. If the specific fault section cannot be quickly and accurately located, it will not only significantly increase the workload of maintenance personnel during line inspections and reduce fault handling efficiency, but may also trigger secondary accidents such as arcing short circuits and overvoltages, disrupting the stable operation of the power distribution network and seriously affecting power supply reliability. Therefore, precise location technology for low-current grounding faults has become a key research focus in the field of power distribution network automation.

[0003] Currently, the industry's detection technologies for low-current grounding faults mainly focus on fault line selection. Mainstream techniques include traditional methods such as steady-state zero-sequence current phase comparison, amplitude and phase comparison, and zero-sequence impedance comparison, as well as improved methods such as transient line selection and injected current line selection. Some technologies also incorporate artificial intelligence models to identify fault types by extracting zero-sequence current waveform features. Additionally, some technologies improve line selection accuracy through dual-criteria verification and feature quantity integration. These existing technologies all rely on detecting the amplitude, phase, or transient characteristics of zero-sequence current and voltage to identify the faulty line, but none have overcome the limitation of only being able to select the line but not pinpoint the specific faulty section.

[0004] Existing low-current grounding fault detection technology can only determine the line to which the fault belongs, and cannot achieve precise location at the section level. This core problem means that maintenance personnel need to conduct full-line inspections of the faulty line, which not only consumes a lot of manpower and time costs and has low inspection efficiency, but also causes serious delays in the fault handling process, prolonging the impact time of the fault. At the same time, fault handling based on this technology can only be carried out around the entire line, making it difficult to achieve targeted fault isolation and load transfer, which can easily expand the power outage area and further reduce the power supply reliability of the distribution network, failing to meet the needs of automated and refined operation and maintenance of the distribution network. Summary of the Invention

[0005] In view of this, this application provides a method, apparatus, storage medium and electronic equipment for selecting sections of low-current grounding faults, which can achieve accurate location of low-current grounding faults at the section level.

[0006] According to a first aspect of this application, a method for selecting sections of a low-current grounding fault is provided, comprising: Real-time acquisition of electrical data of the distribution network busbars; based on the electrical data of the distribution network busbars, triggering a ground fault judgment program to determine the fault type of low current grounding in the distribution network; The distribution network is divided into several independent distribution network sections by using the distribution switch as the dividing point. An equivalent circuit model adapted to the fault type is built for each of the independent distribution network sections based on the basic parameters of the distribution network. The zero-sequence current measured data of each of the independent distribution network sections is collected based on the equivalent circuit model. The basic parameters of the distribution network, the fault type, and the model features of the equivalent circuit model are input into the trained artificial intelligence model, and the zero-sequence current prediction distribution of each independent section of the distribution network is output. The similarity between the zero-sequence current prediction distribution and the measured zero-sequence current data is calculated, and the fault section is determined based on the similarity calculation results. Based on the faulty section, a switch operation command is generated. The switch operation command is used to instruct the physical isolation of the faulty section and automatically realize the load transfer.

[0007] According to a second aspect of this application, a low-current ground fault selection device is provided, comprising: The judgment module is used to collect electrical data of the distribution network bus in real time, and trigger a ground fault judgment program based on the electrical data of the distribution network bus to determine the fault type of small current grounding in the distribution network. The data acquisition module is used to divide the distribution network into several independent distribution network sections with the distribution switch as the dividing point, build an equivalent circuit model adapted to the fault type for each of the independent distribution network sections in combination with the basic parameters of the distribution network, and collect the measured zero-sequence current data of each of the independent distribution network sections based on the equivalent circuit model. The calculation module is used to input the basic parameters of the distribution network, the fault type, and the model features of the equivalent circuit model into the trained artificial intelligence model, output the zero-sequence current prediction distribution of each independent section of the distribution network, perform similarity calculation between the zero-sequence current prediction distribution and the measured zero-sequence current data, and determine the fault section based on the similarity calculation results. The generation module is used to generate switch operation instructions based on the faulty section. The switch operation instructions are used to instruct the completion of physical isolation of the faulty section and automatic load transfer.

[0008] According to a third aspect of this application, a storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the above-described low-current grounding fault selection method.

[0009] According to a fourth aspect of this application, an electronic device is provided, including a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor, wherein the processor executes the program to implement the above-described low-current ground fault selection method.

[0010] Using the above technical solution, this application provides a method, device, storage medium, and electronic equipment for selecting low-current grounding fault sections. This method triggers fault analysis and accurately determines the type of low-current grounding fault by real-time acquisition of electrical data from the distribution network bus. Independent sections of the distribution network are demarcated using distribution switches as boundaries. Equivalent circuit models adapted to the fault type are built for each section, and corresponding zero-sequence current measurement data are collected. The basic parameters of the distribution network, the fault type, and the characteristics of the equivalent circuit model are then input into a trained artificial intelligence model. The model outputs the predicted distribution of zero-sequence current in each section and performs similarity calculations with the measured data to accurately determine the fault section. Finally, based on the fault section, switch operation commands are generated to achieve physical isolation of the fault section and automatic load transfer. This technical solution overcomes the limitations of existing low-current grounding fault detection technologies, which can only select lines but not precise sections. It enables precise section-level location of low-current grounding faults in distribution networks, reducing the patrol range of maintenance personnel to a single fault section. This significantly reduces manpower and time costs while improving patrol efficiency, effectively solving the problems of delayed fault handling and long impact time. At the same time, it enables targeted isolation and load transfer operations around the precisely located fault section, avoiding the need for handling the entire line, minimizing the power outage area, significantly improving the power supply reliability of the distribution network, and fully meeting the needs of automated and refined operation and maintenance of distribution networks.

[0011] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0012] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings: Figure 1 A flowchart illustrating a method for selecting a low-current grounding fault segment according to an embodiment of this application is shown. Figure 2 A flowchart illustrating a low-current grounding fault segmentation method according to another embodiment of this application is shown; Figure 3 This paper illustrates a schematic diagram of the principle of a low-current grounding fault segmentation method provided in an embodiment of this application. Figure 4 A schematic diagram of a low-current grounding fault selection device provided in an embodiment of this application is shown. Detailed Implementation

[0013] The present application will be described in detail below with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in the embodiments of the present application can be combined with each other.

[0014] Currently, the industry's detection technologies for low-current grounding faults mainly focus on fault line selection. Mainstream techniques include traditional methods such as steady-state zero-sequence current phase comparison, amplitude and phase comparison, and zero-sequence impedance comparison, as well as improved methods such as transient line selection and injected current line selection. Some technologies also incorporate artificial intelligence models to identify fault types by extracting zero-sequence current waveform features. Additionally, some technologies improve line selection accuracy through dual-criteria verification and feature quantity integration. These existing technologies all rely on detecting the amplitude, phase, or transient characteristics of zero-sequence current and voltage to identify the faulty line, but none have overcome the limitation of only being able to select the line but not pinpoint the specific faulty section.

[0015] Existing low-current grounding fault detection technology can only determine the line to which the fault belongs, and cannot achieve precise location at the section level. This core problem means that maintenance personnel need to conduct full-line inspections of the faulty line, which not only consumes a lot of manpower and time costs and has low inspection efficiency, but also causes serious delays in the fault handling process, prolonging the impact time of the fault. At the same time, fault handling based on this technology can only be carried out around the entire line, making it difficult to achieve targeted fault isolation and load transfer, which can easily expand the power outage area and further reduce the power supply reliability of the distribution network, failing to meet the needs of automated and refined operation and maintenance of the distribution network.

[0016] To address the aforementioned technical problems, embodiments of the present invention provide a method for selecting fault segments in low-current grounding faults, such as... Figure 1 As shown, the method includes: Step 110: Collect electrical data of the distribution network bus in real time, trigger the ground fault judgment program based on the electrical data of the distribution network bus, and determine the fault type of small current grounding in the distribution network.

[0017] Among them, the electrical data of the distribution network bus refers to various data at the bus of the distribution network substation that can reflect the electrical status of the power grid operation. It is the core monitoring data for identifying ground faults. The core includes the three-phase voltage data of the bus, and can also cover derived electrical parameters such as voltage phase and voltage distortion rate. The ground fault judgment program is a special program in the distribution network used to identify the occurrence of small current ground faults and determine the specific fault type. It can distinguish ground faults from other power grid anomalies and identify the specific type of fault through preset electrical characteristic judgment logic. The fault type of small current grounding in the distribution network refers to the specific type of single-phase ground fault when a small current grounding system in the distribution network occurs. The core is divided into complete ground faults and incomplete ground faults. Due to the significant differences in electrical characteristics and current distribution characteristics, the two require targeted subsequent fault handling.

[0018] During the detection of low-current grounding faults in distribution networks, electrical data at the distribution network busbars can be continuously collected in real time. The collected busbar electrical data is used as the core basis for fault judgment. The data is analyzed through preset judgment logic. When the data characteristics meet the triggering conditions of grounding faults, the grounding fault judgment program is automatically started. Based on the electrical characteristics presented by the busbar electrical data, the program accurately determines the specific type of low-current grounding fault that is currently occurring in the distribution network, providing the core fault type basis for subsequent adaptive fault modeling and fault location work.

[0019] This technical step involves real-time acquisition of electrical data from the distribution network busbars and using this data to trigger a ground fault assessment procedure. This enables rapid identification of low-current ground faults, avoids false triggering of non-ground faults, and improves the accuracy of fault assessment. Furthermore, the assessment procedure accurately determines the specific fault type of low-current grounding, providing crucial information for building suitable equivalent circuit models for different fault types and conducting targeted zero-sequence current acquisition and fault location. This effectively avoids the judgment bias caused by adapting a single model to different fault types.

[0020] Step 120: Divide the distribution network into several independent distribution network sections using the distribution switch as the dividing point. Based on the basic parameters of the distribution network, build an equivalent circuit model for each independent distribution network section that is adapted to the fault type. Collect the measured zero-sequence current data of each independent distribution network section based on the equivalent circuit model.

[0021] Among them, distribution switches refer to switching equipment used for line on / off control and electrical quantity monitoring in the distribution network. They are the core boundary nodes for dividing distribution network sections and can realize segmented management and control of distribution network lines and electrical data collection. Independent distribution network sections refer to independent distribution network line units that can realize electrical quantity monitoring, divided according to the actual line routing and electrical topology of the distribution network with distribution switches as the boundary. They are the basic units for accurate location of fault sections in the distribution network. Distribution network basic parameters refer to the core parameters that reflect the inherent electrical characteristics of distribution network lines and equipment, including line parameters and equipment parameters. Equivalent circuit models refer to circuit models that match the actual electrical characteristics of the distribution network based on distribution network basic parameters. They can accurately simulate the electrical operating status and current distribution characteristics of independent distribution network sections. Zero-sequence current measured data refer to the zero-sequence current related data collected on independent distribution network sections for small current grounding faults in the distribution network. They are measured data that reflect the fault electrical characteristics of independent distribution network sections.

[0022] In the embodiments of this disclosure, during the selection of low-current grounding fault sections, the distribution switch is used as the core dividing point. According to the actual line layout and electrical topology of the distribution network, the entire distribution network is divided into multiple independent distribution network sections. Then, the basic data of the distribution network is integrated and sorted out to extract the basic parameters of the distribution network. Combined with the determined low-current grounding fault type, an equivalent circuit model adapted to each distribution network section is built. This allows the model to accurately match the current distribution characteristics of the distribution network section under the corresponding fault type. Finally, based on the completed equivalent circuit model, zero-sequence current measurement data that reflects its electrical operating status is collected on each distribution network section, providing a measurement data basis for subsequent intelligent matching of fault sections.

[0023] This technical step, by dividing the distribution network into independent sections using distribution switches as boundaries, enables segmented management and control of the distribution network. It breaks through the limitations of traditional techniques that can only identify faulty lines, laying a structural foundation for accurate fault location. By combining basic distribution network parameters, it builds equivalent circuit models for each section, adapting to different fault types. This allows the models to accurately reflect the electrical characteristics of independent distribution network sections under different fault types, effectively improving the model's simulation accuracy of fault states. Simultaneously, by collecting measured zero-sequence current data from each independent distribution network section based on the equivalent circuit models, it ensures the matching of the collected data with the fault characteristics of the independent distribution network sections. This provides high-quality measured data support for subsequent accurate fault location through data comparison.

[0024] Step 130: Input the basic parameters of the distribution network, the fault type and the model features of the equivalent circuit model into the trained artificial intelligence model, output the zero-sequence current prediction distribution of each independent section of the distribution network, calculate the similarity between the zero-sequence current prediction distribution and the measured zero-sequence current data, and determine the fault section based on the similarity calculation results.

[0025] Among them, the model characteristics of the equivalent circuit model refer to the key attributes that reflect the core architecture and electrical characteristics of the equivalent circuit model, including the topology of the equivalent circuit, the parameter configuration adapted to the fault type, and the matching parameters with the actual electrical characteristics of the distribution network; the artificial intelligence model refers to an intelligent computing model that has been trained on a large number of fault samples and has the ability to predict the zero-sequence current distribution of the distribution network. It can accurately output the zero-sequence current distribution trend of different sections of the distribution network based on the input distribution network parameters and fault information; the zero-sequence current prediction distribution refers to the distribution state and numerical characteristics of the zero-sequence current when a small current grounding fault occurs in each independent section of the distribution network, based on the input parameters of the artificial intelligence model; similarity calculation refers to the process of quantitatively calculating the degree of fit between the zero-sequence current prediction distribution and the measured zero-sequence current data through a specific algorithm. The calculation results can intuitively reflect the matching between the two; the fault section refers to the independent section of the distribution network where a small current grounding fault actually occurs. It is the line unit that best matches the fault electrical characteristics after similarity calculation.

[0026] In the embodiments of this disclosure, during the small current grounding fault section determination stage, the basic parameters of the distribution network, the determined small current grounding fault type, and the model features of the equivalent circuit model built for each independent section of the distribution network are input into the trained artificial intelligence model. The artificial intelligence model combines the input information to perform fault simulation and calculation, and outputs the predicted distribution of zero-sequence current corresponding to each independent section of the distribution network when a fault occurs. Then, the predicted distribution of zero-sequence current of each independent section of the distribution network is compared with the previously collected measured data of the corresponding zero-sequence current for the entire domain. Finally, based on the similarity calculation results of each independent section of the distribution network, the independent section of the distribution network that best matches the actual fault characteristics is determined as the fault section.

[0027] This technical step inputs multi-dimensional information, including distribution network basic parameters, fault types, and equivalent circuit model characteristics, into a trained artificial intelligence model. This allows the generation of zero-sequence current prediction distributions to possess accurate distribution network attributes and fault adaptability, ensuring the scientific rigor and relevance of the prediction data. By calculating the similarity between the zero-sequence current prediction distribution and measured data, and using this similarity to determine the fault section, quantitative judgment of the fault section can be achieved. This replaces the traditional method of relying solely on electrical quantity amplitude and phase judgment, effectively improving the accuracy of fault section determination. Simultaneously, leveraging the rapid computing power of the artificial intelligence model, the time for fault section determination can be significantly shortened, achieving a breakthrough from line-level identification to precise section-level location of small-current grounding faults in the distribution network. This technically solves the core problem of traditional line selection techniques being unable to locate specific fault sections.

[0028] Step 140: Generate switch operation instructions based on the faulty section. The switch operation instructions are used to indicate the completion of physical isolation of the faulty section and automatic load transfer.

[0029] Among them, the switch operation command refers to the command generated based on the location information of the fault section, which is used to control the operation of the switch equipment in the distribution network. It can guide the switch equipment to complete the on and off operations, and realize the automated execution of fault section isolation and load transfer. The physical isolation of the fault section refers to the operation of operating the switch equipment in the distribution network to electrically disconnect the independent section of the distribution network with a small current grounding fault from the healthy distribution network line, so that the fault section is removed from the normal power supply network and the fault range is prevented from expanding further. The automatic load transfer refers to the process of automatically switching the power load originally powered by the fault section to the healthy independent section of the distribution network after the physical isolation of the fault section is achieved by operating the tie switch and other equipment in the distribution network, so as to restore the normal power supply to users.

[0030] In this embodiment of the present disclosure, after accurately determining the low-current grounding fault section, a corresponding switch operation command can be automatically generated based on the determined fault section location information. This command plans the operation of the switch equipment in the distribution network according to the topology of the distribution network and the preset fault handling principles. First, the physical isolation between the fault section and the healthy lines of the distribution network is completed through the operation of the relevant switch equipment. Then, after the isolation is completed, the corresponding tie switch is controlled according to the command to re-plan the power supply path for the load of the fault section and automatically transfer the load to the healthy independent section of the distribution network, thereby realizing the automated operation of fault handling.

[0031] This technology generates specific switching operation commands based on accurately identified faulty sections, automating the generation of fault handling instructions and replacing the traditional manual approach. This significantly reduces instruction preparation time and, through the switching operation commands, physically isolates the faulty section, quickly disconnecting the electrical connection between the faulty section and healthy lines, effectively preventing the fault from spreading and avoiding secondary accidents. Furthermore, this instruction enables automatic load transfer, rapidly restoring power to the faulty section after isolation, minimizing the outage area and duration, improving the continuity of power supply in the distribution network, and automating the entire process from section location to handling of low-current grounding faults. This greatly improves fault handling efficiency, enhances the reliability of the distribution network, and meets the needs of automated and refined operation and maintenance of the distribution network.

[0032] In summary, the low-current grounding fault segment selection method provided in this application triggers fault analysis and accurately determines the type of low-current grounding fault by real-time acquisition of electrical data of the distribution network bus. It divides the distribution network into independent segments with the distribution switch as the boundary, builds an equivalent circuit model adapted to the fault type for each segment, and collects corresponding zero-sequence current measured data. Then, it inputs the basic parameters of the distribution network, the fault type, and the characteristics of the equivalent circuit model into the trained artificial intelligence model, outputs the predicted distribution of zero-sequence current in each segment, and performs similarity calculation with the measured data to accurately determine the fault segment. Finally, it generates switch operation instructions based on the fault segment to realize physical isolation of the fault segment and automatic load transfer. This technical solution overcomes the limitations of existing low-current grounding fault detection technologies, which can only select lines but not precise sections. It enables precise section-level location of low-current grounding faults in distribution networks, reducing the patrol range of maintenance personnel to a single fault section. This significantly reduces manpower and time costs while improving patrol efficiency, effectively solving the problems of delayed fault handling and long impact time. At the same time, it enables targeted isolation and load transfer operations around the precisely located fault section, avoiding the need for handling the entire line, minimizing the power outage area, significantly improving the power supply reliability of the distribution network, and fully meeting the needs of automated and refined operation and maintenance of distribution networks.

[0033] Furthermore, as a refinement and extension of the specific implementation of the above embodiments, and to fully illustrate the implementation of this embodiment, this embodiment also provides another method for selecting segments in low-current grounding faults, such as... Figure 2 As shown, the method includes: Step 210: Collect the three-phase voltage data of the 10kV distribution network bus of the substation and set the voltage threshold of the three-phase voltage data.

[0034] Among them, the 10kV distribution network bus refers to the core electrical node in the substation of a 10kV voltage level distribution network, which is used to collect and distribute electrical energy. It is a key monitoring point for electrical data such as voltage and current of the distribution network and directly reflects the overall operating status of the distribution network. The three-phase voltage data refers to the voltage values ​​and related electrical characteristics of phases A, B, and C at the distribution network bus. It is the core electrical monitoring data for identifying grounding faults in the distribution network and can intuitively reflect the voltage operating status of the distribution network. The voltage threshold is a pre-set voltage critical value for determining whether the operating status of the distribution network is abnormal. It is an important basis for triggering the grounding fault assessment procedure. By comparing it with the actual monitored three-phase voltage data, the abnormal voltage state of the distribution network can be identified.

[0035] In the early monitoring stage of low-current grounding faults in the distribution network, the three-phase voltages of A, B, and C at the 10kV distribution network bus of the substation can be continuously collected in real time to obtain continuous and complete three-phase voltage data. At the same time, based on the rated operating voltage of the distribution network, corresponding voltage critical judgment criteria, i.e., voltage thresholds, are set for the three-phase voltage data to form the basis for judging the abnormal state of the three-phase voltage data, providing basic data and judgment criteria for subsequent grounding fault identification and analysis.

[0036] This technical step involves real-time acquisition of three-phase voltage data from the 10kV distribution network busbar of the substation, which can accurately and timely capture the voltage operating status of the distribution network, providing real and effective basic data support for the early identification of ground faults. By setting corresponding voltage thresholds for the three-phase voltage data, a quantitative judgment standard for distribution network voltage anomalies can be established, providing a clear and unified basis for subsequent ground fault triggering and judgment. This can effectively avoid the subjectivity and randomness of fault identification and improve the accuracy and standardization of initial ground fault identification.

[0037] Step 220: When an anomaly is detected in the three-phase voltage data that simultaneously exceeds the upper and lower limits based on the voltage threshold, the ground fault assessment procedure is triggered.

[0038] Among them, exceeding the upper limit means that the voltage values ​​of some phases in the monitored three-phase voltage data exceed the preset upper voltage threshold, indicating a high voltage operating state; exceeding the lower limit means that the voltage values ​​of some phases in the monitored three-phase voltage data are lower than the preset lower voltage threshold, indicating a low voltage operating state.

[0039] In this embodiment of the present disclosure, after the three-phase voltage data acquisition and voltage threshold setting are completed, the real-time monitored three-phase voltage data of the 10kV distribution network bus of the substation can be continuously compared and analyzed with the preset voltage threshold. When the monitoring finds that some phase voltages exceed the upper limit and some phase voltages exceed the lower limit in the three-phase voltage data at the same time, the system will automatically trigger the preset ground fault judgment program and start the subsequent specific analysis and type determination work for small current ground faults.

[0040] This technical step establishes an automated triggering mechanism for the ground fault assessment procedure by continuously comparing real-time monitored three-phase voltage data with voltage thresholds, replacing the traditional method of manual monitoring and judgment. This significantly improves the response speed of fault assessment. At the same time, using the simultaneous exceeding of the upper and lower limits of the three-phase voltage as the triggering condition can accurately capture the typical electrical characteristics of small-current ground faults, effectively distinguish ground faults from other power grid problems with single voltage anomalies, avoid false triggering of the assessment procedure, and improve the accuracy of fault identification.

[0041] Step 230: Analyze the voltage distortion of the three-phase voltage data through the ground fault assessment procedure, and determine whether the fault type of the small current grounding in the distribution network is a complete ground fault or an incomplete ground fault based on the voltage distortion.

[0042] Among them, voltage distortion refers to the degree to which the three-phase voltage of the distribution network bus deviates from the rated operating state. It is an important indicator reflecting the abnormal electrical operation of the distribution network and can intuitively reflect the voltage change characteristics after a small current grounding fault occurs. Complete grounding fault is a type of single-phase grounding fault in a small current grounding system of the distribution network. It is characterized by a significant drop in the voltage of the faulty phase to near zero, while the voltage of the non-faulty phase rises to near the line voltage, and the voltage distortion is significant. Incomplete grounding fault is another type of single-phase grounding fault in a small current grounding system of the distribution network. It is characterized by a significant decrease in the voltage of the faulty phase but not to zero, while the voltage of the non-faulty phase rises but is lower than the line voltage. The voltage distortion is less severe than that of a complete grounding fault.

[0043] In specific application scenarios, after the ground fault assessment program is triggered, it will comprehensively analyze the real-time collected three-phase voltage data of the 10kV distribution network bus of the substation, accurately capture the numerical changes and relative relationships of the three-phase voltage, and quantify the degree of voltage distortion of the distribution network bus. Then, based on the different characteristics of the degree of voltage distortion, it will accurately determine the type of small current ground fault that is currently occurring in the distribution network, clarify whether the fault is a complete ground fault or an incomplete ground fault, and provide accurate fault type basis for subsequent fault handling work.

[0044] This technical step analyzes the voltage distortion of three-phase voltage data through a ground fault assessment procedure and determines the fault type accordingly. It enables accurate identification of specific types of low-current ground faults, allowing subsequent equivalent circuit modeling, zero-sequence current acquisition, and other tasks to be carried out in a targeted manner around the specific fault type. This effectively avoids the judgment bias caused by adapting a single model to different fault types, improving the accuracy and adaptability of the overall fault segment selection. At the same time, the objective judgment method based on the degree of voltage distortion replaces subjective judgment, reducing the possibility of misjudgment in fault type determination. This lays a key foundation for accurate fault location and automated fault handling at the section level, ensuring the efficient and orderly progress of the entire fault segment selection and handling process.

[0045] Step 240: Divide the distribution network into several independent distribution network sections using the distribution switch as the dividing point. Based on the basic parameters of the distribution network, build an equivalent circuit model for each independent distribution network section that is adapted to the fault type. Collect the measured zero-sequence current data of each independent distribution network section based on the equivalent circuit model.

[0046] In the embodiments of this disclosure, when dividing the distribution network into several independent distribution network sections with the distribution switch as the dividing point, the primary and secondary integrated switches in the distribution network can be selected as the core distribution switches first. Then, combined with the actual line laying direction and overall electrical topology of the distribution network, two adjacent primary and secondary integrated switches are used as line boundaries. Based on this, the entire distribution network is divided into multiple independent distribution network sections. All the divided sections are electrically independent of each other, and each section can realize real-time monitoring of its electrical operating status.

[0047] Accordingly, step 240 of the embodiment may include: selecting a primary and secondary integrated switch in the distribution network as the distribution switch; and dividing the entire distribution network into several independent and real-time monitorable distribution network sections, using two adjacent distribution switches as boundaries, based on the actual line routing and electrical topology of the distribution network. Here, the primary and secondary integrated switch refers to a distribution network switch that integrates primary switching equipment and secondary monitoring and control units, possessing both line on / off control and real-time electrical quantity monitoring functions, and is the core equipment for distribution network section division and data acquisition; the electrical topology refers to the connection methods and layout relationships between various electrical devices and lines in the distribution network, intuitively reflecting the circuit structure and power supply path of the distribution network.

[0048] In the embodiments of this disclosure, when constructing equivalent circuit models adapted to fault types for each independent section of the distribution network based on the basic parameters of the distribution network, and collecting measured zero-sequence current data of each independent section of the distribution network based on the equivalent circuit models, the basic parameters of the distribution network, including line parameters and equipment parameters, can be extracted from the basic data of the distribution network ledger. The basic parameters of the distribution network are then input into the modeling system, which matches the actual electrical characteristics of the distribution network based on the parameters. Then, combined with the determined small current grounding fault type, a suitable equivalent circuit model is constructed for each independent section of the distribution network, so that the model can accurately match the current distribution characteristics of the corresponding fault type. Subsequently, the zero-sequence current acquisition node of each distribution switch is determined based on the completed equivalent circuit model. Based on the acquisition node, transient zero-sequence current data and steady-state zero-sequence current data of each independent section of the distribution network are collected to address the differences in current characteristics between complete grounding faults and incomplete grounding faults. Finally, the two types of zero-sequence current data are integrated to form complete measured zero-sequence current data for each independent section of the distribution network.

[0049] Accordingly, step 240 of the embodiment may include: extracting basic distribution network parameters based on the basic distribution network ledger data, the basic distribution network parameters including line parameters and equipment parameters; inputting the basic distribution network parameters into the modeling system, and using the modeling system to determine the actual electrical characteristics of the distribution network that match the basic distribution network parameters; based on the actual electrical characteristics of the distribution network and the fault type, building an equivalent circuit model that is compatible with each independent section of the distribution network, so that the equivalent circuit model matches the current distribution characteristics of the corresponding fault type; using the equivalent circuit model as the basis for data collection, determining the zero-sequence current collection node for each distribution switch; based on the zero-sequence current collection node, collecting transient zero-sequence current data and steady-state zero-sequence current data for each independent section of the distribution network according to the current characteristic differences of the fault type being a complete grounding fault or an incomplete grounding fault; integrating the collected transient zero-sequence current data and steady-state zero-sequence current data to form the measured zero-sequence current data for each independent section of the distribution network.

[0050] Among them, the basic data of the distribution network refers to the basic information recording the inherent attributes and parameters of the distribution network lines and equipment; line parameters refer to the parameters reflecting the electrical and physical characteristics of the distribution network lines, which are one of the core basic parameters for constructing the equivalent circuit model and determine the electrical operating characteristics of the lines; equipment parameters refer to the parameters reflecting the performance and specifications of various power equipment in the distribution network, which, together with the line parameters, constitute the basic parameters of the distribution network and support the equivalent circuit model to fit the actual operating state of the equipment; the modeling system refers to the dedicated system used to build and calculate the equivalent circuit model of the distribution network, which can match the actual electrical characteristics of the power grid based on the input basic parameters of the distribution network to achieve accurate construction of the circuit model; the actual electrical characteristics of the distribution network refer to the actual operating characteristics of the distribution network. The electrical operating patterns and characteristics exhibited during the process are the core attributes that the equivalent circuit model must conform to, ensuring the simulation realism of the model; the zero-sequence current acquisition node refers to the specific monitoring point set in the distribution network for acquiring zero-sequence current data. It is the physical location for obtaining measured zero-sequence current data and determines the targeting and accuracy of data acquisition; transient zero-sequence current data refers to the instantaneous change data of zero-sequence current in the early stage of a small current ground fault in the distribution network. It reflects the current characteristics in the early stage of the fault and is an important data dimension for fault judgment; steady-state zero-sequence current data refers to the numerical data of zero-sequence current tending to a stable state after a small current ground fault in the distribution network. It reflects the current characteristics in the stable stage of the fault and complements the transient data.

[0051] This technical process accurately extracts basic distribution network parameters from the basic data ledger of the distribution network and inputs them into the modeling system to match actual electrical characteristics. This ensures that the constructed equivalent circuit model has a basis that closely reflects the actual operating state of the distribution network. Combined with the fault type, suitable equivalent circuit models are built for each independent section of the distribution network. This enables the model to accurately simulate the current distribution characteristics of each section under different fault types, improving the simulation accuracy and fault adaptability of the model. Based on the equivalent circuit model, the zero-sequence current acquisition nodes are determined, making the data acquisition points more scientific. Transient and steady-state zero-sequence current data are collected in a targeted manner and integrated to form measured data. This allows the collected zero-sequence current data to comprehensively reflect the electrical characteristics of different stages of the fault, providing complete, accurate, and actual measured data support for subsequent fault matching of artificial intelligence models.

[0052] Step 250: Input the basic parameters of the distribution network, the fault type and the model features of the equivalent circuit model into the trained artificial intelligence model, output the zero-sequence current prediction distribution of each independent section of the distribution network, calculate the similarity between the zero-sequence current prediction distribution and the measured zero-sequence current data, and determine the fault section based on the similarity calculation results.

[0053] In the embodiments of this disclosure, when inputting the basic parameters of the distribution network, fault types, and model features of the equivalent circuit model into the trained artificial intelligence model, and outputting the zero-sequence current prediction distribution of each independent section of the distribution network, model features containing matching parameters between the model structure and the electrical characteristics of the distribution network can be extracted from the completed equivalent circuit model first. Then, the basic parameters of the distribution network, the determined small current grounding fault types of the distribution network, and the extracted equivalent circuit model features are integrated and used as unified input features to be input into the trained artificial intelligence model. Based on the various input feature information, the artificial intelligence model performs simulation calculations of small current grounding faults for each independent section of the distribution network, and finally outputs the zero-sequence current prediction distribution corresponding to each independent section of the distribution network when a fault occurs, providing data reference for subsequent fault section determination.

[0054] Accordingly, step 250 of the embodiment may include: extracting model features of the equivalent circuit model, the model features including the matching parameters of the equivalent circuit model structure and the electrical characteristics of the distribution network; inputting the basic parameters of the distribution network, the fault type and the model features as input features into the trained artificial intelligence model; performing fault simulation calculations on each independent section of the distribution network through the artificial intelligence model, and outputting the predicted distribution of zero-sequence current when a fault occurs in each independent section of the distribution network.

[0055] Among them, model features refer to the core attributes and key parameters of the equivalent circuit model, which are key information that reflects the core architecture and adaptability of the model. Specifically, they include the equivalent circuit model structure and distribution network electrical characteristic matching parameters; the equivalent circuit model structure refers to the overall topology layout and connection method of the equivalent circuit model, which is the basic framework for the model to simulate the electrical operation state of the distribution network and determines the degree to which the model restores the distribution network structure; the distribution network electrical characteristic matching parameters refer to the parameters that enable the equivalent circuit model to conform to the actual electrical operation law of the distribution network, which are the key indicators for the model to achieve accurate matching with the actual electrical characteristics of the distribution network; the zero-sequence current prediction distribution refers to the distribution state and numerical characteristics of the zero-sequence current when a small current grounding fault occurs in each independent section of the distribution network, as output by the artificial intelligence model after fault simulation calculation.

[0056] In the embodiments of this disclosure, when calculating the similarity between the predicted distribution of zero-sequence current and the measured data of zero-sequence current, and determining the fault section based on the similarity calculation results, a similarity algorithm can be used to quantify the degree of fit between the predicted distribution of zero-sequence current and the measured data of zero-sequence current for each independent distribution network section, so as to obtain a unique similarity value for each independent distribution network section. Then, by comparing and analyzing the similarity values ​​of all independent distribution network sections, the independent distribution network section with the highest similarity value is determined as the fault section where the small current grounding fault actually occurred.

[0057] Accordingly, step 250 of the embodiment may include: using a similarity algorithm to calculate the similarity value between the predicted distribution of zero-sequence current and the corresponding measured data of zero-sequence current for each independent section of the distribution network; and determining the independent section of the distribution network with the highest corresponding similarity value as the fault section.

[0058] Among them, similarity algorithm refers to the algorithm used to quantify the degree of matching between the distribution characteristics of two sets of data, such as cosine similarity algorithm, which can intuitively reflect the degree of matching between the zero-sequence current prediction distribution and the measured data through numerical calculation, and provide a quantitative standard for fault section judgment; similarity value refers to the specific value that is calculated by similarity algorithm to characterize the degree of matching between the zero-sequence current prediction distribution and the corresponding measured data. The value directly reflects the fault matching probability of independent sections of the distribution network.

[0059] Step 260: Generate switch operation instructions based on the faulty section. The switch operation instructions are used to indicate the completion of physical isolation of the faulty section and automatic load transfer.

[0060] In this embodiment of the present disclosure, after accurately determining the low-current grounding fault section, the specific location information of the fault section can be determined first. Then, based on the location information and the overall topology of the distribution network, a corresponding switch operation command is automatically generated. This command instructs the distribution switch to complete the switching operation in a step-by-step manner from the end of the fault section to the power supply side, thereby achieving physical isolation between the fault section and other healthy lines in the distribution network. At the same time, the command also instructs the system to search for and determine the alternative power supply path for the load of the fault section according to the distribution network topology. By closing the corresponding tie switch, the load originally powered by the fault section is smoothly transferred to the healthy independent section of the distribution network, thus completing the power supply restoration after the fault is handled.

[0061] Accordingly, step 260 in the embodiment may include: determining the location information of the faulty section; generating a corresponding switch operation instruction based on the location information, the switch operation instruction being used to instruct the distribution switch to be operated in the order of advancing from the end of the faulty section to the power supply side, to complete the physical isolation of the faulty section, and to find alternative power supply paths based on the distribution network topology, close the relevant tie switches, and transfer the load of the faulty section to an independent section of the healthy distribution network.

[0062] Location information refers to key information such as the specific location, line affiliation, and connection relationship with surrounding distribution switches and healthy sections of the distribution network where a small current grounding fault actually occurs; switch operation instructions refer to a set of instructions automatically generated based on the fault section location information to control the opening and closing of distribution switches in the distribution network, which can accurately guide the switching equipment to complete the relevant operations of fault isolation and load transfer; the end of the fault section refers to the line endpoint of the fault section far from the power supply side of the distribution network, which is the location of the distribution switch to be operated first when carrying out physical fault isolation; alternative power supply path refers to a healthy line path in the distribution network that can provide power supply to the load of the fault section in addition to the original power supply line of the fault section, which is automatically retrieved and determined by the system based on the distribution network topology; tie switch refers to the switching equipment in the distribution network used to connect different distribution network lines and realize power interconnection between lines, which can complete the transfer of load between different lines after closing; healthy independent distribution network section refers to an independent distribution network section in the distribution network that has not experienced a fault and is in normal electrical operation status, which is the receiving end of the load transfer of the fault section.

[0063] This technology first determines the location of the faulty section and then generates targeted switching operation instructions, giving the switching equipment precise target direction. This effectively avoids the risks to power grid operation caused by blind operation. By operating the switches in a sequence from the end of the faulty section towards the power source, physical isolation is achieved, which can quickly cut off the fault propagation path, prevent the fault range from expanding further, and ensure the normal operation of healthy lines in the distribution network. At the same time, based on the distribution network topology, alternative power supply paths are found and tie switches are closed to transfer the load, enabling rapid power restoration of the load in the faulty section, minimizing the power outage area and reducing the impact of the power outage. The entire process, through the automated generation of switching operation instructions and guidance to complete fault isolation and load transfer, replaces the traditional manual operation plan mode, which can significantly shorten the fault handling time, improve the automation and intelligence level of distribution network fault handling, and significantly enhance the power supply reliability and operational stability of the distribution network.

[0064] The precise fault segment selection method for low-current grounding faults based on artificial intelligence and equivalent circuits proposed in this application is applicable to 10kV low-current grounding distribution networks, such as... Figure 3 As shown, the entire process follows a technical route, sequentially carrying out the following steps: monitoring the 10kV bus voltage of the substation, initiating the ground fault analysis procedure, determining the fault type, building the equivalent circuit model of the distribution network, performing calculations on the large-scale artificial intelligence model, determining the fault section, and generating fault handling strategies. First, by monitoring the three-phase voltage of the 10kV bus in real time, the ground fault analysis procedure is initiated when the voltage exceeds the double limit conditions. Then, based on the degree of voltage distortion, two fault types are accurately determined: complete ground fault or incomplete ground fault. Next, an equivalent circuit model of the distribution network is built based on this fault type, and line parameters and equipment parameters are input into the model. The system collects core data such as data from various switches, and simultaneously gathers transient and steady-state zero-sequence current data. Then, it inputs line parameters, equivalent circuit models, and fault types into a large-scale artificial intelligence model for computation. The model predicts the current distribution trend of each fault section based on the zero-sequence current calculation model. It then calculates the cosine similarity between the predicted zero-sequence current distribution and the measured data, identifying the section with the highest similarity as the fault section. Finally, based on the judgment results, it generates a fault handling strategy following the principle of "from back to front, branch lines first then main lines, isolation first then transfer," thereby achieving precise segment selection and automated handling of small-current grounding faults. "From back to front" means operating the distribution switches sequentially from the end of the fault section furthest from the power source towards the side closest to the power source. "Branch lines first then main lines" means prioritizing isolation of distribution branch lines during fault handling, ensuring the power supply stability of the main lines, and preventing the fault from affecting the normal operation of the main lines. "Isolation first then transfer" means first completing the electrical and physical isolation between the fault section and the healthy lines of the distribution network to cut off the fault propagation path, and then carrying out load transfer operations.

[0065] In summary, the technical solution in this application triggers fault analysis and accurately determines the type of low-current grounding fault by real-time acquisition of electrical data from the distribution network bus. Independent sections of the distribution network are delineated using distribution switches as boundaries. Equivalent circuit models adapted to the fault types are built for each section, and corresponding zero-sequence current measurement data is collected. Then, the basic parameters of the distribution network, fault types, and equivalent circuit model features are input into the trained artificial intelligence model. The model outputs the predicted distribution of zero-sequence current in each section and performs similarity calculations with the measured data to accurately determine the faulty section. Finally, based on the faulty section, switch operation commands are generated to achieve physical isolation of the faulty section and automatic load transfer. This technical solution overcomes the limitations of existing low-current grounding fault detection technologies, which can only select lines but not precise sections. It enables precise section-level location of low-current grounding faults in distribution networks, reducing the patrol range of maintenance personnel to a single fault section. This significantly reduces manpower and time costs while improving patrol efficiency, effectively solving the problems of delayed fault handling and long impact time. At the same time, it enables targeted isolation and load transfer operations around the precisely located fault section, avoiding the need for handling the entire line, minimizing the power outage area, significantly improving the power supply reliability of the distribution network, and fully meeting the needs of automated and refined operation and maintenance of distribution networks.

[0066] Furthermore, as Figure 1 and Figure 2 The specific implementation of the method shown in this embodiment provides a small current grounding fault segmentation device, such as... Figure 4 As shown, the device includes: a determination module 41, an acquisition module 42, a calculation module 43, and a generation module 44.

[0067] The judgment module 41 can be used to collect electrical data of the distribution network bus in real time, trigger a ground fault judgment program based on the electrical data of the distribution network bus, and determine the fault type of small current grounding in the distribution network. The acquisition module 42 can be used to divide the distribution network into several independent distribution network sections with the distribution switch as the dividing point, build an equivalent circuit model for each independent distribution network section adapted to the fault type based on the basic parameters of the distribution network, and collect the measured zero-sequence current data of each independent distribution network section based on the equivalent circuit model. The calculation module 43 can be used to input the basic parameters of the distribution network, the fault type and the model features of the equivalent circuit model into the trained artificial intelligence model, output the zero-sequence current prediction distribution of each independent section of the distribution network, perform similarity calculation between the zero-sequence current prediction distribution and the measured zero-sequence current data, and determine the fault section based on the similarity calculation results. The generation module 44 can be used to generate switch operation instructions based on the fault section. The switch operation instructions are used to indicate the completion of physical isolation of the fault section and automatic load transfer.

[0068] In some embodiments of this application, the determination module 41 can be specifically used to collect three-phase voltage data of the 10kV distribution network bus of the substation and set voltage thresholds for the three-phase voltage data; when an anomaly of exceeding both the upper and lower limits of the three-phase voltage data is detected based on the voltage thresholds, a ground fault judgment procedure is triggered; the degree of voltage distortion of the three-phase voltage data is analyzed through the ground fault judgment procedure, and the fault type of the small current grounding in the distribution network is determined to be a complete ground fault or an incomplete ground fault based on the degree of voltage distortion.

[0069] In some embodiments of this application, when the power distribution network is divided into several independent power distribution network sections with the distribution switch as the dividing point, the acquisition module 42 can be specifically used to select the primary and secondary integrated switches in the power distribution network as the distribution switches; based on the actual line route and electrical topology of the power distribution network, with two adjacent distribution switches as the boundary, the entire power distribution network is divided into several independent power distribution network sections that are mutually independent and can be monitored in real time.

[0070] In some embodiments of this application, when constructing equivalent circuit models adapted to fault types for each independent section of the distribution network based on the distribution network basic parameters, and collecting measured zero-sequence current data for each independent section of the distribution network based on the equivalent circuit models, the acquisition module 42 can be specifically used to extract distribution network basic parameters based on the distribution network basic ledger data. The distribution network basic parameters include line parameters and equipment parameters; input the distribution network basic parameters into the modeling system, and use the modeling system to determine the actual electrical characteristics of the distribution network that match the distribution network basic parameters; based on the actual electrical characteristics of the distribution network and the fault type, for each Equivalent circuit models were constructed for each independent section of the distribution network to match the current distribution characteristics of the corresponding fault types. Based on the equivalent circuit models, zero-sequence current acquisition nodes for each distribution switch were determined. Based on the zero-sequence current acquisition nodes, transient and steady-state zero-sequence current data were collected for each independent section of the distribution network, taking into account the differences in current characteristics between complete grounding faults and incomplete grounding faults. The collected transient and steady-state zero-sequence current data were integrated to form the measured zero-sequence current data for each independent section of the distribution network.

[0071] In some embodiments of this application, when the basic parameters of the distribution network, the fault type, and the model features of the equivalent circuit model are input into the trained artificial intelligence model, and the zero-sequence current prediction distribution of each independent section of the distribution network is output, the calculation module 43 can be used to extract the model features of the equivalent circuit model, which include the matching parameters between the equivalent circuit model structure and the electrical characteristics of the distribution network; input the basic parameters of the distribution network, the fault type, and the model features as input features into the trained artificial intelligence model; perform fault simulation calculations on each independent section of the distribution network through the artificial intelligence model, and output the zero-sequence current prediction distribution corresponding to the fault occurrence in each independent section of the distribution network.

[0072] In some embodiments of this application, when calculating the similarity between the predicted distribution of zero-sequence current and the measured data of zero-sequence current, and determining the fault section based on the similarity calculation result, the calculation module 43 can be specifically used to calculate the similarity value between the predicted distribution of zero-sequence current and the corresponding measured data of zero-sequence current for each independent section of the distribution network using a similarity algorithm; and determine the independent section of the distribution network with the highest corresponding similarity value as the fault section.

[0073] In some embodiments of this application, the generation module 44 can be used to determine the location information of the faulty section; generate corresponding switch operation instructions based on the location information, the switch operation instructions are used to instruct the distribution switch to be operated in the order of advancing from the end of the faulty section to the power supply side, to complete the physical isolation of the faulty section, and to find alternative power supply paths based on the distribution network topology, close relevant tie switches, and transfer the load of the faulty section to an independent section of the healthy distribution network.

[0074] It should be noted that other corresponding descriptions of the functional units involved in the low-current grounding fault selection device provided in this embodiment can be found in [reference]. Figure 1 and Figure 2 The corresponding description in [the document] will not be repeated here.

[0075] Based on the above, Figure 1 and Figure 2 Accordingly, this embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the above-described method. Figure 1 and Figure 2 The method for selecting sections of a low-current grounding fault is shown.

[0076] Based on this understanding, the technical solution of this application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as CD-ROM, USB flash drive, mobile hard drive, etc.) and includes several instructions to cause an electronic device (such as personal computer, server, or network device, etc.) to execute the methods of various implementation scenarios of this application.

[0077] Based on the above, Figure 1 and Figure 2 The method shown, and Figure 4 To achieve the above objectives, the present application also provides an electronic device, specifically a personal computer, tablet computer, server, or other network device, as shown in the virtual device embodiment. This device includes a storage medium and a processor; the storage medium stores a computer program; the processor executes the computer program to achieve the above-described objectives. Figure 1 and Figure 2 The method for selecting sections of a low-current grounding fault is shown.

[0078] Optionally, the aforementioned physical devices may also include a user interface, a network interface, a camera, radio frequency (RF) circuitry, sensors, audio circuitry, a Wi-Fi module, etc. The user interface may include a display screen, input units such as a keyboard, etc., and optional user interfaces may also include USB interfaces, card reader interfaces, etc. The network interface may optionally include standard wired interfaces, wireless interfaces (such as Wi-Fi interfaces), etc.

[0079] Those skilled in the art will understand that the physical device structure provided in this embodiment does not constitute a limitation on the physical device, and may include more or fewer components, or combine certain components, or have different component arrangements.

[0080] The storage medium may also include an operating system and a network communication module. The operating system is a program that manages the hardware and software resources of the aforementioned physical device, supporting the operation of information processing programs and other software and / or programs. The network communication module is used to enable communication between the various components within the storage medium, as well as communication with other hardware and software in the information processing physical device.

[0081] Through the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary general-purpose hardware platform, or it can be implemented by hardware.

[0082] This invention, through real-time acquisition of electrical data from the distribution network bus, triggers fault analysis and accurately determines the type of low-current grounding fault. It divides the distribution network into independent sections using distribution switches as boundaries, builds equivalent circuit models for each section to adapt to the fault type, and collects corresponding zero-sequence current measured data. Then, it inputs the distribution network's basic parameters, fault type, and equivalent circuit model features into a trained artificial intelligence model, outputting the predicted distribution of zero-sequence current in each section and performing similarity calculations with the measured data to accurately determine the faulty section. Finally, it generates switch operation commands based on the faulty section to achieve physical isolation of the faulty section and automatic load transfer. This technical solution overcomes the limitations of existing low-current grounding fault detection technologies, which can only select lines but not precise sections. It enables precise section-level location of low-current grounding faults in distribution networks, reducing the patrol range of maintenance personnel to a single fault section. This significantly reduces manpower and time costs while improving patrol efficiency, effectively solving the problems of delayed fault handling and long impact time. At the same time, it enables targeted isolation and load transfer operations around the precisely located fault section, avoiding the need for handling the entire line, minimizing the power outage area, significantly improving the power supply reliability of the distribution network, and fully meeting the needs of automated and refined operation and maintenance of distribution networks.

[0083] Those skilled in the art will understand that the accompanying drawings are merely schematic diagrams of a preferred embodiment, and the modules or processes shown in the drawings are not necessarily essential for implementing this application. Those skilled in the art will understand that the modules in the apparatus of the embodiment can be distributed within the apparatus of the embodiment as described, or can be modified to be located in one or more apparatuses different from this embodiment. The modules of the above-described embodiment can be combined into one module, or further divided into multiple sub-modules.

[0084] The serial numbers in this application are for descriptive purposes only and do not represent the superiority or inferiority of any particular implementation scenario. The above disclosures are merely a few specific implementation scenarios of this application; however, this application is not limited thereto, and any variations conceived by those skilled in the art should fall within the protection scope of this application.

Claims

1. A method for selecting a section of a low-current grounding fault, characterized in that, include: Real-time acquisition of electrical data of the distribution network busbars; based on the electrical data of the distribution network busbars, triggering a ground fault judgment program to determine the fault type of low current grounding in the distribution network; The distribution network is divided into several independent distribution network sections by using the distribution switch as the dividing point. An equivalent circuit model adapted to the fault type is built for each of the independent distribution network sections based on the basic parameters of the distribution network. The zero-sequence current measured data of each of the independent distribution network sections is collected based on the equivalent circuit model. The basic parameters of the distribution network, the fault type, and the model features of the equivalent circuit model are input into the trained artificial intelligence model, and the zero-sequence current prediction distribution of each independent section of the distribution network is output. The similarity between the zero-sequence current prediction distribution and the measured zero-sequence current data is calculated, and the fault section is determined based on the similarity calculation results. Based on the faulty section, a switch operation command is generated. The switch operation command is used to instruct the physical isolation of the faulty section and automatically realize the load transfer.

2. The method according to claim 1, characterized in that, The real-time acquisition of distribution network bus electrical data, based on which a ground fault analysis program is triggered, determines the type of low-current grounding fault in the distribution network, including: Collect three-phase voltage data of the 10kV distribution network bus of the substation, and set the voltage threshold of the three-phase voltage data; When an anomaly is detected in the three-phase voltage data simultaneously exceeding the upper and lower limits based on the voltage threshold, the ground fault assessment procedure is triggered. The grounding fault assessment procedure analyzes the voltage distortion of the three-phase voltage data and determines the fault type of small current grounding in the distribution network as either a complete grounding fault or an incomplete grounding fault based on the voltage distortion.

3. The method according to claim 1, characterized in that, The division of the distribution network into several independent distribution network sections using the distribution switch as the dividing point includes: Select the primary and secondary integrated switch in the power distribution network as the power distribution switch; Based on the actual route and electrical topology of the distribution network, the entire distribution network is divided into several independent and real-time monitorable distribution network sections, with two adjacent distribution switches as boundaries.

4. The method according to claim 1, characterized in that, Based on the basic parameters of the distribution network, an equivalent circuit model adapted to the fault type is constructed for each independent section of the distribution network. Based on the equivalent circuit model, measured zero-sequence current data for each independent section of the distribution network is collected, including: Basic parameters of the distribution network are extracted from the basic ledger data of the distribution network, which include line parameters and equipment parameters. The basic parameters of the distribution network are input into the modeling system, and the modeling system is used to determine the actual electrical characteristics of the distribution network that match the basic parameters of the distribution network. Based on the actual electrical characteristics of the distribution network and the fault type, an equivalent circuit model is built for each independent section of the distribution network to match the current distribution characteristics of the corresponding fault type. Based on the equivalent circuit model, the zero-sequence current acquisition nodes of each power distribution switch are determined; Based on the zero-sequence current acquisition node, and considering the current characteristic differences between the fault types of complete grounding fault and incomplete grounding fault, transient zero-sequence current data and steady-state zero-sequence current data of each independent section of the distribution network are collected. The collected transient zero-sequence current data and the steady-state zero-sequence current data are integrated to form the measured zero-sequence current data of each independent section of the distribution network.

5. The method according to claim 1, characterized in that, The step of inputting the basic parameters of the distribution network, the fault type, and the model features of the equivalent circuit model into the trained artificial intelligence model, and outputting the zero-sequence current prediction distribution of each independent section of the distribution network, includes: Extract the model features of the equivalent circuit model, which include matching parameters between the equivalent circuit model structure and the electrical characteristics of the distribution network; The basic parameters of the distribution network, the fault type, and the model features are used as input features and input into the trained artificial intelligence model. The artificial intelligence model is used to perform fault simulation calculations on each of the independent sections of the distribution network, and the predicted distribution of zero-sequence current when a fault occurs in each of the independent sections of the distribution network is output.

6. The method according to claim 1, characterized in that, The similarity calculation is performed between the predicted distribution of the zero-sequence current and the measured data of the zero-sequence current. Based on the similarity calculation results, the fault segment is determined, including: A similarity algorithm is used to calculate the similarity value between the predicted distribution of zero-sequence current and the corresponding measured data of zero-sequence current for each independent section of the distribution network. The independent distribution network segment with the highest similarity value is identified as the fault segment.

7. The method according to claim 1, characterized in that, Based on the faulty section, a switching operation command is generated. This switching operation command instructs the completion of physical isolation of the faulty section and automatic load transfer, including: Determine the location information of the faulty section; Based on the location information, a corresponding switch operation instruction is generated. The switch operation instruction is used to instruct the distribution switch to be operated in the order of advancing from the end of the fault section to the power supply side, to complete the physical isolation of the fault section, and to find alternative power supply paths based on the distribution network topology, close relevant tie switches, and transfer the load of the fault section to an independent section of the healthy distribution network.

8. A low-current grounding fault segment selection device, characterized in that, include: The judgment module is used to collect electrical data of the distribution network bus in real time, and trigger a ground fault judgment program based on the electrical data of the distribution network bus to determine the fault type of small current grounding in the distribution network. The data acquisition module is used to divide the distribution network into several independent distribution network sections with the distribution switch as the dividing point, build an equivalent circuit model adapted to the fault type for each of the independent distribution network sections in combination with the basic parameters of the distribution network, and collect the measured zero-sequence current data of each of the independent distribution network sections based on the equivalent circuit model. The calculation module is used to input the basic parameters of the distribution network, the fault type, and the model features of the equivalent circuit model into the trained artificial intelligence model, output the zero-sequence current prediction distribution of each independent section of the distribution network, perform similarity calculation between the zero-sequence current prediction distribution and the measured zero-sequence current data, and determine the fault section based on the similarity calculation results. The generation module is used to generate switch operation instructions based on the faulty section. The switch operation instructions are used to instruct the completion of physical isolation of the faulty section and automatic load transfer.

9. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method of any one of claims 1 to 7.

10. An electronic device comprising a storage medium, a processor, and a computer program stored on the storage medium and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method of any one of claims 1 to 7.

Images