High voltage switch cabinet with fault monitoring remote fast handling function

Abstract

The application relates to the technical field of high-voltage switch cabinet fault processing, in particular to a high-voltage switch cabinet with a fault monitoring and remote rapid processing function, which comprises a data acquisition module for acquiring data in the high-voltage switch cabinet; various physical paths of the high-voltage switch cabinet are set, and each physical path comprises different nodes; a path transmission mismatch condition analysis module is arranged for acquiring equivalent state quantities of each node on each physical path, and then obtaining transmission mismatch intensities of the physical paths; a path fault restriction analysis module is arranged for acquiring cross-path induction quantities and fault restriction coefficients of the physical paths; a fault processing module is arranged for acquiring residual path mismatch vectors after execution of each action of the high-voltage switch cabinet; then, correction benefits of each action are acquired, and the action to be executed is selected according to the correction benefits to process the fault. The application can effectively process the high-voltage switch cabinet fault process.

Description

Technical Field

[0001] This invention relates to the field of high-voltage switchgear fault handling technology, and specifically to a high-voltage switchgear with remote rapid fault monitoring and handling functions. Background Technology

[0002] High-voltage switchgear is a crucial electrical device in power distribution systems used for power distribution, line control, fault isolation, and equipment protection. Its interior typically includes functional compartments such as busbar compartments, circuit breaker compartments, cable compartments, instrument compartments, handcart compartments, and operating mechanism compartments. Because high-voltage switchgear operates under conditions of high voltage, high current, enclosed spaces, and periodic load fluctuations, faults such as overheating, insulation dampness, surface discharge, arc flashover, mechanism jamming, and abnormal positioning are prone to occur in areas such as cable terminals, busbar connection points, contact boxes, insulators, circuit breaker contacts, handcart mechanisms, and grounding switch mechanisms.

[0003] Existing high-voltage switchgear typically includes modules for temperature, current, voltage, humidity, partial discharge, arcing, smoke monitoring, and remote communication to collect operational status data and trigger remote alarms. However, most current technologies rely on a "sensor acquisition—value exceeding limits—remote alarm—manual judgment" model. While this approach can detect some anomalies, its judgments are often based on single-point thresholds or simple trends, failing to reflect the propagation of faults along conductive paths, insulation creepage paths, heat diffusion paths, and mechanical execution paths within the switchgear. It also struggles to determine whether the current fault can be quickly addressed through remote cooling, dehumidification, load reduction, interlocking, power transfer, or circuit breaker tripping. Therefore, existing remote monitoring systems often only indicate faults and cannot directly formulate safe, interpretable, and executable remote rapid response strategies. Summary of the Invention

[0004] To address the aforementioned technical problems, the present invention aims to provide a high-voltage switchgear with remote and rapid fault monitoring and handling functions. The specific technical application adopted is as follows:

[0005] One embodiment of the present invention provides a high-voltage switchgear with remote rapid fault monitoring and handling function, comprising:

[0006] The data acquisition module is used to collect visible light images, infrared thermal images, electrical operation data, insulation environment data, and actuator status data inside the high-voltage switchgear; and to set various physical paths for the high-voltage switchgear, with different nodes on each physical path.

[0007] The path transmission mismatch analysis module is used to obtain the equivalent state quantity of each node on various physical paths using visible light images, infrared thermal images, electrical operation data, insulation environment data, and actuator status data; and to obtain the transmission mismatch intensity of a physical path based on the changes in the equivalent state quantity of each node on the physical path.

[0008] The path fault constraint analysis module is used to set the structural coupling coefficient between every two physical paths, obtain the cross-path induction of a physical path based on the structural coupling coefficient between a physical path and other physical paths and the transmission mismatch strength of each physical path, and obtain the fault constraint coefficient based on the transmission mismatch strength and cross-path induction of a physical path.

[0009] The fault handling module is used to construct an action influence matrix for each action based on the remote actions that can be performed by the high-voltage switchgear; obtain the remaining path mismatch vector after the execution of the action by using the transmission mismatch strength of all paths and the action influence matrix of the action; obtain the correction benefit of the action by using the transmission mismatch strength of all paths and the remaining path mismatch vector after the execution of the action; and select the action to be executed to handle the fault based on the correction benefit.

[0010] Preferably, the various physical paths of the high-voltage switchgear are set, including:

[0011] The various physical paths include conductive paths, creepage paths, thermal diffusion paths, and mechanical execution paths.

[0012] Preferably, the equivalent state quantities of each node on various physical paths are obtained using visible light images, infrared thermal images, electrical operation data, insulation environment data, and actuator state data, including:

[0013] For a node in a conductive path, the temperature is extracted pixel by pixel along the conductor skeleton line corresponding to the node in the registered infrared image, and the temperature is corrected using the material emissivity to obtain the temperature corresponding to each pixel. The maximum value between the temperature of a pixel and the ambient temperature and 0 is obtained as the temperature rise of the pixel. The actual physical length of each pixel is obtained, and the product of the temperature rise of each pixel and the actual physical length is summed to obtain the conductive node feature. The conductive node feature is divided by the sum of the current of the main loop corresponding to the conductive path and the constant term to obtain the equivalent state quantity of the node in the conductive path.

[0014] For a node in a creepage path, the visible light image is segmented to obtain the wetted region. The wetted region is projected onto the surface of the corresponding insulating component. The length of the maximum wetted connected region along the creepage direction on the surface of the corresponding insulating component is calculated as the ratio of the length of the maximum wetted connected region along the creepage direction to the total creepage length on the surface of the corresponding insulating component, thus obtaining the wetted continuity of the insulating surface. The mean values ​​of the TEV amplitude and the mean values ​​of the ultrasonic amplitude collected by the transient ground voltage sensor and the ultrasonic partial discharge sensor within the detection range of the node within a preset time period are normalized and then weighted and summed to obtain the partial discharge activity. The average temperature of the surface of the corresponding insulating component is extracted using an infrared thermal image. The dew point temperature is calculated using the Magnus formula based on the temperature and humidity sensor data inside the cabinet. The difference between the average temperature and the dew point temperature of the surface of the corresponding insulating component is calculated to obtain the dew point margin. The product of the wetted continuity of the insulating surface and the partial discharge activity is calculated and divided by the sum of the dew point margin and a constant term to obtain the equivalent state quantity of the node in the creepage path.

[0015] For a node in a heat diffusion path, in the registered infrared thermal image, the structural boundary corresponding to the node is selected, and two adjacent regions of interest are delineated on both sides of the boundary. The average temperature of the two regions of interest is calculated, and the larger value is recorded as the average temperature on the hot side and the smaller value is recorded as the average temperature on the cold side. The difference between the average temperature on the hot side and the average temperature on the cold side is recorded as the hot-cold temperature difference. The equivalent state quantity of the node is obtained by dividing the hot-cold temperature difference corresponding to the node by the sum of the reference hot-cold temperature difference and the constant term.

[0016] For a node in an organization's execution path, the time from the completion of the action of the previous node to the change of the state of this node is determined by the event sequence record and recorded as the actual response time of this node. The absolute value of the difference between the actual response time and the standard response time of this node is calculated and divided by the standard response time and a constant term of this node to obtain the equivalent state quantity of this node in the organization's execution path.

[0017] Preferably, obtaining the transmission mismatch intensity of a physical path based on the changes in the equivalent state variables of each node along the physical path includes:

[0018] Obtain the transmission coefficient between every two adjacent nodes under fault-free conditions; obtain the transmission mismatch residual between two adjacent nodes by subtracting the equivalent state quantity of the previous node from the equivalent state quantity of the next node in a physical path and the product of the transmission coefficient between the two adjacent nodes; calculate the average of the absolute values ​​of the transmission mismatch residuals between every two adjacent nodes in the physical path, denoted as the first average value; calculate the average of the product of the transmission mismatch residuals between every two adjacent nodes in the physical path and the path direction sign, and take the absolute value to obtain the second average value; calculate the sum of the first average value and the second average value to obtain the transmission mismatch intensity of the physical path.

[0019] Preferably, obtaining the cross-path induction of a physical path based on the structural coupling coefficient between a physical path and other physical paths and the transmission mismatch strength of each physical path includes:

[0020] The cross-path induction of a physical path is obtained by weighting the transmission mismatch strength of the other physical paths with the structural coupling coefficient between a physical path and other physical paths, and then summing the weighted sum of the transmission mismatch strength of the physical path and the constant term.

[0021] Preferably, the fault constraint coefficient is obtained based on the transmission mismatch intensity and cross-path induction of a physical path, including:

[0022] The fault limitation coefficient of a physical path is obtained by dividing the transmission mismatch intensity of a physical path by a first preset value and the sum of the cross-path inductance of the physical path.

[0023] Preferably, the remaining path mismatch vector after the execution of an action is obtained by utilizing the transmission mismatch intensity of all paths and the action influence matrix of an action, including:

[0024] The current transmission mismatch strength vector is composed of the transmission mismatch strengths of all paths; then, the remaining path mismatch vector after the execution of an action is specifically:

[0025] ,

[0026] in, Indicates action The remaining path mismatch vector after the execution of this action, where k is the index of the action; Indicates action Action influence matrix; Represents the identity matrix; This represents the current propagation mismatch strength vector.

[0027] Preferably, the corrective gain for an action is obtained by utilizing the transmission mismatch intensity of all paths and the remaining path mismatch vector after the action is executed, including:

[0028] The correction gain of an action is obtained by comparing the difference between the sum of the absolute values ​​of the transmission mismatch intensity of each physical path and the sum of the absolute values ​​of each element in the remaining path mismatch vector after the execution of an action, and the sum of the sum of the absolute values ​​of the transmission mismatch intensity of each path and the constant term.

[0029] Preferably, the action to be performed to handle the fault is selected based on the corrective benefits, including:

[0030] Select the action that yields the greatest benefit to address the current fault.

[0031] The embodiments of the present invention have at least the following beneficial effects: This application collects visible light images, infrared thermal images, electrical operation data, insulation environment data and actuator status data inside the high-voltage switchgear, and then sets various physical paths of the high-voltage switchgear, each physical path including different nodes;

[0032] Furthermore, the equivalent state quantities of each node on various physical paths are obtained by using visible light images, infrared thermal images, electrical operation data, insulation environment data, and actuator status data. The equivalent state quantities are not simply temperature, humidity, area, or distance, but scenario-based quantities constructed by combining path type and fault mechanism. The transmission mismatch intensity of a physical path is obtained by observing the changes in the equivalent state quantities of each node on the physical path. This not only determines whether the physical path is abnormal, but also whether the abnormality has a tendency to propagate along the path.

[0033] Next, based on the structural coupling coefficient between a physical path and other physical paths and the transmission mismatch strength of each physical path, the cross-path induction of the physical path is obtained, and the fault is upgraded from "single path anomaly" to "inter-path propagation relationship judgment"; then the fault constraint coefficient is obtained, which can distinguish two types of situations that are difficult to distinguish by traditional methods: high-intensity but constrained faults.

[0034] Finally, the remaining path mismatch vector after the execution of an action is obtained by using the transmission mismatch intensity of all paths and the action influence matrix of an action; the correction benefit of an action is obtained by using the transmission mismatch intensity of all paths and the remaining path mismatch vector after the execution of an action; and the action to be executed is selected to handle the fault based on the correction benefit. This processing method improves the remote action selection from "fault type - fixed action" to "current path mismatch status - action matrix correction capability - action optimization", thereby forming an effective monitoring and effective handling of faults in high-voltage switchgear. Attached Figure Description

[0035] To more clearly illustrate the technical applications and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the 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.

[0036] Figure 1 This is a module block diagram of a high-voltage switchgear with remote rapid fault monitoring function provided in an embodiment of the present invention. Detailed Implementation

[0037] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a high-voltage switchgear with remote rapid fault monitoring function proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0038] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0039] The following description, in conjunction with the accompanying drawings, details the specific application of the present invention for a high-voltage switchgear with remote rapid fault monitoring and processing function.

[0040] Example:

[0041] The main application scenario of this invention is as follows: This application mainly obtains various physical paths in high-voltage switchgear, and then analyzes, monitors and processes faults on each physical path, thereby achieving the purpose of remote monitoring and rapid processing of faults in high-voltage switchgear.

[0042] Please see Figure 1 The diagram illustrates a module block diagram of a high-voltage switchgear with remote rapid fault monitoring function provided by an embodiment of the present invention, including the following modules:

[0043] The data acquisition module is used to collect visible light images, infrared thermal images, electrical operation data, insulation environment data, and actuator status data inside the high-voltage switchgear; and to set various physical paths for the high-voltage switchgear, with different nodes on each physical path.

[0044] High-voltage switchgear operates under complex conditions, and a single sensor cannot fully represent the fault status. Therefore, this application requires the collection of the following multi-source data: Visible light image acquisition units are installed in the busbar compartment, cable compartment, circuit breaker compartment, mechanism compartment, and observation window to acquire visible light images. These images are used to identify information such as condensation, moisture, creepage marks, the position of mechanism indicators, the position of the handcart, the position of the grounding switch, and the opening and closing positions of the circuit breaker. Infrared thermal imaging units are installed near cable terminals, busbar connection points, contact boxes, circuit breaker moving and stationary contacts, insulators, wall bushings, and current transformer connection terminals to collect infrared thermal images of the high-voltage switchgear. These images are used to describe conductors, connection points, insulators, and compartment boundaries. The system first collects thermal data; then, it collects electrical operation data, including three-phase current, circuit breaker opening and closing status, energy storage status, remote / local status, trolley working position, trolley test position, and grounding switch status; next, it collects insulation environment data, including cabinet internal temperature, humidity, ultrasonic partial discharge, and transient ground voltage partial discharge, which are used to describe insulation status and discharge activity; finally, it collects actuator status data of the high-voltage switchgear, including the status of cooling fans, dehumidifiers, low-power heaters, electric operating mechanisms, interlocking relays, communication links, and control power supplies.

[0045] Visible light images, infrared thermal images, electrical operation data, insulation environment data, and actuator status data are synchronized using a unified timestamp. If different devices have different sampling frequencies, the image acquisition time is used as the master time point, and sensor data within adjacent time ranges are mapped to that image frame. The system provides unified time synchronization for all acquisition units via NTP or GPS to ensure that all types of data have consistent time tags. The acquisition time of the visible light or infrared image is used as the alignment reference. For other rapidly changing data, the effective value is calculated from the sampled values ​​within the half-cycle window before and after the reference time as the mapping value for that time. For slowly changing data such as temperature, humidity, and partial discharge, the most recent sampling result at the reference time is directly used. For switch quantities such as actuator status, the last updated state before the reference time is used. If there is no hard trigger synchronization between the visible light and infrared cameras, the adjacent frame with the smallest time difference from the reference time is selected for alignment. After the above calculations, all multi-source data are unified to the same time segment, forming a logically consistent data group for subsequent processing.

[0046] The visible light images undergo distortion correction, brightness equalization, reflection suppression, noise filtering, and edge enhancement. This processing reduces the impact of cabinet lighting, metal reflections, and camera noise on subsequent path projection. The infrared thermal images undergo temperature calibration, bad pixel repair, thermal noise filtering, and pseudo-color consistency processing. This processing ensures the comparability of infrared thermal images at different times. Based on the high-voltage switchgear's factory structural diagram, on-site calibration diagrams, or manual calibration points, the visible light and infrared thermal images are mapped to a unified cabinet structural coordinate system. After registration, components such as cable terminals, busbar joints, contact boxes, insulators, mechanism links, and auxiliary contacts in the images correspond one-to-one with the structural template.

[0047] Furthermore, this application argues that high-voltage switchgear faults do not propagate randomly, but rather spread along specific physical paths within the cabinet. Based on the structure and fault type of the high-voltage switchgear, the state within the cabinet is abstracted into the following four physical paths:

[0048] 1. Conductive Path: The conductive path describes the main circuit path through which current flows, including: cable terminal — cable connection terminal — busbar lap joint — contact box — circuit breaker moving and stationary contacts — outgoing busbar. This path mainly corresponds to faults such as contact overheating, contact erosion, loose connection, and abnormal load.

[0049] 2. Creepage Path: The creepage path describes the path through which surface discharge may occur on the insulation surface, including: insulator surface—outer edge of contact box—outer surface of bushing—insulation layer of cable terminal—condensation-covered area. This path mainly corresponds to the risks of insulation moisture, condensation, partial discharge, surface discharge, and flashover.

[0050] 3. Heat Dissipation Path: The heat dissipation path describes the route of heat transfer from the fault location to the surrounding structure, including: heat source point—adjacent conductor—insulator edge—partition—adjacent compartment boundary—vent or enclosed cavity. This path mainly corresponds to the risk of joint overheating propagation, heat transmission across compartments, and thermal runaway.

[0051] 4. Mechanism Execution Path: The mechanism execution path describes the transmission process from remote operation commands to mechanical motion feedback, including: remote control command—energy storage status—operating coil—operating linkage—circuit breaker position indication—auxiliary contact feedback—remote signal return. This path mainly corresponds to mechanism jamming, false positioning, auxiliary contact abnormality, handcart position inconsistency, and remote operation failure.

[0052] Each path is discretized into multiple physically meaningful nodes, denoted as: , Indicates the first A physical path; This represents the j-th node in the u-th path; This indicates the number of nodes contained in the physical path. Nodes can be actual components such as cable terminals, insulator surface sections, busbar connection points, partition edges, mechanism links, or auxiliary contacts.

[0053] Multi-source data is projected onto the physical path to form path-specific observations. Conductive path observations: Linear thermal response is extracted from infrared thermal images along the conductive path and combined with corresponding current values ​​to reflect the unit current heat load at the conductive path nodes. Creepage path observations: Condensation continuity, visible light surface bright spots, partial discharge activity, and dew point margin are extracted along the insulating surface to reflect whether continuous discharge conditions have formed on the insulating surface. Heat diffusion path observations: Heat transmission status is extracted along partitions, vents, and adjacent structural boundaries to reflect whether heat has crossed structural boundaries from the original fault location. Mechanism execution path observations: The timing of actions at each node is extracted along the path from control command to mechanical feedback to reflect whether there are delays, blockages, or false arrivals in the remote operation chain.

[0054] The core of this processing method lies in the fact that multi-source data is not simply used in parallel, but rather projected onto the actual physical paths of the high-voltage switchgear, giving subsequent judgments structural semantics and interpretability. It should be noted that the connection methods of nodes in various paths in this application are illustrative and generalized topological frameworks, not fixed. Specific connection methods need to be determined based on the actual situation of the high-voltage switchgear. The core of this application is to subdivide the four physical paths into nodes with clear physical meaning. The specific links included in each path can be added, deleted, or modified according to actual conditions. This facilitates adaptation to different switchgear models; only the coupling coefficient between the node definitions and the paths needs to be adjusted.

[0055] The path transmission mismatch analysis module is used to obtain the equivalent state quantity of each node on various physical paths using visible light images, infrared thermal images, electrical operation data, insulation environment data, and actuator status data; and to obtain the transmission mismatch intensity of a physical path based on the changes in the equivalent state quantity of each node on a physical path.

[0056] Existing fault monitoring methods for high-voltage switchgear largely rely on single sensor thresholds or simple feature aggregation, such as alarms triggered by temperature, humidity, or partial discharge exceeding a threshold. These methods lack an understanding of the internal structural relationships within the switchgear, making it difficult to determine whether a fault is propagating along a specific physical path, and even more difficult to reliably determine remote intervention actions. This application, based on the fixed internal structure and well-defined fault propagation paths of high-voltage switchgear, proposes a path-based fault monitoring and remote intervention logic. The core idea is that under healthy conditions, adjacent nodes in each physical path of the high-voltage switchgear have relatively stable state transmission relationships. If a node experiences poor contact, moisture discharge, thermal diffusion, or mechanical jamming, the transmission relationship between adjacent nodes on that path will become mismatched. If the fault further develops, this mismatch will be sensed from one path to adjacent paths. Therefore, the fault location, nature, and extent of expansion can be determined through path transmission mismatch, and the most suitable remote intervention action can be selected through an action influence matrix.

[0057] Different faults in high-voltage switchgear manifest differently on different paths. Overheating mainly manifests as abnormal thermal response on the conductive path; moisture-induced partial discharge mainly manifests as abnormal coupling between the wettability continuity and partial discharge activity on the creepage path; thermal runaway mainly manifests as enhanced heat diffusion path transmission across compartments; and mechanical jamming mainly manifests as timing mismatch in the mechanism execution path. Therefore, this application first constructs a unified equivalent state quantity for different physical paths.

[0058] Therefore, the equivalent state quantities of each node on various physical paths are obtained by using visible light images, infrared thermal images, electrical operation data, insulation environment data, and actuator state data.

[0059] Specifically, for a node in a conductive path, the temperature is extracted pixel by pixel along the conductor skeleton line corresponding to the node in the registered infrared image, and the temperature is corrected using the material emissivity to obtain the temperature corresponding to each pixel. The maximum value between the temperature of a pixel and the ambient temperature and 0 is obtained as the temperature rise corresponding to that pixel. The actual physical length corresponding to each pixel is obtained, and the product of the temperature rise corresponding to each pixel and the actual physical length is summed to obtain the conductive node feature. The conductive node feature is divided by the sum of the current of the loop and the constant term to obtain the equivalent state quantity of the node in the conductive path.

[0060] For a node in a creepage path, the visible light image is segmented to obtain the wetted region. The wetted region is projected onto the surface of the corresponding insulating component. The length of the maximum wetted connected region along the creepage direction on the surface of the corresponding insulating component is calculated as the ratio of the length of the maximum wetted connected region along the creepage direction to the total creepage length on the surface of the corresponding insulating component, thus obtaining the wetted continuity of the insulating surface. The mean values ​​of the TEV amplitude and the mean values ​​of the ultrasonic amplitude collected by the transient ground voltage sensor and the ultrasonic partial discharge sensor within the detection range of the node within a preset time period are normalized and then weighted and summed to obtain the partial discharge activity. The average temperature of the surface of the corresponding insulating component is extracted using an infrared thermal image. The dew point temperature is calculated using the Magnus formula based on the temperature and humidity sensor data inside the cabinet. The difference between the average temperature and the dew point temperature of the surface of the corresponding insulating component is calculated to obtain the dew point margin. The product of the wetted continuity of the insulating surface and the partial discharge activity is calculated and divided by the sum of the dew point margin and a constant term to obtain the equivalent state quantity of the node in the creepage path.

[0061] For a node in a heat diffusion path, in the registered infrared thermal image, the structural boundary corresponding to the node is selected, and two adjacent regions of interest are delineated on both sides of the boundary. The average temperature of the two regions of interest is calculated, and the larger value is recorded as the average temperature on the hot side and the smaller value is recorded as the average temperature on the cold side. The difference between the average temperature on the hot side and the average temperature on the cold side is recorded as the hot-cold temperature difference. The equivalent state quantity of the node is obtained by dividing the hot-cold temperature difference corresponding to the node by the sum of the reference hot-cold temperature difference and the constant term.

[0062] For a node in an organization's execution path, the time from the completion of the action of the previous node to the change of the state of this node is determined by the Sequence of Events (SOE), and is recorded as the actual response time of this node. The absolute value of the difference between the actual response time and the standard response time of this node is calculated and divided by the standard response time and a constant term of this node to obtain the equivalent state quantity of this node in the organization's execution path.

[0063] The specific calculation model is as follows:

[0064] ,

[0065] in, Let ε be the equivalent state variable of the j-th node of the u-th physical path; ε is a constant term, which is a very small positive number to prevent the denominator from being 0.

[0066] Let be the equivalent state quantity of the j-th node of the u-th physical path when the path is a conductive path. Let be the conductive node feature quantity of the j-th node of the u-th physical path; represent the heat integral quantity obtained near this node along the conductive skeleton line. After image and cabinet structure registration, the correspondence between each pixel of the infrared thermal image and the actual cabinet size (mm) can be obtained. (The text then abruptly shifts to a different topic:) ...nodes along the conductive path... The corresponding conductor skeleton line (cable terminal, busbar lap joint, contact conductive rod, etc.) is assumed to have a total of M pixels, and the pixels are sorted according to the path direction as follows: For pixels , From this point to the next pixel The actual spatial distance between the pixels is denoted as the actual physical length corresponding to that pixel. Temperature is extracted pixel by pixel along the conductor skeleton line corresponding to that node, and the temperature is corrected using the material emissivity to obtain the temperature corresponding to each pixel. (Temperature of the Mth pixel) is taken as the ambient reference temperature measured by the temperature sensor inside the cabinet. Calculate pixel points Corresponding temperature rise The path integral (summation) of the temperature rise along the skeleton line yields the following result. Because infrared thermometry is affected by the emissivity of the object's surface, the temperature needs to be corrected based on the known emissivity ϵ of the conductor material (such as copper, aluminum, or silver plating). It has the dimension [K·m] (temperature × length), which comprehensively reflects the degree of temperature rise accumulation caused by resistive heating in the node region, and the integral length clearly defines the node range. When the contact resistance increases, the local temperature rise at the node becomes steep. The temperature rise will be significantly higher; while when the overall load current increases, although the temperature rise will also change, The ratio can effectively isolate the influence of current, thereby highlighting the characteristics of contact degradation. The current in this circuit refers to the current flowing through the main circuit corresponding to the u-th conductive path. In engineering, it is the effective value of the three-phase current in this switchgear bay (or feeder), obtained from the secondary side of the current transformer via the acquisition unit, and is an existing electrical quantity. The reason is that conductor heating is related to the square of the current. If the influence of the current is not eliminated, the increase in load may be misjudged as a contact fault.

[0067] Let be the equivalent state quantity of the j-th node of the u-th physical path when the path is a creepage path. The measurement of continuous wetting on the insulating surface is obtained through the following steps: The visible light image is segmented to identify bright wetting areas such as condensation and water film. These wetting areas are then projected onto the surface of the corresponding insulating component (e.g., insulator skirts, contact box outer wall). The length of the largest connected region (the largest area of ​​which is the same as the area of ​​the largest connected region) along the creepage direction within the surface of the corresponding insulating component is calculated. Total creepage length of the insulating surface of this node The ratio is used to obtain the continuous amount of wetting on the insulating surface. This is not the total path length. The total creepage length is the shortest surface creepage distance that the insulating component corresponding to the node provides for current in a healthy state. Assuming the node being analyzed is a post insulator, the total creepage length of the node refers to the nominal creepage distance of this insulator as defined by national standards (such as GB / T 2900.8-2009), measured from the end connected to the live conductor along the surface of its skirts to the end of its grounding metal mounting plate. This value is fixed and can be found directly in the technical manual of the insulator or accurately measured using a three-dimensional structural template. This indicates the partial discharge activity near the node. The acquisition process is as follows: Ultrasonic partial discharge sensors and transient ground voltage (TEV) sensors are arranged according to the measurement points within the cabinet. Their detection range corresponds to specific creepage path nodes. The average TEV amplitude is taken from the TEV sensors and ultrasonic partial discharge sensors within a preset time period (e.g., within the most recent minute) that include the node in the detection range. and the mean of ultrasonic amplitude The localized activity is obtained by weighted synthesis after normalization. , If the node has no direct sensor coverage, then distance-weighted interpolation is performed based on the values ​​of nearby sensors. Dew point margin is represented by the following process: node extraction using infrared thermal imaging. Average temperature of insulating surface Based on data from the temperature and humidity sensors inside the cabinet, Formula for calculating dew point temperature ,calculate A negative value indicates that the temperature is below the dew point, posing a risk of condensation; this is the difference between the insulation surface temperature and the dew point temperature. If the continuous wetting amount is high, the partial discharge activity is high, and the dew point margin is small, then the node has a strong risk of surface discharge.

[0068] The equivalent state quantity of the j-th node of the u-th physical path when the path is a heat diffusion path. The temperature difference between the hot and cold sides of the node represents the amount of heat transmitted across the structural boundary. The acquisition process is as follows: In the registered infrared thermal image, select the structural boundary (such as the edge of a partition or the location of a through-wall sleeve) corresponding to the heat diffusion path node. Define two adjacent regions of interest on both sides of the boundary. Calculate the average temperature of the two regions of interest and record the larger value as the average temperature on the hot side. The smaller value is recorded as the average temperature on the cold side. , (If thickness differences are involved, they can be normalized) to characterize the strength of heat crossing the structural boundary; This represents the reference heat transmission under similar operating conditions in a healthy state, also known as the baseline hot and cold temperature difference. The acquisition process involves recording the hot and cold temperature difference at that node under operating conditions similar to the current load current and ambient temperature during historical periods when the switchgear is confirmed to be fault-free and operating normally. This data is then used to establish a reference temperature difference based on current. and ambient temperature For the input reference table or regression model, query the baseline hot and cold temperature difference corresponding to the current operating condition as... If historical data is insufficient, the steady-state temperature difference value from the factory test can also be used. This value is used to determine whether heat is being transferred from a local fault point to other structures.

[0069] Let be the equivalent state quantity of the j-th node of the u-th physical path when the path is the execution path of the mechanism. This indicates the actual response time of the node. The acquisition process is as follows: the time from the completion of the action of the previous node to the change of the state of this node is determined by the SOE (Sequence of Events) function of the control system. This represents the standard response time of the node under healthy conditions. The acquisition process is as follows: After the equipment is put into operation or maintenance and the mechanism is confirmed to be normal, perform multiple opening and closing operations, record the actual response time of each node, remove outliers, and take the average value as the standard response time. The data is stored in the device's non-volatile memory, which describes whether the mechanism's action is delayed or stuck.

[0070] This equivalent state quantity is not simply temperature, humidity, area, or distance, but a scenario-based quantity constructed by combining path type and fault mechanism. Conductive paths focus on the thermal burden after current normalization; creepage paths focus on the coupling between wetted continuity, partial discharge, and dew point margin; thermal diffusion paths focus on heat transfer across structural boundaries; and mechanical paths focus on action timing deviations.

[0071] In a healthy state, there should be a relatively stable transmission relationship between adjacent path nodes. Taking a conductive path as an example, if the cable terminal, busbar joint, and contact box are all in a healthy state, the normalized thermal response of the current should change relatively smoothly along the path; if the contact resistance of a certain joint is abnormal, the transmission relationship between that node and its adjacent nodes will be disrupted.

[0072] Therefore, the transmission mismatch intensity of a physical path can be obtained based on the changes in the equivalent state variables of each node along the physical path.

[0073] Specifically, the transmission coefficient between every two adjacent nodes in a fault-free state is obtained; the transmission mismatch residual between two adjacent nodes is obtained by subtracting the equivalent state quantity of the previous node from the equivalent state quantity of the latter node and the product of the transmission coefficient between the two adjacent nodes; the average of the absolute values ​​of the transmission mismatch residuals between every two adjacent nodes in the physical path is calculated and denoted as the first average value; the average of the product of the transmission mismatch residuals between every two adjacent nodes in the physical path and the path direction sign is calculated and the absolute value is taken to obtain the second average value; the sum of the first average value and the second average value is obtained to obtain the transmission mismatch intensity of the physical path.

[0074] The specific calculation model for the propagation mismatch residual is as follows:

[0075] ,

[0076] in, This represents the propagation mismatch residual between the j-th node and the (j+1)-th node in the u-th physical path; This represents the equivalent state quantity of the previous node; This represents the equivalent state quantity of the next node; This represents the normal propagation coefficient between two adjacent nodes in a fault-free state. Multiple sets of healthy samples can be calculated by selecting continuous operating data from the initial stage of switchgear commissioning, after confirmation that there are no abnormalities. Robust regression (such as the least absolute residual method) is used to obtain the linear relationship. Thus, the transmission coefficient is obtained. .

[0077] like A value close to 0 indicates that the state transfer of the physical path conforms to healthy laws; if... A significant deviation from 0 indicates the presence of abnormal gain, abnormal obstruction, or abnormal discharge in the physical path. For example, a sudden increase in heat generation at a joint in a conductive path can cause heat transfer mismatch; enhanced partial discharge due to wetting in an insulation section of a creepage path can cause discharge risk transmission mismatch; and a delay in the movement of a link in a mechanism path can cause timing transmission mismatch.

[0078] The residual reflects whether the relationship between path nodes is abnormal, rather than whether the value of a single node exceeds the limit. This processing can effectively avoid false alarms of single-point thresholds. For example, an increase in overall load may cause the temperature of the entire path to rise, but the transmission relationship between adjacent nodes remains healthy, so it will not be misjudged as poor contact; while a local contact fault, even if the temperature has not exceeded the maximum threshold, will disrupt the transmission relationship between nodes, thus being identified in advance.

[0079] The abnormality level of the entire path is described, and the propagation mismatch strength of the path is constructed. The specific calculation model for the propagation mismatch strength is as follows:

[0080] ,

[0081] in, Indicates the first The transmission mismatch intensity of each physical path; the first term is the first average value, representing the average of the absolute values ​​of the mismatch residuals of all adjacent nodes along the path, used to reflect the overall degree of mismatch. Indicates the first The number of nodes on each physical path. The first term represents the propagation mismatch residual between the j-th node and the (j+1)-th node in the u-th physical path; the second term represents the directional mismatch term. The path direction sign of the j-th node is used to indicate whether the residual continuously increases along the main path direction; to determine whether the mismatch propagates consistently along the path direction, this application will... The value is uniformly set to +1 to represent the positive direction along the path from the beginning to the end; in this case, the directional term... The second average value degenerates into the absolute value of the residual mean; when the residuals of adjacent segments have the same sign, the absolute value of the residual mean is large; if the positive and negative signs cancel each other out, the value is small, thus effectively reflecting the consistency of the direction of fault propagation. In other embodiments, the path physical flow direction can also be preset. Whether the value is +1 or -1, it falls within the scope of this application. Path-transmitted mismatch residual. and mismatch strength The calculation method remains unchanged.

[0082] If the fault is only local noise or measurement disturbance, the directions of the residuals in different segments are usually inconsistent, and the directional terms will cancel each other out. If the fault is developing along a path, such as heat continuously spreading along the conductor or insulating surface, then multiple residuals will show the same direction, and the directional terms will increase.

[0083] This formula not only determines whether a path is abnormal, but also whether the abnormality has a tendency to propagate along the path. Therefore, it can distinguish between "isolated disturbances" and "developing faults," which is of great significance for remote and rapid handling of high-voltage switchgear.

[0084] The path fault constraint analysis module is used to set the structural coupling coefficient between every two physical paths, obtain the cross-path induction of a physical path based on the structural coupling coefficient between a physical path and other physical paths and the transmission mismatch strength of each physical path, and obtain the fault constraint coefficient based on the transmission mismatch strength and cross-path induction of a physical path.

[0085] When a physical path mismatch occurs, it doesn't necessarily mean that the circuit breaker must be tripped immediately. The key to rapid remote handling lies in determining whether the fault remains confined to the original path or has already triggered anomalies in other paths. If the fault remains confined to the conductive path, it may be mitigated through heat dissipation, load reduction, or load transfer. If the fault has already triggered anomalies in creepage paths or heat dissipation paths from the conductive path, it indicates that the fault is expanding and should be isolated first. Therefore, this application constructs a cross-path inductance parameter to determine the extent to which the fault extends from the main path to adjacent paths.

[0086] Let the u-th physical path be... With the vth physical path The structural coupling coefficient between them is This coefficient is determined by the cabinet structure, electrical connections, thermal proximity, and insulation proximity. For example, the coupling coefficient between conductive paths and thermal diffusion paths is relatively high; the coupling between creepage paths and partial discharge-related paths is relatively high; there is no direct thermal coupling between the mechanism execution path and the conductive path, but it is highly related to the safety of remote opening and closing; the coupling between the conductive path in the busbar compartment and the thermal diffusion path in the adjacent cable compartment depends on the partition structure and ventilation path.

[0087] Based on the assembly drawings, thermal simulation models, and electrical wiring of the high-voltage switchgear, determine whether there is a heat conduction path, shared insulation space, or electrical connection between any two physical paths. For physical path pairs with direct coupling, their structural coupling coefficient can be set. For two physical paths without direct physical connection, their structural coupling coefficient is set to 0.

[0088] Specifically:

[0089] 1. Conductive Path ↔ Thermal Diffusion Path (Strong Coupling): This is the most direct coupling relationship. Abnormal contact resistance along the conductive path directly generates heat, which is conducted along the structural components. This is exactly what the thermal diffusion path describes. Therefore, the structural coupling coefficient is very high and can be set to a certain value. .

[0090] 2. Conductive Path ↔ Creepage Path (Weak Coupling / Conditional Coupling): The heating of the conductive path itself does not directly cause surface discharge, but it may accelerate the aging of surrounding insulating materials or induce creepage when condensation forms on their surface. Therefore, there is no forced direct physical coupling between the two, and the structural coupling coefficient can be set to... .

[0091] 3. Conductive Path ↔ Mechanism Execution Path (Uncoupled): There is no direct physical connection between the state of the conductive circuit and the mechanical transmission chain of the operating mechanism. Overheating at a contact point will not directly cause the mechanism linkage to jam. Therefore, the structural coupling coefficient is set to 0.

[0092] 4. Thermal diffusion path ↔ Creepage path (strong coupling): This is the second most important correlation after "conductivity-heat". Heat propagation across compartments changes the temperature of the insulation components, directly affecting their surface "dew point margin". When the temperature drops below the dew point, condensation forms, directly constituting the initiation condition for a creepage path. Therefore, the structural coupling coefficient is very high and can be set as... .

[0093] 5. Heat diffusion path ↔ Mechanism execution path (uncoupled): Heat diffusion typically does not directly cause mechanical jamming. Anomalies in the mechanism path primarily stem from mechanical faults or control loop problems, rather than thermal effects. Therefore, the structural coupling coefficient is set to 0.

[0094] 6. Creepage Path ↔ Mechanism Execution Path (Uncoupled): There is no direct physical linkage between the discharge on the insulating surface and the mechanism's movement. Therefore, the structural coupling coefficient is set to 0.

[0095] 7. Coupling between the path and itself: The structural coupling coefficient of a physical path is 1, because the state change of a path completely determines its own state. As shown in Table 1.

[0096] Table 1

[0097] Conductive path 1.0 0.8 (Strong) 0.2(weak) 0 Thermal diffusion path 0.8 (Strong) 1.0 0.7 (Strong) 0 creepage path 0.2(weak) 0.7 (Strong) 1.0 0 Institutional Implementation Path 0 0 0 1.0

[0098] It should be noted that the mechanism execution path is a special case. Although its structural coupling coefficient with other paths is 0, it is a prerequisite for performing any electrical remote action (such as tripping or load reduction). The logic originates from custom engineering parameters based on well-known physical principles. The logic for setting the coupling coefficient is based on recognized physical laws in electrical engineering and thermodynamics, for example:

[0099] The law of thermal conduction states that heat will be conducted from the high-temperature area along the structural components to the low-temperature area, therefore the conductive path and the thermal diffusion path must be strongly coupled.

[0100] Insulation and Discharge Physics: Condensation will occur when the temperature of the insulating surface is below the dew point, which constitutes the conditions for creepage. Therefore, there is a strong coupling between the heat diffusion path and the creepage path.

[0101] Mechanical and electrical isolation: The jamming of the mechanism is a purely mechanical fault and has no necessary physical linkage with conductive heating or insulation discharge. Therefore, the coupling coefficient between the mechanism path and other paths is 0.

[0102] Therefore, the cross-path induction of a physical path is obtained based on the structural coupling coefficient between a physical path and other physical paths and the transmission mismatch strength of each physical path.

[0103] Specifically, the cross-path induction of a physical path is obtained by weighting the transmission mismatch strength of the other physical paths with the structural coupling coefficient between a physical path and other physical paths, and then summing the weights of the transmission mismatch strength of the physical path and the constant term.

[0104] The specific calculation model is as follows:

[0105] ,

[0106] Let be the cross-path sensing quantity of the u-th physical path, representing the relative intensity of its induction of anomalies on other paths when the u-th physical path is the main abnormal path. This represents the transmission mismatch strength of the u-th physical path; Indicates the intensity of mismatch in other paths; This represents the structural coupling coefficient between the u-th physical path and other physical paths; ε is a constant term, a very small positive number. If A smaller value indicates that although the main path is significantly abnormal, adjacent paths have not yet been significantly affected, and the fault may still be in a locally controllable state. A large value indicates that the anomaly on the main path has already sensed other paths, and the fault is spreading across structures.

[0107] This formula elevates the fault diagnosis from "single-path anomaly" to "inter-path propagation relationship judgment." For example, if overheating only causes a mismatch in the conductive path, it can still be addressed by remote load reduction or heat dissipation; however, if it simultaneously causes mismatches in the thermal diffusion path and creepage path, it may have already affected the insulation condition. In this case, continued operation should be avoided, and isolation should be prioritized.

[0108] Furthermore, the fault constraint coefficient is obtained based on the transmission mismatch intensity and cross-path induction of a physical path.

[0109] Specifically, the fault limitation coefficient of a physical path is obtained by dividing the transmission mismatch intensity of a physical path by a first preset value and the sum of the cross-path induction of the physical path.

[0110] The specific calculation model is as follows:

[0111] ,

[0112] Indicates the first Fault limitation coefficient of each physical path; This indicates the strength of the transmission mismatch along the physical path; This represents the cross-path sensing value of the physical path; 1 is the first preset value; when higher and A lower value indicates a relatively high fault intensity, but it is still concentrated within the main path. Higher levels are suitable for considering localized relief exercises; when higher and A higher level indicates that the fault has spread to other paths. If the condition worsens, it is not advisable to continue with slow mitigation measures. Instead, isolation, lockdown, or manual intervention should be prioritized.

[0113] This coefficient can distinguish between two types of situations that are difficult to differentiate using traditional methods: high-intensity but limited faults, such as a single cable termination contact overheating that has not yet affected the insulation or adjacent compartments, which can be addressed by remote load reduction or heat dissipation; and medium-intensity but spreading faults, such as partial discharge that has not yet reached a particularly high level but has already creeped along the insulation path and induced an arcing signal, in which case rapid isolation is necessary. Therefore, this application no longer judges the severity of a fault solely by its numerical value, but rather by whether the fault is spreading and thus the urgency of its handling.

[0114] The fault handling module is used to construct an action influence matrix for each action based on the remote actions that can be performed by the high-voltage switchgear; obtain the remaining path mismatch vector after the execution of the action by using the transmission mismatch strength of all paths and the action influence matrix of the action; obtain the correction benefit of the action by using the transmission mismatch strength of all paths and the remaining path mismatch vector after the execution of the action; and select the action to be executed to handle the fault based on the correction benefit.

[0115] Following the above analysis, the transmission mismatch intensity, main abnormal path, cross-path induction, and fault limitation coefficient for each physical path have been obtained. The next problem to be solved is: which remote processing action should be executed at this time?

[0116] Existing technologies typically employ fixed rules, such as activating fans when temperatures are high, dehumidifiers when humidity is high, alarms when partial discharge is high, and tripping when arcing occurs. These rules cannot handle complex faults or determine whether actions will have side effects on other paths. For example, heaters can reduce the continuous wetting of creepage paths, but may increase the burden on localized heat diffusion paths; fans can reduce heat diffusion path mismatch, but have limited effect on creepage paths already exhibiting surface discharge; load reduction can reduce thermal mismatch in conductive paths, but has no effect on mechanical jamming; remote tripping can cut off energy input to conductive paths, but requires confirmation of mechanical path reliability, otherwise there may be a risk of failure to operate.

[0117] Therefore, this application constructs an action influence matrix for each action based on the remote actions that the high-voltage switchgear can perform, in order to express the ability of different remote actions to correct different path mismatches.

[0118] Let the set of remote actions that the current high-voltage switchgear can perform be:

[0119] ,

[0120] in, This can indicate that the fan is being started; This indicates that the dehumidifier has been activated; This indicates that low-power heating has been activated; This indicates a reduction in load. Indicates load transfer; Indicates remote tripping; Indicates that reclosing is blocked; This indicates that repeated manipulation is prohibited; Indicates the number of actions that can be performed.

[0121] For each action Establish an action influence matrix:

[0122] ,

[0123] In the formula, Indicates action The influence matrix; Indicates action For the Physical path mismatch towards the first The degree of impact of the physical path correction effect; the acquisition process is as follows: for each cabinet type, action response data is obtained through special tests before leaving the factory (such as recording temperature changes in each path after the fan is started, recording humidity and partial discharge changes after the dehumidifier is started, etc.); the equivalent state variables of each physical path after the action are calculated. The decrease in the value, after normalization, is used as the direct impact coefficient; for indirect impact (p≠q), it is calculated based on the physical correlation, calculated from the direct coefficient. Proportional conversion; for switching actions such as tripping and blocking, the impact is mainly reflected in the energy cutoff of the conductive path and heat dissipation path, and the correction coefficient is set according to the energy cutoff ratio; the matrix can be preset offline and fine-tuned during on-site trial operation to verify the effectiveness of the actions. Indicates the number of physical paths.

[0124] For example, starting the fan has a strong positive correction on the heat diffusion path and an indirect positive correction on the conductive path, but has almost no effect on the mechanism execution path; starting the dehumidifier has a strong positive correction on the creepage path, but a weak effect on the conductive path; remote tripping has a strong correction effect on the conductive path and the heat diffusion path, but its execution depends on the reliability of the mechanism path.

[0125] Furthermore, the remaining path mismatch vector after the execution of an action is obtained by utilizing the transmission mismatch intensity of all paths and the action influence matrix of an action.

[0126] Specifically, the propagation mismatch intensities of all paths are combined to form the current propagation mismatch intensities vector; then, the remaining path mismatch vector after the execution of an action is specifically:

[0127] ,

[0128] in, Indicates action The remaining path mismatch vector after the execution of this action, where k is the index of the action; Indicates action Action influence matrix; Represents the identity matrix; This represents the current propagation mismatch strength vector.

[0129] If an action can effectively reduce the mismatch of a certain path, the corresponding element in the influence matrix will be large; if an action has no significant effect on a certain path, the corresponding element will be close to 0; if an action may cause side effects, the corresponding element can be reflected as an adverse effect.

[0130] This approach upgrades remote action selection from "fault type - fixed action" to "current path mismatch state - action matrix correction capability - action optimization". Essentially, it evaluates the corrective effect of actions on the faulty structural state, rather than simply matching rules.

[0131] Therefore, the corrective benefit of an action is obtained by using the transmission mismatch intensity of all paths and the remaining path mismatch vector after the execution of an action.

[0132] Specifically, the correction gain of an action is obtained by comparing the difference between the sum of the absolute values ​​of the transmission mismatch intensity of each physical path and the sum of the absolute values ​​of each element in the remaining path mismatch vector after the execution of an action, and the sum of the sum of the absolute values ​​of the transmission mismatch intensity of each path and the constant term.

[0133] The specific calculation model is as follows:

[0134] ,

[0135] in, Indicates action Benefits of correcting the current fault state; It represents the sum of the absolute values ​​of the transmission mismatch intensity of each physical path before the action is executed, which is also the sum of the absolute values ​​of each element in the current transmission mismatch intensity vector; Indicates action The sum of the absolute values ​​of each element in the remaining path mismatch vector after execution; ε is a constant term, usually a very small positive number.

[0136] like The larger the value, the better the action reduces overall path mismatch; if... If the value is close to 0, it means the action is essentially ineffective; if A negative value indicates that the action may worsen the state of the faulty structure.

[0137] Therefore, the action that yields the greatest benefit is selected as the action to be executed to handle the current fault.

[0138] Typical fault handling logic:

[0139] When encountering overheating faults, if the conductive path mismatch intensity is significantly increased and the cross-path induction is low, it indicates that the fault is mainly limited to the cable terminal, busbar connection point, or contact area. In this case, the system prioritizes comparing the optimal values ​​of actions such as fan, load reduction, and load transfer. If the load reduction action is significantly better than the fan action, it indicates that the heat generation is mainly driven by the load, and load reduction or power transfer should be prioritized. If the benefit of local mitigation action is low, while the benefit of remote tripping is high, it indicates that the overheating can no longer be controlled by ordinary mitigation measures, and the faulty circuit should be isolated.

[0140] For surface discharge faults caused by moisture, when the creepage path mismatch intensity increases significantly and the coupling between the continuous moisture content, partial discharge activity, and dew point margin strengthens, the system determines that there is a risk of surface discharge on the insulation surface. If the cross-path induction is low, it indicates that the fault is still in the early stage of moisture absorption, and dehumidification and low-power heating can be prioritized. If the cross-path induction increases and is accompanied by arcing or current disturbance, it indicates that the discharge risk has expanded, and reclosing should be blocked and preparations should be made for remote isolation.

[0141] In cases of heat diffusion faults affecting multiple compartments, when the heat diffusion path mismatch increases and both the conductive and creepage paths exhibit induced anomalies, it indicates that heat has spread from the local heat source to adjacent insulation components or compartments. In this situation, simple fan cooling may be insufficient; therefore, load descent, power transfer, and remote tripping actions should be prioritized. If remote tripping is the preferred option, the faulty circuit should be directly isolated to prevent further thermal damage.

[0142] In cases of mechanism jamming or false positioning faults, when the mechanism execution path mismatch increases (manifested as inconsistencies between control commands, energy storage status, mechanism position indication, and auxiliary contact feedback), the system will not directly repeat the opening and closing operations. Instead, it will prioritize actions such as prohibiting repeated operation, blocking reclosing, and status verification. If the mechanism path mismatch is not eliminated, a field maintenance work order will be generated to prevent further mechanical damage due to repeated remote operations. When selecting actions, the corrective benefits must be considered.

[0143] Furthermore, the remote rapid processing strategy is translated into actual control actions, and the effectiveness of these actions in suppressing the fault is verified. This operation is the final step in this application's return to the "remote rapid fault monitoring and processing function."

[0144] Before executing a remote action, the system performs the following checks:

[0145] The high-voltage switchgear is in a remote control-enabled state; the communication link is normal; the control power supply is normal; the circuit breaker, handcart, grounding switch, and energy storage meet the operational requirements; the five-proof interlocking logic is satisfied; and the protection devices are not in an abnormally locked state. If any condition is not met, the remote action will not be executed, and a field handling prompt will be automatically generated. If the condition is met, the system will execute the corresponding action according to the action preference value, including but not limited to:

[0146] Start the fan to dissipate heat from the conductive and heat diffusion paths; start the dehumidifier to reduce the continuous moisture in the creepage path; start the low-power heater to increase the dew point margin and prevent continuous condensation; reduce the load or transfer the load to reduce the thermal burden on the conductive path; remotely trip the circuit breaker to isolate the faulty circuit; lock the reclosing circuit to prevent repeated power supply when the fault has not been cleared; prohibit repeated operation to prevent further mechanical damage in case of mechanical jamming; switch to the backup circuit to ensure continuous power supply.

[0147] After the action is executed, the system re-acquires visible light images, infrared thermal images, electrical data, partial discharge data, and mechanism status data, and recalculates the mismatch intensity of each physical path to obtain the transmitted mismatch intensity vector after the action:

[0148] ,

[0149] Then, the effectiveness index is calculated:

[0150] ,

[0151] Formula explanation: Indicates the validity of remote processing actions; This represents the sum of the absolute values ​​of each element in the transmission mismatch intensity vector before the action; This represents the sum of the absolute values ​​of each element in the transmitted mismatch intensity vector after the action; ε is a constant term.

[0152] If V is large, it indicates that the mismatch between paths has significantly converged after the action, and the handling is effective; if V is small or negative, it indicates that the action has failed to effectively suppress the fault, and the system should stop repeatedly executing the same mitigation action and upgrade to isolation, interlocking or manual on-site handling.

[0153] This verification method does not simply determine whether the temperature has decreased, the humidity has decreased, or the current has disappeared, but rather reassesses whether the transmission relationships of each physical path have returned to normal. Therefore, even if a single indicator temporarily improves, but path mismatch still exists, the system can still determine that the fault has not been truly resolved, avoiding misjudgment.

[0154] If the total path mismatch decreases significantly after the action, and the cross-path sensing data decreases synchronously, the system will maintain enhanced monitoring and continuously verify the data over a certain period of time. If the situation remains stable, the high-level alarm will be deactivated.

[0155] If the conductive path mismatch decreases but the creepage path mismatch remains high, it indicates that the fault may have multiple causes, such as overheating of the contact simultaneously causing the risk of insulation moisture. In this case, the system recalculates and selects a secondary remote action to handle the fault.

[0156] If the path mismatch does not decrease or continues to increase after the action, the system will not repeat the same mitigation action, but will instead upgrade to remote tripping, blocking reclosing, or manual on-site handling.

[0157] If the original path mismatch decreases after the action, but the mismatch of other paths increases, it indicates that the remote action has side effects or a fault transfer has occurred. For example, if the conductivity path mismatch of the backup circuit increases after the load transfer, the system will indicate that the backup circuit's carrying capacity is insufficient and will regenerate the handling strategy.

[0158] After each fault is handled, the system generates a fault evidence package, which includes: fault occurrence time; main abnormal path; mismatch intensity of each path; cross-path induction; fault limitation coefficient; action correction benefit of each candidate action; final executed action; changes in path mismatch before and after the action; visible light image, infrared thermogram, current waveform, partial discharge record, and mechanism status record; and subsequent maintenance suggestions. This evidence package can be uploaded to the remote operation and maintenance platform for operation and maintenance personnel to review and generate maintenance work orders.

[0159] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0160] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0161] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A high-voltage switchgear with remote rapid fault monitoring and handling function, characterized in that, The high-voltage switchgear includes: The data acquisition module is used to collect visible light images, infrared thermal images, electrical operation data, insulation environment data, and actuator status data inside the high-voltage switchgear; and to set various physical paths for the high-voltage switchgear, with different nodes on each physical path. The path transmission mismatch analysis module is used to obtain the equivalent state quantity of each node on various physical paths using visible light images, infrared thermal images, electrical operation data, insulation environment data, and actuator status data; and to obtain the transmission mismatch intensity of a physical path based on the changes in the equivalent state quantity of each node on the physical path. The path fault constraint analysis module is used to set the structural coupling coefficient between every two physical paths, obtain the cross-path induction of a physical path based on the structural coupling coefficient between a physical path and other physical paths and the transmission mismatch strength of each physical path, and obtain the fault constraint coefficient based on the transmission mismatch strength and cross-path induction of a physical path. The fault handling module is used to construct an action influence matrix for each action based on the remote actions that can be performed by the high-voltage switchgear; obtain the remaining path mismatch vector after the execution of the action by using the transmission mismatch strength of all paths and the action influence matrix of the action; obtain the correction benefit of the action by using the transmission mismatch strength of all paths and the remaining path mismatch vector after the execution of the action; and select the action to be executed to handle the fault based on the correction benefit.

2. A high-voltage switchgear with remote rapid fault monitoring and processing function according to claim 1, characterized in that, The various physical paths for setting up the high-voltage switchgear include: The various physical paths include conductive paths, creepage paths, thermal diffusion paths, and mechanical execution paths.

3. A high-voltage switchgear with remote rapid fault monitoring and processing function according to claim 1, characterized in that, The method of obtaining the equivalent state quantities of each node on various physical paths using visible light images, infrared thermal images, electrical operation data, insulation environment data, and actuator state data includes: For a node in a conductive path, the temperature is extracted pixel by pixel along the conductor skeleton line corresponding to the node in the registered infrared image, and the temperature is corrected using the material emissivity to obtain the temperature corresponding to each pixel. The maximum value between the temperature of a pixel and the ambient temperature and 0 is obtained as the temperature rise of the pixel. The actual physical length of each pixel is obtained, and the product of the temperature rise of each pixel and the actual physical length is summed to obtain the conductive node feature. The conductive node feature is divided by the sum of the current of the main loop corresponding to the conductive path and the constant term to obtain the equivalent state quantity of the node in the conductive path. For a node in a creepage path, the visible light image is segmented to obtain the wetted region. The wetted region is projected onto the surface of the corresponding insulating component. The length of the maximum wetted connected region along the creepage direction on the surface of the corresponding insulating component is calculated as the ratio of the length of the maximum wetted connected region along the creepage direction to the total creepage length on the surface of the corresponding insulating component, thus obtaining the wetted continuity of the insulating surface. The mean values ​​of the TEV amplitude and the mean values ​​of the ultrasonic amplitude collected by the transient ground voltage sensor and the ultrasonic partial discharge sensor within the detection range of the node within a preset time period are normalized and then weighted and summed to obtain the partial discharge activity. The average temperature of the surface of the corresponding insulating component is extracted using an infrared thermal image. The dew point temperature is calculated using the Magnus formula based on the temperature and humidity sensor data inside the cabinet. The difference between the average temperature and the dew point temperature of the surface of the corresponding insulating component is calculated to obtain the dew point margin. The product of the wetted continuity of the insulating surface and the partial discharge activity is calculated and divided by the sum of the dew point margin and a constant term to obtain the equivalent state quantity of the node in the creepage path. For a node in a heat diffusion path, in the registered infrared thermal image, the structural boundary corresponding to the node is selected, and two adjacent regions of interest are delineated on both sides of the boundary. The average temperature of the two regions of interest is calculated, and the larger value is recorded as the average temperature on the hot side and the smaller value is recorded as the average temperature on the cold side. The difference between the average temperature on the hot side and the average temperature on the cold side is recorded as the hot-cold temperature difference. The equivalent state quantity of the node is obtained by dividing the hot-cold temperature difference corresponding to the node by the sum of the reference hot-cold temperature difference and the constant term. For a node in an organization's execution path, the time from the completion of the action of the previous node to the change of the state of this node is determined by the event sequence record and recorded as the actual response time of this node. The absolute value of the difference between the actual response time and the standard response time of this node is calculated and divided by the standard response time and a constant term of this node to obtain the equivalent state quantity of this node in the organization's execution path.

4. A high-voltage switchgear with remote rapid fault monitoring and processing function according to claim 1, characterized in that, The step of obtaining the transmission mismatch strength of a physical path based on the changes in the equivalent state variables of each node along the physical path includes: Obtain the transmission coefficient between every two adjacent nodes under fault-free conditions; obtain the transmission mismatch residual between two adjacent nodes by subtracting the equivalent state quantity of the previous node from the equivalent state quantity of the next node in a physical path and the product of the transmission coefficient between the two adjacent nodes; calculate the average of the absolute values ​​of the transmission mismatch residuals between every two adjacent nodes in the physical path, denoted as the first average value; calculate the average of the product of the transmission mismatch residuals between every two adjacent nodes in the physical path and the path direction sign, and take the absolute value to obtain the second average value; calculate the sum of the first average value and the second average value to obtain the transmission mismatch intensity of the physical path.

5. A high-voltage switchgear with remote rapid fault monitoring and processing function according to claim 1, characterized in that, The step of obtaining the cross-path induction of a physical path based on the structural coupling coefficient between a physical path and other physical paths and the transmission mismatch strength of each physical path includes: The cross-path induction of a physical path is obtained by weighting the transmission mismatch strength of the other physical paths with the structural coupling coefficient between a physical path and other physical paths, and then summing the weighted sum of the transmission mismatch strength of the physical path and the constant term.

6. A high-voltage switchgear with remote rapid fault monitoring and processing function according to claim 1, characterized in that, The method for obtaining the fault constraint coefficient based on the transmission mismatch strength and cross-path induction of a physical path includes: The fault limitation coefficient of a physical path is obtained by dividing the transmission mismatch intensity of a physical path by a first preset value and the sum of the cross-path inductance of the physical path.

7. A high-voltage switchgear with remote rapid fault monitoring and processing function according to claim 1, characterized in that, The method of obtaining the remaining path mismatch vector after the execution of an action by utilizing the transmission mismatch intensity of all paths and an action influence matrix includes: The current transmission mismatch strength vector is composed of the transmission mismatch strengths of all paths; then, the remaining path mismatch vector after the execution of an action is specifically: ， in, Indicates action The remaining path mismatch vector after the execution of this action, where k is the index of the action; Indicates action Action influence matrix; Represents the identity matrix; This represents the current propagation mismatch strength vector.

8. A high-voltage switchgear with remote rapid fault monitoring and processing function according to claim 1, characterized in that, The method of obtaining the corrected benefit of an action by utilizing the transmission mismatch strength of all paths and the remaining path mismatch vector after the action is executed includes: The correction gain of an action is obtained by comparing the difference between the sum of the absolute values ​​of the transmission mismatch intensity of each physical path and the sum of the absolute values ​​of each element in the remaining path mismatch vector after the execution of an action, and the sum of the sum of the absolute values ​​of the transmission mismatch intensity of each path and the constant term.

9. A high-voltage switchgear with remote rapid fault monitoring and processing function according to claim 1, characterized in that, The step of selecting the action to be performed based on the corrected benefit to handle the fault includes: Select the action that yields the greatest benefit to address the current fault.

Images