JP2008290549A - Abnormality detection system for crossing gate - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To accurately detect abnormality of a crossing gate without receiving influence of the natural environment. <P>SOLUTION: The abnormality detection system for the crossing gate is provided with a detection device 1 for detecting a motor current when an interruption rod 4 driven by a motor 5 is operated; and a processing device 2 for detecting existence of abnormality of the crossing gate based on the current data. The processing device 2 collects the current data when the interruption rod 4 is normally operated, and cuts off the current data in a regular operation section to prepare time series data. The processing device 2 calculates a linear statistic amount and a non-linear statistic amount of the time series data and determines determination reference of existence of abnormality based on the linear statistic amount and the non-linear statistic amount. After determination of the determination reference, the processing device 2 detects existence of abnormality according to the determination reference based on the motor current detected every when the interruption rod 4 is operated. When the abnormality is detected, the processing device 2 informs a management device 1 of the matter. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to an anomaly detection system for detecting anomalies such as railroad crossings, breakage of breakers such as gates, and notifying them.

  A railroad crossing is equipped with a circuit breaker to restrict traffic of people and cars. In the breaker, the breaker is raised and lowered by a motor. The circuit breaker is closed by lowering the barrier in response to the passage of the train.

  However, when the barrier fence is going down, the car may enter the railroad crossing, and the barrier fence may be broken or torn off. Detection of such anomaly such as breakage of the barrier is performed by a train crew member or passerby confirming and reporting.

However, this lacks quick reporting at the time of occurrence of an abnormality, and there is a safety problem. Therefore, in order to automatically detect breakage of the barrier, use of an impact sensor and a weight sensor, use of a camera, use of a sound sensor, use of an optical fiber and an optical sensor are disclosed in Patent Document 1, for example. Are listed.
Japanese Patent Laid-Open No. 2005-145213

  Since the circuit breaker is installed outdoors, when using the above various sensors, it is affected by the natural environment such as rain and dew. That is, these sensors are liable to malfunction and may not be correctly detected when an abnormality occurs. In addition, every time the barrier rod is replaced, it is necessary to adjust the sensor and check the operation. Therefore, the work time becomes long, and the work may not be completed within a limited time.

  In view of the above, an object of the present invention is to provide an abnormality detection system that can accurately detect an abnormality of a circuit breaker without using various sensors and without being affected by the natural environment.

  The present invention is an abnormality detection system for a circuit breaker operated by a motor to limit the passage of a moving body such as a train, a car, and a person outdoors, and a detection device for detecting a motor current when the circuit breaker operates And a processing device for detecting the presence or absence of an abnormality of the circuit breaker based on the detected current data and notifying the abnormality.

  The processing unit creates time-series data by cutting out the current data in the steady operation section from the collection section that collects current data when the circuit breaker is operating normally and the operation section when the circuit breaker is operating A cut-out unit for calculating, a calculation unit for calculating linear statistics and nonlinear statistics of time series data, a criterion determining unit for determining a criterion for presence / absence of abnormality based on the linear statistics and nonlinear statistics, and a criterion And a monitoring unit that detects the presence or absence of an abnormality in accordance with a criterion based on the motor current detected each time the circuit breaker operates after the determination.

  Since the circuit breaker operates normally when the circuit breaker is installed or replaced, the collection unit collects current data during normal operation. Then, the cut-out unit cuts out current data in a steady operation section that is not affected by the inrush current at the start, and the calculation unit calculates linear statistics and nonlinear statistics based on the cut-out time series data . By combining the two types of statistics, the influence of noise due to wind and the like can be eliminated. The reference determination unit determines a determination criterion based on two types of statistics.

  In this way, a determination criterion is determined according to the breaker that actually operates. The monitoring unit determines whether or not there is an abnormality based on the linear statistics and the nonlinear statistics calculated every time the circuit breaker operates according to this criterion. When there is an abnormality, the monitoring unit notifies the occurrence of the abnormality.

  An abnormal circuit breaker occurs when the circuit breaker operates. Therefore, by detecting an abnormality every time the circuit breaker operates, the occurrence of the abnormality can be detected immediately and can be reported quickly.

  Here, when determining the criterion, the criterion determining unit plots the linear statistic and the nonlinear statistic in the multidimensional state space, and uses the normal area as a criterion as an area formed by the plotting. That is, it is normal if each statistic obtained when the circuit breaker operates is inside the normal area, and abnormal if it is outside the normal area.

  The processing device updates the criterion based on the detected motor current when the circuit breaker operates normally. Accurate judgment criteria can be determined according to the characteristics of the circuit breaker by accumulating data obtained when the circuit breaker is normal and monitoring judgment criteria during abnormality monitoring. This increases the accuracy of abnormality detection.

  In addition, a sensitivity coefficient for the determination criterion is set, and the criterion determination unit changes the determination criterion according to the sensitivity coefficient. By adjusting the sensitivity coefficient as necessary, even if the characteristics of the circuit breaker change, it can be easily handled and erroneous detection can be prevented.

  The circuit breaker has a motor-driven liftable barrier rod, and the processing device determines a criterion based on current data when the barrier rod is raised. When an abnormality such as breakage of the barrier rod occurs, the influence of the abnormality on the operation of the barrier rod is greater when the barrier rod is raised than when the barrier rod is lowered. Therefore, a highly accurate determination criterion can be obtained by determining the determination criterion based on the data when the barrier rod is raised.

  According to the present invention, it is possible to determine a determination criterion according to the characteristics of an actually used breaker, so that it is possible to improve the accuracy of abnormality detection and eliminate false alarms. In addition, since an impact sensor and an optical sensor that are easily affected by the natural environment are not used, highly accurate abnormality detection can be performed without being affected by the natural environment.

  FIG. 1 shows an abnormality detection system for a crossing breaker according to this embodiment. The abnormality detection system includes a detection device 1 that detects a motor current when the circuit breaker operates, a processing device 2 that detects whether there is an abnormality in the circuit breaker based on the detected current data, and notifies the abnormality. And a management device 3 that manages the device 2.

  The circuit breaker is provided with a vertically movable barrier 4 that opens and closes the railroad crossing, a motor 5 that raises and lowers the barrier 4, and a drive device 6 that drives and controls the motor 5. The blocking rod 4 is rotatably supported by a speed reduction mechanism such as a gear connected to the motor 5. As the motor 5 is driven, the shut-off rod 4 is lowered to perform a shut-off operation for closing the crossing, and a shut-off rod 4 is lifted and a return operation for opening the crossing is performed. When the approach of the train is detected by the train approach detection device 7, the drive device 6 energizes the motor 5 and lowers the blocking bar 4. When the passage of the train is detected, the motor 5 is energized to raise the barrier 4.

  The detection device 1 is a current detection sensor such as a current transformer or a hall sensor, and is attached to a cable that supplies power from the drive device 6 to the motor 5. The detection device 1 detects a current flowing through the motor 5 when the circuit breaker operates, that is, when the circuit breaker 4 moves up and down.

  The processing device 2 includes a microcomputer having a CPU, a ROM, and a RAM. The detection device 1 is connected to the processing device 2 and outputs the detected current data to the processing device 2. The processing device 2, the detection device 1, and the driving device 6 are housed in an instrument box 8 installed at a railroad crossing.

  The processing device 2 is connected to a communication line such as an optical fiber via the communication interface 9. The communication line is connected to the IP network. A management device 3 is connected to the IP network, and the management device 3 and the processing device 2 can communicate with each other through the Internet. The communication line connected to the network may be a wireless line such as a wireless LAN, and the communication interface 9 performs wireless communication.

  The management device 3 is a host computer installed in the management center. The management device 3 communicates with a plurality of level crossing processing devices 2 to monitor the state of each level crossing. When there is a report of an abnormality from the processing device 2, the management device 3 instructs a measure for the abnormality such as dispatching a worker.

  And in order to detect breaker abnormalities, such as breakage of the breaking rod 4 or the raising and lowering of the breaking rod 4, the processing device 2 is a chaotic time series based on the motor current when the breaking device operates. The normal / abnormal of the circuit breaker is detected based on the statistics of the data. That is, as shown in FIG. 2, the processing device 2 includes a collection unit 10 that collects current data when the circuit breaker is operating normally, and a steady operation period among the operation periods when the circuit breaker operates. A cut-out unit 11 that cuts out current data and generates time-series data; a calculation unit 12 that calculates linear and nonlinear statistics of the time-series data; and whether there is an abnormality based on the linear and nonlinear statistics And a monitoring unit 14 for detecting the presence or absence of an abnormality in accordance with the determination criterion based on the motor current detected each time the circuit breaker operates after the determination criterion is determined.

  Then, the reference determination unit 13 plots the linear statistics and the nonlinear statistics in the multidimensional space, sets the area formed by the plot as a normal area, sets the normal area inside as normal, and sets the normal area outside as abnormal as a determination standard. Decide. When each of the above units operates according to software stored in the ROM, the abnormality detection of the circuit breaker is executed.

  Here, as the elements necessary for determining the determination criterion, there are a section for cutting out current data, a linear statistic, and a non-linear statistic. These elements are selected in advance based on normal / abnormal discrimination ability. Selected intervals, linear statistics and nonlinear statistics are used during actual detection.

  A procedure for selecting each section and each element of each statistic will be described with reference to FIG. In order to select them, an experiment is conducted for the case where the breaker is normal and the case where the breaker is abnormal. First, current data when the circuit breaker operates is collected (S1). As shown in FIG. 4, a plurality of current data during normal operation is collected by performing a shut-off operation and a return operation using a breaker 4 that is not broken. When there is an abnormality in the circuit breaker, the breaking operation 4 is performed using the broken breaker 4. At this time, assuming various abnormalities, the blocking rod 4 having a different length or the blocking rod 4 camouflaged in a bent state is used. Thereby, a plurality of current data when there is an abnormality is collected.

  Next, current data in a predetermined section is cut out from the operation section when the circuit breaker operates (S2). As shown in FIG. 5, when the circuit breaker operates, the motor current changes in time series. The operation section includes a non-operation section, an inrush current section, and a steady operation section. In order to perform normal / abnormal identification, it is necessary to select a section that is most effective for cutting out current data. Therefore, the time series data cut out from the entire operation section of A to E is CASE1, and the time series data cut out from the section of C to E excluding the inrush current section is CASE1STB, from the inrush current peak of B to D to the end of the steady operation. The time series data cut out from this section is CASE2, and the time series data cut out from the steady operation sections C to D is CASE2STB.

  Even in time series data during normal operation, the aspects of each data are greatly different. Also, there is no frequency band that seems to be common noise. Therefore, a frequency filter used for general preprocessing cannot be used for the time series data. Therefore, as preprocessing, known envelope processing using a sigmoid function is performed on the time-series data cut out from each of the above sections (S3).

  The time-series data that has been subjected to the envelope processing is referred to as CASE1_SGENV5, CASE1STB_SGENV5, CASE2_SGENV5, and CASE2STB_SGENV5, respectively. Note that pre-processing is performed as necessary.

  Linear statistics and nonlinear statistics are calculated for the eight types of time series data cut out from each section (S4). The linear statistics are for each time series data. The nonlinear statistic is a trajectory parallel measure obtained based on each time series data.

  The linear statistic is selected from each statistic such as the minimum value, maximum value, average value, median value, standard deviation, and total value of the time series data of the motor current. The nonlinear statistic is selected from each statistic such as the minimum value, maximum value, average value, median value, kurtosis, and chaos information criterion (CIC value) of the orbital parallel measure. The orbit parallel measure is described in, for example, Japanese Patent Application Laid-Open No. 2007-48097 and Japanese Patent Application Laid-Open No. 10-134034. Even if the time series data contains a little noise, it cannot be distinguished at first glance whether it contains noise, but if you apply the orbital parallel measure using chaos theory, the effect of noise due to wind and mischief Can be detected. The chaos information standard is described in, for example, Japanese Patent Laid-Open No. 2000-48097. In addition to the orbital parallel measure, a Lyapunov exponent, fractal dimension, KS entropy, or the like may be used as the nonlinear statistic.

  An evaluation function used for creating a criterion is selected from various statistics or a combination thereof, and normal / abnormal discrimination ability is evaluated (S5). That is, among the various statistics or combinations thereof, the one having the highest normal / abnormal discrimination ability is used as the evaluation function.

  As a result of the evaluation, the combination of the total value of the linear statistics and the average value of the nonlinear statistics has the highest normal / abnormal discrimination ability. Therefore, the evaluation function A is an average value of nonlinear statistics (orbit parallel measure), and the evaluation function B is a total value of linear statistics (time series data of motor current).

  As a method for evaluating normal / abnormal discrimination ability, various linear statistics and combinations thereof are plotted in a multidimensional state space for each of eight types of time-series data. Then, the normal / abnormal separation situation is judged from the distribution of the plotted statistics, and the discrimination ability is evaluated.

  As a specific example, FIGS. 6 to 21 show distribution diagrams of respective statistics when the evaluation function A is an average value of nonlinear statistics and the evaluation function B is a total value of linear statistics. A list of knowledge obtained from each distribution chart is shown in FIG. From these, it can be seen that a higher discrimination ability can be obtained by collecting current data during the return operation than during the interruption operation. In particular, it is found that there is a high possibility that normal / abnormal can be identified in the data in the cut-out sections of CASE2 and CASE2STB.

  The operation of the breaker is related to the weight of the breaker 4. During the shut-off operation of the shut-off rod 4, since the weight of the shut-off rod 4 acts, the motor current becomes lower or the operation time becomes shorter than in the return operation. Here, when the blocking rod 4 is broken, the weight of the blocking rod 4 is reduced as compared with the normal state. The influence of this weight reduction is more noticeable during the return operation than during the shut-off operation. That is, during the return operation, the load on the motor 5 is reduced due to the weight reduction, and the motor current is reduced or the operation time is shortened. On the other hand, during the shut-off operation, the weight of the shut-off rod 4 is not a load on the motor 5, but since the weight of the shut-off rod 4 decreases, the motor current becomes slightly higher or the operation time becomes slightly longer. The difference in motor current or operation time due to weight reduction during each operation is larger during the return operation than during the shut-off operation. Therefore, the separation status of the normal data distribution and the abnormal data distribution during the return operation is separated from the separation between the normal data distribution and the abnormal data distribution during the shut-off operation. The degree increases. This means that normal / abnormal can be easily identified.

  Further, when the blocking bar 4 is bent, the weight of the blocking bar 4 does not change. However, the position of the center of gravity of the blocking rod 4 approaches the rotation center side of the blocking rod 4. Therefore, during the return operation, the torque of the motor 5 can be smaller than that during normal operation, and the motor current becomes lower. During the shut-off operation, the weight of the shut-off rod 4 acts in the same way as during normal operation, so there is no significant change in the motor current. As described above, even when the blocking rod 4 is bent, it is possible to identify normality / abnormality by collecting current data during the return operation.

  Here, the normal / abnormal discrimination ability is evaluated in detail. The normal / abnormal discrimination ability is an ability that does not misrecognize abnormality as normal and does not misrecognize normal as abnormality. In order to verify this, each point of the evaluation function value is plotted, for example, as shown in FIG. And as shown in FIG.23 (b), the point which is the outermost of each point is extracted as a vertex of a polygon. A region is formed by connecting the vertices. Here, when the evaluation function values during normal operation are plotted, the formed area is defined as a normal area. An evaluation function value at the time of normal operation exists inside the normal area, and an evaluation function value at the time of abnormality exists outside.

  For this normal area, the evaluation function value at the time of abnormality existing inside is counted, and the evaluation function value at the time of normal operation existing outside is counted. FIG. 24 shows the misidentification rate calculated based on the count number. From this, it can be seen that a high correct diagnosis rate is obtained in the data in the cut-out sections of CASE2STB, CASE2_SGENV5, and CASE2STB_SGENV5. Based on the above evaluation results, high normal / abnormal discrimination ability can be obtained by using the current data detected during the return operation and cutting out the current data in the steady operation section. By setting the cut-out section as a steady operation section, it is possible to eliminate the influence of the inrush current when starting the operation in collecting current data.

  In the above evaluation method, the case where the evaluation function A is the average value of the nonlinear statistics and the evaluation function B is the total value of the linear statistics has been described. However, as the evaluation function, another statistic or each statistic is used. Even when combined, the most appropriate evaluation function can be selected by evaluating in the same manner.

  Software is created based on the cut-out section and the evaluation function selected as described above, and is installed in the processing device 2. The processing device 2 starts detecting an abnormality of the circuit breaker when the circuit breaker is newly installed or when the circuit breaker is replaced. This detection operation will be described with reference to FIG.

  When the breaker is newly installed or when the breaker 4 is replaced, the breaker 4 operates normally. Therefore, the collection unit 10 of the processing device 2 collects current data during operation detected by the detection device 1 when the circuit breaker is operating normally, and stores it in the memory (S10). Continue collecting until the breaker operates a specified number of times. When data is collected, it is collected during the shut-off operation and the return operation, but may be collected only during the return operation.

  The processing device 2 determines a criterion based on data collected when the circuit breaker is normal (S11). That is, as shown in FIG. 26, the cutout unit 11 cuts out current data in the steady operation section from the collected current data in all operation sections, and stores it in the memory as time series data (S20). The calculation unit 12 calculates a linear statistic of the time series data of the motor current, calculates a total value of the linear statistic, calculates a trajectory parallel measure based on the time series data, and calculates a nonlinear statistic. Then, the average value of the nonlinear statistics is obtained (S21). The reference determining unit 13 plots the total value of the linear statistics and the average value of the nonlinear statistics as evaluation functions in the multidimensional state space. Here, since there are two evaluation functions, they are plotted in a two-dimensional space. Then, a region formed by the plotted outer points is set as a normal area (S22). The determination criterion is determined by determining the normal area in this way.

  After the determination criterion is determined, the monitoring unit 14 monitors whether an abnormality has occurred each time the breaker operates (S12). That is, current data when the circuit breaker operates is collected and stored in a memory. Based on the motor current during the return operation of the shut-off rod 4, the total value of the linear statistics and the average value of the nonlinear statistics are obtained. Then, it is determined whether the points plotted in the two-dimensional space exist inside or outside the normal area. If it exists outside, it is determined that there is an abnormality in the circuit breaker, such as breakage of the barrier 4. At this time, the monitoring unit 14 notifies the management device 3 of the occurrence of an abnormality (S14). The management device 3 deals with a breaker abnormality.

  If it exists inside, the monitoring part 14 will judge that the circuit breaker is operating normally. At this time, the reference determination unit 13 updates the determination reference based on the collected current data (S15). That is, the total value of the linear statistics and the average value of the nonlinear statistics obtained by this operation are plotted, and the normal area is determined again including the added points. Thereby, the criterion is updated. The monitoring unit 14 monitors the presence or absence of an abnormality according to the updated determination criterion.

  Thus, by accumulating data when the circuit breaker is normal during monitoring, the normal area can be determined accurately, and the accuracy of detecting the presence or absence of an abnormality is increased. Therefore, the number of false detections for determining normality as abnormal is reduced, the frequency of workers going to the site can be reduced, and the number of false detections for determining abnormality as normal is reduced, thereby improving safety.

  In addition, if an abnormality is detected based on the data at the time of the return operation, even if an abnormality such as a breakage occurs when the blocking rod 4 is lowered, the occurrence of the abnormality is detected immediately when the blocking rod 4 is raised. Can be reported quickly. Therefore, it is possible to quickly cope with an abnormality in the circuit breaker, and safety is improved.

  In addition, since the motor current can be measured in the instrument box 8, the motor current can be detected without being affected by rain or wind. Therefore, abnormality detection can be performed without being affected by the natural environment, and detection accuracy can be improved.

  By the way, there is a possibility that erroneous detection may occur due to the installation environment of the circuit breaker and deterioration of mechanical parts such as a motor over time. Therefore, it is preferable to set a sensitivity coefficient with respect to the criterion. The reference determination unit 13 of the processing device 2 changes the determination reference by enlarging or reducing the normal area according to the sensitivity coefficient. For example, when the sensitivity coefficient = 1, the normal area is an area determined based on current data. If the sensitivity coefficient is less than 1, the normal area is reduced, and the criterion is strict. If the sensitivity coefficient is greater than 1, the normal area is enlarged and the criterion is relaxed.

  Therefore, when erroneous detection occurs after the installation of the circuit breaker, it can be changed to a determination standard corresponding to the current circuit breaker by adjusting the sensitivity coefficient in the field. Thereby, false detection can be reduced and high detection accuracy can be maintained over a long period of time.

  In addition, this invention is not limited to the said embodiment, Of course, many corrections and changes can be added to the said embodiment within the scope of the present invention. The above abnormality detection system may be applied not only to a level crossing but also to a circuit breaker that restricts the passage of a moving body such as a gate of a parking lot where a vehicle enters and exits. The processing device may be provided in the management device. The processing device collects current data by communicating with the detection device through a network.

  For example, an evaluation function related to operation time may be added as an evaluation function for determining a determination criterion. A normal area is determined in a multidimensional state space using two or more evaluation functions. In the above embodiment, the number of determination criteria is one. However, the presence / absence of an abnormality may be detected using a plurality of determination criteria using another evaluation function. By using a plurality of determination criteria, it is possible to accurately detect an abnormality without being affected by noise due to wind or the like. In addition, the type of abnormality is identified by selecting an evaluation function for each type of breaker abnormality such as breakage of the breaking fence, bending of the breaking fence, and obstruction of the breaking of the breaking fence, and determining a plurality of criteria. It becomes possible. As a result, an appropriate countermeasure can be taken against an abnormality of the circuit breaker.

Overall configuration diagram of an abnormality detection system for a circuit breaker according to the present invention Block diagram showing schematic configuration of processing equipment Flowchart for selecting elements necessary for determination criteria determination Diagram showing the experimental conditions related to the blocking rod for element selection The figure which shows the time-series change of the motor current at the time of operation of a circuit breaker Distribution chart of linear statistics and nonlinear statistics in CASE1 during shut-off operation Distribution chart of linear statistics and nonlinear statistics in CASE1STB during shut-off operation Distribution chart of linear statistics and nonlinear statistics in CASE2 during blocking operation Distribution chart of linear statistics and nonlinear statistics in CASE2STB during shut-off operation Distribution diagram of linear statistics and nonlinear statistics in CASE1_SGENV5 during shut-off operation Distribution diagram of linear statistics and nonlinear statistics in CASE1STB_SGENV5 during shut-off operation Distribution chart of linear statistics and nonlinear statistics in CASE2_SGENV5 during shut-off operation Distribution chart of linear statistics and nonlinear statistics in CASE2STB_SGENV5 during shut-off operation Distribution chart of linear statistics and nonlinear statistics in CASE1 during return operation Distribution chart of linear statistics and nonlinear statistics in CASE1STB during return operation Distribution chart of linear statistics and nonlinear statistics in CASE2 during return operation Distribution chart of linear statistics and nonlinear statistics in CASE2STB during return operation Distribution chart of linear statistics and nonlinear statistics in CASE1_SGENV5 during return operation Distribution chart of linear statistics and nonlinear statistics in CASE1STB_SGENV5 during return operation Distribution chart of linear statistics and nonlinear statistics in CASE2_SGENV5 during return operation The figure which shows the normal / abnormal identification result in CASE2STB_SGENV5 at the time of return operation | movement Diagram showing normal / abnormal discriminability in experiments (A) is the figure which plotted the evaluation function value, (b) is a figure which shows the area | region used as a normal area. The figure which shows the evaluation result of normal / abnormal discrimination in experiment Flowchart when detecting breaker malfunction Flow chart for determining criteria

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Detection apparatus 2 Processing apparatus 3 Management apparatus 4 Blocking rod 5 Motor 6 Drive apparatus 7 Instrument box 10 Collection part 11 Cutting part 12 Calculation part 13 Reference | standard determination part 14 Monitoring part

Claims (5)

An abnormality detection system for a circuit breaker operated by a motor to restrict the passage of a moving object outdoors, a detection device for detecting a motor current when the circuit breaker operates, and a block based on the detected current data A processing device that detects whether there is an abnormality in the machine and notifies the abnormality, and the processing device collects current data when the circuit breaker is operating normally, and when the circuit breaker operates Out of the operating sections, a cutting section that cuts out current data in a steady operating section to create time series data, a calculating section that calculates linear statistics and nonlinear statistics of the time series data, linear statistics and nonlinear statistics A criterion determination unit that determines a criterion for the presence / absence of an abnormality based on the amount, and a motor current detected each time the circuit breaker operates after the determination criterion is determined. Breaker anomaly detection system characterized by having a monitoring unit for detecting a. 2. The circuit breaker abnormality detection system according to claim 1, wherein when the circuit breaker operates normally, the processing device updates the determination criterion based on the detected motor current. 3. The circuit breaker abnormality detection system according to claim 1, wherein a sensitivity coefficient with respect to the determination criterion is set, and the criterion determination unit changes the determination criterion according to the sensitivity coefficient. The standard decision unit plots linear statistics and nonlinear statistics in a multidimensional state space, and determines the criteria for determining that the area formed by the plot is a normal area, the inside of the normal area is normal, and the outside of the normal area is abnormal An abnormality detection system for a circuit breaker according to any one of claims 1 to 3. The circuit breaker has a motor-driven liftable breaker, and the processing device determines a determination criterion based on current data when the breaker is raised. Anomaly detection system for circuit breakers described in 1.

Images