Substation fault remote management and control method based on WAPI network

Abstract

The invention relates to the technical field of power system monitoring, and discloses a WAPI network-based substation fault remote management and control method, which comprises the following steps that: a center node periodically broadcasts heartbeat beacons, and an edge unit immediately packages the serial number of the heartbeat beacons into a trigger response packet to return after detecting the transient sudden change of a power line; the center node constructs a disturbance propagation causal chain according to the response timestamps of the multiple edge units to the same beacon, and the disturbance propagation causal chain is matched with a preset fault mode to generate a diagnosis instruction. And meanwhile, a non-response mechanism and physical layer mark verification are utilized, so that disturbance energy level discrimination and causal authenticity guarantee capabilities are achieved under the condition that hardware complexity does not need to be increased, and the fault positioning precision and the system reliability are improved.

Description

TECHNICAL FIELD

[0001] The application relates to a substation fault remote management and control method based on a WAPI network and belongs to the technical field of power system monitoring. BACKGROUND

[0002] Current substation fault monitoring mainly relies on distributed sensors to collect power line waveform data and upload the data to a central node for analysis. This single-point collection-central processing mode has a fundamental technical bottleneck. When the power grid is disturbed, a large amount of waveform data transmission is triggered by multiple sensors at the same time, which not only causes communication bandwidth overload, but also causes critical fault information to be submerged in redundant data. Although the existing technology attempts to alleviate the data pressure by improving single-point sampling accuracy or optimizing compression algorithms, it cannot solve two essential problems: one is that it cannot distinguish between local noise and disturbance propagation events caused by real faults; the other is that it lacks the ability to capture the temporal and spatial propagation characteristics of disturbances in the substation physical network.

[0003] More deeply, the core defect of the existing technology is that the sensor is regarded as an independent data source, and the networked correlation characteristics formed by the coupling of power equipment through the line are completely ignored. For example, when local discharge is caused by insulation deterioration, the disturbance will propagate along a certain path and trigger multiple sensors in sequence, but the existing system can only obtain discrete alarm points and cannot reconstruct the propagation path to locate the fault source. In addition, external power grid disturbances often trigger simultaneous alarms of all station sensors, which is highly similar to the representation of station faults. The existing method relies on a single criterion of threshold comparison, which is prone to misjudgment.

[0004] The industry recently attempts to introduce a GPS synchronous clock to build a time reference, and to infer the disturbance direction by comparing the time difference of sensor triggering, but this scheme requires a high-precision clock for each node, which greatly increases the hardware cost and is limited by satellite signal coverage. More importantly, the propagation speed of electromagnetic waves in the cable is affected by the medium parameters and will produce subtle deviations, and the existing time synchronization accuracy cannot meet the centimeter-level fault positioning requirements. Therefore, how to build a time reference transfer mechanism based on the existing communication network to realize the dynamic reconstruction of the disturbance propagation path and the multi-dimensional discrimination of the fault nature through the simple cooperation of edge nodes has become a technical problem to be solved by the application. SUMMARY

[0005] The application provides a substation fault remote management and control method based on a WAPI network, which mainly aims to solve the problem that the existing monitoring system cannot capture the power disturbance propagation path and distinguish the fault nature at low cost and high reliability.

[0006] To achieve the above purpose, the application provides a substation fault remote management and control method based on a WAPI network, which comprises the following steps: Periodic broadcast heartbeat beacon, the center control node broadcasts the content unified heartbeat beacon data packet to the plurality of edge monitoring units through the WAPI network with a preset fixed period, and the heartbeat beacon data packet is used for providing a unified relative time reference for the plurality of edge monitoring units; Transient mutation detection and instant response, each edge monitoring unit continuously monitors the physical signal of the power line connected thereto, and detects whether the transient zero rate of the high-frequency component of the power line signal exceeds a specific threshold through a hardware comparator or an algorithm to determine whether a transient mutation exists; when the edge monitoring unit detects a transient mutation, the edge monitoring unit encapsulates the sequence number of the latest received heartbeat beacon data packet in a trigger response packet in less than 100 nanoseconds, and immediately sends it back to the center control node through the WAPI network, and the trigger response packet does not contain the waveform data of the power line physical signal; Causal chain construction, the center control node accurately records the arrival time stamp of each received trigger response packet; when trigger response packets responding to the same heartbeat beacon sequence number from at least two different edge monitoring units are received within a preset fixed heartbeat beacon broadcast period, the center control node sorts these trigger response packets according to the order of their arrival time stamps, and constructs a causal chain representing the propagation path of the transient mutation in the transformer station network according to the sorting and the time difference between the trigger response packets. Intelligent mode matching and control instruction generation, the center control node matches the constructed causal chain with a preset fault mode knowledge base, and the fault mode knowledge base includes fault rules defined based on physical topology and disturbance propagation characteristics; based on the matching result, a diagnostic report containing fault properties and segment positioning is generated and pushed to remote control personnel.

[0007] Preferably, in the transient mutation detection and instant response, when the edge monitoring unit detects that the transient mutation is triggered for the first time, it is recorded as moment, instead of immediately sending the trigger response packet, it enters a preset duration of refractory period, and the refractory period duration satisfies the expression , wherein represents the time required for the transient disturbance energy caused by normal operation to decay to the background noise level, and represents the minimum time for the transient disturbance energy caused by potential faults to continue oscillating; the transient detector is temporarily shielded by software logic during the refractory period, and does not respond to any signal; after the refractory period ends, the shielding is removed and a second inquiry is immediately made to detect whether the transient mutation exists again; according to the result of the second inquiry, an identifier representing the energy level of the transient mutation is attached to the trigger response packet, indicating that the transient mutation is a low-energy disturbance, and if the transient detector has returned to a calm state or a high-energy disturbance, if the transient detector is still in a triggered state.

[0008] Preferably, in the intelligent pattern matching and control instruction generation, the fault mode knowledge base comprises at least the following rules: a line fault location rule: if the edge monitoring unit sequence in the constructed causal chain is consistent with the known power cable physical path in the substation, and the time difference of the trigger response packet arrival of adjacent edge monitoring units on the chain and the theoretical propagation time deviation of electromagnetic waves on the corresponding physical path is less than the preset time error threshold, it is determined that a fault occurs in the corresponding physical path section; an internal and external disturbance distinguishing rule: if the trigger response packets from more than 90% of the edge monitoring units in the substation arrive almost simultaneously within the WAPI network communication jitter error range in a preset fixed heartbeat beacon broadcast period, it is determined that it is a wide-range disturbance conducted by the external power grid, and the alarm is suppressed.

[0009] Preferably, in the intelligent pattern matching and control instruction generation, the fault mode knowledge base further comprises a slow disease evolution identification rule: the central control node long-term statistics a specific causal chain mode, such as the appearance frequency of the causal chain between specific two edge monitoring units on the bus, and according to the appearance frequency, if the appearance frequency in the continuous three and more than three preset statistical periods all presents a growth trend, and the growth rate satisfies , wherein is a preset frequency growth rate threshold, to identify a slow development type fault hidden danger existing in the corresponding section.

[0010] Preferably, it further comprises: the edge monitoring unit first triggered by the transient mutation, while sending the trigger response packet to the central control node, injects a preset marker signal into the power line through the instantaneously activated execution circuit, the marker signal superimposes a small and unique electrical characteristic on the original disturbance waveform, the electrical characteristic includes a voltage depression or a current protrusion with a specific amplitude, duration and shape, to physically mark the wave front of the transient mutation; the transient detection logic of other edge monitoring units is configured as a double feature matching logic, and only when the transient mutation carrying the marker signal is monitored, the trigger response packet is sent to the central control node, and the marker signal is a preset electrical characteristic immediately following the rising edge of the transient waveform.

[0011] Preferably, the edge monitoring unit obtains the high-frequency component of the power line signal through an RC high-pass filter, and compares the high-frequency component with a threshold dynamically adjusted based on the historical background noise level through a comparator, to determine whether the instantaneous zero-crossing rate exceeds a specific threshold.

[0012] Preferably, the central control node ensures that the recording accuracy of the trigger response packet arrival time stamp reaches the microsecond level through a high-precision timing chip or an atomic clock synchronization module.

[0013] Preferably, the edge monitoring unit adopts a microcontroller as its core processing unit, and the microcontroller only performs signal discrimination, data packet packaging and communication tasks.

[0014] Compared with the prior art, the application has the following advantages: 1. By converting the transmission delay characteristics inherent in the WAPI communication network into a distributed timing measurement reference, the center node can naturally construct a causal chain representing the disturbance propagation path based on the asynchronous response timestamps of the edge units to the same heartbeat beacon. This mechanism sublimates traditional isolated single-point monitoring into networked spatiotemporal correlation analysis, making the propagation process of weak disturbances in the substation physical network observable for the first time, and fundamentally solving the cognitive blind spot of the disturbance propagation path between devices in the prior art. Through the cooperation of the unresponsive period mechanism and the secondary interrogation, the edge unit uses the comparison between the physical attenuation characteristics of the disturbance energy and the preset time window to achieve the essential distinction between low-energy operating disturbances and high-energy fault disturbances without the need to increase high-precision sampling circuits. This mechanism is deeply coupled with the causal chain construction step, enabling the system to automatically superimpose energy level identifiers on the chain nodes while locating the propagation path, thereby avoiding alarm storms triggered by benign disturbances at the root cause and improving the decision value of the diagnostic conclusion.

[0015] 2. The active marking behavior of the earliest responding unit to the disturbance wavefront forms a closed-loop verification with the dual feature matching logic of other edge units. By implanting identifiable electrical feature markers at the power line signal level, the successor nodes only respond to disturbance events carrying the same source marker. This mechanism and the timestamp analysis at the communication layer constitute a double mutual verification, ensuring that each link of the causal chain is established based on the blood relationship of the physical layer signal, and avoiding the risk of false correlation caused by multiple disturbance concurrency.

[0016] 3. When the center node performs trend analysis on the long-term accumulated causal chain data, it identifies specific propagation path patterns, such as the frequency growth characteristics of bus A→B in consecutive statistical periods, and converts the evolution of slow variables into quantifiable fault warning indicators. This mechanism extends the transient detection capability to the field of device life cycle management, enabling early identification of slow-changing faults such as early insulation degradation before the threshold alarm is triggered, avoiding the response lag limitations of traditional monitoring systems for slow disease-type hidden dangers. BRIEF DESCRIPTION OF DRAWINGS

[0017] Fig. 1 Figure 1 is a structural schematic diagram of the substation fault remote management and control system based on the WAPI network of the application; Fig. 2 Figure 3 is a timing diagram of disturbance event response and causal chain construction based on physical markers and dual feature matching of the application.

[0018] The purpose realization, functional characteristics, and advantages of the application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0019] It should be understood that the specific embodiments described herein are merely illustrative of the present application and should not be used to limit the present application in any manner.

[0020] The substation fault remote management and control method based on the WAPI network disclosed in the present application adopts a microcontroller as the core processing unit of the edge monitoring unit, and the microcontroller only performs signal discrimination, data packet packaging and communication tasks; the center control node continuously stores and analyzes the trend of the constructed causal chain, which is used to assist in predicting the fault evolution path, optimizing the fault mode knowledge base, and managing the operation state of the substation equipment throughout its life cycle; all edge monitoring units in the WAPI network are configured as a multi-point cooperative data acquisition network, and the communication delay between network nodes is used as a timing measurement reference, so that the WAPI network itself is converted into a distributed timing sensor array; at the same time, unlike the traditional method of judging the severity of disturbance based on the peak value of instantaneous amplitude, the present application adopts a time window-based disturbance response strategy, taking the continuity and change trend of the disturbance in a specific time period as the classification basis, compared with the existing method of judging only according to the amplitude change of the voltage or current instantaneous waveform, such method is easily disturbed by instantaneous sharp peak noise in actual application, resulting in an increase in the misjudgment rate, and by observing the continuous state of the disturbance signal within the preset time window and combining its stability and attenuation characteristics for identification, the recognition ability of the system to the physical nature of the disturbance event is effectively enhanced, thereby improving the distinguishing precision of the fault and non-fault disturbance, and avoiding the misidentification of short-time low-energy noise as a fault signal.

[0021] Embodiment 1: The application proposes a substation fault remote management and control method based on WAPI network, which converts the inherent transmission delay characteristics of the communication network into a time sequence measurement benchmark for constructing the disturbance propagation path. At the same time, through the lightweight discrimination mechanism of the edge node and the physical layer signal marking mechanism, high-precision identification, path tracking and property classification of the disturbance event in the substation power network are realized, thereby avoiding the technical bottlenecks existing in the data flooding processing, space-time correlation perception and disturbance discrimination of the existing scheme. The application includes the following key steps: first, the center control node broadcasts heartbeat beacons with uniform content at a fixed period through the WAPI network, providing a consistent relative time benchmark for multiple edge monitoring units; second, the edge monitoring unit monitors the physical signal connected to the power line in real time, and uses the instantaneous zero-crossing rate of the high-frequency component as the basis for discrimination to identify whether there is a transient mutation; after identifying the mutation, the edge node encapsulates the latest received heartbeat beacon sequence number into a trigger response packet within a response time of less than 100 nanoseconds, and sends it to the center control node through the WAPI network, and the response packet does not contain the original waveform data; finally, the center node constructs the causal chain of disturbance propagation based on the time stamp sequence of multiple edge nodes responding to the same beacon within one heartbeat broadcast period, and matches it with the preset fault mode knowledge base to generate diagnostic information containing fault properties and segment positioning. In the above technical path, the role of the heartbeat beacon is to provide a unified time reference point for the system, and the WAPI network serves as a communication medium and provides a natural time jitter benchmark using its delay characteristics. The edge monitoring unit, as the response end, is the core of the transient mutation detection task. In the context of the application, the transient mutation specifically refers to a short-time high-frequency disturbance event that occurs in the power line signal. The determination basis is whether the instantaneous zero-crossing rate of the high-frequency component exceeds the dynamic threshold. The dynamic threshold is adjusted by the edge node in real time based on the historical background noise level. Specifically, the high-frequency component of the signal is extracted through the RC high-pass filter, and compared with the dynamically set determination threshold through the software algorithm. Once the instantaneous zero-crossing rate of the high-frequency component exceeds the threshold, it is considered that a transient mutation has occurred. At this time, the edge node does not transmit the complete waveform, but only encapsulates the current beacon sequence number, maximally compresses the communication data volume, and improves the response efficiency. The center control node records the high-precision time stamp of all received trigger response packets through the built-in high-precision timing module, including the crystal oscillator time base or the synchronous clock with network synchronization function. If response packets from two or more edge nodes are detected within a beacon period, the center node sorts them according to their arrival time difference, and combines with the physical topology of the station power network to construct the causal chain of the disturbance propagation path, thereby locating the disturbance starting point and propagation direction.

[0022] The application introduces a refractory period mechanism and a secondary inquiry judgment logic to enhance the disturbance discrimination accuracy, when the edge monitoring unit detects a transient mutation for the first time (recorded as T1 moment), it does not immediately send a trigger response packet, but enters a preset refractory period time window Trefractory, which needs to meet the following conditions: Trefractory is greater than the natural energy decay time of benign disturbance (Tbenign decay), and less than the lower limit of the typical fault disturbance energy maintenance time (Tmalicious duration), during the refractory period, the transient detection logic of the edge node is temporarily software shielded to avoid false response to short-time random noise, after the refractory period ends, the system immediately performs a secondary inquiry to re-determine whether the transient mutation still exists, according to the inquiry result, a disturbance energy level identifier is attached in the response packet: if the mutation has subsided, it is marked as low-energy disturbance; if the mutation still exists, it is marked as high-energy disturbance, the above mechanism constructs the disturbance energy differentiation capability based on the time dimension, and strengthens the classification and identification of the nature of the disturbance event by the system, the edge monitoring unit triggered by the disturbance sends a trigger response packet at the same time, and injects a preset marker signal into the monitored power line through the instantaneously activated execution circuit, the signal is a specific voltage depression or current protrusion superimposed on the original power disturbance waveform with clear amplitude, duration and shape, such physical layer signal is used for disturbance bloodline feature labeling, other edge nodes introduce a double matching mechanism in the response logic: only when the disturbance waveform carrying the above marker signal is detected, it is considered as a homologous disturbance and a response packet is sent to the center node, the physical layer feature marker and the time stamp analysis mechanism form a double mutual evidence, which significantly improves the accuracy and robustness of the cause-effect chain construction, and avoids false link establishment caused by concurrent multi-source disturbance; the above content also improves the pattern matching ability of the center control node in intelligent diagnosis, which includes the following aspects: on the one hand, the built-in fault mode knowledge base of the center control node contains rules such as line path consistency matching and time difference tolerance analysis, which are used to judge whether the disturbance is the actual fault of a certain physical path section; on the other hand, when more than 90% of the edge nodes respond almost simultaneously within a single heartbeat period, and the time difference is within the communication jitter error range, the system can judge that it is a wide-area disturbance caused by external power grid interference, and automatically suppress the related alarm, thereby effectively reducing the false alarm probability, further, through the continuous statistics and trend analysis of historical cause-effect chain data, the center control node can identify slow variable evolution characteristics, for example, when the cause-effect chain of a certain path continuously increases in frequency in consecutive statistical periods, and the growth rate Ri is higher than the preset threshold Rth, it can be judged that there may be insulation aging and other slow-changing fault hidden dangers in the path, this strategy extends the disturbance chain construction capability originally used for instantaneous response to the full life cycle health management dimension of the substation operation equipment.The above mechanism constitutes a complete closed-loop system with heartbeat beacon as timing reference, minimal trigger logic as information entrance, causal chain construction as propagation tracking core, and physical layer signal marking and long-term trend analysis as auxiliary, wherein the delay characteristics of WAPI communication network are converted into measurement reference; the disturbance energy judgment relies on time window strategy rather than traditional amplitude comparison; the traceability of disturbance event is realized through physical signal identification. This multi-dimensional element reconstruction and system function reuse substantially improves the substation fault diagnosis capability.

[0023] In a typical system operation scenario, the central control node broadcasts a unified heartbeat beacon through the WAPI network every fifty milliseconds, each beacon packet carrying a monotonically increasing sequence number to provide a unified time reference for each edge monitoring unit. The edge monitoring unit continuously samples the instantaneous voltage signal of the connected power line and uses an RC high-pass filter in the hardware structure to extract the high-frequency component of the signal. The cutoff frequency of the filter is set according to the line voltage rating and the interference spectrum characteristics, typically not less than ten kilohertz in a typical one hundred and ten kilovolt system, to effectively filter out the power frequency component and retain the transient high-frequency disturbance. The filter output signal is connected to the comparison module and compared with the dynamic judgment threshold based on the historical background noise level. The threshold is dynamically adjusted according to the system safety factor to ensure that the discrimination logic has reasonable sensitivity under different electromagnetic backgrounds. If the instantaneous zero-crossing rate exceeds the dynamic threshold, the system preliminarily identifies the presence of a disturbance event. At this time, the edge monitoring unit does not respond immediately but enters a preset refractory period, typically ranging from three to eight milliseconds. The specific value should satisfy the condition that it is greater than the natural decay time of benign disturbance and less than the lower limit of the typical fault disturbance duration. During the refractory period, the transient detection logic is software-shielded to avoid false positives due to short-term interference. After the refractory period ends, a secondary inquiry is immediately executed, i.e., the current high-frequency state is sampled and judged again. If the disturbance has subsided, it is marked as a low-energy disturbance. If the disturbance is still present, it is marked as a high-energy disturbance. The energy level identifier is packaged into the response packet along with the latest received heartbeat sequence number. If this edge node is the first to be disturbed, the system will simultaneously activate the execution circuit to inject a marked signal with specific physical characteristics into the monitored power line. The signal is a micro-amplitude, short-time voltage depression or current protrusion, with unique amplitude, shape, and duration to facilitate subsequent node identification. A double-condition matching mechanism is introduced into the detection logic of other edge nodes, i.e., only when the detection signal contains the above-mentioned marker characteristics, is the disturbance confirmed as a homogenous event and a trigger response is sent out based on the high-frequency discrimination. At the same time, during implementation, the edge monitoring unit does not respond immediately after detecting a transient disturbance but enters a preset refractory period. After the refractory period ends, a secondary inquiry is performed to determine whether the disturbance is still present. Specifically, if it is determined to be a high-energy disturbance, i.e., the transient detector is still in the triggered state, the edge monitoring unit packages the current heartbeat sequence number and disturbance energy identifier into a trigger response packet and sends it to the central control node through the network. If it is determined to be a low-energy disturbance, i.e., the transient detector has returned to a calm state, the system considers that the disturbance has naturally decayed and does not send a trigger response packet to avoid false positives or redundant transmissions.

[0024] All valid response packets are sent to the central control node through the WAPI network. The central control node is equipped with a high-precision timestamp recording module, which can ensure that the receiving time of each response packet has a resolution of microseconds. Within a single heartbeat cycle, if response packets from two or more edge nodes are received and the response beacon serial numbers are consistent, the central control node will sort them in chronological order according to the timestamps. Combined with the physical topology of the station cable, the central control node constructs a causal chain of disturbance propagation paths. If the response time difference between adjacent nodes and the difference between the theoretical propagation time of electromagnetic waves on the path are within the preset tolerance threshold, the system determines that the path segment is a valid transmission path. If the difference exceeds the threshold, it is excluded to eliminate non-real causal relationships. The completed causal chain is input into the fault mode matching module. This module has multiple standard fault rules, including matching mechanisms based on path consistency, response timing characteristics, and physical marker identification. The matching results are used to generate diagnostic information containing fault properties, segments, and confidence levels. The information is provided to the operation and maintenance personnel through the system interface or remote platform for reference. In addition, the system supports long-term archiving of all disturbance chain records and trend modeling by day, week, month, etc. If the causal chain on a fixed path shows an increasing frequency trend in consecutive statistical periods, and the growth rate exceeds the system's set threshold, it is automatically marked as a possible slow-changing hidden danger, such as insulation aging or contact resistance increase. This prompts the user to further check.

[0025] Example 3: This example aims to investigate and verify the practical application effect of the substation fault remote management and control method based on WAPI network in capturing power disturbance propagation path, distinguishing fault nature and improving fault location accuracy and system reliability through systematic verification in a simulated substation power network environment. The test scheme design aims to clarify the implementation details of the method, including using communication network delay characteristics as a time and space measurement benchmark, constructing a disturbance propagation causal chain based on heartbeat beacon serial number and timestamp, and combining refractory period mechanism and physical layer marking for disturbance discrimination and energy level determination. To simulate the complexity of real substation power networks and the dynamic characteristics of communication environment, a configurable test platform is built, which includes: a programmable simulated power network unit to reproduce power transmission paths of different lengths and impedance characteristics, and can inject multiple typical disturbance signals as needed, with parameters such as occurrence location, duration, energy intensity and propagation speed can be finely set; a WAPI communication network simulation module based on software-defined network architecture to simulate WAPI communication network with controllable communication delay, jitter and packet loss rate; and a deployed central control node and edge monitoring unit array, where the central control node integrates a high-precision timing chip, responsible for periodic broadcasting of heartbeat beacons, receiving and processing trigger response packets, executing causal chain construction, fault mode matching and diagnosis instruction generation, the edge monitoring unit is connected to the monitoring points in the simulated power network, configured with RC high-pass filter and comparator for real-time detection of whether the high-frequency component transient zero-crossing rate of power line signal exceeds the dynamically adjusted threshold, the microcontroller mainly performs signal discrimination, data packet packaging and communication tasks, in addition, human-computer interaction and data visualization interface are also provided.

[0026] Three typical substation disturbance events are selected for verification: local short circuit fault, external grid wide range disturbance and internal device insulation deterioration caused by slow fault. Each disturbance is repeated on multiple energy levels and propagation paths. In the establishment of heartbeat beacon and time reference, the central control node broadcasts heartbeat beacons at a fixed period of 50 milliseconds, each beacon carries a unique serial number, multiple edge monitoring units record the time points of receiving consecutive heartbeat beacons, and compare them with the sending timestamp of the central node to evaluate the accuracy of the relative time reference. The test observed that even with the inherent communication jitter of the WAPI network, the time stamps of the same heartbeat beacon received by each edge unit showed high consistency, with a relative deviation from the central node's sending timestamp in the order of microseconds, which is much smaller than the typical time delay of power disturbance propagation, indicating that it is feasible to use the communication delay characteristics of the WAPI network itself as a time and space measurement benchmark, In the transient mutation detection and trigger response mechanism verification, for different positions, energy levels of local short circuit fault in the simulation power network, the transient mutation signal is injected, when the edge monitoring unit detects the transient mutation, the time when the instantaneous zero rate exceeds the dynamic threshold is recorded, and the time from detection to the generation and sending of the trigger response packet is measured, which is usually less than 100 nanoseconds, the verification of the refractory period mechanism includes: when the edge unit first detects the disturbance (T time), it does not immediately send a response packet, but enters a preset time (for example, 3 milliseconds) of refractory period, after the refractory period ends, the system immediately performs secondary interrogation, detects the high frequency state again, according to the secondary interrogation result, the disturbance energy level identifier is attached in the trigger response packet: if the disturbance has subsided, it is marked as low energy disturbance; if the disturbance is still in, it is marked as high energy disturbance, the test data shows that the edge monitoring unit can quickly identify various transient power disturbances, under the action of the refractory period mechanism, the system suppresses the false alarm caused by short-time random noise or transient electromagnetic interference.

[0027] In terms of causal chain construction and disturbance propagation path tracking, disturbance signals with specific propagation time delays are injected on different physical paths of the simulated power network in sequence, the central control node receives trigger response packets from different edge units and accurately records their arrival time stamps, and within a single heartbeat beacon cycle, if trigger response packets from at least two edge units responding to the same heartbeat beacon sequence number are received, the central control node will construct a disturbance propagation causal chain according to the chronological order and mutual time difference of these time stamps, in combination with the preset physical topology of the power network, and the constructed causal chain will be mapped to the simulated power network topology on the visualization interface to achieve dynamic tracking of the disturbance propagation path. Test results show that the central control node can identify the propagation order and path of the disturbance; in terms of intelligent mode matching and diagnostic instruction generation, multiple typical fault modes are preset and their corresponding causal chain features are stored in the fault mode knowledge base, the central control node matches the constructed causal chain with the knowledge base to generate a diagnostic report containing fault properties, segment positioning and corresponding control instructions, the internal and external disturbance differentiation rules are tested: when more than 90% of the edge monitoring units respond almost simultaneously within the preset heartbeat beacon broadcast cycle, and the time difference is within the WAPI network communication jitter error range, the system judges it as an external power grid disturbance and suppresses the related alarm, the slow evolution identification rule is tested: the frequency of occurrence of a specific causal chain mode is statistically long-term, and according to the growth trend of the frequency of occurrence within a continuous statistical cycle, when the growth rate Ri reaches or exceeds a certain preset threshold Rth, a slowly developing fault hazard is identified, and test results show that the system can identify the fault type and generate a corresponding diagnostic report according to the causal chain features and knowledge base content; in terms of physical layer marking and causal authenticity guarantee, in the simulated power network, multiple concurrent disturbance signals are injected at multiple sources and multiple time points, the edge monitoring unit triggered by the disturbance first sends a trigger response packet while injecting a preset marker signal into the power line through the instantaneous activation execution circuit, and other edge monitoring units are configured with double feature matching logic, and only when a high-frequency abrupt signal containing the above marker signal is detected, a trigger response packet is sent, and test data shows that in the multi-disturbance concurrent scenario, the introduction of the physical layer marker signal enhances the accuracy of the causal chain, ensures that each link of the causal chain is established based on the blood relationship of the physical layer signal, and enhances the authenticity of the diagnostic result.

[0028] Embodiment 4: The embodiment combines Figs. 1-2 A WAPI network-based substation fault remote control method is implemented and described. As shown in Fig. 1As shown, first, the central control node undertakes the core functions of heartbeat beacon broadcast, causal chain construction, fault mode matching, etc., and periodically sends heartbeat beacons to all edge monitoring units through the WAPI network to build a unified relative time reference. Each edge monitoring unit is responsible for transient detection and trigger response tasks. When a transient disturbance is detected, the latest heartbeat beacon sequence number received is packaged into a trigger response packet and returned to the central control node through the WAPI network. The figure also shows the disturbance propagation path on the power line, i.e., the propagation trajectory of the disturbance between different edge monitoring units. The central node establishes the causal chain of propagation according to the time difference of multiple response packets and matches it to the built-in fault knowledge base for intelligent analysis. The diagnosis report is generated by the central control node and transmitted to the remote control personnel for receiving the diagnosis report for remote intervention.

[0029] As shown, Fig. 2 The first stage is edge site detection and physical marking. The central control node initiates broadcast heartbeat beacons. When edge unit A (the first trigger) detects an anomaly, it enters a refractory period and performs secondary interrogation logic to confirm the disturbance according to the first mutation detected, and determines whether to send a trigger response packet (A) or suppress the response according to the judgment of high-energy disturbance or low-energy disturbance, respectively. If it is a high-energy disturbance, a preset physical marker signal is injected into the power line (physical channel), and then the second stage of subsequent unit synchronization verification and response is entered. That is, when edge unit B (the subsequent trigger) receives the disturbance waveform and marker signal transmission, it starts the double feature matching logic, i.e., it judges the disturbance waveform feature and the physical marker from unit A at the same time. If the double feature matching is successful, a trigger response packet (B) is sent. If the matching fails, it is determined as a non-homogeneous event and the response is suppressed. Finally, the third stage of central node causal chain construction is entered. That is, the central control node receives the response packets (A) and (B) in sequence according to the received response packets, records the respective arrival times, and constructs the causal transmission chain (A→B) according to the response order and physical marker.

[0030] In the application background of old substations, the system operation and maintenance unit has long been faced with the problem of concurrent slow faults caused by intermittent false alarms of grounding systems and aging of cable insulation within the station. In particular, during the spring-summer transition and in the stage of severe environmental humidity fluctuations, inductive coupling paths formed by cable sheaths between some devices are prone to induce short-time disturbances. Such disturbances generally decay naturally within a few milliseconds and do not constitute electrical faults. However, existing monitoring systems have difficulty in distinguishing such events from substantive faults, often leading to an explosion in the number of alarms and subsequent diagnostic distortion. In this application background, this embodiment refines the key functional modules and implementation logic based on the framework of the aforementioned technical solution, and clearly defines the reasonable basis for setting each parameter and its adaptation relationship with the specific application scenario, to construct a technical path that can be actually deployed.

[0031] First, for the first trigger response mechanism of power disturbance event, in the embodiment, each edge monitoring unit is configured with a set of disturbance preliminary judgment module run by microcontroller, which receives high frequency component signal from high pass filter output, and dynamically counts the instantaneous zero-crossing number of the signal with five milliseconds as sliding window. To ensure the adaptability of the judgment, the module calculates the dynamic threshold value for the current judgment based on the root mean square amplitude of the high frequency signal counted in the window, and multiplies a preset sensitivity coefficient. The setting principle of the sensitivity coefficient is to reflect the change trend of the electromagnetic background disturbance level in the station, which is usually between one point five to one point eight. The specific value is dynamically adjusted by the system upper platform combined with the disturbance activity parameter to achieve the balance control of sensitivity and stability. When a certain edge unit detects more than two times of instantaneous high frequency zero-crossing events exceeding the dynamic threshold value in the current sliding window, it is considered that preliminary disturbance occurs, and the current heartbeat beacon serial number is recorded immediately, and the refractory period timer is started. The refractory period set in the embodiment is six milliseconds, which is set on the basis of the technical safety margin between the typical benign disturbance natural decay period (about three milliseconds) reflected by the historical operation data of the substation and the minimum maintenance time (generally not less than eight milliseconds) of common persistent discharge event, aiming to avoid repeated triggering of the diagnosis process due to transient interference. After the refractory period ends, the edge unit immediately starts the secondary inquiry mechanism, samples the high frequency signal again and repeats the above instantaneous zero-crossing rate judgment logic. If it still exceeds the dynamic threshold value, the system determines that the current disturbance is high-energy disturbance. If the zero-crossing rate has returned to below the warning threshold, it is determined to be low-energy disturbance. The disturbance level result together with the corresponding heartbeat beacon serial number is packaged as a response data packet and returned to the center node.

[0032] To achieve the unique identification and physical tracing of the chain propagation path, in this embodiment, only when the edge unit first detects the disturbance and confirms it as a high-energy disturbance through secondary interrogation, the embedded physical marker signal injection circuit will be triggered. This circuit is controlled by the integrated control logic of the communication module and can inject a set of standardized voltage notch signals on the monitored cable with a duration of zero point eight milliseconds. The amplitude is about zero point one five times the power frequency voltage, with fixed edge characteristics and time domain form. The injection signal parameters have been set before the system is deployed, taking into account the substation cable structure, electrical parameters and load characteristics, to ensure that it will not affect the system stability under normal operating conditions, while having characteristics that can be identified by other edge units. In the subsequent local judgment process of other edge monitoring units, in addition to completing the initial judgment of the disturbance based on the high-frequency zero-crossing rate, they also need to perform feature recognition operations on the received signal based on template matching. The matching template is composed of the reference waveform of the system preset marker signal. The edge unit performs correlation matching calculation with a weighted sliding window. If the correlation between the current sampling signal and the template exceeds zero point eight five, it is considered that there is a marker signal and it is considered to be a disturbance homologous. If no marker features are detected, it is determined that the disturbance is an independent event and does not participate in the construction of the current propagation chain, effectively preventing path confusion caused by multiple unrelated disturbance sources concurrent.

[0033] At the central control node end, to realize the construction and propagation path determination of the causal chain, the node is integrated with a high-precision timing recording chip with a temperature-compensated crystal oscillator, and the sampling resolution is better than 500 nanoseconds. All received edge response packets will be uniformly converted into time offset values relative to the first response packet in the current period. The control node maintains a node path mapping table based on the physical topology data built in the station, recording the actual line distance between each edge unit and its corresponding theoretical propagation delay. In the processing process, the control node compares the relative delay value reported by each unit with the corresponding theoretical delay in the mapping table. If the error is within ten percent of the theoretical delay, it is determined that the physical propagation path is valid, and the related node information is included in the current disturbance causal chain. Considering the actual situation that multiple source disturbances may occur concurrently, in this embodiment, the central control node groups and clusters the heartbeat sequence numbers of all response data packets in each heartbeat beacon period. Only the response data under the same sequence number is considered as homologous response. If more than eighty percent of the edge nodes respond in a single heartbeat period, and the standard deviation of the response delay is lower than the normal jitter range of the network, the system will preliminarily determine that the event is an external uniform disturbance signal propagation, and stop the causal chain analysis in the current period to avoid introducing non-local chain response misjudgment. In terms of long-period trend monitoring, the central control node statistics the disturbance causal chain occurrence frequency on each physical path per day, and generates a visual monitoring map. If the disturbance chain occurrence frequency of a transmission path increases by more than thirty percent in the previous period in the last three daily periods, the system will issue a slow fault risk warning based on the preset rules, and upload it to the upper layer operation and maintenance platform for manual diagnosis and confirmation.

[0034] In this embodiment, for the determination process of the dynamic threshold, the edge monitoring unit continuously samples the high-frequency components of the connected power line and statistically characterizes the instantaneous changes based on a sliding time window. To ensure that the judgment basis is adaptive, the process does not use static fixed values, but uses a setting method that is dynamically related to the background noise level. Specifically, the edge unit records the change amplitude of the high-frequency signal at each sampling point in a sliding time window of no less than two seconds, calculates the root mean square value, and multiplies this value by the sensitivity factor set by the system to generate the judgment baseline. The sensitivity factor is pre-set by the system as a constant in a limited interval, for example, between one and two. The value logic can be based on the activity of electromagnetic interference and the common amplitude variation range of high-frequency disturbances in historical data of the site, or can be dynamically adjusted in combination with the operation cycle. This dynamic baseline serves as the upper limit of the threshold, and is further used to determine whether the instantaneous zero-crossing rate exceeds the normal fluctuation level, thereby determining whether there is a sudden event. In the non-response period mechanism, to avoid false triggering caused by external short-time noise, the system sets a non-response interval. This time length does not use an absolute fixed value, but is defined by a reference boundary related to the operating characteristics of the site. The system limits the setting of the non-response period to a time period between the time required for natural decay of benign disturbance and the minimum time of typical fault disturbance. For example, benign disturbance can be obtained by analyzing non-fault disturbance signals collected by the edge unit during normal operation, which generally decays within three milliseconds. Typical electrical faults such as grounding or discharge generally last for six to eight milliseconds or more. Therefore, the non-response period can be set to about five milliseconds. In this interval, the transient detection logic is suspended, ensuring the robustness and anti-interference ability of the system. On this basis, the system designs a clear disturbance energy level division mechanism to avoid the influence of judgment ambiguity when triggering on the diagnosis result. After the non-response period ends, the edge unit immediately starts the detection process again to re-evaluate whether the current high-frequency disturbance is still active. If the disturbance has significantly weakened and does not trigger the judgment threshold, it is identified as a low-energy disturbance. Otherwise, if the disturbance is still significantly higher than the baseline, it is considered a high-energy disturbance. This level identifier is packaged in the trigger response packet in a clear field to facilitate the central control node to add judgment dimensions in subsequent analysis.

[0035] For the construction of the disturbance propagation causal chain, the central control node constructs a response chain according to the received response packet time stamps. Each response packet is recorded at the time of arrival at the central control node, and the time value is referenced to the heartbeat beacon as the reference point, rather than the absolute time, thereby avoiding the error introduced by the slight deviation of the system clock of each edge unit. In the construction of the causal chain, the system compares the response time difference between two adjacent nodes with the propagation delay on the physical path. The propagation delay is estimated from the known cable length and the signal propagation speed corresponding to the medium parameters. Generally, the dielectric constant and the conventional electromagnetic wave propagation speed can be referred to for estimation. In order to avoid misjudgment caused by delay drift of individual nodes, the system allows a limited tolerance between the propagation delay and the actual response time difference. The tolerance is set to not more than 10% of the theoretical value of the path propagation time, ensuring that the matching process has a reasonable physical consistency basis. Further, to enhance the accuracy of the judgment of disturbance homology, the edge unit is configured with a physical layer feature recognition mechanism. After the first detection unit injects a physical marker signal, other nodes need to compare the specific form feature in the signal when executing the trigger logic, in addition to meeting the high-frequency judgment. If the feature is highly consistent with the preset marker template, it is determined to be a node in the same propagation chain. Such feature comparison does not rely on complex Fourier transform and other high-computing methods, but is completed based on pattern matching operation in the sliding window. The template is trained by the operation and maintenance personnel in combination with the cable characteristics before system deployment, and is uniformly stored in each node to ensure the consistency and anti-interference ability of the response. In addition, in the causal chain trend analysis process, to avoid single disturbance statistical anomalies interfering with the overall judgment, the system uses a daily cycle level frequency statistical window to record the number of occurrences and change trend of the same physical path causal chain. If the system finds that the response frequency of a path causal chain is continuously increasing within a three-day cycle, and the increase is higher than the system set growth limit value, the system will generate a preliminary warning information for the path and submit it to the remote platform for reference by the operation and maintenance personnel. The growth limit does not use a fixed proportion threshold, but is dynamically calculated in combination with the global average disturbance growth level of the system in the past period, ensuring that the warning mechanism has context sensitivity and engineering adaptability. All of the above are the extended implementation manners known to those skilled in the art.

[0036] It is apparent for those skilled in the art that the present application is not limited to the details of the above exemplary embodiments, and the present application can be implemented in other specific forms without departing from the spirit or essential characteristics of the present application.

[0037] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not limiting. Although the present application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or replaced by equivalents without departing from the spirit and scope of the present application.

Claims

1. A method for remote management and control of substation faults based on a WAPI network, characterized in that, The method includes: The heartbeat beacon is broadcast periodically. The central control node broadcasts a unified heartbeat beacon data packet to multiple edge monitoring units through the WAPI network at a preset fixed period. The heartbeat beacon data packet is used to provide a unified relative time reference for multiple edge monitoring units. Transient change detection and immediate response: Each edge monitoring unit continuously monitors the physical signals of the power lines it is connected to, and uses hardware comparators or algorithms to detect whether the instantaneous zero-crossing rate of the high-frequency components of the power line signals exceeds a specific threshold to determine whether a transient change exists. When an edge monitoring unit detects a transient change, it encapsulates the sequence number of the latest received heartbeat beacon data packet in a trigger response packet within less than 100 nanoseconds and immediately sends it back to the central control node through the WAPI network. The trigger response packet does not contain waveform data of the power line physical signals. The causal chain is constructed by the central control node accurately recording the arrival timestamp of each triggered response packet received. When a triggered response packet with the same heartbeat beacon sequence number is received from at least two different edge monitoring units within a preset fixed heartbeat beacon broadcast period, the central control node sorts these triggered response packets according to their arrival timestamps and constructs a causal chain representing the propagation path of transient changes in the substation network based on the sorting and the time difference between each triggered response packet. Intelligent pattern matching and control command generation: The central control node matches the constructed causal chain with the preset fault mode knowledge base, which contains fault rules defined based on physical topology and disturbance propagation characteristics; based on the matching results, a diagnostic report containing the fault nature and segment location is generated and pushed to remote control personnel.

2. The method for remote management and control of substation faults based on a WAPI network according to claim 1, characterized in that, In the transient mutation detection and immediate response, when the edge monitoring unit detects the first triggering of a transient mutation, it records it as... At any given moment, instead of immediately sending a trigger response packet, it enters a refractory period of a preset duration. Satisfy expression ,in This represents the time required for the energy of a transient disturbance caused by normal operation to decay to the level of background noise, while This indicates the minimum duration of continuous oscillation of transient disturbance energy caused by a potential fault; during the refractory period, the transient detector is temporarily shielded by software logic and does not respond to any signal; after the refractory period ends, the shielding is removed and a second interrogation is immediately performed to check for the existence of the transient mutation again; based on the result of the second interrogation, an identifier representing the energy level of the transient mutation is added to the trigger response packet, indicating that the transient mutation is a low-energy disturbance if the transient detector has returned to a calm state or a high-energy disturbance, or if the transient detector is still in the triggered state.

3. The method for remote management and control of substation faults based on a WAPI network according to claim 1, characterized in that, In the intelligent pattern matching and control instruction generation, the fault mode knowledge base contains at least the following rules: Line fault location rule: If the sequence of edge monitoring units in the constructed causal chain is consistent with the physical path of the known power cable in the substation, and the arrival time difference of the trigger response packet of the adjacent edge monitoring unit on the chain is less than the theoretical propagation time of the electromagnetic wave on the corresponding physical path, then it is determined that a fault has occurred in the corresponding physical path segment. Internal and external disturbance distinction rules: If, within the preset fixed heartbeat beacon broadcast cycle, trigger response packets from 90% or more of the edge monitoring units within the substation arrive simultaneously within the WAPI network communication jitter error range, it is determined to be a wide-range disturbance transmitted from the external power grid, and the alarm is suppressed.

4. The method for remote management and control of substation faults based on a WAPI network according to claim 1, characterized in that, In the intelligent pattern matching and control command generation process, the fault mode knowledge base also includes chronic disease evolution identification rules: the central control node performs long-term statistical analysis of specific causal chain patterns, such as the frequency of occurrence of causal chains between two specific edge monitoring units on the bus, and the frequency of occurrence shows an increasing trend in three or more consecutive preset statistical periods, and the growth rate is... satisfy ,in A preset frequency growth rate threshold is used to identify potential slow-developing faults in the corresponding sections.

5. A method for remote management and control of substation faults based on a WAPI network according to claim 1, characterized in that, Also includes: The edge monitoring unit, which is first triggered by the transient change, sends a trigger response packet to the central control node and injects a preset marking signal into the power line through the instantaneously activated execution circuit. The marking signal superimposes a small and unique electrical feature on the original disturbance waveform. The electrical feature includes a voltage dip or current bulge with a specific amplitude, duration and shape to physically mark the wavefront of the transient change. The transient detection logic of other edge monitoring units is configured as dual feature matching logic, which only sends a trigger response packet to the central control node when a transient change carrying a marker signal is detected. The marker signal is a preset electrical feature that follows the rising edge of the transient waveform.

6. The method for remote management and control of substation faults based on a WAPI network according to claim 1, characterized in that, The edge monitoring unit obtains the high-frequency components of the power line signal through an RC high-pass filter, and compares the high-frequency components with a threshold dynamically adjusted based on the historical background noise level through a comparator to determine whether the instantaneous zero-crossing rate exceeds a specific threshold.

7. A method for remote management and control of substation faults based on a WAPI network according to claim 1, characterized in that, The central control node uses a high-precision timing chip or atomic clock synchronization module to ensure that its recording accuracy of the arrival timestamp of the trigger response packet reaches the microsecond level.

8. The method for remote management and control of substation faults based on a WAPI network according to claim 1, characterized in that, The edge monitoring unit uses a microcontroller as its core processing unit. The microcontroller only performs signal discrimination, data packet encapsulation, and communication tasks.

Images