Transformer area topology identification method based on dual-mode communication technology

Abstract

The application discloses a transformer area topology identification method based on a dual-mode communication technology. The method is based on a white list to issue characteristic current identification instructions to each station, and divides the stations into an identified station set and an unidentified station set. The corresponding identified neighbor station set of the unidentified station is determined, and a transformer area identification object set is constructed. The stations in the transformer area identification object set are instructed to collect power frequency zero-crossing period data, and report the data to a central coordinator through a dual-mode communication mode. The central coordinator performs time alignment and similarity analysis on the collected power frequency zero-crossing period data, obtains a power frequency zero-crossing similarity analysis result for the unidentified station, and determines a transformer area attribution determination result of the unidentified station, so that the transformer area topology identification is completed. The method combines characteristic current detection and neighbor station power frequency zero-crossing data analysis, can realize fast and accurate identification of the transformer area topology in a complex power grid environment, and improves the reliability of transformer area management and data acquisition.

Description

Technical Field

[0001] This invention relates to the field of power line carrier communication technology, and in particular to a method for identifying transformer substation topology based on dual-mode communication technology. Background Technology

[0002] In low-voltage power consumption data acquisition systems, transformer substation topology identification typically relies on power line carrier communication or wireless communication, with a central coordinator managing all types of electrical equipment within the substation area. Existing technologies commonly employ methods for transformer substation topology identification, including manual record-keeping, static configuration based on equipment installation information, automatic identification methods based on characteristic current or power frequency cycle data, and topology inference based on communication network relationships. These methods typically involve a concentrator collecting communication status, electrical characteristics, or timing data from each station and analyzing the connection relationships of equipment within the substation area using preset rules to form the substation topology.

[0003] However, in practical applications, existing technologies often suffer from significant impacts on identification results due to on-site installation errors, changes in the communication environment, and fluctuations in power grid operating conditions. This is especially true when the distribution area is large, the number of stations is numerous, or the electromagnetic environment is complex. Problems such as incorrect station attribution, inaccurate topology relationships, or low identification efficiency can easily occur, thereby affecting the reliability of electricity consumption data and the accuracy of distribution area operation and maintenance management.

[0004] Therefore, it is necessary to provide a new technical solution to improve the applicability and reliability of existing transformer substation topology identification methods in complex low-voltage power distribution environments. Summary of the Invention

[0005] This application provides a method for identifying transformer substation topology based on dual-mode communication technology, so as to achieve fast and accurate identification of transformer substation topology.

[0006] This application provides a method for identifying transformer substation topology based on dual-mode communication technology, including: The central coordinator issues characteristic current identification instructions to each station in the whitelist based on the whitelist, and receives characteristic current monitoring results reported by the monitoring module. Then, based on the characteristic current monitoring results, the stations in the whitelist are divided into a set of identified stations and a set of unidentified stations. Based on the set of unidentified sites, determine the set of identified neighboring sites for each unidentified site, and construct the set of identification objects for the transformer area using the unidentified sites and their sets of identified neighboring sites. The central coordinator issues a transformer area identification instruction to the set of transformer area identification objects, instructing each station in the set of transformer area identification objects to collect power frequency zero-crossing cycle data within the same time window, and report the power frequency zero-crossing cycle data to the central coordinator through dual-mode communication. The central coordinator performs time alignment processing on the power frequency zero-crossing cycle data to obtain aligned power frequency zero-crossing cycle data pairs for each unidentified site, wherein each aligned power frequency zero-crossing cycle data pair includes the power frequency zero-crossing cycle data of the unidentified site and the power frequency zero-crossing cycle data of at least one identified neighbor site in its set of identified neighbor sites. The central coordinator performs similarity analysis on the aligned power frequency zero-crossing cycle data pairs to obtain power frequency zero-crossing similarity analysis results for each unidentified site. The central coordinator determines the substation affiliation of each unidentified site based on the power frequency zero-crossing similarity analysis results, and completes the substation topology identification based on the substation affiliation results.

[0007] The beneficial effects of the technical solution provided in this application include: (1) By introducing a characteristic current identification mechanism, the stations in the whitelist are initially screened and divided into a set of identified stations and a set of unidentified stations, so that the transformer area topology identification process has a clear hierarchical processing path, avoiding repeated judgment of all stations and effectively improving the overall efficiency of transformer area topology identification. (2) For unidentified stations, identified neighboring stations are introduced as reference objects. By constructing a set of transformer area identification objects formed by unidentified stations and their identified neighboring stations, the transformer area attribution determination is based on the relationship between stations with close physical distance and stable communication conditions, thereby reducing the impact of complex power grid operating conditions on the identification results and improving the accuracy of transformer area topology identification. (3) By collecting and aligning the power frequency zero-crossing cycle data of unidentified stations and their identified neighboring stations in the same time window, the subsequent similarity analysis has a unified time benchmark, reducing the timing deviation caused by load fluctuations, line differences and other factors, which is conducive to obtaining more stable and reliable transformer area attribution determination results. (4) Combined with dual-mode communication, the transmission and reporting of power frequency zero-crossing cycle data can be completed. Even when the single communication mode is interfered with or the communication quality deteriorates, the continuous acquisition and analysis of data can still be guaranteed, thereby improving the reliability and robustness of the transformer topology identification process in complex low-voltage power distribution environment. Attached Figure Description

[0008] Figure 1 This is a flowchart of a transformer topology identification method based on dual-mode communication technology provided in the first embodiment of this application. Detailed Implementation

[0009] Many specific details are set forth in the following description to provide a full understanding of this application. However, this application can be implemented in many other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of this application; therefore, this application is not limited to the specific embodiments disclosed below.

[0010] The first embodiment of this application provides a method for identifying transformer topology based on dual-mode communication technology. Please refer to... Figure 1 This figure is a schematic diagram of the first embodiment of this application. The following is in conjunction with... Figure 1 The first embodiment of this application provides a detailed description of a method for identifying transformer topology based on dual-mode communication technology.

[0011] Step S101: The central coordinator sends characteristic current identification instructions to each station in the whitelist based on the whitelist, and receives the characteristic current monitoring results reported by the monitoring module. Then, based on the characteristic current monitoring results, the stations in the whitelist are divided into a set of identified stations and a set of unidentified stations.

[0012] In this invention, step S101 is the starting and foundational step of the entire transformer substation topology identification method. Its core function is to perform a preliminary screening of the sites within the whitelist managed by the central coordinator based on physical and electrical characteristics, thereby providing a clear and reliable scope for subsequent refined transformer substation identification targeting specific sites. The central coordinator refers to the communication and control unit deployed on the transformer substation concentrator side, which functions as the main node of the low-voltage power line communication network. It is responsible for maintaining the network topology, site management information, and data scheduling strategies of the transformer substation communication network. The whitelist is a set of site identifiers maintained by the central coordinator, typically identified by the unique hardware address of the communication unit. This whitelist represents the set of sites allowed to access and participate in the current transformer substation communication network.

[0013] In the specific implementation process, the central coordinator first issues characteristic current identification instructions to each station on the whitelist. These instructions can be sent via low-voltage power line carrier communication and include at least a characteristic current trigger identifier, characteristic current transmission parameters, and identification timing control information. This instructs the station, upon receiving the instruction, to inject a specific form of current disturbance signal into the connected low-voltage power supply line within a predetermined time period. The characteristic current referred to here is an artificially constructed current signal distinct from the normal load current. It possesses pre-defined characteristics in amplitude, frequency components, modulation method, or timing structure, and is used to reliably detect and distinguish it at the transformer or line side. The specific implementation of the characteristic current can employ short-time pulse current, modulated current of a specific frequency band, or an identification signal superimposed on the power frequency current, as long as it can be distinguished from background current and noise signals at the target monitoring point.

[0014] Upon receiving the characteristic current identification command, each substation transmits the characteristic current signal to the low-voltage line it is connected to, according to the parameters specified in the command. This transmission process is typically coordinated by the substation's internal communication and control module to ensure that the characteristic current is injected into the line within a specified time window, thereby avoiding time overlap with the characteristic current signals of other substations. Synchronously with the substations transmitting their characteristic current signals, monitoring modules located on the transformer side of the distribution area or at key nodes of the line monitor current changes on the low-voltage line in real time. These monitoring modules can be based on current transformers, Hall effect sensors, or other current detection devices. The current signals they collect are filtered and feature extracted to determine whether a characteristic current signal corresponding to a specific substation has been detected.

[0015] When the monitoring module successfully detects a characteristic current signal from a certain site, it generates a corresponding characteristic current monitoring result and reports it to the central coordinator. The characteristic current monitoring result includes at least site identification information and confirmation that the corresponding characteristic current has been detected, indicating that the characteristic current signal injected by the site can be transmitted along its power supply path to the distribution area where the central coordinator is located. After receiving the characteristic current monitoring results for each site, the central coordinator processes each site in the whitelist.

[0016] Specifically, for stations whose characteristic current monitoring results are successfully reported and confirmed within the predetermined identification period, the central coordinator assigns them to the identified station set. The identified station set indicates that the station not only exists in the whitelist, but its physical power supply path also maintains connectivity with the distribution area to which the central coordinator belongs; that is, the station can be confirmed to belong to the current distribution area. Conversely, for stations whose characteristic current monitoring results are not received within the predetermined identification period, the central coordinator assigns them to the unidentified station set. The unidentified station set does not directly indicate that the station necessarily does not belong to the current distribution area, but rather that its distribution area affiliation cannot be confirmed based solely on the characteristic current criterion. This could be due to installation errors, incorrect distribution area configuration, signal interference, line attenuation, or transient noise, which may have caused the characteristic current to fail to be detected.

[0017] Through the above method, step S101 achieves rapid initial screening of sites within the whitelist, clearly dividing the sites into a set of identified sites and a set of unidentified sites. This provides a clear, controllable, and reliable data source for subsequent steps that focus only on unidentified sites for further neighbor association analysis and power frequency zero-crossing cycle data comparison. This hierarchical identification mechanism effectively avoids repeatedly performing high-cost judgment operations on all sites, ensuring that the entire transformer area topology identification method still has good feasibility and engineering applicability in large-scale site scenarios.

[0018] Furthermore, the central coordinator issues characteristic current identification instructions to each station in the whitelist based on the whitelist, receives characteristic current monitoring results reported by the monitoring module, and then divides each station in the whitelist into a set of identified stations and a set of unidentified stations based on the characteristic current monitoring results, including: The central coordinator generates a sequence of sites to be identified based on a whitelist, and assigns a unique identification number and a characteristic current time slot parameter corresponding to the identification number to each site in the sequence. The identification number and the characteristic current time slot parameter are the components of the characteristic current identification instruction. The central coordinator issues a characteristic current identification command to the corresponding station based on the identification sequence number and characteristic current time slot parameters, instructing the station to inject a characteristic current signal in its corresponding characteristic current time slot and generate a transmission identifier corresponding to the identification sequence number. The monitoring module detects the line current signal within a monitoring window covering the characteristic current time slot, and generates a characteristic current monitoring result record when a characteristic current signal corresponding to the identification number is detected. The characteristic current monitoring result record includes at least the identification number. The central coordinator receives the characteristic current monitoring result record and maps the characteristic current monitoring result record to the corresponding station based on the identification number. Then, depending on whether the characteristic current monitoring result corresponding to the station is received, the station is divided into the identified station set or the unidentified station set.

[0019] In this invention, the process of the central coordinator issuing characteristic current identification instructions to each station in the whitelist based on the whitelist and receiving characteristic current monitoring results reported by the monitoring module, and then dividing each station in the whitelist into a set of identified stations and a set of unidentified stations based on the characteristic current monitoring results, is not a simple one-time detection process, but a hierarchical identification process with a clear organizational structure, time constraints and result judgment rules.

[0020] In practice, the central coordinator first reads and parses its internally maintained whitelist. This whitelist is a set of site identifiers stored and managed by the central coordinator, typically identified by the unique hardware address of the site's communication unit, representing all sites configured to be allowed access to the current area's communication network. Based on this whitelist, the central coordinator generates a sequence of sites to be identified. This sequence is an ordered list of sites within the whitelist, sorted according to a preset rule. This sorting rule can be the natural order of the site identifiers, the order of site access time, or other fixed order maintained internally by the central coordinator. Its purpose is to provide a deterministic scheduling basis for the subsequent identification process.

[0021] After generating the sequence of sites to be identified, the central coordinator assigns a unique identification number to each site in the sequence. This identification number is a logical identifier used to distinguish different sites during the characteristic current identification process. It remains unique within the same characteristic current identification cycle and is used to establish a one-to-one correspondence between sites, characteristic current signals, and monitoring results. In this invention, the identification number is not only used to identify site identity but also serves as a core index for subsequent characteristic current time slot allocation, monitoring result matching, and site division determination.

[0022] Simultaneously, the central coordinator generates corresponding characteristic current time slot parameters for each identification number. These characteristic current time slot parameters define the temporal behavior of the site's injected characteristic current signal, and include at least a characteristic current start offset and a characteristic current duration. The characteristic current start offset represents the delay time relative to the start of the characteristic current identification cycle, and the characteristic current duration represents the length of time the site continuously injects the characteristic current signal within that time slot. By assigning non-overlapping characteristic current time slot parameters to different identification numbers, it can be ensured that the characteristic current signals from different sites are distinguishable from each other on the time axis within the same identification cycle, thereby avoiding signal superposition and misjudgment caused by concurrent injection from multiple sites.

[0023] After allocating the identification sequence number and characteristic current time slot parameters, the central coordinator encapsulates these parameters into a characteristic current identification command and distributes the command to the corresponding stations based on the identification sequence number. Upon receiving the characteristic current identification command, each station parses the included identification sequence number and characteristic current time slot parameters and, strictly adhering to the constraints of these parameters, injects a characteristic current signal into the low-voltage power supply line it is connected to within the corresponding characteristic current time slot. This characteristic current signal is an artificially constructed current disturbance signal, distinct from the normal load current, possessing pre-defined characteristics in amplitude, spectrum, or modulation scheme, for reliable identification at the monitoring side. After injecting the characteristic current signal, each station generates a transmission identifier corresponding to its identification sequence number, indicating that the station has completed the characteristic current transmission operation as instructed.

[0024] While injecting a characteristic current signal at the substation, the monitoring module continuously monitors the line current signal within a monitoring window covering the characteristic current time slot. This monitoring module is typically deployed at the transformer side of the distribution area or other key nodes capable of covering the power supply line of the distribution area. It collects the line current signal through a current transformer or equivalent detection device and performs feature matching analysis on the collected current signal. When the monitoring module detects a characteristic current signal corresponding to a specific identification number within the monitoring window, it generates a characteristic current monitoring result record. This record includes at least the corresponding identification number, indicating that the characteristic current signal injected by the substation corresponding to that identification number has been successfully transmitted along the power supply path to the monitoring location.

[0025] In a preferred embodiment, the characteristic current monitoring result record may further include the detection time, the characteristic current signal strength, and the detection confidence level calculated based on the signal and background noise within the monitoring window. The detection confidence level quantifies the reliability of the monitoring module's detection result of the characteristic current signal, and can be obtained as follows: within the characteristic current time slot, the ratio of the detected characteristic current signal amplitude to the background noise amplitude during the same period is calculated, for example, by dividing the effective amplitude of the characteristic current by the effective amplitude of the noise. The result is used as the numerical representation of the detection confidence level. When the detection confidence level is higher than a preset threshold, the characteristic current detection result can be considered to have high reliability.

[0026] After receiving the characteristic current monitoring result records reported by the monitoring module, the central coordinator maps the characteristic current monitoring result records to the corresponding stations based on the identification serial numbers contained therein. For a station that receives at least one characteristic current monitoring result record corresponding to its identification serial number within a predetermined identification period, the central coordinator determines that the characteristic current injection behavior of that station has been successfully monitored, and thus classifies the station into the identified station set. Conversely, for a station that does not receive any characteristic current monitoring result records corresponding to its identification serial number within the predetermined identification period, the central coordinator determines that its characteristic current has not been successfully monitored, and thus classifies the station into the unidentified station set.

[0027] Through the above methods, this implementation plan achieves structured initial screening of whitelisted sites without directly relying on historical topology information or manual configuration results. It clearly distinguishes sites into a set of identified sites and a set of unidentified sites, providing a clear, reliable, and engineering-feasible input basis for subsequent refined determination of transformer area affiliation based on power frequency zero-crossing cycle data of neighboring sites.

[0028] Furthermore, the central coordinator generates a sequence of sites to be identified based on a whitelist, and assigns a unique identification number and a corresponding characteristic current time slot parameter to each site in the sequence. The identification number and characteristic current time slot parameter are components of the characteristic current identification instruction, including: The central coordinator performs status labeling on the whitelist to distinguish between sites in the identified site set and sites in the unidentified site set. Based on the status labeling results, only sites in the unidentified state are extracted to form an initial set of sites to be identified. The central coordinator sorts the sites according to the preset identification scheduling rules based on the initial set of sites to be identified, and generates a sequence of sites to be identified. The identification scheduling rules take into account at least the number of historical identification failures of the sites and the number of communication adjacencies with the set of identified sites. The central coordinator assigns identification numbers to each site in sequence according to the sequence of sites to be identified, and generates a unique set of characteristic current time slot parameters based on the identification number. The set of characteristic current time slot parameters includes at least the characteristic current start offset and the characteristic current duration, and the characteristic current time slot parameter sets corresponding to different identification numbers do not overlap with each other on the time axis. The central coordinator writes the identification sequence number and the corresponding characteristic current time slot parameters into the characteristic current identification instruction, and caches the characteristic current identification instruction as an identification scheduling mapping table, so that the subsequently received characteristic current monitoring results can be accurately back-mapped to the corresponding site to be identified through the identification sequence number, thereby ensuring the determinism and consistency of the subsequent site set division process.

[0029] In this invention, in order to ensure that the characteristic current identification process has good controllability, scalability and result determinism, the central coordinator does not simply schedule all whitelisted sites indiscriminately before issuing characteristic current identification instructions to the sites. Instead, it constructs a structured and traceable identification and scheduling system by analyzing the status of the whitelist, prioritizing the sites and allocating time-series resources.

[0030] In the specific implementation process, the central coordinator first performs status labeling on the whitelist. The whitelist, as referred to here, is a set of site identifiers maintained by the central coordinator that are allowed to participate in the current area communication and identification process. In its internal storage structure, the central coordinator maintains at least one status identifier for each site in the whitelist, indicating whether the site has completed area attribution confirmation. By querying the execution results of previous identification processes, the central coordinator marks sites that have been successfully assigned to the identified site set as identified, and marks sites that have not yet completed area attribution confirmation or have previously failed identification as unidentified. After completing the status labeling, the central coordinator only extracts sites in the unidentified state to form an initial set of sites to be identified. This initial set of sites to be identified clearly defines the scope of sites that need to participate in the characteristic current identification and scheduling process, thereby avoiding unnecessary time and communication resources wasted by confirmed sites repeatedly participating in the identification process.

[0031] After obtaining the initial set of sites to be identified, the central coordinator performs a site sorting process based on this set to generate a sequence of sites to be identified. This sequence is a list of sites with a defined order, used to constrain the subsequent allocation order of identification numbers and time slot resources. The sorting process is carried out according to preset identification scheduling rules. These identification scheduling rules are a set of rules predefined by the central coordinator during the system design phase, used to reflect the priority requirements of different sites in the identification process. In this invention, the identification scheduling rules comprehensively consider at least two factors: the number of historical identification failures of a site and the number of communication adjacencies between the site and the set of already identified sites.

[0032] The historical identification failure count refers to the number of times a site has failed to be successfully identified as belonging to the current transformer area in previous characteristic current identification or other transformer area identification processes. The central coordinator can maintain a failure counter internally for each site, incrementing the counter each time the site fails to be identified in an identification process. The number of communication adjacencies refers to the number of direct communication links that a site can establish with other sites in the identified site set within the communication network. This number can be obtained by statistically analyzing the neighbor information periodically reported by the sites. By considering both factors simultaneously, the central coordinator can prioritize scheduling sites with higher identification difficulty or higher correlation with already identified sites, thereby improving overall identification efficiency.

[0033] For example, in one specific embodiment, if site A has three historical identification failures and has four direct communication neighbors within the set of identified sites, while site B has one historical identification failure and only one communication adjacency with an identified site, then during the sorting process, the central coordinator can prioritize site A over site B, thereby allocating identification resources to site A first.

[0034] After generating the sequence of sites to be identified, the central coordinator assigns identification numbers to each site sequentially according to the order of the sequence. These identification numbers are logical identifiers used to distinguish different sites within a single characteristic current identification cycle. They remain unique within that cycle and are used to establish a one-to-one correspondence between sites, characteristic current time slots, and subsequent monitoring results. The allocation order of identification numbers strictly follows the order of the site sequence to be identified, thus ensuring that higher-priority sites receive earlier or more stable identification resources.

[0035] While assigning identification numbers to each site, the central coordinator generates a unique set of characteristic current time slot parameters based on each identification number. This set of characteristic current time slot parameters defines the temporal behavior of the site injecting characteristic current signals, and includes at least a characteristic current start offset and a characteristic current duration. The characteristic current start offset represents the delay time at which the site begins injecting characteristic current signals relative to the start of the identification period, and the characteristic current duration represents the length of time the site continuously injects characteristic current signals. When generating the characteristic current time slot parameter set, the central coordinator ensures that the characteristic current time slots corresponding to different identification numbers do not overlap on the time axis, thereby preventing multiple sites from concurrently injecting characteristic current signals within the same time period, which could lead to the monitoring side being unable to distinguish the signal source.

[0036] After generating the identification sequence number and the characteristic current time slot parameter set, the central coordinator writes the identification sequence number and the corresponding characteristic current time slot parameter into the characteristic current identification instruction, and caches the generated characteristic current identification instruction as an identification scheduling mapping table. The identification scheduling mapping table is a data structure used to record the correspondence between the identification sequence number, the site identifier, and the characteristic current time slot parameter. Its function is to provide a reliable reverse mapping basis for subsequently received characteristic current monitoring results. When the central coordinator receives the characteristic current monitoring result record reported by the monitoring module in subsequent steps, it can quickly and accurately locate the corresponding site to be identified through the identification sequence number contained therein, thereby ensuring the determinism and consistency of the site set division process and the traceability of the overall identification process.

[0037] The central coordinator completes the fine-grained scheduling plan for the whitelisted sites before the characteristic current identification process begins. This ensures that the subsequent characteristic current injection, monitoring, and site division processes are all based on a clear, orderly, and verifiable scheduling structure, thereby significantly improving the stability, efficiency, and engineering feasibility of transformer topology identification in complex low-voltage power distribution environments.

[0038] Step S102: Based on the set of unidentified sites, determine the set of identified neighboring sites for each unidentified site, and construct the set of identification objects for the transformer area using the unidentified sites and their sets of identified neighboring sites.

[0039] After completing step S101, the central coordinator, based on the characteristic current monitoring results, has clearly divided the stations in the whitelist into an identified station set and an unidentified station set. The stations in the unidentified station set are the core objects for further determination of substation affiliation. The purpose of step S102 is to introduce reference stations that are closer to the unidentified stations in terms of physical connection and communication conditions, without directly relying on the electrical distance between the central coordinator and the unidentified stations. This establishes a reliable comparative basis for subsequent substation affiliation determination based on power frequency zero-crossing cycle data.

[0040] In this step, the central coordinator first takes the set of unidentified sites as input and, for each unidentified site, determines its corresponding set of identified neighboring sites. Here, neighboring sites refer to sites in the low-voltage power line communication network that can directly establish a communication link with the unidentified site; that is, communication between them does not require forwarding through a proxy node. This direct communication relationship can be achieved through high-speed carrier communication over low-voltage power lines or high-speed wireless communication over low-voltage power lines. As long as direct data exchange is possible at the communication protocol level, it can be considered a neighboring site relationship.

[0041] In practice, each station continuously maintains its neighboring station information during network operation. This information is typically obtained through periodic broadcast beacon frames or active probing mechanisms. For example, in high-speed carrier communication over low-voltage power lines, stations can discover other stations nearby with whom they can communicate directly by sending and receiving beacon frames, and record the corresponding station's identification information, communication quality parameters, and phase line attribute information. The central coordinator can collect the neighboring station information reported by each station to construct a complete network communication topology view, thereby determining whether direct communication relationships exist between any two stations.

[0042] When determining neighboring sites, this step further specifies that the set of identified neighboring sites must originate from the set of identified sites; that is, only sites that have been confirmed as belonging to the current transformer area through step S101 are selected as neighboring reference objects. This limitation avoids introducing sites whose transformer area affiliation is unclear or that do not belong to the current transformer area into the subsequent determination process, thereby reducing the risk of misjudgment. Furthermore, in a preferred embodiment, identified neighboring sites and unidentified sites can be further limited to being located on the same phase line to ensure higher consistency in their power supply path and electrical characteristics. This phase line information can be obtained through site installation configuration or communication parameter negotiation.

[0043] After determining the set of identified neighboring sites for each unidentified site, the central coordinator constructs a set of identification objects for the transformer area, using the unidentified site and its corresponding set of identified neighboring sites as the basic unit. This set of identification objects is not simply a collection of sites, but rather consists of multiple identification objects. Each identification object includes at least one unidentified site and one or more identified neighboring sites directly associated with that unidentified site. This object-oriented construction method ensures that subsequent steps, such as the collection, alignment, and similarity analysis of power frequency zero-crossing cycle data, can all revolve around a clearly defined set of sites, avoiding data overlap and logical confusion between different unidentified sites.

[0044] For example, in a specific implementation scenario, if an unidentified station can directly establish communication with two identified stations, the central coordinator can group the unidentified station and the two identified stations together into a single identification area object. If another unidentified station only has a direct communication relationship with one identified station, its corresponding identification area object only includes that one identified station. In this way, step S102, based on step S101, further refines the unidentified stations from the "set to be confirmed" to "identification objects with clear comparison references," laying a clear, stable, and operable data organization foundation for subsequently collecting power frequency zero-crossing cycle data within the same time window and performing accurate comparisons.

[0045] Furthermore, the step of determining the set of identified neighboring sites for each unidentified site based on the set of unidentified sites, and constructing a set of identification objects for the transformer area using the unidentified sites and their sets of identified neighboring sites, includes: The central coordinator takes the set of unidentified sites as input, and for each unidentified site, retrieves the neighbor site information maintained and periodically reported by each site in the communication network to obtain a set of candidate neighbor sites that can directly establish a communication link with the unidentified site. The central coordinator performs identified site filtering based on the candidate neighbor site set to remove sites that are not assigned to the identified site set, thereby obtaining a candidate identified neighbor site set that contains only the identified sites. The central coordinator performs phase line consistency verification on each site in the candidate identified neighbor site set to determine whether it is located on the same phase line as the corresponding unidentified site. The site that passes the phase line consistency verification is identified as the target identified neighbor site, thereby forming the identified neighbor site set corresponding to the unidentified site. The central coordinator constructs a set of identification objects for a transformer area, using unidentified sites and their corresponding sets of identified neighboring sites as basic units. Each identification object for a transformer area includes at least one unidentified site and at least one identified neighboring site that is consistent with its phase line and has been confirmed to belong to the current transformer area.

[0046] In this invention, after the initial division of stations based on characteristic current identification results, the central coordinator has obtained a set of unidentified stations. This set of unidentified stations does not necessarily mean that these stations do not belong to the current transformer area, but rather that their area affiliation cannot be confirmed solely through characteristic current monitoring; therefore, a further determination mechanism is needed.

[0047] In practice, the central coordinator first takes the set of unidentified sites as input. For each unidentified site, it retrieves neighbor site information maintained and periodically reported by each site in the communication network. This neighbor site information refers to the set of information about other sites with which a site can establish a direct communication link, obtained through the direct communication discovery mechanism during network operation. Establishing a direct communication link means that data transmission between two sites does not require forwarding through a proxy node, but can be directly completed via high-speed low-voltage power line carrier communication or high-speed low-voltage power line wireless communication. Neighbor site information is typically obtained by sites at the communication protocol level through broadcast beacons, active probing, or link maintenance mechanisms, and is periodically reported to the central coordinator in the form of neighbor site identifiers, communication reachability indicators, and necessary communication attributes.

[0048] After receiving and aggregating the neighbor site information reported by each site, the central coordinator can filter out a set of sites with direct communication relationships with a given unidentified site from the neighbor site information, forming a candidate neighbor site set. This candidate neighbor site set only reflects direct reachability relationships at the communication layer and does not contain any determination results regarding the substation's affiliation; therefore, it cannot be directly used for subsequent substation identification.

[0049] Based on this, the central coordinator performs an identified site filtering process on the candidate neighbor site set. Specifically, the central coordinator compares each site in the candidate neighbor site set with the identified site set obtained through the characteristic current identification step, eliminating sites that were not assigned to the identified site set, and retaining only those sites that have been confirmed to belong to the current transformer area. Through this filtering process, a candidate identified neighbor site set containing only identified sites is obtained. The sites in this set not only have a direct communication relationship with the unidentified sites, but their transformer area affiliation has also been confirmed through the previous steps, thus meeting the basic conditions for being used as a reference object for transformer area affiliation determination.

[0050] To further improve the reliability of reference sites, the central coordinator also needs to perform phase line consistency verification on each site in the candidate identified neighbor site set. The phase line referred to here is the specific phase line supplying power to the site in the low-voltage distribution system, such as phase A, phase B, or phase C in a three-phase four-wire system. The purpose of phase line consistency verification is to determine whether the candidate identified neighbor site and its corresponding unidentified site are connected to the same phase line. Since sites on the same phase line have higher consistency in electrical paths, load disturbances, and power frequency signal propagation characteristics, selecting only identified sites on the same phase line as unidentified sites as neighbor references can significantly reduce interference factors in subsequent power frequency zero-crossing cycle data analysis.

[0051] Phase line consistency verification can be implemented in various ways, such as comparing phase line information recorded in the site installation configuration or verifying based on phase line identifiers included in the communication parameters. As long as it can be clearly determined whether sites are on the same phase line, the implementation requirements of this step are met. The central coordinator identifies sites that pass the phase line consistency verification as target identified neighbor sites, thus forming a set of identified neighbor sites corresponding to the unidentified site.

[0052] After completing the above processing, the central coordinator constructs a set of identification objects for a transformer area, using unidentified sites and their corresponding sets of identified neighboring sites as basic units. This set of identification objects consists of multiple identification objects, each including at least one unidentified site and at least one identified neighboring site that is connected to the unidentified site on a phase line, is communicatively reachable, and has been confirmed to belong to the current transformer area. In this way, unidentified sites, which originally existed as isolated entities, are embedded into an identification object structure with a clear reference relationship. This allows subsequent steps to collect power frequency zero-crossing cycle data within the same time window and perform targeted comparative analysis around this structure.

[0053] Without relying on manual configuration or network-wide comparison, the central coordinator constructs a set of reference sites with communication correlation, electrical consistency, and transformer area confirmation attributes for each unidentified site. This provides a stable, controllable, and highly targeted analytical foundation for subsequent transformer area attribution determination, significantly improving the reliability and feasibility of transformer area topology identification in complex low-voltage power distribution environments.

[0054] Furthermore, the central coordinator performs phase line consistency verification on each station in the candidate identified neighbor station set to determine whether it is located on the same phase line as the corresponding unidentified station, and identifies the station that passes the phase line consistency verification as the target identified neighbor station, thereby forming the identified neighbor station set corresponding to the unidentified station, including: For each unidentified site, the central coordinator retrieves the phase line behavior feature data generated during the historical operation of the unidentified site. The phase line behavior feature data includes at least the power frequency zero-crossing phase offset sequence of the unidentified site in multiple sampling periods and the time stamp corresponding to the power frequency zero-crossing phase offset sequence. For each candidate identified neighbor site in the set of candidate identified neighbor sites, the central coordinator extracts the phase line reference feature data formed by the identified neighbor sites within the same or overlapping sampling period as the unidentified sites. The composition of the phase line reference feature data is consistent with the phase line behavior feature data to ensure the consistency of the feature dimension and time scale in the subsequent comparison process. The central coordinator calculates the phase line offset consistency metric based on phase line behavior feature data and phase line reference feature data. The phase line offset consistency metric is used to characterize the consistency of the power frequency zero-crossing phase offset change trend between unidentified sites and candidate identified neighbor sites in multiple sampling periods. The phase line offset consistency metric is obtained by statistically aggregating the phase offset difference under the same time index. The central coordinator compares the phase line offset consistency metric with the preset phase line consistency judgment condition. When the phase line offset consistency metric meets the phase line consistency judgment condition, it determines that the corresponding candidate identified neighbor station and the unidentified station are located on the same phase line, and adds the candidate identified neighbor station to the target identified neighbor station set. Otherwise, the candidate identified neighbor station is removed from the phase line consistency verification process.

[0055] In this invention, phase line consistency verification does not rely on static configuration parameters of the site or manually entered information, but is completed by analyzing the power frequency zero-crossing phase behavior formed by the site during actual operation, thereby avoiding the problem of phase line information distortion caused by on-site wiring changes, inconsistent files or configuration errors.

[0056] In the specific implementation process, the central coordinator first retrieves the phase line behavior characteristic data generated during the historical operation of each unidentified site. The "historical operation process" refers to the process by which a site continuously collects and reports power frequency-related data according to a system-preset sampling strategy under normal power supply and communication conditions; the "phase line behavior characteristic data" is a set of data reflecting the electrical behavior characteristics of a site under its connected phase lines. In this invention, the phase line behavior characteristic data includes at least a power frequency zero-crossing phase offset sequence and a time stamp corresponding to that sequence.

[0057] The power frequency zero-crossing phase offset refers to the time offset of the zero-crossing moment of the power frequency voltage detected at a certain station relative to a reference time base. This reference time base can be a unified sampling starting point issued by the central coordinator or the first zero-crossing moment recorded locally at the station. The power frequency zero-crossing phase offset sequence is formed by arranging the phase offset values ​​from multiple consecutive sampling periods in chronological order, used to describe the changing trend of the power frequency phase at the station over a period of time; the timestamp is used to indicate the sampling period or specific sampling time corresponding to each phase offset value, thereby ensuring the alignment of data from different stations in the time dimension.

[0058] After acquiring the phase line behavior feature data of the unidentified sites, the central coordinator extracts the phase line reference feature data formed by each identified neighbor site in the candidate identified neighbor site set within the same or at least partially overlapping sampling period as the unidentified site. The phase line reference feature data described here maintains the same data structure as the phase line behavior feature data of the unidentified sites, also including the power frequency zero-crossing phase offset sequence and the corresponding time stamp. By ensuring the consistency of the two types of data in feature dimensions and time scales, the central coordinator can directly compare them item by item in subsequent processing without additional normalization or dimension transformation operations.

[0059] After completing the above data preparation, the central coordinator calculates the phase line offset consistency metric based on the phase line behavior characteristic data of unidentified sites and the phase line reference characteristic data of candidate identified neighbor sites. The phase line offset consistency metric is a numerical indicator used to quantify the consistency of the power frequency zero-crossing phase offset change trends of two sites over multiple sampling periods, reflecting whether they are affected by the same electrical environment under the same phase line. The calculation process of the phase line offset consistency metric can be implemented as follows: First, under the premise of consistent or aligned time stamps, the difference between the power frequency zero-crossing phase offset values ​​of the unidentified site and the identified neighbor site at the same time index position is calculated to obtain a phase offset difference sequence; then, the phase offset difference sequence is statistically aggregated, for example, by calculating the absolute value of each difference and taking its arithmetic mean or weighted average, thereby obtaining a single phase line offset consistency metric.

[0060] For example, if, within a certain time interval, the power frequency zero-crossing phase offset sequence for an unidentified station is {1.2 ms, 1.3 ms, 1.1 ms, 1.2 ms}, while the power frequency zero-crossing phase offset sequence for a candidate identified neighbor station at the same time index is {1.25 ms, 1.35 ms, 1.15 ms, 1.25 ms}, then the corresponding phase offset difference sequence is {0.05 ms, 0.05 ms, 0.05 ms, 0.05 ms}. Taking the absolute value of these differences and averaging them yields a phase line offset consistency metric of 0.05 ms. The smaller this value, the higher the consistency in power frequency phase changes between the two stations.

[0061] After obtaining the phase line offset consistency metric, the central coordinator compares this metric with preset phase line consistency judgment conditions. These conditions are threshold rules pre-set by the central coordinator during the system design phase, used to distinguish between stations on the same phase line and stations on different phase lines. When the phase line offset consistency metric is not greater than the threshold, the central coordinator determines that the corresponding candidate identified neighbor station and the unidentified station are on the same phase line, and adds the candidate identified neighbor station to the target identified neighbor station set. Conversely, when the phase line offset consistency metric is greater than the threshold, it determines that the two are not on the same phase line, and removes the candidate identified neighbor station from the phase line consistency verification process.

[0062] The central coordinator can dynamically and objectively determine phase line consistency based on the power frequency phase behavior characteristics formed by the site in the real operating environment without relying on static configuration or manual verification. This ensures that neighboring sites entering the subsequent transformer area identification process have a high degree of consistency with unidentified sites at the electrical topology level, providing a reliable, robust and engineering-feasible prerequisite for subsequent similarity analysis based on power frequency zero-crossing cycle data.

[0063] Step S103: The central coordinator issues a transformer area identification instruction to the transformer area identification object set, instructing each station in the transformer area identification object set to collect power frequency zero-crossing cycle data within the same time window, and report the power frequency zero-crossing cycle data to the central coordinator through dual-mode communication.

[0064] After completing step S102, the central coordinator has constructed a corresponding set of identification objects for each unidentified site. Each set of identification objects consists of one unidentified site and at least one identified neighboring site with which it has a direct communication relationship. The purpose of step S103 is to organize relevant sites to collect power frequency zero-crossing cycle data with electrical discrimination significance under strictly controlled time conditions around the set of identification objects for the substation, and reliably transmit the collected data to the central coordinator, providing the original data foundation for subsequent data alignment and similarity analysis.

[0065] In this step, the central coordinator first issues a transformer area identification instruction to each unidentified site and its identified neighboring sites within the transformer area identification object set. This instruction, sent by the central coordinator via the communication network, includes at least a data acquisition trigger identifier, a unified data acquisition start time, a data acquisition duration, and a data type identifier. This instruction explicitly instructs each site to perform power frequency zero-crossing cycle data acquisition within the same time window. This same time window refers to a time interval uniformly set and issued by the central coordinator. The start and end times of this time interval are consistent across all relevant sites in the transformer area identification object set, ensuring the comparability of the power frequency zero-crossing cycle data acquired by each site in terms of time.

[0066] Power frequency zero-crossing cycle data refers to the time interval between two adjacent zero-crossing points of the same polarity in a power frequency voltage signal. Specifically, substations sample the voltage signal on their connected lines, identify the zero-crossing moments when the voltage signal transitions from positive to negative or vice versa, and record the time interval between adjacent zero-crossing moments, thus forming a continuous sequence of power frequency zero-crossing cycle data. This data reflects the combined effects of load changes, line impedance, and electromagnetic environment on voltage propagation along power lines, and therefore has significant reference value in determining substation affiliation.

[0067] When collecting power frequency zero-crossing cycle data, both unidentified stations and their identified neighboring stations independently complete sampling and zero-crossing detection operations within the same time window, as required by the station identification instructions. Since there may be slight deviations in the local clocks of each station, the unified collection time window does not require absolute physical synchronization among the stations. Instead, it ensures, through unified scheduling by the central coordinator, that the data collected by each station covers the same power frequency operating range, thus creating conditions for subsequent time alignment processing at the central coordinator.

[0068] After completing the acquisition of power frequency zero-crossing cycle data, each station reports the acquired power frequency zero-crossing cycle data to the central coordinator via dual-mode communication. The dual-mode communication method refers to the combined use of high-speed low-voltage power line carrier communication and high-speed low-voltage power line wireless communication. Each station can select one communication method for data transmission based on current communication quality, signal-to-noise ratio, or link status, or automatically switch to the other communication method when one is unavailable, thereby ensuring that the power frequency zero-crossing cycle data can be stably and completely transmitted to the central coordinator.

[0069] Through the above process, step S103 realizes the centralized collection and uploading of power frequency zero-crossing cycle data of each station in the set of identified transformer areas under the conditions of unified scheduling and reliable communication guarantee. This enables the central coordinator to obtain original power frequency zero-crossing cycle data with clear sources and consistent time ranges, providing a necessary and sufficient data input basis for subsequent steps to perform time alignment, similarity analysis and transformer area attribution determination.

[0070] Step S104: The central coordinator performs time alignment processing on the power frequency zero-crossing cycle data to obtain an aligned power frequency zero-crossing cycle data pair for each unidentified site, wherein each aligned power frequency zero-crossing cycle data pair includes the power frequency zero-crossing cycle data of the unidentified site and the power frequency zero-crossing cycle data of at least one identified neighbor site in its set of identified neighbor sites.

[0071] After completing step S103, the central coordinator has received power frequency zero-crossing cycle data from each station in the set of identified transformer areas. However, due to objective differences in physical location, power supply path, communication delay, and local clock among the stations, even if data acquisition is completed within the same time window, the power frequency zero-crossing cycle data reported by each station may still have offsets in terms of time series starting point and sample alignment. Therefore, the core of step S104 is to perform unified time alignment processing on the received power frequency zero-crossing cycle data to construct a standardized data structure that can be directly used for similarity analysis.

[0072] In this step, the central coordinator first extracts the power frequency zero-crossing cycle data reported by each unidentified site, as well as the power frequency zero-crossing cycle data reported by each identified neighboring site in its corresponding set of identified neighboring sites. This power frequency zero-crossing cycle data is typically in time series format, with each data item representing the time interval between two adjacent voltage zero-crossing points. This time interval can be expressed in milliseconds, microseconds, or other suitable time units. Since the power frequency zero-crossing cycle data sequences from different sites are not entirely consistent at the start time of acquisition, the central coordinator needs to perform time alignment processing on these data sequences to eliminate the impact of the initial offset on subsequent analysis.

[0073] Time alignment refers to the process by which the central coordinator, through a unified alignment rule, establishes a correspondence between the zero-crossing cycle data of unidentified stations and the zero-crossing cycle data of at least one identified neighboring station within the same time reference frame. Specifically, the central coordinator can truncate, shift, or rearrange the zero-crossing cycle data sequences of different stations based on the start identifier of the acquisition time window specified in the station identification instruction, or based on the timestamp information contained in the data reported by each station, ensuring that they cover the same time interval length. In cases where the number of sampling points differs or a small number of missing points exists, the central coordinator can discard redundant data or retain the common coverage interval to ensure that the data segments being compared are consistent in the time dimension.

[0074] After time alignment, the central coordinator constructs a corresponding aligned power frequency zero-crossing cycle data pair for each unidentified site. This aligned power frequency zero-crossing cycle data pair refers to a set of data combinations formed by the central coordinator after time alignment processing. Each set includes at least one power frequency zero-crossing cycle data sequence from the unidentified site and one power frequency zero-crossing cycle data sequence from one of its identified neighboring sites. If an unidentified site corresponds to multiple identified neighboring sites, multiple aligned power frequency zero-crossing cycle data pairs can be constructed separately for similarity analysis in subsequent steps.

[0075] Through the above method, step S104 transforms the power frequency zero-crossing cycle data, which originally came from different sources and may have inconsistent start times, into aligned power frequency zero-crossing cycle data pairs that can be directly compared under a unified time reference. This effectively eliminates the interference of factors such as time offset and differences in acquisition start points on the analysis results. This processing not only provides the necessary data prerequisites for conducting stable and reliable similarity analysis in subsequent steps, but also ensures that the determination of transformer substation affiliation is based on strictly consistent data, thereby improving the credibility and repeatability of the overall transformer substation topology identification results.

[0076] Furthermore, the process of time-aligning the power frequency zero-crossing cycle data by the central coordinator to obtain aligned power frequency zero-crossing cycle data pairs for each unidentified station includes: For each unidentified site, the central coordinator extracts the power frequency zero-crossing cycle data of the unidentified site and constructs a time reference sequence for the unidentified site. The time reference sequence for the unidentified site is composed of a sequence of relative zero-crossing times obtained by accumulating the power frequency zero-crossing cycle data in the order of collection. For each identified neighboring site of an unidentified site, the central coordinator extracts the power frequency zero-crossing cycle data of that identified neighboring site and constructs a candidate sequence of neighboring times. The candidate sequence of neighboring times is composed of a sequence of relative zero-crossing times obtained by accumulating the power frequency zero-crossing cycle data of the identified neighboring sites in the order of collection. The central coordinator calculates the candidate time offset based on the unidentified site time reference sequence and the neighbor time candidate sequence, and performs time translation processing on the neighbor time candidate sequence based on the candidate time offset to generate an aligned neighbor time sequence. The candidate time offset is the time translation amount that minimizes the average absolute time difference between the aligned neighbor time sequence and the unidentified site time reference sequence within a preset overlap length range. The central coordinator extracts corresponding power frequency zero-crossing cycle data segments from the overlapping interval of the unidentified site time base sequence and the aligned neighbor time sequence to form aligned power frequency zero-crossing cycle data pairs. The aligned power frequency zero-crossing cycle data pairs include power frequency zero-crossing cycle data segments of unidentified sites and power frequency zero-crossing cycle data segments of identified neighbor sites that are time-aligned with them.

[0077] In this invention, the time alignment processing of power frequency zero-crossing cycle data by the central coordinator is a key intermediate link connecting power frequency data acquisition and subsequent similarity analysis. Its core purpose is to eliminate the time misalignment caused by different sampling starting points, communication delays, and local clock deviations at different sites, so that power frequency zero-crossing cycle data from different sites can be compared under a unified time reference framework.

[0078] In the specific implementation process, the central coordinator first extracts the power frequency zero-crossing cycle data reported by each unidentified site within the same time window. This power frequency zero-crossing cycle data refers to the data sequence formed by sampling the voltage signal of the line connected to the site and detecting the time interval between two adjacent voltage zero-crossing points. This time interval is usually expressed in milliseconds or microseconds. Since the site records continuous cycle lengths rather than absolute timestamps during the acquisition process, the central coordinator needs to convert the power frequency zero-crossing cycle data into a relative zero-crossing time sequence through accumulation. Specifically, the central coordinator accumulates the duration of each cycle item according to the acquisition order of the power frequency zero-crossing cycle data, thereby obtaining a time sequence representing the relative zero-crossing times. This time sequence constitutes the unidentified site time reference sequence. This unidentified site time reference sequence takes the first zero-crossing point as the time starting point, and each subsequent element represents the time offset of the corresponding zero-crossing point relative to the starting point, serving as a unified reference for subsequent time alignment processing.

[0079] After constructing the time reference sequence for unidentified sites, the central coordinator extracts the power frequency zero-crossing cycle data reported by each identified neighbor site within the same acquisition window for each unidentified site, and constructs a neighbor time candidate sequence in the same manner as for the unidentified sites. The neighbor time candidate sequence is also obtained by accumulating the power frequency zero-crossing cycle data of the identified neighbor sites sequentially, with its starting point being the first zero-crossing point of that neighbor site itself. Since the acquisition starting point and local clock of the unidentified site and the identified neighbor site are not completely synchronized, the relative zero-crossing time sequences constructed by the two usually exhibit an overall offset on the time axis.

[0080] To eliminate the aforementioned offset, the central coordinator calculates candidate time offsets based on the unidentified site time reference sequence and neighboring candidate time sequences. The candidate time offset referred to here is the time shift that achieves optimal alignment with the unidentified site time reference sequence within a preset overlap length after shifting the neighboring candidate time sequence forward or backward by a certain time. Specifically, the central coordinator can perform multiple hypothetical shifts on the neighboring candidate time sequence within a reasonable time offset search range. Under each hypothetical shift condition, it calculates the absolute difference between the shifted neighboring candidate time sequence and the unidentified site time reference sequence at each corresponding zero-crossing time within the overlap interval, and then averages these absolute differences. The average absolute time difference is used to measure the alignment degree of the two time sequences under this offset condition. When the average absolute time difference reaches its minimum value, the corresponding time shift is determined as the candidate time offset.

[0081] For example, in one specific embodiment, if the relative zero-crossing times of the unidentified station's time reference sequence within a certain overlapping interval are 10 milliseconds, 30 milliseconds, and 50 milliseconds, respectively, and the zero-crossing times of the corresponding neighbor time candidate sequences of an identified neighbor station after the assumed shift are 12 milliseconds, 32 milliseconds, and 52 milliseconds, then the corresponding absolute differences are 2 milliseconds, 2 milliseconds, and 2 milliseconds, respectively, and the average absolute time difference is 2 milliseconds. If, under another assumed shift condition, the corresponding zero-crossing times are 15 milliseconds, 35 milliseconds, and 55 milliseconds, then the corresponding absolute differences are 5 milliseconds, 5 milliseconds, and 5 milliseconds, and the average absolute time difference is 5 milliseconds. Obviously, the average absolute time difference under the former assumed shift condition is smaller, therefore its corresponding time shift amount is selected as the candidate time offset.

[0082] After determining the candidate time offset, the central coordinator performs time shifting on the candidate neighbor time series based on the candidate time offset, thereby generating the aligned neighbor time series. At this point, the aligned neighbor time series and the unidentified station time reference series are under the same time reference frame, and their relative zero-crossing times have a basis for direct comparison in terms of overall trends and local changes.

[0083] After time alignment, the central coordinator further extracts corresponding power frequency zero-crossing cycle data segments within the overlapping interval of the unidentified site's time reference sequence and the aligned neighbor's time sequence to form aligned power frequency zero-crossing cycle data pairs. Each aligned power frequency zero-crossing cycle data pair consists of two parts: one part is the power frequency zero-crossing cycle data segment corresponding to the unidentified site within the overlapping interval, and the other part is the power frequency zero-crossing cycle data segment of the identified neighbor site within the same overlapping interval, aligned after time shifting. In this way, the two sets of power frequency zero-crossing cycle data that originally had an overall offset on the time axis are transformed into a data structure that corresponds periodically under a unified time reference.

[0084] The central coordinator can achieve precise time alignment between unidentified sites and identified neighboring sites based on the timing characteristics of the power frequency zero-crossing cycle data without relying on absolute time synchronization of the sites. This provides a strictly consistent, repeatable, and engineering-feasible input data foundation for subsequent similarity analysis, thereby significantly improving the stability and reliability of the transformer area topology identification results.

[0085] Furthermore, the process of the central coordinator calculating candidate time offsets based on unidentified site time reference sequences and neighbor candidate time sequences, and performing time shifting processing on the neighbor candidate time sequences based on the candidate time offsets to generate aligned neighbor time sequences, includes: The central coordinator first determines the time offset search interval based on the time span of the two sequences for the unidentified site time reference sequence and the neighboring time candidate sequence. The time offset search interval is limited to the time shift range that can ensure that the two sequences effectively overlap under the condition of at least satisfying the preset overlap length. The central coordinator performs multiple hypothetical time shifts on the neighbor time candidate sequence within the time offset search interval according to a preset time step. After each hypothetical time shift, the point-by-point time difference sequence between the shifted neighbor time candidate sequence and the unidentified station time reference sequence in the corresponding overlapping interval is calculated. The point-by-point time difference sequence is used as the error characterization result of the hypothetical time shift. The central coordinator performs statistical processing on the point-by-point time difference sequence corresponding to each hypothetical time shift, calculates the average absolute time difference corresponding to the hypothetical time shift, and establishes a mapping relationship between the average absolute time difference and the time shift amount corresponding to the hypothetical time shift, thereby forming a candidate time offset evaluation set. The central coordinator selects the time shift that minimizes the mean absolute time difference from the candidate time shift evaluation set as the candidate time shift, and performs unified time shift processing on the candidate time shift based on the candidate time shift to generate the aligned neighbor time series that is optimally aligned with the unidentified station time reference series on the time axis.

[0086] The time alignment process is executed uniformly by the central coordinator based on the previously constructed unidentified station time reference sequence and neighbor time candidate sequence. Its purpose is to achieve precise alignment of data collected from different stations on the time axis by analyzing the time structure characteristics inherent in the power frequency zero-crossing cycle data itself, in the absence of a globally unified clock or high-precision synchronization conditions, thereby providing a reliable data foundation for subsequent similarity analysis.

[0087] In this invention, the unidentified station time reference sequence refers to the relative zero-crossing time sequence formed by continuously acquiring power frequency zero-crossing cycle data from an unidentified station within the same acquisition window, and accumulating these data item by item according to their acquisition order. The "relative zero-crossing time" referred to here is not an absolute timestamp, but rather a relative time position obtained by accumulating the period length of each subsequent power frequency zero-crossing cycle, with the first zero-crossing point within the station's acquisition window as the time zero point. For example, under 50Hz power frequency conditions, if an unidentified station acquires power frequency zero-crossing cycle data of 19.98ms, 20.02ms, and 20.01ms during continuous sampling, the corresponding relative zero-crossing time sequence can be represented as 0ms, 19.98ms, 39.99ms, and 60.00ms. This sequence fully reflects the power frequency rhythm changes of the station within the sampling window.

[0088] The construction method of the neighbor time candidate sequence is exactly the same as that of the unidentified site time reference sequence. It is also a relative zero-crossing time sequence formed by accumulating the power frequency zero-crossing cycle data acquired by the identified neighbor sites in the corresponding acquisition window in the order of acquisition. Due to the different local clock starting points of different sites, the delay in communication triggering, and the transmission uncertainty of dual-mode communication links, there is usually an unknown overall translational deviation between the unidentified site time reference sequence and the neighbor time candidate sequence on the time axis. Therefore, it is necessary to compensate for this deviation by calculating the candidate time offset.

[0089] To this end, the central coordinator first determines a time offset search interval based on the time spans of the unidentified site time base sequence and the neighboring candidate time sequences. This time offset search interval is limited to a range where, under any given time shift, the neighboring candidate time sequence and the unidentified site time base sequence can form an effective overlap interval that meets a preset overlap length requirement. The preset overlap length refers to the minimum effective comparison length required in time alignment analysis; for example, it can be set to at least ten consecutive power frequency zero-crossing cycles to avoid statistical instability due to excessively short overlapping data. This method significantly reduces the search space for candidate time offsets while maintaining computational accuracy.

[0090] After determining the time offset search interval, the central coordinator performs multiple hypothetical time shifts on the neighboring time candidate sequences according to a preset time step. The time step can be set according to system accuracy requirements, for example, 0.1 ms or less, to balance computational complexity and alignment accuracy. After each hypothetical time shift, the central coordinator extracts the overlapping intervals on the time axis between the neighboring time candidate sequences and the unidentified station time reference sequences under the current shift conditions, and calculates the difference for each relative zero-crossing moment at the corresponding position within the overlapping interval, thereby generating a point-by-point time difference sequence. This point-by-point time difference sequence reflects the time deviation of the two sequences at each corresponding zero-crossing point under the current hypothetical time shift conditions.

[0091] Subsequently, the central coordinator statistically processes the point-to-point time difference sequence corresponding to each hypothetical time shift. Specifically, this can be achieved by averaging the absolute values ​​of the differences to obtain an average absolute time difference. This average absolute time difference is used to quantify the alignment degree between the two sequences under the current time shift. The smaller the value, the higher the degree of matching between the neighboring time candidate sequence and the unidentified station time reference sequence on the time axis under the influence of that time shift. The central coordinator establishes a one-to-one mapping relationship between each time shift and its corresponding average absolute time difference, thereby forming a complete evaluation set of candidate time offsets.

[0092] After the candidate time offset evaluation set is constructed, the central coordinator selects the time shift that minimizes the mean absolute time difference as the final candidate time offset. Statistically, this candidate time offset represents the optimal overall time offset compensation between the neighboring candidate time sequences and the unidentified site time reference sequences. Subsequently, the central coordinator performs a uniform time shift on the neighboring candidate time sequences based on this candidate time offset, thereby generating aligned neighboring time sequences that are optimally aligned with the unidentified site time reference sequences on the time axis. This aligned neighboring time sequence will serve as the sole time alignment result for subsequent extraction of aligned power frequency zero-crossing cycle data segments and similarity analysis, ensuring the consistency, repeatability, and determinism of the subsequent transformer substation topology identification process in the time dimension.

[0093] Step S105: The central coordinator performs similarity analysis on the aligned power frequency zero-crossing cycle data pairs to obtain power frequency zero-crossing similarity analysis results for each unidentified station.

[0094] After completing step S104, the central coordinator has constructed at least one set of aligned power frequency zero-crossing cycle data pairs for each unidentified site. These aligned power frequency zero-crossing cycle data pairs are within the same reference frame in the time dimension, providing the basic conditions for quantitative comparison and consistency determination. The purpose of step S105 is to perform similarity analysis on the aligned power frequency zero-crossing cycle data pairs to form an analysis result that objectively reflects the power supply consistency between the unidentified site and its identified neighboring sites, thereby providing a direct basis for subsequent transformer substation affiliation determination.

[0095] In this step, the central coordinator analyzes the aligned power frequency zero-crossing cycle data pairs, performing similarity analysis on the two power frequency zero-crossing cycle data sequences within each pair. This similarity analysis compares the power frequency zero-crossing cycle data sequences of unidentified sites with those of identified neighboring sites in terms of numerical trends, periodic stability, and overall fluctuation characteristics to assess whether they exhibit highly consistent power frequency operating characteristics. Because sites located in the same transformer area and physically close to each other are typically affected by the same load changes and line conditions at the power frequency level, their power frequency zero-crossing cycle data exhibit high similarity in time series characteristics.

[0096] Similarity analysis can be achieved by comparing two power frequency zero-crossing cycle data sequences item by item. For example, it can be done by calculating the time difference distribution between corresponding sampling points to determine whether the difference remains within a preset allowable range over a long period. Alternatively, it can be achieved by comparing the overall statistical characteristics of the two data sequences, such as comparing their average period value, period fluctuation amplitude, or trend to see if they are consistent. As long as the analysis method used can stably and repeatedly reflect the consistency between the two power frequency zero-crossing cycle data sequences, the implementation requirements of this step can be met. This invention does not limit the specific mathematical form of similarity analysis, but emphasizes the use of unified analysis rules to evaluate the consistency of all aligned power frequency zero-crossing cycle data pairs.

[0097] When a pair of aligned power frequency zero-crossing cycle data meets a preset similarity condition after similarity analysis, the central coordinator considers the unidentified site and its corresponding identified neighboring site to be consistent in power frequency zero-crossing characteristics. Conversely, if the similarity analysis result does not meet the preset condition, the two are considered not to be consistent in power frequency zero-crossing characteristics. For the same unidentified site, when there are multiple pairs of aligned power frequency zero-crossing cycle data, the central coordinator can perform similarity analysis on each pair of data separately and obtain the corresponding analysis results.

[0098] After completing the above analysis, the central coordinator summarizes the similarity analysis results between each unidentified site and its identified neighboring sites, thus forming the power frequency zero-crossing similarity analysis result for that unidentified site. The power frequency zero-crossing similarity analysis result can be represented by logical judgment results, flag bits, or other equivalent forms, used to characterize whether there is consistency in power frequency zero-crossing characteristics between the unidentified site and at least one identified neighboring site. Through this step, the power frequency zero-crossing cycle data, which originally only had time alignment relationships, is further transformed into analysis results with clear judgment significance, providing a direct, reliable, and reusable data foundation for subsequent steps in determining transformer substation affiliation.

[0099] Furthermore, the central coordinator performs similarity analysis on the aligned power frequency zero-crossing cycle data pairs to obtain power frequency zero-crossing similarity analysis results for each unidentified site, including: For each unidentified site, the central coordinator acquires at least one set of aligned power frequency zero-crossing cycle data pairs. Using the unidentified site power frequency zero-crossing cycle data segments in the aligned power frequency zero-crossing cycle data pairs as a reference sequence, a periodic change feature sequence of unidentified sites is constructed. The periodic change feature sequence of unidentified sites is composed of the changes between adjacent power frequency zero-crossing cycle data in chronological order, and is used to characterize the power frequency fluctuation behavior of unidentified sites within the acquisition window. For each pair of aligned power frequency zero-crossing cycle data for unidentified sites, the central coordinator constructs a corresponding neighbor cycle change feature sequence based on the power frequency zero-crossing cycle data segments of identified neighbor sites. The construction method of the neighbor cycle change feature sequence is consistent with that of the unidentified site cycle change feature sequence, thereby forming a feature sequence pair that can be directly compared in feature dimension. The central coordinator performs item-by-item difference calculations on the periodic change feature sequences of unidentified sites and their neighbors to generate a periodic change difference sequence. Based on this periodic change difference sequence, a periodic consistency index is calculated. The periodic consistency index is used to quantify the degree of consistency between unidentified sites and identified neighboring sites in the trend of power frequency zero-crossing periodic changes. The central coordinator aggregates the periodic consistency indices calculated by the central coordinator with those of multiple identified neighboring sites for the same unidentified site to generate the power frequency zero-crossing similarity analysis result corresponding to the unidentified site. When at least one periodic consistency index meets the preset consistency condition, the power frequency zero-crossing similarity analysis result is marked as passed; otherwise, it is marked as failed.

[0100] In this invention, after completing the time alignment processing of the power frequency zero-crossing cycle data, the central coordinator has obtained at least one set of aligned power frequency zero-crossing cycle data pairs for each unidentified site. These aligned power frequency zero-crossing cycle data pairs are data structures formed under a unified time reference frame, where the power frequency zero-crossing cycle data of the unidentified site corresponds to the power frequency zero-crossing cycle data of its identified neighboring sites on a cycle-by-cycle time axis, providing the basic conditions for conducting trend consistency analysis.

[0101] In the specific implementation process, the central coordinator first acquires at least one set of aligned power frequency zero-crossing cycle data pairs for each unidentified site, and extracts power frequency zero-crossing cycle data segments from each set of aligned power frequency zero-crossing cycle data pairs as the raw input for subsequent feature construction. The power frequency zero-crossing cycle data segments mentioned here refer to a continuous power frequency zero-crossing cycle duration sequence located within the overlapping interval after time alignment. This sequence reflects the actual situation of voltage signal periodic changes at the unidentified site within the same acquisition window.

[0102] The central coordinator constructs a periodic variation characteristic sequence for the unidentified stations based on the power frequency zero-crossing cycle data segments. This periodic variation characteristic sequence refers to a numerical sequence formed by calculating the changes between two adjacent power frequency zero-crossing cycle data segments and arranging them in chronological order. It describes the trend of power frequency cycle changes over time, rather than the absolute value of a single cycle. Specifically, if the power frequency zero-crossing cycle data segments for the unidentified stations are sequentially T1, T2, T3, ..., T... nThe corresponding periodic change characteristic sequence can be represented as ΔT2, ΔT3, ..., ΔT n , where ΔTᵢ = Tᵢ − Tᵢ -1 In this way, the periodic variation characteristic sequence can highlight the fluctuation behavior of the power frequency cycle, such as the period lengthening or shortening caused by sudden load changes, while weakening the fixed offset effect caused by differences in line length or reference between different stations.

[0103] After constructing the periodic variation feature sequence of the unidentified site, the central coordinator, for each pair of aligned power frequency zero-crossing cycle data for that unidentified site, further generates the corresponding neighbor periodic variation feature sequence based on the power frequency zero-crossing cycle data segments of the identified neighbor sites, using the same construction rules as the unidentified site. By ensuring the consistency of the feature construction method, the periodic variation feature sequence of the unidentified site and the neighbor periodic variation feature sequence remain strictly consistent in feature dimension, temporal order, and numerical meaning, thus forming a feature sequence pair that can be directly compared item by item.

[0104] After obtaining the aforementioned feature sequence pairs, the central coordinator performs item-by-item difference calculations on the periodic variation feature sequences of unidentified sites and their neighboring sites, generating a periodic variation difference sequence. Item-by-item difference calculation refers to performing a difference operation on the feature values ​​of the two feature sequences at the same time index position, for example, taking the absolute value of the difference, to quantify the degree of deviation of the power frequency periodic variation behavior of the two sites at the corresponding time. The periodic variation difference sequence thus reflects the similarity or deviation between the power frequency fluctuation trends of unidentified sites and identified neighboring sites throughout the entire acquisition window.

[0105] Based on this, the central coordinator calculates a periodic consistency index based on the periodic variation difference sequence. This periodic consistency index compresses the entire periodic variation difference sequence into a single numerical value with clear physical meaning, quantifying the overall consistency between unidentified sites and identified neighboring sites in the power frequency zero-crossing periodic variation trend. The periodic consistency index can be obtained by statistically calculating each difference value in the periodic variation difference sequence, for example, by calculating the average or weighted average of all difference values. When the overall periodic variation difference is small, it means that the two sites are highly consistent in their power frequency periodic variation trends, and the corresponding periodic consistency index will also exhibit characteristics that meet preset consistency conditions.

[0106] For example, if the difference sequence of periodic changes between an unidentified site and its identified neighboring site within the overlapping interval is {0.02 ms, 0.01 ms, 0.03 ms, 0.02 ms}, then the periodic consistency index of 0.02 ms can be obtained by calculating the arithmetic mean of the above difference values. If the preset consistency condition is that the periodic consistency index is not greater than 0.05 ms, then it can be determined that the unidentified site and the identified neighboring site meet the consistency requirements in the trend of zero-crossing periodic changes at the power frequency.

[0107] Since the same unidentified site may correspond to multiple identified neighbor sites, the central coordinator calculates multiple periodic consistency indices for each unidentified site with each of the identified neighbor sites. After completing the above calculations, the central coordinator aggregates these periodic consistency indices to generate the power frequency zero-crossing similarity analysis result for the unidentified site. This aggregation process is not a simple numerical average, but rather a judgmental fusion based on the transformer area identification logic. That is, when at least one periodic consistency index meets a preset consistency condition, the unidentified site is considered to have a high degree of consistency with at least one identified neighbor site within the current transformer area in terms of power frequency zero-crossing behavior, and the power frequency zero-crossing similarity analysis result for the unidentified site is marked as passed. Conversely, when the periodic consistency indices between the unidentified site and all its identified neighbor sites do not meet the aforementioned consistency condition, its power frequency zero-crossing similarity analysis result is marked as failed.

[0108] The central coordinator no longer relies on a single cycle value or simple time alignment results to determine the transformer area. Instead, it constructs and compares the power frequency zero-crossing cycle change characteristic sequence to conduct in-depth analysis of the power frequency fluctuation trend between sites. This enables it to reliably distinguish between sites in the same transformer area and sites in different transformer areas even under complex load changes and noise environments, significantly improving the accuracy, robustness, and engineering applicability of transformer area topology identification.

[0109] Furthermore, the central coordinator performs item-by-item difference calculations on the periodic variation feature sequences of unidentified sites and their neighbors to generate a periodic variation difference sequence. Based on this difference sequence, a periodic consistency index is calculated. This index quantifies the degree of consistency between the unidentified sites and their identified neighboring sites in the trend of power frequency zero-crossing periodic variation, including: After obtaining the periodic variation feature sequence of the unidentified site and the corresponding periodic variation feature sequence of the neighbor, the central coordinator first performs a length consistency check on the two feature sequences. In the case of inconsistent lengths, based on the time alignment result, only the effective overlapping part of the two feature sequences within the same time coverage area is retained, thereby forming an aligned feature subsequence pair for difference calculation. The central coordinator performs item-by-item difference calculations on the periodic changes of the corresponding positions for the alignment feature subsequence pairs in chronological order. Specifically, it calculates the numerical difference between the periodic changes of the unidentified station and the periodic changes of its neighbors, and writes this difference as a difference item into the periodic change difference sequence. The periodic change difference sequence fully reflects the deviation of the two stations in the continuous power frequency zero-crossing periodic changes in chronological order. The central coordinator extracts a difference symbol sequence and a difference amplitude sequence based on the periodic change difference sequence. The difference symbol sequence is used to characterize whether the change direction of the unidentified site and the neighboring site is consistent in the corresponding period, and the difference amplitude sequence is used to characterize the absolute degree of change deviation. The difference symbol sequence and the difference amplitude sequence are used as joint inputs to construct the intermediate evaluation features required for periodic consistency calculation. The central coordinator performs statistical aggregation processing on the periodic variation difference sequence based on intermediate evaluation characteristics to generate a periodic consistency index. The periodic consistency index at least comprehensively reflects the proportion of consistent difference signs and the proportion of difference amplitude within a preset allowable range.

[0110] In this invention, the unidentified station periodic variation characteristic sequence refers to the sequence formed by the change in the length of two adjacent zero-crossing cycles in a power frequency zero-crossing cycle data segment generated by an unidentified station within a data acquisition window, arranged in chronological order. The "periodic variation" refers to the difference in length between two adjacent zero-crossing cycles. For example, if a station obtains zero-crossing cycles of 19.98ms, 20.02ms, and 20.01ms in continuous sampling, the corresponding periodic variation sequence can be represented as +0.04ms and -0.01ms. This sequence no longer reflects the absolute size of the cycle but rather the small fluctuation trend of the cycle over time, which is more conducive to eliminating fixed deviations introduced by differences in measurement benchmarks between different stations.

[0111] The construction method of the neighbor periodic variation feature sequence is completely consistent with that of the unidentified site periodic variation feature sequence. It is also calculated from the power frequency zero-crossing periodic data segments formed by the corresponding identified neighbor sites under the same time alignment condition. By maintaining complete consistency between the two types of feature sequences in construction logic, data source and time order, it can be ensured that the subsequent difference calculation process has clear physical meaning and comparability.

[0112] In practice, after obtaining both the unidentified site periodic variation feature sequence and the neighboring periodic variation feature sequence simultaneously, the central coordinator first performs a length consistency check on the two feature sequences. The purpose of the length consistency check is to avoid sequence misalignment problems caused by sampling loss, communication interruption, or boundary truncation differences. When the lengths of the two feature sequences are not completely consistent, the central coordinator does not interpolate or pad them. Instead, based on the aforementioned time alignment results, it only retains the effective overlapping portion that the two feature sequences cover on the time axis, thus forming an aligned feature subsequence pair for difference calculation. For example, when the unidentified site periodic variation feature sequence contains 12 variations, while the neighboring periodic variation feature sequence contains only 10 variations, only the 10 variations within the same time coverage range of both are taken as valid input.

[0113] After constructing the aligned feature subsequence pairs, the central coordinator performs difference calculations on the periodic changes at corresponding positions in chronological order. Item-by-item difference calculation means that, under the same time index, the periodic change of the unidentified station is subtracted from the periodic change of the neighboring station to obtain a difference value, which is then written as a difference item into the periodic change difference sequence. The resulting periodic change difference sequence fully reflects the deviations in the trend and magnitude of change between the two stations during continuous power frequency zero-crossing periodic changes. For example, if at a certain moment the periodic change of the unidentified station is +0.03ms, while the periodic change of the neighboring station is +0.02ms, then the corresponding difference item at that moment is +0.01ms.

[0114] After obtaining the periodic variation difference sequence, the central coordinator further extracts two types of intermediate features with different physical meanings. One type is the difference sign sequence, which consists of the sign of each difference term in the periodic variation difference sequence. This sequence characterizes whether the periodic variation directions of the two stations are consistent within the corresponding period. When the sign of the difference term is positive or negative but the absolute value is small, and the two stations have the same variation direction, it indicates that their power frequency fluctuation trends are consistent within that period. When the sign of the difference term reflects opposite variation directions, it indicates that their power frequency variation behaviors within that period are significantly different. The other type is the difference amplitude sequence, which consists of the absolute value of each difference term in the periodic variation difference sequence. This sequence quantifies the magnitude of the variation deviation, thus avoiding the problem of relying solely on sign judgments and ignoring the deviation amplitude.

[0115] The central coordinator uses the difference symbol sequence and difference amplitude sequence as joint inputs to construct the intermediate evaluation features required for period consistency calculation. These intermediate evaluation features logically constrain both "whether the direction is consistent" and "whether the deviation is acceptable," ensuring that the period consistency index not only reflects trend consistency but also has the ability to suppress abnormal fluctuations.

[0116] Based on this, the central coordinator performs statistical aggregation on the periodic variation difference sequence to generate the final periodic consistency index. Specifically, a tolerance range for the difference amplitude can be pre-defined in the implementation, such as ±0.05ms. When the change direction corresponding to a difference item is consistent and its absolute difference value falls within this tolerance range, the difference item is determined to be a "consistent item". The central coordinator calculates the proportion of consistent items in the periodic variation difference sequence and uses this proportion as the periodic consistency index, or introduces weight correction to form a continuous index value. Therefore, the higher the proportion of difference items in the periodic variation difference sequence that meet the requirements of consistent change direction and amplitude difference within the tolerance range, the larger the calculated periodic consistency index value, and vice versa.

[0117] In this way, the cycle consistency index can quantitatively characterize the degree of consistency between unidentified sites and identified neighboring sites in the trend of power frequency zero-crossing cycle changes in a clear, calculable and physically meaningful manner, providing a stable, reliable and repeatable basis for subsequent determination of transformer area affiliation based on this index.

[0118] Step S106: The central coordinator determines the substation affiliation result of each unidentified site based on the power frequency zero-crossing similarity analysis result, and completes the substation topology identification based on the substation affiliation result.

[0119] After completing step S105, the central coordinator has obtained the corresponding power frequency zero-crossing similarity analysis results for each unidentified site. These results characterize the degree of consistency between the unidentified site and each identified neighbor site in its set of identified neighbor sites in terms of power frequency zero-crossing characteristics. Step S106 transforms the above analysis results into clear and executable substation affiliation determination results, and on this basis, completes the final determination and update of the substation topology identification results.

[0120] In this step, the central coordinator first uses unidentified sites as the judgment object and reads the power frequency zero-crossing similarity analysis results corresponding to the unidentified site. For unidentified sites with multiple sets of aligned power frequency zero-crossing cycle data pairs, the central coordinator will make a judgment by combining the similarity analysis results between the unidentified site and multiple identified neighboring sites. Specifically, when the power frequency zero-crossing similarity analysis results show that the unidentified site and at least one identified neighboring site meet the preset similarity conditions in power frequency zero-crossing characteristics, the central coordinator determines that the unidentified site and the identified neighboring site are in the same power supply area, thereby determining that the unidentified site belongs to the area corresponding to the central coordinator; conversely, when the power frequency zero-crossing similarity analysis results between the unidentified site and all its identified neighboring sites do not meet the aforementioned similarity conditions, the central coordinator determines that the unidentified site does not belong to the area.

[0121] After obtaining the substation affiliation determination result for each unidentified site, the central coordinator uses the determination result as the direct basis for substation topology identification and updates the site affiliation relationship of the current substation. For unidentified sites determined to belong to the current substation, the central coordinator can include them in the set of identified sites and establish connections between them and their corresponding neighboring sites in the substation topology structure; for unidentified sites determined not to belong to the current substation, the central coordinator can mark them as non-substation sites and remove the connections related to them from the substation topology structure, thereby preventing their electricity consumption data and communication behavior from being incorrectly included in the statistics and management scope of the current substation.

[0122] In a preferred embodiment, the central coordinator can also synchronously update the whitelist based on the distribution area attribution determination results. Specifically, when an unidentified site is determined not to belong to the current distribution area, its corresponding identifier is removed from the whitelist or transferred to the whitelist of another distribution area; when an unidentified site is determined to belong to the current distribution area, its valid status in the whitelist is maintained. This whitelist update operation ensures that the whitelist content remains consistent with the actual distribution area topology, reducing the risk of distribution area data confusion caused by incorrect site configurations during subsequent operation.

[0123] Through the above-mentioned judgment and update process, step S106 finally solidifies the power frequency zero-crossing similarity analysis results obtained in the previous steps into a clear distribution area affiliation judgment result, and completes the closed-loop processing of distribution area topology identification accordingly. This enables the distribution area topology structure maintained by the central coordinator to truly reflect the actual power supply relationship in the low-voltage distribution network, thereby providing accurate and reliable basic data support for subsequent applications such as power consumption data collection, distribution area operation and maintenance management, and line loss analysis.

[0124] A second embodiment of this application provides an electronic device, the electronic device comprising: processor; The memory is used to store a program, which, when read and executed by the processor, executes the transformer topology identification method based on dual-mode communication technology provided in the first embodiment of this application.

[0125] The third embodiment of this application provides a computer-readable storage medium storing a computer program thereon. When the program is executed by a processor, it executes a transformer topology identification method based on dual-mode communication technology provided in the first embodiment of this application.

[0126] Although this application discloses preferred embodiments as described above, it is not intended to limit this application. Any person skilled in the art can make possible changes and modifications without departing from the spirit and scope of this application. Therefore, the scope of protection of this application should be determined by the scope defined in the claims of this application.

Claims

1. A method for identifying transformer substation topology based on dual-mode communication technology, characterized in that, include: The central coordinator issues characteristic current identification instructions to each station in the whitelist based on the whitelist, and receives characteristic current monitoring results reported by the monitoring module. Then, based on the characteristic current monitoring results, the stations in the whitelist are divided into a set of identified stations and a set of unidentified stations. Based on the set of unidentified sites, determine the set of identified neighboring sites for each unidentified site, and construct the set of identification objects for the transformer area using the unidentified sites and their sets of identified neighboring sites. The central coordinator issues a transformer area identification instruction to the set of transformer area identification objects, instructing each station in the set of transformer area identification objects to collect power frequency zero-crossing cycle data within the same time window, and report the power frequency zero-crossing cycle data to the central coordinator through dual-mode communication. The central coordinator performs time alignment processing on the power frequency zero-crossing cycle data to obtain aligned power frequency zero-crossing cycle data pairs for each unidentified site, wherein each aligned power frequency zero-crossing cycle data pair includes the power frequency zero-crossing cycle data of the unidentified site and the power frequency zero-crossing cycle data of at least one identified neighbor site in its set of identified neighbor sites. The central coordinator performs similarity analysis on the aligned power frequency zero-crossing cycle data pairs to obtain power frequency zero-crossing similarity analysis results for each unidentified site. The central coordinator determines the substation affiliation of each unidentified site based on the power frequency zero-crossing similarity analysis results, and completes the substation topology identification based on the substation affiliation results.

2. The method for identifying transformer topology based on dual-mode communication technology according to claim 1, characterized in that, The central coordinator issues characteristic current identification instructions to each station in the whitelist based on the whitelist, and receives characteristic current monitoring results reported by the monitoring module. Then, based on the characteristic current monitoring results, it divides each station in the whitelist into a set of identified stations and a set of unidentified stations, including: The central coordinator generates a sequence of sites to be identified based on a whitelist, and assigns a unique identification number and a characteristic current time slot parameter corresponding to the identification number to each site in the sequence. The identification number and the characteristic current time slot parameter are the components of the characteristic current identification instruction. The central coordinator issues a characteristic current identification command to the corresponding station based on the identification sequence number and characteristic current time slot parameters, instructing the station to inject a characteristic current signal in its corresponding characteristic current time slot and generate a transmission identifier corresponding to the identification sequence number. The monitoring module detects the line current signal within a monitoring window covering the characteristic current time slot, and generates a characteristic current monitoring result record when a characteristic current signal corresponding to the identification number is detected. The characteristic current monitoring result record includes at least the identification number. The central coordinator receives the characteristic current monitoring result record and maps the characteristic current monitoring result record to the corresponding station based on the identification number. Then, depending on whether the characteristic current monitoring result corresponding to the station is received, the station is divided into the identified station set or the unidentified station set.

3. The method for identifying transformer topology based on dual-mode communication technology according to claim 1, characterized in that, The process of determining the set of identified neighboring sites for each unidentified site based on the set of unidentified sites, and constructing a set of identification objects for the transformer area using the unidentified sites and their sets of identified neighboring sites, includes: The central coordinator takes the set of unidentified sites as input, and for each unidentified site, retrieves the neighbor site information maintained and periodically reported by each site in the communication network to obtain a set of candidate neighbor sites that can directly establish a communication link with the unidentified site. The central coordinator performs identified site filtering based on the candidate neighbor site set to remove sites that are not assigned to the identified site set, thereby obtaining a candidate identified neighbor site set that contains only the identified sites. The central coordinator performs phase line consistency verification on each site in the candidate identified neighbor site set to determine whether it is located on the same phase line as the corresponding unidentified site. The site that passes the phase line consistency verification is identified as the target identified neighbor site, thereby forming the identified neighbor site set corresponding to the unidentified site. The central coordinator constructs a set of identification objects for a transformer area, using unidentified sites and their corresponding sets of identified neighboring sites as basic units. Each identification object for a transformer area includes at least one unidentified site and at least one identified neighboring site that is consistent with its phase line and has been confirmed to belong to the current transformer area.

4. The method for identifying transformer topology based on dual-mode communication technology according to claim 1, characterized in that, The process of time-aligning the power frequency zero-crossing cycle data by the central coordinator to obtain aligned power frequency zero-crossing cycle data pairs for each unidentified station includes: For each unidentified site, the central coordinator extracts the power frequency zero-crossing cycle data of the unidentified site and constructs a time reference sequence for the unidentified site. The time reference sequence for the unidentified site is composed of a sequence of relative zero-crossing times obtained by accumulating the power frequency zero-crossing cycle data in the order of collection. For each identified neighboring site of an unidentified site, the central coordinator extracts the power frequency zero-crossing cycle data of that identified neighboring site and constructs a candidate sequence of neighboring times. The candidate sequence of neighboring times is composed of a sequence of relative zero-crossing times obtained by accumulating the power frequency zero-crossing cycle data of the identified neighboring sites in the order of collection. The central coordinator calculates the candidate time offset based on the unidentified site time reference sequence and the neighbor time candidate sequence, and performs time translation processing on the neighbor time candidate sequence based on the candidate time offset to generate an aligned neighbor time sequence. The candidate time offset is the time translation amount that minimizes the average absolute time difference between the aligned neighbor time sequence and the unidentified site time reference sequence within a preset overlap length range. The central coordinator extracts corresponding power frequency zero-crossing cycle data segments from the overlapping interval of the unidentified site time base sequence and the aligned neighbor time sequence to form aligned power frequency zero-crossing cycle data pairs. The aligned power frequency zero-crossing cycle data pairs include power frequency zero-crossing cycle data segments of unidentified sites and power frequency zero-crossing cycle data segments of identified neighbor sites that are time-aligned with them.

5. The method for identifying transformer topology based on dual-mode communication technology according to claim 1, characterized in that, The central coordinator performs similarity analysis on the aligned power frequency zero-crossing cycle data pairs to obtain power frequency zero-crossing similarity analysis results for each unidentified site, including: For each unidentified site, the central coordinator acquires at least one set of aligned power frequency zero-crossing cycle data pairs. Using the unidentified site power frequency zero-crossing cycle data segments in the aligned power frequency zero-crossing cycle data pairs as a reference sequence, a periodic change feature sequence of unidentified sites is constructed. The periodic change feature sequence of unidentified sites is composed of the changes between adjacent power frequency zero-crossing cycle data in chronological order, and is used to characterize the power frequency fluctuation behavior of unidentified sites within the acquisition window. For each pair of aligned power frequency zero-crossing cycle data for unidentified sites, the central coordinator constructs a corresponding neighbor cycle change feature sequence based on the power frequency zero-crossing cycle data segments of identified neighbor sites. The construction method of the neighbor cycle change feature sequence is consistent with that of the unidentified site cycle change feature sequence, thereby forming a feature sequence pair that can be directly compared in feature dimension. The central coordinator performs item-by-item difference calculations on the periodic change feature sequences of unidentified sites and their neighbors to generate a periodic change difference sequence. Based on this periodic change difference sequence, a periodic consistency index is calculated. The periodic consistency index is used to quantify the degree of consistency between unidentified sites and identified neighboring sites in the trend of power frequency zero-crossing periodic changes. The central coordinator aggregates the periodic consistency indices calculated by the central coordinator with those of multiple identified neighboring sites for the same unidentified site to generate the power frequency zero-crossing similarity analysis result corresponding to the unidentified site. When at least one periodic consistency index meets the preset consistency condition, the power frequency zero-crossing similarity analysis result is marked as passed; otherwise, it is marked as failed.

6. The method for identifying transformer topology based on dual-mode communication technology according to claim 2, characterized in that, The central coordinator generates a sequence of sites to be identified based on a whitelist, and assigns a unique identification number and a corresponding characteristic current time slot parameter to each site in the sequence. The identification number and characteristic current time slot parameter are components of the characteristic current identification command, including: The central coordinator performs status labeling on the whitelist to distinguish between sites in the identified site set and sites in the unidentified site set. Based on the status labeling results, only sites in the unidentified state are extracted to form an initial set of sites to be identified. The central coordinator sorts the sites according to the preset identification scheduling rules based on the initial set of sites to be identified, and generates a sequence of sites to be identified. The identification scheduling rules take into account at least the number of historical identification failures of the sites and the number of communication adjacencies with the set of identified sites. The central coordinator assigns identification numbers to each site in sequence according to the sequence of sites to be identified, and generates a unique set of characteristic current time slot parameters based on the identification number. The set of characteristic current time slot parameters includes at least the characteristic current start offset and the characteristic current duration, and the characteristic current time slot parameter sets corresponding to different identification numbers do not overlap with each other on the time axis. The central coordinator writes the identification sequence number and the corresponding characteristic current time slot parameters into the characteristic current identification instruction, and caches the characteristic current identification instruction as an identification scheduling mapping table, so that the subsequently received characteristic current monitoring results can be accurately back-mapped to the corresponding site to be identified through the identification sequence number, thereby ensuring the determinism and consistency of the subsequent site set division process.

7. The method for identifying transformer topology based on dual-mode communication technology according to claim 3, characterized in that, The process of the central coordinator performing phase line consistency verification on each site in the candidate identified neighbor site set to determine whether it is located on the same phase line as the corresponding unidentified site, and identifying the site that passes the phase line consistency verification as the target identified neighbor site, thereby forming the identified neighbor site set corresponding to the unidentified site, includes: For each unidentified site, the central coordinator retrieves the phase line behavior feature data generated during the historical operation of the unidentified site. The phase line behavior feature data includes at least the power frequency zero-crossing phase offset sequence of the unidentified site in multiple sampling periods and the time stamp corresponding to the power frequency zero-crossing phase offset sequence. For each candidate identified neighbor site in the set of candidate identified neighbor sites, the central coordinator extracts the phase line reference feature data formed by the identified neighbor sites within the same or overlapping sampling period as the unidentified sites. The composition of the phase line reference feature data is consistent with the phase line behavior feature data to ensure the consistency of the feature dimension and time scale in the subsequent comparison process. The central coordinator calculates the phase line offset consistency metric based on phase line behavior feature data and phase line reference feature data. The phase line offset consistency metric is used to characterize the consistency of the power frequency zero-crossing phase offset change trend between unidentified sites and candidate identified neighbor sites in multiple sampling periods. The phase line offset consistency metric is obtained by statistically aggregating the phase offset difference under the same time index. The central coordinator compares the phase line offset consistency metric with the preset phase line consistency judgment condition. When the phase line offset consistency metric meets the phase line consistency judgment condition, it determines that the corresponding candidate identified neighbor station and the unidentified station are located on the same phase line, and adds the candidate identified neighbor station to the target identified neighbor station set. Otherwise, the candidate identified neighbor station is removed from the phase line consistency verification process.

8. The method for identifying transformer topology based on dual-mode communication technology according to claim 4, characterized in that, The process involves the central coordinator calculating candidate time offsets based on unidentified site time reference sequences and neighbor candidate time sequences, and then performing time shifting processing on the neighbor candidate time sequences based on these candidate time offsets to generate aligned neighbor time sequences. This includes: The central coordinator first determines the time offset search interval based on the time span of the two sequences for the unidentified site time reference sequence and the neighboring time candidate sequence. The time offset search interval is limited to the time shift range that can ensure that the two sequences effectively overlap under the condition of at least satisfying the preset overlap length. The central coordinator performs multiple hypothetical time shifts on the neighbor time candidate sequence within the time offset search interval according to a preset time step. After each hypothetical time shift, the point-by-point time difference sequence between the shifted neighbor time candidate sequence and the unidentified station time reference sequence in the corresponding overlapping interval is calculated. The point-by-point time difference sequence is used as the error characterization result of the hypothetical time shift. The central coordinator performs statistical processing on the point-by-point time difference sequence corresponding to each hypothetical time shift, calculates the average absolute time difference corresponding to the hypothetical time shift, and establishes a mapping relationship between the average absolute time difference and the time shift amount corresponding to the hypothetical time shift, thereby forming a candidate time offset evaluation set. The central coordinator selects the time shift that minimizes the mean absolute time difference from the candidate time shift evaluation set as the candidate time shift, and performs unified time shift processing on the candidate time shift based on the candidate time shift to generate the aligned neighbor time series that is optimally aligned with the unidentified station time reference series on the time axis.

9. The method for identifying transformer topology based on dual-mode communication technology according to claim 5, characterized in that, The central coordinator performs item-by-item difference calculations on the periodic variation characteristic sequences of unidentified sites and their neighbors to generate a periodic variation difference sequence. Based on this sequence, a periodic consistency index is calculated. This index quantifies the degree of consistency between the unidentified sites and their identified neighboring sites in the trend of power frequency zero-crossing periodic variation, including: After obtaining the periodic variation feature sequence of the unidentified site and the corresponding periodic variation feature sequence of the neighbor, the central coordinator first performs a length consistency check on the two feature sequences. In the case of inconsistent lengths, based on the time alignment result, only the effective overlapping part of the two feature sequences within the same time coverage area is retained, thereby forming an aligned feature subsequence pair for difference calculation. The central coordinator performs item-by-item difference calculations on the periodic changes of the corresponding positions for the alignment feature subsequence pairs in chronological order. Specifically, it calculates the numerical difference between the periodic changes of the unidentified station and the periodic changes of its neighbors, and writes this difference as a difference item into the periodic change difference sequence. The periodic change difference sequence fully reflects the deviation of the two stations in the continuous power frequency zero-crossing periodic changes in chronological order. The central coordinator extracts a difference symbol sequence and a difference amplitude sequence based on the periodic change difference sequence. The difference symbol sequence is used to characterize whether the change direction of the unidentified site and the neighboring site is consistent in the corresponding period, and the difference amplitude sequence is used to characterize the absolute degree of change deviation. The difference symbol sequence and the difference amplitude sequence are used as joint inputs to construct the intermediate evaluation features required for periodic consistency calculation. The central coordinator performs statistical aggregation processing on the periodic variation difference sequence based on intermediate evaluation characteristics to generate a periodic consistency index. The periodic consistency index at least comprehensively reflects the proportion of consistent difference signs and the proportion of difference amplitude within a preset allowable range. Specifically, the higher the proportion of difference items in the periodic variation difference sequence that meet the requirements of consistent change direction and amplitude difference within the allowable range, the larger the corresponding periodic consistency index value, thereby realizing a quantitative characterization of the consistency of the zero-crossing periodic change trend between unidentified sites and identified neighboring sites.

Images