A rapid fault location method for power distribution feeders based on 5G communication slicing

Abstract

This invention discloses a rapid fault location method for distribution feeders based on 5G communication slices, specifically relating to the field of distribution network fault detection and location technology. It addresses the problems of existing distribution feeder fault location methods, which fail to incorporate segment-level topology, unified timing, and communication quality characteristics, leading to inaccurate fault segment identification, excessively large isolation ranges, and prolonged power outages for non-faulty users. By constructing a segment-level topology model at the distribution master station, using a unified timing reference, and combining 5G communication slices for fault services, local operation slices, and a fault evidence fingerprint matrix, the method performs constraint matching with segment-level fault theoretical event patterns and solves for the minimum set of fault segments and the minimum set of disconnecting switches. This allows for rapid and accurate location of fault segments and control of isolation ranges even under fluctuating communication conditions or local measurement deficiencies, shortening fault handling time and reducing power outages for non-faulty users.

Description

Technical Field

[0001] This invention relates to the field of power distribution network fault detection and location technology, specifically a method for rapid fault location of power distribution feeders based on 5G communication slicing. Background Technology

[0002] Currently, fault location in distribution networks mainly relies on protection device operation information, switch position information, and a small amount of voltage and current measurement data. At the substation side, fault sections are determined through topological reasoning or simple rule-based logic. Some systems employ single-ended or double-ended ranging algorithms, combining line parameters with measurements before and after the fault to estimate the fault distance. Others utilize tripping and closing signals from feeder automation devices (FTU, DTU) to locate sections based on "nearest tripping point + upstream and downstream voltage loss range." These solutions can achieve basic fault section identification when the feeder structure is simple and communication conditions are relatively stable. However, they typically only describe the network structure at the switch or station level, lacking sufficient fine-grained segmentation of feeders. Furthermore, the time references between different terminals often rely on their respective local clocks or coarse-grained time synchronization mechanisms, resulting in limited alignment accuracy of measurements and operation information before and after the fault on the time axis.

[0003] As urban power distribution networks become increasingly complex, a single feeder often simultaneously carries industrial, residential, and public service loads, incorporating multi-level ring networks and backup power supply paths. While existing technologies are beginning to utilize private wireless networks or 5G networks to transmit data from distribution terminals, they often treat the communication network as a transparent channel, focusing primarily on message delivery. They rarely incorporate communication quality indicators such as end-to-end transmission delay, packet loss, and retransmission behavior into the unified fault location model. Furthermore, they lack practices for segment-level refinement and aggregation of measurement waveforms, protection actions, and switch states based on a unified timing reference. Consequently, in actual operation, when individual terminals experience inaccurate timing, message delays, or loss, the master station still infers faults based on ideal synchronization and reliable communication assumptions. This easily leads to discrepancies between the identified faulty section and the actual fault location, requiring a larger isolation range to ensure fault isolation, resulting in excessive power outages for non-faulty users.

[0004] In summary, among the existing distribution network fault location technologies, there is a lack of a segment-level fault location method that is based on a unified segment-level topology and a unified timing reference, can simultaneously consider voltage and current characteristics, protection actions and switch states, and incorporate communication quality differences into the same evidence system for reasoning. This makes it difficult to locate faulty segments in a timely and accurate manner while also taking into account the practical operational need to reduce power outages for non-faulty users. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a rapid fault location method for power distribution feeders based on 5G communication slicing, in order to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for rapid fault location of power distribution feeders based on 5G communication slicing, comprising:

[0007] S1. Construct a segment-level topology model at the power distribution master station and bind segment identifiers and terminal identifiers, and configure a unified timing reference and 5G communication slice parameters for fault services based on the segment-level topology.

[0008] S2. Set waveform buffers and status buffers in each segment terminal device. When the local fault criterion is triggered, extract the data within the fault time window according to the unified time synchronization and generate a running slice with segment identifier and time window identifier.

[0009] S3. Each segment terminal device sends a running slice message through the fault service 5G communication slice. The message carries voltage and current characteristic values, protection actions, switch status and slice delay packet loss indicators.

[0010] S4. The power distribution master station aggregates operation slices according to the fault time window and constructs a fault evidence fingerprint matrix with segments as rows and measurement characteristics, action status, and communication quality indicators as columns.

[0011] S5. Based on the topology model and protection setting, the distribution master station derives the section-level fault theoretical event patterns, performs constraint matching between the fault evidence fingerprint matrix and each theoretical event pattern, calculates the matching score, and obtains candidate fault sections.

[0012] S6. The power distribution master station solves the minimum fault section set and minimum disconnect switch set on the candidate fault section according to the evidence constraints, generates the fault isolation operation sequence, and solidifies the operation slice, fault evidence fingerprint matrix, matching results, and isolation results to form a replayable fault handling evidence chain.

[0013] Furthermore, S1 includes:

[0014] The power distribution master station constructs a segment-level topology model based on switches and cable joints, and maps segment identifiers to terminal identifiers;

[0015] Time references are sent to terminals based on a unified time synchronization benchmark;

[0016] Configure 5G communication slice parameters for faulty services to carry out slice operation messages;

[0017] When using a dedicated power communication network to carry operation slice messages, the 5G communication slice parameters of faulty services are mapped to the service level within the dedicated power communication network, and the operation slice is uploaded and the fault evidence fingerprint matrix is ​​constructed under the constraints of a unified timing reference and 5G communication slice parameters of faulty services.

[0018] Furthermore, S2 includes:

[0019] Set waveform buffers and status buffers in the terminal device corresponding to the segment;

[0020] The terminal device calculates the effective value of current, effective value of voltage, and zero-sequence component based on the current and voltage sample values ​​and executes local fault criteria.

[0021] When a local fault criterion is triggered, a fault time window identifier is generated based on a unified timing reference. The data corresponding to the fault time window in the waveform cache and status cache are organized into running slices and appended with segment identifiers, terminal identifiers and fault time window identifiers.

[0022] Furthermore, S3 includes:

[0023] Within each terminal device, the operational slices organized around the fault time window are encapsulated into operational slice messages;

[0024] The operation slice message includes segment identifier, terminal identifier, fault time window identifier, voltage and current characteristic values, protection action flag set, switch status change record, slice delay packet loss index and message version number, and is sent to the distribution master station through fault service 5G communication slice.

[0025] Furthermore, the power distribution master station is equipped with an interface for receiving operation slice messages;

[0026] The running slice message receiving interface parses the segment identifier, terminal identifier, fault time window identifier, voltage and current characteristic value set, protection action flag set, switch status change record, slice delay packet loss index and message version number in the running slice message, and returns the reception result with error code field. The terminal device adjusts the local time synchronization configuration, running slice message encapsulation method, field padding and retransmission strategy according to the error code field.

[0027] Furthermore, S4 includes:

[0028] The power distribution master station merges the operation slice messages according to the fault time window identifier and aligns the message time information under a unified time synchronization reference.

[0029] Segment identifiers are used to summarize voltage and current characteristic values, protection action flags, switch status change records, and slice delay packet loss indicators to construct a fault evidence fingerprint matrix with segments as rows and measurement characteristics, action status, and communication quality indicators as columns.

[0030] The communication quality index is mapped from the slice latency and packet loss index to the communication quality weight. The communication quality weight is divided into high confidence level, medium confidence level and low confidence level, and the matrix version number and timestamp are written to the read-only log.

[0031] Furthermore, S5 includes:

[0032] The distribution master station constructs a theoretical event pattern of section-level faults based on the section-level topology model and protection setting parameters, indexed by section identifier and fault type;

[0033] Based on the communication quality weight, the fault evidence fingerprint matrix units are divided into high-confidence entries, medium-confidence entries, and low-confidence entries;

[0034] When traversing each segment, the distribution master station performs constraint matching between the fault evidence fingerprint of the segment and the corresponding segment-level fault theoretical event mode. The matching score is determined based on the proportion of units in the high-confidence entries that are consistent with the theoretical expectation. When the matching score reaches the matching score threshold, the segment is marked as a candidate fault segment.

[0035] Furthermore, S6 includes:

[0036] The power distribution master station applies evidence constraints and operational constraints to the set of candidate fault sections;

[0037] Evidence constraints require that the combination of faulty sections covers all high-confidence entries in the fault evidence fingerprint matrix and is consistent with the electrical connection relationships recorded in the segment-level topology model;

[0038] The operational constraints are based on a static parameter library and an operational rule library. The allowable current of the line, the allowable current of the backup line, and the closing and opening capacity of the switch are checked to determine the minimum set of fault sections and the minimum set of disconnecting switches. The target combination is selected based on the number of users experiencing power outages, in conjunction with the user ledger.

[0039] Furthermore, after determining the minimum fault section set and the minimum disconnect switch set, the power distribution master station generates a fault isolation operation sequence based on the current switch status. Each step in the fault isolation operation sequence includes the target switch identifier, target status, and sequence number.

[0040] The power distribution master station executes opening and closing commands through the control interface and according to the error code feedback control terminal device. Based on the recorded command sequence number and execution result, it generates a replayable fault handling evidence chain indexed by the fault event identifier and evidence chain version number.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] 1. By constructing a segment-level topology model at the power distribution master station, unifying the timing reference, and combining fault service 5G communication slices, local operation slices, and fault evidence fingerprint matrix, constraint matching is performed with the segment-level fault theoretical event mode to solve the minimum fault segment set and the minimum disconnect switch set. Thus, even under fluctuating communication conditions or even local measurement loss, the fault segment can still be quickly and accurately located and the isolation range controlled, shortening the fault handling time and reducing power outages for non-faulty users.

[0043] 2. By implementing version locking and error code and idempotent control on segment identifiers, terminal identifiers, timing references, communication slice configurations, fault evidence fingerprint matrices, theoretical event mode versions, and fault isolation operation sequences throughout the entire process, and generating a replayable fault handling evidence chain, the fault location decision-making and switching operation process can be made traceable, verifiable, and reusable, thereby improving the operational reliability and maintenance management level of the distribution automation system in different distribution network scenarios. Attached Figure Description

[0044] Figure 1 This is a flowchart illustrating the rapid fault location method for power distribution feeders based on 5G communication slicing according to the present invention. Detailed Implementation

[0045] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0046] Example: Figure 1 A flowchart illustrating the rapid fault location method for distribution feeders based on 5G communication slicing according to the present invention is provided. The rapid fault location method for distribution feeders based on 5G communication slicing includes:

[0047] S1. Construct a segment-level topology model at the distribution master station and bind segment identifiers and terminal identifiers. Configure a unified timing reference and 5G communication slice parameters for fault services based on the segment-level topology. The specific implementation is as follows:

[0048] In a real urban power distribution network, a medium-voltage urban power distribution feeder is taken as the application object, and the power distribution master station is used as the centralized computing and coordination unit. The feeder is modeled and its communication and timing environment is configured. An urban power distribution feeder refers to a collection of lines that supply power to multiple loads from the substation busbar through multiple switches and cables. In actual operation, it often simultaneously bears industrial loads, residential loads, and public service loads.

[0049] To facilitate subsequent fault location and communication dispatch, the distribution master station first divides the feeder according to the primary switch, cable joint, and branch point. The conductor intervals that are physically continuous and topologically located in the same switch section are defined as a segment. Each segment is assigned a unique segment identifier in the distribution master station. The segment identifier can be a code composed of line number, interval number, and sequence number, which is used to uniquely indicate the corresponding physical segment within the master station.

[0050] Meanwhile, the feeder terminals, circuit breaker control devices, and intelligent switch control units already installed on the feeder are collectively referred to as terminal devices. Each terminal device is assigned a terminal identifier, which corresponds one-to-one with the actual installation location, access voltage level, and the switchgear it manages. The distribution master station associates each section with at least one terminal device through the correspondence between the maintenance section identifier and the terminal identifier.

[0051] The segment-level topology model is a summary description of the segment identifiers and their electrical connections within the distribution master station. It includes at least the connection methods between segments via switches, cable joints, and busbars, the power supply direction, and the connection relationship with the superior substation. In engineering practice, the distribution master station can review the primary wiring diagram, geographic information system data, and equipment ledgers in the existing dispatch automation system, and combine the results of on-site verification to confirm the switch ownership, line route, and load access point for each segment. The confirmed topology information is then solidified into a segment-level topology model for subsequent fault evidence reasoning and communication resource orchestration.

[0052] To ensure accurate comparison of voltage and current sampling values, protection action status, and switch opening / closing status from various terminal devices on the time axis, the distribution master station configures a unified time reference for the feeder and the terminal devices associated with it. The unified time reference refers to a time reference signal generated and distributed by the same time source, used to constrain the time deviation of the internal clocks of each device. This time source can be obtained through satellite time synchronization equipment, a dedicated power system time synchronization network, or other devices that can provide stable time signals. The terminal devices periodically calibrate their local clocks through local time synchronization modules, so that the fault occurrence time, protection action time, and waveform sampling time recorded by the terminal devices remain comparable within the maintained time error limits. Preferably, in the medium-voltage distribution network scenario, the synchronization accuracy can be set to be significantly less than the time order of one power frequency cycle, so as to distinguish the response sequence of each segment within the fault time window composed of several power frequency cycles.

[0053] To address fault-related communication needs, the distribution master station configures fault service 5G communication slice parameters in the configuration unit that interfaces with the 5G network operator. These fault service 5G communication slice parameters refer to a set of service rules used in the 5G network to carry fault-related messages. These include at least an end-to-end transmission latency limit, an allowable packet loss ratio limit, a single feeder parallel message limit, and a service priority level. These parameters ensure that slice messages running within the fault time window can be delivered to the distribution master station with deterministic service quality. In the aforementioned feeder scenario, the fault service 5G communication slice parameters can be set to a latency limit of 50 milliseconds, a packet loss rate limit of 0.1%, and a limit of dozens of parallel slice messages per feeder. Segments with a wide power supply range and concentrated important users in the segment-level topology model are given higher priority. Key segments can be those simultaneously handling hospital, rail transit, power dispatching, communication equipment room, or other public service loads, or segments with significantly higher power outage losses than ordinary user segments in historical fault statistics. These segments are marked as key segments in the slice configuration to prioritize the transmission quality of their running slice messages during large-scale faults.

[0054] The distribution master station records the version numbers of the segment identifier, terminal identifier, and fault service 5G communication slice parameters in a read-only log. The read-only log is an internal recording area of ​​the distribution master station used to record key configuration changes that can only be added and not modified. Each time the segment-level topology model or fault service 5G communication slice parameters are modified, a new version number is generated and written to the log. At the same time, the corresponding segment-level topology model and slice configuration are stored in the local storage system with the version number as the index. This allows each subsequent fault event to be traced back to the segment division, device mapping, and communication rules that were in effect at that time by the correspondence between the event occurrence time and the version number, which facilitates the analysis and evaluation of the historical fault handling process.

[0055] When some substations do not yet have the conditions for 5G network access, dedicated power fiber optic communication networks, multi-protocol label switching bearer networks, or other dedicated communication networks built and managed by power companies can be used as the bearer method. As long as the bearer method can achieve the same constraint capability as the aforementioned fault service 5G communication slice in terms of transmission delay control, packet loss rate control, and service priority control through network planning and equipment configuration, it can be regarded as having equivalent delay and reliability constraint capability. The equivalent constraint capability preferably refers to the fact that, under typical power distribution fault scenarios, the actual transmission delay and packet loss of the slice message in the dedicated communication network are not statistically inferior to the upper limit of delay and upper limit of packet loss rate specified in the fault service 5G communication slice parameters.

[0056] In this scenario, the construction method of the segment-level topology model, the association method of segment identifiers and terminal identifiers, and the distribution method of the unified timing reference remain consistent. Only the 5G communication slice parameters of the fault service are mapped to the service level, queue scheduling strategy, and bandwidth reservation strategy within the power dedicated communication network. During project implementation, the distribution master station still records the version number in the above manner and locks the corresponding configuration in the read-only log, thereby providing a consistent time and resource foundation for subsequent operation slice uploading and fault evidence fingerprint construction under different communication bearer modes.

[0057] Preferably, when implementing the above configuration on an urban power distribution feeder with a rated voltage of 10 kV, a total length of approximately 20 kilometers, and divided into 20 sections, the section-level topology model of the entire feeder can be established on the power distribution master station side. The synchronization accuracy of the unified time reference is set to better than one millisecond, the end-to-end latency limit of the fault service 5G communication slice is set to tens of milliseconds, the packet loss rate limit is set to thousands, and the number of running slice messages received in parallel by a single feeder is set to dozens. In the section-level topology model, the sections that bear the load of public services such as hospitals and rail transit stations are marked as key sections. In the fault service 5G communication slice parameters, the terminals corresponding to the key sections are given higher service priority. Based on meeting the above configuration conditions, subsequent steps can be carried out on the feeder according to the unified time reference and reserved communication resources to perform running slice interception, fault evidence fingerprint matrix construction, and fault section inference. Engineering technicians can arrange on-site wiring verification, terminal device identification planning, time device installation, and communication slice application based on this, so that this step has directly implementable process capabilities and a clear scope of application.

[0058] S2. Set waveform buffers and status buffers in each segment terminal device. When the local fault criterion is triggered, extract data within the fault time window according to the unified time synchronization to generate a running slice with segment identifier and time window identifier. The specific implementation is as follows:

[0059] After completing the configuration based on the aforementioned segment-level topology model and unified timing reference, waveform buffers and status buffers are set up inside the terminal device corresponding to each segment to save the operating status of the segment, so as to locally identify suspected faults and capture the operating records around the fault time window under actual power distribution conditions.

[0060] Waveform buffer refers to the storage area inside the terminal device used to continuously retain current and voltage sample values ​​according to a unified timing reference. It can write the sampling points marked with the timing reference in sequence according to the preset sampling rhythm and overwrite the earliest sample value in time order when the capacity is full. Status buffer refers to the storage area inside the terminal device used to record the operation status and switch opening / closing status of the protection device associated with this terminal, as well as the corresponding time mark. It can register the operation type, the status before and after the change, and the time of change when the protection device operates or the switch status changes.

[0061] The current and voltage sampling values ​​are numerical sequences obtained by analog-to-digital converters from the outputs of current transformers and sensors connected to the terminal device, following the changes in the power grid waveform. Each sampling point is accompanied by a time stamp derived from a unified time reference, which is used to accurately locate the sampling point in the power frequency cycle. The operation status of the protection device refers to whether the overcurrent protection, grounding protection, and other protection units configured on the terminal device have issued an operation command at a certain moment. The switch opening and closing status refers to the indication of whether the circuit breaker, load switch, and sectionalizing switch managed by the terminal device are in the closed or open state at a certain moment.

[0062] In order to obtain continuous and complete waveform segments within subsequent fault time windows, the capacity of the waveform buffer can be set to cover sampling points of several complete power frequency cycles, and the capacity of the status buffer can be set to cover all protection actions and switch operation records that may occur within a fault location cycle. The fault location cycle refers to the time interval from the triggering of the local fault criterion to the completion of a fault detection, operation slice upload and main station fault location processing, which is used to limit the range of protection actions and switch operation records that need to be retained.

[0063] The start and end times of the fault time window are defined in the terminal device according to the unified time reference. The fault time window is a duration interval centered on the time when a suspected fault is detected locally. It is used to focus on observing sudden changes in voltage and current waveforms and operating states within this interval.

[0064] In medium-voltage distribution network scenarios, the fault time window can be set as a combination of several preceding and several following power frequency cycles, with a unified time reference timestamp representing the specific start and end boundaries. To enable terminal devices to independently determine whether a suspected fault has occurred during field operation, a local fault criterion is set within each terminal device. The local fault criterion judges based on physical quantities such as the amplitude changes and zero-sequence component changes of the collected current and voltage. For example, if, within several consecutive power frequency cycles, the effective value of the current shows a significant increase compared to the previous stable operation phase, the effective value of the voltage shows a significant decrease, or the zero-sequence current component changes from near zero to significantly deviating from zero, and the duration of the above changes exceeds a preset number of cycles, this situation is considered a suspected fault.

[0065] The effective current value used in local fault criteria refers to the value representing the magnitude of the current in one or more complete power frequency cycles, calculated based on the current sampling value. The effective voltage value is the value representing the magnitude of the voltage, calculated based on the voltage sampling value in a similar cycle. The phase change refers to the quantitative description of the change in the phase angle of voltage or current between adjacent cycles or before and after a fault, obtained by comparing the phase position difference of the same phase in the stable operation phase before the fault with that in the fault time window.

[0066] In the terminal device, the local fault criterion is used to make a judgment at the end of each power frequency cycle using the current RMS value, voltage RMS value and zero sequence component change of the most recent several cycles. When several consecutive judgments show that the change exceeds the set threshold, a suspected fault flag is triggered to avoid false judgments triggered by a single sampling noise or brief disturbance.

[0067] After triggering the local fault criterion, the terminal device reads the current time value corresponding to the unified time reference, and copies out the waveform cache content and status cache content centered on this time and covering the start and end boundaries of the predefined fault time window to form a local record of the terminal within the fault time window.

[0068] To maintain the correspondence with the segment-level topology model in the distribution master station, when organizing the above local records, the terminal device will attach the segment identifier and terminal identifier issued during the configuration phase of the distribution master station to the local record. At the same time, it will calculate a unique fault time window identifier based on the start and end times of the fault time window. The fault time window identifier can be formed by combining a unified time synchronization reference time and a preset window length, which is used to distinguish different fault events and different time windows during the subsequent aggregation process on the master station side.

[0069] The terminal device organizes this set of local records, including segment identifier, terminal identifier, fault time window identifier, current effective value sequence, voltage effective value sequence, phase change sequence, protection action flag, and switch status change record, into an operation slice. The operation slice is a local operation segment organized around the same fault time window, which centrally describes the measurement changes and actions of the segment in several power frequency cycles before and after the fault.

[0070] To reduce the impact of intermittent noise, short-term switching jitter, and communication jitter on the content of the operational slice, the terminal device can first fine-tune the sampling time axis according to a unified timing reference when copying data from the waveform buffer and state buffer. This aligns the current and voltage sampling values ​​to the boundary of the complete power frequency cycle. Isolated sampling values ​​that are significantly outside the measurement range are discarded, and switching state changes that are instantaneous and immediately recover are filtered out. Continuous changes in the same direction are treated as valid state change records and retained.

[0071] In the practical application of the aforementioned 10 kV feeder, the waveform buffer capacity in each corresponding terminal device can be set to a sampling point of no less than six power frequency cycles, the status buffer capacity can be set to be able to store all protection actions and switching operation records within one fault location cycle, the fault time window can be set to a combination of three power frequency cycles before and three power frequency cycles after the time when a suspected fault is detected locally, the thresholds for changes in current effective value, voltage effective value, and zero-sequence component in the local fault criterion can be set to representative thresholds obtained from long-term operation data of the feeder, and the number of continuous cycles can be set to trigger a suspected fault only when the above threshold conditions are met for at least two consecutive power frequency cycles. This can cover the obvious waveform changes caused by common faults such as single-phase grounding and phase-to-phase short circuits, and can also avoid misjudgment of short-term load fluctuations or non-faulty operations.

[0072] Under the above configuration, the operational slicing in this embodiment covers all twenty sections of the 10 kV feeder. When the terminal device corresponding to each section detects a suspected fault locally, it can extract the same length of fault time window according to the unified timing reference and organize it into an operational slice with a consistent structure. When implemented on feeders of other voltage levels or with different configurations, it is only necessary to reset the sampling rhythm, waveform buffer capacity and fault time window length according to the voltage level of the feeder, the expected fault duration and the sampling capability of the terminal, and adjust the thresholds of the current RMS value, voltage RMS value and zero-sequence component in the local fault criteria. This method can be extended to different operating conditions without changing the meaning of operational slices. This makes the process directly implementable and has a clear scope of application in different distribution network scenarios. Distribution automation engineers can use this to complete the terminal device software configuration and field parameter tuning.

[0073] S3. Each segment terminal device sends an operation slice message through the fault service 5G communication slice. The message carries voltage and current characteristic values, protection actions, switch status, and slice delay and packet loss indicators. The specific implementation is as follows:

[0074] After completing the aforementioned operational slice organization, the terminal devices corresponding to each segment need to encapsulate the locally generated operational slices into operational slice messages that can be transmitted in the communication network using a unified format, and send them to the distribution master station through the configured fault service 5G communication slice. An operational slice message is a message unit that uses the operational slice as its main content, along with necessary identification fields and transmission quality record fields, and is used to centrally transmit the measurement characteristics and operational status of this segment to the distribution master station within the fault time window.

[0075] To ensure that the distribution master station can accurately correspond the operation slices with the segment-level topology model, the operation slice message must contain at least the segment identifier, terminal identifier, and fault time window identifier. The segment identifier and terminal identifier are consistent with the codes agreed upon in the aforementioned segment-level topology model. The fault time window identifier is obtained by combining the start and end times of the fault time window with the window number, and is used to distinguish different fault events and different time slices.

[0076] In terms of measurement content, the operation slice message includes voltage and current characteristic values ​​and action status information. The voltage and current characteristic values ​​are a set of representative physical quantities calculated based on the current and voltage sampling values ​​in the waveform buffer within the fault time window. They are used to reflect the changing trend of electrical quantities before and after the fault. In the power distribution feeder scenario, they can be set as the effective current value, effective voltage value and the difference between the power frequency cycle before the fault and the power frequency cycle containing the fault. The negative sequence current and zero sequence current amplitude within the fault cycle can also be added according to the project needs to enhance the characterization capability of unbalanced faults. The protection action is used to indicate whether each protection unit operates within the fault time window and the operation time. It is preferably represented by a combination of protection type flag and operation time. For example, it indicates whether the overcurrent protection or grounding protection issues an operation at a certain unified time stamp. The switch status is used to indicate whether the switch changes from closed to open or from open to closed before and after the fault, and carries the corresponding time stamp so that the master station can distinguish the order of switch operations when analyzing the fault development process.

[0077] To ensure that the distribution master station can consider communication quality when constructing the fault evidence fingerprint matrix, the operation slice message also includes a slice delay and packet loss index. The slice delay and packet loss index refers to the record of the actual transmission performance of the operation slice message in the fault service 5G communication slice. It includes at least the end-to-end delay estimate from the time the terminal device sends the message to the time the distribution master station receives the message, whether message retransmission occurs during transmission, and whether message loss occurs. The end-to-end delay can be measured in milliseconds. The terminal records a unified timing reference timestamp when sending the message, and the distribution master station records the corresponding timestamp when receiving the message. The delay is obtained by the difference between the two or by network-side delay statistics. The number of retransmissions can be recorded using natural numbers. Each retransmission increments the count by one. When the network side detects that a message fails to arrive within a preset time limit, it can be marked as lost.

[0078] To ensure the idempotency of the operation slice messages on the master station side and avoid duplicate accumulation, the terminal device associates a message version number with each segment identifier, terminal identifier, and fault time window identifier when encapsulating the operation slice message. The message version number is used to identify the version of the operation slice content. When the terminal needs to retransmit the same operation slice due to network jitter or receiving error codes, the above three identifiers and the message version number remain unchanged, so that the distribution master station can identify that the message is a re-arrival of the same operation slice.

[0079] The power distribution master station sets up a running slice message receiving interface in the fault service 5G communication slice. This receiving interface stipulates that the running slice message should include a segment identifier, terminal identifier, fault time window identifier, voltage and current characteristic value set, protection action flag set, switch status change record, slice delay packet loss index, and message version number, and is accompanied by an error code field to indicate the reception result. The error code field is a status flag returned by the master station to the terminal device after parsing the message, used to indicate whether the reception was successful and the error type. Preferably, it includes a success code to indicate that the parsing is complete and the message content has been adopted; a timing error code to indicate that the message timestamp deviates from the unified timing reference from the allowable range; a format error code to indicate that the message format does not conform to the agreed field order or field type; a field missing error code to indicate that the necessary field is missing; and a replay error code to indicate that the received message already exists in the combination of segment identifier, terminal identifier, fault time window identifier, and message version number and the content has not changed. The terminal device decides whether to adjust the local timing configuration, repackage the message, supplement the missing field, or stop retransmitting the same running slice based on the error code type.

[0080] To ensure that the operational slices within the fault time window can be centrally analyzed at the distribution master station, the fault service 5G communication slice provides communication resources with explicit delay and retransmission limits for operational slice messages. In the preferred embodiment of medium-voltage urban distribution feeders, the end-to-end delay of a single operational slice message can be set to no more than 50 milliseconds, and the number of retransmissions can be set to no more than three. When the terminal still receives a non-successful error code or fails to receive a reception confirmation message from the distribution master station within the set maximum number of retransmissions, the current operational slice message is marked as having poor communication quality, and an alarm record is generated in local storage. The alarm record is attached to the corresponding operational slice when the fault handling evidence chain is subsequently generated, to indicate that the communication performance of this segment in this fault may affect the credibility of the evidence.

[0081] In practical engineering applications, on a city distribution feeder with a rated voltage of 10 kV divided into 20 segments, each segment terminal device can package its local operation slice into operation slice messages that meet the communication bandwidth and real-time requirements according to the above field format. These messages are then sent to the distribution master station via a fault service 5G communication slice or a dedicated power communication network with equivalent delay and reliability constraints. The distribution master station parses each field according to the agreed receiving interface and provides feedback on the terminal device's behavior based on the error codes. Distribution automation engineers can implement field filling, message encoding, error code processing, and retransmission control logic in the terminal software according to this agreement. In network planning, they can ensure that the fault service slice has the carrying capacity to meet the above delay and retransmission constraints. This ensures that this step has clear process capabilities and applicable scope in the actual operating environment, providing a complete and traceable source of operation slice messages for subsequent fault evidence fingerprint matrix construction and segment-level fault theoretical event pattern matching.

[0082] S4. The power distribution master station aggregates operational slices according to the fault time window, and constructs a fault evidence fingerprint matrix with segments as rows and measurement characteristics, action status, and communication quality indicators as columns. The specific implementation is as follows:

[0083] After the aforementioned operational slice messages are sent from each segment terminal to the distribution master station via the fault service 5G communication slice, the distribution master station first merges these operational slice messages according to the fault time window identifier within the same operational cycle. All operational slices with the same fault time window identifier are grouped into a set of records corresponding to the same fault event, and the time information attached to each operational slice message is aligned according to the unified time synchronization benchmark.

[0084] The unified time reference has been distributed to the terminal devices and the distribution master station from the same time source in the aforementioned steps, providing the prerequisite for the alignment of measurement time. When the distribution master station converges, it will calculate the reference time of each operating slice message based on the unified time reference, regard the operating slice with time deviation within the allowable range as having achieved synchronization, and mark the operating slice with time deviation exceeding the preset time error limit as a time synchronization anomaly record.

[0085] The timing error limit is a numerical limit set internally by the distribution master station to determine whether the timing deviation is acceptable. In the case of medium-voltage distribution feeders, it can be set to a time order of significantly less than one power frequency cycle. For example, if the accuracy of the unified timing reference is better than one millisecond, the timing error limit can be configured to several milliseconds. If the timestamp in the running slice message deviates from the master station's reference time by more than the timing error limit, it is recorded as a timing anomaly in this aggregation. The distribution master station assigns a lower weight to these running slices in the subsequent fault evidence matching to reflect that their time reference may have a large deviation.

[0086] After completing the merging and timing alignment by fault time window identifier, the distribution master station organizes the operating slices according to the segment identifier, and collects the voltage and current characteristic values, protection action flags, switch status change records, and slice delay packet loss indicators of each segment within the fault time window into the records at the segment level, thus constructing a fault evidence fingerprint matrix.

[0087] The fault evidence fingerprint matrix is ​​a two-dimensional recording unit in the power distribution master station used to centrally characterize the performance of each segment in a fault event. The segment corresponding to the segment identifier is the row, and the columns are the measurement characteristics, action status and communication quality indicators of the segment within the fault time window. Each matrix unit corresponds to the value of a certain segment in a certain feature dimension. For example, it can include the effective value of the current before and after the fault and its difference, the effective value of the voltage before and after the fault and its difference, the phase change, whether overcurrent protection action occurred within the fault time window, whether grounding protection action occurred, whether the switch opened or closed, and communication quality marks composed of slice delay packet loss indicators.

[0088] To ensure that the fault evidence fingerprint matrix can reflect differences in communication conditions in subsequent segment-level fault theoretical event pattern matching, the distribution master station generates communication quality weights for each segment based on the segment delay and packet loss index in the operation segment message. The communication quality weight is a graded label for the communication performance of a segment in this fault, which is used to differentiate the credibility of the evidence of that segment in the matching process.

[0089] In this preferred embodiment, when the slice latency packet loss index shows an end-to-end latency of less than ten milliseconds with no packet loss and zero retransmissions, the segment is marked as highly reliable, and the corresponding communication quality weight field is recorded as a high reliability level. When the end-to-end latency is in the range of ten to fifty milliseconds or a retransmission occurs but the packet is successfully delivered within the specified time limit, the segment is marked as moderately reliable. When packet loss or retransmissions exceed a preset upper limit, the segment is marked as low reliability, and the low reliability level is recorded in the communication quality index column of the fault evidence fingerprint matrix.

[0090] In this way, within the same fault time window, the evidence formed by different segments in terms of measurement characteristics, protection actions and switch status is clearly visible in the matrix. At the same time, the communication environment status is clearly shown through the hierarchical weights of high confidence, medium confidence and low confidence, which makes it easier to weaken the influence of segments with poor communication conditions to a certain extent and strengthen the judgment role of segments with better communication conditions.

[0091] Each time a fault evidence fingerprint matrix is ​​constructed, the distribution master station assigns a matrix version number and a timestamp to the matrix. The matrix version number is used to identify the combination of feature fields, communication quality grading rules, and timing error limit configuration used in this construction. The timestamp is used to record the time when the matrix construction is completed. The distribution master station writes the matrix version number, timestamp, and corresponding rule summary to a read-only log to ensure that when reviewing a fault event in the future, the fault evidence fingerprint matrix construction method and weight division standard used at that time can be accurately restored, avoiding confusion in interpretation caused by subsequent system upgrades or rule adjustments.

[0092] The fault evidence fingerprint matrix can be stored using a relational storage system or a time-series storage system. In a typical embodiment, a set of records with the fault event identifier as the primary key can be established for the fault evidence fingerprint matrix of each fault event in the relational storage system. The segment identifier, fault time window identifier, each feature value, action flag, and communication quality weight are stored by field. Alternatively, the values ​​of each segment in each feature dimension can be sorted and stored by time in the time-series storage system to support subsequent playback by time slice.

[0093] Regardless of the storage method used, the association between the fault evidence fingerprint matrix and the segment identifier, fault time window identifier, and matrix version number remains stable, and is provided for subsequent segment-level fault theoretical event pattern matching through a unified access interface.

[0094] In a preferred embodiment of a city power distribution feeder with a rated voltage of 10 kV divided into 20 segments, the timing error limit can be configured to be on the order of several milliseconds, and the communication quality weighting rules can be configured as follows: high reliability when the end-to-end delay is less than 10 milliseconds and there is no packet loss; medium reliability when the end-to-end delay is between 10 and 50 milliseconds or one retransmission occurs; and low reliability when packet loss or more than three retransmissions are detected. Under this configuration, when a fault occurs, the operation slice messages sent from the terminals of each segment to the distribution master station will be converged and aligned within a few hundred milliseconds. The fault evidence fingerprint matrix constructed by the distribution master station according to the above rules can be formed within a time of one power frequency cycle, providing a clear, structured, and reliable fault evidence basis for subsequent segment-level fault theoretical event pattern matching and fault segment judgment. Based on this, power distribution automation engineers in this field can implement the fault evidence fingerprint matrix construction logic in database models and fault analysis applications, and extend this step to a wider range of operating conditions by adjusting the timing error limit and communication quality grading rules in power distribution networks with different voltage levels and network conditions.

[0095] S5. Based on the topology model and protection setting derivation, the distribution master station performs constraint matching between the fault evidence fingerprint matrix and each theoretical event pattern, calculates the matching score, and obtains candidate fault sections. The specific implementation is as follows:

[0096] In this embodiment, units with a high confidence level in the communication quality weight of the fault evidence fingerprint matrix are designated as high-confidence entries, units with a medium confidence level are designated as medium-confidence entries, and units with a low confidence level are designated as low-confidence entries. After the aforementioned fault evidence fingerprint matrix is ​​constructed, the distribution master station derives and matches the theoretical event patterns of segment-level faults based on this matrix and the segment-level topology model. The theoretical event patterns of segment-level faults are expected response descriptions derived by combining the feeder segment-level topology model and the protection setting parameters of each segment, under the assumption that a specific type of fault (such as single-phase grounding or phase-to-phase short circuit) occurs in a certain segment. They are used to characterize the sequence of changes in voltage and current characteristic values, protection operation status, and switch status of each segment within the fault time window under this assumption.

[0097] The segment-level topology model provides the upstream and downstream relationships, branch locations, and electrical distances to the substation outlet for each segment in terms of physical connection. The protection setting parameters for each segment include the setting current, operating time, directional attributes, and blocking conditions of the overcurrent protection, grounding protection, and other units configured on each segment. The distribution master station associates these parameters with the topology during the rule configuration phase, so that when a fault occurs in a segment, it can infer the relative change trends of the current amplitude, voltage amplitude, and zero-sequence component of the fault segment and its upstream and downstream segments, and infer which protections should operate, which protections should remain inactive, and which switches should trip and which switches should remain unchanged.

[0098] When implementing the segment-level fault theoretical event mode, it can be set as a mode record indexed by segment identifier and fault type. Each mode record lists the direction of change (increase, decrease, or remain basically unchanged) of the effective values ​​of voltage and current in each segment compared to before the fault when a given fault type occurs in that segment, whether the zero-sequence component deviates significantly, whether the corresponding protection should operate within the fault time window and the order of its operation, whether the relevant switches should open or close within the fault time window, and the basic assumptions for communication quality requirements, such as assuming that the measurement and operation records of the high-reliability communication segment should be able to fully reflect the mode.

[0099] During operation, the power distribution station traverses each row of the fault evidence fingerprint matrix according to the segment identifier. It performs constraint matching on the fault evidence fingerprint of each segment within the fault time window and the segment-level fault theoretical event pattern with that segment as the fault hypothesis segment. Constraint matching means that during the comparison process, high-confidence, medium-confidence, and low-confidence entries are distinguished according to the aforementioned communication quality weights. For high-confidence entries, the actual characteristics and theoretical expectations are required to be consistent in terms of value and direction of change. For medium-confidence entries, a certain tolerance is allowed in terms of amplitude deviation or time alignment, but the trend of change should still be consistent. For low-confidence entries, missing or contradictory entries are allowed but are not used as direct evidence to reject the pattern.

[0100] To facilitate the quantification of matching results, the distribution station counts the number of units whose actual characteristics match the theoretical expectations in the high-confidence entries and the total number of high-confidence entries during each pattern matching. The ratio of the former to the latter is used as one of the core indicators to measure the ability of the theoretical event mode of the fault at this stage to interpret the current fault evidence fingerprint matrix. The number of consistent units in the medium-confidence entries can also be considered as an auxiliary reference.

[0101] In the preferred embodiment of the aforementioned urban power distribution feeder with a rated voltage of 10 kV and divided into 20 segments, the power distribution master station can set a matching score threshold of at least 80% of the consistent units in the high-confidence entries. That is, only when the coverage rate of high-confidence entries in the corresponding candidate fault segment of a certain segment-level fault theoretical event mode reaches or exceeds this proportion will the mode be considered to have reasonable explanatory power for the current fault event, and the corresponding segment or the area composed of several consecutive segments will be marked as a candidate fault segment. The matching score value of the mode, the number of high-confidence entries explained, and the consistency status of medium-confidence entries will be attached to the candidate fault segment record so as to facilitate subsequent comparison and screening among multiple candidate segments.

[0102] To ensure consistency between the theoretical event modes of segment-level faults and the actual protection configuration, the distribution master station updates the theoretical event modes of segment-level faults through offline calculation or semi-automatic verification when the system is put into operation, protection settings are modified, or topology is adjusted. The update content includes the combination of fault types in each segment, the current and voltage change trends in each segment, the sequence of protection actions, and the circuit breaker tripping logic. After each update, a new mode version number is assigned to the mode library, and the currently effective mode version number and the matching scoring threshold value are written to the read-only log. At the same time, the mode version number is appended to the supplementary record of the fault handling evidence chain, so that when reviewing past fault events, it is possible to clearly distinguish which version of the segment-level fault theoretical event mode and matching scoring rules were used at that time.

[0103] In practical engineering applications, for a typical 10kV urban distribution feeder, at least two types of segment-level fault theoretical event modes—single-phase grounding and phase-to-phase short circuit—can be constructed for each segment, resulting in a model library of dozens to hundreds of modes. During operation, the aforementioned matching scoring threshold with 80% high reliability coverage is adopted. After the fault evidence fingerprint matrix is ​​constructed, the constraint matching of all modes and the screening of candidate fault segments are completed within one power frequency cycle. This allows distribution automation engineers to derive and match segment-level fault theoretical event modes by maintaining the segment-level topology model and protection setting parameters without relying on complex numerical simulations. Furthermore, by adjusting the model library content and matching scoring threshold according to different voltage levels and protection configuration information, this process can be extended to more distribution conditions.

[0104] S6. The distribution master station solves for the minimum fault section set and the minimum disconnect switch set on the candidate fault sections based on evidence constraints, generates a fault isolation operation sequence, and solidifies the operation slice, fault evidence fingerprint matrix, matching results, and isolation results to form a replayable fault handling evidence chain. The specific implementation is as follows:

[0105] After obtaining the candidate fault section set and completing the section-level fault theoretical event mode matching, the distribution master station further applies evidence constraints and operational constraints to this candidate set in order to select the minimum fault section set and minimum disconnecting switch set from multiple candidate schemes that can fully explain the fault evidence and meet the field operation boundary.

[0106] Evidence constraints refer to the requirement that when selecting a combination of fault sections, the distribution master station must ensure that the selected combination covers all high-confidence entries in the fault evidence fingerprint matrix and does not contradict the electrical connection relationships recorded in the section-level topology model.

[0107] Specifically, the distribution master station first extracts the measurement features and action states marked as high confidence from the fault evidence fingerprint matrix. For example, certain segments may show a significant increase in current, a significant decrease in voltage, a deviation in zero-sequence components, and protection actions or switch tripping within the fault time window. These high confidence entries are grouped according to segment identifiers to form a set of high confidence evidence that needs to be interpreted.

[0108] Then, the distribution master station verifies the logical consistency between the candidate fault sections and the high-confidence evidence based on the power supply direction, branch point location, and adjacent section connection relationship recorded in the section-level topology model. For example, when a section is assumed to be a fault section, its upstream section should show an increase in current, possibly triggering some protection or switch tripping within the fault time window, and the downstream section should show a voltage drop or complete loss of voltage. If the fault sections included in a candidate combination cannot explain the behavior of the sections corresponding to certain high-confidence entries in terms of topology, then the combination is considered to not meet the conditions in terms of evidence constraints.

[0109] In addition, the distribution master station also checks whether the candidate fault section combination contradicts the ring network switch position and feeder segmentation method in the segment-level topology model. For example, it cannot be that only the downstream end segment is selected as the fault section but attempts to interpret the protection action of the upstream segment.

[0110] Operational constraints refer to the further checks by the distribution master station on the minimum fault section set and minimum disconnecting switch set corresponding to each combination after screening through evidence constraints, to ensure that they meet the operating conditions such as the allowable current of the line, the allowable current of the standby line, and the closing and opening capacity of the switches, and do not exceed the safety boundaries set by the operating procedures.

[0111] To this end, the distribution station maintains a static parameter database and an operation rule database for lines and switches during long-term operation. The static parameter database records the allowable long-term current, short-time current carrying capacity, rated current, ultimate breaking current, and thermal stability limit of each conductor segment. The operation rule database records the loop patterns that are not allowed to be formed, the combination of key equipment that is prohibited from being shut down simultaneously, and the power supply guarantee requirements for important users.

[0112] When the distribution station performs operational constraint verification on a candidate combination of faulty sections, it first simulates the power supply path of the remaining network after these faulty sections are removed from the operating network, based on the section-level topology model and the current operating mode. It then roughly estimates the current distribution on each power supply path. This estimation can be based on the current measurement of the time slice before the fault, obtained by redistributing the load that originally flowed through the faulty section along the standby line and parallel path. This is then compared with the allowable current of the line and the allowable current of the standby line recorded in the static parameter library. If the expected current of a certain line after isolation exceeds the allowable value, then the candidate combination is considered to not meet the operational constraints.

[0113] Meanwhile, the power distribution station checks the rated current and closing / opening capacity of each switch based on the list of switches included in the minimum disconnecting switch set. If a switch needs to bear a current exceeding its rated capacity under the proposed disconnection scheme or is marked in the operating procedures as equipment that is not suitable for frequent operation under specific load conditions, then the scheme is also eliminated.

[0114] Provided that the above-mentioned evidence constraints and operational constraints are met, the distribution master station sets optimization targets among the candidate schemes. In a typical 10 kV feeder scenario, the targets can be set as no more than three fault sections and no more than four disconnect switches that need to be operated. Under these constraints, the combination with fewer power outage users is given priority.

[0115] The number of users experiencing power outages can be obtained by linking the segment-level topology model with the user ledger. The user ledger records the number of users connected to each segment, the identifiers of important users, and their load levels. The user ledger is a set of user information records maintained by the distribution master station, which includes at least the user access segment identifier, load level, and whether it belongs to an important user. It is used to count the number of users experiencing power outages during operation mode analysis and fault isolation optimization. The distribution master station calculates the number of users who will lose power after isolation and the number of important users for each candidate combination. Among multiple candidate combinations, the scheme with fewer fault segments, simpler isolation operation, and the fewest number of users experiencing power outages is selected first.

[0116] After the minimum fault section set and minimum disconnect switch set are finally determined, the distribution master station generates a fault isolation operation sequence based on the current switch status. The fault isolation operation sequence is a series of opening or closing instructions that need to be executed sequentially to achieve the proposed isolation scheme based on the current operating mode. Each instruction includes at least the target switch identifier, the target status (open or closed), and the sequence number. The sequence number is used to ensure that each step of the operation is strictly executed in accordance with the predetermined order in the case of concurrent operation and communication retransmission.

[0117] To ensure idempotency and the ability to continue execution after an interruption, the distribution master station records the sequence number of each instruction and its execution result locally. When communication is interrupted or the terminal temporarily refuses to execute, the distribution master station will issue instructions starting from the instruction after the largest sequence number that has not yet been confirmed as successfully executed. For instructions that have been confirmed as successfully executed by the terminal, even if they arrive again due to the retransmission mechanism, the terminal will identify them as executed instructions based on the sequence number and return a success error code without repeating the physical operation, thereby avoiding repeated opening or closing of the circuit breaker.

[0118] The fault isolation operation sequence is issued through the control interface corresponding to the terminal. The control interface follows the aforementioned message structure convention. After each instruction is issued, the terminal checks the target switch status and current operating conditions locally. If the safety conditions are met, the operation is executed. When the operation is completed or it is determined that it cannot be executed, the execution result and error code are returned. The error codes preferably follow the aforementioned success code, timing error code, format error code, field missing error code, and replay error code, which are used to indicate whether the timing deviation is acceptable, whether the instruction format conforms to the convention, whether the necessary fields are complete, and whether a duplicate instruction has been received.

[0119] After a fault location and isolation is completed, the power distribution master station will summarize all the operation slice messages sent by each terminal in the fault time window, the fault evidence fingerprint matrix constructed in the master station, the version number of the segment-level fault theoretical event mode used for matching, the matching score results of each candidate fault segment, the final determined minimum fault segment set and minimum disconnect switch set, and the records of each switch status change during the execution process, and generate a replayable fault handling evidence chain.

[0120] The replayable fault handling evidence chain uses fault event identifiers and evidence chain version numbers as indexes to record information on the entire process from fault occurrence, fault evidence formation, theoretical event pattern matching, candidate section screening to isolation operation and results. This allows operators to review the data and decisions of each key node in chronological order after the fact. In long-term operation, the evidence chain is used to statistically analyze indicators such as fault location time, fault section determination accuracy, and average power outage time for non-faulty users to evaluate and adjust the field performance of this method on different feeders.

[0121] In a preferred embodiment of a 10 kV urban distribution feeder, it is possible to set up an operation that collects no fewer than several dozen real fault events and drill events over a year. The goal is to achieve a time of several seconds from the completion of the fault evidence fingerprint matrix construction to the generation of the fault isolation operation sequence. The accuracy rate of fault section identification is controlled at over 95%, and the statistical value of the average power outage time for non-faulty users is significantly lower than the historical statistical value of similar feeders without this method. When the above indicators stably meet the objectives under a certain statistical sample size, it can be considered that this method has repeatable field application capability under the current operating conditions. Based on this, distribution automation engineers in the field can extend and implement this step together with the aforementioned steps in different regions, different voltage levels, and different network structures of distribution systems by adjusting the upper limit of the number of fault sections, the upper limit of the number of disconnecting switches, and related operating constraint parameters.

[0122] In the operational scenario shown in this embodiment: on an urban power distribution feeder with a rated voltage of 10 kV, a total length of approximately 20 kilometers, and divided into 20 sections, before commissioning, the power distribution company, through its main distribution station, completes the establishment of a section-level topology model, planning of section and terminal identifiers, configuration of a unified timing reference, and application for 5G communication slice parameters for fault services according to the aforementioned steps. Several sections bearing public service loads such as hospitals, rail transit stations, and power dispatching agencies are marked as key sections in the section-level topology model, and higher service priority is configured for the corresponding terminals in the 5G communication slice parameters for fault services. Waveform buffers and status buffers are pre-set inside the terminal devices installed in each section of the feeder, and local fault criteria and fault time window lengths are configured, so that each terminal can independently complete suspected fault identification and operation slice organization under the constraints of a unified timing reference.

[0123] During daily operation, this feeder carries several industrial loads and a large number of residential loads. The distribution substation collects voltage and current measurement data and monitors the online status of terminals according to a predetermined cycle. The 5G communication slice for fault services remains idle and ready. During a peak electricity consumption period on a certain day, a single-phase ground fault occurred at a cable joint located in the twelfth section. After the fault occurred, the corresponding terminal in the twelfth section detected a significant increase in the effective current value, a significant decrease in the effective voltage value, and a sudden change in the zero-sequence component within several power frequency cycles. The upstream eleventh section terminal detected an increase in current with a small change in voltage, and the downstream thirteenth section terminal detected a sudden drop in voltage and the disappearance of the load-side current. Each terminal, based on its local fault criteria, sets the local suspected fault mark after the judgment results exceed the preset threshold for two consecutive power frequency cycles. It also extracts waveform and status cache contents from the three preceding and three following power frequency cycles around the fault time under a unified timing reference, and organizes them into an operation slice containing section identifier, terminal identifier, fault time window identifier, effective current value sequence, effective voltage value sequence, phase change sequence, protection action flag, and switch status change record. Within a very short time after the fault occurred, the terminal devices of the eleventh, twelfth, and thirteenth segments, as well as other segments, encapsulated their local operation slices into operation slice messages and sent them to the distribution master station through the faulty service 5G communication slice. In this fault, the 5G cell link of the twelfth segment terminal was in good condition, and its operation slice message was delivered once within tens of milliseconds without packet loss. The eleventh segment terminal message was delivered after one retransmission within the allowed time. The thirteenth segment terminal message was retransmitted multiple times due to local wireless environment interference. While receiving the operation slice message, the distribution master station generated a slice delay and packet loss index based on the end-to-end delay estimate, retransmission count, and loss marker carried in the message, and fed back the timing error, format error, or replay to the terminal through the error code mechanism. The terminal recorded the communication performance in this fault locally accordingly.

[0124] The distribution master station aggregates operation slice messages from various segments under the same fault time window identifier. It aligns the time information in the messages according to a unified time synchronization benchmark, considering records with time deviations within the time synchronization error limit as normal, and marking records with deviations exceeding the limit as abnormal. Subsequently, it organizes the voltage and current characteristic values, protection action flags, switch status change records, and slice delay packet loss indicators for each segment within the fault time window according to segment identifiers. This constructs a fault evidence fingerprint matrix with segments as rows and measurement characteristics, action status, and communication quality indicators as columns. According to preset rules, the twelfth segment record is marked as high confidence, the eleventh segment record as medium confidence, and the thirteenth segment record as low confidence due to multiple retransmissions. After construction, a matrix version number and timestamp are assigned to the matrix, written to a read-only log, and stored in a relational storage system.

[0125] The distribution master station reads the segment-level topology model and the protection setting parameters of each segment. For each segment assuming a single-phase grounding or phase-to-phase short-circuit fault, it calls the pre-established segment-level fault theoretical event mode. It then performs constraint matching between the fault evidence fingerprint of each segment in the matrix and the corresponding theoretical event mode. For high-confidence items (such as current, voltage, zero-sequence changes and protection operation in the twelfth segment), the actual characteristics are required to be consistent with the theoretical expectations. For medium-confidence items (such as current changes and protection failure in the eleventh segment), amplitude deviation is allowed but trend consistency is required. For low-confidence items (such as missing measurements in the thirteenth segment), it is not used as a negative criterion. The matching score of each hypothetical fault segment mode is calculated, and the mode with a high-confidence item consistency rate of not less than 80% is selected as a candidate. The corresponding twelfth segment and several consecutive segments composed of its adjacent segments are marked as candidate fault segments, and the matching score value and the number of high-confidence items interpreted are recorded for each candidate segment.

[0126] Subsequently, the distribution station superimposes evidence constraints and operational constraints onto the candidate fault section set. The evidence constraints check whether each fault section combination can reasonably explain the measurement characteristics and action states corresponding to all high-confidence entries under the upstream and downstream relationships shown in the section-level topology model. The operational constraints use the static parameter library and operation rule library to evaluate whether the current distribution in the remaining network after the proposed fault section is cut off and the corresponding disconnecting switch is operated exceeds the line allowable current and the standby line allowable current, whether the closing and opening capacity of the proposed switch meets the requirements, and whether the prohibited operation mode is triggered. After filtering out candidate combinations that do not meet the evidence constraints or operational constraints, an optimal solution is selected according to the objectives of no more than three fault sections, no more than four disconnecting switches to be operated, and minimizing the number of power outage users. The twelfth section and one or two sections consisting of its upstream and downstream adjacent sections are listed as the minimum fault section set, and the several switches used to disconnect the fault section from the main feeder and restore power to some users through the standby feeder are listed as the minimum disconnecting switch set.

[0127] The distribution master station generates a fault isolation operation sequence based on the current switch status. For each switch requiring operation, it generates a tripping or closing command with a sequence number and sends it to the field terminal through the control interface corresponding to the terminal. The terminal locally verifies the target status and current operating conditions. If the safety premise is met, it executes the tripping and closing operations in sequence and returns the execution result and error code after each step. The distribution master station records the execution status of each sequence number command. After the communication interruption is restored, it continues to issue subsequent commands from the highest sequence number that has not yet been successfully confirmed. The sequence number and error code are used to ensure that commands that have been successfully executed are not executed repeatedly.

[0128] After the fault is isolated and power is restored to the non-faulty sections through reconfiguration, the distribution master station summarizes all operational slice messages within the fault time window, the constructed fault evidence fingerprint matrix, the version numbers of the segment-level fault theoretical event modes participating in the matching, the matching score results of each candidate faulty section, the final determined minimum faulty section set and minimum disconnecting switch set, and the records of each switch status change during the isolation process. This generates a replayable fault handling evidence chain indexed by the fault event identifier and the evidence chain version number. Based on this evidence chain, the operation statistics module calculates the time from the first detection of the suspected fault by the terminal to the generation of the fault isolation operation sequence, the degree of match between the minimum faulty section set and the actual fault point, and the average power outage time for non-faulty users. After accumulating a certain number of real fault events and drill events, the method is evaluated to determine whether the fault location time on this feeder is stable in the range of a few seconds, whether the faulty section determination accuracy is stable above 95%, and whether the average power outage time for non-faulty users is significantly better than the historical statistical values ​​when this method is not used. This provides quantifiable field performance data for the promotion of this method on other feeders.

[0129] All calculations involved in the embodiments are dimensionless numerical calculations, and the preset parameters and thresholds in the calculations are set by those skilled in the art according to the actual situation.

[0130] It should be noted that this invention can be deployed on the device itself to realize embedded applications, or it can run on a PC or other terminal with a user interface, thereby meeting various hardware environments and usage requirements.

[0131] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wireless or wired transmission; wired transmission methods include optical fiber, twisted pair, coaxial cable, etc.; wireless transmission includes infrared, microwave, etc. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center containing one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0132] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and modules described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0133] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0134] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0135] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0136] If the aforementioned functions are implemented as software functional modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0137] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0138] In conclusion, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for rapid fault location of power distribution feeders based on 5G communication slicing, characterized in that, include: S1. Construct a segment-level topology model at the power distribution master station and bind segment identifiers and terminal identifiers, and configure a unified timing reference and 5G communication slice parameters for fault services based on the segment-level topology. S2. Set waveform buffers and status buffers in each segment terminal device. When the local fault criterion is triggered, extract the data within the fault time window according to the unified time synchronization and generate a running slice with segment identifier and time window identifier. S3. Each segment terminal device sends a running slice message through the fault service 5G communication slice. The message carries voltage and current characteristic values, protection actions, switch status and slice delay packet loss indicators. S4. The power distribution master station aggregates operation slices according to the fault time window and constructs a fault evidence fingerprint matrix with segments as rows and measurement characteristics, action status, and communication quality indicators as columns. The communication quality index is mapped from the slice latency and packet loss index to the communication quality weight. The communication quality weight is divided into high confidence level, medium confidence level and low confidence level, and the matrix version number and timestamp are written to the read-only log. S5. Based on the topology model and protection setting, the distribution master station derives the section-level fault theoretical event patterns, performs constraint matching between the fault evidence fingerprint matrix and each theoretical event pattern, calculates the matching score, and obtains candidate fault sections. S6. The power distribution master station solves the minimum fault section set and minimum disconnect switch set on the candidate fault section according to the evidence constraints, generates the fault isolation operation sequence, and solidifies the operation slice, fault evidence fingerprint matrix, matching results, and isolation results to form a replayable fault handling evidence chain.

2. The method for rapid fault location of power distribution feeders based on 5G communication slicing according to claim 1, characterized in that, S1 includes: The power distribution master station constructs a segment-level topology model based on switches and cable joints, and maps segment identifiers to terminal identifiers; Time references are sent to terminals based on a unified time synchronization benchmark; Configure 5G communication slice parameters for faulty services to carry out slice operation messages; When using a dedicated power communication network to carry operation slice messages, the 5G communication slice parameters of faulty services are mapped to the service level within the dedicated power communication network, and the operation slice is uploaded and the fault evidence fingerprint matrix is ​​constructed under the constraints of a unified timing reference and 5G communication slice parameters of faulty services.

3. The method for rapid fault location of power distribution feeders based on 5G communication slicing according to claim 1, characterized in that, S2 include: Set waveform buffers and status buffers in the terminal device corresponding to the segment; The terminal device calculates the effective value of current, effective value of voltage, and zero-sequence component based on the current and voltage sample values ​​and executes local fault criteria. When a local fault criterion is triggered, a fault time window identifier is generated based on a unified timing reference. The data corresponding to the fault time window in the waveform cache and status cache are organized into running slices and appended with segment identifiers, terminal identifiers and fault time window identifiers.

4. The method for rapid fault location of power distribution feeders based on 5G communication slicing according to claim 1, characterized in that, S3 includes: Within each terminal device, the operational slices organized around the fault time window are encapsulated into operational slice messages; The operation slice message includes segment identifier, terminal identifier, fault time window identifier, voltage and current characteristic values, protection action flag set, switch status change record, slice delay packet loss index and message version number, and is sent to the distribution master station through fault service 5G communication slice.

5. The method for rapid fault location of power distribution feeders based on 5G communication slicing according to claim 4, characterized in that: The power distribution master station is equipped with an interface for receiving operation slice messages; The running slice message receiving interface parses the segment identifier, terminal identifier, fault time window identifier, voltage and current characteristic value set, protection action flag set, switch status change record, slice delay packet loss index and message version number in the running slice message, and returns the reception result with error code field. The terminal device adjusts the local time synchronization configuration, running slice message encapsulation method, field padding and retransmission strategy according to the error code field.

6. The method for rapid fault location of power distribution feeders based on 5G communication slicing according to claim 1, characterized in that, S4 include: The power distribution master station merges the operation slice messages according to the fault time window identifier and aligns the message time information under a unified time synchronization reference. By summarizing voltage and current characteristic values, protection action flags, switch status change records, and slice delay and packet loss indicators according to segment identifiers, a fault evidence fingerprint matrix is ​​constructed with segments as rows and measurement characteristics, action status, and communication quality indicators as columns.

7. The method for rapid fault location of power distribution feeders based on 5G communication slicing according to claim 1, characterized in that, S5 include: The distribution master station constructs a theoretical event pattern of section-level faults based on the section-level topology model and protection setting parameters, indexed by section identifier and fault type; Based on the communication quality weight, the fault evidence fingerprint matrix units are divided into high-confidence entries, medium-confidence entries, and low-confidence entries; When traversing each segment, the distribution master station performs constraint matching between the fault evidence fingerprint of the segment and the corresponding segment-level fault theoretical event mode. The matching score is determined based on the proportion of units in the high-confidence entries that are consistent with the theoretical expectation. When the matching score reaches the matching score threshold, the segment is marked as a candidate fault segment.

8. The method for rapid fault location of power distribution feeders based on 5G communication slicing according to claim 1, characterized in that, S6 include: The power distribution master station applies evidence constraints and operational constraints to the set of candidate fault sections; Evidence constraints require that the combination of faulty sections covers all high-confidence entries in the fault evidence fingerprint matrix and is consistent with the electrical connection relationships recorded in the segment-level topology model; The operational constraints are based on a static parameter library and an operational rule library. The allowable current of the line, the allowable current of the backup line, and the closing and opening capacity of the switch are checked to determine the minimum set of fault sections and the minimum set of disconnecting switches. The target combination is selected based on the number of users experiencing power outages, in conjunction with the user ledger.

9. The method for rapid fault location of power distribution feeders based on 5G communication slicing according to claim 8, characterized in that: After determining the minimum fault section set and the minimum disconnect switch set, the power distribution master station generates a fault isolation operation sequence based on the current switch status. Each step in the fault isolation operation sequence includes the target switch identifier, the target status, and the sequence number. The power distribution master station executes opening and closing commands through the control interface and according to the error code feedback control terminal device. Based on the recorded command sequence number and execution result, it generates a replayable fault handling evidence chain indexed by the fault event identifier and evidence chain version number.

Images