A data acquisition and association method and system for a substation secondary circuit

Abstract

The application discloses a kind of data acquisition and association method and system of substation secondary circuit, belong to substation technical field, by establishing structure topological graph and generating data label for node, using event-driven mechanism to capture the data slice of state change moment.Based on the establishment of verification task based on theoretical connection relationship, the multi-dimensional response characteristics of the node are extracted, a multi-dimensional physical constraint model integrating time dimension, electrical logic dimension and signal quality dimension is constructed, and the confidence of node connection is quantified by calculating comprehensive verification indicators.When the connection is found to be untrustworthy, the system automatically scans all network nodes to establish a temporary verification task, locates and corrects the real connection nodes, and completes the topology update.The technical problems of static label unable to track physical connection changes, passive diagnosis unable to actively prevent, and unable to automatically repair after problem discovery are solved.

Description

A method and system for data acquisition and correlation of secondary circuits in substations Technical Field

[0001] This invention relates to the field of substation technology, specifically to a data acquisition and correlation method and system for substation secondary circuits. Background Technology

[0002] Substation secondary circuits are low-voltage circuits used for controlling, protecting, measuring, and monitoring primary equipment to ensure the safe and stable operation of the power system. To achieve digital and visual management of the complex connections in substation secondary circuits and improve the efficiency and safety of operation and maintenance, the industry commonly uses a technique of parsing SCD (Substation Configuration Description) files to extract the defined devices, terminals, and their connections, thereby constructing a theoretical topology model. Subsequently, based on this model, smart tags, such as QR codes, containing the identity and connection information of each node in the circuit are generated and affixed to the physical equipment in the field, serving as a digital guide for daily inspections, maintenance, and troubleshooting by operation and maintenance personnel. In related technologies, for example, Chinese patent document with authorization announcement number CN-109241065B discloses a smart substation cable operation information identification system and its identification method. It discloses setting QR codes containing smart substation secondary circuit diagrams, virtual terminal diagrams, physical network addresses and operation and maintenance logs on the cables, solving the technical problem that it is inconvenient to fully understand the business information transmitted by communication cables during the operation and maintenance of smart substation cables; and realizing the monitoring of detailed cable information and operation and maintenance records.

[0003] However, during the long-term operation and maintenance of substations, the physical connections of secondary circuits frequently change due to equipment modifications, routine maintenance, or temporary jumpers. Because existing technology constructs a static model, the theoretical model cannot be updated synchronously with physical reality. This causes the smart tag information relied upon by maintenance personnel to gradually become outdated or even incorrect, not only affecting the efficiency of troubleshooting and maintenance but also potentially leading to misoperation due to erroneous information, posing a significant safety hazard. Summary of the Invention

[0004] In view of the above-mentioned problems, the present invention provides a method and system for data acquisition and correlation of secondary circuits in substations.

[0005] Therefore, the technical problems solved by this invention are: First, the inability of static tags to be dynamically verified. Existing technologies construct one-time static models, lacking the ability to continuously verify the authenticity of node connections, and cannot detect changes in physical connections in a timely manner, resulting in long-term inconsistencies between tag information and actual connections. Second, the inability of passive diagnosis to proactively prevent problems. Existing technologies can only discover connection anomalies through manual troubleshooting after a fault occurs, and cannot proactively identify and warn of potential connection errors during operation. Third, the inability to automatically repair problems after they are discovered. Even if connection anomalies are discovered through certain means, existing technologies cannot automatically search for and locate the corresponding connection nodes, and can only rely on manual on-site verification and manual correction, resulting in low automation, low efficiency, and a high risk of errors.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a method for data acquisition and correlation of secondary circuits in a substation, comprising,

[0007] Establish a structural topology diagram to describe the connection relationship of secondary circuit nodes in a substation, generate data labels for each node that include input / output semantics, theoretical time delay windows, and local connection information, and establish an event-driven slice acquisition mechanism. When a change in node status is detected, the time window is extracted based on the time of change to obtain an event slice containing dynamic response data of multiple nodes.

[0008] Based on the theoretical connection relationship in the structural topology diagram, a verification task is established for the central node in the event slice and its theoretical downstream node. Multi-dimensional response features of the nodes in the verification task are extracted. For the verification task, a multi-dimensional physical constraint model with integrated constraints is constructed. The constraints include time dimension constraints, electrical logic dimension constraints, and signal quality dimension constraints.

[0009] Based on the node response characteristics and the multidimensional physical constraint model, the comprehensive verification index of the verification task is calculated, and the confidence level of the connection of the central node is obtained according to the distribution of the verification index of all verification tasks of the central node.

[0010] When the confidence level indicates a connection failure, the central node automatically locates and corrects itself to the corresponding connection node by establishing a temporary verification task with other nodes in the network and calculating the verification index, thus completing the adaptive update of the topology.

[0011] As a preferred embodiment of the data acquisition and association method for the secondary circuit of a substation described in this invention, the establishment of the event-driven slice acquisition mechanism includes: deploying a data acquisition system to synchronize the time of all nodes in the station; establishing event triggering rules, triggering data acquisition when the system detects any state change or value jump of a node, and capturing preset time windows forward and backward from the triggering time to obtain the data of all nodes within the time period, thus forming an event slice.

[0012] As a preferred embodiment of the data acquisition and association method for substation secondary circuits described in this invention, the step of establishing a verification task between the central node in the event slice and its theoretical downstream node includes taking the node that triggers the event in the event slice as the central node and obtaining the direct downstream node of the central node in the structural topology diagram; the central node and each direct downstream node form a verification task.

[0013] As a preferred embodiment of the data acquisition and association method for the secondary circuit of a substation according to the present invention, the step of extracting multi-dimensional response features of nodes in the verification task includes: acquiring the response data of the direct downstream node corresponding to the verification task after the event is triggered; extracting multi-dimensional response features, wherein the response features include at least: the rate of change of response data, the linear fitting quality features, the amplitude change direction features, the frequency domain energy distribution features, and the fluctuation features of the response stabilization period.

[0014] As a preferred embodiment of the data acquisition and association method for the secondary circuit of a substation described in this invention, the time dimension constraint includes obtaining the actual response time of the direct downstream node corresponding to the verification task.

[0015] Based on the theoretical delay window in the node labels, a Gaussian function is used to establish a time constraint gate to quantitatively evaluate the deviation between the actual response time and the theoretical delay window. A high score is given when the response time falls within the theoretical window, and a penalty is imposed when it falls outside the window.

[0016] As a preferred embodiment of the data acquisition and association method for substation secondary circuits described in this invention, the electrical logic dimension constraints include predicting the expected change direction and trend of node response based on the input and output semantics in the node labels.

[0017] Establish logical constraint gates and quantify the logical compliance of the response by comparing the consistency between the actual response characteristics and the expected response characteristics.

[0018] As a preferred embodiment of the data acquisition and association method for the secondary circuit of a substation described in this invention, the signal quality dimension constraint includes: establishing a quality constraint gate, and comprehensively evaluating the linear fitting goodness of the response data, the regularity of the frequency domain energy distribution, and the stability after the response.

[0019] Response signals with high linear fit and concentrated frequency domain energy distribution are selected by mass constraint gate.

[0020] As a preferred embodiment of the data acquisition and association method for the secondary circuit of a substation described in this invention, the calculation of the comprehensive verification index of the verification task includes multiplying and combining the change rate of the node response with the evaluation results of the time constraint gate, the logic constraint gate, and the quality constraint gate to obtain the comprehensive verification index of the verification task.

[0021] The multiplication relationship ensures that only connections that simultaneously satisfy the requirements of reasonable timing, logical correctness, and reliable quality can achieve a high check index.

[0022] As a preferred embodiment of the data acquisition and association method for the secondary circuit of a substation described in this invention, the automatic positioning and correction to the corresponding connection node includes establishing temporary verification tasks between the central node and the non-direct downstream nodes in the structural topology diagram, and calculating the comprehensive verification index of each temporary verification task.

[0023] Based on the distribution of the verification index of the temporary verification task, nodes with verification indices better than the theoretical connection are identified, and these nodes are determined as the corrected downstream nodes of the central node, and the data labels of the nodes are updated.

[0024] This invention provides a data acquisition and correlation system for the secondary circuit of a substation.

[0025] To solve the above technical problems, the present invention provides the following technical solution: a data acquisition and association system for secondary circuits of substations, comprising: a construction module, which establishes a structural topology diagram for describing the connection relationship of nodes in the secondary circuits of substations, generates data labels for each node containing input and output semantics, theoretical time delay windows and local connection information, establishes an event-driven slice acquisition mechanism, and when a change in node status is detected, extracts a time window based on the time of change to obtain an event slice containing dynamic response data of multiple nodes;

[0026] The extraction module establishes verification tasks for the central node and its theoretical downstream nodes in the event slice based on the theoretical connection relationship in the structural topology diagram, and extracts the multi-dimensional response features of the nodes in the verification tasks. For the verification tasks, a multi-dimensional physical constraint model integrating time dimension, electrical logic dimension and signal quality dimension is constructed.

[0027] The calculation module calculates the comprehensive verification index of the verification task based on the node response characteristics and the multidimensional physical constraint model, and obtains the confidence level of the connection of the central node according to the distribution of the verification index of all verification tasks of the central node.

[0028] The update module automatically locates and corrects the connection node to the corresponding connection node when the confidence level indicates a connection failure. This is achieved by establishing a temporary verification task between the central node and other nodes in the network and calculating the verification index, thus completing the adaptive update of the topology.

[0029] The beneficial effects of this invention are as follows: This invention monitors key moments of node state changes through an event-driven slice acquisition mechanism, reducing the invalid processing of massive steady-state data; by constructing a multi-dimensional physical constraint model that integrates time, electrical logic, and signal quality dimensions, it obtains multi-angle verification of the authenticity of node connections, enabling accurate differentiation between real physical causal connections and spurious correlations caused by noise; by calculating a comprehensive verification index and quantifying the confidence level of node connections, it obtains continuous dynamic monitoring of connection status, fundamentally solving the problem that static labels cannot track changes in physical connections; when connection loss of confidence is detected, it can automatically search all network nodes and locate the real connection point, completing adaptive updates of the topology and reducing misoperations and security risks caused by "discrepancies between the map and reality". Attached Figure Description

[0030] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0031] Figure 1 is a flowchart of a data acquisition and association method for a substation secondary circuit according to an embodiment of the present invention.

[0032] Figure 2 is a computer equipment diagram of a data acquisition and association method for a substation secondary circuit provided in an embodiment of the present invention.

[0033] Figure 3 is a schematic diagram of a partial topology of nodes in a data acquisition and association method for a substation secondary circuit according to an embodiment of the present invention. Detailed Implementation

[0034] To make the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0035] Example 1, referring to Figures 1-3, is an embodiment of the present invention. This embodiment provides a data acquisition and correlation method for secondary circuits in a substation, including:

[0036] S100: Establish a structural topology diagram to describe the connection relationship of secondary circuit nodes in the substation, generate data labels for each node containing input / output semantics, theoretical time delay windows and local connection information, establish an event-driven slice acquisition mechanism, and when a change in node status is detected, extract a time window based on the time of change to obtain an event slice containing dynamic response data of multiple nodes.

[0037] S200: Based on the theoretical connection relationship in the structural topology diagram, establish verification tasks for the central node in the event slice and its theoretical downstream nodes, and extract the multi-dimensional response features of the nodes in the verification tasks. For the verification tasks, construct a multi-dimensional physical constraint model that integrates the time dimension, electrical logic dimension and signal quality dimension.

[0038] S300: Based on the node response characteristics and physical constraint model, calculate the comprehensive verification index of the verification task, and obtain the confidence level of the connection of the central node according to the distribution of the verification index of all verification tasks of the central node.

[0039] S400: When the confidence level indicates a connection failure, the central node establishes a temporary verification task with other nodes in the network and calculates the verification index to automatically locate and correct the corresponding connection node, thus completing the adaptive update of the topology.

[0040] It should be noted that in actual operation, substation secondary circuits experience frequent changes in physical connections due to equipment modifications, maintenance, and temporary jumpers, leading to a continuous evolution of physical connections and the inability to dynamically verify static tags. Existing technologies construct one-time static models, lacking the ability to continuously verify the authenticity of node connections. This makes it impossible to detect changes in physical connections in a timely manner, resulting in long-term inconsistencies between tag information and actual connections. There is also the problem of passive diagnosis failing to proactively prevent issues. Existing technologies can only detect connection anomalies through manual inspection after a fault occurs, unable to proactively identify and warn of potential connection errors during operation. Finally, there is the problem of the inability to automatically repair problems after they are discovered. Even if connection anomalies are detected through certain means, existing technologies cannot automatically search for and locate the corresponding connection nodes, relying solely on manual on-site verification and correction, resulting in low automation, inefficiency, and a high risk of errors.

[0041] To address the aforementioned issues, the S100-S400 steps are implemented. S100 establishes a structural topology diagram and data labels, employing an event-driven mechanism to capture key data changes, achieving efficient data acquisition and reducing the processing of massive amounts of invalid data. S200 establishes verification tasks for theoretical connections based on the structural topology diagram, extracts multi-dimensional response features, and constructs a physical constraint model integrating time, logic, and quality dimensions. Through the strict AND relationship of the three-dimensional constraints, it can distinguish between real physical connections and spurious correlations caused by system disturbances. S300 calculates a comprehensive verification index and confidence level to obtain a quantitative assessment of connection authenticity. Upon detecting connection failures, S400 automatically searches all network nodes and locates the real connections, completing an adaptive update of the topology, thus achieving a process from passive diagnosis to proactive repair.

[0042] Example 2 is an embodiment of the present invention.

[0043] In this embodiment, the event-driven slice acquisition mechanism is established in step S100, including the following steps A1-A2:

[0044] A1: Deploy a data acquisition system to synchronize the time of all nodes on the site;

[0045] Specifically, a structural topology diagram is established to describe the connection relationship of secondary circuit nodes in the substation, and labels are generated for each node. The labels should at least include the node's input / output semantics, theoretical time delay window, and local topology diagram.

[0046] It should be noted that the SCD file is a standardized document defined according to DL / T243-2012. It is used to describe the instance configuration, communication parameters, communication relationships between devices, and primary system structure of IEDs (Intelligent Electronic Devices) in the substation. It is written by the system integration vendor and its uniqueness is guaranteed for the entire substation.

[0047] It should be noted that the label needs to carry enough information to support subsequent verification. The purpose of verification is to check whether the label is correct and whether the node is connected to the correct position. Determining the position requires combining the nodes before and after the node. Therefore, the node information and the local position of the node are used as the information stored in the node label.

[0048] Specifically, nodes and edges between nodes are extracted based on the SCD file. Nodes include all secondary equipment in the substation, such as protection devices, monitoring and control devices, intelligent terminals, and switches. Edges are obtained by connecting nodes that have relationships such as message sending / receiving or port correspondence. Nodes with edges are connected, with the direction of the edges pointing towards the data receiving node. If two nodes connected by an edge transmit data to each other, the direction of the edge points towards both nodes. The network formed after connecting all nodes is the structural topology diagram of the substation's secondary circuit.

[0049] Obtain node information for each node, including input / output semantics. Obtain the local topology graph for each node, and store the node information and local topology graph in the node's label. The label also includes the node's theoretical delay window, which is the estimated time window for the signal to reach the current node from the upstream node, based on the topology path and device manual. This theoretical delay window is denoted as... ,in, The value at the center of the window. Define the window width. Encode the labels into QR codes, print them, and affix them to the corresponding physical devices and cables on-site to complete the initial deployment. Figure 3 shows a schematic diagram of a partial node topology.

[0050] At this point, the structural topology diagram of the substation's secondary circuit has been established, and the labels for each node have been generated.

[0051] A2: Deploy a data acquisition system to synchronize the time of all nodes on the site, establish event triggering rules, and trigger data acquisition when the system detects any state change or value jump of a node. Take the trigger time as the center and capture the preset time window forward and backward to obtain the data of all nodes within the time period, forming an event slice.

[0052] Specifically, a data acquisition system is deployed within the substation; an event-driven triggering mechanism is established, whereby the data acquisition system operates in event-driven mode. When the data acquisition system detects a change in node status or a value jump, it records the time of the status change or value jump. As a baseline, the time window is truncated forward. Capture the time window backward Get time period The data from all nodes within the substation will be used to create a dataset that is recorded as an event slice of the substation.

[0053] In this embodiment, step S200 establishes a verification task for the central node and its theoretical downstream nodes in the event slice based on the theoretical connection relationship in the structural topology diagram, and extracts multi-dimensional response features of the nodes in the verification task. For the verification task, a multi-dimensional physical constraint model integrating time dimension, electrical logic dimension and signal quality dimension is constructed, including the following steps B1-B2:

[0054] B1: Establishing a verification task for the central node in an event slice and its theoretical downstream nodes includes taking the node that triggered the event in the event slice as the central node and obtaining the direct downstream nodes of the central node in the structural topology graph; the central node and each direct downstream node form a verification task.

[0055] Specifically, based on the real-time status changes of the substation's secondary circuits, event slices of the substation are collected; according to the structural topology of the nodes in the event slices, the verification tasks of the event slices are obtained; and verification feature vectors of the verification tasks are established.

[0056] It should be noted that data transmission and status changes in substations exhibit strict causal relationships. For example, power outages cause changes in circuit data; power outage is the cause, and the change in circuit data is the effect. In the connection between any node and its upstream and downstream nodes, the transmitted data has certain characteristics, such as constant voltage. Therefore, when the constant voltage changes, it can be inferred that there is a cause in the circuit leading to the voltage change. Analyzing the cause can determine whether the tag is still accurate. Therefore, this invention uses causal analysis to synchronously verify the data changes and tags of each node during the real-time operation of the substation.

[0057] It should be further explained that under the steady-state operation of a substation, a large number of nodes generate massive amounts of data. However, the effective information that can be directly used for causal verification is relatively sparse in this data. Since unchanging data means there is no effect, and therefore no cause, it is necessary to extract the effective information first, that is, the data segments with obvious data changes. For example, after a protection relay issues an action command, the circuit breaker receives the command and executes the action during the brief, dynamically changing time period. This is to avoid a large number of invalid verification behaviors that would cause a waste of resources.

[0058] Specifically, a data acquisition system is deployed within the substation. This system uses a precise time protocol to accurately synchronize the time of all secondary equipment in the substation. An event-driven triggering mechanism is established, and the data acquisition system operates in event-driven mode. When the data acquisition system detects changes in node status or value jumps, it records the time of the status change or value jump. As a baseline, the time window is truncated forward. Capture the time window backward Get time period The data from all nodes within the dataset will be recorded as an event slice.

[0059] It's important to note that precise time protocols, such as IEEE 1588, are existing standards for achieving high-precision time synchronization in distributed networks. Specifically, a master clock broadcasts a synchronization message to the entire network, and each slave clock device exchanges a series of timestamp messages with the master clock. This allows for precise calculation of line delays and the deviation of its own clock, enabling dynamic adjustments and achieving sub-millisecond synchronization accuracy. The time synchronization accuracy is better than 1 millisecond. The data integrity guarantee mechanism under network latency emphasizes that "the timestamps of all events must be added at the source of the data, i.e., at the front-end acquisition unit, not after the data is transmitted to the central server." Then, on the central processing server side, a sorting buffer queue based on event timestamps is established. All data packets uploaded from different acquisition points are not processed in arrival order after entering the queue, but are reordered according to their source timestamps. This eliminates the problem of out-of-order data packets caused by different network transmission paths and latency jitter, ensuring strict processing according to the true time sequence of events.

[0060] It should be noted that establishing an event-driven triggering mechanism, for example, involves different change detection methods for different data types. For analog signals, a fixed-length time window, such as 100ms, is defined, and the standard deviation of the data within the window is calculated in real time. When the standard deviation exceeds the trigger threshold, a significant change in value is considered to have occurred. For digital signals or state signals, due to mechanical contacts or signal interference, state signals may rapidly change back and forth within a short period. A "confirmation delay," such as 5ms, can be set. After a state change is detected, the system waits for this delay and reconfirms whether the state remains in the new state. Only after confirmation is it considered a valid event trigger, thus filtering out jitter noise. An example trigger threshold setting method could be to collect data from one hour of normal operation of the device and calculate the average value of the analog signal during that period. and standard deviation Then, the trigger threshold is dynamically set to The trigger threshold can be updated periodically to adapt to equipment aging or environmental changes. In addition, to avoid noise interference, associated confirmation rules can be set for critical times. For example, only when the two events "circuit breaker A trips" and "circuit A current drops to zero" are detected simultaneously within a short time window, such as 2ms, is it confirmed as a real circuit breaker tripping event.

[0061] It should be noted that, , These are preset values, determined by the implementers based on the theoretical time window. And the actual implementation settings.

[0062] First, determine based on the physical characteristics of signal propagation in the secondary circuit of the substation. The range of values; due to the event triggering time There may be millisecond-level detection latency, and the state changes of upstream nodes may have already begun before the detection is triggered. Therefore, Sufficient time needs to be traced back to capture the complete causal chain, for example, for a protection loop. It can be set to 50-100ms; for the measurement and control loop, It can be set to 100-200ms. During actual deployment, implementers can dynamically adjust it according to different voltage levels and equipment types. and The value; for critical protection circuits, it is recommended to use a shorter time window to improve response speed; for non-critical monitoring circuits, the time window can be appropriately extended.

[0063] At this point, the event slices of the substation have been obtained.

[0064] It should be noted that the structural topology graph provides the most direct prediction of causal paths. When an upstream event occurs, the focus should be on the response of its theoretical downstream nodes. Therefore, this step creates a verification task based on the structural topology graph, and then establishes the verification feature vector of the verification task node based on the changes in the data of the downstream nodes.

[0065] It should be further explained that the actual signal response process is not an ideal step, but a dynamic process with noise, oscillations, and delays. In order to extract stable and reliable response features, feature extraction of the original data is required. A complete data response process contains information in multiple dimensions, such as linear trends in the time domain, oscillation characteristics in the frequency domain, and details of non-stationary abrupt changes. Therefore, this invention combines the multi-dimensional features of node data to establish a verification feature vector for the verification task node.

[0066] Preferably, based on the structural topology of the nodes in the event slice, the verification task of the event slice is obtained; a verification feature vector is established for the verification task, which describes the data changes of the nodes from multiple dimensions, including:

[0067] Taking the node that triggers the event in the event slice as the central node, we obtain the direct downstream nodes of the central node in the structural topology graph. The central node and each direct downstream node form a verification task. For example, if A is the central node in Figure 3, then... , , , Four verification tasks for this event slice.

[0068] Establish the verification feature vector for the verification task, including: obtaining the direct downstream nodes of the central node. Then, the data within the preset time window is recorded as the reference data for the direct downstream nodes;

[0069] Let any verification task be denoted as ,Establish Verification feature vector:

[0070]

[0071] in, To verify the feature vector; To indicate The first verification feature; express The i-th verification feature, This indicates the total number of verification features.

[0072] The verification features must include at least: The slope, goodness of fit, and amplitude direction of the straight line obtained by linear fitting of the reference data of the direct downstream node of the corresponding central node are calculated. Wavelet packet decomposition is performed on the reference data to calculate the distribution entropy of its energy in different frequency bands, which is denoted as the wavelet capability entropy of the direct downstream node. The standard deviation of the data in the next preset time window of the reference data is calculated and denoted as the stable period standard deviation of the direct downstream node. This value quantifies the signal stability after the response process of the direct downstream node ends. An ideal response should eventually return to stability. It should be noted that linear fitting, wavelet packet decomposition, and distribution entropy calculation are existing technologies and will not be elaborated here. In addition, since different verification features have different importance and may have correlation and redundancy, multiple verification features can be extracted from historical data for principal component analysis. The principal components with the highest contribution rates are extracted as new feature vectors. This not only reduces dimensionality but also eliminates correlation interference.

[0073] At this point, the verification task for the event slice and the verification feature vector of the verification task have been obtained.

[0074] B2: Extracting multi-dimensional response features of nodes in the verification task includes obtaining the response data of the direct downstream nodes corresponding to the verification task after the event is triggered; extracting multi-dimensional response features, which include at least: the rate of change of response data, the linear fitting quality features, the amplitude change direction features, the frequency domain energy distribution features, and the fluctuation features of the response stabilization period.

[0075] Specifically, based on the response time range of the nodes in the verification task, a Gaussian function is used to establish a time constraint gate; based on the degree to which the verification feature vectors of the nodes in the verification task conform to electrical logic, a logic constraint gate is established; based on the stability of the node data, a quality constraint gate is established; and based on the data change rate of the nodes, the verification index of the verification task is obtained by combining the time constraint gate, the logic constraint gate, and the quality constraint gate.

[0076] It should be noted that a true causal connection must simultaneously satisfy multiple physical constraints such as time rationality, logical correctness, and signal quality. Therefore, this invention establishes a logical filter composed of multidimensional constraint gates combined by multiplication. Multiplication means that the three constraint dimensions are in an "AND" relationship. If any dimension fails to meet the conditions, the final total constraint gate will be closed, thus enabling accurate verification.

[0077] It should be noted that, firstly, there is the issue of time reasonableness. Verifying whether the response delay falls within a reasonable response time range cannot be done solely by distinguishing between 0 and 1, as measurement errors exist. A good model should award high scores to responses near the center of the response time range, lower but non-zero scores to responses at the edges of the response time range, and severe penalties to responses outside the response time range. The shape of the Gaussian function perfectly matches this penalty mechanism. Therefore, this invention, based on verifying the data changes and response time range of the task nodes, uses the Gaussian function to establish a time constraint gate.

[0078] B3: Construct a multi-dimensional physical constraint model that integrates time dimension constraints, electrical logic dimension constraints, and signal quality dimension constraints. The steps for time dimension constraints in the multi-dimensional physical constraint model include obtaining the actual response time of the direct downstream node corresponding to the verification task.

[0079] Based on the theoretical delay window in the node labels, a Gaussian function is used to establish a time constraint gate to quantitatively evaluate the deviation between the actual response time and the theoretical delay window. A high score is given when the response time falls within the theoretical window, and a penalty is imposed when it falls outside the window.

[0080] Specifically, based on the response time range of the nodes in the verification task, a Gaussian function is used to establish a time constraint gate, including:

[0081] Get The corresponding direct downstream node is The time corresponding to the extreme value within a time period is denoted as . The response time of the corresponding direct downstream node;

[0082] The time constraint gate satisfies the expression:

[0083] ;

[0084] In the formula, express Time-constrained gates; express The response time of the corresponding direct downstream node; express The center value of the theoretical time window corresponding to the direct downstream node. for The window width corresponding to the theoretical time window of the direct downstream node; This indicates the preset delay tolerance parameter; This represents the natural exponential function.

[0085] In the formula, This represents the distance by which the actual delay exceeds half the width of the theoretical time window. If the delay is within the theoretical time window, this value is negative. It controls the numerical range of the penalty function. The penalty term is positive only when the actual delay falls outside the theoretical time window; otherwise, it is 0. The severity of the latency penalty was controlled, when When a small value, such as 1ms, is set, the time constraint gate has a low tolerance for delay deviations. Even a slight deviation from the theoretical time window will result in a severe penalty. This is suitable for protection loops with extremely high timing requirements. When a larger value, such as 3ms, is set, the time constraint gate has a higher tolerance for time delay deviations. It is believed that even if the theoretical time window is exceeded, there is still a certain degree of reliability. It is suitable for monitoring or signal loops with relatively relaxed timing requirements.

[0086] It should be noted that it is also necessary to verify whether the data direction and strength of the node response conform to the electrical logic. For example, a trip signal should cause a drop in current, i.e., the response slope should be negative.

[0087] B4: Construct a multi-dimensional physical constraint model that integrates time dimension constraints, electrical logic dimension constraints, and signal quality dimension constraints. The steps for the electrical logic dimension constraints in the multi-dimensional physical constraint model include predicting the expected direction and trend of node response changes based on the input and output semantics in the node labels.

[0088] Establish logical constraint gates and quantify the logical compliance of the response by comparing the consistency between the actual response characteristics and the expected response characteristics.

[0089] Specifically, based on the degree to which the verification feature vectors of the nodes in the verification task conform to the electrical logic, logic constraint gates are established, including:

[0090] based on Obtain the input / output semantics of the labels corresponding to the direct downstream nodes. The predicted slope and predicted amplitude direction of the reference data corresponding to the direct downstream node; it should be noted that, for example, the state quantity of the trip output of the logic node of the protection device changes from 0 to 1.

[0091] Logic constraint gates satisfy the expression:

[0092] ;

[0093] In the formula, express Logical constraint gates; , express The slope and magnitude direction of the straight line obtained by linear fitting the reference data of the direct downstream node; , express The predicted slope and predicted amplitude direction of the reference data corresponding to the direct downstream nodes; This represents the normalization function.

[0094] In the formula, This value is positive when the slope of the line is in the same direction as the prediction, and negative when it is not. This indicates that the value is positive when the amplitude direction is consistent with the prediction, and negative when they are inconsistent. If the slope direction and amplitude direction are both consistent with the predicted situation, then the larger the value, the better. The closer the value of the logic constraint gate is to 1.

[0095] It should be noted that the quality of the data itself is also extremely important. Responses caused by real physical connections should have relatively clean and stable waveforms. Conversely, spurious responses caused by noise or interference typically exhibit poor linearity, strong non-stationarity, and instability in later stages. Therefore, this invention establishes a quality constraint gate based on the stability of node data.

[0096] B5: Construct a multi-dimensional physical constraint model that integrates time dimension constraints, electrical logic dimension constraints, and signal quality dimension constraints. The steps for signal quality dimension constraints in the multi-dimensional physical constraint model include: establishing a quality constraint gate, and comprehensively evaluating the linear fit goodness of the response data, the regularity of the frequency domain energy distribution, and the stability after the response.

[0097] Response signals with high linear fit and concentrated frequency domain energy distribution are selected by mass constraint gate.

[0098] Specifically, based on the stability of the node data, a quality constraint gate is established to satisfy the expression:

[0099] ;

[0100] In the formula, express Quality constraint gate; express The goodness of fit is obtained by linear fitting of the reference data corresponding to the direct downstream nodes; express The wavelet capability entropy corresponding to the direct downstream node; express The standard deviation of the steady-state period corresponding to the direct downstream node; This represents the natural exponential function.

[0101] In the formula, the greater the goodness of fit, the higher the linearity, and therefore the higher the base quality score; at the same time, The larger, The larger the value, the more chaotic the signal, the more mutations, and the more violently the signal continues to fluctuate after the response ends. The lower the quality.

[0102] It should be noted that combining time constraint gates, logic constraint gates, and quality constraint gates yields a verification score describing the rationality of the verification task. Furthermore, considering the response triggered by a strong causal connection, such as a circuit breaker tripping directly from the main output, its rate of change is typically extremely fast. Conversely, a weak correlation or indirect influence will have a much slower response rate; for example, a change in a remote signaling quantity might indirectly trigger a fine-tuning of another quantity through background logic. Therefore, the rate of change of the response is the most crucial and intuitive quantitative indicator for measuring the strength of causal relationships. Thus, this invention combines time constraint gates, logic constraint gates, and quality constraint gates on top of response speed to obtain a verification index for the verification task.

[0103] It should be noted that, considering the diverse equipment types, varying circuit characteristics, and variable operating conditions in substations, to ensure the universality and high accuracy of the method in different scenarios, key parameters in the time, logic, and quality constraint gates, such as delay tolerance, prediction slope, and prediction amplitude direction, are not fixed but dynamically learned and optimized through an adaptive adjustment mechanism. Specifically, for parameter initialization, during the initial equipment commissioning phase, the system enters a short-term learning mode, rapidly collecting event slice data and calculating features through multiple known, manually executed correct operations such as manual opening and closing, thereby assigning a reliable initial value to the constraint gate parameters of that node; for parameter learning based on historical data... Online optimization allows the system to periodically, for example weekly, process all manually verified correct event data from the past week as a batch for offline statistical analysis. This enables a global optimization and calibration of the constraint gate parameters of all nodes in the entire secondary circuit network to correct long-term accumulated deviations. For multi-scenario parameter configuration strategies, since the behavior characteristics of the same equipment may differ significantly under different operating conditions such as normal operation, annual maintenance, and live testing, multiple parameter sets can be established for the same node. By receiving the operating mode flag from the SCADA system or through internal logic, such as detecting that the "grounding switch" is closed, the system can dynamically switch to the parameter set that matches the current operating condition, thereby achieving higher-precision scenario-based verification.

[0104] In this embodiment, the calculation of the comprehensive verification index of the verification task in step S300 includes the following steps C1-C2:

[0105] C1: The rate of change of the node response is multiplied and combined with the evaluation results of the time constraint gate, logic constraint gate, and quality constraint gate to obtain the comprehensive verification index of the verification task.

[0106] Specifically, will The absolute value of the slope of the straight line obtained by linear fitting to the reference data of the direct downstream node, and... Time constraint gates Logical constraint gates Multiplying the mass constraint gates, denoted as... The verification index.

[0107] Based on the data change rate of the nodes, and combined with time constraint gates, logical constraint gates, and quality constraint gates, the verification index of the verification task is obtained, including:

[0108] The verification exponent of any verification task satisfies the expression:

[0109] ;

[0110] In the formula, express The verification index; express The slope of the straight line obtained by linear fitting the reference data corresponding to the direct downstream node; express Time-constrained gates; express Logical constraint gates; express The mass constraint gate; represents the absolute value function.

[0111] At this point, the verification index for each verification task has been obtained.

[0112] C2: The multiplication relationship ensures that only connections that are time-reasonable, logically correct, and of reliable quality can achieve a high check index.

[0113] Specifically, based on the overall value of the verification index of each verification task in the event slice, the confidence level of the central node of the event slice is obtained; based on the confidence level range, the label of the node is corrected, including: forming a temporary verification task between the central node and each non-direct downstream node, calculating the verification index of each temporary verification task, and obtaining the corrected downstream node of the central node based on the value range of the verification index of the temporary verification task.

[0114] Specifically, the mean of the verification indices of all verification tasks in an event slice is obtained and positively correlated and normalized to obtain the confidence level of the central node of the event slice. A first threshold is set. When the confidence level is greater than the first threshold, the label of the central node of the event slice is recorded as a trusted label; when the confidence level is less than or equal to the first threshold, the label of the central node of the event slice is recorded as a untrustworthy label. It should be noted that the first threshold is a preset value, which is set by the implementers according to the actual implementation situation. For example, the first threshold can be set to 0.5.

[0115] It should be noted that when a central node's label is determined to be untrustworthy, the most likely reason is a change in its outgoing connection; that is, it is no longer connected to a theoretical downstream node, but to some other unknown node. Therefore, the key to correction lies in identifying the corrected downstream node of the central node in the structural topology diagram.

[0116] Preferably, the label of the central node is corrected, including:

[0117] The central node is paired with each non-direct downstream node to form a temporary verification task. The verification index of each temporary verification task is calculated. Based on the numerical range of the verification index of the temporary verification task, the corrected downstream nodes of the central node are obtained.

[0118] It should be noted that, based on the numerical range of the verification index of the temporary verification tasks, the corrected downstream nodes of the central node are obtained. For example, k-means clustering can be performed on the verification indices of all temporary verification tasks to obtain two clusters of temporary verification tasks. The node corresponding to the temporary verification task in the cluster with the largest verification index is recorded as the corrected downstream node of the central node. Alternatively, a second threshold can be set, and temporary verification tasks with verification indices greater than the second threshold can be recorded as the corrected downstream nodes of the central node.

[0119] The corrected downstream nodes of the central node are used as the new labels to complete the label correction of the nodes.

[0120] This completes the verification and correction of the node labels.

[0121] In this embodiment, the automatic positioning and correction of the corresponding connection node in step S400 includes the following steps D1-D2:

[0122] D1: Establish temporary verification tasks between the central node and the non-direct downstream nodes in the structural topology diagram, and calculate the comprehensive verification index of each temporary verification task.

[0123] Specifically, when the confidence level of the central node falls below the first threshold and is identified as a defaulter, the system initiates an automatic correction process. First, it acquires all other nodes in the structural topology graph, excluding the direct downstream nodes of the central node, and records them as a candidate node set. For each node in the candidate node set, a temporary verification task is established between it and the central node. The establishment of the temporary verification task is exactly the same as the verification task for theoretical connections. The response data of the candidate node after the event is triggered is extracted, and its multi-dimensional response feature vector is calculated, including the rate of change feature, linear fitting quality feature, amplitude change direction feature, frequency domain energy distribution feature, and fluctuation feature during the response stabilization period.

[0124] Next, for each temporary verification task, its time constraint gate, logic constraint gate, and quality constraint gate are calculated. The time constraint gate evaluates the degree to which the actual response time of the candidate node matches the theoretical delay window preset in its label; the logic constraint gate evaluates the consistency between the response characteristics of the candidate node and its input-output semantics; and the quality constraint gate evaluates the stability and purity of the candidate node's response signal. The response change rate of the candidate node is multiplied and combined with the three constraint gates to obtain the comprehensive verification index of the temporary verification task. By traversing all candidate nodes, a set containing all temporary verification tasks and their corresponding verification indices is obtained.

[0125] D2: Based on the distribution of the verification index of the temporary verification task, identify nodes whose verification index is better than the theoretical connection, determine them as the corrected downstream nodes of the central node, and update the data labels of the nodes.

[0126] Specifically, a statistical analysis is performed on the verification indices of all temporary verification tasks. First, the average verification index of the verification tasks of the central node and its theoretically direct downstream nodes is calculated and denoted as the baseline index. Then, nodes with verification indices higher than the baseline index are selected from all temporary verification tasks. A clustering method can be used to perform k-means clustering on the verification indices of all temporary verification tasks, resulting in high-scoring and low-scoring clusters. The nodes with the highest verification indices in the high-scoring clusters are the most likely true downstream connection nodes of the central node. Alternatively, a second threshold can be set, identifying nodes corresponding to temporary verification tasks with verification indices greater than the second threshold and significantly better than the baseline index as corrected downstream nodes. The second threshold is determined by statistically analyzing the verification indices of all verification tasks of the central node and its theoretically direct downstream nodes, calculating their mean and standard deviation, expressed as:

[0127]

[0128] in, The second threshold; The mean; Standard deviation; This is the sensitivity coefficient, typically ranging from 1.5 to 3.0. When... When a smaller value, such as 1.5, is used, the system has higher sensitivity to detecting potential real connections, making it suitable for scenarios with frequent connection changes; when... When a larger value, such as 3.0, is used, the system has stricter requirements for the verification results, which can reduce the false positive rate and is suitable for critical loops with extremely high reliability requirements.

[0129] Based on historical data, the comprehensive verification index of genuine physical connections is typically above 0.7, while the verification index of weak or accidental correlations is usually below 0.4. Therefore, a fixed value between 0.6 and 0.7 can be directly set for the second threshold. For example, in a practical application at a substation, by analyzing 100 known connection change events, the average verification index of genuine connections was found to be 0.82, while the average verification index of false correlations was 0.35. Based on this, a second threshold of 0.65 can be set, which can maintain an accuracy rate of over 95% while controlling the false positive rate below 5%.

[0130] After identifying the corrected downstream nodes, the data labels of the central node are automatically updated. The direct downstream nodes in its local topology graph are modified to the newly discovered real nodes. Simultaneously, the corresponding edge connections are updated in the structural topology graph, deleting invalid theoretical connections and adding new real connections. The updated label information can be re-encoded into QR codes to prompt maintenance personnel for on-site verification and physical label replacement, thus completing closed-loop management from digital model to physical labels.

[0131] In summary, the event-driven slicing acquisition mechanism achieves high precision and efficiency in data acquisition; by deploying a data acquisition system based on a precise time protocol, sub-millisecond time synchronization of all secondary devices across the entire station is achieved, ensuring time consistency of data output from different nodes; by establishing event triggering rules for different data types, including standard deviation jump detection for analog quantities and confirmation delay mechanisms for digital quantities, each event slice can be output as containing real and valid causal information; by extracting multi-dimensional response features, including rate of change, linear fitting quality, amplitude direction, frequency domain energy distribution, and steady-state fluctuations, the dynamic process of node response is displayed to maintenance personnel, providing input information for multi-dimensional physical constraint models.

[0132] Example 3, referring to Figures 2 and 3, is an embodiment of the present invention. This embodiment provides a data acquisition and association system for the secondary circuit of a substation, including a construction module, which establishes a structural topology diagram to describe the connection relationship of nodes in the secondary circuit of the substation, generates data labels for each node containing input and output semantics, theoretical time delay windows and local connection information, and establishes an event-driven slice acquisition mechanism. When a change in the node state is detected, a time window is extracted based on the time of change to obtain an event slice containing dynamic response data of multiple nodes.

[0133] The extraction module establishes verification tasks for the central node and its theoretical downstream nodes in the event slice based on the theoretical connection relationship in the structural topology diagram, and extracts the multi-dimensional response features of the nodes in the verification tasks. For the verification tasks, a multi-dimensional physical constraint model integrating time dimension, electrical logic dimension and signal quality dimension is constructed.

[0134] The calculation module calculates the comprehensive verification index of the verification task based on the node response characteristics and physical constraint model, and obtains the confidence level of the connection of the central node according to the distribution of the verification index of all verification tasks of the central node.

[0135] The update module automatically locates and corrects the connection node to the corresponding connection node when the confidence level indicates a connection failure. This is achieved by establishing a temporary verification task between the central node and other nodes in the network and calculating the verification index, thus completing the adaptive update of the topology.

[0136] This embodiment also provides an electronic device applicable to a data acquisition and association method for a substation secondary circuit, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the data acquisition and association method for a substation secondary circuit as proposed in the above embodiment.

[0137] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, it implements a data acquisition and association method for a substation secondary circuit as proposed in the above embodiment.

[0138] The storage medium proposed in this embodiment and the data acquisition and association method for a substation secondary circuit proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0139] Based on the above description of the implementation methods, those skilled in the art can clearly understand that the present invention can be implemented using software and necessary general-purpose hardware, and of course, it can also be implemented using hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory, hard disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of the various embodiments of the present invention.

[0140] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A method for data acquisition and correlation of secondary circuits in a substation, characterized in that: This includes establishing a structural topology diagram to describe the connection relationships of secondary circuit nodes in a substation; generating data labels for each node that include input / output semantics, theoretical time delay windows, and local connection information; establishing an event-driven slice acquisition mechanism, which, when a change in node state is detected, extracts a time window based on the time of the change to obtain an event slice containing dynamic response data from multiple nodes; the establishment of the event-driven slice acquisition mechanism includes deploying a data acquisition system to synchronize the time of all nodes in the substation; and establishing event triggering rules, which trigger data acquisition when the system detects any state change or numerical jump in a node, extracting slices forward and backward from the trigger time. A preset time window is used to acquire data from all nodes within the time period, forming an event slice. Based on the theoretical connection relationships in the structural topology graph, a verification task is established between the central node in the event slice and its theoretical downstream nodes. Multi-dimensional response features of the nodes in the verification task are extracted. For the verification task, a multi-dimensional physical constraint model with integrated constraints is constructed, including time dimension constraints, electrical logic dimension constraints, and signal quality dimension constraints. Establishing a verification task between the central node in the event slice and its theoretical downstream nodes includes taking the node that triggers the event in the event slice as the central node and acquiring the direct downstream nodes of that central node in the structural topology graph. The central node and each direct downstream node form a verification task; The extraction of multi-dimensional response features of nodes in the verification task includes: obtaining the response data of the direct downstream nodes corresponding to the verification task after the event is triggered; extracting multi-dimensional response features, which at least include: the rate of change of response data, linear fitting quality features, amplitude change direction features, frequency domain energy distribution features, and fluctuation features during the stable period of the response; calculating the comprehensive verification index of the verification task based on the node response characteristics and the multi-dimensional physical constraint model, and obtaining the confidence level of the connection of the central node according to the distribution of the verification index of all verification tasks of the central node; when the confidence level indicates connection failure, automatically locating and correcting the central node to the corresponding connection node by establishing a temporary verification task with other nodes in the network and calculating the verification index, thus completing the adaptive update of the topology.

2. The data acquisition and correlation method for a substation secondary circuit as described in claim 1, characterized in that: The time dimension constraint includes obtaining the actual response time of the direct downstream node corresponding to the verification task; Based on the theoretical delay window in the node labels, a Gaussian function is used to establish a time constraint gate to quantitatively evaluate the deviation between the actual response time and the theoretical delay window. A high score is given when the response time falls within the theoretical window, and a penalty is imposed when it falls outside the window.

3. The data acquisition and correlation method for a substation secondary circuit as described in claim 2, characterized in that: The electrical logic dimension constraints include predicting the expected direction and trend of node response changes based on the input-output semantics in the node labels; establishing a logic constraint gate to quantify the logical compliance of the response by comparing the consistency between the actual response characteristics and the expected response characteristics.

4. The data acquisition and correlation method for a substation secondary circuit as described in claim 3, characterized in that: The signal quality dimension constraints include establishing a quality constraint gate to comprehensively evaluate the linear fit goodness of the response data, the regularity of the frequency domain energy distribution, and the stability after the response; and using the quality constraint gate to screen response signals with good linear fit and concentrated frequency domain energy distribution.

5. The data acquisition and correlation method for a substation secondary circuit as described in claim 4, characterized in that: The calculation of the comprehensive verification index of the verification task includes multiplying and combining the change rate of the node response with the evaluation results of the time constraint gate, logic constraint gate, and quality constraint gate to obtain the comprehensive verification index of the verification task; the multiplication relationship ensures that only connections that simultaneously meet the requirements of reasonable time, correct logic, and reliable quality can obtain a high verification index.

6. The data acquisition and correlation method for a substation secondary circuit as described in claim 5, characterized in that: The automatic location and correction to the corresponding connection node includes: establishing temporary verification tasks between the central node and non-direct downstream nodes in the structural topology diagram; calculating the comprehensive verification index of each temporary verification task; identifying nodes with verification indices better than the theoretical connection based on the verification index distribution of the temporary verification tasks; determining them as the corrected downstream nodes of the central node; and updating the node's data label.

7. A data acquisition and association system for substation secondary circuits, employing the data acquisition and association method for substation secondary circuits as described in any one of claims 1 to 6, characterized in that, include: The construction module establishes a structural topology diagram describing the connection relationships of secondary circuit nodes in a substation. It generates data labels for each node, including input / output semantics, theoretical time delay windows, and local connection information. It also establishes an event-driven slice acquisition mechanism. When a change in node status is detected, a time window is extracted based on the time of change to obtain an event slice containing dynamic response data of multiple nodes. The extraction module establishes verification tasks for the central node and its theoretical downstream nodes in the event slice based on the theoretical connection relationships in the structural topology diagram. It extracts multi-dimensional response features of nodes in the verification tasks and constructs a multi-dimensional physical constraint model that integrates time, electrical logic, and signal quality dimensions for the verification tasks. The calculation module calculates the comprehensive verification index of the verification task based on the node response characteristics and the multidimensional physical constraint model, and obtains the confidence level of the connection of the central node according to the distribution of the verification index of all verification tasks of the central node; the update module automatically locates and corrects the central node to the corresponding connection node by establishing a temporary verification task with other nodes in the network and calculating the verification index when the confidence level indicates a connection failure, thus completing the adaptive update of the topology.

Images