KR102427630B1 - Apparatus and method for predicting failure of thermal image system - Google Patents

Abstract

The present invention relates to a state checking device for checking the state of a piece of thermal imaging monitoring equipment. The state checking device includes: a data acquisition unit for acquiring temperature data of a detector installed in the thermal imaging monitoring equipment; a reference data generation unit for performing deep learning on the basis of temperature data collected from other pieces of thermal imaging monitoring equipment to generate reference data for comparison with the temperature data acquired by the data acquisition unit; a determination unit for comparing the temperature data acquired by the data acquisition unit with the reference data generated by the reference data generation unit, and determining the state of the thermal imaging monitoring equipment; and a failure prediction unit for predicting an expected failure time point of the thermal imaging monitoring equipment according to a determination result of the determination unit. It is possible to check the state of a piece of thermal imaging monitoring equipment and to predict the time point of failure of the thermal imaging monitoring equipment.

Description

A device and method for predicting the failure point of a thermal monitoring system

The present invention relates to an apparatus and method for predicting a failure point of a thermal imaging device, and more particularly, to an apparatus and method for predicting a failure time of a thermal imaging device, which can check the state of the thermal imaging device and predict the expected failure time of the thermal imaging device it's about

In general, monitoring equipment is used to monitor a specific space. Surveillance equipment can be divided into camera surveillance equipment using a CCD (Chang Coupled Device) camera, and thermal imaging equipment using a thermal camera.

CCD cameras can only monitor where there is light, so they are not suitable for use in places where there is no light or at night. Thermal cameras can be used for night surveillance by detecting heat emitted from objects. Therefore, in the military, since it is necessary to monitor the enemy's approach to the front, the shore, the river, etc. not only during the daytime but also at night, a thermal imaging system using a thermal camera is widely used.

At this time, if the thermal monitoring system malfunctions while performing its mission, there is a void in the power. Therefore, the operator periodically visually checks the thermal imaging device and performs signal inspection of the thermal imaging device using the test equipment. However, since this type of inspection is affected by the operator's experience or tendency, there is a problem in that it is difficult to confirm the exact state of the thermal imaging device.

US 10-0919834 B

The present invention provides an apparatus and method for predicting a failure time of a thermal imaging device capable of confirming the state of the thermal imaging device and predicting an expected failure time of the thermal imaging device.

The present invention provides an apparatus and method for predicting a failure point of a thermal imaging device capable of stably operating the thermal imaging monitoring device.

The present invention provides a state checking device for checking the state of a thermal imaging device, comprising: a data acquisition unit for acquiring temperature data of a detector mounted on the thermal imaging device; a reference data generator for deep learning the temperature data collected from other thermal imaging devices and generating reference data for comparison with the temperature data acquired by the data acquisition unit; a determination unit for comparing the temperature data acquired by the data acquisition unit with the reference data generated by the reference data generation unit, and determining a state of the thermal imaging device; and a failure prediction unit for predicting an expected failure time of the thermal imaging device according to the determination result of the determination unit.

The data acquisition unit may include: an operation sensor for detecting the operation of the thermal imaging device; and a data collector for collecting temperature data measured by the temperature of the detector for a preset time from the time the thermal imaging device operates.

The data acquisition unit further includes a data processor for removing noise from the temperature data collected by the data collector.

The data processor may include: a noise identifier for identifying noise among the temperature values by comparing temperature values included in the temperature data with a preset temperature set value; and a noise canceller for replacing the noise identified by the noise identifier with another temperature value.

The reference data generation unit generates the reference data by machine learning the temperature data collected from other thermal imaging devices using a convolutional neural network model.

The determination unit may include: a comparator configured to compare the temperature data and the reference data to calculate a matching rate; and a state determiner for judging the state of the thermal imaging device according to the coincidence rate calculated by the comparator as one of a normal operable state, a state in which operating performance is lower than the normal state, and a failure in an inoperable state; includes

The failure prediction unit may include: an information storage in which temperature data collected from other thermal imaging devices is stored; and a failure predictor for predicting when the state of the thermal monitoring equipment is determined to be a failure by using the temperature data stored in the information storage device when the state determiner determines that the state of the thermal monitoring equipment is abnormal.

The failure predictor may include: a change amount predictor for predicting a degree to which the temperature data obtained from the data acquisition unit changes with time using the temperature data stored in the information storage unit; and an expected failure time selector for selecting a time when the temperature data obtained from the data acquisition unit is changed and the state of the thermal monitoring equipment is changed from abnormal to failure as an expected failure time of the thermal monitoring equipment.

The present invention provides a status checking method for checking the status of a thermal imaging device, comprising: acquiring temperature data of a detector mounted in the thermal imaging device; generating reference data for comparison with temperature data obtained by deep learning the temperature data collected from other thermal imaging devices; comparing the obtained temperature data with the reference data, and determining the state of the thermal imaging device; and predicting an expected failure time of the thermal imaging device according to the state of the thermal imaging device.

A temperature sensor is mounted on the thermal imaging device to measure the temperature of the detector, and the process of acquiring the temperature data of the detector includes the process of receiving the temperature data measured by the temperature sensor.

The process of receiving the temperature data measured by the temperature sensor may include: confirming the operation of the thermal imaging device; and collecting temperature data measured by the temperature sensor for a preset time from the time of operation of the thermal imaging device when the thermal imaging device starts to operate.

The process of receiving the temperature data measured by the temperature sensor further includes a process of removing noise from the temperature data.

The process of removing noise from the temperature data may include: comparing temperature values included in the temperature data with a preset temperature set value; determining, among the temperature values, a temperature value equal to or greater than the set temperature value as noise; and replacing the temperature value determined as noise with another temperature value.

The process of determining the state of the thermal imaging device may include: calculating a coincidence rate by comparing the graph form of the temperature data and the reference data; and a process of determining the state of the thermal imaging device as one of a normal operable state, a state in which the operating performance is lower than the normal state, and a failure in an inoperable state according to the coincidence rate.

The process of judging the state of the thermal imaging device as one of normal, abnormal, and faulty is that if the coincidence rate is 75% or more, it is determined that it is normal, and if the coincidence rate is 50% or more to less than 75%, it is determined that it is abnormal, and the If the coincidence rate is less than 50%, it includes the process of judging a failure.

The process of predicting the expected failure time of the thermal imaging device may include: acquiring temperature data collected from other thermal imaging devices; a process of estimating the degree to which the temperature data of the thermal imaging device judged to be abnormal will change with time, using temperature data of other thermal imaging devices, if it is determined that the state of the thermal imaging device is abnormal; and a process of predicting a time when the temperature data of the thermal imaging device determined as above is changed and the coincidence rate with the reference data changes to less than 50% as the expected failure time of the thermal imaging device.

According to embodiments of the present invention, it is possible to accurately and automatically check the state of the thermal imaging device. Accordingly, it is possible to easily determine whether the state of the thermal imaging device is normal, abnormal, or malfunctioning. In addition, when the state of the thermal imaging device is abnormal, it is possible to predict the expected failure time of the thermal imaging device. Therefore, since the maintenance is performed while monitoring the state of the thermal imaging device, it is possible to stably operate the thermal imaging device.

1 is a view showing the structure of a thermal imaging device failure time prediction apparatus according to an embodiment of the present invention.
2 is a diagram illustrating an operation process of a state determiner and a failure predictor according to an embodiment of the present invention.
3 is a flowchart illustrating a method for predicting a failure point of a thermal imaging device according to an embodiment of the present invention.
4 is a diagram illustrating a process of removing noise from temperature data according to an embodiment of the present invention.
5 is a graph comparing reference data and temperature data according to an embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in more detail with reference to the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but will be implemented in various different forms, and only these embodiments allow the disclosure of the present invention to be complete, and the scope of the invention to those of ordinary skill in the art will be completely It is provided to inform you. In order to explain the invention in detail, the drawings may be exaggerated, and like numerals refer to like elements in the drawings.

1 is a diagram showing the structure of a failure time prediction apparatus for thermal imaging monitoring equipment according to an embodiment of the present invention, and FIG. 2 is a diagram showing an operation process of a state determiner and a failure prediction unit according to an embodiment of the present invention. Hereinafter, an apparatus for predicting a failure point of a thermal imaging device according to an embodiment of the present invention will be described.

The thermal imaging device failure time prediction device according to an embodiment of the present invention is a state checking device for checking the state of the thermal imaging device. Referring to FIG. 1 , the apparatus 100 for predicting a failure point of a thermal imaging device includes a data acquisition unit 110 , a reference data generation unit 120 , a determination unit 130 , and a failure prediction unit 140 .

At this time, in order to understand the present invention, the thermal imaging device 50 will be described first. The thermal imaging device 50 is arranged to monitor a specific space. For maintenance, it is possible to take a break every predetermined time while operating the thermal monitoring system 50 . Accordingly, the start and end of the operation of the thermal imaging device 50 may be performed periodically. The thermal imaging device 50 includes a detector 51 and a temperature sensor 52 .

The detector 51 is a device mounted on the thermal imaging device 50 . The detector 51 may detect heat emitted from the object. The detector 51 includes a detection member 51a and a cooling member 51b.

The detection member 51a detects infrared rays and outputs them as electrical signals. Accordingly, an image representing the temperature distribution of the object may be generated using the electrical signal.

The cooling member 51b is installed to cool the detection member 51a. That is, since the detection member 51a operates sensitively to the ambient temperature, there is a problem that the value output by the detection member 51a increases as the ambient temperature increases. Accordingly, by cooling the detection member 51a to a set temperature with the cooling member 51b, it is possible to adjust the detection member 51a to output a constant value regardless of the ambient temperature.

The temperature sensor 52 is a device mounted on the thermal imaging device 50 . The temperature sensor 52 is installed to measure the temperature of the detector 51 . In detail, the temperature sensor 52 may measure the temperature of the detection member 51a. For example, the temperature sensor 52 may measure an absolute temperature. Accordingly, it can be confirmed whether the detection member 51a is cooled to a set temperature by the cooling member 51b. Accordingly, when the detection member 51a reaches a set temperature, it is activated to detect infrared rays. However, the unit of the temperature measured by the temperature sensor 52 is not limited thereto and may vary.

At this time, when a failure occurs in the detection member 51a or the cooling member 51b when the thermal imaging monitoring device 50 performs a task, there is a void in the power. Accordingly, it is possible to monitor the state of the thermal imaging device 50 by using the device for predicting the failure time of the thermal imaging device 100 . The thermal imaging device failure time prediction device 100 uses the temperature sensor 52 mounted in the thermal imaging device 50 without adding a separate sensor to check the state of the thermal imaging device 50 . Accordingly, even if there is a restriction that a sensor cannot be added to the thermal imaging device 50 , the state of the thermal imaging device 50 can be easily monitored.

The data acquisition unit 110 may be connected to the thermal imaging device 50 wirelessly or by wire. Accordingly, the data acquisition unit 110 may acquire temperature data of the detector 51 measured by the temperature sensor 52 . The data acquisition unit 110 includes an operation sensor 111 and a data collector 112 .

The operation sensor 111 may detect the operation of the thermal imaging device 50 . For example, the operation sensor 111 may detect that power is supplied to the detector 51 , and may determine when the thermal imaging device 50 operates.

The data collector 112 may be connected to transmit and receive a signal to and from the temperature sensor 52 . Accordingly, when the operation sensor 111 detects the operation of the thermal imaging device 50 , the data collector 112 may start to receive the temperature value measured by the temperature sensor 52 . Accordingly, the data collector 112 may collect temperature data obtained by measuring the temperature of the detector 51 for a preset time from the time the thermal imaging device 50 operates. In addition, the data collector 112 may set a period in which the temperature sensor 52 measures the temperature. For example, the set time is 10 minutes and the temperature sensor 52 is set to measure the temperature in units of 1 second, and the temperature of the detector 51 is set in units of 1 second for 10 minutes from the time when the thermal imaging device 50 is operated. The measured temperature data may be collected by the data collector 112 . However, the set time and the period in which the temperature sensor 52 is set to measure the temperature is not limited thereto and may vary.

Meanwhile, the data acquisition unit 110 may further include a data processor 113 . The data processor 113 may be connected to send and receive a signal to and from the data collector 112 . Accordingly, the data processor 113 may receive the temperature data collected by the data collector 112 and remove noise from the temperature data collected by the data collector 112 . Accordingly, it is possible to more accurately check the state of the thermal imaging device 50 using the temperature data from which the noise is removed. The data processor 113 includes a noise identifier 113a and a noise canceller 113b.

The noise identifier 113a may compare temperature values included in the temperature data with a preset temperature set value. Accordingly, among the temperature values of the temperature data, a temperature value greater than or equal to the set temperature may be determined as noise. Accordingly, noise can be identified among the temperature values.

The noise remover 113b may remove the noise by replacing the noise identified by the noise identifier 113a with another temperature value. For example, the noise remover 113b replaces the next or previous temperature value of the noise with the noise among the temperature values included in the temperature data, or an average value of two temperature values selected from the temperature values included in the temperature data. can be replaced with noise.

The reference data generation unit 120 may generate reference data for comparison with the temperature data obtained by the data acquisition unit 110 by deep learning the temperature data collected from other thermal imaging devices. In detail, the reference data generator 120 may perform machine learning with a convolutional neural network (CNN) model, collect temperature data from other thermal imaging devices, and generate reference data.

The determination unit 130 may be connected to send and receive signals to and from the data acquisition unit 110 and the reference data generation unit 120 . Accordingly, the determination unit 130 may compare the temperature data obtained by the data obtaining unit 110 with the reference data generated by the reference data generation unit 120 and determine the state of the thermal imaging monitoring device 50 . . The determination unit 130 includes a comparator 131 and a state determiner 132 .

The comparator 131 may compare the temperature data and the reference data to calculate a matching rate between the two. For example, the temperature data and the reference data may include graphs representing the amount of change in temperature over time. Accordingly, the comparator 131 may compare the graph form of the temperature data and the reference data. Accordingly, by superimposing the two graphs, it is possible to easily calculate the coincidence rate between the two graphs.

The state determiner 132 may be connected to transmit and receive a signal to and from the comparator 131 . Accordingly, the state determiner 132 determines the state of the thermal imaging device 50 according to the coincidence rate calculated by the comparator 131 , which is an operable state, a state in which the operating performance is lower than that of the normal state, and an inoperable state. It can be judged as one of the phosphorus failures. For example, as shown in FIG. 2 , if the coincidence rate is 75% or more, it is determined as normal; Normal means a state in which the thermal monitoring device 50 can operate normally, failure means a state in which a failure occurs in the thermal imaging monitoring device 50 and requires repair, and abnormal is an intermediate state between normal and failure. Although the monitoring equipment 50 can operate, it means a state with a high probability of failure while performance degradation occurs. Therefore, if the state determiner 132 determines that it is normal, the thermal monitoring device 50 may continue to be operated, and if it is determined that it is malfunctioning, the operation of the thermal imaging monitoring device 50 may be stopped and repaired. If the state determiner 132 determines that there is an abnormality, it is possible to perform repair work before an actual failure occurs while continuously monitoring the state of the thermal imaging device 50 .

As such, it is possible to accurately and automatically check the state of the thermal imaging device 50 . Therefore, it is possible to easily determine whether the state of the thermal imaging device 50 is normal, abnormal, or malfunctioning, and select a response according to the state of the thermal imaging device 50 to stably operate the thermal imaging device.

The failure prediction unit 140 may be connected to transmit and receive signals to and from the determination unit 130 . Accordingly, the failure prediction unit 140 may predict the expected failure time at which the failure of the thermal imaging device 50 will occur according to the determination result of the determination unit 130 . That is, when the state determiner 132 determines that the state of the thermal monitoring device 50 is abnormal as shown in FIG. 2 , the failure prediction unit 140 predicts the expected failure time of the thermal monitoring device 50 . can do. The failure predictor 140 includes an information storage 141 and a failure predictor 142 .

The information storage 141 may store temperature data collected from other thermal imaging devices. In detail, temperature data according to time of other thermal imaging devices may be stored. Accordingly, it is possible to check how the temperature data of other thermal imaging devices change over time.

The failure predictor 142 may be connected to send and receive signals to and from the state determiner 132 and the information storage 141 . Accordingly, when the failure predictor 142 determines that the state of the thermal monitoring equipment 50 is abnormal, the failure predictor 142 uses the temperature data stored in the information storage 141 to change the state of the thermal monitoring equipment to failure. The timing of judgment can be predicted. The failure predictor 142 includes a change amount predictor 142a and an expected failure time selector 142b.

The change amount estimator 142a may use the temperature data stored in the information storage 141 to estimate the degree to which the temperature data acquired from the data acquisition unit 110 changes with time. For example, the temperature data obtained from the data acquisition unit 110 and the temperature data obtained from the data acquisition unit 110 among the temperature data stored in the information storage 141 are searched for temperature data that is 95% or more, and the amount of change of the temperature data is determined to be abnormal. It can be determined that the future change amount of the temperature data of (50). In this case, when there are a plurality of temperature data obtained from the data acquisition unit 110 and a matching rate of 95% or more among the temperature data stored in the information storage 141 , the temperature data having the highest matching rate may be selected.

The predicted failure point selector 142b may be connected to transmit and receive a signal to and from the change amount predictor 142a. Accordingly, the predicted failure time selector 142b determines the time when the temperature data acquired from the data acquisition unit 110 changes and the state of the thermal monitoring equipment 50 changes from abnormal to failure. You can choose the point in time. In detail, the change amount predictor 142a calculates a coincidence rate with the reference data while changing the temperature data according to the expected change amount, and as time passes, it is possible to find a time point at which the state is determined to be a failure. Therefore, it can be predicted that the failure of the thermal imaging device 50 may occur after the time taken until the time when it is determined as a failure.

As such, when the state of the thermal imaging device 50 is abnormal, the expected failure time of the thermal imaging device may be predicted. Accordingly, while monitoring the state of the thermal imaging device 50, it is possible to preemptively repair the thermal imaging device 50 before failure actually occurs. Accordingly, the maintenance of the thermal imaging monitoring device 50 is facilitated, and it is possible to prevent the occurrence of a power gap due to a sudden failure of the thermal imaging monitoring device 50 .

On the other hand, the thermal imaging device failure time prediction apparatus 100 may further include a display unit (not shown). The display unit may be a display device, and may be connected to transmit and receive a signal with at least one of the determination unit 130 and the failure prediction unit 140 . Accordingly, the display unit may receive and visually display at least one of the information about the state of the thermal imaging device 50 determined by the determination unit 130 and the predicted failure time predicted by the failure prediction unit 140 . Therefore, the operator can easily check the state of the thermal imaging device 50 to perform appropriate maintenance work.

3 is a flowchart illustrating a method for predicting a failure point of a thermal imaging device according to an embodiment of the present invention, FIG. 4 is a diagram illustrating a process of removing noise from temperature data according to an embodiment of the present invention, and FIG. 5 is this It is a graph comparing reference data and temperature data according to an embodiment of the present invention. Hereinafter, a method for predicting a failure point of a thermal imaging device according to an embodiment of the present invention will be described.

The method for predicting the failure time of the thermal imaging device according to an embodiment of the present invention is a state checking method for checking the state of the thermal imaging device. Referring to FIG. 3 , the method for predicting the failure point of the thermal imaging device is a process of acquiring temperature data of a detector mounted in the thermal imaging device (S110), deep learning the temperature data collected from other thermal imaging devices, and obtained The process of generating reference data for comparison with the temperature data (S120), the process of comparing the obtained temperature data with the reference data, and determining the state of the thermal imaging device (S130), and the thermal image according to the state of the thermal imaging device It includes the process of predicting the expected failure time of the monitoring equipment (S140).

In this case, the method for predicting the failure point of the thermal imaging device may be performed by the device for predicting the failure point of the thermal imaging device according to an embodiment of the present invention having a structure as shown in FIG. 1 . In addition, when the thermal imaging device 50 starts to operate, the cooling member 51b provided in the detector 51 operates to cool the detection member 51a to a set temperature. If there is a problem with the thermal imaging device 50, the detection member 51a may not be properly cooled by the cooling member 51b when the thermal imaging device 50 is operated. Since the thermal imaging monitoring device 50 is operated for maintenance and taking a break every predetermined time, whenever the thermal imaging monitoring device 50 is operated, the temperature state of the detector 51 is checked and cooled by a method for predicting the failure point of the thermal imaging device 50 It can be checked whether the detection member 51a is normally cooled by the member 51b. However, the present invention is not limited thereto, and the method for predicting the failure time of the thermal monitoring equipment may be performed by various thermal monitoring equipment failure time prediction devices.

First, the temperature data of the detector mounted on the thermal imaging device is acquired (S110). In detail, temperature data of the detector 51 measured by the temperature sensor 52 mounted in the thermal imaging device 50 may be acquired. At this time, the operation of the thermal imaging device 50 can be confirmed. Accordingly, when the thermal imaging device 50 starts to operate, the temperature data measured by the temperature sensor 52 may be collected for a preset time from the operation time of the thermal imaging device 50 . That is, since the detector 51 collects temperature data that can confirm the time it takes for the detector 51 to cool to the set temperature by the operation of the thermal imaging device 50, after the detector 51 is cooled to the set temperature, the temperature data is no need to collect Accordingly, temperature data may be collected only for a set time after the thermal imaging device 50 starts to operate.

Meanwhile, after collecting the temperature data, a process of removing noise from the temperature data may be further performed. That is, some of the temperature values included in the temperature data may be abnormally measured as shown in FIG. 4( a ) due to a problem with the temperature sensor 52 . Accordingly, the abnormally measured temperature value can be identified and removed as noise.

To this end, temperature values included in the temperature data may be compared with a preset temperature set value. The noise may be identified by determining a temperature value equal to or greater than the temperature set value among the temperature values as noise. For example, the temperature set value may be 300K. Therefore, a temperature value of 300K or higher can be identified as noise. However, the temperature set value is not limited thereto and may vary.

Also, a temperature value determined as noise may be replaced with another temperature value and removed. For example, if the next measured temperature value of the noise is less than the temperature set value, the noise may be replaced with the next measured temperature value as shown in FIG. 4B . Alternatively, if the next measured temperature value of the noise is greater than or equal to the temperature set value, and the previously measured temperature value of the noise is less than the temperature set value, replace the noise with the previously measured temperature value as shown in FIG. 4(b) can do.

On the other hand, if both the next measured temperature value and the previously measured temperature value of the noise are equal to or greater than the temperature set value, the average value of the two temperature values selected from the temperature values included in the temperature data as shown in FIG. can be replaced with For example, it is possible to select two temperature values measured before and after the same time interval with noise interposed therebetween, and obtain an average value of the two temperature values to replace the noise.

Then, the temperature data collected from other thermal imaging devices is deep-learned, and reference data for comparison with the acquired temperature data is generated (S120). In detail, by machine learning with a convolutional neural network model, reference data can be generated based on temperature data collected from other thermal imaging devices. Accordingly, it is possible to determine the state of the thermal imaging device 50 by analyzing the temperature data based on the reference data. In this case, the process of generating the reference data may be performed after the process of acquiring the temperature data, before the process of acquiring the temperature data, or both processes may be performed together.

Then, the temperature data and the reference data are compared, and the state of the thermal imaging device is determined (S130). For example, the temperature data and the reference data may include graphs representing the amount of change in temperature over time. Accordingly, as shown in FIG. 5 , the matching rate can be calculated by comparing the graph form of the temperature data and the reference data. That is, it is possible to calculate how different the graph form of the temperature data is from the graph form of the reference data in a state in which the two graphs are superimposed.

In addition, according to the coincidence rate, it is possible to determine the state of the thermal imaging device 50 as one of a normal operable state, a state in which the operating performance is lower than the normal state, and a failure in an inoperable state. For example, if the coincidence rate is 75% or more, it is determined as normal; That is, the higher the coincidence rate, the faster the detector 51 is cooled to the set temperature, and the lower the coincidence rate, the longer it takes for the detector 51 to cool to the set temperature. Accordingly, when the matching rate is low, it can be determined that a problem occurs in the cooling member 51b or the detection member 51a, and the cooling rate is slowed down. Therefore, if it is determined that the state of the thermal imaging device 50 is normal, the thermal imaging device 50 may be continuously operated, and if it is determined that the thermal imaging device 50 is malfunctioning, the operation of the thermal imaging device 50 may be stopped and repaired. If it is determined that the state of the thermal imaging device 50 is abnormal, it is possible to perform repair work before an actual failure occurs while continuously monitoring the state of the thermal imaging device 50 . However, the criteria for judging normal, abnormal, and failure are not limited thereto and may vary.

As such, it is possible to accurately and automatically check the state of the thermal imaging device 50 . Therefore, it is possible to easily determine whether the state of the thermal imaging device 50 is normal, abnormal, or malfunctioning, and select a response according to the state of the thermal imaging device 50 to stably operate the thermal imaging device.

Next, the predicted failure time of the thermal imaging device is predicted according to the state of the thermal imaging device (S140). In order to predict the expected failure time of the thermal imaging device 50, temperature data collected from other thermal imaging devices may be acquired. Specifically, it is possible to acquire temperature data that changes over time while operating other thermal imaging devices. When it is determined that the state of the thermal imaging device 50 is abnormal, it is possible to estimate the degree to which the temperature data of the thermal imaging device determined to be abnormal changes with time using the temperature data of other thermal imaging devices. For example, among the temperature data of other thermal monitoring equipment, the temperature data of the thermal imaging device 50 determined to be abnormal is found, and the temperature data having a coincidence rate of 95% or higher, and the thermal monitoring device determined to be abnormal in the amount of change of the corresponding temperature data ( 50) can be determined as the future change of temperature data.

In addition, if the future change amount of the temperature data of the thermal imaging monitoring device 50 determined to be abnormal is determined, the temperature data of the thermal imaging monitoring device 50 judged to be abnormal is changed, so that the coincidence rate with the reference data is changed to less than 50%. The time may be predicted as the expected failure time of the corresponding thermal imaging device 50 . In detail, by changing the temperature data of the thermal imaging monitoring device 50 determined to be abnormal according to the expected amount of change, the coincidence rate with the reference data is calculated, and the time point at which the condition is determined to be a failure can be found. Therefore, it can be predicted that the failure of the thermal imaging device 50 may occur after the time taken until the time when it is determined as a failure.

As such, when the state of the thermal imaging device 50 is abnormal, the expected failure time of the thermal imaging device may be predicted. Accordingly, while monitoring the state of the thermal imaging device 50, it is possible to preemptively repair the thermal imaging device 50 before failure actually occurs. Accordingly, the maintenance of the thermal imaging monitoring device 50 is facilitated, and it is possible to prevent the occurrence of a power gap due to a sudden failure of the thermal imaging monitoring device 50 .

On the other hand, after determining the state of the thermal imaging device 50 and predicting the expected failure time of the thermal imaging device 50, among the information about the determined state of the thermal imaging device 50 and the predicted predicted failure time At least one may be visually displayed. Therefore, the operator can easily check the state of the thermal imaging device 50 to perform appropriate maintenance work.

As such, although specific embodiments have been described in the detailed description of the present invention, various modifications are possible without departing from the scope of the present invention, and various combinations between the embodiments are possible. Therefore, the scope of the present invention should not be limited to the described embodiments, but should be defined by the claims to be described below as well as the claims and equivalents.

50: thermal monitoring equipment 51: detector
52: temperature sensor 100: thermal imaging device failure point prediction device
110: data acquisition unit 120: reference data generation unit
130: determination unit 140: failure prediction unit

Claims (16)

As a prediction device for predicting the time of failure by checking the state of the thermal imaging device,
a data acquisition unit for acquiring temperature data of a detector mounted on the thermal imaging device;
a reference data generator for deep learning the temperature data collected from other thermal imaging devices and generating reference data for comparison with the temperature data obtained by the data acquisition unit;
a determination unit for comparing the temperature data acquired by the data acquisition unit with the reference data generated by the reference data generation unit, and determining a state of the thermal imaging device; and
and a failure prediction unit for predicting an expected failure time of the thermal imaging device according to the determination result of the determination unit;
The temperature data includes temperature values measured in a preset time unit for a preset time,
The data acquisition unit,
a data processor for removing noise from the acquired temperature data;
The data processor is
a noise identifier for comparing the temperature values with a preset temperature set value to identify a temperature value equal to or greater than the temperature set value among the temperature values as noise; and
A thermal imaging device failure point prediction device comprising a noise remover for replacing a temperature value identified as noise with any one of a temperature value measured immediately following the noise and a temperature value measured immediately before the noise. The method according to claim 1,
The data acquisition unit,
an operation sensor for detecting the operation of the thermal imaging device; and
A thermal imaging device failure time prediction device comprising a; a data collector for collecting temperature data of the temperature of the detector for a preset time from the time the thermal imaging device operates. delete delete The method according to claim 1,
The reference data generation unit,
A thermal imaging device failure time prediction device for generating the reference data by machine learning temperature data from other thermal imaging devices using a convolutional neural network model. The method according to claim 1,
the judging unit,
a comparator for calculating a matching rate by comparing the temperature data and the reference data; and
A state determiner for judging the state of the thermal monitoring device as one of a normal operable state, a state in which the operating performance is lower than the normal state, and a failure in an inoperable state according to the coincidence rate calculated by the comparator; Including thermal imaging device failure point prediction device. 7. The method of claim 6,
The failure prediction unit,
an information storage in which temperature data collected from other thermal imaging devices is stored; and
When the state determiner determines that the state of the thermal monitoring equipment is abnormal, a failure predictor for predicting when the state of the thermal monitoring equipment is determined to be a failure using the temperature data stored in the information storage device; point of time prediction. 8. The method of claim 7,
The failure predictor,
a change amount estimator for estimating the degree to which the temperature data obtained from the data acquisition unit changes with time using the temperature data stored in the information storage unit; and
Thermal monitoring equipment failure time including; an expected failure time selector for selecting a time when the temperature data obtained from the data acquisition unit is changed and the state of the thermal monitoring equipment is changed from abnormal to malfunctioning as an expected failure time of the thermal monitoring equipment prediction device. As a prediction method for predicting the time of failure by checking the state of the thermal monitoring system,
acquiring temperature data of a detector mounted on the thermal imaging device;
Deep learning the temperature data collected from other thermal monitoring equipment, and generating reference data for comparison with the obtained temperature data;
comparing the obtained temperature data with the reference data, and determining the state of the thermal imaging device; and
The process of predicting an expected failure time of the thermal imaging device according to the state of the thermal imaging device; including,
The temperature data includes temperature values measured in a preset time unit for a preset time,
The process of acquiring the temperature data of the detector is
Including the process of removing noise from the temperature data,
The process of removing noise from the temperature data is
comparing the temperature values with a preset temperature set value;
determining, among the temperature values, a temperature value equal to or greater than the set temperature value as noise; and
A method of predicting a failure point of a thermal imaging monitoring system, comprising: replacing the temperature value determined as noise with any one of a temperature value measured immediately after the noise and a temperature value measured immediately before the noise. 10. The method of claim 9,
A temperature sensor is mounted on the thermal imaging device to measure the temperature of the detector,
The process of acquiring the temperature data of the detector is
A method for predicting a failure point of a thermal imaging device, comprising the step of receiving the temperature data measured by the temperature sensor. 11. The method of claim 10,
The process of receiving the temperature data measured by the temperature sensor is,
confirming the operation of the thermal imaging device; and
When the thermal imaging device starts to operate, the process of collecting temperature data measured by the temperature sensor for a preset time from the operating time of the thermal imaging device; delete delete 10. The method of claim 9,
The process of determining the state of the thermal imaging device is,
calculating a matching rate by comparing the graph form of the temperature data and the reference data; and
The process of determining the state of the thermal imaging device as one of a normal operable state, a state in which the operating performance is lower than the normal state, and a failure in an inoperable state according to the coincidence rate; Prediction method. 15. The method of claim 14,
The process of determining the state of the thermal imaging device as any one of normal, abnormal, and malfunctioning is,
If the coincidence rate is 75% or more, it is determined that it is normal; . 16. The method of claim 15,
The process of predicting the expected failure time of the thermal imaging device is,
acquiring temperature data collected from other thermal imaging devices;
a process of estimating the degree to which the temperature data of the thermal imaging device judged to be abnormal will change with time by using the temperature data of other thermal imaging devices when it is determined that the state of the thermal imaging device is abnormal; and
The process of predicting the time when the temperature data of the thermal imaging device judged above is changed and the coincidence rate with the reference data changes to less than 50% as the expected failure time of the thermal imaging device; .

Images