阅读说明：本技术 一种开关柜非介入式监测方法、装置、存储介质 (Non-intrusive monitoring method and device for switch cabinet and storage medium ) 是由 郭晨华 潘晨曦 宁松浩 汪俊 杨志强 于 2021-08-23 设计创作，主要内容包括：本发明公开了一种开关柜非介入式监测方法、装置、存储介质,本方法包括以下步骤：步骤S1、获取拍摄的红外图像与可见光图像,融合所述红外图像与所述可见光图像,形成双频融合图像；步骤S2、识别所述双频融合图像、红外图像与可见光图像,进行图像处理,以提取设备状态信息；步骤S3、根据所述设备状态信息及温度分布信息,根据非介入测温模型计算开关柜的集总热源温度；步骤S4、将所述非介入测温模型计算得到的开关柜的集总热源温度,与预设故障诊断阈值进行比较,对所述开关柜的运行状态进行诊断。本发明解决传统高压开关柜的温度监测中的传感器的数量多和安装工程复杂等问题。(The invention discloses a non-intrusive monitoring method, a non-intrusive monitoring device and a non-intrusive monitoring storage medium for a switch cabinet, wherein the method comprises the following steps of: s1, acquiring a shot infrared image and a shot visible light image, and fusing the infrared image and the visible light image to form a double-frequency fused image; step S2, recognizing the dual-frequency fusion image, the infrared image and the visible light image, and performing image processing to extract equipment state information; step S3, calculating the temperature of the lumped heat source of the switch cabinet according to the equipment state information and the temperature distribution information and a non-intrusive temperature measurement model; and step S4, comparing the temperature of the lumped heat source of the switch cabinet calculated by the non-intrusive temperature measurement model with a preset fault diagnosis threshold value, and diagnosing the operation state of the switch cabinet. The invention solves the problems of large quantity of sensors, complex installation engineering and the like in the temperature monitoring of the traditional high-voltage switch cabinet.)

1. A non-intrusive monitoring method for a switch cabinet is characterized by comprising the following steps:

s1, acquiring a shot infrared image and a shot visible light image, and fusing the infrared image and the visible light image to form a double-frequency fused image;

step S2, recognizing the dual-frequency fusion image, the infrared image and the visible light image, and performing image processing to extract equipment state information;

step S3, calculating the temperature of the lumped heat source of the switch cabinet according to the equipment state information and the temperature distribution information and a non-intrusive temperature measurement model;

and step S4, comparing the temperature of the lumped heat source of the switch cabinet calculated by the non-intrusive temperature measurement model with a preset fault diagnosis threshold value, and diagnosing the operation state of the switch cabinet.

2. The non-intrusive monitoring method for a switchgear cabinet as recited in claim 1, wherein the step S2 includes the steps of:

step S21, recognizing the dual-frequency fusion image, the infrared image and the visible light image, and performing image processing to extract equipment state information;

step S22, recognizing characters of the equipment label according to the target detection frame image of the visible light image, and naming the target detection frame image of the double-frequency fusion image, the infrared image and the visible light image according to the character recognition result;

step S23, performing equipment region segmentation according to the double-frequency fusion image and the target detection frame image of the infrared image to obtain an infrared equipment region segmentation image; and extracting temperature distribution information of the infrared equipment region segmentation image through an infrared image temperature algorithm.

3. The non-intrusive monitoring method for a switchgear cabinet as recited in claim 1, wherein the step S3 further comprises the steps of:

and step S31, selecting a dynamic non-intervention temperature measurement model or a static non-intervention temperature measurement model according to the equipment state information and the temperature distribution information of the switch cabinet, and calculating the temperature of the switch cabinet, wherein the non-intervention temperature measurement model comprises a dynamic non-intervention temperature measurement model and a static non-intervention temperature measurement model.

4. The non-intrusive monitoring method of claim 3, wherein the dynamic non-intrusive temperature measurement model is

Wherein, theta₁(t) is the temperature of the lumped heat source in a certain interval inside the high-voltage switch cabinet; theta₂(t) is the highest temperature of each area at the top of the switch cabinet; theta₀(t) taking the lowest temperature of a cabinet surface on the front side of the cabinet body as the ambient temperature of the cabinet body of the switch cabinet; a. the₁₂The heat transfer thermal resistance coefficient between the lumped heat source of the switch cabinet and the cabinet top surface is constant; a. the₂₃The coefficient of heat transfer thermal resistance between the highest temperature of each area at the top of the switch cabinet and the ambient temperature of the switch cabinet is constant; b is₁₂The thermal capacitance coefficient is the sum of all substances in the highest temperature isothermal surface of each area of the top of the switch cabinet and is a constant; t is the double-view image sampling time, then d θ₂(t)/dt represents the derivative of temperature with respect to time.

5. The non-intrusive method of monitoring a switchgear as recited in claim 3, wherein the simplified form of the static non-intrusive thermometry model is:

this formula can be expressed in simplified form as:

θ₁(t)＝θ₂(t)+kΔθ₂(t)

wherein the content of the first and second substances,static model coefficients of a non-intrusive thermometry algorithm; a. the₁₂The heat transfer thermal resistance coefficient between the lumped heat source of the switch cabinet and the cabinet top surface is constant; a. the₂₃The coefficient of heat transfer thermal resistance between the highest temperature of each area at the top of the switch cabinet and the ambient temperature of the switch cabinet is constant; theta₁(t) is the temperature of the lumped heat source in a certain interval inside the high-voltage switch cabinet; theta₂(t) is the highest temperature of each area at the top of the switch cabinet; theta₀And (t) taking the lowest temperature of the cabinet surface on the front side of the cabinet body as the ambient temperature of the cabinet body of the switch cabinet.

6. The non-intrusive monitoring method for the switch cabinet according to claim 1, wherein in step S4, the lumped heat source temperature of the switch cabinet calculated by the non-intrusive temperature measurement model is compared with a preset fault diagnosis threshold value, so as to diagnose the operating state of the switch cabinet, specifically, a first diagnosis method is used to diagnose the operating state of the switch cabinet, and the first diagnosis method is:

when in useOrIn time, the switch cabinet runsThe diagnosis result of the state is normal;

when in useOrThen, the diagnosis result of the running state of the switch cabinet is early warning;

when in useOrThen, the diagnosis result of the running state of the switch cabinet is an alarm;

wherein theta is₁(t) is the temperature of the lumped heat source in a certain interval inside the switch cabinet; delta theta₁(t) is θ₁(t)-θ₀(t)；θ₀(t) taking the lowest temperature of the cabinet surface on the front side of the cabinet body as the cabinet body environment temperature;a temperature threshold for an "over-temperature warning" condition for the lumped heat source;a temperature threshold for a lumped heat source "over-temperature alarm" condition;a temperature rise threshold for an integrated heat source 'over-temperature warning' state;is the temperature rise threshold value of the lumped heat source overheat alarm state.

7. The method according to claim 6, wherein the step S4 of comparing the temperature of the lumped heat source of the switch cabinet calculated by the non-intrusive temperature measurement model with a preset fault diagnosis threshold to diagnose the operating state of the switch cabinet further comprises: diagnosing the operation state of the switch cabinet by adopting a second diagnosis method, wherein the second diagnosis method comprises the following steps:

when in useOrWhen the operation state of the switch cabinet is normal, the diagnosis result of the operation state of the switch cabinet is normal;

when in useOrThen, the diagnosis result of the running state of the switch cabinet is early warning;

when in useOrThen, the diagnosis result of the running state of the switch cabinet is an alarm; theta₂(t) is the highest temperature of each area at the top of the switch cabinet;

wherein, Delta theta₂(t) is θ₂(t)-θ₀(t)；θ₀(t) taking the lowest temperature of the cabinet surface on the front side of the cabinet body as the cabinet body environment temperature;a temperature threshold for an "over-temperature warning" state for a non-intervention point;a temperature threshold for a non-intrusive point "overheat alarm" condition;a temperature rise threshold value in an overheat early warning state of a non-intervention point;a temperature rise threshold value in an overheat alarm state of a non-intervention point;from said first diagnostic method according to a non-invasive algorithmic modelAnd (3) reversely calculating to obtain:

8. the non-intrusive monitoring method for switch cabinets of claim 7, wherein the step S4 further comprises: diagnosing the operating state of the switch cabinet by adopting a transverse comparison method and a longitudinal comparison method, wherein the transverse comparison method is to compare the operating state of the switch cabinet with the same type of equipment on the basis of the second diagnosis method so as to confirm the equipment with the highest temperature rise; the longitudinal comparison method is to compare the current monitoring data of the equipment with the historical data of the equipment so as to judge whether the equipment is abnormal or not.

9. The device is characterized by comprising a camera module and a processing module, wherein the camera module comprises an infrared camera and a visible light camera, the angles of the lenses of the infrared camera and the visible light camera are the same, and a preset distance exists between the infrared camera and the visible light camera and the top of a switch cabinet so as to shoot real-time images or videos of the switch cabinet; the real-time image or video at least comprises a preset area of the switch cabinet; the preset distance is 0.5-1 m;

the processing module monitors the working state of the switch cabinet according to the real-time image or video of the camera module; the processing module comprises a processor, a memory and a computer program which is stored on the memory and can run on the processor, and when the processor executes the computer program, the non-intrusive monitoring method for the switch cabinet is realized according to any one of claims 1 to 8.

10. A storage medium having stored thereon a computer program which, when executed, implements the method of non-intrusive monitoring of a switchgear panel as claimed in any of claims 1 to 8.

Technical Field

The invention relates to the technical field of switch cabinet monitoring, in particular to a non-intrusive monitoring method and device for a switch cabinet and a storage medium.

Background

With the rapid development of economy in China, electric power safety plays a crucial role in the national security and security system and economic development, and how to ensure the safety, stability and economic operation of the electric power safety system and prevent catastrophic accidents is one of the major problems to be solved urgently. The distribution room is a place of main electrical equipment for distributing and transmitting electric energy in a power distribution network at the tail end of an electric power system, is an important component in an urban power distribution network system, and has the characteristics of wide distribution, large quantity, complex management and the like. With the increasing power supply load density, the number of the power distribution rooms is larger and larger.

At present, the distribution room mainly adopts a mode of routing inspection and first-aid repair according to regions for operation and maintenance management, and the so-called manual routing inspection is that routing inspection is carried out by using operating personnel and regular sampling inspection is carried out by testing personnel. The traditional state detection and analysis method comprises the steps of ultrasonic partial discharge, infrared temperature measurement, field environment observation and the like. However, with the increasing of the number of power distribution rooms and the increasing of the equipment capacity in recent years, the traditional equipment inspection maintenance method cannot meet the actual requirement of safe operation of modern power equipment, and is difficult to truly reflect the operation state and the fault state of various power equipment, so that the method has great limitation.

The replacement of regular maintenance by condition maintenance has become an essential trend of power system equipment maintenance, equipment damage and unplanned power failure caused by faults can be avoided or reduced, and the basis of condition maintenance is an online monitoring technology. However, the existing distribution room equipment state monitoring sensors are various in types, large in monitoring data quantity, inconvenient in communication connection, difficult in maintenance of inspection personnel and the like, and the problems of equipment loss, faults, low voltage quality, poor stability of a distribution network and the like caused by untimely maintenance are caused.

Disclosure of Invention

In order to overcome the defects of the prior art, one of the purposes of the invention is to provide a non-intrusive monitoring method for a switch cabinet, which solves the problems of large number of sensors, complex installation engineering and the like in the traditional temperature monitoring of the switch cabinet.

The invention also aims to provide a device for implementing the non-intrusive monitoring method of the switch cabinet, and solve the problems of large quantity of sensors, complex installation engineering and the like in the temperature monitoring of the traditional high-voltage switch cabinet.

The invention also aims to provide a storage medium for implementing a non-intrusive monitoring method of a switch cabinet, which solves the problems of large quantity of sensors, complex installation engineering and the like in the temperature monitoring of the traditional high-voltage switch cabinet

One of the purposes of the invention is realized by adopting the following technical scheme:

a non-intrusive monitoring method for a switch cabinet comprises the following steps:

s1, acquiring a shot infrared image and a shot visible light image, and fusing the infrared image and the visible light image to form a double-frequency fused image;

step S2, recognizing the dual-frequency fusion image, the infrared image and the visible light image, and performing image processing to extract equipment state information;

step S3, calculating the temperature of the lumped heat source of the switch cabinet according to the equipment state information and the temperature distribution information and a non-intrusive temperature measurement model;

and step S4, comparing the temperature of the lumped heat source of the switch cabinet calculated by the non-intrusive temperature measurement model with a preset fault diagnosis threshold value, and diagnosing the operation state of the switch cabinet.

Further, the step S2 includes the following steps:

step S21, recognizing the dual-frequency fusion image, the infrared image and the visible light image, and performing image processing to extract equipment state information;

step S22, recognizing characters of the equipment label according to the target detection frame image of the visible light image, and naming the target detection frame image of the double-frequency fusion image, the infrared image and the visible light image according to the character recognition result;

step S23, performing equipment region segmentation according to the double-frequency fusion image and the target detection frame image of the infrared image to obtain an infrared equipment region segmentation image; and extracting temperature distribution information of the infrared equipment region segmentation image through an infrared image temperature algorithm.

Further, the step S3 further includes the following steps:

and step S31, selecting a dynamic non-intervention temperature measurement model or a static non-intervention temperature measurement model according to the equipment state information and the temperature distribution information of the switch cabinet, and calculating the temperature of the switch cabinet, wherein the non-intervention temperature measurement model comprises a dynamic non-intervention temperature measurement model and a static non-intervention temperature measurement model.

Further, the dynamic non-invasive thermometry model is

Wherein, theta₁(t) is the temperature of the lumped heat source in a certain interval inside the high-voltage switch cabinet; theta₂(t) is the highest temperature of each area at the top of the switch cabinet; theta₀(t) taking the lowest temperature of a cabinet surface on the front side of the cabinet body as the ambient temperature of the cabinet body of the switch cabinet; a. the₁₂The heat transfer thermal resistance coefficient between the lumped heat source of the switch cabinet and the cabinet top surface is constant; a. the₂₃The coefficient of heat transfer thermal resistance between the highest temperature of each area at the top of the switch cabinet and the ambient temperature of the switch cabinet is constant; b is₁₂The thermal capacitance coefficient is the sum of all substances in the highest temperature isothermal surface of each area of the top of the switch cabinet and is a constant; t is the double-view image sampling time, then d θ₂(t)/dt represents the derivative of temperature with respect to time.

Further, the simplified form of the static non-invasive thermometry model is:

this formula can be expressed in simplified form as:

θ₁(t)＝θ₂(t)+kΔθ₂(t)

wherein the content of the first and second substances,static model coefficients of a non-intrusive thermometry algorithm; a. the₁₂The heat transfer thermal resistance coefficient between the lumped heat source of the switch cabinet and the cabinet top surface is constant; a. the₂₃The coefficient of heat transfer thermal resistance between the highest temperature of each area at the top of the switch cabinet and the ambient temperature of the switch cabinet is constant; theta₁(t) is the temperature of the lumped heat source in a certain interval inside the high-voltage switch cabinet; theta₂(t) is the highest temperature of each area at the top of the switch cabinet; theta₀And (t) taking the lowest temperature of the cabinet surface on the front side of the cabinet body as the ambient temperature of the cabinet body of the switch cabinet.

Further, in the step S4, the lumped heat source temperature of the switch cabinet calculated by the non-intrusive temperature measurement model is compared with a preset fault diagnosis threshold value, and the operation state of the switch cabinet is diagnosed, specifically, the operation state of the switch cabinet is diagnosed by using a first diagnosis method, where the first diagnosis method is:

when in useOrWhen the operation state of the switch cabinet is normal, the diagnosis result of the operation state of the switch cabinet is normal;

when in useOrThen, the diagnosis result of the running state of the switch cabinet is early warning;

when in useOrThen, the diagnosis result of the running state of the switch cabinet is an alarm;

wherein theta is₁(t) is the temperature of the lumped heat source in a certain interval inside the switch cabinet; delta theta₁(t) is θ₁(t)-θ₀(t)；θ₀(t) taking the lowest temperature of the cabinet surface on the front side of the cabinet body as the cabinet body environment temperature;a temperature threshold for an "over-temperature warning" condition for the lumped heat source;a temperature threshold for a lumped heat source "over-temperature alarm" condition;a temperature rise threshold for an integrated heat source 'over-temperature warning' state;is the temperature rise threshold value of the lumped heat source overheat alarm state.

Further, the step S4 of comparing the lumped heat source temperature of the switch cabinet calculated by the non-intrusive temperature measurement model with a preset fault diagnosis threshold value to diagnose the operating state of the switch cabinet further includes: diagnosing the operation state of the switch cabinet by adopting a second diagnosis method, wherein the second diagnosis method comprises the following steps:

when in useOrWhen the operation state of the switch cabinet is normal, the diagnosis result of the operation state of the switch cabinet is normal;

when in useOrThen, the diagnosis result of the running state of the switch cabinet is early warning;

when in useOrThen, the diagnosis result of the running state of the switch cabinet is an alarm; theta₂(t) is the highest temperature of each area at the top of the switch cabinet;

wherein, Delta theta₂(t) is θ₂(t)-θ₀(t)；θ₀(t) taking the lowest temperature of the cabinet surface on the front side of the cabinet body as the cabinet body environment temperature;a temperature threshold for an "over-temperature warning" state for a non-intervention point;a temperature threshold for a non-intrusive point "overheat alarm" condition;a temperature rise threshold value in an overheat early warning state of a non-intervention point;a temperature rise threshold value in an overheat alarm state of a non-intervention point;from said first diagnostic method, based on a non-invasive algorithmic modelAnd (3) reversely calculating to obtain:

further, the step S4 further includes: diagnosing the operating state of the switch cabinet by adopting a transverse comparison method and a longitudinal comparison method, wherein the transverse comparison method is to compare the operating state of the switch cabinet with the same type of equipment on the basis of the second diagnosis method so as to confirm the equipment with the highest temperature rise; the longitudinal comparison method is to compare the current monitoring data of the equipment with the historical data of the equipment so as to judge whether the equipment is abnormal or not.

The third purpose of the invention is realized by adopting the following technical scheme:

the device comprises a camera module and a processing module, wherein the camera module comprises an infrared camera and a visible light camera, the angles of the infrared camera and the visible light camera are the same, and a preset distance exists between the infrared camera and the visible light camera and the top of a switch cabinet so as to shoot real-time images or videos of the switch cabinet; the real-time image or video at least comprises a preset area of the switch cabinet; the preset distance is 0.5-1 m;

the processing module monitors the working state of the switch cabinet according to the real-time image or video of the camera module; the processing module comprises a processor, a memory and a computer program stored on the memory and capable of running on the processor, and the processor executes the computer program to realize the non-intrusive monitoring method for the switch cabinet.

The third purpose of the invention is realized by adopting the following technical scheme:

a storage medium having stored thereon a computer program which, when executed, implements a switchgear non-intrusive monitoring method as defined in any of the above.

Compared with the prior art, the invention has the beneficial effects that:

the invention provides a switch cabinet non-intervention monitoring method, a switch cabinet non-intervention monitoring device and a storage medium, which are used for comprehensively and intelligently analyzing visible light images and infrared images, capturing abnormity in the operation of equipment, realizing intelligent fault diagnosis and alarm of power equipment and greatly reducing the number of monitoring equipment and occupied space. The method has the advantages that operation and maintenance decision and fault analysis are assisted by workers, power supply reliability is improved, equipment perception is ubiquitous, application analysis is intelligent, and accordingly management service efficiency and intelligent level of the power equipment are improved.

Drawings

FIG. 1 is a schematic structural diagram of an embodiment of the present invention;

FIG. 2 is a schematic flow chart illustrating a method for monitoring and diagnosing a process module according to an embodiment of the present invention;

fig. 3 is a flowchart illustrating step S2 according to an embodiment of the present invention.

Detailed Description

The present invention will be further described with reference to the accompanying drawings and the detailed description, and it should be noted that any combination of the embodiments or technical features described below can be used to form a new embodiment without conflict.

As shown in fig. 1, the application provides a non-intrusive monitoring method for a switch cabinet, which is applied to a non-intrusive monitoring device for a non-switch cabinet, and solves the problems of large number of sensors, complex installation engineering and the like in the temperature monitoring of the traditional high-voltage switch cabinet.

Specifically, the non-intervention monitoring device of the switch cabinet comprises a camera shooting module and a processing module, the camera shooting comprises a first camera and a second camera, and the lenses of the first camera and the second camera are all in a preset distance with the top of the switch cabinet so as to shoot real-time images or videos of the switch cabinet. The real-time image or video at least comprises a preset area of the switch cabinet.

The positions of the first camera and the second camera and the preset distance existing at the top of the switch cabinet can not be too small, and the top of the switch cabinet needs to be ensured to have a sufficient shooting angle, so that a shot real-time image or video is clear and can cover a preset area. Therefore, in the present embodiment, the first camera is an infrared camera, and the second camera is a visible light camera. The preset distance between the infrared camera, the visible light camera and the top of the switch cabinet is more than 0.3 m. The preset distance is preferably 0.5-1 meter, and in the embodiment, is preferably 0.6 meter, so that the whole area of the top surface of the switch cabinet and the area of the instrument room above the front surface of the switch cabinet can be completely photographed, and the partition information at the top of the switch cabinet, the label information at the front surface of the switch cabinet and the information of the detection instrument can be completely extracted through real-time images or videos.

And the infrared camera shoots at the same angle with the lens of the visible light camera to obtain two images or videos of visible light and infrared light at the same position. In this embodiment, the straight line where the cameras of the infrared camera and the visible light camera are located and the drooping line form an included angle of 45 °. When the video shot by the camera module is taken, video screenshot with a fixed time interval is required to obtain a static real-time image. In this example, the time interval is taken to be 3 minutes. In practical application, a worker can adjust the time interval according to the load change speed of the switch cabinet, and the general time interval is 0.5-10 minutes.

The time of shooting, the shooting angle and the height of the infrared camera or the visible light camera are marked in the real-time image or the video. The infrared camera and the visible light camera can be fixedly arranged on a wall or the movement or the direction of the infrared camera and the visible light camera is controlled by a track, and when the infrared camera and the visible light camera are controlled by the track, the shooting positions of the infrared camera and the visible light camera during shooting need to be marked in an image.

The processing module is in communication connection with the camera module. The processing module is used for monitoring the working state of the switch cabinet according to the real-time image or video of the camera module. The processing module may be a remote computer or a cloud platform, or the processing module may be built in the camera module to facilitate rapid processing of the real-time image or video. And after receiving the real-time image or video, the processing module extracts required equipment state information, wherein the equipment state information comprises but is not limited to the mark number of the detected switch cabinet, the highest temperature data of a plurality of areas at the top, the temperature data of the switch cabinet body and the temperature sampling time point. And further calculating and processing the equipment state information to obtain the diagnosis conclusion of the operating state of the switch cabinet, thereby realizing monitoring and diagnosis of the operating state of the high-voltage switch cabinet.

The processing module comprises an online monitoring unit, an image processing unit, a calculation diagnosis unit and a communication unit. The online monitoring unit is connected with the camera module, is in two-way communication with the camera module, and is used for receiving and storing the real-time image or video and the mark information thereof sent by the camera module. The image processing unit is used for extracting the equipment state information in the real-time image and the video according to the real-time image or the video received by the online monitoring unit. The calculation diagnosis unit is used for calculating the temperature of the lumped heat source of the conductor in each interval region of the switch cabinet by using a non-intrusive algorithm according to the equipment state information of the image processing unit; managing the model parameters of the non-intervention algorithm of each switch cabinet, judging the model of each switch cabinet according to the label information of each switch cabinet, calling the matched non-intervention model parameters, and calculating and diagnosing the state of each switch cabinet.

As shown in fig. 2, the process schematic diagram of the method for monitoring and diagnosing by the processing module according to the real-time image of the camera module specifically includes the following steps:

and step S1, acquiring the infrared image and the visible light image shot by the camera module, and fusing the infrared image and the visible light image to form a dual-frequency fused image.

The infrared image and the visible light image have complementarity in content, the infrared image sensor performs imaging by acquiring infrared radiation of a target, temperature information of the target can be acquired through the infrared image, the infrared imaging is less influenced by ambient light, the imaging is generally dark, the resolution ratio is low, the edge is not sensitive, and background information is blurred. The user can see the portions of the device that are overheated through the infrared image, but cannot accurately identify the specific locations of overheating in the device. The visible light image is rich in spectral information, can retain more detail and texture information, but needs to have good illumination and no temperature information.

The fusion of the two can effectively improve the description capability of the image on the scene details and the hot target so as to obtain more accurate, reliable and comprehensive scene description, is convenient for human visual perception or further processing and analysis by a computer, and has higher use value on the functions of target detection, object identification and the like. The infrared image and the visible light image are fused, and the method aims to solve the problem that the detection of the infrared image on the equipment region target is inaccurate, so that the detail display in the infrared image is more definite.

Fusing the infrared image and the visible light image, specifically, acquiring the infrared image and the visible light image at the same visual angle; cutting the visible light image to ensure that the visual field of the visible light image is completely the same as that of the infrared image; performing smooth filtering and embossment processing on the visible light image to obtain a corresponding embossment image; and carrying out fusion processing on the relief image and the infrared image, extracting image contour information of the relief image, and transferring the image contour information to the infrared image to obtain a dual-frequency fusion image.

And step S2, recognizing the dual-frequency fusion image, the infrared image and the visible light image, and performing image processing to extract equipment state information. The image processing unit runs an image processing program to extract the required device state information. Specifically, as shown in fig. 3, the step S2 includes the following steps:

step S21, carrying out target detection of the outline of the power equipment on the dual-frequency fusion image, the infrared image and the visible light image, and respectively segmenting a target detection frame image; firstly, automatic target detection and region segmentation are carried out on the dual-frequency fusion image. According to the segmentation result of the dual-frequency fusion image, the infrared image and the visible light image are segmented in the same region, direct and accurate data support is provided for a subsequent diagnosis method, and independent image documents are established and subjected to filing management for the segmented regions for the next processing. For example, for a withdrawable circuit breaker switchgear, the top partition areas that can be divided are: circuit breaker room pressure release window, female chamber pressure release window, cable chamber pressure release window of arranging.

In the present application, the target detection comprises the following steps: firstly, a candidate area set is generated, the candidate area is obtained by finding out the possible positions of targets in the image in advance by utilizing the information of textures, edges, colors and the like in the image, all the candidate areas are used as training samples and input into a Convolutional Neural Network (CNN) for training, then the CNN characteristics extracted from each candidate area are input into a classifier SVM for training, finally the classified candidate areas of the classifier SVM are subjected to frame regression to correct the candidate areas, and the condition that the window extracted from the candidate areas is more consistent with a target real window is met.

And step S22, recognizing characters of the equipment label according to the target detection frame image of the visible light image, and naming the target detection frame image of the double-frequency fusion image, the infrared image and the visible light image according to the character recognition result. The switch cabinet sign characters of the target detection frame image in the visible light image are automatically recognized, the sign comprises a nameplate and a number plate, the model and specification information of the switch cabinet is recognized and extracted through the nameplate, and the name and number information of the switch cabinet are recognized and extracted through the number plate. In this embodiment, the feature extraction and template matching method based on the OPENCV software development environment is used for character recognition.

Step S23, performing equipment region segmentation according to the double-frequency fusion image and the target detection frame image of the infrared image to obtain an infrared equipment region segmentation image; and extracting temperature distribution information of the infrared equipment region segmentation image through an infrared image temperature algorithm. And calculating and outputting the highest temperature of the infrared segmentation maps of the interval areas according to the existing temperature extraction algorithm, wherein the highest temperature is used as the characteristic temperature of the interval areas. In addition, a lowest temperature is calculated in the original infrared image and is used as the cabinet body environment temperature of the switch cabinet. In the application, an infrared pixel polynomial fitting calibration algorithm is adopted to carry out the method for extracting the temperature.

And step S3, calculating the temperature of the lumped heat source of the switch cabinet according to the equipment state information and the temperature distribution information and a non-intrusive temperature measurement model.

The surface of the cabinet body of the high-voltage switch cabinet has certain correlation with the temperature of a conductor in the cabinet. Based on the principle of heat transfer science, the heat transfer process from the inner conductor of the cabinet to the environment outside the cabinet is researched and analyzed, and a temperature relation model between the surface of the cabinet body and the inner conductor of the cabinet is established. According to the first law of thermodynamics, the sum of the net heat flow introduced into the object and the heat value of the heat source in the object is equal to the increase of the internal energy of the object, so that the heat balance is obtained: and (4) leading in heat and internal heat source heat productivity which are the increment of the internal energy, and further deducing a non-intervention temperature measurement model.

The temperature rise on the surface of the cabinet body of the switch cabinet can be influenced by physical quantities such as the heating power of a load loop conductor, the indirect electric shock resistance heating power of the conductor, the environmental temperature of the cabinet body, the heat dissipation structure and materials of the cabinet body, the heat balance time and the like. The possible temperature of the conductor in the switch cabinet under a certain condition can be calculated from the temperature distribution data on the surface of the switch cabinet through a non-intervention temperature measurement model, so that the state of equipment in the switch cabinet can be diagnosed.

In the non-intervention temperature measurement model, the dynamic process of the thermal system of the switch cabinet is described, so that the temperature state of the thermal system of the switch cabinet in the change process can be accurately reflected and calculated. However, the model of the dynamic process has more model parameters, requires more monitoring data, and has higher requirements on the model parameters and the monitoring data than the model of the static process. Therefore, the non-invasive thermometry model includes a dynamic non-invasive thermometry model and a static non-invasive thermometry model. The application is also provided with:

and S31, selecting a dynamic non-intrusive temperature measurement model or a static non-intrusive temperature measurement model according to the equipment state information and the temperature distribution information of the switch cabinet, and calculating the temperature of the switch cabinet. In actual use, a dynamic algorithm or a static algorithm is adopted, analysis and selection are carried out according to the use conditions of engineering equipment, and a dynamic non-intervention temperature measurement model is adopted for scenes which are frequently in load change or environment temperature change; for scenes with less load change and slow environmental temperature change, a static non-intrusive temperature measurement model is adopted.

The dynamic non-intervention temperature measurement model comprises the following steps:

wherein, theta₁(t) is the temperature of a lumped heat source (hereinafter referred to as "temperature of the lumped heat source") in a certain interval inside the high-voltage switch cabinet; theta₂(t) is the highest temperature of each area at the top of the switch cabinet; theta₀(t) taking the lowest temperature of a cabinet surface on the front side of the cabinet body as the ambient temperature of the cabinet body of the switch cabinet; a. the₁₂The heat transfer thermal resistance coefficient between the lumped heat source of the switch cabinet and the cabinet top surface is constant; a. the₂₃The coefficient of heat transfer thermal resistance between the highest temperature of each area at the top of the switch cabinet and the ambient temperature of the switch cabinet is constant; b is₁₂The thermal capacitance coefficient is the sum of all substances in the highest temperature isothermal surface of each area of the top of the switch cabinet and is a constant; t is the double-view image sampling time, then d θ₂(t)/dt represents the derivative of temperature with respect to time. The non-intrusive temperature measurement model calculates the temperature of the lumped heat source according to the highest temperature data of each area at the top of the switch cabinet, the environmental temperature data of the cabinet body and the sampling time point of each temperature, and reveals the temperature distribution of the space (inside and outside) of the switch cabinet as a whole and the rule of the temperature distribution along with the change of the space.

The static working environment means that the physical quantity does not change with time, is in a static state, and is a relatively ideal state. The simplified form of the static non-invasive thermometry model is:

this formula can be expressed in simplified form as:

θ₁(t)＝θ₂(t)+kΔθ₂(t)

wherein the content of the first and second substances,the model coefficients are static model coefficients of a non-intrusive temperature measurement algorithm, and are simply referred to as model coefficients.

The model coefficient k is a physical characteristic coefficient of the switch cabinet equipment, and in general engineering application, the model coefficient k is roughly taken as a constant coefficient, and fitting and verification are performed through an experimental method. In some special applications, the model coefficient k can also be used as a variable coefficient, and the functional relationship between the variable coefficient and the material temperature and the cabinet body environment temperature can be fitted through experiments. Different interval areas of the switch cabinet have different model coefficients k; the same interval area of the switch cabinets with the same specification has similar model coefficients k. And the calculation and diagnosis unit of the processing module is used for matching and selecting the corresponding model coefficient k according to the specification and the interval area of the switch cabinet.

In practical applications, a plurality of heat sources with different central hot spots exist in the switch cabinet, and the heat sources have different mutual influences, such as: the A-phase busbar and the A-phase breaker contact are connected through the metal copper conductor, and have larger heat conduction correlation, so that the same-phase metal conductor can be taken as a heat source; and the A-phase breaker contact and the B-phase breaker contact are heat sources of two different hot spots, and are close to each other in distance and insulated and isolated from each other. For a high-voltage switch cabinet, any one interval area can be roughly divided into three different physical heat sources according to A, B, C three-phase metal conductors, and because the three-phase metal conductors have similar volume and heat capacity among phases, in the process of diagnosis, the three-phase heat sources are regarded as the same heat capacity, so that the approximation is as follows: the lumped heat source temperature is equivalent to an average of the maximum temperatures of the three-phase metal conductors. The non-intrusive temperature measurement model cannot judge the condition of three-phase heating imbalance. When the three-phase heating value in the switch cabinet is seriously unbalanced, the situation that the temperature of the lumped heat source calculated according to a standard model is lower than the actual highest temperature can occur.

For universal applicability and reliability of engineering applications, certain assumptions are usually made about the three-phase heating balance state of the switchgear, for example, preferably, 10% of the three-phase heating imbalance probability of the switchgear under normal conditions is large, and a non-intrusive model parameter, i.e., a model coefficient k, is appropriately adjusted high so that the model coefficient is suitable for most switchgear devices of the same model.

And step S4, comparing the temperature of the lumped heat source of the switch cabinet calculated by the non-intrusive temperature measurement model with a preset fault diagnosis threshold value, and diagnosing the operation state of the switch cabinet.

According to the non-intervention temperature measurement model, the temperature of a lumped heat source of the metal conductor in the cabinet body can be calculated according to the temperature of the non-intervention point, or the temperature of the non-intervention point can be calculated according to the temperature of the lumped heat source of the metal conductor in the cabinet body, and the diagnosis threshold value of the result standard is judged according to the calculated temperature or the measured temperature, so that the diagnosis of the operation temperature rise state of the equipment can be made.

And comparing the calculated temperature of the lumped heat source with a preset fault diagnosis threshold value, wherein the specific two diagnosis methods are shown in the following table.

The first diagnostic method is as follows:

TABLE 1

θ₁(t) is the temperature of the lumped heat source in a certain interval inside the high-voltage switch cabinet;

Δθ₁(t) is θ₁(t)-θ₀(t)；θ₀(t) taking the lowest temperature of the cabinet surface on the front side of the cabinet body as the cabinet body environment temperature;

a temperature threshold for an "over-temperature warning" condition for the lumped heat source;

a temperature threshold for a lumped heat source "over-temperature alarm" condition;

a temperature rise threshold for an integrated heat source 'over-temperature warning' state;

is the temperature rise threshold value of the lumped heat source overheat alarm state.

The threshold can be set according to the related temperature and temperature rise limit standards in the technical requirements of GBT 11022 and 2011 standards for common use of high-voltage switching equipment and control equipment, and different threshold settings are made according to different types of cabinets and different metal conductor materials. For example, for a circuit breaker contact of bare copper material, the diagnostic thresholds set are:

the second diagnostic method is as follows: and calculating the temperature value of the surface of the cabinet body corresponding to the lumped heat source temperature diagnosis threshold value according to the non-intrusive temperature measurement model, and taking the temperature value as the diagnosis threshold value of the actually measured temperature of the surface of the cabinet body, thereby diagnosing the operation state of the switch cabinet.

TABLE 2

θ₂(t) is the highest temperature of each area at the top of the switch cabinet;

Δθ₂(t) is θ₂(t)-θ₀(t)；θ₀(t) taking the lowest temperature of the cabinet surface on the front side of the cabinet body as the cabinet body environment temperature; the actual measurement value given by the image processing program should be taken in the application, and if the cabinet body environment temperature is lacked, the cabinet body environment temperature theta can be optimized₀＝40℃。

A temperature threshold for an "over-temperature warning" state for a non-intervention point;

a temperature threshold for a non-intrusive point "overheat alarm" condition;

a temperature rise threshold value in an overheat early warning state of a non-intervention point;

the temperature rise threshold value is the temperature rise threshold value of the non-intervention point overheat alarm state.

Wherein the content of the first and second substances,based on non-invasive algorithmic models, from diagnostic method oneReverse calculation is carried out; the specific process is as follows, in the static non-intervention temperature measurement model:

wherein:is a known set value, A₂₃，A₁₂Is a constant number, theta₀In order to monitor the value of the measurement,then the corresponding can be found

Both of the above-mentioned diagnostic methods, which are based on non-invasive models of heat transfer science, monitor and diagnose the operating state of the installation only on the basis of the temperature or temperature rise data of the installation, have the limitation that the diagnosis can only be made if the switchgear installation has a sufficiently high operating temperature, i.e. is overheated or tends to be overheated.

Therefore, in the present embodiment, the lateral comparison method and the longitudinal comparison method are employed at the same time to increase the reliability of the fault diagnosis. The transverse comparison method is characterized in that on the basis of the method, transverse comparison diagnosis is performed on the same type of equipment in a power distribution room or a transformer substation, so that the equipment with the highest temperature rise can be found conveniently, and the important monitoring range of the equipment is reduced. The longitudinal comparison method is a comparison diagnosis of the current monitoring data of the equipment and the historical data of the equipment, and is convenient for assisting in judging whether the equipment is abnormal recently. In order to increase the data dimension of fault diagnosis and widen the diagnosable range, on the basis of the method, the monitoring of the load data of the equipment can be increased, the load data of the equipment can participate in calculation and diagnosis, and the diagnosis of the health and the fault hidden danger state of the equipment can be realized without the actual large load or high temperature of the equipment.

And S5, classifying the diagnosis result of the switch cabinet, and providing matched alarm information according to the classification result. And providing corresponding alarm information according to different classification results, such as various modes of alarm sound, alarm light, alarm short message and the like, so as to remind workers.

The invention discloses a switch cabinet non-intervention monitoring method and a switch cabinet non-intervention monitoring device, which are used for comprehensively and intelligently analyzing visible light images and infrared images, capturing abnormity in equipment operation, realizing intelligent fault diagnosis and alarm of power equipment and greatly reducing the number of monitoring equipment and occupied space. The method has the advantages that operation and maintenance decision and fault analysis are assisted by workers, power supply reliability is improved, equipment perception is ubiquitous, application analysis is intelligent, and accordingly management service efficiency and intelligent level of the power equipment are improved.

In addition, the invention also provides a storage medium, wherein the storage medium stores a computer program, and the computer program realizes the steps of the switch cabinet non-intrusive monitoring method when being executed by a processor.

The invention is operational with numerous general purpose or special purpose computing system environments or configurations. For example: personal computers, server computers, hand-held or portable devices, tablet-type devices, multiprocessor systems, microprocessor-based systems, set top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments that include any of the above systems or devices, and the like.

The above embodiments are only preferred embodiments of the present invention, and the protection scope of the present invention is not limited thereby, and any insubstantial changes and substitutions made by those skilled in the art based on the present invention are within the protection scope of the present invention.