KR102570528B1 - Smart switchboard fusing heterogeneous sensor data - Google Patents

Abstract

A smart switchboard that distributes power provided through incoming lines to loads according to an embodiment of the present invention comprises: a power quality monitoring unit that detects the incoming lines and monitors the quality of the power; a facility status sensor unit that detects the status of facilities included in the smart switchboard through at least one sensor; and a smart recognition module that determines the status of the smart switchboard by fusing sensor signals provided from the power quality monitoring unit and the facility status sensor unit. The smart recognition module normalizes the sensor signals into a common format, generates event signals by applying weights to the sensor signals normalized to the common format, fuses a first event signal and a second event signal among the event signals to generate a first fusion event signal, fuses a third event signal and a fourth event signal among the event signals to generate a second fusion event signal, and fuses the first fusion event signal and the second fusion event signal to generate a third fusion event signal. According to the above-described embodiment of the present invention, high-quality situation information can be obtained through horizontal fusion processing and vertical layering processing of heterogeneous multiple situation data detected in the switchboard.

Description

Smart switchboard that fuses heterogeneous sensor data {SMART SWITCHBOARD FUSING HETEROGENEOUS SENSOR DATA}

The present invention relates to a switchboard system, and more particularly, to a smart switchboard that obtains high-quality situation information by fusing and hierarchically processing data from different types of sensors sensed within a switchboard.

A switchboard is a facility that receives power and supplies power to load facilities installed in each power consumer. The switchboard can monitor, control, and protect the power system, and can accommodate a switch or various electrical devices for this purpose. Most electrical equipment inevitably develops aging or fatigue over time. Over time, fatigue builds up, and aging electrical equipment can pose a number of serious hazards, including overheating and electric shock from ground faults. In particular, a switchboard is a place where a common electricity supply network such as an overhead line and facilities of a power receiving location are connected, and it is necessary to monitor the state in particular for safety.

Recently, according to the development of artificial intelligence technology, research on intelligent monitoring of electrical equipment is also being actively conducted. In addition, smart switchboards equipped with sensors and networks for intelligent monitoring of switchboards are being actively researched. In a smart switchboard, sensors and control devices can be connected in real time through a communication network such as the Internet of Things.

The goal of situational awareness detected by a plurality of sensors in a smart switchboard is to prevent fatal risks such as fire and electric shock by detecting fatigue and failure of electrical equipment in advance. However, the detected situations are only fragmentary situations, and in order to accurately predict failures or abnormal symptoms caused by complex factors, it is necessary to interpret and process the various detected situations in a complex and hierarchical manner.

(1) Korea Patent Registration No. 10-1510676 (2015.04.03) (2) Korean Patent Publication No. 0-2019-0130953 (2019.11.25)

An object of the present invention is to provide a smart switchboard system and its operating method for obtaining high-quality context information through horizontal convergence and vertical layering of heterogeneous context data sensed by the switchboard system.

Another object of the present invention is to provide contextual information with high reliability through a hierarchy and convergence of heterogeneous sensor data detected in a switchboard system.

A smart switchboard for distributing power provided through an incoming line to loads according to an embodiment of the present invention is included in the smart switchboard through a power quality monitoring unit that monitors the quality of the power by detecting the incoming line, and at least one sensor. a facility state sensor unit for detecting the state of equipment that is being controlled, and a smart recognition module that determines the state of the smart switchboard by fusing sensor signals provided from the power quality monitoring unit and the facility state sensor unit, wherein the smart recognition module includes: The module normalizes the sensor signals to a common format, generates event signals by applying weights to the sensor signals normalized to the common format, and fuses a first event signal and a second event signal of the event signals. to generate a first fusion event signal, fuse a third event signal and a fourth event signal among the event signals to generate a second fusion event signal, and generate the first fusion event signal and the second fusion event signal is fused to generate 3 fusion event signals.

In this embodiment, the first event signal is generated from at least one of sensor signals provided from the power quality monitoring unit, and the second event signal is generated from at least one of sensor signals provided from the facility state sensor unit. is created

In this embodiment, the smart recognition module applies a format conversion unit that performs a normalization operation for converting the format of the sensor signals into the common format, and a weight of each of the normalized sensor signals output from the format conversion unit. and an event fusion processing unit that performs a fusion operation on the event signals.

In this embodiment, the event convergence processing unit performs the convergence operation by applying a machine learning or artificial intelligence algorithm to the event signals.

In this embodiment, when the first event signal corresponds to an overcurrent event and the second event signal includes a temperature rise or carbon monoxide generation, the first fusion event signal is determined to be an insulator curing event, and the third event When the signal includes an arc generation event and the fourth event signal includes vibration or sound, the second fusion event signal is determined to be a facility fatigue increase event.

According to the above-described embodiment of the present invention, high-quality context information can be obtained through horizontal convergence and vertical layering of heterogeneous multiple context data sensed in a switchboard.

1 is a block diagram exemplarily showing a smart switchboard according to an embodiment of the present invention.
2 is a block diagram showing the configuration of a smart recognition module of the present invention by way of example.
3 is a diagram exemplarily illustrating processing operations in a weight application unit and an event convergence processing unit of a smart recognition module according to an embodiment of the present invention.
4 is a flowchart exemplarily showing a method of operating a smart recognition module according to an embodiment of the present invention.
5 is a block diagram briefly showing another embodiment of an event fusion method performed by an event fusion processing unit of the present invention.
FIG. 6 is a diagram showing an example in which the fusion operation of events of FIG. 5 is specifically applied.
7 is a table briefly showing an example of a fusion operation for sensor signals or event signals.
8 is a flowchart exemplarily showing a method of operating a smart recognition module according to another embodiment of the present invention.

Hereinafter, some embodiments of the present invention will be described in detail with reference to exemplary drawings. In adding reference numerals to components of each drawing, the same components may have the same numerals as much as possible, even if they are displayed on different drawings. In addition, in describing the present invention, if it is determined that a detailed description of a related known configuration or function may obscure the gist of the present invention, the detailed description may be omitted.

1 is a block diagram exemplarily showing a smart switchboard according to an embodiment of the present invention. Referring to FIG. 1 , the smart switchboard 100 may stably distribute power of incoming lines R1 , S1 , T1 , and N1 provided from a power source to loads 10 and 20 . To this end, the smart switchboard 100 includes a plurality of breakers 110, 112, and 114, a power quality monitoring unit 120, a facility state sensor unit 130, a smart recognition module 140, and a communication unit 150. can do.

The plurality of circuit breakers 110, 112, and 114 are wire circuit breakers (MCCB), such as a first distribution circuit breaker 110 connected to the incoming lines R1, S1, T1, and N1, and second and third circuit breakers connected to the load side. (112, 114). A circuit breaker (MCCB) is a device that prevents device failure or fire by cutting off power when a circuit malfunction occurs, and serves to block fault currents such as overload and short circuit current. A circuit breaker (MCCB) mainly uses a thermal type that applies the principle of bimetal, or an electronic type that operates using the magnetomotive force of a coil generated in the event of overcurrent or short circuit.

The power quality monitoring unit 120 monitors the quality of the voltage or current of the three-phase line drawn through the first distribution circuit breaker 110 and transmits the result to the smart recognition module 140 . That is, the power quality monitoring unit 120 detects the incoming lines (R1, S1, T1, N1) coming into the smart switchboard 100 to detect leakage current, abnormal current, abnormal voltage, partial discharge, and overcurrent to electrical equipment or loads. , ground faults, harmonics, and phase imbalances can be detected. To this end, the power quality monitoring unit 120 may include at least one of a current detection sensor, a voltage detection sensor, and a partial discharge detection sensor. In addition, the power quality monitoring unit 120 may monitor the plurality of circuit breakers 110 , 112 , and 114 . For example, the power quality monitoring unit 120 may detect an arc occurring inside the plurality of circuit breakers 110 , 112 , and 114 . Unlike shown in the drawing, the power quality monitoring unit 120 may be installed outside the smart switchboard 100 to detect leakage current, abnormal current, abnormal voltage, partial discharge, overcurrent, ground fault, harmonics, phase imbalance, etc. .

The facility state sensor unit 130 senses the state of facilities included in the smart switchboard 100 and provides the sensed state to the smart recognition module 140 . The facility state sensor unit 130 includes a plurality of sensors that sense the state or situation of various electrical facilities included in the smart switchboard 100 . For example, the facility state sensor unit 130 may include one or more vibration sensors that detect vibration of the smart switchboard 100 due to an earthquake or impact. In addition, the facility state sensor unit 130 may include at least one temperature sensor, deterioration sensor, arc detection sensor, sound sensor, gas concentration sensor, etc. for detecting deterioration, fire, or discharge in the facility. In addition, the facility state sensor unit 130 may further include a humidity sensor for measuring humidity or an image sensor for detecting an image. The sensor configuration of the facility state sensor unit 130 is not limited to the above-described technology, and various types of sensor combinations may be used to monitor the status of the smart switchboard 100 .

The smart recognition module 140 processes sensor signals provided from the power quality monitoring unit 120 and the facility state sensor unit 130 to determine the state or situation of the smart switchboard 100 . The smart recognition module 140 may recognize the quality of incoming power and the state of the energization system using a signal provided from the power quality monitoring unit 120 . In addition, the smart recognition module 140 may recognize a situation inside the smart switchboard 100 using sensor signals provided from the facility state sensor unit 130 .

In particular, the smart recognition module 140 may acquire high-quality situation information by fusing sensor data provided from heterogeneous sensors constituting the power quality monitoring unit 120 and the facility state sensor unit 130 . It is not possible to express all complex and diverse situations occurring in the smart switchboard 100 only with sensor data sensed from each of the heterogeneous sensors. Since each of the sensors provides only dedicated sensing data, it is difficult to detect the situation of the smart switchboard 100 in high resolution only with these uniform sensing data. The smart recognition module 140 of the present invention may horizontally converge and process sensing data or event information provided by different types of sensors. In addition, the smart recognition module 140 can generate high-quality context information by hierarchically processing event information generated through convergence.

The communication unit 150 may transmit the situation judgment value transmitted from the smart recognition module 140 to a control server or control device that manages the smart switchboard 100 . For example, the communication unit 150 supports standards such as 3rd Generation Partnership Project (3GPP), Long Term Evolution (LTE), World Interoperability for Microwave Access (WIMAX), Wi-Fi, 3G, 4G, 5G, and 6G. It may include a wired and wireless communication module that supports. Alternatively, the communication unit 150 may use at least one communication method among Ethernet communication, CDMA, RS485 communication, Bluetooth communication, PLC (Power Line Communication) communication, Zigbee communication, or other IoT communication.

Although the convergence of sensing data for monitoring the status of the smart switchboard 100 has been described as an example in FIG. 1 described above, it will be well understood that the present invention is not limited thereto. That is, the horizontal convergence and vertical layering of sensor data according to the present invention can be used as a method of obtaining high-quality context information in various fields of monitoring the state of a specific target using heterogeneous sensors.

2 is a block diagram showing the configuration of a smart recognition module of the present invention by way of example. Referring to FIG. 2 , the smart recognition module 140 may include a format conversion unit 141, a weight application unit 143, and an event fusion processing unit 145.

The format conversion unit 141 receives event signals or sensor signals S ₁ to S _k provided from the power quality monitoring unit 120 and the facility state sensor unit 130 . Here, event signals or sensor signals are collectively referred to as sensor signals S ₁ to S _k for simplicity of description. The sensor signals S ₁ to S _k may be signals in the form of raw data output from each of a plurality of sensors included in the power quality monitoring unit 120 and the facility state sensor unit 130 . Accordingly, each of the sensor signals S ₁ to S _k may be analog or digital signals representing different physical or chemical properties or values. In order to fuse these heterogeneous sensor signals S ₁ to S _k , the format conversion unit 141 performs a format conversion operation or normalization operation on the sensor signals S ₁ to S _k .

The format conversion operation performed by the format conversion unit 141 may be exemplarily performed as follows. Each of the sensor signals S ₁ to S _k representing different physical or chemical characteristics may be represented by Equation 1 below.

The plurality of heterogeneous sensor signals S ₁ to S _k are provided to the format converter 141 as values that change over time. The format conversion unit 141 may normalize the size of each of the sensor signals S ₁ to S _k by dividing the size of each sensor signal by a reference value S _b assigned to each sensor. The reference value (S _b ) for each sensor can be expressed by Equation 2 below.

The size of the reference value S _b for each sensor may be determined according to sensor characteristics and based on accumulated experimental results or actually measured experience values.

For normalization of the sensor signals S ₁ to S _k , the normalized sensor signal N _k divided by the reference value S _b for each sensor may be expressed by Equation 3 below.

The sensor signals S ₁ to S _k processed by the format conversion unit 141 are output as normalized sensor signals N _k . And whether an event of each sensor has occurred can be determined according to the level of the normalized sensor signal N _k . For example, when the normalized sensor signal N _k is 1 or more, it may be determined that an event corresponding to the corresponding sensor has occurred. However, when the normalized sensor signal N _k is less than 1, it may be determined that an event corresponding to the corresponding sensor has not occurred.

The weight application unit 143 applies weights (W ₁ , W ₂ , ..., W _k ) to each of the normalized sensor signals (N _k ) having a unified format. This is because when a certain event occurs, a sensing activity of a sensor is performed to recognize the situation, and values of sensor signals obtained as a result of the sensing activity of the sensor should not be equally applied. The application of the weights (W ₁ , W ₂ , ..., W _k ) may be provided in a predetermined size by a manager, or may be applied variably based on information on the maximum sensor value change rate that increases when a specific event occurs. There will be. The weight application unit 143 outputs the weighted sensor signal (W _k *N _k ) as an event signal (E _k ) and transmits it to the event fusion processor 145 .

The event fusion processing unit 145 fusion-processes the event signals E _k to which the weights W ₁ , W ₂ , ..., W _k are applied, and outputs a high-dimensional situation determination result. In one embodiment, the event convergence processing unit 145 may output a high-dimensional event (HD_Event) by combining the magnitudes of two or more event signals (E _k ). The high-level event (HD_Event) may be determined by linking sensor data on power supply quality and an event on a facility state. For example, when an overcurrent is detected and a temperature rise is detected, it can be determined that fatigue of the equipment and hardening of the wire coating have occurred.

3 is a diagram exemplarily illustrating processing operations in a weight application unit and an event convergence processing unit of a smart recognition module according to an embodiment of the present invention. Referring to FIG. 3 , the weight application unit 143 and the event convergence processing unit 145 may derive a high-order event (HD_Event) called an insulator curing situation by fusing overcurrent detection, carbon monoxide detection, and temperature rise detection results.

First, a plurality of sensor signals S 1 to S _k will be provided to the smart recognition module 140 (see FIG _. 1 ) from the power quality monitoring unit 120 and the facility state sensor unit 130 . For example, overcurrent detection signal (S ₁ ), ground fault detection signal (S ₂ ), arc detection signal (S ₃ ), gas detection signal (S ₄ ), vibration detection signal (S ₅ ), temperature detection signal (S ₆ ), a sound detection signal (S _k ), and the like will be provided. These plurality of sensor signals (S ₁ to S _k ) are overcurrent detection signals (N ₁ ), ground detection signals (N ₂ ), and arc detection signals (which are signals of the same format) by the format conversion unit 141 (not shown). N ₃ ), a gas detection signal (N ₄ ), a vibration detection signal (N ₅ ), a temperature detection signal (N ₆ ), and a sound detection signal (N _k ). A plurality of sensor signals (N ₁ to N _k ) whose formats are converted are provided to the weight application unit 143 and filtered according to their relative importance. For example, the normalized overcurrent detection signal N ₁ , gas detection signal N ₄ , and temperature detection signal N ₆ may be selected as meaningful context information by a weight. The selected situation information is transmitted to the event fusion processing unit 145 as event signals P ₁ , P ₄ , and P ₆ .

The event fusion processor 145 may perform fusion processing of the overcurrent event (E ₁ ), the gas event (E ₄ ), and the temperature rise event (E ₆ ), and output a high-order event as a final result. The event fusion processing unit 145 may process an overcurrent event (E ₁ ), a gas event (E ₄ ), and a temperature rise event (E ₆ ) using artificial intelligence (AI) calculation. The artificial intelligence (AI) operation may be performed by a processor or a machine learning algorithm such as, for example, a deep neural network or a convolutional neural network. Alternatively, the event fusion processing unit 145 may determine a corresponding high-order event (HD_Event) according to a combination of signal levels of the overcurrent event (E ₁ ), the gas event (E ₄ ), and the temperature rise event (E ₆ ).

4 is a flowchart exemplarily showing a method of operating a smart recognition module according to an embodiment of the present invention. Referring to FIG. 4 , the smart recognition module 140 may detect a high-level event (HD_Event) by converging heterogeneous sensor signals.

In step S110, the smart recognition module 140 receives sensor signals (S ₁ to S _k ) provided from the power quality monitoring unit 120 and the facility state sensor unit 130, and defines an event for each sensor. . Here, the sensor signals S ₁ to S _k may be signals in the form of raw data output from the power quality monitoring unit 120 and the facility state sensor unit 130 , respectively.

In step S120, in order to combine these disparate sensor signals (S ₁ to S _k ) and integrate them with each other, the format conversion unit 141 performs a normalization operation on the sensor signals (S ₁ to S _k ). can The heterogeneous sensor signals S ₁ to S _k are provided to the format conversion unit 141 as values that change over time. The format conversion unit 141 normalizes the magnitudes of each of the sensor signals S ₁ to S _k by dividing them by a reference value S _b assigned to each sensor. The size of the reference value S _b for each sensor may be provided as a predetermined value. For normalization of the sensor signals S ₁ to S _k , a normalized sensor signal N _k will be generated by dividing by the reference value S _b for each sensor.

In step S130, the weight application unit 145 (see FIG. 2) applies weights (W ₁ , W ₂ , ..., W _k ) to each of the normalized sensor signals (N _k ) having a unified format. This is because when a certain event occurs, a sensing activity of a sensor is performed to recognize the situation, and values of sensor signals obtained as a result of the sensing activity of the sensor should not be equally applied. The application of the weights (W ₁ , W ₂ , ..., W _k ) may be provided in a predetermined size by a manager, or may be applied variably based on information on the maximum sensor value change rate that increases when a specific event occurs. There will be. The weight application unit 143 outputs weighted event signals E _k and transfers them to the event fusion processor 145 .

In step S140, the event fusion processing unit 145 fuses the event signals E _k . In one embodiment, the event convergence processing unit 145 may determine a high-level event (HD_Event) by combining two or more event signals (E _k ). The high-level event (HD_Event) may be determined by linking sensor data on power supply quality and an event on a facility state. For example, when an overcurrent is detected and a temperature rise is detected, it can be determined that fatigue of the equipment and hardening of the wire coating have occurred.

In step S150, the event convergence processing unit 145 outputs a high-dimensional event (HD_Event) generated according to the result of the convergence processing operation. The event convergence processing unit 145 may transmit a situation corresponding to a high-level event (HD_Event) to a control center or a separately defined target node. Alternatively, the event convergence processing unit 145 may connect to a wired or wireless network after recognizing the occurrence of a situation corresponding to the high-level event (HD_Event) and transmit the event to a user or a control center.

5 is a block diagram briefly showing another embodiment of an event fusion method performed by an event fusion processing unit of the present invention. Referring to FIG. 5 , an event convergence method according to another embodiment may generate a high-level event (HD_Event) by processing event convergence at a plurality of levels.

The event fusion processing unit 145 performs a multi-layered event fusion operation on the weighted event signals E ₁ to E _N transmitted from the weight application unit 143 . First, at least two or more sensor signals (E ₁ to E _N ) are grouped through the first fusion operation. A fusion operation is performed on the grouped sensor signals. For example, sensor signals E ₁ to E ₃ may be selected as one group for fusion. In addition, the sensor signals E ₄ to E ₅ are classified into another group, and the sensor signals E ₆ to E ₉ are classified into another group and applied to the convergence operation. In this way, a plurality of groups are output as primary fusion events (FE ₁₁ , FE ₁₂ , FE ₁₃ , ..., FE _1M ) through the primary fusion operation.

Subsequently, a secondary fusion operation is performed on the primary fusion events FE ₁₁ , FE ₁₂ , FE ₁₃ , ..., FE _1M . The secondary fusion events (FE ₂₁ , FE ₂₂ , FE ₂₃ , …, FE _2K ) are performed through the secondary fusion operation on the primary fusion events (FE ₁₁ , FE ₁₂ , FE ₁₃ , …, FE _1M ). ) is output. In addition, a tertiary fusion operation is performed on the secondary fusion events (FE ₂₁ , FE ₂₂ , FE ₂₃ , ..., FE _2K ), and a high-order event (HD_Event) may be generated as a result.

FIG. 6 is a diagram showing an example in which the fusion operation of events of FIG. 5 is specifically applied. Referring to FIG. 6 , the event convergence processing unit 145 may identify a high-level event (HD_Event) by performing an event convergence operation in two stages.

Sensor signals S ₁ to S ₆ are transmitted to the smart recognition module 140 from the power quality monitoring unit 120 (see FIG. 1 ) and the facility state sensor unit 130 (see FIG. 1 ). Then, the format conversion unit 141 and the weight application unit 143 of the smart recognition module 140 detect an event by applying a format conversion operation and a weight for each sensor.

For example, the overcurrent is detected as an event (E ₁ ) from the sensor signal (S ₁ ), the detection of carbon monoxide gas is detected as an event (E ₄ ) from the sensor signal (S ₄ ), and the sensor signal (S ₆ ), it is assumed that the temperature rise is detected as an event (E ₆ ). In addition, arc generation is detected as an event (E ₃ ) from the sensor signal (S ₃ ), vibration is detected as an event (E ₅ ) from the sensor signal (S ₅ ), and a specific event is detected from the sensor signal (S ₂ ). Assume that sound is detected as an event E ₂ .

When the detection of the basic events (E ₁ to E ₆ ) is completed, a first fusion operation is performed on the detected events (E ₁ to E ₆ ). The first fusion operation is performed in units of two groups. That is, an overcurrent event (E ₁ ), a carbon monoxide gas event (E ₄ ), and a temperature rise event (E ₆ ) may be selected as the first group. An arc generation event (E ₃ ), a vibration detection event (E ₅ ), and a sound detection event (E ₂ ) may be selected as another second group.

The first fusion event FE ₁ may be determined as a result of the primary fusion operation of the first group events _of the overcurrent event (E 1 ), the carbon monoxide gas event (E ₄ ), and the temperature rise event (E ₆ ). For example, the accompanying occurrence of overcurrent detection, generation of carbon monoxide, and temperature rise of a facility may be determined as a first convergence event (FE ₁ ) of insulator hardening. Also, the second fusion event FE ₂ may be determined as a result of the primary fusion operation of second group events of the arc generation event E ₃ , the vibration detection event E ₅ , and the acoustic detection event E ₂ . . That is, generation of arc, vibration, and sound may be determined as a second convergence event (FE ₂ ), which is an increase in facility fatigue. Therefore, as a result of the primary fusion operation, a first fusion event (FE ₁ ) of insulator hardening and a second fusion event (FE ₂ ) of equipment fatigue increase may be derived.

The first convergence event (FE ₁ ) of insulator hardening and the second convergence event (FE ₂ ) of equipment fatigue increase, which are results of the first convergence calculation, are applied to the second convergence calculation. When the first convergence event (FE ₁ ) of insulator hardening and the second convergence event (FE ₂ ) of facility fatigue increase are fused according to the secondary fusion operation, a high-level event (HD_Event) of weakening switchboard safety can be derived.

7 is a table briefly showing an example of a fusion operation for sensor signals or event signals. Referring to FIG. 7 , information obtained from sensor signals or information obtained from a result of a fusion operation is described as an example.

A sensor signal sensed from at least one temperature sensor installed in the smart switchboard 100 indicates a temperature value within the switchboard. If a sudden increase in the sensor signal value sensed by the temperature sensor occurs, this may be determined as an event in which heat is generated in the smart switchboard 100 .

A vibration sensor installed inside or outside the smart switchboard 100 may detect vibration caused by an earthquake or external impact. For example, as the vibration sensor, at least one of a seismic accelerometer generally used to detect an earthquake, a 3-axis gyro sensor, and an acceleration sensor for detecting intensity of vibration may be used. An increase in the vibration sensor value may be determined as a vibration event of the switchboard due to an earthquake or external impact.

The current sensor is composed of a zero current transformer (Zero Current Transition) and its peripheral circuits, and can detect leakage current flowing in a power line such as an incoming line inside the smart switchboard 100. In addition, the current sensor is composed of a shunt resistor and its peripheral circuit, and can detect the occurrence of overcurrent or abnormal current by monitoring the operating current of the power applied from the input terminal to the output terminal of the power terminal through the power line.

When an abnormal current event and a temperature rise event are simultaneously detected by the current sensor and the temperature sensor, the two event signals may be fused by the event fusion processor 145 . As a result of the convergence of the current anomaly event and the temperature rise event, it can be judged that the fatigue of the equipment and the coating hardening factor of the converter exist.

In addition, if the case where the current sensor and the temperature sensor simultaneously detect an abnormal current event and a temperature rise event are repeated N times or more, the event convergence processing unit 145 determines that the fatigue of the equipment and the hardening of the coating of the converter are in progress. can

Additionally, when a specific time T elapses after the current abnormal event and the temperature rise event are repeated N or more times, the event convergence processing unit 145 may determine that the risk of leakage current in the facility has increased.

The event fusion processing unit 145 can identify and generate a high-level event (HD_Event) through such sensor signal or event convergence calculation.

8 is a flowchart exemplarily showing a method of operating a smart recognition module according to another embodiment of the present invention. Referring to FIG. 8 , the smart recognition module 140 may horizontally and hierarchically fuse heterogeneous sensor signals to detect a high-level event (HD_Event).

In step S210, the smart recognition module 140 receives sensor signals (S ₁ to S _k ) provided from the power quality monitoring unit 120 and the facility state sensor unit 130, and defines an event for each sensor. . Here, the sensor signals S ₁ to S _k may be signals in the form of raw data output from the power quality monitoring unit 120 and the facility state sensor unit 130 , respectively.

In step S220, the format conversion unit 141 converts the sensor signals S ₁ to S _k into a common format in order to process these heterogeneous sensor signals S ₁ to S _k with the same operation. That is, the format conversion unit 141 may perform a normalization operation on the sensor signals S ₁ to S _k . The heterogeneous sensor signals S ₁ to S _k are provided to the format conversion unit 141 as values that change over time. For format conversion, the format conversion unit 141 may normalize the size of each of the sensor signals S ₁ to S _k by dividing the size of each sensor signal by a reference value S _b allocated to each sensor. For normalization of the sensor signals S ₁ to S _k , a normalized sensor signal N _k will be generated by dividing by the reference value S _b for each sensor.

In step S230, the weight application unit 145 (see FIG. 2) applies weights (W ₁ , W ₂ ,..., W _k ) to each of the normalized sensor signals (N _k ) having a unified format. This is because when a certain event occurs, a sensing activity of a sensor is performed to recognize the situation, and values of sensor signals obtained as a result of the sensing activity of the sensor should not be equally applied. The application of the weights (W ₁ , W ₂ ,..., W _k ) may be provided in a predetermined size by a manager, or may be applied variably based on information on the maximum sensor value change rate that increases when a specific event occurs. There will be. The weight application unit 143 outputs weighted event data (E _k ) and transfers it to the event fusion processor 145 .

In step S240, the event convergence processing unit 145 processes the event data (E _k ) through the first convergence operation. In an embodiment, the event fusion processing unit 145 may generate a plurality of primary fusion events FE _1i through a primary fusion operation. The first fusion event FE _1k , for example, when an overcurrent is detected and a temperature rise is detected, it may be determined that fatigue of equipment and a hardening factor of a wire coating have occurred.

In step S250 , the event fusion processing unit 145 may execute a secondary fusion operation on the primary fusion events FE _1i . The event fusion processor 145 may generate a plurality of secondary fusion events FE _2j through secondary fusion calculations.

In step S260 , the event fusion processing unit 145 may execute a tertiary fusion operation on the secondary fusion events FE _1j . The event convergence processing unit 145 may generate a plurality of high-dimensional events (HD_Event) through a tertiary convergence operation.

In step S270, the event fusion processing unit 145 outputs a high-order event (HD_Event) generated according to the result of the fusion processing operation on the secondary fusion events (FE _2j ). The event convergence processing unit 145 may transmit a situation corresponding to a high-level event (HD_Event) to a control center or a separately specified target node. Alternatively, the event convergence processing unit 145 may connect to a wired or wireless network after recognizing the occurrence of a situation corresponding to the high-level event (HD_Event) and transmit the event to a user or a control center.

The foregoing are specific embodiments for carrying out the present invention. The present invention will include not only the above-described embodiments, but also embodiments that can be simply or easily changed in design. In addition, the present invention will also include techniques that can be easily modified and practiced using the embodiments. Therefore, the scope of the present invention should not be limited to the above-described embodiments and should not be defined by the following claims as well as those equivalent to the claims of this invention.

Claims (4)

In a smart switchboard that distributes power provided through an incoming line to loads:
a power quality monitoring unit that detects the incoming line and monitors the quality of the power;
a facility state sensor unit that senses states of facilities included in the smart switchboard through at least one sensor; and
Including a smart recognition module for determining the status of the smart switchboard by fusing sensor signals provided from the power quality monitoring unit and the facility state sensor unit,
The smart recognition module:
normalize the sensor signals to a common format;
Generating event signals by applying weights to sensor signals normalized to the common format;
Fusing a first event signal and a second event signal among the event signals to generate a first fusion event signal;
A third event signal and a fourth event signal among the event signals are fused to generate a second fusion event signal, and
Fusing the first fusion event signal and the second fusion event signal to generate 3 fusion event signals;
The smart recognition module:
When the first event signal corresponds to an overcurrent event and the second event signal includes a temperature rise or carbon monoxide generation, determining the first fusion event signal as an insulator curing event;
When the third event signal includes an arc generation event and the fourth event signal includes vibration generation or sound generation, determining the second fusion event signal as a facility fatigue increase event;
A smart switchboard for determining the third convergence event signal as a weakening of switchboard safety. According to claim 1,
The first event signal is generated from at least one of sensor signals provided from the power quality monitoring unit, and the second event signal is generated from at least one of sensor signals provided from the facility state sensor unit. According to claim 1,
The smart recognition module:
a format conversion unit performing a normalization operation to convert the format of the sensor signals into the common format;
a weight application unit for applying a weight to each of the normalized sensor signals output from the format conversion unit and outputting the result as event signals; and
A smart switchboard comprising an event fusion processing unit that performs a fusion operation on the event signals.



delete

Images