Remote monitoring and fault diagnosis method and system for 10kV switch cabinet

Abstract

The invention relates to a remote monitoring and fault diagnosis method and system for 10kV switch cabinets. The remote monitoring and fault diagnosis method comprises the following steps: S1, synchronously acquiring infrared thermal images, visible light images and electrical parameters and environmental parameters related to equipment operation states in a plurality of switch cabinets; s2, converging data from different switch cabinets, performing protocol conversion, preprocessing and timestamp alignment on the converged data, and extracting key feature data from the converged data to form a structured data packet; and S3, based on the structured data packet, generating an equipment health state assessment result and hierarchical early warning information by using a multi-channel image fusion algorithm and a multi-level early warning model of a state stability index. The method has the advantages that early warning can be given out at the initial stage of equipment abnormity, so that fault processing is changed from post-maintenance to pre-prevention, and the equipment fault rate is reduced; the unmanned automatic inspection is realized, the manual inspection frequency is reduced, the labor cost is reduced, and the monitoring precision and the diagnosis reliability are greatly improved.

Description

Technical Field

[0001] This invention relates to the field of power supply technology, and in particular to a method and system for remote monitoring and fault diagnosis of 10kV switchgear. Background Technology

[0002] As a critical node in urban power distribution networks, 10kV switching stations bear load currents for extended periods. Due to increased contact resistance, mechanical loosening, or insulation aging, parts such as circuit breaker contacts, busbar connection points, and cable joints within the cabinets are prone to localized overheating, potentially leading to equipment burnout, short circuits, or even fires. Therefore, real-time, accurate, and reliable monitoring of the temperature status of critical components inside the switchgear is essential for ensuring the safe and stable operation of the power distribution network.

[0003] Currently, the industry mainly uses the following types of switchgear temperature monitoring technologies: I. Manual Infrared Spot Temperature Inspection Maintenance personnel regularly use handheld infrared thermometers to measure the surface temperature of switchgear. This method is flexible and has low initial investment, but it has obvious drawbacks: First, it cannot penetrate the metal cabinet and can only measure the outer shell temperature, making it difficult to reflect the actual internal hot spots. If the cabinet is opened for inspection, the process is cumbersome and has a large impact area due to the high-voltage distribution cabinet, and there are also certain safety hazards. Second, it relies on manual labor, and the frequency of inspection is limited, making continuous monitoring impossible. Third, the measurement results are affected by factors such as operational experience and environmental reflection, resulting in poor data comparability and making it difficult to support trend analysis and early warning.

[0004] II. Fiber Optic Temperature Measurement Technology High-precision, electromagnetic interference-resistant temperature monitoring can be achieved by laying distributed optical fiber or fluorescent optical fiber probes in key areas. However, this technology requires the sensing optical fiber to be installed directly near high-voltage live parts, which places extremely high demands on insulation design and construction technology. At the same time, the internal space of switchgear is small and the structure is complex, making optical fiber wiring difficult and prone to breakage due to equipment operation or vibration, resulting in high maintenance costs and making large-scale promotion difficult.

[0005] III. Wireless Temperature Sensor Wireless temperature measurement nodes powered by batteries or current transformers (such as the DS18B20 and ATE series) are attached to the temperature measurement point and transmit data via frequency bands such as 433MHz / 2.4GHz. While this approach eliminates the hassle of wiring, it still has significant limitations: battery life is typically only 3–5 years, and replacement is difficult in confined, high-pressure environments; current transformer power cannot provide stable power under light or no-load conditions; furthermore, in environments with strong electromagnetic interference, wireless communication reliability decreases, and data packet loss or bit errors are prone to occur.

[0006] IV. Fixed Infrared Thermal Imaging Technology In recent years, some manufacturers have attempted to install infrared thermal imagers inside the observation windows of switchgear to achieve non-contact internal temperature measurement. However, existing solutions generally suffer from the following bottlenecks: 1. The switch cabinet is a fully enclosed metal structure. The imaging effect of the infrared thermal imager is greatly affected by the temperature and humidity of the working environment. Low temperature can easily cause condensation on the lens, while high humidity or high temperature will affect the imaging accuracy of the instrument. Moreover, most infrared thermal imaging terminals lack environmental adaptability. 2. A single infrared image lacks information about the device structure, making it difficult to accurately locate heat-generating components, especially in areas with dense components, where misjudgment is likely to occur; 3. Data processing remains at the threshold alarm level, without integrating multi-dimensional operating parameters such as current, voltage, and vibration, making it impossible to distinguish between "normal temperature rise" and "fault overheating," resulting in a high false alarm rate; 4. The system architecture is loose, with fragmented data acquisition, transmission, and analysis processes, lacking a unified time synchronization and feature extraction mechanism, making it difficult to support intelligent diagnosis.

[0007] In summary, existing technologies cannot meet the urgent needs of 10kV switchgear for all-weather, highly reliable, and intelligent remote monitoring. Based on this, this case is proposed. Summary of the Invention

[0008] The purpose of this invention is to provide a method and system for remote monitoring and fault diagnosis of 10kV switchgear. By constructing a three-layer collaborative architecture of "data acquisition - data preprocessing and transmission - intelligent diagnosis", it realizes all-weather, high-precision, non-contact remote monitoring and fault early warning of the internal status of 10kV switchgear, which significantly improves the safety, reliability and intelligent operation and maintenance level of the distribution network.

[0009] To achieve the above objectives, the technical solution of the present invention is as follows: A method for remote monitoring and fault diagnosis of 10kV switchgear includes the following steps: S1. Simultaneously collect infrared thermal imaging data, visible light image data, and electrical and environmental parameters related to equipment operating status inside multiple switch cabinets; S2. Aggregate data from different switch cabinets, perform protocol conversion, preprocessing and timestamp alignment on the aggregated data, and extract key feature data to form structured data packets; S3. Based on the structured data packet, use a multi-channel image fusion algorithm and a multi-level early warning model based on the state stability index to generate equipment health status assessment results and graded early warning information.

[0010] Furthermore, the protocol conversion includes the process of uniformly converting the raw data from Modbus-RTU, IEC 60870-5-104 or proprietary wireless sensing protocols into standard Modbus-TCP or JSON format.

[0011] Furthermore, the preprocessing includes non-uniformity correction, noise suppression, and temperature calibration of the infrared thermal imaging data, and brightness equalization and edge enhancement of the visible light image data.

[0012] Furthermore, in step S2, the extraction of key feature data includes: Extract the highest temperature, average temperature, and temperature distribution histogram of each heat-generating area from infrared thermal imaging data; The spatial coordinates of key components are identified from visible light image data. The types of key components are preset by the operation and maintenance personnel during the parameter setting stage, and at least include circuit breaker contacts, busbar connection points and cable joints. Extract the effective value of current, voltage fluctuation rate, and harmonic distortion rate from electrical parameters.

[0013] Furthermore, in step S2, the timestamp alignment adopts a clock synchronization mechanism to align the infrared thermal imaging frame, the visible light image frame, and the electrical parameter sampling points according to the same time reference, with an alignment error not exceeding ±100 milliseconds.

[0014] Furthermore, in step S3, the multi-channel image fusion algorithm includes the following steps: S31. Perform quality assessments on both infrared thermal imaging data and visible light image data: Brightness uniformity analysis is performed on visible light image data to determine whether it contains clear device structure information; Analyze the thermal signal coverage of the infrared thermal imaging data to determine whether the key parts are all within the effective thermal imaging area of ​​the infrared field of view; if the thermal signal coverage of any key part is insufficient, mark the area as "infrared invisible" and record its missing status. S32. Based on the spatial coordinates of the key parts extracted in step S2, perform pixel-level spatial registration between the infrared thermal imaging data and the visible light image data to ensure that the same physical component is spatially aligned in the two modal images. S33. For each key area, dynamically calculate the fusion weight based on the quality assessment results in S32: If the visible light image in the area is uniform in brightness and unobstructed, and the infrared thermal signal coverage is complete, then balanced weight fusion is used, that is, the weight of the visible light image data is equal to the weight of the infrared thermal imaging data. If the visible light image quality of the area is good but the infrared thermal signal coverage is insufficient, then the weight of the visible light image data will be greater than the weight of the infrared thermal imaging data. If the area has insufficient light or is obstructed, or the visible light image is blurry, but the infrared thermal signal coverage is complete, then the weight of the infrared thermal imaging data is greater than the weight of the visible light image data. If both visible light and infrared light are unreliable in this area, then the area is marked as "data missing"; S34. Based on the weight settings in step S33, the registered infrared and visible light images are weighted and fused to generate a fused image stream.

[0015] Based on state stability index The multi-level early warning model is calculated as follows: ; in, Defined as the weighted average of the difference between the current temperature of each key component and its ambient reference temperature; Defined as a weighted average of the deviation of the current value of each state parameter from its normal range, and then normalized to obtain the value in the interval of 0~1; α and β are weighting coefficients customized according to the device type.

[0016] Furthermore, the state stability index is divided into five levels: like If the value is ≤0.3, the status is displayed as normal; If 0.3 < If the value is ≤0.5, the status will be displayed as "Caution". If 0.5 < If the value is ≤0.7, the status will be displayed as abnormal; If 0.7 < If the value is ≤0.9, the status is displayed as severely abnormal; like If the value is greater than 0.9, the status is displayed as critical and abnormal.

[0017] A remote monitoring and fault diagnosis system for 10kV switchgear based on the above method includes: The terminal acquisition layer, located inside the switch cabinet, is used to acquire infrared thermal imaging data, visible light image data, and electrical and environmental parameters related to the equipment's operating status. It includes at least an infrared thermal imaging module, a visible light camera module, a temperature and humidity sensing module, a current and voltage transformer module, and a vibration sensing module. The network transport layer is used to aggregate data from different switch cabinets, perform protocol conversion, preprocessing and timestamp alignment on the aggregated data, and extract key feature data to form structured data packets. It includes at least a data receiving module, a protocol conversion module, a preprocessing module, a time synchronization module and a key feature data extraction module. The platform application layer, deployed in the monitoring center, is used to generate equipment health status assessment results and graded early warning information, and to output them in a visual format.

[0018] Furthermore, the infrared thermal imaging module includes a closed housing, in which an infrared thermal imaging unit, a heating unit, and a temperature and humidity sensing unit are fixed. A through hole is provided on the front side wall of the housing, and the through hole is sealed and covered with an infrared-transparent layer. The optical lens of the infrared thermal imaging unit is directly facing the through hole. The top surface of the housing is provided with a mounting box. The top plate of the mounting box is exposed above the top surface of the housing and can be opened and closed. The bottom plate of the mounting box is located inside the housing and is a perforated plate. A desiccant is placed inside the mounting box. A semiconductor cooling unit is provided on the back of the housing, with the cold end of the semiconductor cooling unit located inside the housing and the hot end located outside the housing; The bottom surface of the housing is fixed with a mounting base for fixing the housing to the switch cabinet.

[0019] The advantages of this invention are: 1. Proactive equipment fault warning: By collecting real-time infrared thermal image data and visible light image data inside the switch cabinet, and combining multi-source sensor information (current, voltage, vibration, temperature and humidity, etc.), and combining dynamic image fusion and state stability index model, the system can issue early warnings at the initial stage of equipment abnormality, transforming fault handling from "post-event maintenance" to "pre-event prevention", significantly reducing the equipment failure rate.

[0020] 2. Cost-effective Operation and Maintenance: The system achieves unmanned automated inspection, reducing the frequency of manual inspections and lowering labor costs. Through precise fault location and diagnosis, unnecessary power outages for maintenance are avoided, optimizing the allocation of maintenance resources and improving operational efficiency. The system supports remote monitoring and data querying, reducing the need for personnel to go to the site, making it particularly suitable for remote or harsh environments.

[0021] 3. Significantly improves monitoring accuracy and diagnostic reliability: By introducing manually pre-calibrated coordinates of key components, pixel-level registration of infrared and visible light images is achieved, and the fusion weights are dynamically adjusted based on lighting and occlusion conditions. Simultaneously, intelligent weight reduction is applied to areas with insufficient thermal signal coverage or missing data to avoid misjudgments. This strategy significantly improves the accuracy of status identification for key components such as circuit breaker contacts and cable joints, outperforming traditional single-threshold alarm methods.

[0022] 4. Significantly Enhanced Safety Performance: This invention employs a non-contact fault diagnosis method, eliminating the need for direct wiring on high-voltage equipment and fundamentally solving the high-voltage insulation problem. The system can monitor the internal status of the switchgear in real time, promptly detecting localized overheating caused by poor contact, abnormal load, etc., effectively preventing fires or explosions caused by equipment overheating, and improving the overall safety level of the power system.

[0023] 5. Enhanced support for smart grids: The system adopts a three-layer architecture (terminal acquisition - edge preprocessing - platform intelligent diagnosis), completing protocol conversion, time alignment and feature extraction at the edge to reduce the burden on the cloud; the platform automatically generates temperature trends, equipment health reports and graded early warning suggestions, supporting remote inspection and precise maintenance, reducing the frequency of manual on-site operations, especially suitable for remote or unattended switchgear, significantly improving operation and maintenance efficiency and resource utilization.

[0024] 6. The infrared thermal imaging monitoring terminal adopts a sealed shell design, with the light-transmitting holes covered by high-transmittance infrared wave-transmitting material, and integrates a temperature and humidity control unit. It automatically maintains the stability of the internal working environment in harsh environments such as low temperature, high humidity, and high temperature, ensuring that the temperature measurement accuracy is not disturbed and meeting the requirements of continuous and reliable operation of power equipment 24 / 7. Attached Figure Description

[0025] Figure 1 This is a flowchart illustrating the remote monitoring and fault diagnosis method for a 10kV switchgear in the embodiment. Figure 2 This is a schematic diagram of the architecture of the remote monitoring and fault diagnosis system for a 10kV switchgear in the embodiment. Figure 3 This is a schematic diagram of the structure of the infrared thermal imaging module in the embodiment.

[0026] Label Explanation 1. Housing; 2. Infrared thermal imaging unit; 3. Heating unit; 4. Temperature and humidity sensing unit; 5. Through hole; 6. Infrared transparent layer; 7. Mounting box; 8. Semiconductor cooling unit; 9. Mounting base. Detailed Implementation

[0027] The present invention will be further described in detail below with reference to embodiments. It should be understood that the terms "upper", "lower", "front", "rear", "left", "right", "top", "bottom", "inner", "outer", etc., used in this document indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limiting the present invention.

[0028] This embodiment proposes a method for remote monitoring and fault diagnosis of 10kV switchgear, such as... Figure 1 As shown, the process includes the following steps.

[0029] Step S1: Simultaneously collect infrared thermal imaging data, visible light image data, and electrical and environmental parameters related to the equipment's operating status inside multiple switch cabinets.

[0030] Step S2: Aggregate data from different switch cabinets, perform protocol conversion, preprocessing and timestamp alignment on the aggregated data, and extract key feature data to form structured data packets.

[0031] In this embodiment, the protocol conversion includes the process of uniformly converting the raw data from Modbus-RTU, IEC 60870-5-104, or proprietary wireless sensing protocols into standard Modbus-TCP or JSON format. Since the temperature measurement terminals, electrical parameter acquisition units, and environmental sensors deployed on-site may use different communication protocols, protocol conversion not only standardizes the data structure but also provides a unified data foundation for subsequent timestamp alignment, feature extraction, and multi-source information fusion.

[0032] Preprocessing includes non-uniformity correction, noise suppression, and temperature calibration of infrared thermal imaging data, and brightness equalization and edge enhancement of visible light image data. Since infrared thermal imaging is susceptible to detector non-uniformity and environmental interference, and visible light images often suffer from uneven brightness and blurred details under low-light conditions inside cabinets, preprocessing significantly improves the accuracy of temperature measurement of key components and the clarity of structural identification, providing a high-quality data foundation for subsequent multi-channel image fusion and condition assessment.

[0033] Timestamp alignment employs a clock synchronization mechanism to align infrared thermal imaging frames, visible light image frames, and electrical parameter sampling points to the same time base, with an alignment error not exceeding ±100 milliseconds. Since infrared thermal imaging, visible light video, and electrical parameters are acquired by different sensors at different frequencies, their raw data naturally exhibit temporal discrepancies. Timestamp alignment ensures the consistency of multi-source data in the temporal dimension, providing a reliable temporal foundation for multi-channel image fusion and state stability index modeling.

[0034] In this embodiment, the extraction of key feature data includes: (1) Extract the highest temperature, average temperature and temperature distribution histogram of each heat-generating area from the infrared thermal imaging data; (2) Identify the spatial coordinates of key parts from visible light image data. The types of key parts are preset by the operation and maintenance personnel during the parameter setting stage, including at least circuit breaker contacts, busbar connection points and cable joints. (3) Extract the effective value of current, voltage fluctuation rate and harmonic distortion rate from electrical parameters.

[0035] Step S3: Based on the structured data packet, use a multi-channel image fusion algorithm and a multi-level early warning model based on the state stability index to generate equipment health status assessment results and graded early warning information.

[0036] The multi-channel image fusion algorithm includes the following steps: S31. Perform quality assessments on both infrared thermal imaging data and visible light image data: Brightness uniformity analysis is performed on visible light image data to determine whether it contains clear device structure information; Analyze the thermal signal coverage of the infrared thermal imaging data to determine whether the key parts are all within the effective thermal imaging area of ​​the infrared field of view; if the thermal signal coverage of any key part is insufficient, mark the area as "infrared invisible" and record its missing status. S32. Based on the spatial coordinates of the key parts extracted in step S2, perform pixel-level spatial registration between the infrared thermal imaging data and the visible light image data to ensure that the same physical component is spatially aligned in the two modal images. S33. For each key area, dynamically calculate the fusion weight based on the quality assessment results in S32: If the visible light image in the area is uniform in brightness and unobstructed, and the infrared thermal signal coverage is complete, then balanced weight fusion is used, that is, the weight of the visible light image data is equal to the weight of the infrared thermal imaging data. If the visible light image quality of the area is good but the infrared thermal signal coverage is insufficient, then the weight of the visible light image data will be greater than the weight of the infrared thermal imaging data. If the area has insufficient light or is obstructed, or the visible light image is blurry, but the infrared thermal signal coverage is complete, then the weight of the infrared thermal imaging data is greater than the weight of the visible light image data. If both visible light and infrared light are unreliable in this area, then the area is marked as "data missing"; S34. Based on the weight settings in step S33, the registered infrared and visible light images are weighted and fused to generate a fused image stream.

[0037] This dynamic fusion strategy generates a fused image stream with both high spatial resolution and temperature information, enabling precise identification and visualization of the status of key components such as circuit breaker contacts and cable joints, significantly improving the accuracy of fault diagnosis and maintenance efficiency.

[0038] Based on state stability index The multi-level early warning model is calculated as follows: ; Defined as the weighted average of the difference between the current temperature of each key component and its ambient reference temperature (i.e., "temperature rise"), it is calculated as follows: For each critical component i (including cable joints, insulators, circuit breaker contacts, etc.), obtain its current temperature T. i And obtain the reference temperature T0 of the switch cabinet's internal environment at the same moment (e.g., the air temperature inside the cabinet or the average temperature of the non-heating metal structure), and calculate the temperature rise of that part. : ; ; In the formula, n represents the number of critical parts. The preset weights can be set according to the importance of the components.

[0039] It is defined as a weighted average of the deviation of the current value of each state parameter from its normal range, and then normalized to obtain the value in the interval of 0 to 1.

[0040] Taking current as an example, let the rated current be I. n The lower limit of the current is I. min The upper limit of the current is I. max Given that the current value is I, calculate the current deviation d. I : If I<I min , then d I =I min -I / I min ; If I min ≤I≤I max , then d I =0; If I>I max , then d I =II max / I max Calculate the voltage deviation d sequentially using the method described above. U Vibration deviation d V (Calculation items can be added according to actual needs; this is just an example.) Finally, calculate... .

[0041] =w1*d I +w2*d U +w3*d V ; In the formula, w1, w2, and w3 are preset weighting coefficients.

[0042] α and β are custom weighting coefficients based on the device type.

[0043] The tiered early warning information is divided into five levels by the system based on the state stability index: like If the value is ≤0.3, the status is displayed as normal; If 0.3 < If the value is ≤0.5, the status will be displayed as "Caution". If 0.5 < If the value is ≤0.7, the status will be displayed as abnormal; If 0.7 < If the value is ≤0.9, the status is displayed as severely abnormal; like If the value is greater than 0.9, the status is displayed as critical and abnormal.

[0044] Different alarm levels correspond to different alarm methods and handling suggestions, which realizes the accuracy and differentiation of early warning and effectively reduces false alarms and missed alarms.

[0045] like Figure 2 As shown, this embodiment also proposes a remote monitoring and fault diagnosis system for the above method. The system adopts a three-layer distributed architecture, including: The terminal acquisition layer, located inside the switch cabinet, is used to acquire infrared thermal imaging data, visible light image data, and electrical and environmental parameters related to the equipment's operating status. It includes at least an infrared thermal imaging module, a visible light camera module, a temperature and humidity sensing module, a current and voltage transformer module, and a vibration sensing module. The network transport layer is used to aggregate data from different switch cabinets, perform protocol conversion, preprocessing and timestamp alignment on the aggregated data, and extract key feature data to form structured data packets. It includes at least a data receiving module, a protocol conversion module, a preprocessing module, a time synchronization module and a key feature data extraction module. The platform application layer, deployed in the monitoring center, is used to generate equipment health status assessment results and graded early warning information, and to output them in a visual format.

[0046] The main interface of the platform application layer should clearly display the following information: Real-time monitoring screen: Displays real-time visible light images, infrared thermal images, or fused images of the inside of the switch cabinet in the form of a video stream, and overlays the real-time temperature values ​​of each monitoring point on the image.

[0047] Data trend analysis: Provides a function to query historical temperature curves, allowing users to view the temperature change trend of any monitoring point over a specified period of time (such as 24 hours, one week, or one month).

[0048] Alarm management interface: Set alarm thresholds (including absolute temperature threshold and state stability index Z). twd (Thresholds at various levels). When an alarm is triggered, the system should notify maintenance personnel through various means such as pop-ups, sounds, SMS, or app push notifications. Alarm information must be recorded in a database, supporting queries and statistics by time, device, alarm level, etc.

[0049] Automatic report generation: The system can automatically generate equipment temperature operation reports on a daily, weekly, and monthly basis, and supports exporting and printing, providing data support for equipment status assessment and operation and maintenance decisions.

[0050] To address the issue of high environmental requirements for thermal imagers, such as Figure 3 As shown, this embodiment specifically designs an infrared thermal imaging module (or infrared thermal imaging terminal), including a closed housing 1. An infrared thermal imaging unit 2, a heating unit 3, and a temperature and humidity sensing unit 4 are fixed inside the housing 1. The heating unit 3 is used for temperature stabilization in low-temperature environments. A through-hole 5 is provided on the front side wall of the housing 1, and an infrared-transparent layer 6 is sealed and covered over the through-hole 5. The optical lens of the infrared thermal imaging unit 2 faces the through-hole 5. The combination of the through-hole 5 and the infrared-transparent layer 6 ensures the penetration of infrared signals while achieving physical isolation between the switch cabinet and the infrared thermal imaging unit 2, effectively avoiding interference during the imaging process.

[0051] The top surface of the housing 1 is provided with a mounting box 7. The top plate of the mounting box 7 protrudes from the top surface of the housing 1 and can be opened and closed. The bottom plate of the mounting box 7 is located inside the housing 1 and is a perforated plate. A desiccant is placed inside the mounting box 7 for dehumidification inside the housing 1. The part of the mounting box 7 located outside the housing 1 can be opened, enabling the desiccant to be replaced (generally once a year). The back of the housing 1 is provided with a semiconductor cooling unit 8. The cold end of the semiconductor cooling unit 8 is located inside the housing 1, and the hot end is located outside the housing 1, used to cool the inside of the housing 1 at high temperatures.

[0052] The temperature and humidity sensing unit 4 is used to detect the temperature and humidity inside the housing 1 and report it to the control module in the terminal acquisition layer. When the ambient temperature inside the housing is lower than 5°C, the heating module is activated to raise the temperature; when the ambient temperature is higher than 45°C, the semiconductor cooling unit 8 is activated to actively cool down.

[0053] In addition, a mounting base 9 is fixed on the bottom surface of the housing 1 for fixing the housing 1 to the switch cabinet.

[0054] This embodiment constructs a three-layer collaborative architecture of "terminal acquisition - edge preprocessing - platform intelligent diagnosis," integrating infrared thermal imaging, visible light images, and electrical operating parameters to achieve high-precision, all-weather status monitoring of key parts of 10 kV switchgear (such as circuit breaker contacts and cable joints). With the help of multi-channel image dynamic fusion, time synchronization alignment, structured feature extraction, and state stability index model, the system can not only accurately identify early overheating hazards, but also effectively distinguish between normal load heating and contact degradation faults, significantly reducing false alarm rate and missed alarm rate, and greatly improving the status perception capability, early warning foresight, and intelligent operation and maintenance level of distribution network equipment.

[0055] The above embodiments are only used to explain the concept of the present invention, and are not intended to limit the protection of the present invention. Any non-substantial modifications made to the present invention using this concept should fall within the protection scope of the present invention.

Claims

1. A remote monitoring and fault diagnosis method for a 10 kV switchgear, characterized in that, The method comprises the following steps: S1. synchronously collecting infrared thermal imaging data, visible light image data, and electrical parameters and environmental parameters related to the operation state of the equipment inside multiple switch cabinets; S2. aggregating data from different switch cabinets, performing protocol conversion, preprocessing, and timestamp alignment on the aggregated data, and extracting key feature data therefrom to form a structured data package; S3. based on the structured data package, generating an equipment health state evaluation result and graded warning information using a multi-channel image fusion algorithm and a multi-level early warning model based on a state stability index.

2. A method for remote monitoring and fault diagnosis of a 10 kV switchgear as claimed in claim 1, wherein, The protocol conversion comprises a process of uniformly converting original data from Modbus-RTU, IEC 60870-5-104, or a private wireless sensing protocol into a standard Modbus-TCP or JSON format.

3. A method for remote monitoring and fault diagnosis of a 10 kV switchgear as claimed in claim 1, wherein, The preprocessing comprises non-uniformity correction, noise suppression, and temperature value calibration on the infrared thermal imaging data, and brightness equalization and edge enhancement on the visible light image data.

4. The method for remote monitoring and fault diagnosis of a 10 kV switchgear according to claim 1, characterized in that, In step S2, the extraction of key feature data comprises: extracting the maximum temperature, average temperature, and temperature distribution histogram of each heat-emitting region from the infrared thermal imaging data; identifying the spatial coordinate positions of key parts from the visible light image data, the types of the key parts being pre-set by an operation and maintenance personnel during a parameter setting stage, and at least including circuit breaker contacts, bus connection points, and cable joints; extracting the current effective value, voltage fluctuation rate, and harmonic distortion rate from the electrical parameters.

5. A method for remote monitoring and fault diagnosis of a 10 kV switchgear as claimed in claim 1, wherein, In step S2, the timestamp alignment adopts a clock synchronization mechanism to align the infrared thermal imaging frames, visible light image frames, and electrical parameter sampling points according to the same time reference, with an alignment error of not more than ±100 milliseconds.

6. A method for remote monitoring and fault diagnosis of a 10 kV switchgear as claimed in claim 4, wherein, In step S3, the multi-channel image fusion algorithm comprises the following steps: S31. performing quality evaluation on the infrared thermal imaging data and the visible light image data respectively: performing brightness equalization analysis on the visible light image data to determine whether it has clear device structure information; performing thermal signal coverage range analysis on the infrared thermal imaging data to determine whether the key parts are all in the effective thermal imaging area of the infrared field of view; if the thermal signal coverage of any key part is insufficient, mark the area as "infrared invisible" and record its missing state; S32. based on the spatial coordinate positions of the key parts extracted in step S2, performing pixel-level spatial registration on the infrared thermal imaging data and the visible light image data to ensure spatial alignment of the same physical component in the two modal images; S33. for each key part region, dynamically calculating a fusion weight according to the quality evaluation result in S32: if the visible light image of the region is bright and unobstructed, and the infrared thermal signal coverage is complete, balanced weight fusion is adopted, i.e., the visible light image data weight is equal to the infrared thermal imaging data weight; if the visible light image quality of the region is good but the infrared thermal signal coverage is insufficient, the visible light image data weight is greater than the infrared thermal imaging data weight; if the light of the region is insufficient or there is an obstruction or the visible light image is blurred, but the infrared thermal signal coverage is complete, the infrared thermal imaging data weight is greater than the visible light image data weight; If the region is neither reliable in visible light nor in infrared, mark the region state as "data missing"; S34. Based on the weight setting in step S33, the registered infrared and visible light images are weighted and fused to generate a fused image stream.

7. A method for remote monitoring and fault diagnosis of a 10 kV switchgear as claimed in claim 4, wherein, State stability index based The multi-stage early warning model is calculated as follows: ； wherein, is defined as the weighted average of the difference between the current temperature of each critical point and its environmental reference temperature; is defined as the weighted average of the deviation of each state parameter from its normal range, and is normalized to obtain a value in the interval 0~1; and α and β are weight coefficients defined according to the type of the device.

8. A method for remote monitoring and fault diagnosis of a 10 kV switchgear as claimed in claim 7, wherein, The state stability index is divided into five levels: If ≤ 0.3, the status is shown as normal; If 0.3 < If the value is ≤0.5, the status will be displayed as "Caution". If 0.5 < If the value is ≤0.7, the status will be displayed as abnormal; If 0.7 < If the value is ≤0.9, the status is displayed as severely abnormal; If greater than 0.9, the status is displayed as a critical anomaly.

9. A remote monitoring and fault diagnosis system for 10 kV switchgear based on the method of any one of claims 1 to 8, characterized in that, It includes: The terminal acquisition layer is arranged in the switch cabinet and is used for collecting infrared thermal imaging data, visible light image data, and electrical parameters and environmental parameters related to the operation state of the equipment, and at least includes an infrared thermal imaging module, a visible light camera module, a temperature and humidity sensing module, a current and voltage mutual inductance module, and a vibration sensing module; The network transmission layer is used for gathering data from different switch cabinets, performing protocol conversion, preprocessing, and timestamp alignment on the gathered data, extracting key feature data therefrom, forming a structured data packet, and at least includes a data receiving module, a protocol conversion module, a preprocessing module, a time synchronization module, and a key feature data extraction module; The platform application layer is deployed in the monitoring center and is used for generating device health state evaluation results and graded early warning information and performing visual output.

10. The remote monitoring and fault diagnostic system for 10 kV switchgear as claimed in claim 9, wherein, The infrared thermal imaging module includes a closed shell, an infrared thermal imaging unit, a heating unit, and a temperature and humidity sensing unit fixed in the shell, a through hole is formed in the front side wall of the shell, an infrared wave-transparent layer is sealed and covered on the through hole, and the optical lens of the infrared thermal imaging unit is directly opposite to the through hole; A mounting box is arranged on the top surface of the shell, the top plate of the mounting box is exposed on the top surface of the shell and can be opened and closed, the bottom plate of the mounting box is located in the shell and is a mesh plate, and a drying agent is placed in the mounting box; A semiconductor refrigeration unit is arranged on the back surface of the shell, the cold end of the semiconductor refrigeration unit is located in the shell, and the hot end is located outside the shell; An installation seat is fixed on the bottom surface of the shell and is used for fixedly connecting the shell and the switch cabinet.

Images