Detection system for action state of mechanical connecting rod of ground isolating knife switch in electric power GIS (Geographic Information System)

Abstract

According to the detection system for the action state of the mechanical connecting rod of the isolating knife switch in the electric power GIS system provided by the invention, multi-dimensional monitoring on the action state of the mechanical connecting rod of the isolating knife switch and the grounding knife switch is realized through cooperative configuration of the mechanical state sensing unit and the electrical state sensing unit; the limitation that a traditional method only takes a single-phase state as a criterion is effectively solved. And the data analysis processing unit can perform comprehensive correlation analysis on the connection state, the action synchronism and the driving load condition of the mechanical connecting rod through fusion processing of the visual image information, the position signal and the electrical parameters of the motor, so that the accuracy and the reliability of state diagnosis are remarkably improved. And the early warning output unit outputs graded early warning information based on the severity of the diagnosis result, so that operation and maintenance personnel can identify abnormal phases and fault causes in time, thereby realizing early warning and accurate positioning of potential faults of the mechanical connecting rod, and greatly improving the operation safety and maintenance efficiency of the electric power GIS system.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of high-voltage electrical equipment state monitoring and control technology, and particularly relates to a detection system for the mechanical linkage action state of a grounding knife switch in a GIS system. BACKGROUND

[0002] The B and C phase mechanical linkages of existing GIS (Geographic Information System, power informatization information system) isolation knife switches and grounding knife switches lack effective monitoring means, and only rely on single electrical signal monitoring and manual field observation to determine the action state, which leads to the risk of human error, mechanical injury hazards, and possible equipment damage, significant economic losses and other vicious events, and cannot reliably guarantee accurate determination of the action state of the knife switch and safe operation of the power station. SUMMARY

[0003] Therefore, the present application provides a detection system for the mechanical linkage action state of a grounding knife switch in a GIS system to solve the technical defects in the prior art.

[0004] Specifically, the present application provides a detection system for the mechanical linkage action state of a grounding knife switch in a GIS system, comprising: a mechanical state sensing unit configured to collect action information of the mechanical linkage of the isolation knife switch and the grounding knife switch; an electrical state sensing unit configured to collect electrical parameters of the electric motor driving the isolation knife switch and the grounding knife switch; a data analysis processing unit connected to the mechanical state sensing unit and the electrical state sensing unit and configured to receive and fuse the action information and the electrical parameters, and generate a state diagnosis result of the mechanical linkage through correlation analysis; an early warning output unit connected to the data analysis processing unit and configured to receive the state diagnosis result and output early warning information of a corresponding level based on the severity of the state diagnosis result.

[0005] In some embodiments, the mechanical state sensing unit comprises: a visual monitoring module arranged to be linked with the operating mechanism of the isolation knife switch and the grounding knife switch, for automatically moving to a preset position and collecting action video images of the mechanical linkage when the isolation knife switch or the grounding knife switch is actuated; a position sensing module comprising a position sensor arranged on each phase mechanical linkage of the isolation knife switch and the grounding knife switch, for independently collecting the position signal of each phase.

[0006] In some embodiments, the visual monitoring module is a camera mounted on a slide rail, and the slide rail is arranged such that the camera can move along the movement trajectory direction of the mechanical linkage. The position sensing module includes a first position sensor for the A-phase mechanical link, a second position sensor for the B-phase mechanical link, and a third position sensor for the C-phase mechanical link.

[0007] In some implementations, the data analysis and processing unit is configured as follows: Based on video images, an image recognition algorithm is used to determine whether there is an abnormal connection status of the mechanical link and generate abnormal connection status information. Based on the opening or closing position signal, determine the consistency of the action logic and timing of the three-phase mechanical linkage, and generate signal consistency information; Based on electrical parameters, determine whether the motor load is abnormal and generate load abnormality information; Based on connection status anomaly information, signal consistency information, and load anomaly information, generate status diagnosis results that include anomaly phase identifiers and anomaly cause inferences.

[0008] In some implementations, the data analysis and processing unit is specifically configured to, when performing the judgment on the consistency between the action logic and timing of the three-phase mechanical linkage, as follows: The action of the mechanical linkage in phase A is used as the time reference; Calculate the lag of the action time of the mechanical linkages in phases B and C relative to the time reference; If the hysteresis exceeds the preset hysteresis threshold, it is determined that there is a linkage abnormality in the corresponding B-phase or C-phase mechanical linkage.

[0009] In some implementations, the status diagnosis results include a comprehensive health status score, and the formula for calculating the comprehensive health status score includes:

[0010]

[0011]

[0012]

[0013] Where S is the overall health status score; It is a mechanical health score based on video images; It is a mechanical health score based on the consistency of action time; It is a motor drive health score based on the motor current; These are dimensionless weighting factors pre-stored in the data analysis and processing unit, used to adjust the relative importance and contribution of image evidence, time evidence, and current evidence in the comprehensive health assessment. Represents an exponential function; It is a sensitivity coefficient for image differences ≥0, which is pre-set and stored in the data analysis and processing unit. The larger the value, the more significant the impact of image differences on health scores. is the weight factor of the i-th visual feature point of the mechanical component, which is pre-stored in the data analysis and processing unit and is used to represent the importance of the feature point to the overall mechanical state; N is the total number of mechanical component feature points extracted by the image recognition algorithm from the video image of the current operation cycle and used for comparison with the standard template; i is the feature point index, i=1,2,...,N; It is the normalized difference between the i-th feature point and the standard template image in the current operation cycle. ∈[0,1]; P is an even exponent greater than 1, pre-set and stored in the data analysis and processing unit, used to amplify the impact of abnormal differences; is a sensitivity coefficient for time differences ≥ 0, which is pre-set and stored in the data analysis and processing unit. The larger the value, the more significant the impact of time differences on health scores; j is the phase index, which traverses phases B and C. It is the trigger time of the opening or closing position signal of the j-th phase (B or C phase) collected by the position sensing module; It is the trigger time of the A-phase position signal collected by the position sensing module, which is used as a time reference; It is a weighting factor for the time difference of the j-th phase position signal, which is pre-stored in the data analysis and processing unit; It is the time dead zone threshold pre-stored in the data analysis and processing unit, when At that time, the time difference is considered to be 0, which is used to suppress false alarms caused by minor jitter; It is a time scaling constant pre-stored in the data analysis and processing unit, used to normalize the time difference; max() represents the maximum value function, used for time dead zone processing; It is the peak value of the motor current during the current operating cycle, collected by the electrical state sensing unit; It is a reference value of the peak motor current under normal operating conditions, which is pre-stored in the data analysis and processing unit and obtained through statistical learning of historical normal operating data; It is pre-stored in the data analysis and processing unit, and the peak motor current relative to the data under normal operating conditions. The standard deviation is obtained through statistical learning of historical normal operating data; k is the normal range boundary coefficient, k>0, which is pre-set and stored in the data analysis and processing unit to limit the normal fluctuation range of the current value to ( ); It is a decay coefficient >0 for severe current differences, which is preset and stored in the data analysis and processing unit. The larger the value, the faster the health score drops when the current is severely abnormal.

[0014] In some implementations, the formula for calculating the normalized difference includes:

[0015] Where M is the feature vector dimension used to describe each feature point, determined by the selected image feature descriptor algorithm; k is the feature vector dimension index, traversed from 1 to M; It is the value of the feature vector of the i-th feature point in the video image of the current operation cycle in the k-th dimension, which is extracted by the data analysis and processing unit; Q is the value of the feature vector corresponding to the i-th feature point in the standard template image in the k-th dimension, which is pre-stored in the data analysis and processing unit; Q is an exponent greater than 0, which is pre-set and stored in the data analysis and processing unit to control the sensitivity of the difference. It is a normal value for adjusting the overall sensitivity, which is preset and stored in the data analysis and processing unit; It is a scaling factor used to adjust the decay rate of the exponential function, which is preset and stored in the data analysis and processing unit; exp() represents the exponential function.

[0016] In some implementations, the warning output unit is configured to divide the warning level into multiple levels based on the severity of the fault indicated by the status diagnosis results; The warning message includes the abnormal phase identifier and possible causes of the fault.

[0017] In some implementations, multiple levels include at least a warning and an emergency warning; When it is determined that only a single information source has a minor anomaly, a warning message will be output. When multiple information sources indicate a consistent serious anomaly, an emergency warning is output.

[0018] In some implementations, the linkage control logic between the visual monitoring module and the operating mechanisms of the isolating switch and the grounding switch includes: Receive operating commands from the control system of the isolating switch or grounding switch; Based on the device identifier in the operation instructions, query the pre-stored coordinate mapping table to determine the target position of the camera on the slide rail; Control the camera to move to the target location and start recording.

[0019] At least one embodiment of the present invention achieves multi-dimensional monitoring of the operational status of the mechanical linkages of isolating switches and grounding switches through the coordinated configuration of mechanical and electrical state sensing units, effectively overcoming the limitations of traditional methods that rely solely on single-phase status as the criterion. The data analysis and processing unit, by fusing visual image information, position signals, and motor electrical parameters, can comprehensively analyze the connection status, operational synchronization, and drive load of the mechanical linkages, significantly improving the accuracy and reliability of status diagnosis. The early warning output unit outputs graded early warning information based on the severity of the diagnostic results, enabling maintenance personnel to promptly identify abnormal phases and fault causes, thereby achieving early warning and precise location of potential faults in the mechanical linkages, greatly improving the safety and maintenance efficiency of the power GIS system. Attached Figure Description

[0020] Figure 1 This is a structural block diagram of a detection system for the action status of a grounding switch mechanical linkage in a power GIS system, provided by the present invention. Detailed Implementation

[0021] Many specific details are set forth in the following description to provide a full understanding of this specification. However, this specification can be implemented in many other ways than those described herein, and those skilled in the art can make similar extensions without departing from the spirit of this specification. Therefore, this specification is not limited to the specific implementations disclosed below.

[0022] The terminology used in one or more embodiments of this specification is for the purpose of describing particular embodiments only and is not intended to limit the scope of the one or more embodiments of this specification. The singular forms “a” and “the” as used in one or more embodiments of this specification and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in one or more embodiments of this specification refers to and includes any or all possible combinations of one or more associated listed items. The modifications “a” and “a plurality” as used in this disclosure are illustrative and not restrictive, and those skilled in the art will understand that they should be understood as “one or more” unless the context clearly indicates otherwise.

[0023] It should be understood that although the terms first, second, etc., may be used to describe various information in one or more embodiments of this specification, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, first may also be referred to as second without departing from the scope of one or more embodiments of this specification, and similarly, second may also be referred to as first. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to a determination."

[0024] See Figure 1 , Figure 1 This specification illustrates a structural block diagram of a detection system for the operational status of a grounding switch mechanical linkage in a power GIS system, according to some embodiments thereof. The system includes: a mechanical state sensing unit configured to collect operational information of the mechanical linkages of the isolating switch and the grounding switch; an electrical state sensing unit configured to collect electrical parameters of the motors driving the isolating switch and the grounding switch; a data analysis and processing unit connected to the mechanical state sensing unit and the electrical state sensing unit, configured to receive and fuse the operational information and electrical parameters, and generate a state diagnosis result for the mechanical linkage through correlation analysis; and an early warning output unit connected to the data analysis and processing unit, configured to receive the state diagnosis result and output an early warning message of a corresponding level based on the severity of the state diagnosis result.

[0025] A mechanical status sensing unit can refer to a combination of hardware modules responsible for collecting physical motion information of mechanical components in switching equipment. For example, this unit consists of a visual monitoring module and a position sensing module, providing direct evidence of the mechanical linkage's movement. An electrical status sensing unit can refer to sensing devices responsible for collecting electrical parameters during the operation of a drive motor. For example, this unit collects the motor's operating current through a current transformer or Hall effect sensor to reflect the magnitude and anomalies of the mechanical load. A data analysis and processing unit can refer to a central processing unit that fuses and intelligently analyzes multi-source sensor data. For example, this unit uses a microprocessor to run analysis algorithms, processing images, position signals, and current data to generate comprehensive status diagnostic conclusions. A warning output unit can refer to a human-machine interface that converts analysis results into actionable warning information. For example, this unit outputs warning levels and fault information through an HMI interface or communication interface to remind maintenance personnel to take appropriate measures. Action information can refer to characteristic data describing the motion process and state of a mechanical linkage. For example, this information consists of video streams collected by a camera and switching signals collected by a position sensor, used to record the actual operation process of the mechanical linkage. Electrical parameters refer to physical quantities characterizing the operating characteristics of a motor. For example, this parameter mainly refers to the three-phase operating current of the drive motor, used to indirectly determine whether the mechanical transmission is obstructed. Correlation analysis refers to computational methods for finding the inherent relationships between different data sources. For example, this method uses algorithms to cross-validate the correlation between image, time series, and current data to improve the accuracy and reliability of fault diagnosis. Condition diagnosis results refer to the final judgment on the health status of mechanical linkages. For example, the results can be output in the form of health indicators and abnormal phase identifiers, or as health scores, used to guide subsequent maintenance decisions. Early warning information refers to alarm content graded according to the severity of the fault. For example, this information includes the warning level, abnormal phase, and possible causes, used to achieve differentiated operation and maintenance responses.

[0026] The present invention will be further described below through a detailed embodiment: This solution addresses the issue that the position monitoring of isolating switches and grounding switches in 500kV GIS systems often uses the state of phase A as the criterion for the overall operation result, lacking effective monitoring of the mechanical linkage operation status in phases B and C. By integrating multiple detection methods and developing applications, this solution achieves automatic analysis and status judgment of the mechanical linkage operation process. The specific solution is as follows: The detection methods include: The mechanical state sensing unit collects image data through image recognition. A movable sliding rail camera system is deployed in the switch operation area. When the opening or closing operation of a certain isolating switch or grounding switch is initiated, the system automatically controls the camera to move along the sliding rail to the corresponding monitoring position of the device, capturing the entire movement process of the mechanical linkage in real time. Through video image acquisition, the rotation angle, connection status, and movement trajectory of the linkage are recorded, providing visual evidence for subsequent analysis.

[0027] The electrical status sensing unit detects the open or closed position signal of each phase. It monitors the motor current driving the A-phase disconnector and the linkages of B and C in real time, collecting current change data. When the mechanical linkage jams, detaches, or malfunctions, the motor load changes, and the current value deviates from normal operating conditions (e.g., current increases when jammed, and decreases when detached). This current signal serves as an auxiliary indicator for judging whether the linkage operation is normal.

[0028] The electrical status sensing unit performs travel signal switch detection. Metal probes are installed on the A, B, and C phase mechanical linkages of the isolating switch and grounding switch, respectively, and travel signal switches are installed at the endpoints of the corresponding probes' movement trajectories. When the switch is closed, the mechanical linkage drives the metal probes to rotate, triggering the travel signal switches to output a "closed position signal." When the switch is open, the linkage rotates in the opposite direction, and the probes move accordingly, causing the travel signal switches to output a "open position signal." Signals are collected independently for each of the three phases, enabling accurate detection of whether the linkages have reached their designated positions. The signals for phases B and C can individually reflect the status of their respective linkages.

[0029] Data analysis and processing include: The collected information is then fed into the data analysis and processing unit for analysis. This unit receives and integrates the motion information collected by the mechanical state sensing unit and the electrical parameters collected by the electrical state sensing unit, and generates a state diagnosis result for the mechanical linkage through correlation analysis.

[0030] Video image analysis: The system processes the video captured by the sliding rail camera. Through image comparison technology (comparing with a normal motion trajectory template), it can manually determine whether the motion trajectory of the mechanical link meets the standard and whether the connection is intact. It can also identify abnormal phenomena such as jamming and detachment. The system can further process the image data and finally generate a status diagnosis result for the mechanical link.

[0031] Travel signal verification: Analyze the output status of the travel signal switches of phases A, B, and C. If a phase fails to output the corresponding opening or closing signal within the specified time (e.g., no "closed position signal" when closing), it can be manually determined that the linkage action is not in place. Alternatively, the opening or closing position signal can be further processed, and finally, a status diagnosis result for the mechanical linkage can be generated.

[0032] Current signal analysis: By comparing the real-time current of the motor with the current curve under normal operating conditions, when the current value exceeds the preset threshold range, an abnormal warning is triggered. Combined with video and travel signals, the cause of the fault is further located (such as jamming causing a sudden increase in current).

[0033] Early warning output includes: The early warning output unit receives the status diagnosis results generated by the data analysis and processing unit, and outputs corresponding level early warning information based on the severity of the status diagnosis results. The system classifies the early warning level into multiple levels according to the severity of the diagnostic results. The early warning information includes the phase identifier of the anomaly and possible causes of the fault, providing clear handling guidance for maintenance personnel.

[0034] The beneficial effects of one of the embodiments in this specification include at least the following: Through the coordinated configuration of the mechanical state sensing unit and the electrical state sensing unit, multi-dimensional monitoring of the operational status of the mechanical linkages of isolating switches and grounding switches is achieved, effectively overcoming the limitations of traditional methods that rely solely on single-phase status as the criterion. The data analysis and processing unit, by fusing visual image information, position signals, and motor electrical parameters, can comprehensively analyze the connection status, operational synchronization, and drive load of the mechanical linkages, significantly improving the accuracy and reliability of status diagnosis. The early warning output unit outputs graded early warning information based on the severity of the diagnostic results, enabling maintenance personnel to promptly identify abnormal phases and fault causes, thereby achieving early warning and precise location of potential faults in the mechanical linkages, greatly improving the safety and maintenance efficiency of the power GIS system.

[0035] In some implementations, the mechanical state sensing unit includes: a visual monitoring module, configured to be linked with the operating mechanism of the isolating switch and the grounding switch, for automatically moving to a preset position and acquiring video images of the mechanical linkage's movement when the isolating switch or the grounding switch is activated; and a position sensing module, including position sensors installed on each phase mechanical linkage of the isolating switch and the grounding switch, for independently acquiring the opening or closing position signal of each phase.

[0036] The operating mechanism of isolating and grounding switches can refer to the mechanical transmission device that controls the opening and closing of the contacts of switching equipment. For example, this mechanism consists of a motor, reduction gears, and connecting rods, receiving control commands to drive the contact movement for safe connection or isolation of the circuit. A visual monitoring module can refer to a camera device used to capture visual images of mechanical components. For example, this module uses a high-definition industrial camera, mounted on a slide rail and driven by a servo motor, to acquire high-definition video of the mechanical linkage's operation. A preset position can refer to a pre-set spatial coordinate point to achieve the optimal viewing angle. For example, this position is determined by calibrating key points on the mechanical linkage's movement trajectory, ensuring the camera can record the operation process completely and clearly. Action video images can refer to a continuous frame sequence recording the movement of mechanical linkages. For example, this image is acquired as a 1080p resolution video stream encoded in H.264 format, providing intuitive visual evidence of the action process. A position sensing module can refer to a set of sensors that detects and outputs the physical position of mechanical components. For example, this module uses a high-precision angle sensor or linear displacement sensor to accurately measure the real-time position of each connecting rod. Each phase mechanical linkage can refer to an independent power transmission linkage for each phase in a three-phase AC system. For example, each phase in phases A, B, and C is equipped with an independent insulated linkage mechanism for phase-by-phase operation and monitoring. The open or closed position signal can refer to a digital signal indicating whether the switch contacts are open or closed. This signal can be a dry contact signal or a PROFINET IO fieldbus digital output, used to accurately determine the final operating state of the switch.

[0037] As a concrete example: During the tripping operation of a 330kV GIS device, upon receiving the tripping command, the control unit simultaneously sends a linkage trigger signal to the visual monitoring module. A high-definition network camera (using a 2-megapixel CMOS sensor) mounted on a precision slide rail, driven by a stepper motor, rapidly moves to a preset observation position (e.g., 1.5 meters from the axis of the B connecting rod) based on a pre-stored coordinate mapping table, and begins acquiring video images of the mechanical linkage's movement at a rate of 25 frames per second. Simultaneously, absolute encoders (multi-turn absolute encoders with 17-bit resolution) installed on the shafts of each phase of the A, B, and C mechanical linkage monitor the linkage rotation angle in real time. When the angle reaches a predetermined tripping position threshold (e.g., 88 degrees), a tripping position signal is sent to the position sensing module via the IO-Link interface. The position sensing module performs photoelectric isolation and signal conditioning on the three-phase signals, then uploads them to the data analysis and processing unit via the Modbus TCP protocol, providing accurate timestamp data for analyzing the consistency of the three-phase actions.

[0038] By combining vision and position sensing, multi-angle precise monitoring of the mechanical linkage operation process is achieved, providing sufficient data support for comprehensive analysis and judgment of equipment status, and effectively improving the comprehensiveness and reliability of status detection.

[0039] In some implementations, the visual monitoring module is a camera mounted on a slide rail, the slide rail being arranged so that the camera can move along the movement trajectory of the mechanical link; the position sensing module includes a first position sensor for the A-phase mechanical link, a second position sensor for the B-phase mechanical link, and a third position sensor for the C-phase mechanical link.

[0040] A slide rail can refer to a mechanical guide rail device that provides a directional movement path for a camera. For example, the slide rail may use a linear guide rail structure and be driven by a stepper motor for precise positioning, guiding the camera to move along a predetermined trajectory. The direction of motion trajectory can refer to the direction of the mechanical link's movement path during operation. For example, this direction is determined by analyzing the kinematic model of the link's hinge point, used to plan the optimal tracking path for the camera. A-phase mechanical link can refer to the mechanical transmission link containing phase A in a three-phase power system. For example, this link may be made of insulating material and connect the moving contact of phase A to the operating mechanism, used to transmit the operating torque of phase A. The first position sensor can refer to the first position detection device installed on the phase A mechanical link. For example, this sensor may use an absolute encoder or a linear variable differential transformer (LVDT), used to accurately measure the angular or linear displacement of the phase A link. A phase B mechanical link can refer to the mechanical transmission link containing phase B in a three-phase power system. For example, this link may be structurally symmetrical to phase A but electrically isolated, used to independently transmit the operating torque of phase B. The second position sensor can refer to a position detection device installed on the B-phase mechanical link. For example, this sensor may be the same model and configuration as the first position sensor, used to achieve synchronous monitoring of the position of the B-phase link. The C-phase mechanical link can refer to the mechanical transmission rod containing the C-phase in a three-phase power system. For example, this link, together with phases A and B, forms a complete three-phase mechanical transmission system, used to transmit the operating torque of the C-phase. The third position sensor can refer to a position detection device installed on the C-phase mechanical link. For example, this sensor may use the same specifications as the first position sensor, used to provide accurate position feedback for the C-phase link.

[0041] As a concrete example: In a 750kV GIS substation, the camera of the visual monitoring module is mounted on a 2.5-meter-long linear slide rail, which is arranged parallel to the movement trajectory of the isolating switch operating mechanism. Upon receiving a closing operation command, a servo motor inside the slide rail drives the camera to move at a speed of 0.3 meters per second, ensuring that the lens is always focused on the moving mechanical linkage. Simultaneously, a first position sensor (using a multi-turn absolute encoder and RS485 interface) mounted on the A-phase mechanical linkage shaft, a second position sensor on the B-phase mechanical linkage, and a third position sensor (all using encoders of the same specification) on the C-phase mechanical linkage collect real-time rotation angle data of each linkage. These sensors transmit position signals to the data analysis and processing unit via a PROFIBUS-DP fieldbus at a sampling frequency of 100 times per second, providing high-precision timing data for three-phase action consistency analysis.

[0042] By using a sliding rail-guided camera movement and a coordinated configuration of three-phase position sensors, comprehensive monitoring of the mechanical linkage's motion process is achieved. This provides a complete data foundation for analyzing and judging the synchronicity of the three-phase operation and the mechanical state, effectively ensuring the reliability and accuracy of the detection system.

[0043] In some implementations, the data analysis and processing unit is configured to: determine whether there is an abnormal connection status of the mechanical linkage based on video images using image recognition algorithms, and generate abnormal connection status information; determine the consistency of the action logic and timing of the three-phase mechanical linkage based on the open or close position signal, and generate signal consistency information; determine whether the load of the motor is abnormal based on electrical parameters, and generate load abnormality information; and generate a status diagnosis result containing abnormal phase identifiers and abnormality cause inferences based on the abnormal connection status information, signal consistency information, and load abnormality information.

[0044] Video images can refer to continuous frame sequences of data acquired by a visual monitoring module. For example, the image might be stored as a 1080p resolution video stream in H.264 encoding format, used to record the complete operation process of a mechanical linkage. Image recognition algorithms can refer to computer programs that extract features and analyze patterns from video images. For example, such algorithms might use a Convolutional Neural Network (CNN) model based on deep learning to automatically identify the state of mechanical components in images. Abnormal connection status can refer to abnormal physical states at the connection points of mechanical linkages. This can be determined by analyzing the relative position changes and deformation of feature points, used to detect loosening, deformation, or damage to the linkage. Abnormal connection status information can refer to information about abnormal mechanical connection status obtained through image analysis. For example, the result might be expressed as a Boolean value or confidence score, used to indicate whether there is a loosening or deformation fault in the mechanical linkage. Opening or closing position signals can refer to digital signals indicating whether the switch contacts are in an open or closed state. For example, this signal could be a dry contact signal or a PROFINET IO fieldbus digital output, used to accurately determine the final operating state of the switch. The motion logic and timing consistency of three-phase mechanical linkages refer to the synchronous motion characteristics that the three linkages A, B, and C should follow during operation. For example, this can be analyzed by comparing the trigger time difference of the three-phase signals with a preset threshold to detect abnormal inter-phase linkage. Signal consistency information refers to the information obtained from analyzing the synchronization of the three-phase mechanical linkage actions. For example, the result includes time difference data and consistency judgment flags, reflecting the coordinated performance of the three-phase operation. Motor load refers to the mechanical resistance experienced by the drive mechanism during operation. For example, this can be judged by monitoring the deviation of the motor's operating current from its rated value, used to detect mechanical transmission jamming or overload faults. Load anomaly information refers to information on whether the motor load status is normal. For example, the result is output in the form of anomaly level or deviation degree numerical values, used to indicate whether there is an overload or jamming problem in the mechanical transmission. Abnormal phase identification refers to identification information used to indicate the specific abnormal phase. For example, this can be determined by analyzing the differences between the sensor data of each phase and the reference value, used to accurately locate the phase where the fault occurred. Anomaly cause inference can refer to the analysis and judgment of the root cause of a fault based on multi-source data. For example, it can be used to conduct a comprehensive diagnosis by associating image anomalies, time deviations, and current anomalies, in order to guide maintenance personnel to take the correct handling measures. Condition diagnosis results can refer to the final assessment conclusion of the overall condition of the mechanical linkage. For example, the result includes a health score, abnormal phase information, and fault cause analysis, in order to provide a complete equipment condition assessment report.

[0045] As a specific example: During the operation of a 550kV GIS device, the system first analyzes the connection status of the mechanical linkage based on video images using the YOLOv5 image recognition algorithm, generating connection status anomaly information (such as "connection normal" or "phase B connection loose"). Simultaneously, based on the trip position signal, it calculates the time series differences of phases A, B, and C, generating signal consistency information (such as "time difference within allowable range" or "phase C lags by 25ms"). Based on motor current parameters, it analyzes the deviation between peak current and normal values, generating load anomaly information (such as "load normal" or "load exceeds standard by 30%). Finally, the system integrates these three results to generate a status diagnosis result containing anomaly phase identification (such as "phase C") and anomaly cause inference (such as "mechanical transmission mechanism jammed"), which is then uploaded to the monitoring system via the IEC 61850 protocol.

[0046] By parallel analysis and intelligent fusion of multi-source data, a comprehensive and accurate diagnosis of the mechanical linkage status is achieved. This provides not only specific anomaly location information but also fault cause analysis, offering a complete and reliable technical basis for operation and maintenance decisions and significantly improving the level of intelligence in equipment status management.

[0047] In some implementations, when the data analysis and processing unit performs the judgment on the consistency of the action logic and time of the three-phase mechanical linkage, it is specifically configured to: take the action of the A-phase mechanical linkage as the time reference; calculate the lag of the action time of the B-phase and C-phase mechanical linkages relative to the time reference; if the lag exceeds a preset lag threshold, determine that the corresponding B-phase or C-phase mechanical linkage has a linkage abnormality.

[0048] A time reference can refer to the starting point used for measuring and comparing time series. For example, the rising edge of the trigger signal of the phase A position sensor can be used as the absolute time zero point to provide a unified timing reference for three-phase time consistency analysis. The action time of the mechanical linkages in phases B and C can refer to the duration from the start of movement to reaching the specified position for phases B and C. This can be measured, for example, by recording the timestamps of the position signal transitioning from an invalid to an active state, to quantify the operational timing characteristics of each phase. Lag can refer to the time delay of the subsequent phase's action time relative to the reference phase. For example, calculating the phase B timestamp T... B With A's timestamp T A The difference Δt BA This information is obtained for evaluating the synchronization performance of phase-to-phase actions. The preset hysteresis threshold can refer to the critical time difference value for determining whether an action is abnormal. For example, it can be set to a fixed value, such as 25 milliseconds, based on the mechanical characteristics of the equipment and historical data, to distinguish between normal timing deviations and fault-related hysteresis. Linkage anomalies can refer to a fault state in which the normal coordinated motion relationship between multi-phase mechanical links is lost. For example, when the hysteresis of a certain phase exceeds the threshold, this anomaly is determined to exist, and it is used to detect connection failures or jamming problems in mechanical transmission mechanisms.

[0049] As a specific example: During the closing operation of a 220kV GIS device, the data analysis and processing unit uses the position signal of the mechanical linkage of phase A as the trigger time T. A Using this as a time reference, the system clock count value at that moment is recorded. The unit then acquires the trigger times T of the B-phase and C-phase position signals. B and T C Calculate the lag Δt relative to the time base. B =T B -T A and Δt C =T C -T A These time differences are compared with a preset lag threshold (e.g., 20ms). If the lag of phase B is Δt... B If the threshold is exceeded, it is determined that there is an abnormal linkage in phase B mechanical linkage. At the same time, the unit records the timestamp of the abnormal event, phase information, and the value exceeding the limit, generates a status report containing the diagnostic conclusion of "phase B action is not synchronized", and sends it to the monitoring system through the early warning output unit.

[0050] By establishing precise time benchmarks and quantitative evaluation mechanisms, accurate monitoring and anomaly identification of three-phase mechanical linkage performance have been achieved, providing an effective technical means for early detection of mechanical transmission faults and ensuring the reliable operation of switchgear.

[0051] In some implementations, the status diagnosis results include a comprehensive health status score, and the formula for calculating the comprehensive health status score includes:

[0052]

[0053]

[0054]

[0055] Wherein, S is the overall health status score. The closer the overall health status score is to 1, the healthier the mechanical link is, and the closer it is to 0, the higher the risk of failure. It is a mechanical health score based on video images; It is a mechanical health score based on the consistency of action time; It is a motor drive health score based on the motor current; These are dimensionless weighting factors pre-stored in the data analysis and processing unit, used to adjust the relative importance and contribution of image evidence, time evidence, and current evidence in the comprehensive health assessment. Represents an exponential function; It is a sensitivity coefficient for image differences ≥0, which is pre-set and stored in the data analysis and processing unit. The larger the value, the more significant the impact of image differences on health scores. is the weight factor of the i-th visual feature point of the mechanical component, which is pre-stored in the data analysis and processing unit and is used to represent the importance of the feature point to the overall mechanical state; N is the total number of mechanical component feature points extracted by the image recognition algorithm from the video image of the current operation cycle and used for comparison with the standard template; i is the feature point index, i=1,2,...,N; It is the normalized difference between the i-th feature point and the standard template image in the current operation cycle. ∈[0,1]; P is an even exponent greater than 1, pre-set and stored in the data analysis and processing unit, used to amplify the impact of abnormal differences; is a sensitivity coefficient for time differences ≥ 0, which is pre-set and stored in the data analysis and processing unit. The larger the value, the more significant the impact of time differences on health scores; j is the phase index, which traverses phases B and C. It is the trigger time of the opening or closing position signal of the j-th phase (B or C phase) collected by the position sensing module; It is the trigger time of the A-phase position signal collected by the position sensing module, which is used as a time reference; It is a weighting factor for the time difference of the j-th phase position signal, which is pre-stored in the data analysis and processing unit; It is the time dead zone threshold pre-stored in the data analysis and processing unit, when At that time, the time difference is considered to be 0, which is used to suppress false alarms caused by minor jitter; It is a time scaling constant pre-stored in the data analysis and processing unit, used to normalize the time difference; max() represents the maximum value function, used for time dead zone processing; It is the peak value of the motor current during the current operating cycle, collected by the electrical state sensing unit; It is a reference value of the peak motor current under normal operating conditions, which is pre-stored in the data analysis and processing unit and obtained through statistical learning of historical normal operating data; It is pre-stored in the data analysis and processing unit, and the peak motor current relative to the data under normal operating conditions. The standard deviation is obtained through statistical learning of historical normal operating data; k is the normal range boundary coefficient, k>0, which is pre-set and stored in the data analysis and processing unit to limit the normal fluctuation range of the current value to ( ); It is a decay coefficient >0 for severe current differences, which is preset and stored in the data analysis and processing unit. The larger the value, the faster the health score drops when the current is severely abnormal.

[0056] The sensitivity coefficient can refer to a parameter that controls how sensitive the health score is to changes in variability values, for example... and A larger coefficient value indicates a more significant decrease in score due to the same difference, used to adjust the sensitivity requirements of different pieces of evidence. Visual feature points of mechanical components refer to the characteristic locations in an image that represent the key states of the mechanical component, such as corner points and edges extracted using SIFT or ORB algorithms, used for image comparison and state assessment. Standard template images refer to reference images collected under normal conditions as a comparison benchmark, such as clear images of the connecting rods at the open and closed positions collected during equipment debugging, used to provide a benchmark for image comparison. Normalized difference refers to a metric that normalizes the original difference values ​​to the range of 0-1, such as obtaining D by processing the feature vector distance using an exponential function. i The value is used to eliminate the influence of dimensions for comprehensive evaluation. Even exponents can refer to powers of even numbers such as 2, 4, 6, etc. For example, using an exponent of P=2 or P=4 in calculations can amplify the impact of abnormal differences on health scores. The time dead zone threshold refers to the allowable range of time differences to be ignored, such as setting τ. deadzone =5ms ignores minor time jitter to prevent false alarms caused by normal fluctuations. The time scaling constant can refer to the scaling factor used for time difference normalization, for example, setting τ scale =100ms scales the time difference to an appropriate range to maintain consistency across evaluation metrics. Standard deviation can refer to a statistic that measures the dispersion of data, such as σ. I This indicates the peak current relative to I during normal operation. normal The fluctuation range is used to define the normal variation range of the current. The normal range boundary coefficient can refer to a parameter that defines the boundary multiple of the normal fluctuation range; for example, setting k=3 defines I... normal ±3σ I This is within the normal range, used to distinguish between normal fluctuations and abnormal situations. The attenuation coefficient can refer to a parameter that controls the rate of score decrease during severe anomalies; for example, a larger γ value indicates a more severe current anomaly in H. current The faster the decline, the more sensitive it is to severe anomalies.

[0057] By using a multi-parameter configurable weighted fusion model, a refined and quantitative assessment of the mechanical linkage status is achieved, which not only ensures the accuracy of the assessment but also provides flexible adjustment capabilities, thus providing a scientific basis for equipment status management.

[0058] In some implementations, the formula for calculating the normalized difference includes:

[0059] Where M is the feature vector dimension used to describe each feature point, determined by the selected image feature descriptor algorithm; k is the feature vector dimension index, traversed from 1 to M; It is the value of the feature vector of the i-th feature point in the video image of the current operation cycle in the k-th dimension, which is extracted by the data analysis and processing unit; Q is the value of the feature vector corresponding to the i-th feature point in the standard template image in the k-th dimension, which is pre-stored in the data analysis and processing unit; Q is an exponent greater than 0, which is pre-set and stored in the data analysis and processing unit to control the sensitivity of the difference. It is a normal value for adjusting the overall sensitivity, which is preset and stored in the data analysis and processing unit; It is a scaling factor used to adjust the decay rate of the exponential function, which is preset and stored in the data analysis and processing unit; exp() represents the exponential function.

[0060] By introducing a normalized difference calculation method based on an exponential decay model, the multidimensional differences of image feature vectors are transformed into a unified standardized metric, effectively eliminating the influence of dimensions between different feature dimensions and making the visual feature comparison results more objective and accurate. Combined with adjustable sensitivity parameters and exponential control factors, differentiated responses to subtle anomalies and significant defects are achieved, improving the accuracy and adaptability of image recognition algorithms for detecting abnormalities in mechanical connection states and providing reliable visual evidence support for comprehensive state assessment.

[0061] Feature descriptor algorithms refer to computer vision algorithms that extract and describe feature points from images. For example, the SIFT (Scale-Invariant Feature Transform) algorithm is used to generate feature descriptions that are invariant to scale and rotation. Scaling factors refer to the proportional coefficients used to adjust the decay rate of the exponential function. For instance, a larger λ value means that the difference increases faster with increasing feature distance, and it is used to control the response characteristics of difference calculation.

[0062] In some implementations, the warning output unit is configured to divide the warning level into multiple levels based on the severity of the fault indicated by the status diagnosis results; the warning information includes the phase identifier of the anomaly and the possible cause of the fault.

[0063] Fault severity can refer to the fault level classified according to the comprehensive health status score and other diagnostic parameters. For example, comparing the health index value with a preset threshold can divide it into three levels: minor, moderate, and severe, to distinguish fault states of different urgency. Warning level can refer to an alarm level system classified according to the severity of the fault. For example, using color coding, it can be divided into four levels: green (normal), yellow (indicating), orange (warning), and red (urgent), to achieve a differentiated warning response mechanism. Abnormal phase identification can refer to the specific phase information indicating the fault. For example, analyzing abnormal data from each phase sensor can determine whether the faulty phase is A, B, C, or a combination thereof, to accurately locate the equipment where the fault occurred.

[0064] As a concrete example: During condition diagnosis, the data analysis and processing unit determines the severity of the fault based on the comprehensive health status score. When the S value is between 0.6 and 0.8, it is judged as a minor anomaly, and a warning is output; when the S value is below 0.6, it is judged as a serious anomaly, and an emergency warning is output. The warning information is displayed through the HMI interface, including the specific phase identifier of the anomaly (such as "Phase B anomaly") and possible fault cause inferences (such as "loose mechanical linkage connection"). Simultaneously, the system uploads the warning level and detailed information to the station control system via the IEC 61850 protocol, triggering the corresponding maintenance procedures. Different warning levels correspond to different response time limits: warnings require inspection within 24 hours, and emergency warnings require on-site handling within two hours.

[0065] By establishing a tiered early warning mechanism and a detailed fault description system, we have achieved refined management and rapid response to abnormal equipment conditions. This has not only avoided excessive alarms but also ensured the timely handling of major faults, significantly improving the operation and maintenance efficiency and safety and reliability of power equipment.

[0066] In some implementations, multiple levels include at least a warning and an emergency warning; a warning is output when it is determined that only a single information source has a minor anomaly; an emergency warning is output when it is determined that multiple information sources indicate a consistent serious anomaly.

[0067] Informative alerts can refer to the lowest level alarms targeting minor anomalies, such as when a single sensor's data slightly deviates from the normal range, alerting maintenance personnel to changes in equipment status. Emergency alerts can refer to the highest level alarms targeting severe anomalies, such as when multiple sensors simultaneously detect severe anomalies, requiring immediate emergency response measures. A single information source can refer to data from a single sensor or monitoring channel, such as a judgment based solely on current or image data, describing diagnostic results based on single evidence. Minor anomalies can refer to initial anomalies where equipment parameters deviate from the normal range but have not yet affected functionality, such as a current slightly exceeding a threshold or minor changes in image features, identifying anomalies requiring attention but not immediate intervention. Multiple information sources can refer to multiple data sources from different sensors or monitoring channels, such as comprehensive evidence including current, image, and location data, improving the reliability of diagnostic conclusions. Consistent severe anomalies can refer to multiple independent data sources simultaneously indicating the same severe fault state, such as current anomalies, image anomalies, and time anomalies occurring simultaneously and pointing to the same phase, confirming the occurrence of a major fault.

[0068] As a concrete example: In the early warning level classification, the system sets multiple warning levels. When the data analysis and processing unit determines that only the current sensor detects a minor anomaly, while the image and location data are normal, it outputs a prompt warning to remind maintenance personnel to pay attention to changes in motor load. When the unit determines that multiple information sources, including current, image, and location, have detected serious anomalies, and these anomalies point to the same phase, forming consistent evidence of serious anomalies, it outputs an emergency warning. The system simultaneously issues warning information through audible and visual alarms and the monitoring system. Prompt warnings use yellow indicator lights and text prompts, while emergency warnings use flashing red lights and buzzer alarms to ensure that different levels of warnings receive appropriate attention and response.

[0069] By establishing a multi-level early warning mechanism and a judgment logic based on evidence consistency, we have achieved refined hierarchical management of abnormal equipment status. This avoids overreacting to minor anomalies while ensuring timely early warning of serious faults, thus optimizing the allocation of operation and maintenance resources and the efficiency of emergency response.

[0070] In some implementations, the linkage control logic between the visual monitoring module and the operating mechanisms of the isolating switch and the grounding switch includes: receiving operating instructions from the control system of the isolating switch or the grounding switch; querying a pre-stored coordinate mapping table according to the device identifier in the operating instructions to determine the target position of the camera on the slide rail; and controlling the camera to move to the target position and start recording.

[0071] The linkage control logic can refer to the program control rules that coordinate the visual monitoring module and the operating mechanism to work together. For example, this logic can be implemented in the form of a state machine, defining the correspondence between various operation commands and camera movement to ensure the synchronization of visual monitoring and equipment operation. The control system's operation commands can refer to equipment control commands issued by the superior monitoring system, such as opening and closing control commands transmitted via the IEC 61850 MMS protocol, used to trigger specific operations on the switching equipment. Equipment identification can refer to the unique coding information that identifies a specific switching equipment, such as using a combination of the logical device name (LDName) and logical node name (LNName) based on the IEC 61850 standard, used to accurately locate the specific equipment to be operated. The pre-stored coordinate mapping table can refer to a data table that stores the correspondence between equipment identification and camera position. For example, this mapping table can be stored in the form of a database table, containing fields such as equipment ID, preset coordinates (X, Y, Z values), used for quickly querying the target position. The target position can refer to the specified spatial coordinates that the camera needs to move to on the slide rail, such as a precise position point (x, y, z) defined by a three-dimensional coordinate system, used to ensure the camera obtains the best viewing angle.

[0072] As a concrete example: When the control system of the isolating switch sends an operation command via a GOOSE message, the linkage control logic first parses the device identifier in the message, such as the "QB1 Disconnect" command. The system queries the pre-stored coordinate mapping table and finds the corresponding target position coordinates (1250, 560, 90) based on the device identifier "QB1". Then, it controls the slide rail servo drive via the PROFINET protocol to precisely position the camera at the target location. After the camera is in place, the system sends a StartRecording command via the ONVIF protocol to start recording, and simultaneously sends a signal to the operating mechanism to allow operation. The entire linkage process is completed within milliseconds, ensuring that the camera is in place and has started recording before the mechanical linkage begins to move, thus capturing the complete operation process video.

[0073] Through an intelligent linkage control mechanism, precise synchronization between visual monitoring and equipment operation is achieved, ensuring that no key operation processes are missed in recording. This provides complete and reliable visual evidence support for subsequent status analysis and fault diagnosis, thereby improving the practicality and reliability of the monitoring system.

[0074] The preferred embodiments disclosed above are merely illustrative of this specification. The optional embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this invention. These embodiments are selected and specifically described in this specification to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize this specification. This specification is limited only by the claims and their full scope and equivalents.

Claims

1. A detection system for the mechanical linkage operation status of isolating switches and grounding switches in a power GIS system, characterized in that, include: The mechanical state sensing unit is configured to collect the motion information of the mechanical linkages of the isolating switch and the grounding switch; An electrical status sensing unit is configured to collect electrical parameters of the motors driving the isolating switch and the grounding switch; A data analysis and processing unit is connected to the mechanical state sensing unit and the electrical state sensing unit, and is configured to receive and fuse the motion information and the electrical parameters, and generate a state diagnosis result about the mechanical link through correlation analysis; The early warning output unit is connected to the data analysis and processing unit and is configured to receive the status diagnosis results and output early warning information of the corresponding level based on the severity of the status diagnosis results.

2. The system according to claim 1, characterized in that, The mechanical state sensing unit includes: The visual monitoring module is configured to be linked with the operating mechanism of the isolating switch and the grounding switch, and is used to automatically move to a preset position and collect video images of the mechanical linkage when the isolating switch or the grounding switch is activated; The position sensing module includes position sensors installed on the mechanical linkage of each phase of the isolating switch and the grounding switch, which are used to independently collect the open or closed position signal of each phase.

3. The system according to claim 2, characterized in that, The visual monitoring module is a camera mounted on a slide rail, and the slide rail is arranged so that the camera can move along the movement trajectory of the mechanical linkage. The position sensing module includes a first position sensor for the A-phase mechanical link, a second position sensor for the B-phase mechanical link, and a third position sensor for the C-phase mechanical link.

4. The system according to claim 3, characterized in that, The data analysis and processing unit is configured as follows: Based on the video image, an image recognition algorithm is used to determine whether there is an abnormal connection status of the mechanical link, and abnormal connection status information is generated. Based on the opening or closing position signal, determine the consistency of the action logic and timing of the three-phase mechanical linkage, and generate signal consistency information. Based on the electrical parameters, determine whether the load on the motor is abnormal and generate load abnormality information; Based on the connection status anomaly information, signal consistency information, and load anomaly information, a status diagnosis result containing anomaly phase identifiers and anomaly cause inferences is generated.

5. The system according to claim 4, characterized in that, When the data analysis and processing unit performs the judgment on the consistency between the action logic and timing of the three-phase mechanical linkage, it is specifically configured as follows: The action of the mechanical linkage in phase A is used as the time reference; Calculate the lag of the action time of the mechanical linkages in phase B and phase C relative to the time reference; If the hysteresis exceeds the preset hysteresis threshold, it is determined that the corresponding B-phase or C-phase mechanical linkage has a linkage abnormality.

6. The system according to claim 4, characterized in that, The status diagnosis result includes a comprehensive health status score, and the formula for calculating the comprehensive health status score includes: Wherein, S is the overall health status score; It is a mechanical health score based on video images; It is a mechanical health score based on the consistency of action time; It is a motor drive health score based on the motor current; These are dimensionless weighting factors pre-stored in the data analysis and processing unit, used to adjust the relative importance and contribution ratio of image evidence, time evidence, and current evidence in the comprehensive health assessment; Represents an exponential function; It is a sensitivity coefficient for image differences ≥0, which is pre-set and stored in the data analysis and processing unit. The larger the value, the more significant the impact of image differences on health scores. is the weight factor of the i-th mechanical component visual feature point pre-stored in the data analysis and processing unit, used to represent the importance of the feature point to the overall mechanical state; N is the total number of mechanical component feature points extracted by the image recognition algorithm from the video image of the current operation cycle for comparison with the standard template; i is the feature point index, i=1,2,...,N; It is the normalized difference between the i-th feature point and the standard template image in the current operation cycle. ∈[0,1]; P is an even exponent greater than 1, which is preset and stored in the data analysis and processing unit to amplify the impact of abnormal differences; is a sensitivity coefficient for time differences ≥ 0, which is pre-set and stored in the data analysis and processing unit. The larger the value, the more significant the impact of time differences on health scores; j is the phase index, traversing phases B and C; It is the trigger time of the opening or closing position signal of the j-th phase (B or C phase) collected by the position sensing module; It is the trigger time of the A-phase position signal collected by the position sensing module, which is used as a time reference; It is a weighting factor for the time difference of the j-th phase position signal, which is pre-stored in the data analysis and processing unit; It is the time dead zone threshold pre-stored in the data analysis and processing unit, when At that time, the time difference is considered to be 0, which is used to suppress false alarms caused by minor jitter; It is a time scaling constant pre-stored in the data analysis and processing unit, used to normalize the time difference; max() represents the maximum value function, used for time dead zone processing; It is the peak value of the motor current during the current operating cycle collected by the electrical state sensing unit; It is a reference value of the peak motor current under normal operating conditions, which is pre-stored in the data analysis and processing unit and obtained through statistical learning of historical normal operating data; It is pre-stored in the data analysis and processing unit, and the peak value of the motor current under normal operating conditions is relative to... The standard deviation is obtained through statistical learning of historical normal operating data; k is the normal range boundary coefficient, k>0, which is pre-set and stored in the data analysis and processing unit to limit the normal fluctuation range of the current value to ( ); It is a decay coefficient >0 for severe current differences, which is preset and stored in the data analysis and processing unit. The larger the value, the faster the health score drops when the current is severely abnormal.

7. The system according to claim 6, characterized in that, The formula for calculating the normalized difference includes: Where M is the feature vector dimension used to describe each feature point, determined by the selected image feature descriptor algorithm; k is the feature vector dimension index, traversed from 1 to M; It is the value of the feature vector of the i-th feature point in the current operation cycle video image in the k-th dimension, which is extracted by the data analysis and processing unit; Q is the value of the feature vector corresponding to the i-th feature point in the standard template image in the k-th dimension, which is pre-stored in the data analysis and processing unit; Q is an exponent greater than 0, which is pre-set and stored in the data analysis and processing unit to control the sensitivity of the difference. It is a normal value for adjusting the overall sensitivity, which is preset and stored in the data analysis and processing unit; It is a scaling factor used to adjust the decay rate of the exponential function, which is preset and stored in the data analysis and processing unit; exp() represents the exponential function.

8. The system according to claim 7, characterized in that, The warning output unit is configured to divide the warning level into multiple levels based on the severity of the fault indicated by the status diagnosis result; The warning information includes the abnormal phase identifier and possible causes of the fault.

9. The system according to claim 8, characterized in that, The multiple levels include at least advisory warnings and emergency warnings; When it is determined that only a single information source has a minor anomaly, the aforementioned warning is output; When multiple information sources indicate a consistent serious anomaly, the emergency warning is output.

10. The system according to claim 3, characterized in that, The linkage control logic between the visual monitoring module and the operating mechanisms of the isolating switch and the grounding switch includes: Receive operating instructions from the control system of the isolating switch or grounding switch; Based on the device identifier in the operation instruction, the pre-stored coordinate mapping table is queried to determine the target position of the camera on the slide rail; Control the camera to move to the target location and start recording.

Images