Temperature measurement diagnosis method and system based on box-type substation

Abstract

The invention relates to the technical field of electrical equipment diagnosis, and discloses a temperature measurement diagnosis method and system based on a box-type substation, and the method comprises the following steps: S1, obtaining the partial discharge data, current data and temperature data of the box-type substation; s2, establishing a partial discharge and temperature correlation model according to the partial discharge data and the temperature data; establishing a current and temperature mapping relation according to the current heat effect of the current data; and S3, establishing a processing model according to the partial discharge and temperature correlation model, the current and temperature mapping relation and the temperature data, inputting the partial discharge data, the current data and the temperature data into the processing model, and outputting a diagnosis result and a linkage control signal by the processing model. The problem of low accuracy of temperature measurement diagnosis in the prior art is solved.

Description

Technical Field

[0001] This application relates to the technical field of power equipment diagnosis, and relates to a temperature measurement and diagnosis method and system based on a box-type substation. Background Art

[0002] The temperature measurement and diagnosis of box-type substations generally include real-time monitoring, analysis, and fault warning of the temperature states of transformer windings, switch contacts, cable joints, etc. Temperature measurement and diagnosis is a key module for the intelligent upgrade of box-type substations, and its penetration rate in the box substation market has increased rapidly.

[0003] In the existing temperature measurement and diagnosis devices or methods, there is no physical correlation modeling between partial discharge signals and temperature field changes, and the collaborative diagnosis of insulation deterioration and temperature rise cannot be achieved. Moreover, the real-time impact of current load changes on temperature rise is not quantified, which easily leads to misjudgment of faults. Therefore, the accuracy of temperature measurement and diagnosis is relatively low. Summary of the Invention

[0004] This application provides a temperature measurement and diagnosis method and system based on a box-type substation, aiming to solve the problem of relatively low accuracy of temperature measurement and diagnosis in the prior art.

[0005] In one aspect, a temperature measurement and diagnosis method based on a box-type substation is provided, including the following steps:

[0006] S1. Obtain the partial discharge data, current data, and temperature data of the box-type substation;

[0007] S2. Establish a partial discharge and temperature correlation model according to the partial discharge data and the temperature data; establish a current and temperature mapping relationship according to the current thermal effect of the current data;

[0008] S3. Establish a processing model according to the partial discharge and temperature correlation model, the current and temperature mapping relationship, and the temperature data, input the partial discharge data, the current data, and the temperature data into the processing model, and the processing model outputs a diagnosis result and a linkage control signal.

[0009] In one aspect, the temperature data includes thermal imaging data and sensor temperature data; in step S2, a partial discharge and temperature correlation model is established according to the partial discharge data and the thermal imaging data; in step S3, a processing model is established according to the partial discharge and temperature correlation model, the current and temperature mapping relationship, and the sensor temperature data.

[0010] In one solution, the box-type substation includes a high-voltage chamber, a transformer chamber, and a low-voltage chamber; temperature sensors are installed on the cable head copper busbar of the high-voltage chamber, the windings and iron cores of the transformer chamber, the incoming and outgoing line copper busbars of the transformer chamber, and the incoming and outgoing line copper busbars of the low-voltage chamber; current sensors are provided on the three incoming lines and four outgoing lines of the transformer chamber; a plurality of partial discharge sensors are installed on the inner wall of the transformer; a thermal imager is also provided in the box-type substation.

[0011] Specifically, the partial discharge sensor is a TEV partial discharge sensor.

[0012] In one solution, the temperature sensors for the high-voltage side windings and iron cores of the transformer chamber and the incoming and outgoing line copper busbars of the low-voltage chamber are fiber optic single-point temperature measurement sensors; the temperature sensors for the cable head copper busbar of the high-voltage chamber and the incoming and outgoing line copper busbars of the transformer chamber are RFID temperature sensors; the temperature sensors for the low-voltage side windings and iron cores of the transformer chamber are PT100 sensors.

[0013] In one solution, the steps for establishing the partial discharge and temperature correlation model include: collecting partial discharge characteristic quantities and thermal imaging data for multiple time periods and fitting the Hurst index; establishing a temperature rise mapping formula through the Hurst index; establishing a dynamic temperature rise model according to the temperature rise mapping formula.

[0014] In one solution, the current-temperature mapping relationship is based on the current load Dynamic temperature field formula: t is achieved; where is the equivalent resistance of the device, is the thermal conductivity coefficient of the material, represents the temperature change at time t; represents the current load changing with time.

[0015] In one solution, after performing step S2, the following steps are further included: using the thermal imaging data as the plane coordinates and calibrating the detection positions of each temperature sensor, calibrating the partial discharge source positions according to the partial discharge and temperature correlation model, and calibrating the detection positions of the current sensors according to the current-temperature mapping relationship.

[0016] In one solution, the establishment of the processing model is to dynamically adjust the weight distribution of the output confidence levels of the thermal imaging data, the partial discharge data, the current data, and the sensor temperature data based on the D-S evidence theory, and set the output confidence level threshold to establish a dynamic weight fusion diagnosis mechanism.

[0017] In one solution, the dynamic adjustment of the weight distribution is carried out through the formula: is carried out; where, is the current weight of sensor i; is the previous weight of sensor i; is any temperature sensor, partial discharge sensor, or thermal imager; is the learning rate; is the recent prediction error of sensor i. Among them, the recent prediction error of sensor i is approximately estimated through the covariance and predicted value output by the Kalman filter.

[0018] In one solution, on the other hand, a temperature measurement and diagnosis system based on a box-type substation is provided, and the described temperature measurement and diagnosis method is adopted.

[0019] Advantages of this application: Establishing a partial discharge and temperature correlation model based on partial discharge data and the temperature data can improve the diagnostic accuracy of insulation deterioration and increase the detection rate of instantaneous temperature anomalies. By establishing the mapping relationship between current and temperature, the real-time impact of current load changes on temperature rise can be quantified, so the situation of false fault judgment can be reduced and the accuracy of temperature measurement and diagnosis can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0021] Figure 1 is a schematic flow chart of the temperature measurement and diagnosis method in an embodiment of the present application;

[0022] Figure 2 is a comparison chart of traditional temperature measurement and this method's temperature measurement in an embodiment of the present application;

[0023] Figure 3 is a comparison chart of traditional temperature measurement and this method's temperature measurement and diagnosis in an embodiment of the present application; DETAILED DESCRIPTION OF THE EMBODIMENTS

[0024] The following will further describe in detail the specific embodiments of the present application in conjunction with the drawings and embodiments. The following embodiments are used to illustrate the present application, but are not used to limit the scope of the present application. Similarly, the following embodiments are only some embodiments of the present application rather than all embodiments. All other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present application.

[0025] In the present invention, terms such as "one embodiment", "some embodiments", "example", "specific example", or "some examples" mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0026] This application makes improvements and innovations and proposes the following embodiments.

[0027] In some embodiments, please refer to Figure 1 , on the one hand, a temperature measurement and diagnosis method based on a box-type substation is provided, including the following steps:

[0028] S1. Obtain the partial discharge data, current data, and temperature data of the box-type substation;

[0029] S2. Establish a partial discharge and temperature correlation model based on the partial discharge data and temperature data; establish a current-temperature mapping relationship based on the current thermal effect of the current data;

[0030] S3. Establish a processing model based on the partial discharge and temperature correlation model, the current-temperature mapping relationship, and the temperature data, input the partial discharge data, current data, and temperature data into the processing model, and the processing model outputs a diagnosis result and a linkage control signal.

[0031] Establishing a partial discharge and temperature correlation model based on the partial discharge data and temperature data can improve the diagnostic accuracy of insulation degradation and increase the detection rate of instantaneous temperature anomalies. By establishing a current-temperature mapping relationship, the real-time impact of current load changes on temperature rise can be quantified. Therefore, the situation of false fault judgment can be reduced and the accuracy of temperature measurement and diagnosis can be improved.

[0032] In some embodiments, the temperature data includes thermal imaging data and sensor temperature data; in step S2, a partial discharge and temperature correlation model is established based on the partial discharge data and thermal imaging data; in step S3, a processing model is established based on the partial discharge and temperature correlation model, the current-temperature mapping relationship, and the sensor temperature data. The temperature situation of the box-type substation can be intuitively reflected through the thermal imaging data, while the sensor temperature data can obtain the temperature situation at specific positions in the box-type substation. It is used to improve the temperature detection of the box-type substation, provide more data support for subsequent diagnosis work, and improve the accuracy of diagnosis.

[0033] In some embodiments, a box-type substation includes a high-voltage room, a transformer room, and a low-voltage room; temperature sensors are installed on the cable head copper busbars in the high-voltage room, the windings and cores in the transformer room, and the incoming and outgoing copper busbars in the transformer room, as well as the incoming and outgoing copper busbars in the low-voltage room; current sensors are installed on the three incoming wires and four outgoing wires in the transformer room; multiple partial discharge sensors are installed on the inner wall of the transformer room; and a thermal imager is also installed in the box-type substation. The installation of temperature sensors on the cable head copper busbars in the high-voltage room, the windings and cores in the transformer room, and the incoming and outgoing copper busbars in the transformer room, as well as the incoming and outgoing copper busbars in the low-voltage room, facilitates obtaining temperature data at corresponding locations; current sensors are installed on the three incoming wires and four outgoing wires in the transformer room to obtain current conditions in corresponding circuits; multiple partial discharge sensors are installed on the inner wall of the transformer room to facilitate obtaining partial discharge signals at multiple locations; multiple temperature data, multiple current data, and multiple partial discharge data can provide more data support for subsequent diagnostic work and improve diagnostic accuracy.

[0034] Specifically, the partial discharge sensor is a TEV partial discharge sensor. Using a TEV partial discharge sensor can better obtain the electromagnetic pulse signal of partial discharge and improve the detection sensitivity.

[0035] In some embodiments, the temperature sensors of the high-voltage side winding and iron core of the transformer chamber and the incoming and outgoing copper bars of the low-voltage chamber are optical fiber single-point temperature sensors; the temperature sensors of the cable head copper bars of the high-voltage chamber and the incoming and outgoing copper bars of the transformer chamber are RFID temperature sensors; the temperature sensors of the low-voltage side winding and iron core of the transformer chamber are PT100 sensors.

[0036] Specifically, the fiber optic single-point temperature sensor has high precision and sensitivity, with a measurement accuracy of ±0.1°C‌, and can capture tiny temperature changes (such as fluctuations of 0.1°C); its response speed to local hot spots far exceeds that of traditional thermocouples; and it can operate stably in electromagnetic interference environments such as high-voltage substations and radar stations without the risk of signal distortion.

[0037] RFID temperature sensors (Radio Frequency Identification Temperature Sensors) can withstand ambient temperatures ranging from -40°C to 220°C, making them suitable for high-temperature environments such as substations and metallurgical industries. They draw power from the reader's radio frequency signal, eliminating the challenges of power cable routing. They are particularly suitable for measuring the temperature of high-voltage equipment (such as switchgear contacts) and rotating components.

[0038] PT100 sensors (platinum resistance temperature sensors) can compensate for wire resistance errors through three-wire / four-wire wiring and reduce the impact of line interference.

[0039] In some embodiments, the steps for establishing the partial discharge and temperature correlation model include: collecting partial discharge characteristic quantities and thermal imaging data for multiple time periods and fitting the Hurst exponent; establishing a temperature rise mapping formula through the Hurst exponent; and establishing a dynamic temperature rise model according to the temperature rise mapping formula.

[0040] After establishing the dynamic temperature rise model, the position of the partial discharge can be accurately captured according to the positioning of the partial discharge sensor in combination with the dynamic temperature rise model, improving the diagnostic accuracy of insulation deterioration and the detection rate of instantaneous temperature anomalies.

[0041] Specifically, establishing the temperature rise mapping formula through the Hurst exponent is based on the fact that the partial discharge signal shows a non-uniform and self-similar scatter distribution in the time-frequency domain (such as the PRPD pattern), and its complexity can be described by the fractal dimension (box dimension). The partial discharge signal is converted into a binary image, and an improved grid covering method is used to count the number of grid-covered scatter regions of different sizes, and the fractal dimension is obtained by fitting the slope in the double logarithmic coordinate. Specifically, through the formula , where is the grid side length is the number of non-empty grids covering the discharge area under is the fractal dimension.

[0042] The Hurst exponent quantifies the long-range correlation between the partial discharge time series and the temperature time series in the fractal grid through rescaled range analysis (R / S analysis). The original sequence of length (partial discharge signal and temperature time series) is divided into multiple sub-intervals, each sub-interval with a length of , and there are a total of intervals. The data of each sub-interval is zero-meaned to eliminate the baseline offset interference: where is the sub-interval mean. Calculate the cumulative deviation of each sub-interval , representing the cumulative effect of the sequence deviating from the mean: . Define the range (R) as the difference between the maximum and minimum values of the cumulative deviation: , define S as the standard deviation, and calculate the value for each sub-interval to eliminate the dimension effect: , repeat the above steps for multiple sub-interval lengths , fit the and linear relationship to obtain the formula: where is the intercept; is the slope, i.e., the Hurst exponent.

[0043] Finally, the local temperature rise mapping model formula can be obtained: where, is the box dimension, is the Hurst exponent; is the thermal conductivity of the material; .

[0044] Construct a dynamic temperature rise model according to the local temperature rise mapping model formula: Laplace , where k is the thermal conductivity constant; among them, ; among them, (According to the current thermal effect, is the thermal conductivity of the material, ρ is the density, and c is the specific heat capacity), (Estimation of partial discharge temperature rise).

[0045] Using this formula can quickly obtain the situation of local dynamic temperature rise.

[0046] In some embodiments, the current-temperature mapping relationship is based on the current load dynamic temperature field formula: implemented by t; where is the equivalent resistance of the device, is the thermal conductivity of the material, represents the temperature change at time t; represents the current load changing with time.

[0047] In some embodiments, after performing step S2, the following steps are further included: using the thermal imaging data as the plane coordinates and calibrating the detection positions of each temperature sensor, calibrating the partial discharge source position according to the partial discharge and temperature correlation model, and calibrating the detection positions of the current sensors according to the current-temperature mapping relationship. By constructing a unified spatial reference, the coordinate system difference is eliminated. Since each sensor (such as temperature sensor, current sensor, partial discharge sensor) has an independent coordinate system, it needs to be converted to the global coordinate system through calibration to establish an accurate mapping relationship between temperature data and physical positions. It can avoid the situation where the thermodynamic analysis fails due to the coordinate system splitting.

[0048] Specifically, using the thermal imaging data as the plane coordinates and calibrating the detection positions of each temperature sensor, calibrating the partial discharge source position, and calibrating the detection positions of the current sensors require the following processing: based on the thermal imaging plane coordinates.

[0049] 1. Time alignment: Use the sliding window mechanism to compensate for the transmission delay (window size Δt = 1s) where is the delay time of each sensor, calibrated through the NTP protocol.

[0050] 2. Time-frequency analysis of the partial discharge signal: Extract the equivalent time width and the center frequency wherein is the partial discharge time-domain signal, and is its Fourier transform.

[0051] 3. Calculation of partial discharge spatial matching degree: Coordinates of partial discharge correlation points Coincidence degree with the high-temperature area of the thermal imaging. Among them, by comparing the thermal imaging temperature with the high-temperature threshold , the high-temperature area is identified; calculate the coordinates of the partial discharge correlation points and the central coordinates of the high-temperature area The distance between ) Through a specific function or coefficient and temperature-related constant , calculate the matching degree Match; when Match ≥ 0.9, it is determined as a valid correlation point ( is the high-temperature threshold, ); when the matching degree Match ≥ 0.9, it is considered that there is a valid correlation between the partial discharge point and the high-temperature area, indicating that the equipment has a specific fault or overheating phenomenon.

[0052] Joule heat model of current sensor: Based on current load Dynamic temperature field; wherein is the equivalent resistance of the equipment, is the thermal conductivity of the material.

[0053] 4. Calculation of current spatial matching degree: Coincidence degree between the coordinates of the current correlation point and the high-temperature area of the thermal imaging. Use the spectrogram normalized cross-correlation algorithm to calculate the coordinates of , when Match ≥ 0.9, it is determined as a valid correlation point.

[0054] Specifically, the calculation of the matching degree and the confirmation of the correlation point of other single-point temperature measurement sensors are the same as those of the current sensor.

[0055] After aligning the acquisition times of the sensors, spatial registration under time synchronization can be achieved through coordinate calibration, avoiding the "ghosting" phenomenon in the fused data. Adopting a moving window mechanism to compensate for the transmission delay can eliminate the time sequence misalignment of multi-sensors, suppress network transmission jitter, and support the reconstruction of the dynamic temperature field.

[0056] This partial discharge spatial matching degree calculation formula can evaluate the matching degree between the partial discharge point and the high-temperature area by comprehensively considering the thermal imaging temperature and distance factors.

[0057] In some embodiments, the processing model is established by dynamically adjusting the weight distribution of the output confidence degrees of thermal imaging data, partial discharge data, current data, and sensor temperature data based on the D-S evidence theory, and setting an output confidence degree threshold to establish a dynamic weight fusion diagnosis mechanism.

[0058] Through the improved D-S evidence theory, adaptively and dynamically adjust the weights, fuse multi-source sensor data such as temperature, current, thermal imaging, and single-point temperature measurement, and perform abnormal probability output and abnormal diagnosis.

[0059] Allocate weights according to the data confidence degree, and at the same time, dynamically and adaptively adjust the sensor weights, fuse the output confidence degrees and diagnoses of multi-source data including thermal imaging, partial discharge, current, and single-point temperature measurement: set that within the point , time if the temperature rise is greater than , it is an abnormal temperature change (code name A), and if the temperature is greater than , it is an over-temperature anomaly (code name B), and this point is also a single-point temperature measurement correlation point, a current correlation point, and a partial discharge correlation point. At this time, the evidence body calculation can be obtained; , is . Among them, K is a constant, and the set confidence degree threshold is 0.5. When the abnormal confidence degree is greater than 0.5, a warning alarm is generated, and the alarm point is highlighted.

[0060] In some embodiments, the dynamic adjustment of the weight distribution is carried out through the formula: where, is the current weight of sensor i; is the previous weight of sensor i; is any temperature sensor or partial discharge sensor or thermal imager; is the learning rate; is the recent prediction error of sensor i. Among them, the recent prediction error of sensor i is estimated by the covariance and predicted value output by the Kalman filter. Taking the Kalman filter fusion of the thermal imaging and partial discharge correlation point positions as an example.

[0061] The state transition process is the thermal imaging temperature measurement value of the previous moment plus the partial discharge temperature rise value, and the measurement value is the thermal imaging temperature measurement value of the next moment. Then, the transfer matrix The state quantity x = [ T t -∆ t , Q PD ×∆ t] T , the measurement value Z = , is the measurement covariance, is the thermal imaging and partial discharge covariance matrix, the Kalman gain , the Kalman state update , the Kalman covariance update = 。

[0062] After obtaining the error covariance matrix, the confidence level m of each sensor can be calculated, and the Gaussian distribution integral is used , where is the standard deviation of the error.

[0063] Specifically, after the processing model processes the input data, it outputs three-level conclusions of "normal", "temperature rise warning", and "overtemperature alarm"; at the same time, linkage control signals are output. For example, when an alarm is triggered, the fault information is automatically uploaded to the operation and maintenance platform, and the fan is started for heat dissipation at the same time. If the temperature does not drop within 10 minutes, remote power-off protection is triggered.

[0064] In some embodiments, on the other hand, a temperature measurement and diagnosis system based on a box-type substation is provided, which adopts the temperature measurement and diagnosis method. By establishing a partial discharge and temperature correlation model based on the partial discharge data and temperature data, the diagnosis accuracy of insulation deterioration can be improved, and the detection rate of instantaneous temperature anomalies can be increased. By establishing a mapping relationship between current and temperature, the real-time impact of current load changes on temperature rise can be quantified. Therefore, the situation of false fault judgment can be reduced, and the accuracy of temperature measurement and diagnosis can be improved.

[0065] As Figure 2 shown, the blue line represents the case where only fiber optic single-point temperature sensors are set in the substation, and the temperature fluctuations of its output are large, while the yellow line is the temperature line output after inputting the partial discharge data, current data, and temperature data into the processing model by using the method of the present application. Compared with the blue line, the fluctuation amplitude of its temperature is smaller, and the line is smoother. Combining with Figure 3 shown, when diagnosing the temperature situation by the method of the present application, due to the dynamic weight processing of multi-data, its correct rate is higher than that of the traditional fiber optic single-point temperature sensor.

[0066] The above are only optional embodiments of the present application, and are not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application. Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those of ordinary skill in the art can make changes, modifications, replacements, and variations to the above embodiments within the scope of the present invention.

Claims

1. A temperature measurement and diagnosis method based on a box-type substation, characterized in that, It includes the following steps: S1. Obtain the partial discharge data, current data, and temperature data of the box-type substation; S2. Establish a partial discharge and temperature correlation model based on the partial discharge data and the temperature data; establish a mapping relationship between current and temperature based on the current thermal effect of the current data; S3. Establish a processing model based on the partial discharge and temperature correlation model, the mapping relationship between current and temperature, and the temperature data. Input the partial discharge data, the current data, and the temperature data into the processing model, and the processing model outputs a diagnostic result and a linkage control signal.

2. The temperature measurement and diagnosis method according to claim 1, wherein The temperature data includes thermal imaging data and sensor temperature data; in step S2, establish a partial discharge and temperature correlation model based on the partial discharge data and the thermal imaging data; in step S3, establish a processing model based on the partial discharge and temperature correlation model, the mapping relationship between current and temperature, and the sensor temperature data.

3. The temperature measurement and diagnosis method according to claim 2, wherein, The box-type substation includes a high-voltage chamber, a transformer chamber, and a low-voltage chamber; temperature sensors are installed on the cable head copper busbar of the high-voltage chamber, the windings and iron cores of the transformer chamber, the incoming and outgoing line copper busbars of the transformer chamber, and the incoming and outgoing line copper busbars of the low-voltage chamber; current sensors are provided on the three incoming lines and four outgoing lines of the transformer chamber; a plurality of partial discharge sensors are installed on the inner wall of the transformer; a thermal imager is also provided in the box-type substation.

4. The temperature measurement and diagnosis method according to claim 3, wherein The temperature sensors for the high-voltage side windings and iron cores of the transformer chamber and the incoming and outgoing line copper busbars of the low-voltage chamber are fiber optic single-point temperature measurement sensors; the temperature sensors for the cable head copper busbar of the high-voltage chamber and the incoming and outgoing line copper busbars of the transformer chamber are RFID temperature sensors; the temperature sensors for the low-voltage side windings and iron cores of the transformer chamber are PT100 sensors.

5. The temperature measurement and diagnosis method according to claim 4, wherein The steps for establishing the partial discharge and temperature correlation model include: collecting partial discharge characteristic quantities and thermal imaging data for multiple time periods and fitting the Hurst index; establishing a temperature rise mapping formula through the Hurst index; establishing a dynamic temperature rise model based on the temperature rise mapping formula.

6. The temperature measurement and diagnosis method according to any one of claims 3-5, characterized in that The current-temperature mapping relationship is based on the current load Dynamic temperature field formula: Realized by t; where is the equivalent resistance of the device, is the thermal conductivity of the material, represents the temperature change at time t; represents the current load varying with time.

7. The temperature measurement and diagnosis method according to claim 6, characterized in that After executing step S2, it further includes the steps of: using the thermal imaging data as a plane coordinate and calibrating the detection positions of the temperature sensors, calibrating the positions of partial discharge sources based on the partial discharge and temperature correlation model, and calibrating the detection positions of the current sensors based on the mapping relationship between current and temperature.

8. The temperature measurement and diagnosis method according to claim 7, characterized in that, The establishment of the processing model is through dynamically adjusting the weight distribution of the output confidence degrees of the thermal imaging data, the partial discharge data, the current data, and the sensor temperature data based on the D-S evidence theory, and setting a threshold for the output confidence degrees to establish a dynamic weight fusion diagnosis mechanism.

9. The temperature measurement and diagnosis method according to claim 8, characterized in that The dynamic adjustment of weight distribution is carried out through the formula: wherein, is the current weight of sensor i; is the previous weight of sensor i; is any temperature sensor, or partial discharge sensor, or thermal imager; is the learning rate; is the recent prediction error of sensor i.

10. The temperature measurement and diagnosis system based on the box-type substation is characterized in that, It is carried out by using the temperature measurement and diagnosis method according to any one of claims 1-9.

Images