KR102414080B1 - System and method for predicting cable failure through trend analysis - Google Patents

Abstract

A system and method for predicting cable failure through trend analysis are disclosed. The cable failure prediction system is mounted on a loop consisting of a pair of cables that electrically connect the + and - terminals of facilities included in the electrical and measurement control system in the form of a loop circuit, and measurement information of the loop current value sensed by the loop a current sensing sensor that transmits; and trend reference information of the loop current value during a preset calibration period using the measurement information of the loop current value received from the current detection sensor, the current trend change information for the current unit period, and the previous trend change information for the immediately preceding unit period , and applying the trend reference information, the current trend change information, and the immediately preceding trend change information to a predetermined cable failure prediction method to determine whether the loop equipped with the current detection sensor is in a normal state, or and a monitoring device that transmits a general alarm or an emergency alarm to be inspected to an administrator terminal.

Description

System and method for predicting cable failure through trend analysis

The present invention relates to a system and method for predicting cable failure through trend analysis.

A Distributed Control System (DCS) is a control system in which autonomous controllers are distributed, and is different from a centralized controller system in which supervisory control by a central operator exists.

In such a distributed control system, even if a dangerous situation occurs due to an aging cable at an arbitrary location, this risk is not notified in real time, so that safety measures cannot be taken quickly in many cases.

Korean Patent No. 10-1030589

The present invention can remotely predict the state (normal operation state, general inspection necessary state, or emergency inspection necessary state) of various cables constituting the electrical and instrumentation control system, so that appropriate measures can be quickly taken according to the state of each cable. It is to provide a cable failure prediction system and method through trend analysis.

The present invention predicts cable failure based on the loop current value measured by the current detection sensor and the current measured loop current value using a machine-learning prediction model using cable failure history information corresponding to the inspection result according to the cable failure prediction The purpose of this is to provide a cable failure prediction system and method through trend analysis that maximizes work convenience by allowing the condition of cables to be predicted quickly and accurately, and by selectively requesting a manager's inspection only when necessary.

The present invention relates to the same/similar environment (eg, voltage level, cable information (eg, single/stranded cable, cable diameter), location (eg, , indoor/outdoor, underground/ground), site type (eg, power plant/solar power plant, etc.), temperature, humidity, etc.) can be universally applied to different sites, and the applied prediction model is It is to provide a cable failure prediction system and method through trend analysis that can be further learned individually and improved to be optimized for the electrical and instrumentation control system of the site.

Objects other than the present invention will be easily understood through the following description.

According to one aspect of the present invention, for a facility included in an electrical and measurement control system, it is mounted on a loop consisting of a pair of cables that electrically connect the + terminal and the - terminal of the facility in a loop circuit shape, and the loop a current detection sensor that transmits measurement information of the sensed loop current value; and trend reference information of the loop current value during a preset calibration period using the measurement information of the loop current value received from the current detection sensor, the current trend change information for the current unit period, and the previous trend change information for the immediately preceding unit period , and applying the trend reference information, the current trend change information, and the immediately preceding trend change information to a predetermined cable failure prediction method to determine whether the loop equipped with the current detection sensor is in a normal state, or There is provided a cable failure prediction system through trend analysis including a monitoring device that transmits a general alarm or an emergency alarm to be inspected to an administrator terminal. Here, the trend reference information may include information on a reference average current value, a reference minimum current value, a reference maximum current value, a reference current fluctuation range value, and a reference maximum amount change period value regarding the fluctuation of the loop current value during the calibration period. can In addition, the current trend change information and the immediately preceding trend change information may include information about an average current value, a minimum current value, a maximum current value, and a current fluctuation width value related to a change in a loop current value during each unit period. .

The size of the current unit period and the immediately preceding unit period may be designated to be the same as the size of the reference maximum amount change period value.

The monitoring device may determine that the loop is in a normal state when the current value of the loop currently received from the current detection sensor is smaller than the reference maximum current value and less than or equal to the reference average current value.

When the loop current value currently received from the current detection sensor is equal to or greater than the reference maximum current value, the monitoring device includes a first variation multiplier obtained by dividing the current variation value included in the current trend change information by the reference current variation value and , when the first average current value magnification obtained by dividing the average current value included in the current trend change information by the reference average current value is all less than or equal to a predetermined first magnification value, a general alert for the inspection target may be transmitted.

When at least one of the first variation magnification and the first average current value magnification is greater than the first magnification value, the monitoring device sets the current variation value included in the immediately preceding trend change information to the current trend change information. A second variation multiplier divided by the included current variation value, a second average current value multiplier obtained by dividing the average current value included in the previous trend change information by the average current value included in the current trend change information, and the previous trend change The minimum current value multiplied by dividing the minimum current value included in the information by the minimum current value included in the current trend change information, and the maximum current value included in the immediately preceding trend change information is the maximum current value included in the current trend change information When one or more of the maximum average current value magnifications divided by .

When all of the second variable amplitude magnification, the second average current value magnification, the minimum current value magnification, and the maximum current value magnification are equal to or less than the second magnification value, the monitoring device may include a loop currently received from the current detection sensor. The trend history information corresponding to the current value is derived as comparison reference information, the trend history information corresponding to the comparison reference information is detected from the cable failure history information stored in advance in the storage unit, and the failure rate corresponding to the detected trend history information By detecting the information, it is possible to recognize a failure probability value for the inspection target.

When all of the second variable amplitude magnification, the second average current value magnification, the minimum current value magnification, and the maximum current value magnification are equal to or less than the second magnification value, the monitoring device may include a loop currently received from the current detection sensor. It is possible to derive the trend history information corresponding to the current value, and to recognize the failure probability value for the inspection target by applying the trend history information derived to the machine-learned prediction model in advance. Here, the prediction model is cable failure history information stored in the storage unit to include the trend history information derived to correspond to the loop current value received from each of the current detection sensors provided in the field and the failure rate information corresponding to each of the trend history information. After machine learning using , as learning data, it may be set to derive a failure probability value according to trend history information derived corresponding to a loop current value currently received from the current detection sensor.

The monitoring device may transmit an emergency alert or a general alert for the inspection target to the manager terminal based on whether the failure probability value is equal to or greater than a predetermined threshold value.

When the inspection result information on the inspection target is received from the manager terminal, the monitoring device may update the cable failure history information with reference to normal state confirmation or failure confirmation information included in the inspection result information.

According to another aspect of the present invention, as a computer program stored in a computer-readable medium to perform a cable failure prediction method through trend analysis, the computer program causes the computer to perform the following steps, , using the loop current value measured and received by the current detection sensor mounted on the loop consisting of a pair of cables that electrically connect the + and - terminals of the facilities included in the electrical and instrumentation control system in the form of a loop circuit, generating, respectively, trend reference information of the loop current value during a specified calibration period, current trend change information for a current unit period, and previous trend change information for a previous unit period; If the loop current value currently received from the current detection sensor is equal to or greater than the reference maximum current value, a first variation multiplier obtained by dividing the current variation value included in the current trend change information by the reference current variation value, and included in the current trend change information If the first average current value magnification obtained by dividing the average current value divided by the reference average current value are all less than or equal to the first preset magnification value, transmitting a general alarm to the manager terminal using the loop as an inspection target; When at least one of the first variation magnification and the first average current value magnification is greater than the first magnification value, the current variation value included in the immediately preceding trend change information is divided by the current variation width value included in the current trend change information A second variation width magnification, a second average current value magnification obtained by dividing the average current value included in the immediately preceding trend change information by the average current value included in the current trend change information, and the minimum current value included in the immediately preceding trend change information The minimum current value multiplier obtained by dividing by the minimum current value included in the current trend change information, and the maximum average current value multiplier obtained by dividing the maximum current value included in the immediately preceding trend change information by the maximum current value included in the current trend change information Comprising the step of transmitting an emergency alert for the inspection target to the manager terminal when at least one of them is greater than or equal to a second magnification value specified in advance, wherein the trend reference information is a reference for a change in the loop current value during the calibration period It includes information on an average current value, a reference minimum current value, a reference maximum current value, a reference current variation value, and a reference maximum amount change period value, and the size of the current unit period and the immediately preceding unit period is the reference maximum amount change period value. There is provided a computer program stored in a computer-readable medium, characterized in that it is designated equal in size. .

In the cable failure prediction method, if all of the second variation width magnification, the second average current value magnification, the minimum current value magnification and the maximum current value magnification are less than or equal to the second magnification value, the current received from the current detection sensor Trend history information corresponding to the loop current value is derived as comparison reference information, and trend history information corresponding to the comparison reference information is detected from the cable failure history information stored in advance in the storage unit, and then corresponding to the detected trend history information The method may further include detecting failure rate information to recognize a failure probability value for the inspection target.

In the cable failure prediction method, if all of the second variation width magnification, the second average current value magnification, the minimum current value magnification and the maximum current value magnification are less than or equal to the second magnification value, the current received from the current detection sensor The method may further include deriving trend history information corresponding to the loop current value, and recognizing a failure probability value for the inspection target by applying the trend history information derived in advance to a machine-learned prediction model. Here, the prediction model is cable failure history information stored in the storage unit to include the trend history information derived to correspond to the loop current value received from each of the current detection sensors provided in the field and the failure rate information corresponding to each of the trend history information. After machine learning using , as learning data, it may be set to derive a failure probability value according to trend history information derived corresponding to a loop current value currently received from the current detection sensor.

The cable failure prediction method includes: determining whether the failure probability value is greater than or equal to a predetermined threshold value; and if the failure probability value is less than or equal to the threshold value, transmitting a general alert for the inspection target to the manager terminal, and if the failure probability value is greater than the threshold value, transmitting an emergency alert for the inspection target to the manager terminal may include more.

Other aspects, features and advantages other than those described above will become apparent from the following drawings, claims, and detailed description of the invention.

According to an embodiment of the present invention, it is possible to remotely predict the states (normal operation state, general inspection necessary state, or emergency inspection necessary state) of various cables constituting the electrical and instrumentation control system, so that appropriate measures are taken according to the state of each cable. has the effect of being done quickly.

In addition, based on the loop current value measured by the current detection sensor, the current measured loop current value using the machine-learned prediction model using the cable failure history information corresponding to the cable failure prediction and the inspection result according to the cable failure prediction. The condition of the cable can be predicted quickly and accurately, and the operator's inspection can be requested only when it is selectively necessary, thereby maximizing work convenience.

In addition, the predictive model machine-learned in the electrical and instrumentation control system of a specific site is universally applied to other sites with the same/similar environment (e.g., voltage level, cable information, location, site type, temperature, humidity, etc.) It can be applied, and the applied prediction model is further learned individually at each site, and there is also an effect of being improved to be optimized for the electric and instrumentation control system of the site.

The effects obtainable in the present invention are not limited to the above-mentioned effects, and other effects not mentioned may be clearly understood by those of ordinary skill in the art to which the present invention belongs from the following description. will be.

1 is a block diagram schematically showing a cable failure prediction system according to an embodiment of the present invention.
2 is a block diagram of a monitoring device according to an embodiment of the present invention.
3 is a diagram illustrating a calibration period, a current unit period, and an immediately preceding unit period for cable failure prediction according to an embodiment of the present invention.
4 and 5 are flowcharts illustrating a cable failure prediction method according to an embodiment of the present invention.

Since the present invention can have various changes and can have various embodiments, specific embodiments are illustrated in the drawings and described in detail in the detailed description. However, this is not intended to limit the present invention to specific embodiments, and it should be understood to include all modifications, equivalents, and substitutes included in the spirit and scope of the present invention.

1 is a block diagram schematically showing a cable failure prediction system according to an embodiment of the present invention, FIG. 2 is a block diagram of a monitoring device according to an embodiment of the present invention, and FIG. 3 is an embodiment of the present invention It is a diagram illustrating a calibration period, a current unit period, and an immediately preceding unit period for cable failure prediction according to an embodiment.

Referring to FIG. 1 , the cable failure prediction system may include a current detection sensor 110 and a monitoring device 120 .

The current detection sensor 110 is mounted in a loop consisting of a pair of cables electrically connecting the + and - terminals of the DCS panel in a loop circuit shape, and monitors the measurement information of the loop current value by sensing the loop (120) send to

In FIG. 1, a DCS panel included in an electrical and instrumentation control system is illustrated as an example for convenience of description and described based on this, but it is of course not limited thereto. That is, not only the DCS panel, but also all facilities in the electrical and instrumentation control system where cables such as power, communication, and control are connected to form a loop circuit for cable failure prediction can be expanded to the target. Of course.

Here, the current sensing sensor 110 is connected to a DCS panel such as a distribution box provided in various field types, such as power plants, solar power generation facilities, and general industrial sites, and all types of cables (eg, signal cables, power cables, etc.).

Since the current detection sensor 110 is mounted to monitor the loop, the loop current value is theoretically measured as 0 mA, but when the cable is aged and is affected by external noise or a short circuit or ground fault occurs, the loop current value is higher than 0 mA. measured in large values. In general, when a current of 10mA or more is generated, it can be recognized as a ground fault condition.

However, the loop current value measured by the current detection sensor 110 mounted to monitor the loop is actually a noise, cable resistance, magnetic field/electromagnetic wave in the air, residual current of the power supply, etc. even when the cable is not old. For various reasons, it may be measured with a low current value other than 0mA.

In order to set a reference value according to the environment in which the current detection sensor 110 is mounted in consideration of the discrepancy between the abnormal state and the actual state as described above, the monitoring device 120 according to the present embodiment performs a calibration process as described below. There is a characteristic.

The monitoring device 120 includes a measurement information receiving unit 211 , a trend analyzing unit 213 , a failure prediction unit 215 , an alarm unit 217 , an inspection result receiving unit 219 , a history information updating unit 221 , and a storage unit. (223).

The monitoring device 120 uses the measurement information of the loop current value received from the current detection sensor 110 to determine the trend reference information of the loop current value during the preset calibration period, the current trend change information for the current unit period, and the immediately preceding unit. Generates each of the previous trend change information for the period, and applies the trend reference information, the current trend change information, and the previous trend change information to a predetermined cable failure prediction method to determine whether the loop equipped with the current detection sensor 110 is in a normal state Or, a general alert or an emergency alert targeting a loop may be transmitted to the manager terminal 130 .

In this case, the cable failure history information stored in the storage unit 223 may be referred to, or a cable failure prediction method may be designated in advance so that a prediction model machine-learned in advance using the cable failure history information is used. In addition, the cable failure history information may be updated using the inspection result information and stored in the storage unit 223 .

For reference, in the present specification, cable failure may be interpreted to include all abnormal conditions of the cable, such as cable breakage, cable sheath damage, cable aging, and ground fault, for example.

Specifically, the measurement information receiving unit 211 is connected to the current detection sensor 110 through a wired or wireless communication network, and receives measurement information of the loop current value from the current detection sensor 110 mounted on a pair of cables. .

The electric and instrumentation control system installed in a field such as a power plant may be equipped with a current sensing sensor 110 to individually monitor each of the loops (ie, a pair of cables), and the measurement information receiving unit 211 is a current sensing sensor. Measurement information of the loop current value may be received from each of the 110 .

At this time, the current detection sensor 110 may transmit uniquely predetermined identification information together with the measurement information of the loop current, and the measurement information receiving unit 211 recognizes the installation position of the current detection sensor 110 using the identification information. can do.

However, in the following, for convenience of explanation and understanding, it is a case in which measurement information of a loop current value is received from any one of the current detection sensors 110 installed in an electrical and measurement control system in the field. limitedly described.

The trend analyzer 213 generates trend reference information of the loop current value by using the measurement information of the loop current value continuously received during the preset calibration period from the current detection sensor 110 .

The calibration period may be preset as a period value (eg, 7 days, 300 hours, etc.) for generating trend reference information of the loop current value.

By using the measurement information of the loop current value continuously received during the calibration period from the current detection sensor 110, the trend analyzer 213 as the trend reference information, for example, a reference average current value, a reference minimum current value, a reference The maximum current value, the reference current fluctuation range value, and the reference maximum amount change period value may be calculated. The calculated trend reference information is stored in the storage unit 223 to correspond to the current detection sensor 110, and may be used as reference information for predicting cable failure.

Here, the reference average current value is an average value (eg, 1 mA) of loop current values received during the calibration period. The reference average current value may be an average value for a period before calibration or an average value for a predetermined unit period (eg, an average value for one hour).

The reference minimum current value is a minimum value (eg, 0.5 mA) among loop current values received during the calibration period, and the reference maximum current value is a maximum value (eg, 1.4 mA) among loop current values received during the calibration period. )to be.

The reference current fluctuation value is the difference between the maximum value and the minimum value of the loop current value collected during the calibration period (eg, 1.4mA - 0.5mA = 0.9mA).

The reference maximum amount change period value is the period value between the time when the maximum value of the loop current value occurred and the time when the minimum value of the loop current value occurred (for example, the minimum value of the loop current value occurred within 23 hours after the start of calibration, 120 If the maximum value of the loop current value is generated in time, it is calculated as 97 hours).

The trend analysis unit 213 sets a unit period for predicting cable failure with the time length of the reference maximum amount change period value calculated as the trend reference information, and calculates trend change information for each unit period. The trend change information calculated for each unit period may include, for example, information on an average current value, a minimum current value, a maximum current value, and a current variation width value.

As shown in FIG. 3 , a time interval from a time point earlier by the size of a unit period from the present to the present is referred to as a current unit period, and a time interval from the most advanced time point of the current unit period to a time point further ahead by the size of the unit period. is referred to as the immediately preceding unit period.

The failure prediction unit 215 calculates the loop current value measured and received from the current detection sensor 110, the trend reference information calculated for the calibration period, the current trend change information calculated for the current unit period, and the previous unit period. It is possible to predict whether a cable constituting a loop in which the current detection sensor 110 is mounted has failed by using the previous trend change information. The trend reference information, the current trend change information, and the previous trend change information may be stored in the storage unit 223 to correspond to the current detection sensor 110 .

As an example, the failure prediction unit 215 determines that the loop current value measured and received from the current detection sensor 110 is less than the reference maximum current value included in the trend reference information, but is greater than the reference average current value, according to the judgment of the administrator. It can be determined as a target of a general alarm that allows that loop to be checked.

As another example, if the loop current value measured and received from the current detection sensor 110 is greater than or equal to the reference maximum current value included in the trend reference information, the failure prediction unit 215 may include a first current variation value included in the current trend change information. The first average current value multiplier obtained by dividing by the reference current fluctuation value included in the trend reference information, and the first average current value obtained by dividing the first average current value included in the current trend change information by the reference average current value included in the trend reference information It may be determined whether at least one of the predetermined margin ratios is equal to or greater than the first magnification value P1 in consideration. The first magnification value is a value greater than 1 and may be preset, for example, 1.1 or 1.2.

If both the first variation magnification and the first average current value magnification are not equal to or greater than the first magnification value, the failure prediction unit 215 may determine a general alarm target for checking the corresponding loop according to the judgment of the administrator.

However, if at least one of the first variation magnification and the first average current value magnification is equal to or greater than the first magnification value, the failure prediction unit 215 sets the second current variation value included in the previous trend change information to the current trend change information. a second variation scale multiplier divided by the included first current variation value, and a second average current value multiplier obtained by dividing the second average current value included in the previous trend change information by the first average current value included in the current trend change information; The minimum current value multiplier obtained by dividing the second minimum current value included in the previous trend change information by the first minimum current value included in the current trend change information, and the second maximum current value included in the previous trend change information are calculated as the current trend change information It may be determined whether at least one of the magnifications of the maximum average current value divided by the first maximum current value included in the . The second magnification value may be set equal to the first magnification value, or may be set relatively larger than the first magnification value.

If at least one of the second variation scale magnification, the second average current value magnification, the minimum current value magnification, and the maximum current value magnification is greater than or equal to the second magnification value, the failure predicting unit 215 may include a corresponding current detection sensor 110 installed. Because a loop is likely to fail, it can be determined as an emergency alert target for an administrator to check that loop immediately.

However, if all of the second variation magnification magnification, the second average current value magnification, the minimum current value magnification and the maximum current value magnification are less than the second magnification value, the failure prediction unit 215 is a loop in which the corresponding current detection sensor 110 is mounted. to determine the notification method as a general alarm or emergency alarm for the inspection target by applying a method of referencing the cable failure history information stored in the storage unit 223 or a method using a machine-learned prediction model as an inspection target can

The cable failure history information stored in the storage unit 223 includes a first variation magnification calculated using a loop current value continuously received from current detection sensors 110 provided in an electrical and measurement control system in the field; The first average current value magnification, the second variation width magnification, the second average current value magnification, the minimum current value magnification, and the maximum current value magnification, trend history information, and failure rate information corresponding to each of the trend history information are stored.

The cable failure history information is, for example, {first variable amplitude multiplier (1.1), first average current value multiplier (1.15), second variable amplitude multiplier (1.3), second average current value multiplier (1.1), minimum current value multiplier (1.0), a maximum current value magnification (1.2), a failure rate (35%)}, etc., may be stored in the storage unit 223 .

Of course, as trend history information included in the cable failure history information, trend reference information, current trend change information, and previous trend change information may be stored.

In addition, cable failure history information can be stored and managed in the storage unit 223 by being divided based on the environmental information (eg, above ground/underground, temperature, humidity, etc.) of the DCS panel equipped with the current detection sensor 110 . have.

For the inspection target determined by the failure prediction unit 215, the alarm unit 217 sends a notification of a general alarm or an emergency alarm to the manager terminal 130, and the manager When the inspection result information is transmitted by checking the installed loop), the history information update unit 221 may calculate the failure rate information by using the normal state confirmation or failure confirmation information included in the received inspection result information.

For example, the history information update unit 221 may calculate a failure rate by dividing the number of failure confirmations by the number of notifications, for example. For example, if the number of failure checks is 40 for 100 times of notification, the failure rate may be calculated as 40%. The failure rate information may be updated and stored to correspond to the corresponding trend history information whenever inspection result information is received from the manager terminal 130 .

As a cable failure prediction method that refers to cable failure history information stored in advance, the failure prediction unit 215 derives trend history information according to the loop current value continuously received from any current detection sensor 110 as comparison reference information and , the current detection sensor 110 by detecting the trend history information that is the same as the comparison reference information or matching within a predetermined error range among the cable failure history information stored in advance in the storage unit 223, and detecting the corresponding failure rate It is possible to recognize the failure probability value for the inspection target, which is the installed loop. Here, the failure probability value may be recognized to be the same as the detected failure rate.

Alternatively, as a cable failure prediction method using a machine-learned prediction model in advance, the failure prediction unit 215 derives trend history information corresponding to a loop current value continuously received from an arbitrary current detection sensor 110, and a prediction model It is also possible to recognize the failure probability value for the inspection target by applying the trend history information to the

For example, the prediction model is derived to correspond to the loop current value continuously received from the current detection sensor 110 provided in the electric and instrumentation control system of the field, and each of the trend history information and the trend history information stored in the storage unit 223 . Artificial to generate failure rate information according to the trend history information corresponding to the loop current value currently received from the arbitrary current detection sensor 110 and machine-learned using the cable failure history information including the failure rate information corresponding to the learning data. It can be configured as an intelligence-based built-in AI module.

As for the prediction model, whenever a notification of a general alarm or an emergency alarm is newly sent to the manager terminal 130 , inspection result information is received from the manager terminal 130 , and cable failure history information is updated and stored in the storage unit 223 . It can be set up for real-time machine learning.

Predictive models machine-learned in electrical and instrumentation control systems at a specific site are similar to the same environment (e.g., voltage level, cable information (e.g., solid/stranded, cable diameter), location (e.g., indoor/outdoor) , underground/ground), site type (eg, power plant/solar facility, etc.), temperature, humidity, etc.) can be installed and applied universally, and the applied prediction model is the cable generated at each site By further learning using the failure history information, it can be improved to be optimized for the electrical and instrumentation control system of the site.

Predictive models are generated in other sites in the same environment, for example, using a generative model such as a few-shot learning method that can learn with only a small amount of training data or an adversarial generative network (GAN) with a small amount of training data. It can be implemented to learn by using a method of resolving the lack of learning data by increasing the number of similar data.

Predictive models include, for example, Fully Convolutional Neural Networks, Convolutional Neural Networks, Recurrent Neural Networks, Restricted Boltzmann Machines, and Deep Belief Neural Networks. Network), etc., may be created with one or more deep learning-based models. Of course, it may be implemented as a machine learning technique other than the deep learning technique, or it may be created as a hybrid model in which the deep learning technique and the machine learning technique are combined.

In addition, a method of learning the predictive model may be various, such as supervised learning, unsupervised learning, reinforcement learning, and the like.

If the failure rate of the failure rate information to be derived by the above-described cable failure prediction method for the loop designated as the inspection target is greater than or equal to a predetermined threshold, the failure prediction unit 215 sends a notification of a general alarm or an emergency alarm to the manager terminal 130. can A threshold value for separately notifying a general alert and an emergency alert may be preset, for example, 20% or the like. Of course, it is natural that the threshold value may be designated as 0%, so that an emergency alert may be unconditionally notified when the failure rate is not 0%.

The alarm unit 217 transmits a notification of a general alarm or an emergency alarm regarding a loop determined as an inspection target by the failure prediction unit 215 to the manager terminal 130 connected through a wired or wireless communication network.

The notification transmitted by the alarm unit 217 to the manager terminal 130 may include, for example, classification information of a general alarm or an emergency alarm, installation location information of a loop requiring confirmation, and the like. Here, the manager terminal 130 may be, for example, a digital processing device equipped with a communication function, such as a smart phone, a tablet PC, or a laptop computer.

The inspection result receiving unit 219 receives inspection result information input by the administrator by checking the loop on which the inspection target current detection sensor 110 is mounted, from the administrator terminal 130 .

The history information update unit 221 uses the normal state confirmation or failure confirmation information included in the inspection result information received from the manager terminal 130 , the trend history information and failure rate information of the cable failure history information stored in the storage unit 223 . One or more of these can be updated.

4 and 5 are flowcharts illustrating a cable failure prediction method according to an embodiment of the present invention.

Referring to FIG. 4 , in step 310 , the monitoring device 120 generates trend reference information of the loop current value by using the measurement information of the loop current received from the current detection sensor 110 during a preset calibration period.

The trend reference information may include, for example, information on a reference average current value, a reference minimum current value, a reference maximum current value, a reference current variation value, and a reference maximum amount change period value for the calibration period.

Although not shown in FIG. 4 , after the calibration period, the monitoring device 120 may use the measurement information of the loop current continuously received from the current detection sensor 110 to calculate trend change information for each designated unit period.

Here, the size of the unit period for cable failure prediction may be set as a reference maximum amount change period value included in the trend reference information, and the trend change information for each unit period is, for example, an average current value, a minimum current value, a maximum current It may include information about the value and the current variation value.

As described above, the monitoring device 120 may generate the current trend change information for the current unit period and the previous trend change information for the immediately preceding unit period, respectively.

In step 315 , the monitoring device 120 determines whether a loop current value measured from the current detection sensor 110 and currently received is equal to or greater than a reference maximum current value included in the trend reference information.

Although not shown, prior to step 315, determining whether the loop current value measured from the current detection sensor 110 and currently received is greater than the ground fault determination value set by the administrator for early recognition of a grounding situation. may be executed.

Here, a plurality of ground fault determination values may be designated. For example, if the loop current value is greater than the first ground fault determination value of 10 mA, proceed to step 350 and select the loop equipped with the current detection sensor 110 in which the loop current value is measured in the manager terminal 130 as an inspection target. An emergency alert can be sent. In addition, when the loop current value is within the range of 10 mA, which is the first value, and 5 mA, which is the second value, it may be set to proceed to step 355 to transmit a general alarm for the corresponding inspection target to the manager terminal 130 . .

If, by the determination of step 315, the currently received loop current value is less than the reference maximum current value, in step 380 (see FIG. 5), the monitoring device 120 determines that the loop current value is the reference average current value included in the trend reference information. Determine whether or not

According to the determination of step 380, if it is less than the reference average current value, the monitoring device 120 determines that it is in a normal state in step 385, and proceeds to step 315 (refer to FIG. 4) again.

However, if it is equal to or greater than the reference average current value, the monitoring device 120 proceeds to step 355 and transmits a general alert for the corresponding inspection target to the manager terminal 130 . The general alarm may be a notification for allowing the administrator to determine whether to check the loop on which the current detection sensor 110 is mounted.

Referring back to step 315, if the measured loop current value is equal to or greater than the reference maximum current value, in step 320, the monitoring device 120 determines the first current variation value included in the current trend change information calculated for the current unit period. The first average current value multiplier obtained by dividing the value of ? by the reference current fluctuation value included in the trend reference information, and the first average current value multiplier obtained by dividing the first average current value included in the current trend change information by the reference average current value included in the trend reference information It is determined whether at least one of the first magnification values P1 is set to a value greater than 1 in consideration of a predetermined margin ratio.

By the determination in step 320, if both the first variation magnification and the first average current value magnification are less than the first magnification value, in step 355, the monitoring device 120 issues a general alarm for the corresponding inspection target to the manager terminal 130 send to The general alarm may be a notification in which whether a corresponding loop is checked or not is determined by an administrator.

However, if it is determined in step 320 that at least one of the first variation magnification and the first average current value magnification is equal to or greater than the first magnification value, in step 325, the monitoring device 120 determines the immediately preceding trend change calculated for the immediately preceding unit period. The second variation multiplier obtained by dividing the second current variation value included in the information by the first current variation value included in the current trend change information, and the second average current value included in the previous trend change information The second average current value multiplier divided by the first average current value, the minimum current value multiplier obtained by dividing the second minimum current value included in the previous trend change information by the first minimum current value included in the current trend change information, and the previous trend Whether at least one of the maximum average current value magnifications obtained by dividing the second maximum current value included in the change information by the first maximum current value included in the current trend change information is greater than or equal to the second magnification value P2 in which the predetermined margin ratio is considered judge The second magnification value may be set equal to the first magnification value, or may be set relatively larger than the first magnification value.

If, by the determination of step 325, at least one of the second variation scale magnification, the second average current value magnification, the minimum current value magnification, and the maximum current value magnification is equal to or greater than the second magnification value, the monitoring device 120 proceeds to step 350 An emergency alert for the corresponding inspection target is transmitted to the manager terminal 130 . The emergency alert may be a notification instructing the manager to immediately check the loop because the failure of the loop to which the current detection sensor 110 is mounted is high.

However, by the determination of step 325, if all of the second variation magnification magnification, the second average current value magnification, the minimum current value magnification and the maximum current value magnification are less than the second magnification value, in step 330, the monitoring device 120 performs the check It is determined whether a predictive model trained over a reference value (minimum required number of times) or more is provided to determine whether the target is a general warning target or an emergency warning target.

If the prediction model is provided, in step 335, the monitoring device 120 calculates trend history information corresponding to the loop current value continuously received from the current detection sensor 110 corresponding to the inspection target, and machine learning in advance By applying the trend history information to the predicted model, a failure probability value for the loop equipped with the corresponding current detection sensor 110 is derived.

Here, the predictive model, for example, is derived to correspond to the loop current value continuously received from the current detection sensor 110 provided in the electric and instrumentation control system in the field, and learns the cable failure history information stored in the storage unit 223 . It can be configured with an artificial intelligence-based built-in AI module to be machine-learned as data, and to generate failure rate information according to trend history information corresponding to a loop current value currently received from an arbitrary current detection sensor 110 .

The cable failure history information stored in the storage unit 223 includes the first variation magnification, the first calculated using the loop current value received from the current detection sensors 110 provided in the electric and measurement control system of the field. The average current value magnification, the second variation width magnification, the second average current value magnification, the minimum current value magnification, and the maximum current value magnification may include trend history information and failure rate information corresponding to each trend history information.

The cable failure history information is, for example, {first variable amplitude multiplier (1.1), first average current value multiplier (1.15), second variable amplitude multiplier (1.3), second average current value multiplier (1.1), minimum current value multiplier (1.0), a maximum current value magnification (1.2), a failure rate (35%)}, etc., may be stored in the storage unit 223 .

The current trend information relates to the first variation magnification, the first average current value magnification, the second variation magnification magnification, the second average current value magnification, the minimum current value magnification, and the maximum current value magnification calculated according to the currently received loop current value. may contain information. The trend history information and the current trend information of the cable failure history information may further include the aforementioned trend reference information.

However, by the determination of step 330, if there is no predictive model learned more than the reference value, in step 340, the monitoring device 120 refers to the cable failure history information stored in advance in the storage unit 223 to refer to the loop current value Recognizes a failure probability value for the loop equipped with the current detection sensor 110 measuring .

For example, the monitoring device 120 derives trend history information according to the loop current value continuously received from the corresponding current detection sensor 110 as comparison reference information, and cable failure history information stored in advance in the storage unit 223 . Among them, it is possible to detect trend history information that is the same as the comparison reference information or within a predetermined error range, and to recognize the corresponding failure rate information as a failure probability value for the loop in which the corresponding current detection sensor 110 is mounted.

If the trend history information that is the same as the comparison reference information or coincides within the error range is not yet stored in the storage unit 223, it may be set to apply the failure rate information of the trend history information that is relatively similar among the stored trend history information. have.

In step 345 , the monitoring device 120 determines whether a failure probability value calculated using a predictive model or previously stored cable failure history information is equal to or greater than a predetermined threshold value.

If a failure probability value greater than or equal to the threshold value is calculated, in step 350 , the monitoring device 120 transmits an emergency alert targeting the loop of the current detection sensor 110 in which the corresponding loop current value is measured to the manager terminal 130 . do.

However, if the failure probability value below the threshold value is calculated, in step 355, the monitoring device 120 issues a general alarm to the inspection target of the loop of the current detection sensor 110 in which the corresponding loop current value is measured to the manager terminal 130. send to Of course, it is natural that the threshold value may be designated as 0%, so that an emergency alert may be unconditionally notified when the failure rate is not 0%.

In step 360 , the monitoring device 120 receives inspection result information for an inspection target designated in an emergency alert or a general alert from the manager terminal 130 .

In step 365, the monitoring device 120 updates the cable failure history information with reference to the normal state confirmation or failure confirmation information included in the inspection result information. At this time, the cable failure history information may be newly created in the storage unit 223 so that the trend history information and the failure rate information are newly added, and the failure rate information for the trend history information already stored in the storage unit 223 is changed to change the cable failure history. Information may be updated.

In addition, when cable failure history information is newly created or updated in the storage unit 223 , the predictive model may be set for real-time machine learning using this.

It goes without saying that the above-described method for predicting cable failure may be implemented as a software program, an application, etc. embedded in a digital processing device, and may be performed as an automated procedure according to a time-series sequence. Codes and code segments constituting the program or the like can be easily inferred by a computer programmer in the art. In addition, the program is stored in a computer readable medium (computer readable media), read and executed by the computer to implement the method.

Although the above has been described with reference to the embodiments of the present invention, those of ordinary skill in the art can variously modify the present invention within the scope without departing from the spirit and scope of the present invention described in the claims below. and may be changed.

110: current detection sensor 120: monitoring device
130: manager terminal 211: measurement information receiving unit
213: trend analysis unit 215: failure prediction unit
217: alarm unit 219: inspection result receiving unit
221: history information update unit 223: storage unit

Claims (13)

Measurement information of the loop current value that is mounted on a loop consisting of a pair of cables that electrically connect the + and - terminals of the facility in the form of a loop circuit for a facility included in the electrical and measurement control system, and sensed the loop a current sensing sensor that transmits; and
Using the measurement information of the loop current value received from the current detection sensor, the trend reference information of the loop current value during the preset calibration period, the current trend change information for the current unit period, and the immediately preceding trend change information for the previous unit period By applying the trend reference information, the current trend change information, and the previous trend change information to each generated and predetermined cable failure prediction method, determine whether the loop equipped with the current detection sensor is in a normal state, or check the loop Including a monitoring device for transmitting a target general alarm or emergency alarm to an administrator terminal,
The trend reference information includes information about a reference average current value, a reference minimum current value, a reference maximum current value, a reference current fluctuation range value, and a reference maximum amount change period value with respect to the fluctuation of the loop current value during the calibration period,
The current trend change information and the immediately preceding trend change information include information about an average current value, a minimum current value, a maximum current value, and a current fluctuation width value related to a change in the loop current value during each unit period,
The monitoring device determines that the loop is in a normal state when the loop current value currently received from the current detection sensor is smaller than the reference maximum current value and is less than or equal to the reference average current value,
When the loop current value currently received from the current detection sensor is equal to or greater than the reference maximum current value, the monitoring device includes a first variation multiplier obtained by dividing the current variation value included in the current trend change information by the reference current variation value and , When the first average current value magnification obtained by dividing the average current value included in the current trend change information by the reference average current value is all less than or equal to a predetermined first magnification value, a general warning for the inspection target is transmitted, characterized in that Cable failure prediction system through trend analysis.
According to claim 1,
The cable failure prediction system through trend analysis, characterized in that the current unit period and the size of the immediately preceding unit period are designated to be the same as the size of the reference maximum amount change period value.
delete delete According to claim 1,
When at least one of the first variation magnification and the first average current value magnification is greater than the first magnification value, the monitoring device,
A second variation magnification by dividing the current variation value included in the immediately preceding trend change information by the current variation value included in the current trend change information and the average current value included in the immediately preceding trend change information are included in the current trend change information A second average current value multiplier divided by the average current value obtained, a minimum current value multiplier obtained by dividing the minimum current value included in the immediately preceding trend change information by the minimum current value included in the current trend change information, and the previous trend change information When one or more of the maximum average current value magnifications obtained by dividing the maximum current value included in the current trend change by the maximum current value included in the current trend change information is greater than or equal to a predetermined second magnification value, an emergency alert for the inspection target is transmitted Cable failure prediction system through trend analysis.
6. The method of claim 5,
When all of the second variable amplitude magnification, the second average current value magnification, the minimum current value magnification, and the maximum current value magnification are equal to or less than the second magnification value, the monitoring device,
Deriving trend history information corresponding to the loop current value currently received from the current detection sensor as comparison reference information, detecting trend history information corresponding to the comparison reference information from the cable failure history information stored in advance in the storage unit, and Cable failure prediction system through trend analysis, characterized in that the failure rate information corresponding to the detected trend history information is detected to recognize a failure probability value for the inspection target.
6. The method of claim 5,
When all of the second variable amplitude magnification, the second average current value magnification, the minimum current value magnification, and the maximum current value magnification are equal to or less than the second magnification value, the monitoring device,
Deriving trend history information corresponding to the loop current value currently received from the current detection sensor, and applying the trend history information derived to the machine-learned prediction model in advance to recognize the failure probability value for the inspection target,
The prediction model learns the cable failure history information stored in the storage unit to include the trend history information derived to correspond to the loop current value received from each of the current detection sensors provided in the field and the failure rate information corresponding to each of the trend history information. After machine learning using data, the cable failure prediction system through trend analysis, characterized in that it is set to derive a failure probability value according to trend history information derived corresponding to the loop current value currently received from the current detection sensor.
8. The method according to claim 6 or 7,
The monitoring device transmits an emergency alert or a general alert for the inspection target to the manager terminal based on whether the failure probability value is greater than or equal to a predetermined threshold.
9. The method of claim 8,
When the inspection result information on the inspection target is received from the manager terminal, the monitoring device updates the cable failure history information with reference to the normal state confirmation or failure confirmation information included in the inspection result information Cable failure prediction system through analysis.
A computer program stored in a computer-readable medium to perform a method of predicting cable failure through trend analysis, the computer program causing the computer to perform the following steps, the steps comprising:
Using the loop current value measured and received by the current detection sensor mounted on the loop consisting of a pair of cables that electrically connect the + and - terminals of the facilities included in the electrical and instrumentation control system in the form of a loop circuit, generating, respectively, trend reference information of the loop current value during the calibration period, current trend change information for the current unit period, and last trend change information for the immediately preceding unit period;
If the loop current value currently received from the current detection sensor is equal to or greater than the reference maximum current value, a first variation multiplier obtained by dividing the current variation value included in the current trend change information by the reference current variation value, and included in the current trend change information If the first average current value magnification obtained by dividing the average current value divided by the reference average current value is less than or equal to the first preset magnification value, transmitting a general alarm to the manager terminal using the loop as a check target;
When at least one of the first variation magnification and the first average current value magnification is greater than the first magnification value, the current variation value included in the immediately preceding trend change information is divided by the current variation width value included in the current trend change information A second variation width magnification, a second average current value magnification obtained by dividing the average current value included in the immediately preceding trend change information by the average current value included in the current trend change information, and the minimum current value included in the immediately preceding trend change information The minimum current value multiplier obtained by dividing by the minimum current value included in the current trend change information, and the maximum average current value multiplier obtained by dividing the maximum current value included in the immediately preceding trend change information by the maximum current value included in the current trend change information Comprising the step of transmitting an emergency alert for the inspection target to the manager terminal if one or more of the preset second magnification value or more,
The trend reference information includes information on a reference average current value, a reference minimum current value, a reference maximum current value, a reference current fluctuation range value, and a reference maximum amount change period value regarding the fluctuation of the loop current value during the calibration period,
The computer program stored in the computer-readable medium, characterized in that the size of the current unit period and the immediately preceding unit period are designated to be the same as the size of the reference maximum amount change period value.
11. The method of claim 10,
When all of the second variation magnification, the second average current value magnification, the minimum current value magnification, and the maximum current value magnification are equal to or less than the second magnification value,
After deriving trend history information corresponding to the loop current value currently received from the current detection sensor as comparison reference information, and detecting trend history information corresponding to the comparison reference information from the cable failure history information stored in advance in the storage unit, and detecting failure rate information corresponding to the detected trend history information to recognize a failure probability value for the inspection object.
11. The method of claim 10,
When all of the second variation magnification, the second average current value magnification, the minimum current value magnification, and the maximum current value magnification are equal to or less than the second magnification value,
Deriving trend history information corresponding to the loop current value currently received from the current detection sensor, and applying the trend history information derived to a machine-learned prediction model in advance, further comprising the step of recognizing a failure probability value for the inspection target but,
The prediction model learns the cable failure history information stored in the storage unit to include the trend history information derived to correspond to the loop current value received from each of the current detection sensors provided in the field and the failure rate information corresponding to each of the trend history information. A computer program stored in a computer-readable medium, characterized in that it is set to derive a failure probability value according to trend history information derived corresponding to the loop current value currently received from the current detection sensor after machine learning as data .
13. The method of claim 11 or 12,
determining whether the failure probability value is greater than or equal to a predetermined threshold value; and
If the failure probability value is less than or equal to the threshold value, transmitting a general alert for the inspection target to the manager terminal, and if the failure probability value is greater than the threshold value, transmitting an emergency alert for the inspection target to the manager terminal. A computer program stored on a computer-readable medium comprising:

Images