KR102610309B1 - System for performing thermal imaging monitoring and diagnosis based on artificial intelligence - Google Patents

Abstract

The present invention relates to an artificial intelligence-based thermal imaging surveillance and diagnosis system, which includes at least one thermal imaging camera installed inside a switchboard enclosure to obtain at least one thermal image by photographing surveillance objects within the switchboard; And, by analyzing the thermal image using artificial intelligence, the temperature of each monitored object is detected, and the temperature change pattern is tracked and monitored, and the status of each monitored object is diagnosed through at least one of the temperature and the temperature change pattern, and reporting information is provided. and an analysis device that generates and provides, wherein the unit for providing the reporting information is the switchboard, a facility constituting the switchboard, and at least one piece of equipment provided in the facility.

Description

Artificial intelligence-based thermal imaging monitoring and diagnosis system {System for performing thermal imaging monitoring and diagnosis based on artificial intelligence}

The present invention relates to an artificial intelligence-based thermal image monitoring and diagnosis system that allows simpler and more effective monitoring and diagnosis of the status of equipment in a switchgear.

In general, a switchboard is a device that distributes and provides power supplied from a power supplier to power consumers.

The switchgear may include substation means for converting extra-high voltage power to high voltage, or high voltage power to low voltage, main breaker and branch breaker tripped by overcurrent or overvoltage, and a surge protection device to protect devices from surges. etc. are provided. Typically, each monitoring object within the switchboard is connected by a bus bar.

Conventionally, a method using a temperature sensor has been proposed as a method of diagnosing deterioration or aging of monitoring objects in a switchgear.

However, since the monitoring objects that make up the switchgear all have different deterioration standard temperatures, there is a problem of installation cost in having to install a temperature sensor in each monitoring target area, and it is necessary to manage whether multiple temperature sensors operate normally and determine the temperature damaged by deterioration. Maintenance inconveniences such as having to replace sensors occur.

Domestic registered patent No. 10-1529818 (registration date: 2015.06.11.)

Accordingly, in order to solve the above problems, the present invention is based on artificial intelligence, which acquires thermal images and analyzes them using artificial intelligence to enable simpler and more effective monitoring of the status of the main monitoring object in the switchgear. It is about a thermal imaging surveillance and diagnosis system.

In addition, it is about an artificial intelligence-based thermal imaging surveillance and diagnosis system that allows status monitoring and diagnosis to be performed by subdividing the unit into parts within the facility rather than the switchgear unit.

The purpose of the present invention is not limited to the purposes mentioned above, and other purposes not mentioned will be clearly understood by those skilled in the art to which the present invention pertains from the description below.

As a means for solving the above problem, according to an embodiment of the present invention, at least one thermal imaging camera is installed inside the switchgear enclosure and acquires at least one thermal image by photographing surveillance objects within the switchboard; And, by analyzing the thermal image using artificial intelligence, the temperature of each monitored object is detected, and the temperature change pattern is tracked and monitored, and the status of each monitored object is diagnosed through at least one of the temperature and the temperature change pattern, and reporting information is provided. An artificial intelligence-based thermal image monitoring and diagnosis system comprising an analysis device that generates and provides a unit of reporting information, wherein the unit for providing the reporting information is the switchboard, a facility constituting the switchboard, and at least one equipment provided in the facility. provides.

The analysis device includes a surveillance target identification model that automatically identifies the location and type of each surveillance target included in the thermal image by pre-learning the correlation between the thermal image and the location and type of the surveillance target; An image extraction unit that obtains an image of interest by extracting the area where the surveillance object was filmed based on the location of the surveillance object; A temperature detection model that automatically detects the temperature corresponding to the image of interest of the surveillance target by pre-learning the correlation between the heat distribution of the image of interest and temperature; After tracking and monitoring the temperature change pattern of each monitoring object and diagnosing the status of each monitoring object, reporting information including information on at least one of the current temperature, temperature change pattern, thermal image, and status diagnosis result of each monitoring object. A status reporting unit that generates and provides; and a communication unit supporting communication with the administrator terminal.

The status reporting unit provides the current temperature in the form of at least one of a temperature value and a temperature gauge, and provides a temperature change pattern in the form of at least one of the maximum temperature difference by period, maximum temperature, average temperature, minimum temperature, and temperature change trend. It is characterized by

The status reporting unit further includes a function of generating a bounding box indicating the detection position of the monitoring object and text information guiding the type and temperature of the monitoring object, and then overlaying and displaying the text information on the thermal image.

The status reporting unit may further include a function that stores information related to each monitoring object in a database and then searches and browses the information according to a timeline.

The present invention acquires a thermal image and analyzes it using artificial intelligence to confirm the status of the main monitoring target in the switchboard, thereby ensuring minimization of system size and implementation cost.

In addition, by subdividing the unit of reporting information, the resolution of status monitoring and diagnosis operations can be maximized and the scope of application can be expanded.

1 and 2 are diagrams showing a switchgear equipped with an artificial intelligence-based thermal image monitoring and diagnosis function according to an embodiment of the present invention.
Figure 3 is a diagram illustrating a monitoring target within a switchgear according to an embodiment of the present invention.
Figure 4 is a diagram for explaining a method for diagnosing abnormalities in a monitoring target according to an embodiment of the present invention.
5 to 9 are diagrams for explaining a method of providing reporting information according to an embodiment of the present invention.

Hereinafter, with reference to the attached drawings, preferred embodiments will be described in detail so that those skilled in the art can easily practice the present invention. However, when describing preferred embodiments of the present invention in detail, if it is determined that a detailed description of a related known function or configuration may unnecessarily obscure the gist of the present invention, the detailed description will be omitted. In addition, the same symbols are used throughout the drawings for parts that perform similar functions and actions.

In addition, throughout the specification, when a part is said to be 'connected' to another part, this does not only mean 'directly connected', but also 'indirectly connected' with another element in between. Includes. In addition, 'including' a certain component means that other components may be further included rather than excluding other components, unless specifically stated to the contrary.

1 and 2 are diagrams showing a switchgear equipped with an artificial intelligence-based thermal image monitoring and diagnosis function according to an embodiment of the present invention.

As shown in FIG. 1, the switchgear 100 of the present invention is installed inside the switchgear enclosure 110 and the switchgear enclosure 110 in which a plurality of facilities are built, and captures at least one column by photographing surveillance objects within the switchgear. At least one thermal imaging camera 120 that acquires a video image, analyzes each thermal imaging image using artificial intelligence to identify the temperature of each monitoring object, and checks the status of each monitoring object based on the temperature, and then uses the administrator terminal (200) ), and an analysis device 130 that notifies.

As shown in FIG. 2, the analysis device 130 of the present invention pre-learns the location and type of the surveillance target corresponding to the thermal image image, and automatically identifies the location and type of each surveillance target included in the thermal image image. A monitoring object identification model 131, an image extraction unit 132 that extracts an image of interest by extracting the area where the monitoring object is filmed based on the location of the monitoring object, and a correlation between temperatures corresponding to the heat distribution diagram is learned in advance. , a temperature detection model 133 that detects the temperature of each monitoring object corresponding to the heat distribution of the image of interest, tracks and monitors the temperature change pattern of each monitoring object, diagnoses the status of each monitoring object, and then diagnoses the current temperature of each monitoring object. , a status reporting unit 134 that generates and provides reporting information including information about at least one of a temperature change pattern, a thermal image, and a status diagnosis result, and a communication unit 135 that supports communication with the manager terminal 200. ), etc.

It is most desirable that the monitoring target identification model 131 of the present invention is built based on the learning model of YOLOv5, and the temperature detection model 133 is built based on the learning model of Faster R-CNN, but there is no need to be limited to this. does not exist.

In addition, the status reporting unit 135 can provide reporting information in the form of a file or web page to the manager terminal 200, but if necessary, a program that supports linkage with the analysis device 130 is provided to the manager terminal 200. Install and run it, and provide reporting information through the program.

In addition, the analysis device 130 of the present invention may be implemented as a separate hardware device and embedded in the switchboard, but if necessary, it may be implemented as an embedded device or a software program and installed in combination with the switchgear management server. make it possible

Figure 3 is a diagram illustrating a monitoring target within a switchgear according to an embodiment of the present invention.

As shown in Figure 3, the switchgear may be equipped with facilities such as TR class, LBS&LA class, MOF&PT class, Main VCB class, individual VCB class, VCB&CT class, ACB class, and MCCB class. In the present invention, the locations of the facilities are Considering this, multiple thermal imaging cameras are distributed and distributed within the switchboard.

Then, after subdividing the monitoring target into equipment units rather than switchboard units, each of these temperature standard values is set based on the temperature pattern method of the electrical equipment safety management standards. That is, in the present invention, by subdividing the equipment into equipment units rather than switchboard units, the abnormal state of the switchboard can be monitored, diagnosed, and monitored in more detail.

In addition, by setting the priority of each monitoring target, abnormal conditions of the most important monitoring target can be checked with the highest priority and guided to the monitoring target manager.

In this way, the present invention enables motoring by subdividing the status of the switchgear into equipment units rather than switchboards or facilities.

However, in this case, the number of monitoring objects increases exponentially, and it is realistically impossible to detect the status of all of these monitoring objects using a temperature sensor. In other words, the existing temperature sensor must detect the temperature of the monitored object in a 1:1 ratio, which causes the problem that the number of temperature sensors must be increased by the number of monitored objects.

Accordingly, the present invention proposes a new surveillance object identification model 131 and a temperature detection model 133 implemented in an artificial intelligence method, and through these, the type and number of surveillance objects can be easily expanded through image processing.

Figure 4 is a diagram for explaining a method for identifying a monitoring object and detecting temperature according to an embodiment of the present invention.

First, at least one thermal image obtained through at least one thermal image camera 120 installed inside the switchgear is acquired (S1).

Then, after sequentially selecting one of at least one thermal image image (S2), it is analyzed through the surveillance target identification model 131 to determine the location and type of at least one surveillance target present in the thermal image image (S2). S3).

Then, based on the location of the surveillance object, at least one image of interest is obtained by extracting at least one surveillance target shooting area from one thermal image (S4).

Then, after sequentially selecting one of the images of interest (S5), it is analyzed through the temperature detection model 133 to determine the temperature of the monitored object (S6).

Steps S5 to S7 are repeatedly performed until all images of interest extracted from the currently selected thermal image are selected (S7).

If all images of interest related to the currently selected thermal image have been selected (i.e., analysis of the currently selected thermal image has been completed), step S2 is re-entered and the analysis of the new thermal image is repeated. In other words, the analysis operation for all thermal image images can be performed sequentially (S8).

5 to 9 are diagrams for explaining a method of providing reporting information according to an embodiment of the present invention.

First, the present invention provides a system setting page as shown in FIG. 5, through which it is possible to register and manage the building where the switchboard is installed, the equipment structure of the switchboard, the equipment taken by the camera, etc.

At this time, building information includes information about the building management period, building name, location information (address or latitude/longitude), etc. In particular, building location information is registered by clicking on the building location on the map in conjunction with the map service. This is to maximize user convenience.

Switchboard information includes information about the building where the switchboard is installed, a list of facilities included in the switchboard, and a list of equipment belonging to the facility, and camera shooting information may include information about the correspondence between the bounding box and equipment in the camera image.

And as shown in Figure 6, the reporting unit of the switchboard is subdivided into building, facility, and equipment units to determine and select the monitoring target.

It also provides guidance on at least one of the current temperature, temperature change pattern, thermal image, and status diagnosis results of the monitored object.

At this time, the current temperature is provided in the form of at least one of a temperature value and a temperature gauge, and the temperature change pattern is provided in the form of at least one of the maximum temperature difference by period, maximum temperature, average temperature, minimum temperature, and temperature change trend (graph or table). can be provided.

Additionally, a bounding box indicating the location of the monitoring object in the thermal image and text information guiding the type and temperature of the monitoring object are additionally created, and these information are overlaid and displayed on the thermal image as shown in FIG. 7. It allows managers to more intuitively recognize the fever status of the monitoring target.

Additionally, when you click on a specific surveillance target through the equipment list, only the bounding box corresponding to the surveillance target is displayed or the bounding box is highlighted, enabling more accurate and faster location confirmation on the thermal image.

The status diagnosis result informs whether there is an abnormality in the monitoring object as shown in FIG. 8. In the present invention, the monitoring object is monitored based on not only the temperature of the monitoring object, but also the temperature change pattern that is the result of tracking and monitoring the temperature of the monitoring object over a long period of time. Allows the condition to be diagnosed. For example, the occurrence of an abnormal condition can be confirmed and notified not only when the allowable temperature of the monitored object increases beyond the preset tolerance, but also when the temperature of the monitored object continues to increase or decrease or the temperature fluctuation value becomes abnormally large. Let it happen.

That is, in the present invention, after accumulating data over a long operation period, it is possible to diagnose the status of the monitored object by considering data trends.

In addition, monitoring objects with confirmed abnormalities are displayed at the top of the equipment list, allowing for faster status confirmation of the monitoring objects.

In addition, all information acquired through system operation (i.e., temperature, thermal image, status diagnosis results, etc.) is stored in a database so that this information can be searched and viewed according to the future timeline as shown in Figure 9. .

In the above, preferred embodiments of the present invention have been shown and described, but the present invention is not limited to the specific embodiments described above, and may be used in the technical field to which the invention pertains without departing from the gist of the present invention as claimed in the claims. Of course, various modifications can be made by those skilled in the art, and these modifications should not be understood individually from the technical idea or perspective of the present invention.

Claims (5)

At least one thermal imaging camera installed inside the switchgear enclosure to acquire at least one thermal image by photographing surveillance objects within the switchboard; and
By analyzing the thermal image image using artificial intelligence, the temperature of each monitored object is detected and the temperature change pattern is tracked and monitored. The status of each monitored object is diagnosed and reporting information is provided through at least one of the temperature and temperature change pattern. Contains analysis devices that generate and provide,
The unit of provision of the above reporting information is
The switchboard, a facility constituting the switchboard, and an equipment unit based on the facility,
The analysis device is
Implemented based on the YOLOv5 learning model, it pre-learns the correlation between the location and type of the thermal image and the surveillance target, and automatically identifies the location and type of the surveillance target included in the thermal image using artificial intelligence. Model;
An image extraction unit that obtains an image of interest by extracting the area where the surveillance object was filmed based on the location of the surveillance object;
A temperature detection model implemented based on the Faster R-CNN learning model, which pre-learns the correlation between the heat distribution and temperature of the image of interest and detects the temperature corresponding to the image of interest of the surveillance target using artificial intelligence;
After identifying the location and type of the monitoring object through the above monitoring object identification model, the temperature change pattern of each monitoring object is classified by location and type, tracked and monitored to diagnose the status of each monitoring object, and then the current temperature of each monitoring object is monitored. , a status reporting unit that generates and provides reporting information including information on at least one of a temperature change pattern, a thermal image, and a status diagnosis result; and
An artificial intelligence-based thermal imaging monitoring and diagnosis system comprising a communication unit that supports communication with an administrator terminal. delete The method of claim 1, wherein the status reporting unit
An artificial device characterized in that the current temperature is provided in the form of at least one of a temperature value and a temperature gauge, and the temperature change pattern is provided in the form of at least one of the maximum temperature difference, maximum temperature, average temperature, minimum temperature, and temperature change trend by period. Intelligence-based thermal imaging surveillance and diagnosis system. The method of claim 1, wherein the status reporting unit
Artificial intelligence-based thermal imaging, characterized in that it further includes a function of generating a bounding box that displays the detection position of the monitoring target and text information guiding the type and temperature of the monitoring target, and then overlaying and displaying it on the thermal image. Surveillance diagnostic system. The method of claim 1, wherein the status reporting unit
An artificial intelligence-based thermal imaging surveillance and diagnosis system that further includes a function that allows information related to each surveillance target to be stored in a database and then searched and viewed according to a timeline.