A cable power transmission and transformation equipment diagnosis device based on multispectral image processing

Abstract

The application belongs to the technical field of power transmission and transformation equipment diagnosis, and particularly relates to a cable power transmission and transformation equipment diagnosis device based on multispectral image processing, which comprises a multispectral image processing module, a path detection module, a temperature identification module, a path identification module, an abnormal area identification module, an image interception module and a diagnosis module. The multispectral image processing module is used for acquiring composite image data of the cable power transmission and transformation equipment. The path detection module is used for monitoring the current value change of the cable power transmission and transformation equipment and tracking the current path of the changed current. The temperature identification module is used for extracting infrared light detection data from the composite image data and judging whether there is a temperature abnormal point exceeding a temperature threshold in the infrared light detection data. Through the identification and processing of the temperature, the false positives caused by interference factors are reduced, and the data upload frequency is reduced. The composite image data corresponding to the abnormal area is uploaded, the data upload amount is reduced, and then the bandwidth occupation is reduced.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of power transmission and transformation equipment diagnosis, and particularly relates to a cable power transmission and transformation equipment diagnosis device based on multi-spectral image processing. BACKGROUND

[0002] The cable power transmission and transformation equipment is a key facility for power transmission, distribution and control in a power system, mainly including core components such as transformers, switch devices and protection devices. The cable power transmission and transformation equipment transforms voltage and controls current flow to ensure that power is efficiently and safely transmitted from a power plant to a user end. It is necessary to diagnose the cable power transmission and transformation equipment in real time to ensure stable operation of the power system.

[0003] In the prior art, multi-spectral image processing technology is usually used to collect data of the cable power transmission and transformation equipment, and then the data is uploaded to the cloud for diagnosis. Due to a large amount of uploaded data, a problem of insufficient bandwidth may occur. In turn, the image compression rate is increased, and key details are lost. In severe cases, data dimensionality reduction may occur, spectral features are lost, and the accuracy of diagnosis is reduced. In addition, when multi-spectral image collection is performed, visible light interference (such as strong sunlight and reflected light) in the environment may cause local overexposure and color distortion of the image, affecting the accuracy of diagnosis. It is necessary to increase the data collection frequency, which in turn increases the bandwidth occupation.

[0004] Therefore, the application provides a diagnosis device based on multi-spectral image processing technology, which detects the temperature of the cable power transmission and transformation equipment first, filters out abnormal points, and then uploads data around the abnormal points to reduce the bandwidth occupation, thereby improving the accuracy of diagnosis. SUMMARY

[0005] The application aims to provide a cable power transmission and transformation equipment diagnosis device based on multi-spectral image processing to solve the problem of reduced diagnosis accuracy caused by high bandwidth occupation in data uploading.

[0006] To achieve the above-mentioned purpose, the application provides the following technical scheme: a cable power transmission and transformation equipment diagnosis device based on multi-spectral image processing, comprising:

[0007] A multi-spectral image processing module is configured to acquire composite image data of the cable power transmission and transformation equipment, and the composite image data includes infrared light detection data, ultraviolet light detection data and visible light detection data.

[0008] A path detection module is configured to monitor the change of current value passing through the cable power transmission and transformation equipment, and track the current path of the changed current.

[0009] A temperature identification module is configured to extract the infrared light detection data from the composite image data, and determine whether there is a temperature abnormal point exceeding a temperature threshold.

[0010] A path identification module is configured to track the temperature abnormal point and determine whether the temperature abnormal point is on the current path.

[0011] An abnormal area identification module is configured to determine whether the temperature abnormal point is an abnormal area. When the temperature abnormal point is on the current path, the temperature abnormal point is calculated by the current value, and then the maximum temperature of the temperature abnormal point is calculated as an over-limit threshold value by the heat dissipation capacity of the cable power transmission and transformation equipment and the environmental heat dissipation capacity. Then, it is determined whether the temperature of the temperature abnormal point exceeds the over-limit threshold value. If not, the temperature abnormal point is a normal area. If yes, the temperature abnormal point is an abnormal area. When the temperature abnormal point is not on the current path, the temperature abnormal point is also regarded as an abnormal area.

[0012] An image capturing module is configured to capture data corresponding to the abnormal area in the composite image data and upload the data.

[0013] A diagnosis module is configured to receive the data uploaded by the image capturing module and perform diagnosis.

[0014] Preferably, when the current path is divided, all points affected by the same input current form the same current path.

[0015] When there are multiple current paths, the path detection module can simultaneously monitor the value changes of the currents passing through all the current paths.

[0016] Preferably, the composite image data is subjected to grid processing. When the area where the temperature abnormal point is located is an abnormal area, the dispersion of the temperature abnormal point is calculated. When the temperature abnormal points are dispersed, the temperature abnormal points are located at the center of the grid. When the temperature abnormal points are concentrated, the concentrated temperature abnormal points are concentrated inside the grid. The image capturing module captures the corresponding data in the grid and uploads the data.

[0017] Preferably, the error correction module is further included. When multiple temperature abnormal points exist simultaneously, the error correction module analyzes whether the temperature abnormal points are concentrated. When the temperature abnormal points are concentrated and the center area of the temperature abnormal points corresponds to a normal area, the area corresponding to the concentrated temperature abnormal points is a normal area.

[0018] Preferably, the bandwidth allocation module is further included. The bandwidth allocation module is configured to monitor the changes in the communication environment in real time and dynamically allocate the bandwidth. When the bandwidth is sufficient, the composite image data with ultraviolet light and visible light detection data is automatically uploaded.

[0019] Preferably, the multi-spectral image processing module has an angle adjustment function and can collect the composite image data of the cable power transmission and transformation equipment at different angles.

[0020] Preferably, the temperature threshold is a dynamic threshold, and the normal temperature of the cable power transmission and transformation equipment under different temperature environments is learned by a neural network as the temperature threshold.

[0021] Compared with the prior art, the present application has the following advantages:

[0022] 1) The device divides the current path in the cable power transmission and transformation equipment, and when a temperature abnormal point occurs, the relevance of the temperature abnormal point and the current path is analyzed first, and whether the temperature of the temperature abnormal point exceeds the standard is further judged through the temperature change caused by the current change in the current path, so as to determine whether the temperature abnormal point is an abnormal area. Through the identification and processing of temperature, false positives caused by interference factors are reduced, data upload frequency is reduced, and bandwidth occupation is reduced.

[0023] 2) The device analyzes the temperature abnormal point, finds out the abnormal area, and then uploads the corresponding composite image data of the abnormal area, which can reduce the amount of data uploaded, thereby reducing the bandwidth occupation. BRIEF DESCRIPTION OF DRAWINGS

[0024] Figure 1 The present application is a schematic diagram of the module.

[0025] Figure 2 The present application is a schematic diagram of the process. DETAILED DESCRIPTION

[0026] Please refer to Figure 1 、 2 A cable power transmission and transformation equipment diagnosis device based on multi-spectral image processing, comprising: a multi-spectral image processing module, a path detection module, a temperature identification module, a path identification module, an abnormal area identification module, a correction module, an image interception module, a diagnosis module and a bandwidth allocation module.

[0027] The multi-spectral image processing module uses an existing multi-spectral camera, which can be mounted on a drone, a patrol robot or a fixed monitoring device to capture image information of the cable power transmission and transformation equipment under different spectral bands. It can detect the cable power transmission and transformation equipment through infrared light, ultraviolet light and visible light to obtain composite image data. Among them, the detection data of infrared light is temperature data, the detection data of ultraviolet light is ultraviolet radiation data generated by corona, partial discharge, etc., and the detection data of visible light is detail wheel frame data of the cable power transmission and transformation equipment. When the multi-spectral camera is fixed on the monitoring device, it can be adjusted in angle with the monitoring device, and can collect composite image data of the cable power transmission and transformation equipment at different angles. When the multi-spectral image processing module is currently monitoring the cable power transmission and transformation equipment, the collection position of the multi-spectral image processing module can be adjusted to avoid obstruction.

[0028] The path detection module adopts the existing current transformer. The large current in the cable power transmission and transformation equipment is converted into small current by the current transformer in proportion, and then the ammeter is connected on the secondary side circuit to measure the current, and then the numerical change of the current on the current path is identified. When dividing the current path, all points affected by the same input current form the same current path, that is, the currents passing through all electrical elements in the same current path are related to each other. When the current in an electrical element changes, the current in other electrical elements will also change accordingly, so that the numerical value of the ammeter can be used to calculate the current value in other electrical elements. For example, in a series circuit, the passing current in each electrical element is equal, so the corresponding points of all electrical elements in the series circuit can form a current path. Or the currents in the two parallel circuits are always in a fixed ratio. When the current in one circuit is measured, the current in the other circuit can be calculated, so the two parallel circuits belong to the same current path.

[0029] When there are multiple current paths in the cable power transmission and transformation equipment, multiple sets of path detection modules need to be equipped. When the current changes, the numerical value of the passing current in all current paths can be monitored at the same time. In order to avoid delay in data monitoring, all current paths that need to be found (current paths that change current) are found. When the current changes, the heat generated by the resistance also changes. By identifying the current path passing through the cable power transmission and transformation equipment through current change, the temperature change on the subsequent current path can be predicted.

[0030] The neural network learns the past temperature data of the cable power transmission and transformation equipment, formulates the temperature fluctuation interval of the cable power transmission and transformation equipment in different seasons and different time periods, and uses the temperature fluctuation interval as the temperature threshold. The temperature recognition module extracts the infrared light detection data from the composite image data, identifies the temperature of the cable power transmission and transformation equipment through the infrared light detection data, and then judges whether there is a corresponding point exceeding the temperature threshold in the temperature. The corresponding point exceeding the temperature threshold is a temperature abnormal point.

[0031] A coordinate system is established in the composite image data, so that the temperature data in the infrared light detection data and the image of the cable power transmission and transformation equipment in the visible light detection data have corresponding coordinates. Then the corresponding coordinates of the electrical elements are established by the corresponding positions of the electrical elements in the cable power transmission and transformation equipment. The projection of the current path in the coordinate system is constructed through the connection of the corresponding coordinates of the electrical elements. Then the path recognition module analyzes whether the temperature abnormal point coincides with the projection of the current path. When they coincide, the temperature abnormal point is on the current path.

[0032] Temperature anomaly, which may be caused by current change, or may be caused by shielding of heat dissipation part, corona discharge, etc., the former will not cause cable transmission and transformation equipment anomaly when the current is still within the normal range, the latter is easy to cause cable transmission and transformation equipment anomaly, and the area which can cause cable transmission and transformation equipment anomaly is the abnormal area. When the path recognition module determines that the temperature anomaly point is on the current path, it indicates that the temperature may be caused by current change, at this time, the highest temperature of the electrical element is calculated by Joule's law and heat conduction law as the overrun threshold, and then it is judged whether the temperature of the temperature anomaly point exceeds the overrun threshold, if not, it is a normal area, if yes, it is an abnormal area.

[0033] First, the power of the temperature anomaly point is calculated.

[0034] P W =(k×I D ) 2 ×R W .

[0035] Wherein, P W represents the power of the temperature anomaly point.

[0036] k represents the ratio of the current of the temperature anomaly point to the measured current in the current path.

[0037] I D represents the measured current value in the current path where the temperature anomaly point is located.‌

[0038] R W represents the resistance of the electrical element of the temperature anomaly point.

[0039] Then the theoretically increased temperature of the temperature anomaly point is calculated.

[0040] ΔT=P W ×R θ .

[0041] ΔT represents the increased temperature of the anomaly point.

[0042] R θ represents the thermal resistance of the electrical element of the temperature anomaly point (R θ is taken as the reference of the current temperature of the temperature anomaly point).

[0043] Finally, the highest temperature of the temperature anomaly point is calculated by combining the heat dissipation capacity of the cable transmission and transformation equipment and the environmental heat dissipation capacity.

[0044] T L =ΔT+T a -T b .

[0045] T L represents the highest temperature of the temperature anomaly point that can be theoretically reached.

[0046] T a represents the current ambient temperature.

[0047] T b represents the cable power transmission and transformation equipment heat dissipation capacity and the environmental heat dissipation capacity theoretically reduced temperature (T b When taking value, the current temperature of the temperature anomaly point is taken as reference).

[0048] T L as an over-limit threshold, determine whether the temperature of the temperature anomaly point exceeds the over-limit threshold, if not, it means that the temperature of the temperature anomaly point is within the normal range, and the area corresponding to the temperature anomaly point is a normal area. If it exceeds, it means that the temperature is abnormal, that is, the area corresponding to the temperature anomaly point is an abnormal area. If the temperature anomaly point is not on the current path, it is also regarded as an abnormal area, that is, the temperature anomaly is not caused by current fluctuation at this time.

[0049] The abnormal area indicates that the cable power transmission and transformation equipment is abnormal, at this time it is difficult to judge the specific problem relying on a single temperature data, and it is necessary to upload the composite image data. But the overall upload of the composite image data will excessively occupy the bandwidth, resulting in that the diagnosis cannot be carried out in real time. Therefore, the image cutting module is adopted to cut the data corresponding to the abnormal area from the composite image data, and then upload the data corresponding to the abnormal area to the diagnosis module. The diagnosis module receives the data uploaded by the image cutting module and carries out diagnosis. The diagnosis module uses the existing technology to process multispectral data through the architecture of Transformer, detects whether there is a defect in the uploaded composite image data, and automatically marks the defect.

[0050] The composite image data is subjected to grid processing, and the data in each grid can be independently uploaded. When the area where the temperature anomaly point is located is an abnormal area (that is, the composite image data needs to be cut and uploaded), the dispersion of the coordinate point corresponding to the temperature anomaly point is calculated, and the standard deviation of the x coordinate and the y coordinate of the temperature anomaly point is obtained. Then the ratio of the x coordinate standard deviation to the single grid horizontal length and the ratio of the y coordinate standard deviation to the single grid vertical length are calculated, when both ratios are less than 0.5, it means that the temperature anomaly points are concentrated. Otherwise, it means that the temperature anomaly points are dispersed. When the temperature anomaly points are dispersed, the grid position is adjusted so that each temperature anomaly point is located at the center of the grid. When there is intersection between grids, the boundary of the intersecting grid is taken as a new grid, and the composite image data is cut and processed. When the temperature anomaly points are concentrated, the grid position is adjusted to concentrate the concentrated temperature anomaly points in the inside of the grid as much as possible, and then the composite image data is cut and processed. When the image cutting module cuts the composite image data, the corresponding data in the selected grid is uploaded.

[0051] Due to the problem of heat transfer, when a point has temperature anomaly, the surrounding area can also have temperature anomaly, at this time, only part of the temperature anomaly points can be on the current path, and other temperature anomaly points are not on the current path, if the area where the temperature anomaly point is located on the current path is a normal area, directly positioning other temperature anomaly points in the surrounding area as abnormal areas can easily misjudge. Through the error correction module, whether the temperature anomaly point has aggregation phenomenon is analyzed. First, the temperature of the temperature anomaly point is detected to analyze whether the adjacent point has gradient, if yes, it indicates that the temperature anomaly point has aggregation phenomenon. Then, the highest temperature area is found through the gradient difference, and the area is the heat source area, and the heat source area is usually in the center area of the temperature gradient, and then whether the temperature anomaly point in the heat source area corresponds to the normal area is judged, if yes, it indicates that the corresponding area of the aggregated temperature anomaly point is the normal area. If not, it indicates that the corresponding area of the aggregated temperature anomaly point is the derived area of the abnormal area, and the derived area does not need to intercept data for uploading. If the temperature anomaly point does not have aggregation phenomenon, the temperature anomaly point needs to intercept data for uploading.

[0052] The composite image data obtained by the multispectral image processing module needs to be uploaded, and the bandwidth ratio can be effectively reduced through the application. When there is no abnormal temperature point, less data is uploaded, which can reduce the bandwidth utilization. And the infrared detection can also have some abnormalities that cannot be found in time, and these abnormalities can also cause cable transmission and transformation equipment abnormalities after a long time accumulation. Therefore, the bandwidth allocation module is used to monitor the change of the communication environment in real time, and when the bandwidth is sufficient, the composite image data with ultraviolet light and visible light detection data is automatically uploaded to the diagnosis module, so that the diagnosis module can identify other abnormalities that are difficult to cause temperature change. Because the problems caused by these abnormalities need a certain time to break out, uploading information when idle will not cause information omission, and ensures that the abnormalities can be found. At the same time, it also will not cause the bandwidth to be idle.

Claims

1. A diagnostic device for cable transmission and transformation equipment based on multispectral image processing, characterized in that, include: The multispectral image processing module is used to acquire composite image data of cable power transmission and transformation equipment. The composite image data includes infrared light, ultraviolet light and visible light detection data. The path detection module is used to monitor changes in current values ​​through cable power transmission and transformation equipment and to track the current path through which the changing current passes. The temperature recognition module is used to extract infrared light detection data from composite image data and determine whether there are any temperature anomalies that exceed the temperature threshold. The path identification module is used to track abnormal temperature points and determine whether they are on the current path. The abnormal area identification module is used to determine whether a temperature anomaly point is an abnormal area; When the temperature anomaly is on the current path, the temperature rise of the anomaly is calculated by the current value. Then, the maximum temperature that the temperature anomaly can reach is calculated by taking into account the heat dissipation capacity of the cable transmission and transformation equipment and the heat dissipation capacity of the environment, and used as the over-limit threshold. Then, it is determined whether the temperature of the temperature anomaly exceeds the over-limit threshold. If it does not exceed the threshold, it is in the normal area; if it exceeds the threshold, it is in the abnormal area. When the temperature anomaly is not on the current path, it is also regarded as an abnormal area. The image cropping module is used to crop the data corresponding to abnormal areas in composite image data and upload it. The diagnostic module is used to receive data uploaded by the image capture module and perform diagnostics. The error correction module analyzes whether there is a clustering phenomenon when multiple temperature anomalies exist simultaneously. If temperature anomalies cluster and the temperature anomaly in the central area corresponds to a normal area, then the area corresponding to the clustered temperature anomalies is a normal area.

2. The cable transmission and transformation equipment diagnostic device based on multispectral image processing according to claim 1, characterized in that: When dividing current paths, all points affected by the same input current form the same current path; When there are multiple current paths, the path detection module can simultaneously monitor the changes in the current values ​​passing through all current paths.

3. The cable transmission and transformation equipment diagnostic device based on multispectral image processing according to claim 1, characterized in that: The composite image data is processed into a grid. When the area where the temperature anomaly point is located is an anomaly area, the dispersion of the temperature anomaly point is calculated. When the temperature anomaly points are scattered, the temperature anomaly points are located at the center of the grid. When the temperature anomaly points are concentrated, the clustered temperature anomaly points are concentrated inside the grid. The image cropping module crops the corresponding data of the composite image data in the grid and uploads it.

4. The cable transmission and transformation equipment diagnostic device based on multispectral image processing according to claim 1, characterized in that: It also includes a bandwidth allocation module, which is used to monitor changes in the communication environment in real time and dynamically allocate bandwidth; when bandwidth is sufficient, it automatically uploads composite image data with ultraviolet and visible light detection data.

5. The cable transmission and transformation equipment diagnostic device based on multispectral image processing according to claim 1, characterized in that: The multispectral image processing module has an angle adjustment function, which can acquire composite image data of cable transmission and transformation equipment at different angles.

6. The cable transmission and transformation equipment diagnostic device based on multispectral image processing according to claim 1, characterized in that: The temperature threshold is a dynamic threshold, and the normal temperature of the cable transmission and transformation equipment under different temperature environments is learned through a neural network and used as the temperature threshold.

Images