KR102192717B1 - Device for diagnosing core using cherenkov radiation image and method thereof - Google Patents

Abstract

It relates to a core diagnosis apparatus and a method thereof using a Cherenkov radiation image, wherein the core diagnosis apparatus includes a collection unit that collects in real time a photographed image of a nuclear reactor from a camera mounted in an area near the nuclear reactor, and a Cherenkov copy included in the photographed image. It includes a digital signal conversion unit for converting the brightness and color difference of light into a digital signal, and a core diagnosis unit for diagnosing whether the core output of the nuclear reactor and the internal state of the core are normal according to the digital signal.

Description

Core diagnosis apparatus and method using Cherenkov copy image {DEVICE FOR DIAGNOSING CORE USING CHERENKOV RADIATION IMAGE AND METHOD THEREOF}

A core diagnosis apparatus and method using Cherenkov radiation images are provided.

In order to maximize the reliability of nuclear reactor protection systems such as HANARO, design requirements such as multiplicity, diversity, and independence are applied. In the control system, design requirements such as multiplicity and independence are applied to the implementation of some important functions in order to maximize the reliability of the output control. Accordingly, various reactor power measurement devices enable safe protection and power control of the reactor.

The output of the Hanaro reactor is currently being checked by heat output, neutron output, and gamma output. In the case of heat output, the response speed is very slow, and in the case of neutron output and gamma output, the response speed is very fast. In the case of a neutron output or gamma output measuring instrument, the measuring instrument itself can detect an instantaneous output change, but there are cases in which very short changes are not recorded in the process of processing the measurement signal or the output signal. Therefore, it is useful to have a measuring device with a moderate response speed.

In addition, it is necessary to secure a system that can measure abnormalities such as flow pipes and nuclear fuel inside the core, or a system that can detect changes in nuclear fuel after use.

On the other hand, recently, machine learning through neural networks has been applied to various fields with the advantage of improving the diversity and accuracy of pattern recognition through big data, and research is being conducted. In the nuclear reactor-related field, a technology for monitoring the core using the neural network is being studied. As a related prior document, Korean Patent Registration 1,651,893 discloses a method for synthesizing the axial output distribution of a nuclear reactor core using a neural network circuit and a core monitoring system to which the method is applied.

Korean Patent 1,651,893

An embodiment of the present invention is for diagnosing a normal state or an abnormal state of a core of a nuclear reactor using Cherenkov radiation.

One embodiment of the present invention is to more accurately diagnose a normal state or an abnormal state of a nuclear reactor core through deep learning.

In addition to the above problems, the embodiments according to the present invention may be used to achieve other tasks not specifically mentioned.

The core diagnosis apparatus according to an embodiment of the present invention includes a collection unit that collects in real time a photographed image of a nuclear reactor from a camera mounted in an area near the nuclear reactor, and converts the brightness and color difference of the Cherenkov radiation included in the photographed image into a digital signal. And a digital signal conversion unit for converting, and a core diagnosis unit for diagnosing whether a core output of the nuclear reactor according to the digital signal and an internal state of the core are normal.

The method for diagnosing a core diagnosis apparatus according to an embodiment of the present invention includes photographing Cherenkov radiation emitted from a nuclear reactor through an interlocked camera, converting Cherenkov radiation from the captured image into an RGB histogram, RGB Converting the histogram to a digital signal, analyzing whether the core output of the reactor according to the digital signal and the internal state of the core are normal, and diagnosing the core based on the analyzed output of the reactor and whether the inside of the core is normal. Include.

According to an embodiment of the present invention, by observing a nuclear reactor using Cherenkov radiation observed according to the output of the nuclear reactor, it is possible to observe an instantaneous output change due to the afterimage effect of light.

According to an embodiment of the present invention, by observing the output of a reactor through an image using a general digital camera, it is possible to monitor safe operation of a reactor at low cost and secure convenience of maintenance according to the analysis of changes in the core.

According to an embodiment of the present invention, abnormal conditions can be quickly and accurately detected through deep learning not only of the core of a nuclear reactor, but also of nuclear fuel rods, flow pipes, reactor parts, and spent nuclear fuel of a nuclear reactor in which the Cherenkov radiation phenomenon is observed.

1 is an exemplary view showing a core diagnosis system according to an embodiment of the present invention.
2 is a block diagram illustrating a core diagnosis apparatus according to an embodiment of the present invention.
3 is a flowchart illustrating a method of diagnosing a core diagnosis apparatus according to an exemplary embodiment of the present invention.
4 is an exemplary diagram showing an RGB histogram according to an embodiment of the present invention.

Embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those of ordinary skill in the art may easily implement the present invention. The present invention may be implemented in various different forms, and is not limited to the embodiments described herein. In the drawings, parts irrelevant to the description are omitted in order to clearly describe the present invention, and the same reference numerals are used for the same or similar components throughout the specification. Also, in the case of well-known technologies, detailed descriptions thereof will be omitted.

Throughout the specification, when a part "includes" a certain component, it means that other components may be further included rather than excluding other components unless otherwise stated.

1 is an exemplary view showing a core diagnosis system according to an embodiment of the present invention.

As shown in FIG. 1, the reactor 100 is a structure generated during nuclear fission through a plurality of nuclear fuel rod bundles, and is composed of a moderator, a control rod, a coolant, and various other structures that control the nuclear reactor.

The reactor 100 represents an open top reactor such as a HANARO reactor.

On the other hand, if the reactor is of a closed type, a camera must be installed inside the reactor generated in the reactor. In this case, it is difficult to shoot with a general digital camera due to the high temperature of the reactor, and a camera composed of special equipment is required.

Next, the core diagnosis apparatus 200 analyzes an image captured in real time through the interlocked camera 300 and checks whether the state of the core is normal or abnormal.

The core diagnosis apparatus 200 collects images of Cherenkov radiation observed when the reactor 100 is started.

Here, the Cherenkov radiation is light generated when charged particles pass through an optically transparent medium and the speed of the particles is greater than the speed of light in the medium. Cherenkov radiation has blue light, and the color difference and brightness of Cherenkov radiation appear differently depending on the output of the reactor.

Accordingly, the core diagnosis apparatus 200 may measure the output of the nuclear reactor according to the image of the Cherenkov radiation, and may diagnose one or more of the state of the core, the flow pipe inside the core, or the state of the spent nuclear fuel.

Accordingly, the core diagnosis apparatus 200 may continuously monitor one or more of the state of the inner core of the reactor, the state of the flow pipe of the reactor, the state of the components of the reactor, or the nuclear fuel after use.

Next, the camera 300 represents a general digital camera of 10 million pixels or more, and is installed in an area near the reactor. In more detail, the camera 300 may be installed on an upper open-type reactor or in a space where Cherenkov radiation may be generated in addition to a region near the reactor.

In addition, the camera 300 has a unique ID and may store an image captured in real time or an image captured in a preset time unit in a separate database (not shown).

In addition, the camera 300 may set and store a location, a function, a pixel, a setting condition, a shooting condition, and the like of the camera 300 together with a unique ID.

Although one camera 300 is illustrated in FIG. 1, a plurality of cameras may be installed to photograph the same nuclear reactor.

Meanwhile, the core diagnosis apparatus 200 may be a server, a terminal, or a combination thereof.

The terminal is collectively referred to as a device having computational processing capability by having a memory and a processor, respectively. For example, there are a personal computer, a handheld computer, a personal digital assistant (PDA), a mobile phone, a smart device, a tablet, and the like.

The server is a memory in which a plurality of modules are stored, and a processor that is connected to the memory and reacts to the plurality of modules and processes service information provided to the terminal or action information that controls the service information, communication It may include means, and a UI (user interface) display means.

Memory is a device that stores information, and is non-volatile such as high-speed random access memory, magnetic disk storage, flash memory device, and other non-volatile solid-state memory devices. It may include various types of memories such as volatile memory.

The communication means transmits and receives service information or action information to and from the terminal in real time.

The UI display means outputs system service information or action information in real time. The UI display means may be an independent device that directly or indirectly outputs or displays the UI, or may be a part of the device.

Hereinafter, a core diagnosis apparatus for diagnosing whether a core state is normal by analyzing a captured image will be described using FIG. 2.

2 is a block diagram illustrating a core diagnosis apparatus according to an embodiment of the present invention.

As shown in FIG. 2, the core diagnosis apparatus 200 includes a collection unit 210, a digital signal conversion unit 220, and a core diagnosis unit 230. It may further include a learning unit 240.

First, the collection unit 210 collects the captured image in real time through communication with the camera 300 mounted in the area near the nuclear reactor. When a plurality of cameras are installed in the reactor, the collection unit 210 may collect images from each camera.

Next, the digital signal conversion unit 220 converts the brightness and color difference of the Cherenkov radiation included in the captured image into a digital signal. The digital signal conversion unit 220 may convert the captured image into an RGB histogram and then convert the RGB histogram into a digital signal.

Next, the core diagnosis unit 230 diagnoses whether the core output of the nuclear reactor and the internal state of the core are normal based on the digital signal.

At this time, the core diagnosis unit 230 may diagnose an abnormal state of the core by comparing the image with the normal state, and may diagnose the state of the used nuclear fuel.

For example, the core diagnosis unit 230 stores the image of the Cherenkov radiation light of the core in a normal state or a digital signal based on the image, the image of the Cherenkov radiation light of the core in an abnormal state, or a digital signal light in the image as big data. And, it is possible to estimate and diagnose the state of the core by searching for an image having the highest similarity to the digital signal converted in real time.

In addition, the core diagnosis unit 230 may diagnose whether the core output of the nuclear reactor and the internal state of the core are normal according to a digital signal from an image captured in real time through the learned neural network.

Further, the learning unit 240 trains a neural network that outputs a nuclear reactor for a digital signal value of the Cherenkov radiation and estimates whether the internal state of the core is normal, based on the previously stored learning data. The learning unit 240 performs deep learning through a deep neural network.

Here, the learning data includes a digital signal of Cherenkov radiated light according to a change in output from a zero output to a maximum output at the time of starting the reactor and a change in the core. Accordingly, the learning unit 240 may learn and analyze in units of outputs that can be set according to the Cherenkov state.

In addition, the learning unit 240 may continuously learn the neural network by using data according to the diagnosis result of the core diagnosis unit 230 as learning data.

Hereinafter, a method of diagnosing a core through Cherenkov radiation from a nuclear reactor photographed with a digital camera will be described in detail using FIGS. 3 and 4. Here, a method of diagnosing a core using deep learning, which is one of the methods of diagnosing a core, will be described.

3 is a flowchart illustrating a diagnosis method of a core diagnosis apparatus according to an embodiment of the present invention, and FIG. 4 is an exemplary view showing an RGB histogram according to an embodiment of the present invention.

As shown in FIG. 3, the core diagnosis apparatus 200 photographs Cherenkov radiation emitted from the nuclear reactor through the interlocked camera 300 (S310).

The core diagnosis apparatus 200 may control the camera 300 installed in an area near the nuclear reactor to perform real-time or every set time to transmit a captured image. Then, the camera 300 may transfer the image captured of the nuclear reactor to the core diagnosis apparatus 200 and store it in a separate database.

Next, the core diagnosis apparatus 200 converts the Cherenkov radiation from the captured image into an RGB histogram (S320).

The core diagnosis apparatus 200 may convert the captured image into an RGB histogram based on the amount of data according to the intensity of light.

4 shows an RGB histogram converted according to a reactor output value. The graphs of FIG. 4 show a function representing the number of pixels having the density level for each density level for an image, or a ratio to the number of all pixels.

Accordingly, in the graphs of FIG. 4, the horizontal axis represents brightness, and the vertical axis represents the amount of data, that is, the amount of data having a corresponding brightness value. Accordingly, as the horizontal axis value increases, the degree of brightness increases, and as the vertical axis value increases, the amount of data increases.

The graphs shown in FIG. 4 are graphs showing a process up to a point in time when a reactor is started to perform maximum output and output of the reactor is terminated in an RGB histogram.

For example, (a) of FIG. 4 is a graph obtained by converting an image captured in a situation of 0 MW before the reactor is started into a histogram, and it can be seen that values without a brightness value are the most. 4(b) is a situation in which the output of the reactor is 15MW, and many values of high brightness are displayed, and (c) is a situation where the output of the reactor is 30MW, and the degree of brightness is close to 100%. It can be seen that it is concentrated.

FIG. 4(d) shows that when the output of the nuclear reactor is 25MW, a lot of values close to 100% of the brightness were detected, but many values having a relatively low brightness compared to FIG. 4(c) were converted. Able to know. Next, (e) of FIG. 4 is a graph in which the degree of brightness is uniformly converted to the point when the start of the reactor is finished, and (f) is a graph in which the situation after a certain period of time has elapsed from the point of the start of the reactor is converted into a histogram. to be.

In this way, it can be seen that the characteristics of the histogram appear different according to each time point due to the Cherenkov radiation emitted at the time when the reactor is started and the maximum output is performed and the output of the reactor is terminated. In particular, looking at (e) and (f) of Fig. 4, even after the use of the reactor is completed, the value of the degree of brightness can be confirmed due to the afterimage effect of light, so that the state of the reactor after the use of the reactor is completed can be known.

Next, the core diagnosis apparatus 200 converts the RGB histogram into a digital signal (S330).

The core diagnosis apparatus 200 may monitor a digital signal according to the Cherenkov radiation in real time.

In addition, the core diagnosis apparatus 200 trains the neural network by using the learning data (S340).

The neural network estimates whether the reactor output for the digital signal value of the Cherenkov radiation and the state inside the core are normal, and represents deep learning formed by a deep neural network.

Here, the learning data represents a digital signal of Cherenkov radiation according to a change in output and a change in a core according to the total output of the nuclear reactor or a partial output of the nuclear fuel channel. The core diagnosis apparatus 200 may use the measured value of the reactor core for the total output value of the reactor, and the partial output value of the nuclear fuel channel may use an estimated value calculated through a core calculation program.

In addition, the learning data may include digital signals of Cherenkov radiation according to conditions such as a nuclear reactor core, a flow tube of a nuclear reactor, a reactor component, and a normal state or an abnormal state of the spent nuclear fuel.

For example, the learning data may include a digital signal of Cherenkov radiation measured in abnormal conditions such as a situation in which a part of the nuclear fuel rod bundle is broken, a cooling channel is not normally operated, or a part of the channel is blocked.

In addition, after learning the neural network, the core diagnosis apparatus 200 may retrain the neural network when the result of the learned neural network has a reliability lower than a threshold value.

The core diagnosis apparatus 200 analyzes whether the core output of the nuclear reactor and the internal state of the core are normal by applying the digital signal acquired in real time to the previously learned neural network (S350).

The core diagnosis apparatus 200 may analyze an abnormal state by comparing an image and data in a normal state with an image and data captured in real time through the learned neural network.

Here, the data represents data that is compared and analyzed through an image, and may include reactor control data, reactor record data, inspection data, and driving data of a device interlocked with the reactor. The core diagnosis device 200 may analyze an abnormal state of a nuclear reactor core, and in addition, may analyze an abnormal state of a flow pipe of a nuclear reactor, a reactor component, and used nuclear fuel.

Next, the core diagnosis apparatus 200 diagnoses the core based on the analyzed output of the reactor and whether the inside of the core is normal (S360).

In addition, the core diagnosis apparatus 200 can diagnose whether the core state is a normal state or an abnormal state based on the data analyzed through the neural network, and also the abnormal state of the flow pipe of the reactor, the reactor parts, and the nuclear fuel after use. Can be diagnosed. In this way, the core diagnosis apparatus 200 can check the state of the core of the reactor in real time through a digital signal according to the Cherenkov radiation, and the abnormal state type of the core, the region estimated to be an abnormal state, and an abnormal state through a neural network learned in advance. You can diagnose the degree of condition, etc.

In this case, the core diagnosis apparatus 200 may be set to analyze the abnormal state of the core through the automatically learned neural network only when the monitored digital signal is out of the normal range or the effective range. These settings can be easily changed and designed by the user in the future.

In addition, when the core diagnosis apparatus 200 is in an abnormal state, it may generate and provide a notification message, and transmit the notification message to an interworking manager terminal.

Here, the notification message may include a type of an abnormal state, a device diagnosed as an abnormal state, a location of the corresponding device, and a detected time.

Meanwhile, the core diagnosis apparatus 200 may store a result of analyzing and diagnosing that a core state of a nuclear reactor is a normal or abnormal state, a corresponding reactor image, an RGB histogram of a corresponding image, and a digital signal in a database. Further, the core diagnosis apparatus 200 may improve the reliability of the neural network by learning the neural network using the stored analysis and diagnosis data. A program for executing a method according to an embodiment of the present invention may be recorded on a computer-readable recording medium.

The computer-readable medium may include program instructions, data files, data structures, and the like alone or in combination. The media may be specially designed and configured, or may be known and usable to those skilled in computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical recording media such as CD-ROMs and DVDs, magnetic-optical media such as floppy disks, and ROM, RAM, flash memory, and the like. Hardware devices specially configured to store and execute the same program instructions are included. Here, the medium may be a transmission medium such as an optical or metal wire, a waveguide, etc. including a carrier wave that transmits a signal specifying a program command, a data structure, and the like. Examples of the program instructions include not only machine language codes such as those produced by a compiler, but also high-level language codes that can be executed by a computer using an interpreter or the like.

Although the preferred embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto, and various modifications and improvements by those skilled in the art using the basic concept of the present invention defined in the following claims are also present. It belongs to the scope of rights of

100: reactor 200: core diagnosis device
210: collection unit 220: digital signal conversion unit
230: core diagnosis unit 240: learning unit
300: digital camera

Claims (14)

A collection unit that collects in real time a photographed image of the nuclear reactor from a camera mounted in an area near the reactor,
A digital signal conversion unit converting the brightness and color difference of the Cherenkov radiated light included in the captured image into an RGB histogram, and converting the converted RGB histogram into a digital signal, and
A core diagnosis apparatus comprising a core diagnosis unit for diagnosing whether a core output of the nuclear reactor and an internal state of the core are normal according to the digital signal.
delete In claim 1,
The core diagnosis unit,
By monitoring the digital signal according to the Cherenkov radiated light, the state of the core of the nuclear reactor is checked in real time,
A core diagnosis device that diagnoses the type of abnormal state of the core, the area estimated to be the abnormal state, and the degree of the abnormal state.
In claim 1,
The core diagnosis unit,
A core diagnosis apparatus for diagnosing an abnormal state of a core by comparing a Cherenkov radiation image in a normal state with a Cherenkov radiation image captured in real time when the reactor is started.
In claim 1,
The collection unit,
Collecting Cherenkov radiation images taken in real time of the nuclear fuel that has been used in the reactor,
The core diagnosis unit,
A core diagnostic apparatus for estimating the state of the nuclear fuel by comparing the Cherenkov radiation image of normal nuclear fuel with the photographed Cherenkov radiation image, and diagnosing an abnormal state of the used nuclear fuel.
A collection unit that collects in real time a photographed image of the nuclear reactor from a camera mounted in an area near the reactor,
A digital signal conversion unit for converting the brightness and color difference of the Cherenkov radiation included in the captured image into a digital signal,
A learning unit for learning a neural network for estimating whether a nuclear reactor output for a digital signal value of the Cherenkov radiation and a state inside the core is normal based on the previously stored learning data, and
A core diagnosis apparatus comprising a core diagnosis unit for diagnosing whether a core output of the nuclear reactor and an internal state of the core are normal according to the digital signal through a pre-learned neural network.
In paragraph 6,
The training data,
Core diagnosis apparatus including a digital signal of Cherenkov radiation according to a change in output from a zero output to a maximum output and a change in a core when the reactor is started.
In paragraph 6,
The training data,
The reactor output value includes a total output value that is a measured value of the nuclear reactor core, and a partial output value of a nuclear fuel channel that is an estimated value calculated through a core calculation program.
In paragraph 6,
The learning unit,
A core diagnosis apparatus for continuously learning the neural network by using data according to a diagnosis result of the core diagnosis unit as learning data.
In claim 1,
The camera,
A digital camera, a core diagnostic device installed above the reactor to photograph the reactor.
Photographing the Cherenkov radiation emitted from the reactor through an interlocked camera,
Converting the Cherenkov radiation from the captured image into an RGB histogram,
Converting the RGB histogram into a digital signal,
Analyzing whether the core output of the nuclear reactor and the internal state of the core are normal according to the digital signal, and
Diagnosing the core based on the analyzed output of the nuclear reactor and whether the inside of the core is normal,
Diagnosis method of a core diagnosis device comprising a.
In clause 11,
Analyzing whether the core output and the internal state of the nuclear reactor are normal,
A diagnostic method for a core diagnosis apparatus for comparing and analyzing a Cherenkov radiation image captured in real time with a Cherenkov radiation image in a normal state for the nuclear reactor core.
In clause 11,
Further comprising the step of learning a neural network using digital signals of Cherenkov radiation according to a change in output and a change in a core according to a total output of the nuclear reactor or a partial output of a nuclear fuel channel,
The analyzing step,
A method for diagnosing a core diagnosis apparatus that applies the digital signal to a pre-learned neural network to analyze whether the core output of the nuclear reactor and the internal state of the core are normal.
In claim 13,
The step of training the neural network,
When the diagnosis of the nuclear reactor core is completed, related data is stored in a database and used as data for training the neural network.

Images