A method for monitoring faults of a 35kV looped network cable intermediate joint of urban rail transit

Abstract

This application provides a method for monitoring faults in intermediate joints of 35kV ring network cables in urban rail transit, relating to the field of data processing. In this method, a cable segmentation diagram is constructed, and joint identification nodes and ring network status identifiers are generated. A synchronous observation array is established based on the station to collect transient voltage and current segments to form event evidence packages. Under the same ring network status, a set of propagation operators is constructed, and a virtual playback field is generated through recorrelation and superposition to produce a focused spectrum. Under the same operating context, a reversibility discrimination is performed to screen candidates for saturated critical states. A first focused spectrum and a second focused spectrum are generated, and structural fracture is synchronously checked. Finally, the target joint identification node is located, and the intermediate joint fault monitoring results are output. Implementing the technical solution provided in this application facilitates overcoming the dependence on continuous gradual changes and identifying the electric field reconstruction characteristics caused by space charge trap saturation, thereby improving the accuracy of abrupt degradation judgment.

Description

Technical Field

[0001] This application relates to the technical field of data processing, specifically to a method for monitoring faults in intermediate joints of 35kV ring network cables for urban rail transit. Background Technology

[0002] Urban rail transit 35kV ring network cables operate under long-term high load, frequent start-stop, and complex electromagnetic environment conditions. As a critical weak point in the cable line, the cable joints are subjected to the coupled effects of electric field stress, thermal stress, and mechanical stress, making them highly susceptible to space charge accumulation at the insulation interface. Current technologies for monitoring the operating status of joints largely rely on continuous gradual changes in characteristics such as temperature rise trend analysis, partial discharge amplitude statistics, or traveling wave anomaly frequency statistics. Threshold determination or regression analysis of long-term trend changes in these characteristic quantities is used to identify and warn of deterioration. However, in actual operation, space charge gradually accumulates within the joint at the insulation interface and defect areas, forming traps. When these traps gradually approach saturation, the electric field distribution within the material may be reconstructed in a short period, leading to a redistribution of discharge channels or a sudden increase in discharge intensity. This process often manifests as a long period of relatively stable characteristic quantities followed by a sudden jump at a critical moment, representing a typical abrupt deterioration path.

[0003] In such abrupt degradation processes, because the initial temperature, partial discharge amplitude, or conventional traveling wave characteristics do not show a clear gradual change trend, monitoring models built based on the assumption of continuous change struggle to identify critical degradation states in a timely manner, easily leading to monitoring blind spots of "long-term normal operation followed by instantaneous failure." Simultaneously, the frequent switching of operating modes and severe load fluctuations in urban rail transit ring network cables, coupled with the intertwining and superposition of external disturbances and internal degradation signals, make traditional monitoring methods relying on amplitude thresholds or simple trend judgments more susceptible to interference from changes in operating conditions, thereby reducing the accuracy of identifying abrupt degradation.

[0004] Therefore, there is an urgent need for a fault monitoring method for intermediate joints that can overcome the dependence on continuous gradual changes and identify the electric field reconstruction characteristics caused by the saturation of space charge traps, so as to accurately determine the path of initial stability and subsequent abrupt deterioration. Summary of the Invention

[0005] This application provides a fault monitoring method for intermediate joints of 35kV ring network cables in urban rail transit, which facilitates the identification of electric field reconstruction characteristics caused by space charge trap saturation, thus improving the accuracy of abrupt degradation judgment.

[0006] The first aspect of this application provides a method for monitoring faults in intermediate joints of 35kV ring network cables for urban rail transit. The method includes: acquiring ring network topology data, switch position status data, and cable mileage mapping data of the 35kV ring network cable for urban rail transit to construct a cable segmentation diagram; simultaneously generating a joint identifier node for each intermediate joint in the cable segmentation diagram and generating a ring network status identifier corresponding to the current switch position status; constructing a synchronous observation array based on the stations in the cable segmentation diagram to simultaneously monitor transient events during ring network operation, generating an event identifier for each transient event, and extracting voltage and current segments from the event identifier within the synchronous observation array to form an event evidence package; accumulating the event evidence packages within the same ring network status identifier to construct a propagation operator set; and performing cross-station recorrelation superposition on the propagation operator set after performing amplitude-phase normalization and time-domain inversion processing to form a virtual feedback loop. The system performs a virtual playback field and generates a focusing map indexed by the joint identifier node. Under the same current-carrying range and the same laying environment, it performs a regressibility judgment on the focusing map. When the focusing index set of the same joint identifier node cannot regress to the center of the historical focusing baseline group, the corresponding joint identifier node is marked as a saturation critical state candidate, and the saturation critical state candidate is used to generate a first focusing map and a second focusing map based on voltage and current segments, respectively. When the first focusing map and the second focusing map show structural fracture in the neighborhood of the same joint identifier node, and the structural fracture meets the synchronization condition, the corresponding joint identifier node is determined as a faulty joint node, and a fault status mark bound to the faulty joint node is generated. According to the fault status mark, the target joint identifier node is determined on the cable segmentation diagram, and the intermediate joint fault monitoring result corresponding to the target joint identifier node is output.

[0007] A second aspect of this application provides a fault monitoring device for intermediate joints of 35kV ring network cables in urban rail transit. The device includes an acquisition module and a processing module. The acquisition module acquires ring network topology data, switch position status data, and cable mileage mapping data of the 35kV ring network cable in urban rail transit to construct a cable segmentation diagram. Simultaneously, it generates a joint identification node for each intermediate joint in the cable segmentation diagram and generates a ring network status identifier corresponding to the current switch position status. The processing module constructs a synchronous observation array based on the stations in the cable segmentation diagram, simultaneously monitors transient events during ring network operation, generates an event identifier for each transient event, and extracts voltage and current segments from the event identifiers within the synchronous observation array to form an event evidence package. The processing module further accumulates the event evidence packages within the same ring network status identifier, constructs a propagation operator set, and performs cross-site recorrelation superposition after performing amplitude-phase normalization and time-domain inversion processing on the propagation operator set. A virtual playback field is formed, and a focusing pattern indexed by the joint identification node is generated based on the virtual playback field. The processing module is further configured to perform a regressibility judgment on the focusing pattern under the same current carrying range and the same laying environment. When the focusing index set of the same joint identification node cannot be returned to the center of the historical focusing baseline group, the corresponding joint identification node is marked as a saturation critical state candidate, and the saturation critical state candidate is used to generate a first focusing pattern and a second focusing pattern based on voltage and current segments, respectively. The processing module is further configured to determine the corresponding joint identification node as a faulty joint node when the first focusing pattern and the second focusing pattern have structural fractures in the neighborhood of the same joint identification node, and the structural fractures meet the synchronization condition, and generate a fault status mark bound to the faulty joint node. The processing module is further configured to determine the target joint identification node on the cable segmentation diagram according to the fault status mark, and output the intermediate joint fault monitoring result corresponding to the target joint identification node.

[0008] A third aspect of this application provides an electronic device including a processor, a memory, a user interface, and a network interface. The memory is used to store instructions, and both the user interface and the network interface are used to communicate with other devices. The processor is used to execute the instructions stored in the memory to cause the electronic device to perform the method described above.

[0009] A fourth aspect of this application provides a non-transitory computer-readable storage medium storing instructions that, when executed, perform the method described above.

[0010] In summary, one or more technical solutions provided in this application have at least the following technical effects or advantages: Cable segmentation diagrams are constructed using ring network topology data, switch position status data, and cable mileage mapping data. Joint identification nodes and ring network status identifiers are introduced, ensuring that all subsequent monitoring calculations are performed under clearly defined topological boundary conditions. This approach, using ring network status identifiers as the binning dimension, effectively isolates the impact of different operating modes on the propagation structure, preventing misjudgments of structural changes caused by switching or operating mode changes as joint degradation, thus improving the physical consistency of monitoring results from the source. Voltage and current segments from multiple sites are acquired through a synchronous observation array, and event evidence packages are constructed, ensuring that each transient has a consistent spatiotemporal description across sites, providing stable input for the construction of the propagation operator set. This arrayed, multi-channel evidence organization method improves spatial resolution and anti-interference capabilities compared to single-point monitoring, laying the foundation for subsequent joint-level localization. Furthermore, based on the propagation operator set, a virtual playback field is formed through time-domain reversal and cross-site recorrelation superposition, generating a focused spectrum indexed by joint identification nodes, thus realizing the visualization of the cable's internal scattering structure. This focused pattern is not a simple traveling wave arrival time analysis, but rather a concentrated expression of the energy of potential anomaly scattering centers formed by structural superposition, enabling joint-level anomalies to be characterized in a spatially focused form, thereby enhancing positioning accuracy and structural interpretability.

[0011] Under the same current-carrying range and laying environment, a reversibility discrimination is performed, transforming the monitoring logic from "whether the threshold is exceeded" to "whether it can return to the historical structural center." This reversibility discrimination mechanism can identify changes in dielectric memory caused by the gradual accumulation of space charge. Even when temperature rise and partial discharge have not significantly increased, it can detect potential risks by focusing on structural offsets, thus solving the problem of identifying early-stage stability followed by later abrupt changes. After confirming the candidate saturation critical state, cross-channel consistency verification is achieved by constructing a first and second focusing map and performing synchronous verification, avoiding false alarms caused by single-channel anomalies or measurement link problems. This dual-map synchronous structural fracture criterion improves the reliability of fault confirmation, making the determination of faulty joint nodes physically consistent and supported by multi-source evidence. By performing localized attribution based on fault state markings on the cable segment diagram and determining the target joint identification node, a closed-loop output from structural anomaly to joint-level positioning is achieved. The monitoring results are bound to mileage mapping information, allowing the results to directly serve on-site operation and maintenance decisions. This not only improves the accuracy of identifying abrupt degradation, but also enhances the stability and traceability of fault diagnosis, realizing a technological shift from reactive emergency repairs to predictive maintenance. Attached Figure Description

[0012] Figure 1 A flowchart illustrating a method for monitoring faults in intermediate joints of 35kV ring network cables for urban rail transit, provided as an embodiment of this application; Figure 2 A schematic diagram of a module for monitoring faults in intermediate joints of 35kV ring network cables for urban rail transit, provided in an embodiment of this application; Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0013] Explanation of reference numerals in the attached figures: 21. Acquisition module; 22. Processing module; 31. Processor; 32. Communication bus; 33. User interface; 34. Network interface; 35. Memory. Detailed Implementation

[0014] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments.

[0015] In the description of the embodiments of this application, the words "for example" or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design that is described as "for example" or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design options. Rather, the use of the words "for example" or "for instance" is intended to present the relevant concepts in a specific manner.

[0016] In the description of the embodiments of this application, the term "multiple" means two or more. For example, multiple systems means two or more systems, and multiple screen terminals means two or more screen terminals. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the indicated technical features. Thus, a feature defined with "first" or "second" may explicitly or implicitly include one or more of that feature. The terms "comprising," "including," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0017] To address the aforementioned technical problems, this application provides a method for monitoring faults in intermediate joints of 35kV ring network cables in urban rail transit, referring to... Figure 1 , Figure 1 This is a flowchart illustrating a method for monitoring faults in intermediate joints of 35kV ring network cables in urban rail transit, provided as an embodiment of this application. The method is applied to a server and includes steps S110 to S160, as follows:

[0018] S110. Obtain the ring network topology data, switch position status data and cable mileage mapping data of the 35kV ring network cable of urban rail transit to construct the cable segmentation diagram. At the same time, generate a joint identification node for each intermediate joint in the cable segmentation diagram and generate a ring network status identifier corresponding to the current switch position status.

[0019] Specifically, a server refers to a centralized computing and data processing platform deployed in the urban rail transit power supply monitoring network or dedicated operation and maintenance network. It is used to carry all core algorithms and data storage functions, such as cable segmentation diagram construction, event evidence package management, propagation operator set operation, focus map generation, reversibility discrimination, structural fracture judgment, and joint-level positioning output. The ring network topology data is used to describe the primary electrical connections and node relationships of the 35kV ring network. It comes from primary wiring diagram data and equipment ledger data, and includes at least the connection relationships between site nodes, switch nodes, and cable segment edges. The switch position status data is used to describe the open or closed position of each switch node at the target time. It comes from remote signaling records and operation records of the monitoring system, and requires that each switch node be provided with a unique status bound to a timestamp. The cable mileage mapping data is used to convert the cable laying path into a locationable mileage coordinate expression. It comes from as-built drawings, mileage stake data, and joint construction records, and includes at least the mileage range of each cable segment edge and the joint mileage value corresponding to each intermediate joint. The intermediate joint refers to the connection part between two cable segments in the cable line, and the joint mileage value refers to the position of the joint along the cable laying path in the mileage coordinate system. To ensure consistency in subsequent diagram construction and segmentation, identifier unification is first performed. The equipment numbers in the primary wiring diagram, the equipment numbers in the ledger, and the remote signaling point numbers in the monitoring system are aligned to unified switch node identifiers and site node identifiers. The cable segment numbers in the as-built drawings are also aligned to unified cable segment edge identifiers, so that the ring network topology data, switch position status data, and cable mileage mapping data can be associated under the same identifier system.

[0020] The server defines site nodes as electrical aggregation nodes or endpoint nodes in a ring network, switch nodes as electrically disconnected or closed nodes that can change connectivity, and cable segment edges as cable line edges connecting two nodes. This forms the basic topology graph to support subsequent connected component analysis and subgraph extraction. The basic topology graph is expressed as a graph structure consisting of a set of nodes and a set of edges. The basic topology graph can be represented by the following graph structure expression:

[0021] in, Represents the basic topology graph; This represents a set of site nodes, which are used to characterize the electrical locations of traction substations, switching stations, sectioning stations, or ring main units within the station. This represents a set of switch nodes, which are used to characterize circuit breakers, load switches, or disconnectors that can be switched on or off in a ring network. This represents the set of cable segment edges, where each cable segment edge is used to characterize the electrical connection relationship of a section of cable line; This represents the set of intra-station connection edges. Each intra-station connection edge represents the electrical connection relationship formed by bus segments, bus couplers, or cabinet connections within the same site. Consistency verification is used to eliminate duplicate nodes and broken edges. Duplicate nodes refer to the situation where the same physical device is mapped to multiple node identifiers in different data sources. Broken edges refer to the situation where the nodes at both ends of an edge do not exist in the node set or the endpoint references of the edge are incomplete. Consistency verification is completed through identifier deduplication and endpoint integrity checks, ensuring that each node identifier is unique in the basic topology graph and that the endpoints of each side can be found in the node set. This ensures that subsequent partitioning and connected component resolution will not produce false connectivity or false disconnection due to data anomalies.

[0022] The server reads the mileage range of each cable segment edge and the set of joint mileage values ​​contained within that cable segment edge. It then sorts the joint mileage value set by mileage direction and performs a segmentation operation on the cable segment edge, dividing it into multiple sub-cable segment edges. A joint identifier node is inserted between every two adjacent sub-cable segment edges to indicate the presence of an intermediate joint at that location. This joint identifier node is a structured node specifically used for fault monitoring and location, serving both to identify the "joint-level object" on the graph and to provide an index for subsequent focused graphs. The segmentation and insertion can be described using the following formula to segment the mileage range of the cable segment edge:

[0023] in, This indicates the mileage range corresponding to the edge of the original cable segment. Indicates the starting mileage value. Indicates the destination mileage value; This represents the mileage value of the joint located inside the edge of the cable segment, and is ordered in ascending order of mileage; each of the values ​​in the set on the right... This indicates the mileage range of the sub-cable segment obtained after segmentation. After inserting the connector identifier node, the connector identifier node is connected to the sub-cable segment edges on both sides, thus forming a connection relationship of "sub-cable segment edge - connector identifier node - sub-cable segment edge" in the topology. Inheriting the loop identifier means that the connector identifier node inherits the loop identifier corresponding to the cable segment edge it belongs to, so as to maintain electrical consistency. Inheriting the laying environment segment attribute means that the connector identifier node inherits the laying environment segment attribute corresponding to its mileage location, so as to maintain environmental context consistency. The laying environment segment attribute is used to describe the environmental category such as tunnel segment, pipe gallery segment, trench segment or station segment, as well as environmental characteristics such as ventilation and humidity, so as to provide a basic label for subsequent "same laying environment segment condition" discrimination.

[0024] The server reads the switch position status of each switch node at the target time. Opening constraints indicate that the switch node is in the open position and its associated edges are disconnected. Closure constraints indicate that the switch node is in the closed position and its associated edges remain connected. Then, connected component analysis is performed on the constrained cable segment graph to extract the reachable subgraph set. Connected component analysis is used to identify whether there is a reachable path between any two nodes in the graph structure. To express the change in the edge set after applying constraints, the following formula describes the filtering relationship of the opening constraint on the edge set:

[0025] in, This represents the set of valid edges after applying the break constraint; This represents the set of edges in the cable segment diagram after the cable is cut and the connector identifier node is inserted. This represents the set of switch nodes that are in the quantile position; Indicates the connection between the switch node and the switch node. The set of associated edges, where an edge's endpoint contains the switch node. Connectivity resolution is performed on the valid set of edges. The corresponding graph structure yields several connected components, each corresponding to a reachable subgraph. The current ring network subgraph refers to the subgraph corresponding to the connected component that is identified as being in energized operation or within observable range at the target time. A subgraph identifier is generated for this current ring network subgraph, which uniquely identifies the topology of the subgraph and can be generated by combining the set of switch nodes participating in the subgraph with their switch position states. The set of switch nodes participating in the subgraph refers to the set of switch nodes that constrain connectivity within the subgraph and on its boundaries. The set of reachable connector identifier nodes refers to the set of all connector identifier nodes reachable through connectivity within the subgraph. The subgraph identifier, the set of switch nodes participating in the subgraph, and the set of reachable connector identifier nodes are written into the ring network status identifier, so that the ring network status identifier semantically carries both the "topological boundary of the current operating mode" and the "set of currently monitorable connector objects," thereby providing stable pre-constraints for subsequent bucketing and accumulating event evidence packets according to the ring network status identifier, constructing a set of propagation operators, and generating a focused graph.

[0026] S120. Based on the stations in the cable segmentation diagram, construct a synchronous observation array to simultaneously monitor transient events during the operation of the ring network, generate an event identifier for each transient event, and extract voltage and current segments from the event identifier within the synchronous observation array to form an event evidence package.

[0027] Specifically, a site node refers to a node in the cable segmentation diagram used to represent fixed electrical locations such as traction substations, switching stations, sectioning stations, or ring main units within the station. After extracting the site nodes, a unique site identifier is assigned to each site node. This site identifier is used to uniquely reference the site node during subsequent data acquisition, synchronization, and encapsulation processes. The site identifier is bound to the voltage transformer secondary side access point, the sheath grounding down conductor access point, the communication link identifier, and the timing link identifier. The voltage transformer secondary side access point refers to the physical terminal or sampling interface used to connect to the secondary circuit of the voltage transformer on the site side to acquire the 35kV side voltage observation signal. The sheath grounding down conductor access point refers to the physical terminal or sampling interface used to connect to the cable metal sheath connection point. The grounding lead is the current transformer terminal or shunt sampling interface for collecting the sheath return current observation signal. The communication link identifier refers to the transmission link number or network port identifier from the site-side acquisition device to the server. The time synchronization link identifier refers to the time synchronization interface identifier used by the site side to align with the unified time source. To ensure the binding relationship is available, access point verification is performed. Access point verification verifies the power frequency phase consistency of the access point on the secondary side of the voltage transformer, the return current direction consistency of the access point on the sheath grounding lead, the link connectivity of the communication link identifier, and the time source reachability of the time synchronization link identifier, so that the site identifier logically has both observation and synchronization capabilities.

[0028] The ring network status identifier is an identifier object used to uniquely represent the ring network topology reachability structure under the current switch position combination. It includes at least the set of reachable nodes and the set of reachable edges of the current ring network subgraph. Based on the ring network status identifier, candidate site identifiers located in the current ring network subgraph and in a reachable state are selected from all site identifiers. Spatial separation is used to constrain the interval coverage of candidate site identifiers on the mileage mapping, so that the array member site set spans multiple segments to form a sufficient array baseline. Topology coverage is used to constrain the cross-observation capability of candidate site identifiers on the cable segmentation diagram, so that the array member site set can simultaneously "see" the propagation structure changes of more connector identifier node neighborhoods. In actual implementation, the mileage interval between any two candidate site identifiers is first calculated and a mileage interval matrix is ​​formed. Then, the shortest topological distance between candidate site identifiers on the cable segmentation diagram is calculated and a topological distance matrix is ​​formed. Then, under the constraints of the minimum mileage interval threshold and the minimum topological distance threshold, at least three candidate site identifiers are selected as the array member site set, and a synchronous observation array identifier is assigned to the array member site set. The synchronous observation array identifier is used to uniquely reference the array configuration and bind it to the subsequent event evidence package.

[0029] The voltage observation channel refers to the complete acquisition link from the secondary side access point of the voltage transformer to the digital output of the site-side acquisition device. The current observation channel refers to the complete acquisition link from the sheath grounding lead access point to the digital output of the site-side acquisition device. Channel identifiers are assigned to the voltage observation channel and the current observation channel, and the sampling rate, analog front-end bandwidth, anti-aliasing filter parameters, and measurement range are configured to ensure that the channel configurations of different array member sites are consistent or comparable. When performing time synchronization of array member sites through the time synchronization link identifier, the local clock of each site-side acquisition device is disciplined using the same time source, so that the sampling sequence output by each site has a unified time axis meaning. After time synchronization, a time synchronization health mark is generated. The time synchronization health mark is used to characterize whether the synchronization quality of the array member site in the current period meets the cross-site related calculation requirements. The time synchronization health mark is obtained from the alignment error statistics, which can be defined by the following formula: in, Indicates the first Alignment error of each array member station; Indicates the first The reference time stamp timestamps output by each array member station under the time link identifier constraint; This represents the reference timestamp corresponding to the unified time source; alignment error statistics are used as the basis for determining the generation of time synchronization health markers, and the determination will be made by... If the time synchronization health flag is valid when compared with the preset alignment error threshold, the time synchronization health flag is invalid and the site is frozen from participating in event evidence package encapsulation in this period.

[0030] Transient events refer to short-term voltage and current waveform changes caused by switch operation, load disturbance, or external grid disturbance during 35kV ring network operation. Transient event monitoring rules are used to jointly trigger data streams and log streams. Switch position status changes are triggered by the state transition record of switch position status data, traction load changes are triggered by the load change record output by traction power supply monitoring, regenerative braking feedback is triggered by the regenerative energy feedback record, and external grid disturbances are triggered by the power quality record within the station or the alarm record of the upper-level grid. Event candidates refer to the suspected transient event objects formed by a single trigger. Event candidates are bound to the trigger source, trigger timestamp, and current ring network status identifier. Event windows refer to the time interval that can be used to extract segments by expanding a single event candidate. An event window must contain at least the window start time and the window end time. Multiple triggers of the same transient event within a short period of time are merged into the same event window by deduplication and merging. Then, a unique event identifier is assigned to each event window. The event identifier is used to achieve consistent extraction and consistent encapsulation across stations within the synchronous observation array. The event identifier must contain at least the synchronous observation array identifier, the trigger source identifier, and the event window start time index, thereby ensuring that the event identifier is unique and traceable globally.

[0031] The server locates the corresponding event window based on the event identifier and extracts voltage and current segments along a unified time axis on the voltage and current observation channels of each array member site. Each voltage and current segment includes a pre-transient baseline segment, a main transient response segment, and a transient tail segment. The pre-transient baseline segment characterizes the background noise and steady-state level before the transient occurs; the main transient response segment covers the main energy and propagation structure of the transient; and the transient tail segment covers the reflected wave train and diffusion attenuation structure. Segment quality assessment determines whether the extracted segments meet the requirements for constructing subsequent propagation operator sets. Segment quality assessment includes at least saturation detection, sample loss detection, bandwidth consistency detection, and noise floor anomaly detection. Saturation detection identifies whether sampling has reached its peak or clipped; sample loss detection identifies timestamp gaps; bandwidth consistency detection identifies whether channel configuration has been incorrectly changed; and noise floor anomaly detection identifies abnormal increases in background noise caused by strong external interference. Based on the segment quality assessment, a segment availability flag is output. This flag characterizes whether the voltage and current segments of the site under the given event identifier can participate in subsequent cross-site recorrelation superposition.

[0032] An event evidence package refers to a cross-site joint observation data carrier formed within a synchronous observation array for the same transient event. It is used for accumulation on the server side using the ring network status identifier as the bucket key and serves as the original input for constructing the propagation operator set. During encapsulation, an event evidence package identifier is assigned to the event evidence package, and the event evidence package identifier, the synchronous observation array identifier, the ring network status identifier, the array member site set, the timing health flag, and the availability flag of each site fragment are written into the event evidence package metadata, so that the event evidence package has traceable topological boundary conditions and synchronization quality evidence. When there is a site with a fragment availability flag of unavailable, the placeholder information of the site is retained and the missing mask is recorded. The missing mask is used to assign zero weight to the site in the subsequent construction of the propagation operator set without destroying the site order consistency, thereby ensuring that the event evidence package maintains structural consistency when superimposed across events and across sites and avoids introducing site dimension mismatch.

[0033] S130. Accumulate event evidence packets within the same ring network status identifier, construct a propagation operator set, and perform cross-site recorrelation superposition on the propagation operator set after performing amplitude-phase normalization and time-domain inversion processing to form a virtual playback field. Based on the virtual playback field, generate a focused map indexed by the connector identifier node.

[0034] Specifically, the server uses each ring network status identifier as the key for the status bucket index, and maintains the event evidence packet queue, array member site set, event source identifier distribution, and time coverage distribution under this key, ensuring that subsequent overlay operations are always performed within the same topology boundary. When performing timing health marking and fragment availability marking filtering on the event evidence packet queue, it not only determines whether the timing health marking and fragment availability marking are valid, but also introduces an "array consistency score" to suppress the pollution of the overall structure by single-site clock drift or single-site saturation. The array consistency score can be characterized by cross-site phase consistency and energy consistency, and is used together with site coverage as a filtering condition for the available event evidence packet set. The expression for the array consistency score is:

[0035] in, This represents the array consistency score, with a value range of [value range missing]. The closer the value is to 1, the more consistent the cross-site structure is; This indicates the number of array member stations within the synchronous observation array; Indicates the first The valid mask for each site identifier is determined by the timing health flag and the fragment availability flag. If the requirement is met, the mask is set to 1; otherwise, it is set to 0. Indicates the first Each site is identified by the feature vector extracted from the event evidence package. The feature vector can be obtained by concatenating the complex analytical representations of voltage segments and current segments in the transient main response segment. This represents the L2 norm. When incorporating an event evidence package into the set of available event evidence packages, it is required that the site coverage rate is not lower than a preset threshold and the array consistency score is not lower than a preset threshold, thereby ensuring that the event evidence packages constructed by entering the propagation operator set are both "sufficiently covered by sites" and "consistent across sites".

[0036] The array configuration consistency processing uses the order of the reference station identifiers bound to the synchronous observation array identifier as the unique order. For each event evidence packet, the multi-site voltage and current segments are rearranged according to this order, and empty segments are filled at missing station locations. Simultaneously, a missing mask vector is generated to record the missing positions, ensuring that each event evidence packet is standardized into operator samples of a uniform dimension. Furthermore, to make the propagation operator set more robust to noise and strong interference, the operator samples are elevated from time-domain signals to a "time-frequency complex coefficient matrix." This involves performing a short-time Fourier transform on the voltage and current segments of each station identifier to form a complex spectrum, and then stacking the voltage and current complex spectra according to the station and channel dimensions to form an operator sample tensor. The expression for the short-time Fourier transform is:

[0037] in, Indicates the first Each station is identified in the channel. The time-frequency complex coefficients on, Used to distinguish between voltage observation channels and current observation channels; Indicates the first The original sampling sequence of each site identifier at the sampling point The amplitude at that point; Indicates the window function; Indicates the index of the window's center position; Indicates a discrete frequency point index; Indicates the length of the transformation window; The imaginary unit is represented. Based on this time-frequency representation, each event evidence packet is constructed into an operator sample and aggregated under the ring network state identifier to form a propagation operator set. This propagation operator set not only contains multi-site information but also phase structure information of multi-frequency points, providing a higher resolution computational space for subsequent amplitude-phase normalization, time-domain inversion, and recorrelation superposition.

[0038] Amplitude-phase normalization does not use a single energy scale, but instead employs "cross-site covariance whitening normalization," thereby simultaneously suppressing the contamination of the correlation structure by channel gain differences, event source intensity differences, and strong interference frequencies. During implementation, within each event evidence packet, this is applied to each frequency point. Construct cross-site observation vectors and estimate their covariance matrices, then compute a whitening matrix to map the observation vectors to a unit covariance space. The expression for the cross-site observation vectors is:

[0039] in, Indicates channel At time and frequency points Cross-site complex vectors; This represents the missing mask and the valid mask obtained through quality screening. This indicates transpose. The expressions for the covariance matrix and whitening matrix are:

[0040] in, Indicates channel At frequency The cross-site covariance matrix; This represents the time index set used for calculating covariance, and is taken from the window index set corresponding to the transient main response segment. Represents a set The number of elements; Indicates conjugate transpose; Represents the identity matrix; This represents the regularization coefficient, used to avoid ill-conditioned covariance matrix; its value is determined by the noise floor level and the number of stations. The inverse square root operation of a matrix can be obtained through eigenvalue decomposition or singular value decomposition. The whitened and normalized cross-site vector is... The significance lies in transforming the cross-site correlation structure from "amplitude-dominated" to "phase and propagation structure-dominated." In the time-frequency domain, time-domain inversion is equivalent to performing a conjugate and time-inverse index transformation on the complex spectrum of the mirror site identifier, and constructing a time-inversion input using the mirror site identifier as a reference, so that subsequent cross-correlation is equivalent to performing a phase transformation of generalized cross-correlation. The time-frequency inversion segment of the mirror site identifier can be written as:

[0041] in, Indicates the mirror site identifier In the passage The inverted time-frequency complex coefficients on; This represents the complex coefficient of the mirror site identifier after whitening and normalization; Indicates complex conjugation; This indicates the inversion alignment offset, used to unify the inversion reference zero point. Its value is determined by the center of the event window or the position of the transient main peak. This process ensures that when the inverted segment is cross-correlated with other station identifier positive segments, it can be superimposed in the same reference system to form a focus.

[0042] When performing cross-site recorrelation superposition based on the inverted fragment of the mirror site identifier and the forward fragment of the remaining site identifier to form a virtual playback field, simple cross-correlation superposition is not used. Instead, a "generalized cross-correlation phase transformation and event balance weighting" recorrelation superposition is adopted to ensure that the virtual playback field remains sensitive to the propagation structure even in strong interference scenarios. During implementation, the generalized cross-correlation is calculated for each event evidence packet, each channel, and each site pair, and a phase weighting function is applied in the frequency domain to highlight the propagation delay structure. The expression for the generalized cross-correlation is:

[0043] in, Indicates the mirror site identifier With site identifier In the passage The following generalized cross-correlation sequences; This indicates the set of relevant frequency points selected after removing power frequency and its strong harmonic energy ranges. This represents the phase weighting function, used to enhance sensitivity to specific frequency bands or specific propagation characteristics, and can be determined by noise spectrum estimation and frequency band reliability scoring; This represents the stable term, used to avoid the denominator being zero; its value is obtained from the noise floor statistics. This represents the discrete latency index. Subsequently, a missing mask weighted overlay is performed at the site level, and an event-balanced weighted overlay is performed at the event level to obtain the virtual replay field. The event-balanced weighting is performed by grouping events by event source identifier and then normalizing the events within each group, ensuring that no single type of event source identifier dominates the virtual replay field due to numerical superiority. Its expression is:

[0044] in, Indicates channel Virtual replay field in time delay The superposition result at the location; Represents the set of event source identifiers; The event source identifier is The set of event evidence packages indexes; Indicates the size of the set; Indicates the index of the event evidence package; Indicates evidence package of the incident China Station Identification Valid mask; Indicates evidence package of the incident The selected mirror site identifier; Indicates evidence package of the incident The generalized cross-correlation of corresponding site pairs is then established. When constructing the topological path set between site identifiers and joint identifier nodes based on the cable segmentation diagram, the topological path set provides a unique or shortest path expression for each site identifier and each joint identifier node. This path expression is then converted into a propagation window constraint, which limits the time delay search range corresponding to that path in the virtual playback field. When projecting the virtual playback field onto the joint identifier node, not only is the main peak selected, but a focusing index set is also jointly formed using "focused energy—morphological sparsity—sidelobe suppression—cross-channel consistency," thus making it more sensitive to the joint-level scattering center. (Using the joint identifier node...) The focused energy for the index can be expressed as:

[0045] in, Indicates the connector identification node Focused energy; This represents a set of channels, including voltage observation channels and current observation channels; Indicates the connector identification node The set of delay indices corresponding to the propagation window constraints; This represents a delay weighting function used to enhance reliable delays near the window center and suppress reflection interference at the window edges. The weights are determined by path length uncertainty and noise level. The focused peak position is achieved through... Searching inside The maximum value position is obtained; the focused peak width is obtained through the energy concentration in the neighborhood of the main peak or the equivalent second moment; the peak sidelobe ratio is obtained through the ratio of the main peak amplitude to the statistical amplitude of the sidelobes; and the focused peak energy integral is obtained through... The above indicators are written into the focusing map using the joint identifier node as the index, and the focusing map, along with the ring network status identifier, propagation operator set version number, event source identifier distribution, and whitening parameters, are archived together. This allows subsequent reversibility judgment and structural fracture judgment to be strictly compared under the same topological constraints and the same calculation conditions, avoiding pseudo-structural changes caused by parameter drift or event source bias.

[0046] S140. Under the same current-carrying range and the same laying environment, perform a reversibility judgment on the focusing pattern. When the focusing index set of the same joint identification node cannot be returned to the center of the historical focusing baseline group, mark the corresponding joint identification node as a saturation critical state candidate, and generate the first focusing pattern and the second focusing pattern based on the voltage segment and the current segment, respectively.

[0047] Specifically, the operating context label consists of a current-carrying interval label and a laying environment segment label. The current-carrying interval label is used to manage the focusing structure differences of the same joint identifier node under different current-carrying levels in separate buckets. The laying environment segment label is used to manage the focusing structure differences of the same joint identifier node under different laying environment segment conditions in separate buckets. The current-carrying interval label is obtained by interval quantization of the effective current value of the cable segment adjacent to the joint identifier node within the time window corresponding to the event evidence packet. The laying environment segment label directly reads the laying environment segment attribute bound to the joint identifier node in the cable segmentation diagram and uses the stability verification result as the validity gate when environmental monitoring data exists. Under the same operating context label and the same ring network status identifier, the focusing index set sample corresponding to the joint identifier node is extracted from the historical focusing map database to construct the historical focusing baseline cluster. The historical focusing baseline cluster refers to the statistical cluster of focusing indicators that can be repeated under the same topological boundary and the same operating context. The cluster center is used to characterize the typical structural position of the cluster, and the cluster discrete range is used to characterize the normal fluctuation boundary of the cluster. To make the population center and population dispersion range more robust to outliers, each focused indicator set sample is represented as an eigenvector, and a weighted robust estimation is used to obtain the population center and population covariance matrix. The expression for the focused indicator set sample vector is:

[0048] in, Indicates the connector identification node In the time window The following is a sample vector of the focus indicator set; This indicates the focused peak position, and its value is obtained from the time delay index of the main peak within the propagation window constraint. This represents the width of the focused peak, and its value is estimated from the equivalent width of the energy distribution in the neighborhood of the main peak. This represents the energy integral of the focused peak, and its value is obtained by summing the energies of the main peak and its neighborhood. This represents the peak-to-side-lobe ratio, which is obtained by taking the ratio of the main peak amplitude to the statistical amplitude of the side lobes. This indicates the transpose. The expressions for the population center and population covariance matrix are:

[0049] in, Indicates the connector identification node In the running context label The ethnic center below; Indicates the connector identification node In the running context label The population covariance matrix below; This indicates that the same runtime context label is satisfied. And it satisfies the historical time window index set with the same ring network status identifier; This represents robust weights used to reduce the impact of outliers on the population center and the population covariance matrix. The values ​​can be obtained by iteratively updating the Mahalanobis distance of the samples and are limited to a preset range. This represents the regularization coefficient, used to avoid ill-conditioned covariance matrix; its value is determined by the number of historical samples and the noise floor level. The identity matrix is ​​represented. The population discrete range is expressed as an "ellipsoidal discrete range," which is an equidistant surface centered on the population center and shaped by the population covariance matrix, serving as the normal fluctuation boundary. The discrete range threshold is determined by the Mahalanobis distance quantiles of historical samples. The expressions for Mahalanobis distance and quantile threshold are:

[0050] in, Indicates historical time window Mahalanobis distance of the sample vector relative to the population center; The matrix representing the inverse of the population covariance matrix; Indicates the running context label The discrete range threshold below; This represents the quantile function, used to select the quantile from the historical Mahalanobis distance set. quantile value as threshold The value is used to control the false alarm rate and can be set according to the operation and maintenance risk strategy.

[0051] Under the constraints of the current ring network status identifier and the current operating context label, the server extracts the connector identifier node from the current focused map. In each current time window Corresponding sample vector Then, the Mahalanobis distance of the sample vector relative to the population center is calculated, and it is determined whether it falls within the population's discrete range. If it falls within the population's discrete range, it is considered regression; if it continuously falls outside the population's discrete range, it is considered deviation. To reflect the constraints of "deviation within multiple consecutive time windows" and "failure to regress after the running context label is restored to the same conditions," a deviation indicator sequence and a regression gating sequence are introduced in implementation. The deviation indicator sequence is used to characterize continuity, and the regression gating sequence is used to characterize the non-regression of failure to regress after the running context label is restored. The expression for the deviation indicator sequence is:

[0052] in, Indicates the connector identification node In the time window The deviation indicator below has a value of 1 indicating deviation from the group's discrete range, and a value of 0 indicating falling within the group's discrete range. Indicates an indicator function; and The Mahalanobis distance and runtime context label of the current sample are respectively. The discrete range threshold is defined below. The intensity of continuous deviation employs exponential forgetting accumulation to enhance sensitivity to persistent deviations and suppress isolated deviations. The expression for the intensity of continuous deviation is:

[0053] in, Indicates the connector identification node In the time window The continuous deviation intensity below, with a value range of Furthermore, the closer the value is to 1, the more persistent the deviation. This represents the forgetting factor, used to balance the contributions of historical deviation and current deviation to the strength of continuous deviation. A larger value indicates a greater emphasis on historical continuity; initial value It can be set to 0. The runtime context label recovery gating is used to explicitly state "failure to revert after restoring to the same conditions," and during implementation, the current time window will be used. The runtime context label and the target runtime context label Consistency is encoded as a gating variable, and the strength of consecutive deviations is checked within the interval where the gating variable is 1 to see if it still exceeds a preset threshold. The expression for the gating variable is:

[0054] in, Indicates the connector identification node In the time window The runtime context label recovery gate is set to 1, indicating the current runtime context label. With target running context label Consistency is indicated by a value of 0, while inconsistency is indicated by a value of 0. The reversibility criterion uses the simultaneous fulfillment of "deviation persistence" and "non-regression after recovery" as the non-reversible offset condition, meaning that a continuous set of time windows must satisfy this condition. still meet the requirements If the value is higher than the continuous deviation threshold and the state continues to exceed the preset duration threshold, the joint identifier node is marked as a saturated critical state candidate. The saturated critical state candidate refers to a candidate object that has already shown changes in propagated structural memory in the abrupt degradation path but has not yet experienced structural fracture. The marking is used to drive subsequent cross-channel consistency verification.

[0055] The first propagation operator set is constructed solely from multi-site voltage segments in the event evidence package, and the second propagation operator set is constructed solely from multi-site current segments in the event evidence package. Both sets maintain the same ring network state identifier, the same synchronous observation array identifier, the same event identifier set, the same array configuration consistency processing results, and the same missing mask, ensuring that the differences between the two generation processes originate only from the observation channels and not from topological boundaries or event sets. Within the first propagation operator set, amplitude and phase normalization, mirror site identifier selection, time domain reversal, and cross-site recorrelation superposition are performed in accordance with the propagation operator set to form a first virtual playback field. Based on the first virtual playback field, a first focused spectrum is obtained by projecting it under the same topological path set and the same propagation window constraints. Within the second propagation operator set, a second virtual playback field is formed in the same manner and projected to obtain a second focused spectrum, thus making the first and second focused spectra characterize the structural response of the same joint identifier node under different observation channels. By reconstructing the sub-channels, it is possible to check whether the focusing structure shift of the saturated critical state candidate has cross-channel consistency in subsequent judgments, thereby suppressing pseudo-saturated critical state candidates caused by single observation channel anomalies, single link drifts, or single interference sources, and improving the reliability of the discrimination of abrupt degradation precursors caused by space charge trap saturation.

[0056] S150. When the first focused map and the second focused map show structural fractures in the neighborhood of the same joint identifier node, and the structural fractures meet the synchronization condition, the corresponding joint identifier node is determined as a faulty joint node, and a fault status marker bound to the faulty joint node is generated.

[0057] Specifically, a window management table is maintained for each joint identifier node on the server side. The window management table records the window start construction period, the window end construction period, the version number range of the propagation operator set within the window, and the event identifier set range within the window. The construction period refers to the discrete time unit of the focused map update according to the rolling cycle. The ring network status identifier binding means that the window management table only receives the focused map update results with consistent ring network status identifiers. Once a change in the ring network status identifier is detected, the current structural fracture observation window is closed and a new structural fracture observation window is opened, thereby avoiding misjudging the focused structure changes caused by switching or operation mode adjustment as structural fractures. The joint identifier node neighborhood is used to limit the spatial range. The joint identifier node neighborhood is defined as the set of neighboring nodes on the cable segment diagram whose topological distance from the joint identifier node does not exceed a preset hop count threshold. The topological distance refers to the number of edges traversed along the cable segment edge from one node to another. The neighborhood node set is used to stably characterize the neighborhood focused structure in subsequent discontinuous migration detection rather than relying solely on the fluctuations of a single node.

[0058] The server extracts the focused peak position, focused peak width, focused peak energy integral, and peak-top sidelobe ratio of each node in the neighborhood node set for each construction cycle, and arranges them into a neighborhood feature matrix according to a fixed order of neighborhood nodes. Then, the neighborhood feature matrices of multiple construction cycles are stacked in time to form a neighborhood focusing sequence. The peak position family is used to describe the spatial concentration pattern of the focused peak positions in the neighborhood, the morphology family is used to describe the joint morphological pattern of the focused peak width and peak-top sidelobe ratio, and the energy family is used to describe the energy concentration and gradient pattern of the focused peak energy integral in the neighborhood. The implementation method of discontinuous migration detection is to simultaneously construct a "pre-fracture structural prototype" and a "post-fracture structural prototype", and measure the degree of jump and stability of the structural prototype between adjacent construction cycles. When the degree of jump remains stable in multiple consecutive construction cycles and the regression discrimination fails, it is determined as a structural fracture. To make discontinuous migration detection more robust to noise, the neighborhood focusing structure of each construction cycle can be mapped into a vector and the robust Mahalanobis distance can be calculated as a jump metric. The expression of the neighborhood structure vector is:

[0059] in, Indicates the connector identification node During the construction cycle The neighborhood structure vector; The peak family matrix is ​​composed of the focused peaks of each node in the neighborhood node set in a fixed order; The morphological family matrix is ​​composed of the focal peak width and peak sidelobe ratio of each node in the neighborhood node set in a fixed order; The energy group matrix is ​​composed of the focused peak energy integrals of each node in the neighborhood node set in a fixed order. This represents a vectorization operator used to expand a matrix into column vectors according to fixed rules. This indicates transpose. The pre-fracture structural prototype uses the weighted robust center of the initial construction period within the structural fracture observation window, while the post-fracture structural prototype uses the weighted robust center of the subsequent construction period. The expression is:

[0060] in, This represents the structural prototype before fracture. Indicates the prototype structure after fracture; This represents the set of construction cycle indices for the pre-fracture segment within the structural fracture observation window; This represents the set of construction cycle indices for the later stages of a fracture within the structural fracture observation window. This represents a robust weight used to reduce the impact of outliers on the structural prototype. Its value is iteratively updated by the distance from the neighboring structural vector to the prototype and limited to a preset range. The criterion for stable transitions consists of "large transition amplitude" and "stability after the transition." The transition amplitude is characterized by the Mahalanobis distance between the pre-fracture and post-fracture structural prototypes, while post-transition stability is characterized by the average Mahalanobis distance from the post-fracture segment sample to the post-fracture structural prototype. The relevant expressions are:

[0061] in, This represents the magnitude of the jump; the larger the value, the more significant the difference before and after the fracture. This represents the stability index after a jump; the smaller the value, the more stable the structure is after fracture. This represents the neighborhood structure covariance matrix estimated by the historical focused baseline population under the same running context label and the same ring network state identifier. The neighborhood structure covariance matrix is ​​used to give the relative weights of the changes in each dimension and suppress the noise dimension. This represents the number of elements in the index set of the post-fracture construction cycle. The determination of whether regression to the historical focused baseline population is possible is achieved by comparing the neighborhood structure vector of the post-fracture segment with the center of the historical focused baseline population using Mahalanobis distance. If the post-fracture segment exceeds the historical discrete range threshold in multiple consecutive construction cycles, it is considered irreversible. Exceeding the jump threshold and When the condition falls below the stability threshold and cannot be reversed, it is considered a structural fracture.

[0062] When performing synchronization condition verification on the structural fractures of the first and second focused maps, the synchronization condition is used to confirm that the two observation channels have a common time starting point and a common spatial orientation for structural fractures in the neighborhood of the same joint identifier node, avoiding the triggering of pseudo-structural fractures caused by single-channel link anomalies, single-channel interference, or single-channel threshold drift. Time synchronization means that the difference in the fracture starting point construction period between the first and second focused maps does not exceed the preset tolerance. The fracture starting point construction period is defined as the construction period index that first meets the structural fracture criterion. Spatial synchronization means that the subsets of the main peaks after fracture in the neighborhood of the joint identifier node of the first and second focused maps have significant overlap. In practice, the fracture starting point construction period and the subsets of the main peaks after fracture are calculated for the first and second focused maps respectively, and verified by time difference and set similarity. Time synchronization can be described by the following formula:

[0063] in, Indicates the connector identification node The fracture initiation point has a different construction cycle; This indicates the periodic index for constructing the fracture initiation point under the first focused spectrum; This indicates the periodic index constructed from the fracture initiation points under the second focal map. Spatial synchronicity can be described by the weighted Jaccard similarity of subsets within the neighborhood main peak set:

[0064] in, This represents the spatial synchronization score, with a value range of [value missing]. The closer it is to 1, the more consistent the spatial pointers are; Indicates the connector identification node The connector identifier node neighborhood; This indicates that the first focused graph is in the neighborhood nodes. The focusing weight after fracture is determined by the robust mean of the energy integral of the focusing peak within the fracture window. This indicates that the second focused graph is in the neighborhood nodes. The weighting of the focus after the break. When the synchronization condition is satisfied, it is also required that... Not exceeding the preset time tolerance and The similarity threshold is not lower than the preset threshold, and the connector identifier nodes that meet the conditions are judged as faulty connector nodes. Faulty connector nodes refer to connector identifier node objects that have completed cross-channel synchronization verification and meet the requirements of non-continuous migration and non-reversal, which are used to trigger the generation of fault status markers and subsequent positioning output.

[0065] The server generates a unique fault status marker for each faulty joint node and writes the joint identifier node, fracture initiation timestamp, structural fracture observation window identifier, ring network status identifier, event identifier set, and fracture summaries from the first and second focused maps into the fault status marker. The fracture initiation timestamp is obtained by mapping the fracture initiation construction cycle index to the real time. The structural fracture observation window identifier refers to the unique identifier of the corresponding window in the window management table. The event identifier set is used to reference the event evidence package set involved in fracture determination. The fracture summaries from the first and second focused maps are used to solidify key quantities such as the pre-fracture structural prototype, post-fracture structural prototype, jump amplitude index, post-jump stability index, and spatial synchronization score. The fault status marker is archived together with the version number of the corresponding propagation operator set, the version number of the focused map, and the running context label, so that the fault status marker can fully reflect the state migration process from the historical focused baseline group to the structural fracture confirmation in the life cycle archive.

[0066] S160. Based on the fault status marking, determine the target joint identification node on the cable segmentation diagram, and output the intermediate joint fault monitoring result corresponding to the target joint identification node.

[0067] Specifically, the fault status marker is a solidified carrier of the fault joint node judgment conclusion and its evidence chain index. Among them, the ring network status marker is used to uniquely characterize the topological boundary conditions corresponding to the fault occurrence and judgment time. The structural fracture observation window marker is used to uniquely reference the construction cycle interval and event marker set range participating in the fracture judgment. The first focused map fracture summary and the second focused map fracture summary are used to provide key quantities such as the fracture starting point construction cycle, the structural prototype before fracture, the structural prototype after fracture, and the spatial synchronization score. During implementation, the historical version of the cable segment diagram is retrieved on the server side through the index field of the fault status marker and the graph snapshot consistent with the ring network status marker is read. Then, the reachable node set and reachable edge set corresponding to the ring network status marker are extracted from the graph snapshot to form the ring network subgraph. The ring network subgraph refers to the subgraph range of electrical connections that can be reached under the current switch position state constraint, thereby strictly limiting the subsequent positioning calculation to the reachable range and avoiding false positioning caused by including unreachable joint marker nodes in the candidate set.

[0068] The comprehensive fracture strength index is used to characterize the "fracture saliency" of each joint marker node within the structural fracture observation window. It requires simultaneously reflecting consistent fracture components from both the first and second focused spectra to suppress single-channel spurious fractures. In practice, for each joint marker node, the structural vectors of the post-fracture neighborhood and the pre-fracture neighborhood are extracted from both the first and second focused spectra. Fracture jump amplitude, post-fracture stability, and irreversible strength are calculated, and these quantities are fused into a comprehensive fracture strength index by channel. During the fusion process, time synchronization gating and spatial synchronization gating are introduced, ensuring that only fracture components that simultaneously meet the gating conditions in both channels are included in the comprehensive fracture strength index. The comprehensive fracture strength index can be expressed by the following formula:

[0069] in, Indicates the connector identification node The comprehensive fracture strength index, the larger the value, the more significant the fracture; The jump amplitude index under the first focal spectrum is obtained by the Mahalanobis distance between the pre-fracture and post-fracture structural prototypes. This indicates the jump amplitude index under the second focus spectrum; This represents the post-fracture stability index under the first focused spectrum; the smaller the value, the more stable the structure after fracture. This indicates the post-fracture stability index under the second focused spectrum. This represents the spatial synchronization score between the first and second focused maps in the neighborhood of the connector marker node, with a value range of [value range missing]. ; This indicates the cross-channel consistency strength, used to characterize the degree of consistency between two channels in terms of the fracture initiation period, peak migration direction, and energy concentration region migration direction. Its value range is... And the larger the size, the more consistent they are; This represents the weighting coefficient, the value of which is determined by the historical validation set or operational risk strategy. The fracture weighting graph is a graph structure with joint identifier nodes as the node set and the comprehensive fracture strength index as the node weight. In implementation, the joint identifier nodes in the ring network subgraph are used as nodes in the fracture weighting graph, and the adjacency relationships formed between joint identifier nodes in the ring network subgraph through cable segment edges or intermediate connections are used as edges in the fracture weighting graph, thus obtaining a weighted graph that can perform connected subgraph searches under the same topological constraints.

[0070] A connected subgraph refers to the set of nodes in the fracture weight graph where any two nodes are connected by a path. Preset conditions are used to simultaneously constrain "fracture strength concentration" and "spatial connectivity compactness" to prevent multiple dispersed high-weight nodes from being mistakenly merged into the same fault region. During implementation, the fracture weight graph is first truncated using a threshold, retaining joint identifier nodes whose comprehensive fracture strength index exceeds the preset weight threshold to form a candidate node set. Then, connected component analysis is performed on the candidate node set to obtain multiple connected subgraphs. For each connected subgraph, the sum of its weights and its diameter are calculated. The diameter characterizes the spatial span of the connected subgraph in terms of topological distance, thus prioritizing connected subgraphs with high total weights and small diameters as the target joint set. The selection score for connected subgraphs can be expressed by the following formula:

[0071] in, Representing a connected subgraph Selection rating; Represents the set of connector nodes in a connected subgraph; This represents the sum of the comprehensive fracture strength indices within the connected subgraph; represents the diameter of the connected subgraph, and its value is the maximum of the shortest topological distance between any two nodes within the connected subgraph. After obtaining the target joint set, to determine the unique target joint identifier node, the peak overlap and energy concentration region consistency of the first and second focused spectra within the target joint set are further calculated. Peak overlap characterizes the degree to which the main peak positions of the two channels point to the same joint identifier node after fracture, and energy concentration region consistency characterizes the degree to which the energy concentration regions of the two channels cover the neighborhood of the same joint identifier node after fracture. During implementation, the peak overlap score and energy consistency score are calculated for each joint identifier node within the target joint set and jointly ranked with the comprehensive fracture strength index. The joint identifier node with the highest score is selected as the target joint identifier node. The peak overlap score can be expressed by the following formula:

[0072] in, Indicates the connector identification node The peak overlap score, with a value range of [value missing]. Furthermore, the closer the value is to 1, the more overlapping the peak positions are. Indicates the connector identifier node under the first focused spectrum. The peak position index of the main peak after fracture is obtained from the position of the main peak within the propagation window constraint; Indicates the connector identifier node under the second focused spectrum. Index of the main peak position after fracture; This represents the peak position tolerance scale, used to map peak position differences to similarity. Its value is determined by the sampling rate, propagation window width, and topological path length uncertainty. The consistency score for energy concentration regions can be defined using weighted Jaccard similarity:

[0073] in, Indicates the connector identification node The consistency score of the energy concentration region, with a value range of [value missing]. Furthermore, the closer the value is to 1, the more consistent the energy concentration area. Indicates the connector identification node The connector identifier node neighborhood; This indicates that the first focused graph is in the neighborhood nodes. The energy weight after fracture is obtained by taking the robust mean value of the energy integral of the focused peak within the fracture window; This indicates that the second focused graph is in the neighborhood nodes. The energy weight after fracture. The target joint identifier node can ultimately be determined using a joint sorting score, expressed as:

[0074] in, Indicates the connector identification node The joint ranking score within the target connector set indicates that a higher value signifies a higher priority in being selected as the target connector identifier node; this is achieved by analyzing the identifier nodes within the target connector set. Sort the data and select the node corresponding to the maximum value to achieve unique identification.

[0075] Mileage mapping information is used to convert the node location of the target joint identifier node in the cable segmentation diagram into an executable mileage location on site. The laying environment segment attribute is used to provide environmental context for operation and maintenance and for subsequent statistical analysis of similar faults. During implementation, the joint mileage value of the target joint identifier node and the mileage range of the adjacent cable segment are read from the cable mileage mapping data, and the laying environment segment label bound to the target joint identifier node is read from the cable segmentation diagram. The fault status mark is written as an evidence index into the intermediate joint fault monitoring result, so that the monitoring result can be traced back to the structural fracture observation window identifier, event identifier set, and fracture summary of the first focus map and the fracture summary of the second focus map. The intermediate joint fault monitoring result is generated in the form of structured record on the server side and sent to the operation and maintenance management system through the communication interface. At the same time, the target joint identifier node is located and marked on the cable segmentation diagram visualization interface, and the mileage mapping information and laying environment segment attribute are displayed, thereby realizing closed-loop delivery from structural fracture judgment to joint-level location output, and ensuring that the output result has topological constraint consistency, cross-channel evidence consistency, and location executability.

[0076] This application also provides a fault monitoring device for intermediate joints of 35kV ring network cables in urban rail transit, referring to... Figure 2 , Figure 2This application provides a schematic diagram of a fault monitoring device for intermediate joints of 35kV ring network cables in urban rail transit. The device is a server, comprising an acquisition module 21 and a processing module 22. The acquisition module 21 acquires ring network topology data, switch position status data, and cable mileage mapping data of the 35kV ring network cable in urban rail transit to construct a cable segmentation diagram. It also generates a joint identifier node for each intermediate joint in the cable segmentation diagram and generates a ring network status identifier corresponding to the current switch position status. The processing module 22 constructs a synchronous observation array based on the stations in the cable segmentation diagram, simultaneously monitors transient events during ring network operation, generates an event identifier for each transient event, and extracts voltage and current segments from the event identifiers within the synchronous observation array to form an event evidence package. The processing module 22 also accumulates event evidence packages within the same ring network status identifier, constructs a propagation operator set, and performs amplitude-phase normalization and time-domain inversion processing on the propagation operator set before further processing. The processing module 22 is used to perform regression discrimination on the focus map under the same current carrying range and the same laying environment. When the focus index set of the same joint identification node cannot be returned to the center of the historical focus baseline group, the corresponding joint identification node is marked as a saturation critical state candidate, and the saturation critical state candidate is used to generate a first focus map and a second focus map based on voltage segment and current segment, respectively. The processing module 22 is also used to determine the corresponding joint identification node as a faulty joint node when the first focus map and the second focus map have structural fracture in the neighborhood of the same joint identification node and the structural fracture meets the synchronization condition, and generate a fault status mark bound to the faulty joint node. The processing module 22 is also used to determine the target joint identification node on the cable segmentation diagram according to the fault status mark, and output the intermediate joint fault monitoring result corresponding to the target joint identification node.

[0077] This application also provides an electronic device, with reference to... Figure 3 , Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. The electronic device may include: at least one processor 31, at least one network interface 34, a user interface 33, a memory 35, and at least one communication bus 32.

[0078] The communication bus 32 is used to enable communication between these components.

[0079] The user interface 33 may include a display screen and a camera. Optionally, the user interface 33 may also include a standard wired interface and a wireless interface.

[0080] The network interface 34 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface).

[0081] The processor 31 may include one or more processing cores. The processor 31 connects to various parts of the server via various interfaces and lines, executing instructions, programs, code sets, or instruction sets stored in the memory 35, and calling data stored in the memory 35 to perform various server functions and process data. Optionally, the processor 31 may be implemented using at least one hardware form of Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), or Programmable Logic Array (PLA). The processor 31 may integrate one or a combination of several of the following: Central Processing Unit (CPU), Graphics Processing Unit (GPU), and modem. The CPU primarily handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing the content to be displayed on the screen; and the modem handles wireless communication. It is understood that the modem may also not be integrated into the processor 31 and may be implemented as a separate chip.

[0082] The memory 35 may include random access memory (RAM) or read-only memory. Optionally, the memory 35 may include a non-transitory computer-readable storage medium. The memory 35 can be used to store instructions, programs, code, code sets, or instruction sets. The memory 35 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as touch function, sound playback function, image playback function, etc.), instructions for implementing the above-described method embodiments, etc.; the data storage area may store data involved in the above-described method embodiments, etc. Optionally, the memory 35 may also be at least one storage device located remotely from the aforementioned processor 31. Figure 3 As shown, the memory 35, which serves as a computer storage medium, may include an operating system, a network communication module, a user interface module, and an application program for a method of monitoring faults in intermediate joints of 35kV ring network cables for urban rail transit.

[0083] exist Figure 3 In the electronic device shown, the user interface 33 is mainly used to provide an input interface for the user and to obtain the user input data; while the processor 31 can be used to call the application program stored in the memory 35, which is a method for monitoring faults in intermediate joints of 35kV ring network cables for urban rail transit. When executed by one or more processors, the electronic device executes one or more methods as described in the above embodiments.

[0084] This application also provides a non-transitory computer-readable storage medium storing instructions. When executed by one or more processors, these instructions cause an electronic device to perform one or more of the methods described in the above embodiments.

[0085] The foregoing description is merely an exemplary embodiment of this disclosure and should not be construed as limiting the scope of this disclosure. Any equivalent changes and modifications made in accordance with the teachings of this disclosure shall still fall within the scope of this disclosure. Those skilled in the art will readily conceive of other embodiments of this disclosure upon considering the specification and the disclosure of practical truth. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not described in this disclosure. The specification and embodiments are considered exemplary only, and the scope and spirit of this disclosure are defined by the claims.

Claims

1. A method for monitoring faults in intermediate joints of 35kV ring network cables for urban rail transit, characterized in that, The method includes: The system acquires the ring network topology data, switch position status data, and cable mileage mapping data of the 35kV ring network cable for urban rail transit to construct a cable segmentation diagram. At the same time, it generates a joint identification node for each intermediate joint in the cable segmentation diagram and generates a ring network status identifier corresponding to the current switch position status. Based on the stations in the cable segmentation diagram, a synchronous observation array is constructed to simultaneously monitor transient events during the operation of the ring network. An event identifier is generated for each transient event, and voltage and current segments are extracted from the event identifier within the synchronous observation array to form an event evidence package. Accumulate the event evidence packets within the same ring network status identifier to construct a propagation operator set. After performing amplitude-phase normalization and time-domain inversion processing on the propagation operator set, perform cross-site recorrelation superposition to form a virtual playback field. Based on the virtual playback field, generate a focusing map indexed by the connector identifier node. Under the same current-carrying range and the same laying environment, the regressibility of the focusing pattern is judged. When the focusing index set of the same joint identification node cannot be returned to the center of the historical focusing baseline group, the corresponding joint identification node is marked as a saturation critical state candidate, and the saturation critical state candidate is used to generate the first focusing pattern and the second focusing pattern based on the voltage segment and the current segment, respectively. When the first focused map and the second focused map show structural fractures in the neighborhood of the same joint identifier node, and the structural fractures meet the synchronization condition, the corresponding joint identifier node is determined to be a faulty joint node, and a fault status marker bound to the faulty joint node is generated. Based on the fault status marker, the target joint identification node is determined on the cable segmentation diagram, and the intermediate joint fault monitoring result corresponding to the target joint identification node is output.

2. The method for monitoring faults in intermediate joints of 35kV ring network cables for urban rail transit according to claim 1, characterized in that, The process of acquiring ring network topology data, switch position status data, and cable mileage mapping data of the 35kV ring network cable for urban rail transit to construct a cable segmentation diagram, and generating a joint identification node for each intermediate joint in the cable segmentation diagram, and generating a ring network status identifier corresponding to the current switch position status, specifically includes: Obtain ring network topology data, switch position status data, and cable mileage mapping data for 35kV ring network cables in urban rail transit. Based on the ring network topology data, a basic topology graph containing site nodes, switch nodes and cable segment edges is constructed, and a consistency check is performed on the basic topology graph to eliminate duplicate nodes and broken edges. Based on the cable mileage mapping data, the cable segment edges are divided according to the joint mileage value. The joint identification node is inserted at each division position, so that the joint identification node connects the adjacent sub-cable segment edges and inherits the loop identification and laying environment segment attributes. Based on the switch position status data, open or closed constraints are applied to the switch nodes. Connectivity component parsing is performed on the cable segment diagram after the constraints are applied to obtain the current ring network subgraph. The subgraph identifier of the current ring network subgraph, the set of switch nodes participating in the subgraph, and the set of reachable joint identifier nodes in the current ring network subgraph are written into the ring network status identifier.

3. The method for monitoring faults in intermediate joints of 35kV ring network cables for urban rail transit according to claim 1, characterized in that, The method involves constructing a synchronous observation array based on the stations in the cable segmentation diagram, simultaneously monitoring transient events during the ring network operation, generating an event identifier for each transient event, and extracting voltage and current segments from the event identifier within the synchronous observation array to form an event evidence package, specifically including: Extract station nodes from the cable segmentation diagram and generate a station identifier for each station node. The station identifier is bound to the voltage transformer secondary side access point, the sheath grounding lead-down access point, the communication link identifier, and the time synchronization link identifier. Based on the ring network status identifier, the station identifiers in the current ring network subgraph are filtered, and the array member station set is determined according to the spatial separation degree and topology coverage to form the synchronous observation array identifier; Establish voltage and current observation channels at each array member site, and perform time synchronization on the array member sites through the time synchronization link identifier to generate a time synchronization health mark. The system monitors transient events during the operation of the ring network and generates event candidates based on changes in switch position status, traction load changes, regenerative braking feedback, and external power grid disturbances. Event windows are formed based on the event candidates, and unique event identifiers are assigned. Within the synchronous observation array, voltage and current segments are extracted based on the event identifier. Each voltage and current segment includes a pre-transient baseline segment, a transient main response segment, and a transient tail segment. Segment quality assessment is performed on the voltage and current segments to generate segment availability markers. The multi-site voltage segment, multi-site current segment, segment availability flag, and timing health flag under the same event identifier are encapsulated into the event evidence package.

4. The method for monitoring faults in intermediate joints of 35kV ring network cables for urban rail transit according to claim 1, characterized in that, The process involves accumulating event evidence packets within the same ring network status identifier to construct a propagation operator set. After performing amplitude-phase normalization and time-domain inversion on the propagation operator set, cross-site recorrelation and superposition are performed to form a virtual playback field. Based on this virtual playback field, a focused map indexed by the connector identifier node is generated. Specifically, this includes: A state bucket index is established according to the ring network status identifier to store the corresponding event evidence packet queue, and the event evidence packet queue is filtered by timing health marking and fragment availability marking to form a set of available event evidence packets; The available event evidence package set is subjected to array configuration consistency processing to unify the array member site set and site identifier order, and operator samples containing multi-site voltage segments, multi-site current segments and missing masks are generated to generate the propagation operator set. Amplitude-phase normalization is performed on the operator samples in the propagation operator set, and mirror site identifiers are selected within the operator samples to perform time-domain inversion on the corresponding segments; Based on the inverted segment of the mirror site identifier and the forward segment of the remaining site identifier, cross-site recorrelation superposition is performed to form a virtual playback field. A set of topological paths between site identifiers and joint identifier nodes is established on the cable segmentation diagram. Under the constraint of the propagation window, the virtual playback field is projected onto each joint identifier node to generate a set of focusing indicators including focused peak position, focused peak width, focused peak energy integral and peak sidelobe ratio, so as to form the focusing spectrum indexed by the joint identifier node.

5. The method for monitoring faults in intermediate joints of 35kV ring network cables for urban rail transit according to claim 1, characterized in that, Under the same current-carrying range and the same laying environment, the regressibility of the focusing pattern is determined. When the focusing index set of the same connector identifier node cannot regress to the center of the historical focusing baseline cluster, the corresponding connector identifier node is marked as a saturation critical state candidate. The saturation critical state candidate is then used to generate a first focusing pattern and a second focusing pattern based on voltage and current segments, respectively. Specifically, this includes: For each of the aforementioned joint identification nodes, an operating context label consisting of a current-carrying interval label and a laying environment segment label is established. Under the same operating context label and the same ring network status label, a set of focusing index samples from the historical focusing map is extracted to construct a historical focusing baseline cluster, and the cluster center and cluster discrete range of the historical focusing baseline cluster are calculated. The set of focus indicators corresponding to the joint identifier node in the current focus map is compared with the population center. When the set of focus indicators deviates from the population center in multiple consecutive time windows and does not return to the discrete range of the population after the running context label is restored to the same conditions, the joint identifier node is marked as the saturation critical state candidate. For each of the saturated critical state candidates, a first set of propagation operators is constructed based on voltage segments and a first focusing spectrum is generated; a second set of propagation operators is constructed based on current segments and a second focusing spectrum is generated.

6. The method for monitoring faults in intermediate joints of 35kV ring network cables for urban rail transit according to claim 1, characterized in that, When the first focused map and the second focused map show structural fractures within the neighborhood of the same joint identifier node, and the structural fractures meet the synchronization condition, the corresponding joint identifier node is identified as a faulty joint node, and a fault status marker bound to the faulty joint node is generated. Specifically, this includes: A structural fracture observation window is established for each of the aforementioned joint identification nodes, and the structural fracture observation window is bound to the ring network status identifier; Within the structural fracture observation window, the neighborhood focusing sequence of the first focusing spectrum and the second focusing spectrum in the neighborhood of the joint identifier node is extracted, and discontinuous migration detection is performed on the peak position group, morphological group and energy group based on the neighborhood focusing sequence. When the peak position group, the morphological group and the energy group undergo stable transition in multiple consecutive construction cycles and cannot regress to the historical focusing baseline group, it is determined to be a structural fracture. Synchronization condition verification is performed on the structural fractures of the first focused spectrum and the second focused spectrum. The synchronization condition includes time synchronization and spatial synchronization. When both the time synchronization and the spatial synchronization meet the preset constraints, the corresponding joint identification node is determined to be a faulty joint node. Generate a fault status marker bound to the faulty joint node. The fault status marker includes a joint identifier node, a fracture start time stamp, a structural fracture observation window identifier, a ring network status identifier, an event identifier set, and a first focused spectrum fracture summary and a second focused spectrum fracture summary.

7. The method for monitoring faults in intermediate joints of 35kV ring network cables for urban rail transit according to claim 1, characterized in that, The step of determining the target joint identification node on the cable segmentation diagram based on the fault status marker and outputting the intermediate joint fault monitoring result corresponding to the target joint identification node specifically includes: Based on the fault status marker, read the ring network status identifier, structural fracture observation window identifier, and the first and second focused spectrum fracture summaries, and lock the ring network sub-graph corresponding to the ring network status identifier in the cable segmentation diagram; Within the ring network subgraph, the comprehensive fracture strength index of each joint identification node is calculated based on the first focused spectrum and the second focused spectrum, and a fracture weight map is constructed using the comprehensive fracture strength index as the node weight; Search the fracture weight map for connected subgraphs whose weights meet preset conditions to form a set of target joints, and filter the set of target joints based on the peak overlap and energy concentration region consistency of the first and second focused spectra to determine the target joint identification nodes; The mileage mapping information of the target joint identification node, the laying environment section attributes, and the fault status marker are written into the intermediate joint fault monitoring results and output.

8. A fault monitoring device for intermediate joints of 35kV ring network cables in urban rail transit, characterized in that, The device is used to perform the fault monitoring method for intermediate joints of 35kV ring network cables in urban rail transit as described in any one of claims 1 to 7. The device includes an acquisition module and a processing module, wherein... The acquisition module is used to acquire the ring network topology data, switch position status data and cable mileage mapping data of the 35kV ring network cable of urban rail transit, so as to construct the cable segmentation diagram. At the same time, it generates a joint identification node for each intermediate joint in the cable segmentation diagram and generates a ring network status identifier corresponding to the current switch position status. The processing module is used to construct a synchronous observation array based on the stations in the cable segmentation diagram, simultaneously monitor transient events during the operation of the ring network, generate an event identifier for each transient event, and extract voltage and current segments for the event identifier within the synchronous observation array to form an event evidence package. The processing module is also used to accumulate the event evidence packets within the same ring network status identifier, construct a propagation operator set, and perform cross-site recorrelation superposition after performing amplitude-phase normalization and time-domain inversion processing on the propagation operator set to form a virtual playback field, and generate a focusing map indexed by the connector identifier node based on the virtual playback field. The processing module is also used to perform a regressibility judgment on the focusing pattern under the same current carrying range and the same laying environment section. When the focusing index set of the same joint identification node cannot be returned to the center of the historical focusing baseline group, the corresponding joint identification node is marked as a saturation critical state candidate, and the saturation critical state candidate is used to generate a first focusing pattern and a second focusing pattern based on voltage segment and current segment, respectively. The processing module is further configured to determine the corresponding joint identifier node as a faulty joint node and generate a fault status marker bound to the faulty joint node when the first focused spectrum and the second focused spectrum have structural fractures in the neighborhood of the same joint identifier node and the structural fractures meet the synchronization condition. The processing module is further configured to determine the target joint identification node on the cable segmentation diagram based on the fault status marker, and output the intermediate joint fault monitoring result corresponding to the target joint identification node.

9. An electronic device, characterized in that, The electronic device includes a processor, a memory, a user interface, and a network interface. The memory is used to store instructions. The user interface and the network interface are both used to communicate with other devices. The processor is used to execute the instructions stored in the memory to cause the electronic device to perform the method as described in any one of claims 1 to 7.

10. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium stores instructions that, when executed, perform the method as described in any one of claims 1 to 7.

Images