A high temperature alarm system and processing device for a power distribution station

Abstract

The present invention discloses a high temperature alarm system and processing device for a distribution station, and relates to the technical field of high temperature alarm. The present invention comprises a sensor control module, a dynamic threshold discrimination module, a hierarchical linkage disposal plan library and a disposal backtracking monitoring module; the sensor control module constructs a three-dimensional model of the distribution station by laser scanning and photogrammetry; the present invention, through the sensor control module, utilizes laser scanning and photogrammetry technology to construct a three-dimensional model of the distribution station, accurately analyzes the heating parts of the equipment, and evaluates the installation feasibility and accessibility of the sensors, thereby realizing a reasonable layout of the sensors, which not only improves the comprehensiveness of temperature monitoring and ensures accurate temperature measurement of key parts, but also, through the dynamic threshold discrimination module, uses a deep learning algorithm combined with transfer learning and generative adversarial network technology to break through the limitation of fixed thresholds, more accurately judges temperature anomalies and locates minor faults, and effectively reduces false alarms and missed alarms.

Description

Technical Field

[0001] The present invention relates to the technical field of high temperature alarms, and in particular to a high temperature alarm system and a processing device for a power distribution station. Background Art

[0002] There are many electrical equipment in the distribution station, which generate heat during operation and normally rely on the heat dissipation system to maintain the temperature. However, in reality, environmental factors such as high temperatures in summer, poor ventilation, humidity, and equipment aging, overload, loose connections, etc. can easily hinder heat dissipation, causing high temperature risks, endangering equipment safety and power supply stability.

[0003] Traditional alarm systems rely on manual inspections and single-point sensors. Inspection intervals are long, single-point monitoring is incomplete, and there are blind spots. Alarms are based on a single threshold, lack comprehensive analysis, are prone to false alarms and missed alarms, and cannot accurately locate abnormal points. Disposal relies solely on manual work, lacks effective strategies and effect evaluation, is blind, and makes it difficult to ensure that the equipment temperature returns to a safe range.

[0004] In order to solve the above defects, a technical solution is now provided. Summary of the invention

[0005] The purpose of the present invention is to solve the problems of untimely, inaccurate and limited monitoring of traditional high temperature alarm systems in distribution stations, and lack of effective disposal measures and disposal effect evaluation, and to propose a high temperature alarm system and processing device for distribution stations.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] A high temperature alarm system for a power distribution station, comprising:

[0008] The sensor control module builds a 3D model of the distribution station through laser scanning and photogrammetry, analyzes the heating parts of the equipment, comprehensively evaluates the feasibility and accessibility of sensor installation, and selects control points;

[0009] The dynamic threshold discrimination module uses deep learning algorithms to train temperature threshold models based on historical and real-time data of distribution station equipment, and combines transfer learning and generative adversarial network technology to determine temperature anomalies and locate subtle faults.

[0010] The hierarchical linkage disposal plan library divides the disposal plan into three levels according to the degree of temperature anomaly and the importance of equipment. The abnormality level is judged by comprehensive temperature threshold range, change rate and equipment weight factors, and the corresponding plan is automatically executed;

[0011] After the disposal action is executed, the disposal retrospective monitoring module collects temperature data, calculates the temperature gradient and rate, and evaluates the disposal effect by combining the physical heat conduction model and statistical analysis. Based on the evaluation results, it automatically adjusts the plan or prompts manual intervention to form a closed disposal loop.

[0012] Furthermore, the execution process of the sensor control module is as follows:

[0013] Laser scanners are used to perform all-round scanning of the indoor and outdoor spaces of the power distribution station and various electrical equipment. During the scanning process, multiple scanning positions are set around and on the top of large transformers to obtain complete shape contour data. The resolution of each scanning point when collecting data meets the standard. For the building structure of the power distribution station, walls, doors, windows, beams and columns are scanned from different directions and heights to record the spatial scale and position relationship.

[0014] Photogrammetry is used as an auxiliary supplement. Photographs are taken with photographic equipment at the same time. The shooting angles include overhead shooting, downward shooting and different horizontal perspectives. The equipment identification plate, operation panel and gap texture feature information are recorded from various dimensions. With the help of photogrammetry software, based on the principle of multi-view geometry, the photos are matched, spliced ​​and 3D reconstructed to supplement the color and detail texture parts missed by laser scanning. After fusion, a point cloud data model with preliminary geometric and visual information is formed.

[0015] The collected point cloud data is imported into the 3D modeling software, and the filtering algorithm is used to remove noise points. The point cloud is classified according to the characteristics of the equipment and building structure to distinguish the equipment body, bracket, and building wall. Then, the discrete point cloud is converted into a continuous polygonal mesh model through the triangulation algorithm. In this process, the mesh quality is controlled to meet the preset standard, and the shape and size distribution of the triangle patch are optimized to restore the physical form of the distribution station.

[0016] Based on the grid model, attribute labels are added to different model components, including the equipment name, rated power, operating parameters, and heat dissipation method technical parameters for electrical equipment; the material and thermal insulation performance indicators of the building structure are noted to build a three-dimensional model of the distribution station that integrates geometric shape, appearance texture and physical properties;

[0017] According to the operating principle of electrical equipment and heat transfer theory, analyze the internal heating parts of each device; for transformers, according to the heating mechanism of iron core hysteresis loss and winding copper loss caused by electromagnetic induction, lock the heating nodes between the core laminations, between the turns of each level of winding, and the connection between the winding and the iron core in the three-dimensional model; for switchgear, analyze the resistance heat generated by the current passing through the contacts and busbar connection plug-ins, and mark the contact opening and closing points and the plug-in connection gap positions in the model;

[0018] While determining the key heating locations, the feasibility and accessibility of sensor installation are evaluated based on the actual structural space inside the equipment.

[0019] Furthermore, the sensor control module evaluates the feasibility and accessibility of sensor installation in combination with the actual internal structural space of the equipment. The specific operation steps are as follows:

[0020] Comprehensive evaluation is conducted through measurement parameters, including:

[0021] Installation space margin ratio: Calculate the space required for sensor installation Available space inside the equipment The ratio is given by the formula: Get the installation space margin ratio , and recorded as the margin ratio ;

[0022] Minimum channel width to sensor size ratio: Inside the device, the sensor needs to pass through the channel to reach the installation location; measure the width of the channel at its narrowest point , and the maximum lateral dimension of the sensor By comparison, we can get the ratio of the minimum channel width to the sensor size. , and recorded as the width-to-length ratio ;

[0023] Contact angle deviation range: For contact sensors, the temperature can only be accurately measured by contacting the heat source. The optimal contact angle between the sensor and the heat source is determined in the 3D model. And the actual installation angle deviation ; By simulating the installation path, the maximum value of the angle deviation is calculated , and calculate the optimal contact angle deviation range ratio ; and recorded as the angle deviation ratio ;

[0024] Maintenance operation space coefficient: Considering the convenience of subsequent maintenance operations, the minimum spherical space radius that can meet the operation of maintenance tools is measured with the installation point as the center. The actual operating space radius available inside the equipment By comparison, the maintenance operation space coefficient is obtained. , and recorded as the maintenance ratio ;

[0025] Occlusion rate: Through the sectioning and roaming functions of the 3D model, the proportion of the part blocked by the internal structure of the equipment from the installation location to the sensor data transmission interface and the maintenance operation entrance is counted. The projection area of ​​the blocked part is The total visual area is , then the occlusion rate ;

[0026] The obtained margin ratio , width-to-size ratio , Angle deviation ratio , maintenance ratio And occlusion rate , after normalization, enter the following formula To obtain the comprehensive evaluation value E, Residue ratio , width-to-size ratio , Angle deviation ratio , maintenance ratio And occlusion rate The preset weight coefficient is used; the calculated comprehensive evaluation value E is used as the standard for measuring the feasibility and accessibility of sensor installation;

[0027] The comprehensive evaluation value E calculated for all determined key heating parts is compared with the evaluation threshold. When the comprehensive evaluation value E is greater than the evaluation threshold, it is judged that the feasibility and accessibility of the sensor installation at the key heating part meet the standards, and the key heating parts with a comprehensive evaluation value E greater than the evaluation threshold are selected as the screened control points.

[0028] Furthermore, the sensor control module also includes:

[0029] Arrange sensors at corresponding positions on the 3D model along the heat dissipation fins and ventilation ducts of the equipment shell to monitor the heat dissipation effect;

[0030] This includes placing points at the root of the heat sink of the transformer shell and around the top air outlet to monitor the heat dissipation rate and balance in real time;

[0031] For indoor spaces, based on airflow organization simulation analysis, the sensor position is determined in the cold and hot air intersection area and airflow dead corners to fully control the indoor temperature distribution and the impact of air flow on equipment heat dissipation.

[0032] Furthermore, the execution process of the hierarchical linkage disposal plan library is as follows:

[0033] Relying on the historical operation big data and real-time working conditions of distribution station equipment, the temperature threshold model is trained using deep learning algorithms;

[0034] The model is updated dynamically during operation, distinguishing the temperature ranges during equipment start-up and shutdown, stable operation, and overload conditions. It breaks through the limitations of fixed thresholds, determines temperature anomalies, and locates signs of temperature rise caused by subtle fault points.

[0035] Furthermore, the execution process of the dynamic threshold determination module also includes:

[0036] Transfer learning introduces similar equipment experience: Use transfer learning technology to transfer common feature knowledge from other distribution station equipment data sets of the same type and working conditions but with longer operating history and more fault cases;

[0037] Generate adversarial networks to expand sample diversity: Use adversarial networks to generate pseudo samples that are identical to real fault scenarios and input them into discriminator training together with real samples, prompting the threshold model to learn more comprehensive fault temperature patterns and improve the ability to identify and define thresholds for complex faults with small samples.

[0038] 1. Furthermore, the execution process of the hierarchical linkage disposal plan library is as follows:

[0039] The three-level emergency response plan is divided based on the degree of temperature anomaly and the importance of equipment. Specifically:

[0040] Obtain the temperature threshold interval determined by the dynamic threshold discrimination module and use it to divide the levels; at the same time, analyze the importance of the equipment in the distribution station; by setting different weight coefficients for equipment of different importance, the importance of the equipment is taken into consideration when judging the temperature anomaly level;

[0041] In addition to paying attention to whether the real-time temperature reaches a certain threshold, the temperature change rate is also analyzed; historical temperature data and real-time data are used to predict the temperature development trend, and by fitting the temperature-time curve, it is determined whether the temperature is in a continuous rise, stable or starting to fall stage;

[0042] When it is judged to be a slight temperature rise at level 1, the ventilation optimization in the station is started, including adjusting the opening and closing of the blinds and the speed of the fans, and a pop-up window on the monitoring screen prompts the operation and maintenance personnel to check the corresponding equipment;

[0043] Level 2 moderately high temperature: automatically adjust equipment operating parameters, including load reduction, switching redundant modules, and push SMS or APP to notify the operation and maintenance team to go to the site;

[0044] In the case of level three critical high temperature, the power supply to the equipment is cut off and the fire protection system is pre-started, including spraying inert gas to suppress the fire and reduce the temperature. Simultaneous reporting to multiple levels of dispatch ensures that each link is closely connected and automatically executed.

[0045] Furthermore, the execution process of the disposal backtracking monitoring module is as follows:

[0046] Automatically mark the time point when the hierarchical linkage response plan is launched , and at the same time trigger all the temperature sensors involved in the monitoring to increase the frequency of data collection, comprehensively record the temperature information of all parts of the equipment and the surrounding environment, build a temperature snapshot before disposal, covering the basic data of temperature values ​​and distribution, and store it in a temporary buffer area for subsequent comparative analysis;

[0047] Each time step after the treatment action begins is set to , calculate the temperature gradient for each sensor feedback data; for the structure with obvious heat conduction direction inside the equipment, the temperature gradient vector is obtained by dividing the temperature difference between adjacent monitoring points by the distance to quantify the change in heat transfer direction and intensity; at the same time, according to the formula , where The temperature difference between the current moment and the previous moment is used to calculate the temperature change rate, forming a temperature gradient and rate data set that is dynamically updated over time, clearly showing the temperature change trajectory;

[0048] Import the equipment 3D model, material thermophysical properties and current temperature data into the built-in physical heat conduction numerical simulation software module to simulate the theoretical temperature distribution and change trend of the equipment under the implemented treatment measures;

[0049] Use statistical analysis methods to process temperature data during the treatment process, calculate average temperature, temperature variance, maximum and minimum value statistics, and observe temperature fluctuation characteristics; compare the current actual temperature with the preset safety threshold, while considering the set cooling ratio index;

[0050] When the evaluation determines that the temperature reduction has not reached the expected level, a command is issued according to the pre-programmed logic; on the one hand, it automatically switches to a higher-level disposal plan. If it was previously at the second-level moderate high temperature disposal stage, it is immediately upgraded to the third-level critical high temperature disposal plan, and the monitoring interface is updated to display the warning information simultaneously, highlighting the abnormal temperature stubborn points and new disposal action prompts;

[0051] On the other hand, manual intervention requests are sent to operation and maintenance personnel through multiple channels such as site broadcasts, SMS push, and APP alarms, with current temperature data, handling history, and simulation analysis results information attached. This is used by operation and maintenance personnel to troubleshoot the root cause of the fault and continue monitoring until the temperature is stabilized in the safe zone, completing the handling closed loop.

[0052] A high temperature alarm processing device for a power distribution station, comprising:

[0053] Core processor, data storage unit, communication interface module, human-computer interaction interface and power management module.

[0054] Compared with the prior art, the present invention has the following beneficial effects:

[0055] (1) The present invention uses a sensor control module to construct a three-dimensional model of the power distribution station using laser scanning and photogrammetry technology, accurately analyzes the heating parts of the equipment, and evaluates the installation feasibility and accessibility of the sensors, thereby achieving a reasonable layout of the sensors. This not only improves the comprehensiveness of temperature monitoring and ensures accurate temperature measurement of key parts, but also uses a dynamic threshold discrimination module to use a deep learning algorithm combined with transfer learning and generative adversarial network technology to break through the limitations of fixed thresholds, more accurately judge temperature anomalies and locate subtle faults, and effectively reduce false alarms and missed alarms;

[0056] (2) Compared with the traditional high temperature alarm system, the hierarchical linkage disposal plan library of the present invention can automatically divide the disposal plans according to the degree of temperature anomaly and the importance of the equipment, realizing three-level disposal from slight temperature rise to critical high temperature. This intelligent processing method not only improves the response speed, but also reduces the blindness of manual intervention by automatically executing the corresponding plan, ensuring that the equipment temperature can be quickly restored to a safe range, thereby effectively improving the safety of the distribution station and the stability of power supply;

[0057] (3) In the present invention, after executing the treatment action, the treatment retrospective monitoring module can collect temperature data, calculate the temperature gradient and rate, and evaluate the treatment effect by combining the physical heat conduction model and statistical analysis. This module can automatically adjust the plan or prompt manual intervention according to the evaluation results, forming a complete treatment closed loop; it not only improves the efficiency and effect of the treatment measures, but also provides detailed data support for operation and maintenance personnel through real-time monitoring and simulation analysis, making fault troubleshooting and temperature control more accurate, ensuring the safe operation of the distribution station under various conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In order to facilitate understanding by those skilled in the art, the present invention is further described below in conjunction with the accompanying drawings;

[0059] Figure 1 This is the overall system block diagram of the present invention. DETAILED DESCRIPTION

[0060] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0061] It should be understood that the terms “include” and “comprising” used in the specification and claims of the present disclosure indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or collections thereof.

[0062] It should also be understood that the terms used in this disclosure are only for the purpose of describing specific embodiments and are not intended to limit the disclosure. As used in this disclosure and claims, the singular forms of "a", "an", and "the" are intended to include the plural forms unless the context clearly indicates otherwise. It should also be further understood that the term "and / or" used in this disclosure and claims refers to any combination of one or more of the associated listed items and all possible combinations, including these combinations.

[0063] like Figure 1 As shown, a high temperature alarm system for a power distribution station includes a sensor control module, a dynamic threshold determination module, a hierarchical linkage disposal plan library, and a disposal backtracking monitoring module;

[0064] The sensor control module constructs a 3D model of the power distribution station through laser scanning and photogrammetry, analyzes the heating parts of the equipment, comprehensively evaluates the feasibility and accessibility of sensor installation, selects control points, and realizes a reasonable layout of sensors to ensure accurate temperature measurement and convenient maintenance;

[0065] Laser scanners are used to perform all-round scanning of the indoor and outdoor spaces of the power distribution station and various electrical equipment. During the scanning process, multiple scanning positions are set around and on the top of large transformers to obtain complete shape contour data. Sufficient resolution (up to millimeter level) must be ensured when collecting data at each scanning point to clearly present key geometric information such as equipment surface features and connection structures. For the building structure of the power distribution station, walls, doors, windows, beams and columns are scanned from different directions and heights to record the spatial scale and relative position relationship.

[0066] Photogrammetry is used as an auxiliary supplement. Photographic equipment is used to take a large number of photos from various angles, including overhead and horizontal perspectives, to record the equipment identification plates, operation panels, and texture feature information of tiny gaps from various dimensions. With the help of photogrammetry software, based on the principle of multi-view geometry, the photos are matched, spliced, and 3D reconstructed to supplement the color and texture details missed by laser scanning. After fusion, a point cloud data model with precise geometry and rich visual information is formed.

[0067] The collected point cloud data is imported into 3D modeling software (Autodesk ReCap, Geomagic Wrap, etc.), and the filtering algorithm is used to remove noise points (such as abnormal points caused by ambient light reflection and scanning jitter), and the point cloud is classified according to the equipment and building structure characteristics to distinguish the equipment body, bracket, and building wall categories; the discrete point cloud is then converted into a continuous polygonal mesh model through a triangular meshing algorithm. In this process, the mesh quality is controlled, and the shape and size distribution of the triangle facets are optimized to ensure that the model surface is smooth and the transition is natural, and the physical form of the distribution station is restored; on the basis of the mesh model, attribute labels are added to different model components, including for electrical equipment, marking the equipment name (transformer, switchgear, busbar specific model), rated power, operating parameters, and heat dissipation method technical parameters; the material (concrete, masonry, metal, etc.) and thermal insulation performance indicators of the building structure are noted to construct a 3D model of the distribution station that integrates geometric shape, appearance texture and physical properties.

[0068] According to the operating principle of electrical equipment and heat transfer theory, analyze the parts inside each device that are prone to heat. For transformers, based on the heating mechanism of core hysteresis loss and winding copper loss caused by electromagnetic induction, lock the heating nodes between core laminations, between turns of each level of winding, and at the connection between winding and core in the three-dimensional model. For switchgear, analyze the resistance heat generated by current passing through contacts and busbar connection plug-ins, and mark the contact opening and closing points and plug-in connection gap positions in the model. The temperature changes in these areas are crucial to the healthy operation of the equipment and are the core targets for priority deployment of sensors.

[0069] While determining the key heating parts, the feasibility and accessibility of sensor installation are evaluated in combination with the actual internal structural space of the equipment. The comprehensive evaluation is carried out by measuring the parameters, including:

[0070] Installation space margin ratio: Calculate the space required for sensor installation Available space inside the equipment The ratio is given by the formula: Get the installation space margin ratio , and recorded as the margin ratio ;

[0071] Minimum channel width to sensor size ratio: Inside the device, the sensor needs to pass through the channel to reach the installation location; measure the width of the channel at its narrowest point , and the maximum lateral dimension of the sensor By comparison, we can get the ratio of the minimum channel width to the sensor size. , and recorded as the width-to-length ratio ; Contact angle deviation range: For contact sensors, it is necessary to fit the heat source to accurately measure the temperature. Determine the optimal contact angle between the sensor and the heat source in the three-dimensional model. And the actual installation angle deviation ; By simulating the installation path, the maximum value of the angle deviation is calculated , and calculate the optimal contact angle deviation range ratio ; and recorded as the angle deviation ratio Maintenance operation space coefficient: Considering the convenience of subsequent maintenance operations (such as sensor replacement, calibration, etc.), the minimum spherical space radius that can meet the maintenance tool operation is measured with the installation point as the center. The actual operating space radius available inside the equipment By comparison, the maintenance operation space coefficient is obtained. , and recorded as the maintenance ratio ; Occlusion rate: Through the sectioning and roaming functions of the 3D model, the proportion of the part blocked by the internal structure of the equipment from the installation location to the sensor data transmission interface and the maintenance operation entrance is counted. The projection area of ​​the blocked part is The total visual area is , then the occlusion rate ;

[0072] The obtained margin ratio , width-to-size ratio , Angle deviation ratio , maintenance ratio And occlusion rate , after normalization, enter the following formula To obtain the comprehensive evaluation value E, Residue ratio , width-to-size ratio , Angle deviation ratio , maintenance ratio And occlusion rate The preset weight coefficient is used; the calculated comprehensive evaluation value E is used as the standard for measuring the feasibility and accessibility of sensor installation; the comprehensive evaluation value E calculated for all determined key heating parts is compared with the evaluation threshold. When the comprehensive evaluation value E is greater than the evaluation threshold, it is judged that the feasibility and accessibility of sensor installation in the key heating part meet the standard, and the key heating parts with a comprehensive evaluation value E greater than the evaluation threshold are selected as the screened control points; the above screening is used to ensure that the selected control points can not only fit closely to the heat source to obtain accurate temperature, but also are easy to install and maintain, so as to avoid the inability to construct or subsequent replacement due to cramped space and obstruction.

[0073] Sensors are arranged at corresponding positions of the three-dimensional model along the heat dissipation paths of the cooling fins and ventilation ducts of the equipment casing to monitor the heat dissipation effect. This includes placing sensors at the root of the heat sink of the transformer casing and around the top air outlet to grasp the heat dissipation rate and balance in real time. For indoor spaces, based on airflow organization simulation analysis (using the computational fluid dynamics (CFD) plug-in to simulate the ventilation airflow direction on the three-dimensional model), the sensor positions are determined in the intersection of cold and hot air (below the door and window vents, near the air outlet of ventilation equipment) and in airflow dead corners (the four corners of the room, behind large equipment) to fully control the indoor temperature distribution and the impact of air flow on equipment heat dissipation.

[0074] The dynamic threshold discrimination module uses deep learning algorithms to train temperature threshold models based on historical and real-time data of distribution station equipment. It combines transfer learning and generative adversarial network technology to break through the limitations of fixed thresholds, judge temperature anomalies, and locate subtle faults.

[0075] Relying on the historical operation big data of distribution station equipment (temperature curves under seasonal temperature and load cycle changes) and real-time working conditions (current load, ambient temperature and humidity), the temperature threshold model is trained using a deep learning algorithm. The model is dynamically updated with the operation, distinguishing the reasonable temperature ranges under equipment start-up and stop, stable operation, and overload conditions, breaking through the limitations of fixed thresholds, accurately judging temperature anomalies, and locating the signs of temperature rise caused by subtle fault points.

[0076] Transfer learning introduces similar equipment experience: In the actual operation of distribution stations, some extreme or rare faults occur very rarely, resulting in insufficient local accumulated data to support the deep learning model to accurately learn the temperature thresholds related to such faults. Use transfer learning technology to "migrate" common feature knowledge from other distribution station equipment data sets of the same type and working conditions but with longer operating history and richer fault cases; for example, a new type of switchgear has a special contact welding fault, and local data is scarce. By drawing on the data and model weights of similar switchgear at other sites in terms of contact faults, combined with local fine-tuning of a small number of samples, it can quickly adapt to the needs of rare fault monitoring of local equipment and broaden the fault coverage of the threshold model;

[0077] Generative adversarial networks (GANs) expand sample diversity: In response to the lack of rare fault samples, GAN networks are used to generate "pseudo samples" similar to real fault scenarios. Taking the transformer inter-turn short-circuit fault as an example, based on the normal operating temperature curve and a small number of known inter-turn short-circuit samples, the generator in the GAN creates a temperature change sequence at different stages under similar faults, which is input into the discriminator training together with the real samples, prompting the threshold model to learn a more comprehensive fault temperature pattern, and improving the ability to identify and define thresholds for complex faults under small samples.

[0078] The hierarchical linkage disposal plan library divides the disposal plans into three levels according to the degree of temperature anomaly and the importance of equipment. It judges the abnormality level by comprehensively considering factors such as temperature threshold range, change rate and equipment weight, and automatically executes the corresponding plan to ensure the safety of distribution station equipment.

[0079] The emergency response plan is divided into three levels based on the degree of temperature anomaly and the importance of equipment;

[0080] Get the temperature threshold interval determined by the dynamic threshold discrimination module and use it to divide the levels; at the same time, consider the importance of the equipment in the distribution station; for core equipment such as main transformers and key busbars, the level corresponding to their temperature threshold intervals will be more stringent; by setting different weight coefficients for equipment of different importance, the importance of the equipment is taken into consideration when judging the temperature anomaly level; in addition to paying attention to whether the real-time temperature reaches a certain threshold, the rate of temperature change is also analyzed; use historical temperature data and real-time data to predict the temperature development trend, and by fitting the temperature-time curve, determine whether the temperature is in a continuous rise, stabilization, or beginning to decline. If the temperature is predicted to break through a higher threshold in a short period of time, even if it is currently in a lower temperature range, the level of the disposal plan will be adjusted in advance;

[0081] When it is judged to be a slight temperature rise at level one, the ventilation optimization in the station is started (intelligent control of shutter opening and closing, fan speed), and a window pops up on the monitoring screen to remind the operation and maintenance personnel to check the corresponding equipment; for level two moderate high temperature, in addition to enhanced ventilation, the equipment operating parameters are automatically adjusted (load reduction, switching redundant modules), and SMS / APP notifications are sent to the operation and maintenance team to go to the scene; for level three critical high temperature, the power supply of the equipment is urgently cut off, and the fire protection system is pre-started (inert gas is sprayed to suppress fire and reduce temperature), and reports are made to multi-level dispatchers at the same time, and each link is closely connected and automatically executed.

[0082] After the treatment action is executed, the treatment retrospective monitoring module collects temperature data, calculates the temperature gradient and rate, combines the physical heat conduction model and statistical analysis to evaluate the treatment effect, and automatically adjusts the plan or prompts manual intervention based on the evaluation results to form a closed-loop treatment.

[0083] At the moment when the hierarchical linkage disposal plan is launched, the system automatically marks the time point , and at the same time trigger all the temperature sensors involved in the monitoring to enter the high-frequency data collection mode (such as increasing the collection rate from once a minute to once every 10 seconds), comprehensively record the temperature information of key parts of the equipment and the surrounding environment, and build a temperature "snapshot" before disposal, covering the basic data of temperature values ​​and distribution, and store it in a temporary cache area for subsequent comparative analysis;

[0084] Each time step after the treatment action begins to be implemented (set as , for example, 10 seconds), calculate the temperature gradient for each sensor feedback data; for the structure with obvious heat conduction direction inside the equipment (axial and radial directions of transformer winding), the temperature gradient vector is obtained by dividing the temperature difference of adjacent monitoring points by the distance to quantify the change in heat transfer direction and intensity; at the same time, according to the formula ( The temperature change rate is calculated by using the temperature difference between the current moment and the previous moment, forming a temperature gradient and rate data set that is dynamically updated over time, clearly showing the temperature change trajectory;

[0085] Import the equipment 3D model (a detailed model built in the early stage for sensor control planning), material thermophysical properties (thermal conductivity, specific heat capacity, density, etc.) and current temperature data into the built-in physical heat conduction numerical simulation software module (thermal analysis engine based on finite difference method and finite element method) to simulate the theoretical temperature distribution and change trend of the equipment under the implemented treatment measures (ventilation adjustment, load reduction, power outage, etc.);

[0086] Use statistical analysis methods to process temperature data during the treatment process, calculate average temperature, temperature variance, maximum and minimum value statistics, and observe temperature fluctuation characteristics; compare the current actual temperature (especially the highest temperature of key parts) with the preset safety threshold, while considering the set cooling ratio index;

[0087] When the assessment determines that the temperature drop does not meet expectations, instructions are issued based on pre-programmed logic. On the one hand, it automatically switches to a higher-level disposal plan. If the previous stage was the second-level moderate high temperature disposal stage (adjusting operating parameters, strengthening ventilation), it will be immediately upgraded to the third-level critical high temperature plan (emergency power outage, fire linkage), and the monitoring interface will be updated simultaneously to display early warning information, highlighting stubborn temperature anomalies and new disposal action prompts. On the other hand, detailed manual intervention requests are sent to operation and maintenance personnel through multiple channels such as station broadcasts, SMS push, and APP alarms, with current temperature data, disposal history, and simulation analysis results information attached, so that operation and maintenance can accurately identify the root cause of the fault and continue to monitor until the temperature is stabilized in a safe zone to complete the disposal closed loop.

[0088] A high temperature alarm processing device for a power distribution station, comprising:

[0089] The core processor is used for the control and coordination of the entire device. It integrates high-performance processor cores such as the ARM Cortex-A series, with an operating frequency of more than 1 GHz. It has strong data processing capabilities and can quickly process data from various modules to ensure efficient operation of the system.

[0090] The data storage unit is used to use a large-capacity Flash storage chip with a storage capacity of up to several GB. It is used to store the historical operating data of the distribution station equipment, including temperature curves in different seasons and load cycles, equipment start and stop records, operating hours and other information, to provide data support for the dynamic threshold judgment module. It can also record alarm events, disposal measures and results and other data to facilitate subsequent query and analysis.

[0091] Communication interface module, used to equip with a variety of communication interfaces, such as Ethernet interface, supporting 10 / 100 / 1000Mbps adaptive communication rate, used to interact with the monitoring system and remote dispatch center in the distribution station to achieve remote monitoring and management; RS-485 interface, which can connect multiple sensors or other intelligent devices, has a long communication distance (up to more than 1200 meters) and strong anti-interference ability, ensuring stable transmission of sensor data; wireless communication module (such as 4G / 5G communication module), realizes remote wireless transmission of data, making it convenient for operation and maintenance personnel to grasp the temperature of distribution station equipment anytime and anywhere;

[0092] The human-computer interaction interface consists of a display screen (such as a color LCD screen larger than 7 inches) and operation buttons. The display screen can intuitively display the real-time temperature distribution, temperature curve, alarm information and other contents of the power distribution station equipment. The operation buttons facilitate the operation and maintenance personnel to set system parameters, adjust alarm thresholds, query historical data and other operations, thereby improving the convenience of human-computer interaction;

[0093] The power management module includes a power conversion circuit and an uninterruptible power supply (UPS). The power conversion circuit converts the AC mains power of the distribution station into the DC power (such as 5V, 12V, etc.) required by the device to power each module. It also has functions such as overvoltage protection and overcurrent protection to ensure stable and reliable power supply. When the mains power is interrupted, the UPS can immediately switch to power the device. It uses a lithium battery pack as an energy storage unit with a battery life of several hours, ensuring that the device can still work normally during a power outage and avoiding alarm failure due to power outage.

[0094] The preferred embodiments of the present invention disclosed above are only used to help explain the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to only specific implementation methods. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. A high temperature alarm system for a power distribution station, characterized in that: include: The sensor control module builds a 3D model of the distribution station through laser scanning and photogrammetry, analyzes the heating parts of the equipment, comprehensively evaluates the feasibility and accessibility of sensor installation, and selects the control points. The execution process is as follows: Laser scanners are used to perform all-round scanning of the indoor and outdoor spaces of the power distribution station and various electrical equipment. During the scanning process, multiple scanning positions are set around and on the top of large transformers to obtain complete shape contour data. The resolution of each scanning point when collecting data meets the standard. For the building structure of the power distribution station, walls, doors, windows, beams and columns are scanned from different directions and heights to record the spatial scale and position relationship. Photogrammetry is used as an auxiliary supplement. Photographs are taken with photographic equipment at the same time. The shooting angles include overhead shooting, overhead shooting and different horizontal viewing angles. The equipment identification plate, operation panel and gap texture feature information are recorded from various dimensions. With the help of photogrammetry software, based on the principle of multi-view geometry, the photos are matched, stitched and 3D reconstructed to supplement the color, detail and texture parts missed by laser scanning, and then integrated to form a point cloud data model with preliminary geometric and visual information; The collected point cloud data is imported into the 3D modeling software, and the filtering algorithm is used to remove noise points. The point cloud is classified according to the characteristics of the equipment and building structure to distinguish the equipment body, bracket, and building wall. Then, the discrete point cloud is converted into a continuous polygonal mesh model through the triangulation algorithm. In this process, the mesh quality is controlled to meet the preset standard, and the shape and size distribution of the triangle patch are optimized to restore the physical form of the distribution station. Based on the grid model, attribute labels are added to different model components, including the equipment name, rated power, operating parameters, and heat dissipation method technical parameters for electrical equipment; the material and thermal insulation performance indicators of the building structure are noted to build a three-dimensional model of the distribution station that integrates geometric shape, appearance texture and physical properties; According to the operating principle of electrical equipment and heat transfer theory, analyze the internal heating parts of each device; for transformers, according to the heating mechanism of iron core hysteresis loss and winding copper loss caused by electromagnetic induction, lock the heating nodes between the core laminations, between the turns of each level of winding, and the connection between the winding and the iron core in the three-dimensional model; for switchgear, analyze the resistance heat generated by the current passing through the contacts and busbar connection plug-ins, and mark the contact opening and closing points and the plug-in connection gap positions in the model; While determining the key heating locations, evaluate the feasibility and accessibility of sensor installation based on the actual internal structural space of the equipment; The dynamic threshold discrimination module uses deep learning algorithms to train temperature threshold models based on historical and real-time data of distribution station equipment, and combines transfer learning and generative adversarial network technology to determine temperature anomalies and locate subtle faults. The hierarchical linkage disposal plan library divides the disposal plan into three levels according to the degree of temperature anomaly and the importance of equipment. The abnormality level is judged by comprehensive temperature threshold range, change rate and equipment weight factors, and the corresponding plan is automatically executed; After the disposal action is executed, the disposal retrospective monitoring module collects temperature data, calculates the temperature gradient and rate, and evaluates the disposal effect by combining the physical heat conduction model and statistical analysis. Based on the evaluation results, it automatically adjusts the plan or prompts manual intervention to form a closed disposal loop.

2. A high temperature alarm system for a power distribution station according to claim 1, characterized in that: The specific operation steps of the sensor control module to evaluate the feasibility and accessibility of sensor installation in combination with the actual internal structural space of the equipment are as follows: Comprehensive evaluation is conducted through measurement parameters, including: Installation space margin ratio: Calculate the space required for sensor installation Available space inside the equipment The ratio is given by the formula: Get the installation space margin ratio , and recorded as the margin ratio ; Minimum channel width to sensor size ratio: Inside the device, the sensor needs to pass through the channel to reach the installation location; measure the width of the channel at its narrowest point , and the maximum lateral dimension of the sensor By comparison, we can get the ratio of the minimum channel width to the sensor size. , and recorded as the width-to-length ratio ; Contact angle deviation range: For contact sensors, the temperature can only be accurately measured by contacting the heat source. The optimal contact angle between the sensor and the heat source is determined in the 3D model. And the actual installation angle deviation ; By simulating the installation path, the maximum value of the angle deviation is calculated , and calculate the optimal contact angle deviation range ratio ; and recorded as the angle deviation ratio ; Maintenance operation space coefficient: Considering the convenience of subsequent maintenance operations, the minimum spherical space radius that can meet the operation of maintenance tools is measured with the installation point as the center. The actual operating space radius available inside the equipment By comparison, the maintenance operation space coefficient is obtained. , and recorded as the maintenance ratio ; Occlusion rate: Through the sectioning and roaming functions of the 3D model, the proportion of the part blocked by the internal structure of the equipment from the installation location to the sensor data transmission interface and the maintenance operation entrance is counted. The projection area of ​​the blocked part is The total visual area is , then the occlusion rate ; The obtained margin ratio , width-to-size ratio , Angle deviation ratio , maintenance ratio And occlusion rate , after normalization, enter the following formula To obtain the comprehensive evaluation value E, Residue ratio , width-to-size ratio , Angle deviation ratio , maintenance ratio And occlusion rate The preset weight coefficient is used; the calculated comprehensive evaluation value E is used as the standard for measuring the feasibility and accessibility of sensor installation; The comprehensive evaluation value E calculated for all determined key heating parts is compared with the evaluation threshold. When the comprehensive evaluation value E is greater than the evaluation threshold, it is judged that the feasibility and accessibility of the sensor installation at the key heating part meet the standards, and the key heating parts with a comprehensive evaluation value E greater than the evaluation threshold are selected as the screened control points.

3. A high temperature alarm system for a power distribution station according to claim 1, characterized in that: The sensor control module also includes: Arrange sensors at corresponding positions on the 3D model along the heat dissipation fins and ventilation ducts of the equipment shell to monitor the heat dissipation effect; This includes placing points at the root of the heat sink of the transformer shell and around the top air outlet to monitor the heat dissipation rate and balance in real time; For indoor spaces, based on airflow organization simulation analysis, the sensor position is determined in the cold and hot air intersection area and airflow dead corners to fully control the indoor temperature distribution and the impact of air flow on equipment heat dissipation.

4. A high temperature alarm system for a power distribution station according to claim 1, characterized in that: The execution process of the dynamic threshold discrimination module: Relying on the historical operation big data and real-time working conditions of distribution station equipment, the temperature threshold model is trained using deep learning algorithms; The model is updated dynamically during operation, distinguishing the temperature ranges during equipment start-up and shutdown, stable operation, and overload conditions. It breaks through the limitations of fixed thresholds, determines temperature anomalies, and locates signs of temperature rise caused by subtle fault points.

5. A high temperature alarm system for a power distribution station according to claim 4, characterized in that: The execution process of the dynamic threshold determination module also includes: Transfer learning introduces similar equipment experience: Use transfer learning technology to transfer common feature knowledge from other distribution station equipment data sets of the same type and working conditions but with longer operating history and more fault cases; Generate adversarial networks to expand sample diversity: Use adversarial networks to generate pseudo samples that are identical to real fault scenarios and input them into discriminator training together with real samples, prompting the threshold model to learn more comprehensive fault temperature patterns and improve the ability to identify and define thresholds for complex faults with small samples.

6. A high temperature alarm system for a power distribution station according to claim 1, characterized in that: The execution process of the hierarchical linkage disposal plan library is as follows: The three-level emergency response plan is divided based on the degree of temperature anomaly and the importance of equipment. Specifically: Obtain the temperature threshold interval determined by the dynamic threshold discrimination module and use it to divide the levels; at the same time, analyze the importance of the equipment in the distribution station; by setting different weight coefficients for equipment of different importance, the importance of the equipment is taken into consideration when judging the temperature anomaly level; In addition to paying attention to whether the real-time temperature reaches a certain threshold, the temperature change rate is also analyzed; historical temperature data and real-time data are used to predict the temperature development trend, and by fitting the temperature-time curve, it is determined whether the temperature is in a continuous rise, stable or starting to fall stage; When it is judged to be a slight temperature rise at level 1, the ventilation optimization in the station is started, including adjusting the opening and closing of the blinds and the speed of the fans, and a pop-up window on the monitoring screen prompts the operation and maintenance personnel to check the corresponding equipment; Level 2 moderately high temperature: automatically adjust equipment operating parameters, including load reduction, switching redundant modules, and push SMS or APP to notify the operation and maintenance team to go to the site; In the case of level three critical high temperature, the power supply to the equipment is cut off and the fire protection system is pre-started, including spraying inert gas to suppress the fire and reduce the temperature. Simultaneous reporting to multiple levels of dispatch ensures that each link is closely connected and automatically executed.

7. A high temperature alarm system for a power distribution station according to claim 1, characterized in that: The execution process of the disposal backtracking monitoring module is as follows: Automatically mark the time point when the hierarchical linkage response plan is launched , and at the same time trigger all the temperature sensors involved in the monitoring to increase the frequency of data collection, comprehensively record the temperature information of all parts of the equipment and the surrounding environment, build a temperature snapshot before disposal, covering the basic data of temperature values ​​and distribution, and store it in a temporary buffer area for subsequent comparative analysis; Each time step after the treatment action begins is set to , calculate the temperature gradient for each sensor feedback data; for the structure with obvious heat conduction direction inside the equipment, the temperature gradient vector is obtained by dividing the temperature difference between adjacent monitoring points by the distance to quantify the change in heat transfer direction and intensity; at the same time, according to the formula , where The temperature difference between the current moment and the previous moment is used to calculate the temperature change rate, forming a temperature gradient and rate data set that is dynamically updated over time, clearly showing the temperature change trajectory; Import the equipment 3D model, material thermophysical properties and current temperature data into the built-in physical heat conduction numerical simulation software module to simulate the theoretical temperature distribution and change trend of the equipment under the implemented treatment measures; Use statistical analysis methods to process temperature data during the treatment process, calculate average temperature, temperature variance, maximum and minimum value statistics, and observe temperature fluctuation characteristics; compare the current actual temperature with the preset safety threshold, while considering the set cooling ratio index; When the evaluation determines that the temperature reduction has not reached the expected level, a command is issued according to the pre-programmed logic; on the one hand, it automatically switches to a higher-level disposal plan. If it was previously at the second-level moderate high temperature disposal stage, it is immediately upgraded to the third-level critical high temperature disposal plan, and the monitoring interface is updated to display the warning information simultaneously, highlighting the abnormal temperature stubborn points and new disposal action prompts; On the other hand, manual intervention requests are sent to operation and maintenance personnel through multiple channels such as site broadcasts, SMS push, and APP alarms, with current temperature data, handling history, and simulation analysis results information attached. This is used by operation and maintenance personnel to troubleshoot the root cause of the fault and continue monitoring until the temperature is stabilized in the safe zone, completing the handling closed loop.

8. A processing device for a high temperature alarm system of a power distribution station according to any one of claims 1 to 7, characterized in that: include: Core processor, data storage unit, communication interface module, human-computer interaction interface and power management module.

Images