Substation intelligent agent-based device security management method and device

Abstract

The application provides a device safety management method and device based on a transformer substation intelligent agent, which comprises the following steps: based on each transformer substation device, the running state of the transformer substation device is locally perceived to obtain a perception data set; based on a hidden association perception model, the potential coupling relationship between the running characteristics is analyzed in combination with the perception data set to obtain hidden association characteristics; based on an explicit logic reasoning model, fault risk analysis is performed in combination with the hidden association characteristics to obtain the safety decision result of each transformer substation device; based on the safety decision result of adjacent transformer substation devices, decision cooperation is performed to obtain the cooperative decision result of each transformer substation device; based on the cooperative decision result, potential risk pattern recognition is performed to obtain the safety situation recognition result of each transformer substation device; based on the safety situation recognition result, a safety intervention instruction is generated, and a local safety response is performed on the transformer substation device based on the safety intervention instruction. The application improves the timeliness and accuracy of device safety management and control.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a method and apparatus for equipment safety management based on a substation intelligent agent. Background Technology

[0002] In the field of substation equipment safety management, there exists an automated inspection method based on a fixed rule base. This method periodically compares and judges the operating status of various types of equipment in the substation through a preset set of rules. Once an operating parameter deviates from a set threshold, an alarm is triggered or a preset control command is executed.

[0003] However, the rule base is statically configured and relies solely on explicit logical conditions predefined by humans. It cannot be combined with the implicit correlation characteristics contained in the multi-source sensor data generated during equipment operation for comprehensive judgment. As a result, when faced with atypical anomalies or early compound faults, it is difficult to effectively identify risk precursors, which can easily lead to missed or false alarms, thereby affecting the timeliness and accuracy of equipment safety management. Summary of the Invention

[0004] This invention provides a method and apparatus for equipment safety management based on substation intelligent agents, aiming to solve the problem of missed reports caused by the lack of perception of latent fault precursors in fixed rule base methods, and to improve the timeliness and accuracy of equipment safety management and control.

[0005] In a first aspect, the present invention provides a method for equipment safety management based on a substation intelligent agent, comprising: Based on the local perception of the operating status of each substation equipment, a perception dataset of each substation equipment is obtained. Based on the implicit association perception model and the perception dataset, the potential coupling relationship between the operating characteristics is analyzed to obtain the implicit association characteristics of each substation equipment. Then, based on the explicit logic reasoning model and the implicit association characteristics, the fault risk analysis is carried out to obtain the safety decision results of each substation equipment. Decision collaboration is performed based on the safety decision results between adjacent substation equipment to obtain the collaborative decision results of each substation equipment. Based on the collaborative decision results, potential risk pattern identification is performed to obtain the safety status identification results of each substation equipment. Based on the security situation identification results, a security intervention command is generated, and a local security response is executed on the substation equipment based on the security intervention command.

[0006] In a second aspect, the present invention also provides an equipment safety management device based on a substation intelligent agent, for implementing the equipment safety management method based on a substation intelligent agent as described in the first aspect; the equipment safety management device based on a substation intelligent agent includes: The local sensing module is used to locally sense the operating status of the equipment in each substation based on the equipment in each substation, and obtain the sensing dataset of each substation equipment. The model intelligent analysis module is used to analyze the potential coupling relationship between operating features based on the implicit association perception model and the perception dataset, to obtain the implicit association features of each substation equipment, and to perform fault risk analysis based on the explicit logic reasoning model and the implicit association features, to obtain the safety decision results of each substation equipment. The risk pattern recognition module is used to perform decision collaboration based on the safety decision results between adjacent substation equipment to obtain the collaborative decision results of each substation equipment, and to perform potential risk pattern recognition based on the collaborative decision results to obtain the safety status recognition results of each substation equipment. The safety response module is used to generate safety intervention commands based on the safety situation identification results, and to perform local safety responses on substation equipment based on the safety intervention commands.

[0007] Thirdly, the present invention also provides an electronic device, comprising: a memory for storing computer software programs; and a processor for reading and executing the computer software programs, thereby realizing the equipment safety management method based on substation intelligent agents as described above.

[0008] Fourthly, the present invention also provides a non-transitory computer-readable storage medium storing a computer software program, which, when executed by a processor, implements the equipment safety management method based on a substation intelligent agent as described above.

[0009] Fifthly, the present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the equipment safety management method based on a substation intelligent agent as described above.

[0010] The equipment safety management method based on substation intelligent agents provided in this invention analyzes the potential coupling relationships between operating characteristics using an implicit correlation perception model based on a perception dataset, obtaining implicit correlation features of each substation device. This allows for the mining of implicit correlations contained in multi-source sensor data, overcoming the limitation of fixed rule bases that can only identify manually preset explicit logic, and providing a basis for identifying early complex faults and atypical anomalies. Then, based on an explicit logic reasoning model combined with the implicit correlation features, fault risk analysis is performed to obtain safety decision results for each substation device, achieving a combination of implicit correlation features and explicit logic, avoiding the one-sidedness of single logical judgments, and reducing the possibility of missed or false alarms. Based on the safety decision results, collaborative decision-making between adjacent substation devices is used to obtain collaborative decision results for each substation device. Cross-device collaboration further verifies the accuracy of decisions and avoids the bias of single-device decisions. Based on the collaborative decision results, potential risk pattern recognition is performed to obtain the safety status recognition results for each substation device, accurately capturing risk precursors and solving the pain point of fixed rule bases' difficulty in identifying atypical anomalies and early faults. Based on the security situation identification results, security intervention instructions are generated, and local security responses are executed on substation equipment based on the security intervention instructions. This enables timely handling of risks, solves the problem of missed reporting caused by the lack of perception of latent fault precursors in fixed rule base methods, and improves the timeliness and accuracy of equipment safety management. Attached Figure Description

[0011] Figure 1 This is a flowchart illustrating the equipment safety management method based on a substation intelligent agent provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of the equipment safety management device based on a substation intelligent agent provided in an embodiment of the present invention; Figure 3 An embodiment diagram of the electronic device provided in this invention; Figure 4 An embodiment diagram of a computer-readable storage medium provided in accordance with the present invention. Detailed Implementation

[0012] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0013] See Figure 1 , Figure 1This is a flowchart illustrating the equipment safety management method based on a substation intelligent agent provided by the present invention. In this embodiment, the executing entity of the equipment safety management method based on a substation intelligent agent is an equipment management device. Therefore, the equipment safety management method based on a substation intelligent agent includes: Step 10: Based on the local perception of the operating status of each substation equipment, obtain the perception dataset of each substation equipment.

[0014] Optionally, each substation device can be controlled to perform local sensing operations on the operating status of its own substation equipment, acquiring a real-time sensing dataset for each device. Local sensing refers to the real-time data collection performed by the sensing modules built into the substation equipment at the operating site, without relying on remote data transmission to obtain basic operating data, ensuring the real-time nature and accuracy of the sensing data. Substation equipment refers to various power equipment used for the normal operation of a substation, including transformers, circuit breakers, disconnectors, and instrument transformers. Operating status refers to various real-time operating parameters and working conditions of the substation equipment during operation, including operating voltage, operating current, operating temperature, insulation status, and mechanical action status. The sensing dataset refers to the collection of all data acquired through local sensing that can completely characterize the operating status of each substation device. Data within the dataset must be labeled with the corresponding substation equipment identifier, sensing time, and sensing parameter type.

[0015] Optionally, in the specific sensing process of this embodiment of the invention, a local sensing start command is sent to each substation device, specifying the sensing parameter type, sensing frequency, sensing duration, and other requirements. Upon receiving the command, each substation device activates its own sensing module and collects data on its own operating status locally according to the command requirements. During the collection process, the data is initially verified in real time, and obviously abnormal or invalid data (such as data exceeding the normal operating parameter range of the device) is removed. After the collection is completed, each substation device uploads the verified sensing data. All uploaded sensing data are summarized and grouped according to the substation device identifier to form an independent sensing dataset corresponding to each substation device.

[0016] Step 20: Based on the implicit association perception model and the perception dataset, analyze the potential coupling relationship between operating features to obtain the implicit association features of each substation equipment. Then, based on the explicit logic reasoning model and the implicit association features, conduct fault risk analysis to obtain the safety decision results of each substation equipment.

[0017] Optionally, based on the implicit association perception model and the perception dataset, the potential coupling relationship between the operating characteristics of each substation equipment is analyzed to obtain the implicit association characteristics of each substation equipment, as in steps 201 to 204.

[0018] Among them, the implicit association feature characterizes the state coupling capability of substation equipment with other equipment through the electrical network structure when the equipment is in operation without triggering protection actions or alarm signals.

[0019] Furthermore, based on the implicit correlation features obtained by combining the explicit logical reasoning model, fault risk analysis is performed on each substation equipment to obtain the safety decision results of each substation equipment, as detailed in steps 205 to 208.

[0020] The safety decision results are generated based on the risk consequence level of the hidden risk transmission path and the existence status of the corresponding effective risk source equipment. This includes the type of handling instructions for substation equipment. For example, for a hidden risk transmission path with a high risk consequence level, the safety decision result generated for the effective risk source equipment is "execute local isolation operation"; for a hidden risk transmission path with a medium risk consequence level, the safety decision result generated for the effective risk source equipment is "start enhanced monitoring mode"; and for a hidden risk transmission path with a low risk consequence level, the safety decision result generated for the effective risk source equipment is "maintain current operating status".

[0021] Step 30: Based on the safety decision results between adjacent substation equipment, perform decision collaboration to obtain the collaborative decision results of each substation equipment, and based on the collaborative decision results, perform potential risk pattern identification to obtain the safety status identification results of each substation equipment.

[0022] Optionally, a decision coordination operation is performed based on the safety decision results between adjacent substation equipment to obtain the coordinated decision results for each substation equipment, as described in steps 301 to 304. The coordinated decision result represents the handling instruction corresponding to the final safety response role of each substation equipment. Adjacent substation equipment refers to substation equipment that has a direct electrical connection in the electrical network structure, is geographically adjacent, and whose operating states influence each other.

[0023] Optionally, potential risk pattern identification refers to analyzing collaborative decision-making results to uncover potential, undetected risk patterns that may exist during the operation of various substation equipment, including the patterns of risk occurrence and the characteristics of risk transmission paths. Safety situation identification results refer to comprehensive judgments that fully characterize the current safety status, potential risks, and risk development trends of each substation's equipment.

[0024] Optionally, the collaborative decision-making results are classified and sorted, and the substation equipment is grouped according to the type of disposal instruction (execute local isolation operation, start enhanced monitoring mode, maintain current operating status). At the same time, the adjacent equipment information and implicit correlation characteristics of each group of equipment are marked to clarify the equipment distribution and the correlation between equipment corresponding to different disposal instructions.

[0025] Furthermore, the sorted collaborative decision-making results and related data are compared and matched with a pre-set risk pattern database. The comparison includes the degree of fit between the characteristics of the combination of disposal instructions, the characteristics of equipment relationships, the characteristics of implicit relationships and the risk patterns, to determine the type of potential risk pattern corresponding to the current collaborative decision-making results, including single-equipment implicit risk patterns, multi-equipment related transmission risk patterns, regional concentrated risk patterns, etc.

[0026] Furthermore, the matched potential risk patterns are verified by combining the perception dataset to check whether there are abnormal tendencies in the device operating parameters corresponding to the risk patterns. Risk patterns with incorrect matching or invalid verification are eliminated to confirm the actual potential risk patterns.

[0027] Furthermore, based on the confirmed potential risk patterns, and combined with the risk level classification standards and risk development trend judgment rules in the risk pattern library, the current safety status (safe, general risk, relatively high risk, high risk), the severity of potential risks, the possible transmission paths and development trends of risks of each substation equipment are analyzed one by one to generate the safety status identification results corresponding to each substation equipment.

[0028] In one embodiment, for example, the collaborative decision-making results are: the final disposal instruction for transformer 1 is "execute local isolation operation", the final disposal instruction for transformer 2 is "start enhanced monitoring mode", the final disposal instruction for transformer 3 is "start enhanced monitoring mode", the final disposal instruction for circuit breaker 1 is "start enhanced monitoring mode", and the final disposal instruction for circuit breaker 2 is "maintain current operating state". The adjacent device relationships are: transformer 1 and circuit breaker 1 are directly electrically connected and geographically adjacent, transformer 2 and transformer 3 are directly electrically connected and geographically adjacent, transformer 2 and circuit breaker 2 are directly electrically connected and geographically adjacent, and all devices have no other adjacent relationships.

[0029] The collaborative decision-making results of the above 5 devices are summarized and grouped according to the type of disposal instruction: Group 1 (execute local isolation operation): Transformer 1; Group 2 (start enhanced monitoring mode): Transformer 2, Transformer 3, Circuit Breaker 1; Group 3 (maintain current operating status): Circuit Breaker 2; At the same time, the sensing dataset of step 10 is associated (it is known that the sensing data of Transformer 1 has shown abnormal temperature and has been removed, and the sensing data of the other devices have no obvious abnormalities) and the implicit association features of step 20 (it is known that Transformer 1 and Circuit Breaker 1 have a strong state coupling ability, Transformer 2 and 3 have a strong state coupling ability, and Transformer 2 and Circuit Breaker 2 have a weak state coupling ability).

[0030] A pre-defined risk pattern library is invoked, containing three types of related risk patterns: Pattern 1 (Single Device Implicit Risk Pattern): The single device's handling instruction is "Execute local isolation operation," and the adjacent device's handling instruction is "Activate enhanced monitoring mode," with strong state coupling between the two. The corresponding risk is a tendency for the implicit fault propagation of a single device. Pattern 2 (Multi-Device Associated Transmission Risk Pattern): The handling instructions for multiple adjacent devices are all "Activate enhanced monitoring mode," and there is strong state coupling between the devices. The corresponding risk is a multi-device associated implicit risk with no obvious propagation tendency. Pattern 3 (No Risk Pattern): The device's handling instruction is "Maintain current operating state," adjacent devices have no high-risk handling instructions, and the state coupling is weak, corresponding to no obvious potential risk. Through comparison, the collaborative decision-making results and associated features of transformer 1 and circuit breaker 1 match Pattern 1; the collaborative decision-making results and associated features of transformer 2 and transformer 3 match Pattern 2; and the collaborative decision-making results and associated features of circuit breaker 2 match Pattern 3.

[0031] Verification was performed using the sensing dataset: Transformer 1 experienced temperature anomalies (although invalid data was removed, there was still a potential hidden fault), the sensing data of circuit breaker 1 showed no obvious anomalies, but it was strongly coupled with transformer 1, verifying that mode 1 was true and effective; the sensing data of transformers 2 and 3 showed no obvious anomalies, but they were strongly coupled and both required enhanced monitoring, verifying that mode 2 was true and effective; the sensing data of circuit breaker 2 showed no anomalies, and its coupling with the adjacent transformer 2 was weak, and transformer 2 only required enhanced monitoring without high risk, verifying that mode 3 was true and effective, with no risk of mismatch or invalidity.

[0032] Based on the verified risk patterns and the rules in the risk pattern library, security situation identification results are generated: Transformer 1: The safety level is high risk, the potential risk mode is the hidden risk mode of a single device, and the risk development trend is the possibility of the fault spreading to circuit breaker 1. The focus of subsequent attention is the effectiveness of isolation operation and the investigation of potential faults. Circuit breaker 1: The safety level is high risk, and the potential risk mode is the single device hidden risk mode (affected by transformer 1). The risk development trend is that if the failure of transformer 1 spreads, it is prone to cascading anomalies. The focus is on enhancing the monitoring of parameter changes and coupling status with transformer 1 during the monitoring process. Transformer 2: The safety level is general risk, the potential risk mode is multi-device interconnection and transmission risk mode, the risk development trend is stable risk with no obvious tendency to spread, and the focus of subsequent attention is to enhance the monitoring of operating parameters and the correlation status with transformer 3 during the monitoring process. Transformer 3: The safety level is general risk, the potential risk mode is multi-device interconnection and transmission risk mode, the risk development trend is stable risk with no obvious tendency to spread, and the focus of subsequent attention is to enhance the monitoring of operating parameters and the correlation status with transformer 2 during the monitoring process. Circuit breaker 2: Safety level is safe, potential risk mode is risk-free, risk development trend is stable operation with no potential risks, the focus is on maintaining normal operation and periodically synchronizing sensing data.

[0033] Step 40: Generate a security intervention command based on the security situation identification result, and execute a local security response on the substation equipment based on the security intervention command.

[0034] Optionally, a safety intervention instruction refers to an instruction sent to the substation equipment to adjust its operating status and handle risks in response to existing safety risks. The instruction type corresponds to the handling instruction type in step 20. Local safety response refers to the substation equipment performing the corresponding operation locally according to the instruction requirements after receiving the safety intervention instruction.

[0035] Optionally, the safety situation identification results are analyzed one by one to determine the current safety status, potential risk level, and risk development trend of each substation device. Based on the analysis results, a corresponding safety intervention instruction template is matched, and specific information such as the instruction recipient and execution parameters are filled in to generate a personalized safety intervention instruction. Furthermore, the safety intervention instruction is sent to the corresponding substation device through a dedicated communication channel, while the instruction transmission status is monitored in real time to ensure successful delivery of the instruction.

[0036] After receiving a safety intervention command, the substation equipment parses the command, executes the local safety response operation as required by the command, and sends back the response result after completion. The response effect is verified to ensure that the substation equipment has completed the safety response according to the command.

[0037] In one embodiment, for example, the security situation identification result is: transformer 1 has a high risk and the risk has a further upward trend; circuit breaker 1 has a medium risk and the risk is stable; transformer 2, transformer 3, and circuit breaker 2 have no obvious risks and their operating status is stable.

[0038] The above security situation identification results were analyzed one by one: Transformer 1 is of high risk and the risk is expanding, so isolation measures need to be taken immediately; Circuit breaker 1 is of medium risk and its status is stable, so monitoring needs to be strengthened; The remaining 3 devices are of no risk and can maintain their current operation. Based on the analysis results, three personalized security intervention instructions were generated: Command 1 (Recipient: Transformer 1): "Execute local isolation operation", Execution time: Immediate execution, Execution parameters: Disconnect all incoming and outgoing switches of Transformer 1, cut off its connection with the electrical network, and mark the isolation point. Command 2 (Recipient: Circuit Breaker 1): "Activate enhanced monitoring mode", Execution time: Immediate execution, Execution parameters: Adjust the sensing frequency to collect operating parameters once every 5 seconds, focusing on monitoring the contact resistance and operating status of Circuit Breaker 1; report any abnormal data immediately. Command 3 (Recipient: Transformer 2, Transformer 3, Circuit Breaker 2): "Maintain current operating status", Execution time: Continuous execution, Execution parameters: Collect operating data according to the original sensing frequency (once every 10 seconds); normal operation is sufficient. The three instructions are sent to the corresponding substation equipment through a dedicated communication channel, and the instruction sending status is monitored in real time to confirm that all instructions are successfully delivered.

[0039] Upon receiving the instruction, each substation device executes a local safety response: Transformer 1 immediately disconnects all incoming and outgoing line switches, completes local isolation, and reports "Isolation operation completed"; Circuit breaker 1 adjusts the sensing frequency to once every 5 seconds, starts enhanced monitoring, and reports "Enhanced monitoring mode startup completed"; Transformer 2, Transformer 3, and Circuit breaker 2 maintain their original operating status and report "Current operating status remains normal". Verify the response effect: It is confirmed that transformer 1 is completely isolated and has no electrical connection; circuit breaker 1 has been enhanced monitoring according to the new parameters; the other 3 devices are operating normally, and the local safety response operation has been completed.

[0040] The embodiments of the present invention solve the problem of missed detection caused by the lack of perception of latent fault precursors in fixed rule base methods, thereby improving the timeliness and accuracy of equipment safety management and control.

[0041] Optionally, the processes of steps 201 to 204 include: Step 201: Based on the multi-dimensional operation feature sequence contained in the perception dataset of each substation equipment, determine the set of operation status trajectories of each substation equipment in the time dimension, and extract the feature evolution path of each substation equipment in different operation stages based on the set of operation status trajectories.

[0042] Optionally, the multidimensional operating feature sequence refers to the sequence formed by arranging various types of operating parameters used to characterize the operating status of substation equipment in the sensing dataset according to the sensing time order. Each operating parameter corresponds to a dimension, including operating voltage sequence, operating current sequence, operating temperature sequence, insulation resistance sequence, etc.

[0043] The time dimension refers to the chronological order of the data collection, forming a continuous time axis with the sensing moment as the node; the set of operating status trajectories refers to the set of all trajectories that associate the multidimensional operating feature sequences of each substation device with the corresponding sensing time, forming a complete record of the changes in the operating status of the device throughout the entire sensing period. Each trajectory corresponds to a complete record of the changes in multidimensional operating features over time, and the trajectory must be labeled with the corresponding substation device identifier to ensure that the trajectory can be associated with a specific device.

[0044] The operation phase refers to dividing the entire sensing period into multiple consecutive time periods based on the changing patterns of the substation equipment's operating status. Within each time period, the equipment's operating status is relatively stable without significant abrupt changes. The division criteria include the fluctuation range of equipment operating parameters and the switching of operating modes.

[0045] The feature evolution path refers to the path formed by the changing trend, magnitude and law of the multidimensional operating characteristics of each substation equipment over time within a single operating phase, which can reflect the evolution of the operating status of the equipment within that operating phase.

[0046] Optionally, all multidimensional operating feature sequences are extracted from the sensing dataset of each substation equipment. The sequences are then cleaned to remove minor abnormal data and duplicate data that were not removed during the sensing process, ensuring the integrity and accuracy of the multidimensional operating feature sequences.

[0047] Furthermore, the cleaned multidimensional operational feature sequence is associated with the corresponding sensing time, and the feature data is sorted according to the chronological order of sensing time to form a single operational state trajectory of each substation device in the time dimension. All single operational state trajectories are summarized to obtain the set of operational state trajectories corresponding to each substation device.

[0048] Furthermore, based on the preset operation phase division rules, the set of operation status trajectories for each device is divided into phases. The preset division rules include operation parameter fluctuation thresholds, operation mode switching indicators, etc. When the fluctuation amplitude of operation parameters in a certain time period in the trajectory is less than the preset fluctuation threshold and there is no operation mode switching indicator, the time period is divided into an independent operation phase.

[0049] Furthermore, for each operating stage, information such as the changing trend, magnitude, and sequence of multi-dimensional operating characteristics within that stage is extracted and integrated to form the characteristic evolution path corresponding to that operating stage. The characteristic evolution paths of all operating stages of the same equipment are summarized to obtain the characteristic evolution paths of each substation equipment in different operating stages.

[0050] Step 202: Based on the direction and sequence of change of operating characteristics between adjacent sampling times in the evolution path, determine the temporal dependency relationship between different operating characteristics inside each substation equipment, and identify the feature linkage mode that continuously repeats during the triggering of fault-free alarm signals based on the temporal dependency relationship.

[0051] Optionally, adjacent sampling moments refer to two adjacent sensing moments in the feature evolution path according to the sequential order of sensing time, and the time interval between adjacent sampling moments is consistent with the preset sensing frequency in step 10.

[0052] The direction of change refers to the trend of a certain operating feature’s value change between adjacent sampling times, including three cases: rising, falling, and remaining unchanged; the order of change refers to the sequence in which multiple operating features change between adjacent sampling times; the temporal dependency relationship refers to the mutual dependence and mutual influence between different operating features within each substation equipment in the time dimension, that is, the relationship that a change in a certain operating feature will trigger or accompany other operating features to change accordingly at subsequent sampling times or simultaneously.

[0053] The period during which no fault alarm signal is triggered refers to the period during which the substation equipment does not issue any fault alarm signal or trigger any protection action, and the equipment appears to be in normal operating condition with no obvious signs of fault.

[0054] Feature linkage mode refers to the combination of feature changes that occur repeatedly based on time-series dependencies among different operating features within the equipment during the period when no fault alarm signal is triggered. This combination of features has a fixed pattern and order of change and can reflect the linkage pattern of the operating features within the equipment.

[0055] Optionally, for each feature evolution path of each device, all adjacent sampling times in the path are extracted one by one, and the direction of change of each operating feature between each adjacent sampling time is analyzed to determine whether each feature is rising, falling or remaining unchanged. At the same time, the order in which multiple operating features change is recorded to form the feature change information corresponding to each adjacent sampling time.

[0056] Furthermore, the feature change information of all adjacent sampling times is summarized and analyzed to explore the change correlation between different operating features. When an operating feature changes in a specific direction at a certain sampling time, and another operating feature changes in the corresponding direction at subsequent adjacent sampling times, it is determined that there is a temporal dependency between the two operating features. All operating features are analyzed one by one to obtain the complete temporal dependency between different operating features within each substation equipment.

[0057] Furthermore, the characteristic evolution path during the period of no fault alarm signal triggering is screened out, and the evolution path during the period of fault alarm signal triggering and protection action initiation is excluded.

[0058] Furthermore, based on the temporal dependency, pattern recognition is performed on the feature evolution path during the triggering of fault-free alarm signals to screen out the feature change combinations that appear repeatedly. That is, the same feature change combination appears multiple times at different adjacent sampling times and different operating stages, and the change direction, change order and temporal dependency are consistent. This type of feature change combination is identified as a feature linkage mode.

[0059] Step 203: Based on the co-occurrence of feature linkage modes among multiple substation devices, determine the set of events in which cross-device operating characteristics change synchronously within the same time window, and filter out feature change combinations whose deviation at the start time of change does not exceed a preset deviation threshold based on the event set.

[0060] Optionally, co-occurrence refers to the situation where the characteristic linkage modes of multiple substation devices appear simultaneously and occur synchronously within the same time period. Cross-device operation characteristics refer to the individual operation characteristics of different substation devices, which are interconnected through the electrical network structure. The same time window refers to a preset fixed time period, the duration of which is reasonably set according to the operating characteristics and sensing frequency of the substation devices, and is used to determine whether the characteristic changes of different devices are synchronous.

[0061] An event set refers to a collection of events in which the cross-equipment operating characteristics of multiple substation devices change synchronously within the same time window. Each event contains information such as the equipment identifier, operating characteristic type, direction of change, start time of change, and end time of change of the equipment involved in the synchronous change.

[0062] The change start time refers to the sensing time when a certain operating characteristic begins to change; the preset deviation threshold refers to the maximum time deviation set in advance to determine whether the changes of two or more operating characteristics are synchronized, and is reasonably set according to the sensing frequency and the equipment operation response speed.

[0063] Optionally, the occurrence time, duration, included operational characteristic change information, and corresponding equipment identifier of each feature linkage mode are recorded. Further, a preset time window is set, and the occurrence times of all feature linkage modes are iterated through. The system analyzes whether feature linkage modes of different substation equipment occur simultaneously within the same time window, and statistically analyzes the co-occurrence frequency, co-occurrence duration, and co-occurrence equipment combinations of feature linkage modes among multiple substation devices.

[0064] Furthermore, based on the co-occurrence situation, cross-device operation features that change synchronously within the same time window are extracted. Each synchronously changing scenario is treated as an independent event. Information such as the identification of each substation equipment involved in the synchronous change, the corresponding operation feature type, the direction of feature change, the start time of change, and the end time of change are recorded. All such events are summarized to form a set of events in which cross-device operation features change synchronously within the same time window.

[0065] Finally, the preset deviation threshold is invoked to analyze the cross-device operational feature changes contained in each event in the event set. The start times of the changes of different operational features in the same event are compared, the time deviation between the start times of each change is calculated, and feature changes in which the time deviation between the start times of all changes does not exceed the preset deviation threshold are selected. These feature changes are then integrated into feature change combinations.

[0066] Step 204: Response instance analysis is performed based on the combination of feature changes to obtain the implicit correlation features of each substation device.

[0067] Optionally, response instance analysis can be performed based on the combination of feature changes to obtain the implicit correlation features of each substation equipment, as in steps 2041 to 2044.

[0068] The embodiments of the present invention can accurately perceive the latent operating state of substation equipment when no protection action or alarm signal is triggered, uncover the hidden operating characteristic coupling relationship between equipment and equipment, form a latent correlation feature that can characterize the latent state coupling ability between equipment, make up for the omission defect caused by the lack of perception ability of latent fault precursors, and improve the timeliness and accuracy of equipment safety management and control.

[0069] Optionally, the process of steps 2041 to 2044 includes: Step 2041: Based on the substation equipment pairs corresponding to the combination of feature changes, determine the instances of coordinated response of the operating features between the equipment under the conditions of physical connection or electrical topology proximity.

[0070] Optionally, a substation equipment pair refers to a combination of two interconnected substation devices that participate in the synchronous change of cross-equipment operating characteristics in a characteristic change combination. Each characteristic change combination corresponds to at least one substation equipment pair. If a characteristic change combination includes multiple cross-equipment operating characteristics, then it corresponds to multiple substation equipment pairs. Physical connection refers to the direct connection between two substation devices through physical carriers such as power lines, cables, and busbars, forming a path for transmitting power signals.

[0071] Electrical topology proximity refers to the condition where two substation devices, although not directly physically connected, are at adjacent levels in the substation's electrical topology, indirectly connected through other devices, and their electrical signals can be rapidly transmitted and influence each other.

[0072] A coordinated response instance refers to a specific instance in which the operating characteristics of a pair of equipment from two substations change synchronously based on a combination of characteristic changes, under conditions of physical connection or electrical topology proximity. Each coordinated response instance corresponds to a unique combination of characteristic changes and a pair of substation equipment, and includes detailed information such as equipment pair identification, characteristic change information, change synchronization status, and connection condition type.

[0073] Optionally, each feature change combination is analyzed one by one to extract the identifiers of all substation equipment participating in the synchronous change. Corresponding substation equipment pairs are generated based on the equipment identifier combinations, ensuring that each equipment pair corresponds to two substation devices participating in the same feature change combination, and that no duplicate equipment pairs are generated. Further, a preset substation electrical topology database and equipment physical connection information database are retrieved. The electrical topology database stores information such as the electrical topology location, hierarchical relationship, and indirect connection paths of all substation equipment, while the equipment physical connection information database stores information such as the direct physical connection relationships, connection carriers, and connection status of all substation equipment.

[0074] Furthermore, the connection conditions of each substation equipment pair are checked one by one to determine whether there is a direct physical connection between the equipment pairs or whether the electrical topology proximity condition is met. During the check, the electrical topology and physical connection information are combined to exclude equipment pairs that have no physical connection and do not meet the electrical topology proximity condition. Finally, the substation equipment pairs that meet the physical connection or electrical topology proximity conditions are associated with the corresponding characteristic change combinations. Information such as the connection condition type of the equipment pair and the synchronous change of operating characteristics is recorded. Each such association scenario is taken as a collaborative response instance. All collaborative response instances are summarized to obtain the collaborative response instances of the operating characteristics between equipment under the physical connection or electrical topology proximity conditions.

[0075] Step 2042: Based on the cooperative response instances, exclude common cause responses caused by the same external power grid disturbance event, and retain characteristic coupling instances that occur under the condition that there is a direct electrical path between the devices.

[0076] Optionally, the same external power grid disturbance event refers to an abnormal power grid event occurring outside the substation that synchronously affects the operating status of multiple substation devices. This includes external power grid voltage fluctuations, frequency anomalies, lightning strikes, and external short-circuit faults. Such events cause synchronous characteristic changes in multiple unrelated devices, but are not caused by state coupling between the devices themselves. Common-cause response refers to the synchronous change in the operating characteristics of multiple substation devices, not caused by state coupling between devices, but by a synchronous response phenomenon caused by the same external power grid disturbance event. Direct electrical path condition refers to the existence of a direct path for transmitting electrical signals and energy between two substation devices. This path can be a direct physical connection or an unobstructed, fast-conducting indirect electrical path formed through other devices, and the path must be in normal operating condition, enabling mutual transmission and influence of the operating status between the devices.

[0077] Feature coupling instances refer to instances where, after excluding common cause responses, the operating characteristics of a pair of equipment consisting of two substation devices change synchronously due to the inherent state coupling relationship between the equipment, under the condition of a direct electrical path. These instances can truly reflect the implicit correlation between the equipment.

[0078] Optionally, a pre-defined external power grid disturbance event database stores historical and real-time information on external power grid disturbance events, including disturbance event type, occurrence time, impact range, disturbance intensity, and duration. Therefore, each coordinated response instance is analyzed individually, extracting information such as the occurrence time, duration, and magnitude of synchronous changes in operational characteristics within the instance. This information is then compared with the disturbance event information in the external power grid disturbance event database to determine whether the synchronous changes in characteristics corresponding to the coordinated response instance are caused by the same external power grid disturbance event. During the comparison, the focus is on verifying whether the occurrence time of the synchronous changes in characteristics is consistent with the occurrence time of the disturbance event, whether the duration of the changes matches the duration of the disturbance event, and whether the magnitude of the changes is within the impact range of the disturbance event. If all three conditions are met, the coordinated response instance is determined to be a common-cause response. Further, all coordinated response instances determined to be common-cause responses are excluded, while those not determined to be common-cause responses are retained.

[0079] Furthermore, the substation equipment pairs corresponding to the retained coordinated response instances are re-examined to confirm that there is a direct electrical path between the equipment pairs. Instances without a direct electrical path are excluded, and the final retained instances are determined as characteristic coupling instances.

[0080] Step 2043: Based on the feature coupling instances, determine the set of associated events that reflect the impact of the transmission of operating status between substation equipment through non-control command paths, and construct an initial local implicit association topology with substation equipment as nodes and feature coupling instances as edges based on the associated event set.

[0081] Optionally, a non-control command path refers to a path between two substation devices that transmits operational status influences without relying on sent control commands, but rather through their own electrical connections and operational status coupling relationships. This path is a naturally formed implicit transmission path between devices. Operational status influence transmission refers to the process where, after the operational status of one substation device changes, this change is transmitted to another related substation device through a direct electrical path, causing a corresponding change in the operational status of the other device. An associated event set refers to a collection of multiple associated events, each generated based on a feature coupling instance. It reflects the specific circumstances of the transmission of operational status influences between two substation devices through a non-control command path. Each associated event includes device pair identification, operational status influence transmission direction, transmission strength, transmission duration, and corresponding feature coupling instance information. The initial local implicit association topology refers to a topology constructed using substation devices as nodes and feature coupling instances as edges to reflect the implicit association relationships between substation devices within a local area. Nodes represent individual substation devices, and edges represent the implicit association relationships between two devices, i.e., the coupling relationships corresponding to the feature coupling instances.

[0082] Optionally, each feature coupling instance is analyzed one by one to extract the operating status change information of the two substation devices in the instance, determine the transmission direction of the operating status influence (i.e., which device's status change occurs first and is transmitted to the other device), and analyze the transmission intensity (i.e., the degree of influence of one device's status change on the other device's status change, which is determined by the correlation of the status change amplitude) and the transmission duration (i.e., the duration of the influence transmission).

[0083] Furthermore, the scenario where the operating status is transmitted between devices through non-control command paths, corresponding to each feature coupling instance, is identified as an associated event. Information such as the device pair identifier, the direction of influence transmission, the transmission intensity, the transmission duration, and the corresponding feature coupling instance identifier of the associated event are recorded. All associated events are then aggregated to form an associated event set.

[0084] Furthermore, the identification information of all substation equipment is retrieved, and each substation equipment is treated as an independent node, with the node identification being consistent with the substation equipment identification.

[0085] Furthermore, based on the associated events in the associated event set, the nodes corresponding to two substation devices that have feature coupling instances (i.e., have implicit association relationships) are connected by edges. The attributes of the edges are set to information such as the coupling strength and influence propagation direction of the corresponding feature coupling instances, thus constructing an initial local implicit association topology structure with substation devices as nodes and feature coupling instances as edges. This structure only reflects the association situation between local devices that have implicit association relationships.

[0086] Step 2044: Perform correlation topology analysis based on the initial local implicit correlation topology to obtain the implicit correlation characteristics of each substation device.

[0087] Optionally, correlation topology analysis is performed based on the initial local implicit correlation topology to obtain the implicit correlation characteristics of each substation device, as described in steps 20441 to 20445.

[0088] The embodiments of this invention accurately uncover the implicit relationships between substation equipment when protection actions or alarm signals are not triggered. This overcomes the limitations of fixed rule bases, which cannot capture implicit relationships and are prone to missing latent fault precursors. It can accurately obtain the state coupling capabilities between equipment, thereby improving the timeliness and accuracy of equipment safety management and ensuring that latent fault hazards can be identified in advance.

[0089] Optionally, the processes of steps 20441 to 20445 include: Step 20441: Based on the operational feature type, the temporal order of coupling, and the electrical path type corresponding to each edge in the initial local implicit association topology, determine the coupling behavior label of each associated edge.

[0090] Optionally, the operating characteristic type refers to the specific category of the operating characteristics of the substation equipment involved in the coupling within the characteristic coupling instance corresponding to the associated edge, including operating voltage, operating current, operating temperature, insulation resistance, etc. The timing of coupling refers to the order in which the operating characteristics of the two substation equipment in the characteristic coupling instance corresponding to the associated edge change, that is, which equipment's operating characteristic changes first, and then propagates to the other equipment, triggering a corresponding change.

[0091] Electrical path type refers to the specific category of direct electrical path for transmitting the impact of operating status between two substation devices corresponding to the associated edge. It includes direct physical connection path (path formed by direct connection through physical carriers such as power lines, cables, and busbars) and indirect electrical connection path (path formed by indirect connection through other substation devices, which is unobstructed and can be quickly transmitted).

[0092] Optionally, all associated edges in the initial local implicit association topology are extracted, and the identifier of each associated edge, the corresponding node (substation equipment) identifier, and the attribute information of the edge are recorded one by one.

[0093] Furthermore, each associated edge is associated with the corresponding feature coupling instance in step 2043. The operating feature type and the time sequence of coupling occurrence of the associated edge are extracted from the feature coupling instance. At the same time, the direct electrical path information between the two substation devices corresponding to the associated edge, which was checked in step 2042, is retrieved to determine the electrical path type.

[0094] Optionally, a preset rule for generating coupling behavior labels is established. This rule requires the labels to include the operational characteristic type, coupling time sequence (expressed in "first-last" format, such as "Transformer 1-Circuit Breaker 1," indicating that transformer 1 changes first and the change is transmitted to circuit breaker 1), and electrical path type. Further, a corresponding coupling behavior label is generated for each associated edge according to the preset generation rule. The label corresponds one-to-one with the associated edge, and the coupling behavior label is added to the attribute information of each associated edge.

[0095] Step 20442: Based on the coupling behavior label, perform semantic consistency verification on the association path composed of multiple edges in the topology, remove association paths with conflicting running feature types or contradictory time order, and obtain the target local implicit association topology.

[0096] Optionally, an associated path refers to a path in the initial local implicit associated topology, formed by two or more associated edges connected sequentially, connecting multiple substation equipment nodes in series. Adjacent associated edges in the path share a single node (substation equipment), reflecting the implicit association relationships passed sequentially between multiple substation equipment. Semantic consistency verification refers to a comprehensive check of the coupling behavior labels of all associated edges in the associated path to determine whether the coupling behavior of each associated edge in the path is logical and consistent, ensuring that the associated path can truly reflect the continuous implicit association transmission process between multiple devices.

[0097] Running feature type conflict refers to the situation in which the running feature types of the coupling behavior labels corresponding to different associated edges in the associated path are contradictory. That is, in the same associated path, the running feature types corresponding to the coupling behavior of adjacent associated edges cannot achieve continuous state transmission (e.g., the running feature type of one associated edge is running temperature, the running feature type of the adjacent associated edge is insulation resistance, and there is no reasonable conduction association).

[0098] A temporal sequence contradiction refers to a logical contradiction in the timing of coupling behavior labels corresponding to different associated edges in an associated path. That is, the timing of coupling of each associated edge in the path cannot form a continuous transmission chain (e.g., the timing of associated edge 1 in the path is "device A-device B", the timing of associated edge 2 is "device C-device B", and the coupling change between device C and device B is earlier than the coupling change between device A and device B, so a continuous transmission logic cannot be formed).

[0099] Optionally, all associated paths consisting of multiple associated edges in the initial local implicit association topology are extracted, and the associated edge identifier, node identifier, and coupling behavior label of each associated edge are recorded one by one. Further, a semantic consistency verification rule is preset, which clearly states that for an associated path to pass the verification, it must meet two conditions: first, the operational characteristic types of all associated edges in the path can form a continuous state transmission logic without contradictions; second, the coupling time sequence of all associated edges in the path can form a continuous transmission chain without logical contradictions.

[0100] Furthermore, according to the preset verification rules, each associated path is verified one by one. First, the operational characteristic type of each associated edge in the path is checked to determine if there is a conflict. Then, the coupling time sequence of each associated edge is checked to determine if there is a contradiction. For associated paths with conflicting operational characteristic types or contradictory time sequences, they are determined to be invalid associated paths and are all eliminated. For associated paths that meet both verification conditions and have no conflicts or contradictions, they are determined to be valid associated paths and are retained. Finally, based on the retained valid associated paths, their corresponding associated edges and nodes, the topology is reconstructed, and this structure is the target local implicit associated topology.

[0101] Step 20443: Based on the target local implicit association topology, determine the set of effective state influence channels between each substation device and its adjacent devices in the electrical topology.

[0102] Optionally, in the electrical topology, adjacent devices refer to devices in the target local implicit association topology that are directly connected to a certain substation device node through a single association edge. That is, the two devices have no other intermediate nodes in the topology, are directly adjacent, and have a valid implicit association relationship. An effective state influence channel refers to a channel in the target local implicit association topology between a certain substation device and its adjacent devices in the electrical topology, capable of stable transmission of operational state influence and having passed semantic consistency verification. Each effective state influence channel corresponds to an association edge in the target local implicit association topology, and the channel's attributes are consistent with those of the association edge. The effective state influence channel set refers to the set of effective state influence channels between a certain substation device and all its adjacent devices in the electrical topology. Each substation device corresponds to an independent effective state influence channel set, which contains all effective state influence channels between the device and all its adjacent devices, fully reflecting the implicit association transmission capability between the device and its adjacent devices.

[0103] Optionally, all nodes (substation equipment) and associated edges (effective associated edges) in the target local implicit association topology are extracted, and the adjacent nodes (nodes directly connected by a single associated edge) and their corresponding associated edges for each node are recorded one by one. Further, taking each substation equipment node as the core, the association relationship between this node and all adjacent nodes is analyzed one by one. The associated edge corresponding to each adjacent node is determined as the effective state influence channel between the core equipment and the adjacent equipment. The channel's attributes inherit the attributes of the corresponding associated edge, including coupling behavior label, operating characteristic type, coupling time sequence, electrical path type, coupling strength, etc.

[0104] Furthermore, all effective status impact channels corresponding to each core device are summarized, classified and organized according to the category of adjacent devices, and the adjacent device identifiers and channel attributes corresponding to each channel are labeled to form the set of effective status impact channels for that core device.

[0105] Furthermore, the above operations are performed on all substation equipment nodes one by one to generate a corresponding set of effective state influence channels for each substation equipment, ensuring that there are no omissions or duplicate channels in the set, and fully reflecting the implicit association transmission between each equipment and its adjacent equipment.

[0106] Step 20444: Based on the set of effective state influence channels, summarize the multi-level state propagation mode in which each substation equipment participates through non-explicit control signal paths during operation.

[0107] Optionally, the non-explicit control signal path refers to the path that transmits the influence of the operating status without sending control commands, which is the opposite of the explicit control command path. It is the same concept as the non-control command path in step 2043, that is, the implicit transmission path that is naturally formed between devices.

[0108] Multi-level state propagation mode refers to the specific mode in which each substation device participates in the propagation of continuous operating state influences through non-explicit control signal paths at multiple levels and involving multiple devices during operation. The mode includes different propagation scenarios where the device acts as the propagation start point, intermediate node, or end point. It integrates information such as the devices involved in the propagation, effective state influence channels, propagation direction, and propagation intensity, and can fully reflect the device's state propagation role and capabilities in the implicit association network.

[0109] Optionally, all effective state-affected channels within each set of effective state-affected channels are analyzed one by one, extracting information such as the adjacent device identifier, propagation direction (coupling time order), coupling strength, and operating characteristic type corresponding to each channel. Further, taking each substation device as the core, and combining the associated paths in the target local implicit association topology, all state propagation scenarios in which the device participates are identified. These scenarios include three types: first, the device acts as the propagation starting point, transmitting its own operating state changes to adjacent devices through effective state-affected channels; second, the device acts as a propagation intermediate node, receiving state propagation from one adjacent device and then transmitting it to another adjacent device through other effective state-affected channels; third, the device acts as the propagation endpoint, receiving state changes transmitted by adjacent devices through effective state-affected channels.

[0110] Furthermore, all state propagation scenarios for each core device are classified and summarized, and scenarios with the same propagation characteristics (same type of operating characteristics, similar propagation direction, and similar coupling strength) are integrated. The state propagation rules under each scenario are extracted, and the multi-level devices (starting point, intermediate node, and ending point), effective state influence channel sequence, and propagation level (first-order is direct propagation to adjacent devices, second-order is propagation to indirect adjacent devices through an intermediate node, and so on) involved in the propagation are clarified.

[0111] Furthermore, the summarized multi-level state propagation rules of the device are integrated to form the multi-level state propagation mode of the device. Each substation device corresponds to a set of multi-level state propagation modes, and the state propagation mode can fully characterize the state propagation role of the device in the implicit association network.

[0112] Step 20445: Based on the operational characteristic types, state propagation directions, and electrical path levels involved in each propagation link in the multi-level state propagation mode, determine the implicit association characteristics of each substation device.

[0113] Optionally, a propagation link refers to the process in a multi-stage state propagation mode where the influence of each operating state is transmitted from one device to another, with each propagation link corresponding to an effective state influence channel. The state propagation direction refers to the order in which the operating state influence is transmitted within the propagation link, i.e., from which device to which device, consistent with the chronological order of coupling. The electrical path hierarchy refers to the hierarchical division of the electrical paths corresponding to the effective state influence channels within the propagation link, determined according to the hierarchical relationship of the electrical topology, and divided into Level 1 (the hierarchy corresponding to direct physical connection paths) and Level 2 and above (the hierarchy corresponding to indirect electrical connection paths; the higher the level number, the more indirectly connected devices).

[0114] Optionally, each propagation link in the multi-level state propagation mode of each substation equipment can be analyzed one by one, and the operating characteristic type, state propagation direction and electrical path level involved in each propagation link can be extracted. At the same time, the role of the equipment in each propagation link (start point, intermediate node, end point) can be recorded.

[0115] Optionally, a pre-defined rule for determining implicit association characteristics is included, comprising four core elements: first, the type of operational characteristics that the device can transmit through implicit association, obtained by integrating the operational characteristic types of all propagation links; second, the scope of implicit association of the device, i.e., the number and distribution of all substation devices that can be affected through multi-level state propagation; third, the strength of implicit association of the device, judged comprehensively based on the average coupling strength and propagation levels in the propagation links (the higher the coupling strength and the fewer the propagation levels, the stronger the association); and fourth, the state propagation capability of the device, judged comprehensively based on the number of propagation levels and the propagation range in the multi-level state propagation mode (the more propagation levels and the wider the range, the stronger the propagation capability). Therefore, according to the pre-defined rule, the propagation link information of each substation device is comprehensively analyzed to obtain the implicit association characteristics corresponding to each substation device.

[0116] The embodiments of the present invention accurately uncover the implicit correlation characteristics of substation equipment in the state where protection actions or alarm signals are not triggered, effectively avoiding the problem of missed detection of latent fault precursors due to inaccurate identification of implicit correlations, and improving the timeliness and accuracy of equipment safety management and control.

[0117] Optionally, steps 205 to 208 include: Step 205: Based on the implicit association characteristics, determine the potential abnormal transmission path in which each substation equipment participates in the operating state without triggering an alarm signal, and based on the potential abnormal transmission path, identify the target transmission link that will transmit the abnormal operating characteristics of the starting equipment to other substation equipment along the electrical path.

[0118] Optionally, the operating state without triggering alarm signals refers to the apparent normal operating state of substation equipment, where no fault alarm signals are issued and no protection actions are triggered, consistent with the operating state of equipment corresponding to implicit correlation characteristics. A potential anomaly propagation path refers to a path that each substation device, in its operating state without triggering alarm signals, may participate in based on its own implicit correlation characteristics, potentially propagating an abnormal operating characteristic from one device to other related devices. This path is formed based on the implicit correlation between devices and has not yet resulted in actual anomaly propagation; it is a potential anomaly propagation path. The starting device refers to the substation equipment in the potential anomaly propagation path that first experiences an abnormal operating characteristic, serving as the source of the anomaly propagation.

[0119] Operating characteristic anomalies refer to latent anomalies in which one or more operating characteristic values ​​of substation equipment deviate from their historical steady-state range without triggering alarm signals; these are not considered overt faults. Electrical paths refer to the pathways between substation equipment that enable the transmission of operational status influences, including direct physical connections and indirect electrical connections. Target transmission links refer to complete links selected from potential anomaly transmission paths that clearly transmit the operating characteristic anomaly of the starting device along the electrical path to other substation equipment. These links have clearly defined starting devices, transmission paths, and downstream receiving devices, and enable continuous transmission of the anomaly.

[0120] Optionally, the implicit correlation characteristics of each substation device are analyzed one by one, extracting information such as the implicit correlation range, correlation strength, state propagation capability, and effective state influence channels contained in the characteristics, to identify the associated devices and corresponding electrical paths through which each device can transmit anomalies via implicit correlations. Further, based on the implicit correlation information of each device and combined with preset anomaly propagation path determination rules, the potential anomaly propagation paths that the device may participate in under the operating state without triggering an alarm signal are determined. The determination rules specify that a potential anomaly propagation path must be centered on the device, include at least one associated device, possess an electrical path capable of transmitting anomalies, and the path must conform to the implicit correlation propagation law between devices. Each device can correspond to multiple potential anomaly propagation paths.

[0121] Furthermore, each potential anomaly propagation path is analyzed to identify the starting device (assuming the device that first experiences an anomaly), downstream related devices, the electrical path upon which the propagation relies, and implicit connections. Further, from all potential anomaly propagation paths, paths capable of continuous anomaly transmission are selected; that is, paths where the operational anomaly of the starting device can be gradually propagated to other downstream substation devices along the electrical path and through implicit connections between devices. These paths are identified as target propagation links.

[0122] Step 206: Based on the operational characteristic change patterns and electrical connection relationships of each link in each target transmission link, determine the abnormal transmission result of each target transmission link.

[0123] Optionally, a link in the target transmission link refers to the abnormal transmission process between two adjacent substation devices in the target transmission link. Each link corresponds to an effective state influence channel, connecting two adjacent devices, and is the basic unit of abnormal transmission.

[0124] Operational characteristic change patterns refer to the possible changes in the operational characteristics of substation equipment corresponding to each link in the target transmission link, including the direction, magnitude, and duration of change. These patterns are derived from the operational characteristic types and state propagation laws contained in implicit correlation characteristics and are matched with the implicit coupling relationships between equipment. Electrical connection relationships refer to the specific connection methods between two adjacent substation devices corresponding to each link in the target transmission link, including direct physical connections and indirect electrical connections. These correspond to the electrical path type and determine the speed and intensity of abnormal transmission.

[0125] Anomaly propagation results refer to the judgment results used to characterize whether the anomaly, through implicit coupling, causes the operating characteristics of downstream substation equipment to deviate from the historical steady-state range in subsequent sampling periods when the continuous sampling period of the operating characteristics of the starting device in the propagation link is outside the historical steady-state range. This is categorized as "anomaly propagable" and "anomaly non-propagable." The historical steady-state range refers to the stable numerical range of each operating characteristic of the substation equipment under normal operating conditions, statistically derived from the sensing dataset obtained in step 10 and the equipment's historical operating data. This range has been verified through long-term operation and can accurately characterize the parameter range of normal equipment operation. The continuous sampling period refers to the preset number of continuous sensing periods used to determine whether the operating characteristic is in an abnormal state, reasonably set according to the equipment's operating characteristics and sensing frequency (e.g., 3 sampling periods). Downstream substation equipment refers to the substation equipment located after the starting device in the target propagation link that receives the abnormal signal propagated by the starting device. According to the propagation order, they can be divided into first-level downstream equipment, second-level downstream equipment, etc.

[0126] Optionally, each target transmission link is analyzed one by one, and each transmission link in the link is broken down. The identifiers of the two adjacent devices corresponding to each link, the effective state influence channel, the operating characteristic change mode, and the electrical connection relationship are recorded. Further, a preset historical steady-state interval database is retrieved. The database stores the historical steady-state intervals corresponding to each operating characteristic of each substation device. At the same time, the perception dataset obtained in step 10 is called. Based on the explicit logic reasoning model, combined with the operating characteristic change mode and electrical connection relationship of each link of each target transmission link, anomaly transmission simulation analysis is performed. It is assumed that the operating characteristic of the starting device is outside its historical steady-state interval in the continuous sampling period (i.e., a latent anomaly occurs). It is inferred whether the anomaly can be transmitted to the downstream device through the latent coupling relationship of each link along the electrical connection relationship, and whether the operating characteristic of the downstream device deviates from its own historical steady-state interval in subsequent sampling periods.

[0127] Furthermore, based on the reasoning results, the abnormal transmission result of each target transmission link is determined: if the abnormality can cause the operating characteristics of downstream equipment to deviate from the historical steady-state range through implicit coupling, the abnormal transmission result is "abnormality can be transmitted"; if the abnormality cannot be transmitted through implicit coupling, or does not cause the operating characteristics of downstream equipment to deviate from the historical steady-state range after transmission, the abnormal transmission result is "abnormality cannot be transmitted".

[0128] Step 207: The target transmission link with the abnormal transmission result as the effective result is the risk implicit transmission path, and the substation equipment corresponding to the transmission starting point in the risk implicit transmission path is determined as the risk source candidate equipment set.

[0129] Optionally, a valid result refers to an abnormal transmission result that reflects the precursors of latent faults and may lead to equipment failure risks, i.e., the "abnormally transmissible" result determined in step 206. The target transmission link corresponding to such a result has the actual fault risk transmission capability. A risk latent transmission path refers to a target transmission link that takes the abnormal transmission result as a valid result (abnormally transmissible). This path can realize the continuous transmission of latent anomalies and is the main hidden danger source leading to missed equipment failures.

[0130] The transmission origin refers to the location in the implicit risk transmission path where the operational characteristics first become abnormal, serving as the source of the abnormality transmission; it corresponds to the starting device in the path.

[0131] The risk source candidate equipment set refers to the set of substation equipment corresponding to the starting point of all implicit risk transmission paths. The equipment in the set are all potential risk source equipment that may cause implicit abnormal transmission and lead to risks to downstream equipment. It is the core analysis object for subsequent determination of safety decision results.

[0132] Optionally, the abnormal transmission results of each target transmission link are compared one by one, and target transmission links with abnormal transmission results that are "abnormally transmissible" are selected. These target transmission links are identified as hidden risk transmission paths, and all hidden risk transmission paths are summarized to form a set of hidden risk transmission paths. Further, each path in the set of hidden risk transmission paths is analyzed one by one, and the substation equipment corresponding to the transmission starting point of each path (i.e., the starting equipment in the path) is extracted. The identification of the equipment, the corresponding hidden risk transmission path, and the hidden association characteristics are recorded.

[0133] Furthermore, the substation equipment corresponding to all extracted transmission starting points is deduplicated, removing duplicate equipment identifiers to prevent the same equipment from entering the set multiple times. Then, all deduplicated substation equipment is integrated to form a risk source candidate equipment set, clarifying the implicit risk transmission path corresponding to each equipment in the set.

[0134] Step 208: Based on the relationship between the current operating characteristic values ​​of each substation device in the risk source candidate device set and its historical steady-state interval, determine the safety decision results of each substation device.

[0135] Optionally, based on the relationship between the current operating characteristic values ​​of each substation device in the candidate risk source device set and its historical steady-state interval, the safety decision results of each substation device are determined, as in steps 2081 to 2084.

[0136] Based on explicit logical reasoning models and implicit correlation features, the embodiments of this invention fully cover the identification of implicit anomaly transmission paths, the judgment of anomaly transmission capabilities, the location of risk sources, and the derivation of safety decisions. It accurately captures the implicit fault precursors that have not triggered alarm signals. Therefore, by accurately identifying the implicit risk transmission paths and risk sources, it ensures the accuracy of safety decision results and improves the timeliness and accuracy of equipment safety management.

[0137] Optionally, the process of steps 2081 to 2084 includes: Step 2081: Based on the relationship between the current operating characteristic values ​​of each substation device in the risk source candidate device set and its historical steady-state interval, determine whether there is a judgment result that the continuous sampling period of the operating characteristics of each candidate device is outside the historical steady-state interval.

[0138] Optionally, candidate equipment refers to each piece of substation equipment in the set of candidate risk source equipment, which is a potential risk source equipment.

[0139] The continuous sampling period of the operating characteristic is outside the historical steady-state range. This means that one or more operating characteristic values ​​of the candidate device deviate from their corresponding historical steady-state range in each sampling period within the preset continuous sampling period, and no alarm signal or protection action is triggered. This is a hidden abnormal state.

[0140] The judgment result refers to the conclusion used to characterize whether each candidate device has the above-mentioned hidden abnormal state, including two cases: "there is a continuous sampling period of the operating characteristics outside the historical steady state range" and "there is no continuous sampling period of the operating characteristics outside the historical steady state range".

[0141] Optionally, each candidate device in the risk source candidate device set is processed one by one. For each operating characteristic of the candidate device, the value of the operating characteristic in each sampling period within the most recent continuous sampling period is compared with the corresponding historical steady-state interval to determine whether the value in that sampling period deviates from the historical steady-state interval. Further, the deviation of all operating characteristics of the candidate device within the continuous sampling period is statistically analyzed. If at least one operating characteristic deviates from the historical steady-state interval in each sampling period within the preset continuous sampling period, the candidate device is determined to have "operating characteristics whose continuous sampling period is outside the historical steady-state interval." If none of the above situations occur for any operating characteristic, that is, no operating characteristic continuously deviates from the historical steady-state interval within the preset continuous sampling period, the candidate device is determined to have "no operating characteristics whose continuous sampling period is outside the historical steady-state interval."

[0142] Step 2082: Based on the judgment results, candidate equipment that does not have continuous sampling periods of operating characteristics outside the historical steady-state range is eliminated, and substation equipment that has continuous sampling periods of operating characteristics outside the historical steady-state range is retained to obtain effective risk source equipment.

[0143] Optionally, effective risk source equipment refers to substation equipment that has been screened and retained, currently exhibits persistent latent anomalies, and can trigger the transmission of anomalies in the latent risk transmission path. Each effective risk source equipment corresponds to at least one latent risk transmission path.

[0144] Optionally, the judgment result is associated with the corresponding candidate device identifier, and the candidate devices are grouped according to the judgment result type into a "possibly persistent latent anomaly" group and a "not possessing persistent latent anomaly" group. Further, candidate devices in the "not possessing persistent latent anomaly" group are removed from the risk source candidate device set. Further, the candidate devices in the "possibly persistent latent anomaly" group are verified to confirm that the judgment result for each device is correct. The verification includes the operating characteristic values ​​of the continuous sampling period, historical steady-state range, deviations, etc. Devices with judgment errors discovered during the verification process are removed, resulting in a valid risk source device set.

[0145] Step 2083: Based on the effective risk source equipment and its corresponding implicit risk transmission path, determine the set of downstream substation equipment affected by each effective risk source equipment, and based on the set of downstream substation equipment and the substation primary system topology, determine whether the affected equipment is located on a critical power supply path or a preset load node, and obtain the judgment result.

[0146] Optionally, the downstream substation equipment set refers to the set of all downstream substation equipment that can be affected by the implicit anomaly transmission of each effective risk source equipment through each corresponding implicit risk transmission path. Each effective risk source equipment corresponds to an independent downstream substation equipment set, which includes all downstream equipment in all implicit risk transmission paths of the effective risk source equipment, and marks the implicit risk transmission path, transmission level, and other information corresponding to each downstream equipment.

[0147] The primary system topology of a substation refers to the electrical connection relationship and layout structure of all primary equipment (including transformers, circuit breakers, disconnectors, busbars, power lines, etc.) within the substation. It is the basis for determining critical power supply paths and preset load nodes. A preset substation primary system topology database stores complete topology information.

[0148] The critical power supply path refers to the continuous electrical path from the main power supply input side to the feeder output of the preset load node, and its interruption will cause the preset load node to lose power. This path is the core path of substation power supply and is directly related to the normal power supply of the preset load node. The main power supply input side refers to the input port of the substation that receives power from the external power grid and related equipment nearby, which is the starting point of the substation's power input.

[0149] Preset load nodes refer to pre-defined load nodes with significant power supply needs, including important user power supply nodes and power supply nodes for critical equipment inside substations. Power loss of such nodes will have a significant impact. The specific location and power supply needs of the preset load nodes need to be stored in advance.

[0150] Feeder outlet refers to the outlet location of the feeder that leads from the substation to supply power to the preset load node, and related nearby equipment. Affected equipment refers to each piece of substation equipment in the downstream substation equipment set, that is, equipment that may be affected by latent anomalies of effective risk source equipment.

[0151] The judgment result refers to the conclusion used to characterize whether each affected device is located on a critical power supply path or a preset load node, namely "located on a critical power supply path or a preset load node" and "not located on a critical power supply path or a preset load node". The judgment result corresponds one-to-one with the affected device and is marked with the corresponding critical power supply path identifier or preset load node identifier (if the conditions are met).

[0152] Optionally, the identifier of each effective risk source device and all its corresponding implicit risk transmission paths are extracted. Each effective risk source device is analyzed one by one. Based on each corresponding implicit risk transmission path, all downstream substation devices in the path (i.e., all substation devices in the path other than the effective risk source device) are extracted. These downstream devices are summarized to form the set of downstream substation devices corresponding to the effective risk source device, and the implicit risk transmission path corresponding to each downstream device is marked.

[0153] Furthermore, the system retrieves the preset substation primary system topology database, critical power supply path labeling information, and preset load node information, and analyzes the affected equipment in each downstream substation equipment set one by one to determine whether the affected equipment is located on any critical power supply path or is a preset load node. If the affected equipment is located on a critical power supply path or is itself a preset load node, it is determined that the affected equipment is "located on a critical power supply path or a preset load node". If the affected equipment is neither located on any critical power supply path nor a preset load node, it is determined that the affected equipment is "not located on a critical power supply path or a preset load node".

[0154] Step 2084: Based on the judgment results, determine the risk consequence level of each hidden risk transmission path, and based on the risk consequence level of the hidden risk transmission path and the existence status of the corresponding effective risk source equipment, generate the safety decision results of each substation equipment.

[0155] Optionally, the risk consequence level refers to the degree of power supply impact and risk loss that may be caused after a latent anomaly occurs in each latent risk transmission path. Based on the judgment results of the affected equipment, it is divided into three levels: high, medium, and low, with the level classification standard preset at medium. The existence status of effective risk source equipment refers to the current operating status of effective risk source equipment, that is, the existence of a latent anomaly state where the continuous sampling period of the operating characteristics is outside the historical steady-state range, and no alarm signal or protection action has been triggered.

[0156] Optionally, each implicit risk transmission path is analyzed one by one. Based on the preset risk consequence level classification standard and combined with the judgment results of the affected equipment corresponding to each implicit risk transmission path, the risk consequence level of the implicit risk transmission path is determined: if there are devices located on critical power supply paths or preset load nodes among the affected devices corresponding to the path, and the implicit abnormal transmission may cause the preset load nodes to lose power or the critical power supply path to be interrupted, then the risk consequence level is judged as high; if none of the affected devices corresponding to the path are located on critical power supply paths or preset load nodes, but the implicit abnormal transmission may cause multiple downstream devices to have implicit abnormalities, then the risk consequence level is judged as medium; if the number of affected devices corresponding to the path is small, and the implicit abnormal transmission will only cause a few downstream devices to have slight implicit abnormalities without obvious risk loss, then the risk consequence level is judged as low. Furthermore, based on the risk consequence level of each hidden risk transmission path and the existence status of the corresponding effective risk source device, a corresponding security decision result is generated for each effective risk source device: for high risk consequence level, a "execute local isolation operation" instruction is generated; for medium risk consequence level, a "start enhanced monitoring mode" instruction is generated; and for low risk consequence level, a "maintain current operating status" instruction is generated.

[0157] The embodiments of this invention cover the identification of latent anomalies, the identification of effective risk sources, the analysis of the scope of risk impact, risk classification, and decision generation. It can accurately capture latent fault risks that have not triggered alarm signals. By accurately locating effective risk sources and classifying and determining the consequences of risks, it ensures the pertinence and accuracy of safety decision results, and improves the timeliness and accuracy of equipment safety management.

[0158] Optionally, the processes of steps 301 to 304 include: Step 301: Determine the safety response role of each substation device based on the handling instruction type in the safety decision results of each substation device.

[0159] Optionally, the disposal instruction type refers to the specific disposal requirement category included in the safety decision result, which is completely consistent with the three types of safety decision results, namely "execute local isolation operation", "activate enhanced monitoring mode" and "maintain current operating status". Each disposal instruction type corresponds to a unique safety response role. The safety response role refers to the specific response responsibilities assigned to each substation equipment based on the disposal instruction type, which must be undertaken in the safety management process. It corresponds one-to-one with the disposal instruction type and clarifies the specific safety response actions that the equipment must perform. Specifically, it includes three types: execute local isolation operation, activate enhanced monitoring mode and maintain current operating status. The responsibilities of each role are completely matched with the requirements of the corresponding disposal instruction.

[0160] Optionally, the safety decision results for each substation device are extracted one by one, the corresponding handling instruction type for each device is identified, and the identifier of each device is recorded to ensure a one-to-one correspondence between the handling instruction type and the device identifier. Further, a pre-defined correspondence rule between handling instruction types and safety response roles is established, specifically: a handling instruction type of "execute local isolation operation" corresponds to the safety response role of "execute local isolation operation"; a handling instruction type of "activate enhanced monitoring mode" corresponds to the safety response role of "activate enhanced monitoring mode"; and a handling instruction type of "maintain current operating state" corresponds to the safety response role of "maintain current operating state". Further, according to the pre-defined correspondence rule, a corresponding safety response role is matched for each substation device one by one.

[0161] Step 302: Based on the direct electrical connection relationship between each substation device in the substation primary system topology, determine the set of adjacent substation devices for each substation device.

[0162] Optionally, the set of adjacent substation equipment refers to the set of all adjacent substation equipment that have a direct electrical connection with each substation equipment. Each substation equipment corresponds to an independent set of adjacent substation equipment, which contains the identifiers of all directly adjacent substation equipment of that equipment. If a certain substation equipment has no directly electrically connected adjacent equipment, its set of adjacent substation equipment is an empty set. The set must be labeled with the corresponding core equipment identifier to ensure that it can be mapped to a specific substation equipment.

[0163] Optionally, the identifiers of all substation equipment and all direct electrical connection information are extracted from the substation primary system topology database, and each pair of substation equipment identifiers with a direct electrical connection is sorted out and recorded. Further, each substation equipment is selected as a core device, and using the identifier of the core device as the search condition, all substation equipment identifiers with a direct electrical connection to that core device are retrieved from the sorted direct electrical connection information.

[0164] Furthermore, all the retrieved adjacent substation equipment identifiers are aggregated, deduplicated, and duplicate equipment identifiers are removed to prevent the same adjacent equipment from entering the set multiple times. If no adjacent equipment is retrieved, an empty set is generated.

[0165] Step 303: For any target substation equipment, based on its safety response role and the safety response roles of each adjacent substation equipment in the set of adjacent substation equipment, determine the neighborhood decision distribution within the local electrical area where the target substation equipment is located.

[0166] Optionally, a local electrical area refers to an electrical area centered on the target substation equipment and bounded by all adjacent equipment in its adjacent substation equipment set. This area comprises the target substation equipment and all its directly adjacent equipment. Equipment within this area is interconnected through direct electrical connections, and their operating states influence each other. Neighborhood decision distribution refers to the distribution of equipment within the local electrical area where the target substation equipment is located, specifically whether any adjacent substation equipment is assigned the role of performing local isolation operations. This focuses solely on the safety response role of performing local isolation operations, ignoring the distribution of the other two roles. The distribution must clearly indicate the target substation equipment identifier, adjacent equipment identifier, and corresponding safety response role, reflecting the presence and number (if any) of adjacent equipment performing the local isolation operation role.

[0167] Optionally, taking the target substation equipment as the core, construct the local electrical area where it is located, clarify the target equipment and all adjacent equipment contained in the area, focus on verifying the safety response role of each adjacent equipment in the set of adjacent equipment, determine whether each adjacent equipment is assigned the role of performing local isolation operation, and count the number of adjacent equipment assigned the role (if any).

[0168] Furthermore, the verification and statistical results are integrated to form the neighborhood decision distribution corresponding to the target substation equipment. The distribution content needs to be clearly defined as follows: the target substation equipment identifier, the identifiers of adjacent equipment in the local electrical area and their corresponding safety response roles, whether there are any devices in the adjacent equipment that perform local isolation operations, and the number of devices with this role (if any). The above operations are performed on all target substation equipment one by one to obtain the neighborhood decision distribution corresponding to each target substation equipment.

[0169] Step 304: Based on the existence of adjacent substation equipment in the neighborhood decision distribution that are assigned the role of performing local isolation operations, perform operation continuity analysis to obtain the collaborative decision results of each substation equipment.

[0170] Optionally, based on the existence of adjacent substation equipment in the neighborhood decision distribution that are assigned the role of performing local isolation operations, an operational continuity analysis is performed to obtain the collaborative decision results of each substation equipment, as in steps 3041 to 3043.

[0171] The embodiments of the present invention can accurately capture latent fault risks that have not triggered alarm signals, effectively solving the problem of missed reporting caused by the lack of perception of latent fault precursors in fixed rule base methods. By accurately locating effective risk sources and classifying and judging risk consequences, the invention ensures the pertinence and accuracy of safety decision results, and improves the timeliness and accuracy of equipment safety management.

[0172] Optionally, the processes of steps 3041 to 3043 include: Step 3041: Based on the existence of adjacent substation equipment assigned the role of performing local isolation operation and the safety response role of the target substation equipment in the neighborhood decision distribution, determine the operational continuity result of the target substation equipment.

[0173] Optionally, the existence of adjacent substation equipment assigned the role of performing local isolation operation means that, in the set of adjacent substation equipment of the target substation equipment explicitly marked in the neighborhood decision distribution, at least one adjacent substation equipment is assigned the safety response role of performing local isolation operation. The operational continuity result refers to the conclusion used to characterize the current safety response role of the target substation equipment, and whether its operational status will be affected or whether the local operational continuity of the power grid will be disrupted in the scenario where adjacent equipment with the role of performing local isolation operation exists. It includes "operational continuity is affected" and "operational continuity is not affected." This result only applies to scenarios where adjacent equipment with the role of performing local isolation operation exists. If there are no such adjacent equipment in the neighborhood decision distribution, this result does not need to be determined, and the operational continuity of the target substation equipment is assumed to be unaffected.

[0174] Optionally, each substation device is selected as the target substation device, and the neighborhood decision distribution corresponding to the target substation device is checked to determine whether there are any adjacent substation devices in the set of adjacent substation devices that have been given the role of performing local isolation operations.

[0175] If no such adjacent substation equipment exists in the neighborhood decision distribution, no further analysis is required, and the operational continuity result of the target substation equipment is assumed to be unaffected, and this result is directly recorded. If such adjacent substation equipment exists in the neighborhood decision distribution, the safety response role of the target substation equipment itself is further extracted, and its operational continuity result is determined by combining the preset operational continuity result judgment rules. The preset judgment rules are clear and unique, specifically: if the safety response role of the target substation equipment is to maintain the current operating state, and there is an adjacent substation equipment performing a local isolation operation role, then the operational continuity result is determined to be affected; if the safety response role of the target substation equipment is to perform a local isolation operation or activate the enhanced monitoring mode, and there is an adjacent substation equipment performing a local isolation operation role, then the operational continuity result is determined to be unaffected.

[0176] Step 3042: Based on the operational continuity results and whether the target substation equipment is located on a critical power supply path, determine the collaborative upgrade judgment result of whether the target substation equipment needs to adjust its security response role.

[0177] Optionally, whether the target substation equipment is located on a critical power supply path refers to determining whether the identifier of the target substation equipment is included in the equipment list of any critical power supply path by checking the preset critical power supply path labeling information. The determination conclusion includes "located on a critical power supply path" and "not located on a critical power supply path". The collaborative upgrade determination result refers to the determination conclusion used to characterize whether the target substation equipment needs to adjust its current security response role, including "needs to adjust security response role" and "does not need to adjust security response role". This result directly determines whether the security response role of the target equipment should be optimized in the future, thus affecting the generation of collaborative decision results.

[0178] Optionally, the identification and operational continuity results of each target substation equipment can be extracted one by one, and the preset critical power supply path marking information can be retrieved to check whether each target substation equipment is located on the critical power supply path and record the verification results.

[0179] Furthermore, the system pre-defines collaborative upgrade judgment rules. These rules are based on a comprehensive assessment of operational continuity results and whether the target device is located on a critical power supply path, ensuring no omissions or contradictions. The specific rules are divided into four scenarios, covering all possible situations: Scenario 1: If the operational continuity result of the target substation equipment indicates that operational continuity is affected, and the target equipment is located on a critical power supply path, then the collaborative upgrade judgment result is that the safety response role needs adjustment. Scenario 2: If the operational continuity result of the target substation equipment indicates that operational continuity is affected, and the target equipment is not located on a critical power supply path, then the collaborative upgrade judgment result is that the safety response role does not need adjustment. Scenario 3: If the operational continuity result of the target substation equipment indicates that operational continuity is not affected, regardless of whether the target equipment is located on a critical power supply path, then the collaborative upgrade judgment result is that the safety response role does not need adjustment. Scenario 4: If there are no adjacent devices in the target substation equipment's neighborhood decision distribution that perform the local isolation operation role (i.e., the default operational continuity is not affected in step 3041), regardless of whether the target equipment is located on a critical power supply path, then the collaborative upgrade judgment result is that the safety response role does not need adjustment. Furthermore, in accordance with the preset collaborative upgrade judgment rules, the collaborative upgrade judgment result is determined by combining the operational continuity results of each target substation equipment and the verification conclusion of whether it is located in the critical power supply path.

[0180] Step 3043: Based on the security response role determined by the collaborative upgrade judgment result of each target substation equipment, generate the collaborative decision result of each substation equipment.

[0181] Optionally, the security response role determined based on the collaborative upgrade judgment result means that if the collaborative upgrade judgment result is that no adjustment is needed, the current security response role of the target device will remain unchanged; if the collaborative upgrade judgment result is that adjustment is needed, the current security response role of the target device will be adjusted to the specified security response role according to the preset role adjustment rules, and the adjusted role will still belong to one of the three security response roles.

[0182] Optionally, a pre-defined safety response role adjustment rule is provided. The rule only applies to target equipment whose collaborative upgrade determination result indicates that adjustment is required. Specifically, if the current safety response role of the target substation equipment is to maintain the current operating state, and the collaborative upgrade determination result indicates that adjustment is required, then its safety response role is adjusted to start enhanced monitoring mode. After adjustment, it no longer maintains the current operating state, but instead performs enhanced monitoring, balancing operational continuity and risk control. If the collaborative upgrade determination result indicates that adjustment is not required, then the current safety response role of the target equipment remains unchanged, and no adjustment is made.

[0183] Furthermore, each target substation device is processed one by one, and its final safety response role is determined based on its collaborative upgrade judgment result: for those whose collaborative upgrade judgment result does not require adjustment, the current role is maintained; for those that require adjustment, they are adjusted to the designated role according to the preset adjustment rules.

[0184] The final safety response role of each target substation device is transformed into a corresponding collaborative decision result (i.e., the type of handling instruction corresponding to the safety response role). The collaborative decision result is associated with the target device identifier, collaborative upgrade judgment result, and adjustment basis (if any). The above operation is performed on all substation devices one by one to complete the generation of collaborative decision results for each substation device.

[0185] Through precise scenario analysis and rule determination, the optimized collaborative decision-making results of this invention not only retain the ability to control hidden fault risks, but also take into account the continuity of power grid operation. This effectively makes up for the shortcomings of traditional fixed rule base decision-making, which lacks collaboration and is prone to causing local power grid operation anomalies, and improves the accuracy and timeliness of equipment safety management.

[0186] Furthermore, the equipment safety management device based on substation intelligent agents provided by the present invention will be described below. The equipment safety management device based on substation intelligent agents described below can be referred to in correspondence with the equipment safety management method based on substation intelligent agents described above.

[0187] Optionally, refer to Figure 2 , Figure 2 This is a schematic diagram of the equipment safety management device based on a substation intelligent agent provided by the present invention. The equipment safety management device based on a substation intelligent agent includes: The local sensing module 210 is used to locally sense the operating status of the equipment in each substation based on the equipment in each substation, and obtain the sensing dataset of each substation equipment. The model intelligent analysis module 220 is used to analyze the potential coupling relationship between operating features based on the implicit association perception model and the perception dataset, obtain the implicit association features of each substation equipment, and perform fault risk analysis based on the explicit logic reasoning model and the implicit association features to obtain the safety decision results of each substation equipment. The risk pattern recognition module 230 is used to perform decision collaboration based on the safety decision results between adjacent substation equipment, obtain the collaborative decision results of each substation equipment, and perform potential risk pattern recognition based on the collaborative decision results to obtain the safety status recognition results of each substation equipment. The safety response module 240 is used to generate safety intervention commands based on the safety situation identification results, and to execute local safety responses on substation equipment based on the safety intervention commands.

[0188] The embodiments of the present invention solve the problem of missed detection caused by the lack of perception of latent fault precursors in fixed rule base methods, thereby improving the timeliness and accuracy of equipment safety management and control.

[0189] Please see Figure 3 , Figure 3 An embodiment diagram of an electronic device provided in accordance with the present invention. For example... Figure 3 As shown, an embodiment of the present invention provides an electronic device 300, including a memory 310, a processor 320, and a computer program 311 stored in the memory 310 and executable on the processor 320. When the processor 320 executes the computer program 311, it implements the processes of steps 10 to 40.

[0190] Please see Figure 4 , Figure 4 An embodiment diagram of a computer-readable storage medium provided in accordance with an embodiment of the present invention is shown. Figure 4 As shown, this embodiment provides a computer-readable storage medium 400 on which a computer program 311 is stored. When the computer program 311 is executed by a processor, it implements the processes of steps 10 to 40.

[0191] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the equipment safety management method based on substation intelligent agents provided by the above methods, which includes steps 10 to 40.

[0192] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for equipment safety management based on a substation intelligent agent, characterized in that, include: Based on the local perception of the operating status of each substation equipment, a perception dataset of each substation equipment is obtained. Based on the implicit association perception model and the perception dataset, the potential coupling relationship between the operating characteristics is analyzed to obtain the implicit association characteristics of each substation equipment. Then, based on the explicit logic reasoning model and the implicit association characteristics, the fault risk analysis is carried out to obtain the safety decision results of each substation equipment. Decision collaboration is performed based on the safety decision results between adjacent substation equipment to obtain the collaborative decision results of each substation equipment. Based on the collaborative decision results, potential risk pattern identification is performed to obtain the safety status identification results of each substation equipment. Based on the security situation identification results, a security intervention command is generated, and a local security response is executed on the substation equipment based on the security intervention command.

2. The equipment safety management method based on substation intelligent agents according to claim 1, characterized in that, The steps involved in determining the collaborative decision-making results for each substation device include: Based on the type of handling instruction in the safety decision results of each substation device, the safety response role of each substation device is determined; the safety response role includes performing local isolation operations, activating enhanced monitoring mode, and maintaining the current operating status. Based on the direct electrical connection relationships between various substation equipment in the substation primary system topology, determine the set of adjacent substation equipment for each substation equipment. For any target substation equipment, based on its safety response role and the safety response roles of each adjacent substation equipment in the set of adjacent substation equipment, the neighborhood decision distribution within the local electrical area where the target substation equipment is located is determined; the neighborhood decision distribution characterizes whether there are any adjacent substation equipment that have been assigned the role of performing local isolation operations. Based on the existence of adjacent substation equipment in the neighborhood decision distribution that are assigned the role of performing local isolation operations, an operational continuity analysis is performed to obtain the collaborative decision results of each substation equipment; the collaborative decision results represent the handling instructions corresponding to the final safety response role of each substation equipment.

3. The equipment safety management method based on substation intelligent agents according to claim 2, characterized in that, The steps involved in obtaining the collaborative decision-making results for each substation device through continuity analysis include: Based on the existence of adjacent substation equipment assigned the role of performing local isolation operations and the safety response role of the target substation equipment in the neighborhood decision distribution, the operational continuity result of the target substation equipment is determined; wherein, if the safety response role of the target substation equipment is to maintain the current operating state and there are adjacent substation equipment performing isolation operations, the operational continuity result is determined to be that the operational continuity is affected. Based on the operational continuity results and whether the target substation equipment is located on a critical power supply path, a collaborative upgrade determination result is made to determine whether the target substation equipment needs to adjust its safety response role. Based on the safety response role determined by the collaborative upgrade judgment result of each target substation equipment, collaborative decision results for each substation equipment are generated.

4. The equipment safety management method based on substation intelligent agents according to claim 1, characterized in that, The steps involved in performing feature analysis based on the implicit association perception model include: Based on the multidimensional operation feature sequence contained in the perception dataset of each substation equipment, the set of operation status trajectories of each substation equipment in the time dimension is determined, and the feature evolution path of each substation equipment in different operation stages is extracted based on the set of operation status trajectories. Based on the direction and sequence of change of operating characteristics between adjacent sampling times in the evolution path, the temporal dependency relationship between different operating characteristics inside each substation equipment is determined, and based on the temporal dependency relationship, the characteristic linkage pattern that continuously repeats during the triggering of a fault-free alarm signal is identified. Based on the co-occurrence of the aforementioned feature linkage mode among multiple substation devices, a set of events in which cross-device operating characteristics change synchronously within the same time window is determined, and feature change combinations whose deviation from the start time of change does not exceed a preset deviation threshold are selected based on the event set. Based on the combination of the aforementioned feature changes, response instance analysis is performed to obtain the implicit correlation features of each substation device. The implicit correlation features characterize the state coupling capability of substation devices with other devices through the electrical network structure when the equipment is in operation without triggering protection actions or alarm signals.

5. The equipment safety management method based on a substation intelligent agent according to claim 4, characterized in that, The response instance analysis based on the aforementioned combination of feature changes yields the implicit correlation features of each substation device, including: Based on the substation equipment pairs corresponding to the combination of the aforementioned feature changes, instances of coordinated response of the operating characteristics between the equipment under physical connection or electrical topology proximity conditions are determined. Based on the aforementioned coordinated response instances, common-cause responses caused by the same external power grid disturbance event are excluded, while characteristic coupling instances occurring under conditions where there is a direct electrical path between devices are retained. Based on the aforementioned feature coupling instance, a set of associated events reflecting the impact of the transmission of operating status between substation equipment through non-control command paths is determined, and an initial local implicit association topology structure with substation equipment as nodes and the aforementioned feature coupling instance as edges is constructed based on the associated event set. Based on the initial local implicit association topology, association topology analysis is performed to obtain the implicit association characteristics of each substation device.

6. The equipment safety management method based on a substation intelligent agent according to claim 5, characterized in that, The association topology analysis based on the initial local implicit association topology structure yields the implicit association characteristics of each substation device, including: Based on the operational feature type, the temporal order of coupling, and the electrical path type corresponding to each edge in the initial local implicit association topology, the coupling behavior label of each associated edge is determined; Based on the coupling behavior label, the semantic consistency of the associated paths composed of multiple edges in the topology is checked, and associated paths with conflicting running feature types or contradictory time sequences are eliminated to obtain the target local implicit associated topology. Based on the target local implicit association topology, determine the set of effective state influence channels between each substation device and its adjacent devices in the electrical topology; Based on the set of effective state influence channels, a multi-stage state propagation mode in which each substation device participates through non-explicit control signal paths during operation is summarized. Based on the operational characteristic types, state propagation directions, and electrical path levels involved in each propagation link of the multi-level state propagation mode, the implicit association characteristics of each substation device are determined.

7. The equipment safety management method based on substation intelligent agents according to claim 1, characterized in that, The steps involved in performing fault risk analysis based on an explicit logical reasoning model include: Based on the implicit correlation features, the potential abnormal transmission paths in which each substation device participates in the operating state without triggering an alarm signal are determined, and the target transmission link that transmits the abnormal operating characteristics of the starting device to other substation devices along the electrical path is identified based on the potential abnormal transmission paths. Based on the change patterns of operating characteristics and electrical connection relationships of each link in each target transmission link, the abnormal transmission result of each target transmission link is determined; the abnormal transmission result characterizes whether the abnormality causes the operating characteristics of the downstream substation equipment to deviate from the historical steady-state range in subsequent sampling periods when the continuous sampling period of the operating characteristics of the starting equipment of the transmission link is outside the historical steady-state range through implicit coupling. The target transmission link with the abnormal transmission result as the effective result is the risk implicit transmission path, and the substation equipment corresponding to the transmission starting point in the risk implicit transmission path is determined as the risk source candidate equipment set; Based on the relationship between the current operating characteristic values ​​of each substation device in the candidate risk source device set and its historical steady-state interval, the safety decision results of each substation device are determined.

8. The equipment safety management method based on a substation intelligent agent according to claim 7, characterized in that, The process of generating safety decision results for each substation device based on the risk consequence level of the implicit risk transmission path and the existence status of the corresponding effective risk source equipment includes: Based on the relationship between the current operating characteristic values ​​of each substation device in the candidate device set of risk sources and their historical steady-state intervals, it is determined whether there is a judgment result that the continuous sampling period of the operating characteristics of each candidate device is outside the historical steady-state interval; Based on the judgment results, candidate devices that do not have continuous sampling periods of operating characteristics outside the historical steady-state range are eliminated, and substation devices that have continuous sampling periods of operating characteristics outside the historical steady-state range are retained to obtain effective risk source devices; Based on the effective risk source equipment and its corresponding implicit risk transmission path, the set of downstream substation equipment affected by each effective risk source equipment is determined. Based on the set of downstream substation equipment and the substation primary system topology, it is determined whether the affected equipment is located on a critical power supply path or a preset load node, and the determination result is obtained. The critical power supply path represents a continuous electrical path from the main power supply incoming line to the feeder outlet of the preset load node, and its interruption will cause the preset load node to lose power. Based on the judgment results, the risk consequence level of each hidden risk transmission path is determined, and based on the risk consequence level of the hidden risk transmission path and the existence status of the corresponding effective risk source equipment, the safety decision results of each substation equipment are generated.

9. A device for equipment safety management based on a substation intelligent agent, characterized in that, For implementing the equipment safety management method based on substation intelligent agents as described in any one of claims 1 to 8; The equipment safety management device based on substation intelligent agents includes: The local sensing module is used to locally sense the operating status of the equipment in each substation based on the equipment in each substation, and obtain the sensing dataset of each substation equipment. The model intelligent analysis module is used to analyze the potential coupling relationship between operating features based on the implicit association perception model and the perception dataset, to obtain the implicit association features of each substation equipment, and to perform fault risk analysis based on the explicit logic reasoning model and the implicit association features, to obtain the safety decision results of each substation equipment. The risk pattern recognition module is used to perform decision collaboration based on the safety decision results between adjacent substation equipment to obtain the collaborative decision results of each substation equipment, and to perform potential risk pattern recognition based on the collaborative decision results to obtain the safety status recognition results of each substation equipment. The safety response module is used to generate safety intervention commands based on the safety situation identification results, and to perform local safety responses on substation equipment based on the safety intervention commands.

10. A non-transitory computer-readable storage medium, wherein a computer software program is stored therein, characterized in that, When the computer software program is executed by the processor, it implements the equipment safety management method based on substation intelligent agents as described in any one of claims 1 to 8.

Images