Judgment criterion generation device and status monitoring device

Abstract

To obtain a criterion creation device that can create a criterion used in a state monitoring device without using a display and mixing unauthorized data.SOLUTION: A criterion creation device comprises an operation state acquisition unit, a waveform acquisition unit, an outlier detection unit, and a criterion creation unit. The operation state acquisition unit acquires operation state information indicating whether a monitoring target is in operation. The waveform acquisition unit performs a plurality of times of sampling in one cycle to obtain waveform data indicating the waveform of the operation of the monitoring target, and samples current effective values of the monitoring target in a plurality of cycles when the operation state information indicates that the monitoring target is in operation. The outlier detection unit calculates an outlier calculation reference range based on a statistical technique by using the waveform data in the plurality of cycles, and determines if the waveform data is data falling out of the outlier calculation reference range. The criterion creation unit creates a criterion by using the waveform data in the plurality of cycles determined not to be data falling out of the range.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present disclosure relates to a state monitoring device that monitors abnormalities in devices that repeat the same operations, such as machining centers (MC) and presses, and a determination criterion generation device that generates determination criteria serving as a basis for determining abnormalities, and a state monitoring device.

Background Art

[0002] Patent Document 1 discloses an abnormality monitoring device that monitors abnormalities in rotating equipment that vibrates steadily. The abnormality monitoring device described in Patent Document 1 accumulates a plurality of waveform data when the rotating equipment is in a steady state in a waveform memory, and calculates a standard value and a standard deviation from the characteristic values of the waveform data in the waveform memory. When the characteristic value of newly detected waveform data is not within the standard deviation based on the standard value, the abnormality monitoring device described in Patent Document 1 determines that an abnormality has occurred in the rotating equipment and issues an alarm.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] The conventional technology described above assumes that the waveform data stored in the waveform memory is not abnormal, and calculates the standard value and standard deviation from the characteristic values ​​of the waveform data stored in the waveform memory. However, the conventional technology does not disclose how to store non-abnormal waveform data in the waveform memory. Therefore, there is a problem that unsuitable waveform data, which is not suitable to be used as a standard, may also be stored in the waveform memory. It is also conceivable to display the waveform data to be stored in the waveform memory on a graphical display device and remove the unsuitable data by visual inspection by the user, but in this case, there is a problem that a display device for displaying the waveform data is required.

[0005] This disclosure is made in view of the above, and aims to provide a judgment criterion generation device that can generate judgment criteria used in a status monitoring device without using a display and without introducing false data. [Means for solving the problem]

[0006] To solve the above-mentioned problems and achieve the objective, the judgment criterion generation device according to this disclosure comprises an operating state acquisition unit, a waveform acquisition unit, an outlier detection unit, and a judgment criterion generation unit. The operating state acquisition unit acquires operating state information indicating whether or not a monitored object that performs repetitive operations is in operation. The waveform acquisition unit performs multiple samples in one cycle to obtain waveform data that shows the waveform of the operation of the monitored object, and samples the effective current value, which is the effective value of the load current of the monitored object for multiple cycles, when the operating state information indicates that it is in operation. The outlier detection unit uses the waveform data for multiple cycles to calculate an outlier calculation criterion range, which is a criterion for determining whether the data is suitable to be used as a judgment criterion, based on a statistical method, and determines that the waveform data is data that falls outside the outlier calculation criterion range. It is an outlier. The determination is made. The determination criterion generation unit generates determination criteria for detecting abnormalities in the operation of the monitored target using waveform data from multiple cycles that have been determined by the outlier detection unit not to be outlier data. The outlier detection unit discards all acquired data, or waveform data formed from data sampled in the cycle in which the outlier-identified sampling point was found, if it detects a sampling point that has been determined to be an outlier. [Effects of the Invention]

[0007] This disclosure offers the advantage of generating judgment criteria used in a condition monitoring device without using a display and without introducing fraudulent data. [Brief explanation of the drawing]

[0008] [Figure 1] Block diagram showing an example of the configuration of a state monitoring device including a judgment criterion generation device according to Embodiment 1. [Figure 2] A flowchart showing an example of the processing procedure for the judgment criterion generation method in Embodiment 1. [Figure 3] A flowchart showing an example of the waveform data acquisition process in Figure 2. [Figure 4] A diagram illustrating an example of waveform data acquisition processing. [Figure 5] This figure shows an example of waveform data acquisition information obtained during the waveform data acquisition process. [Figure 6] This figure shows an example of waveform data acquisition information obtained during the waveform data acquisition process. [Figure 7] A flowchart showing an example of the outlier detection process in Figure 2. [Figure 8] A flowchart showing an example of the quartile extraction procedure in Figure 7. [Figure 9] A flowchart showing an example of the procedure for calculating the reference waveform in Figure 2. [Figure 10] A flowchart showing an example of the band generation process in Figure 2. [Figure 11] A diagram showing an example of reference waveform data and bands under normal conditions. [Figure 12] A flowchart showing an example of the processing procedure for the state monitoring method in the state monitoring device according to Embodiment 1. [Figure 13] This figure shows an example of the relationship between the effective current value obtained during the state monitoring process and the band used as the judgment criterion. [Figure 14]Block diagram showing an example of the configuration of a computer system that realizes the judgment criterion generation device according to Embodiment 1. [Modes for carrying out the invention]

[0009] The judgment criteria generation device and the status monitoring device according to the embodiments of this disclosure will be described in detail below with reference to the drawings.

[0010] Embodiment 1. Figure 1 is a block diagram showing an example of the configuration of a condition monitoring device including a judgment criterion generation device according to Embodiment 1. The condition monitoring device 10 is a device that monitors the operation of a device to be monitored 1, which is a device that repeats the same operation at a fixed period, and detects abnormalities in the device to be monitored 1. The device to be monitored 1 is the device that is monitored by the condition monitoring device 10. Here, the device to be monitored 1 is equipment that repeats the same operation. In one example, the device to be monitored 1 is production equipment such as a machining center or press that repeats the same operation at a fixed period. The device to be monitored 1 comprises an operation status output unit 2 and a current sensor 3. The operation status output unit 2 outputs operation status information, which is a signal indicating the operation status of the device to be monitored 1. An example of operation status information is a signal that indicates whether or not the device to be monitored 1 is in operation, i.e., processing. The signal is ON when processing is in progress and OFF when not processing. In the case of a machining center, the operation status output unit 2 outputs operation status information that indicates ON when processing is in progress and operation status information that indicates OFF when not processing. The current sensor 3 detects the current value flowing through the monitored device 1 and outputs the detected current value as an analog signal to the status monitoring device 10. As described above, in this example, the status of the monitored device 1 is monitored by the operating status information indicating the operating state of the monitored device 1 and the current value flowing through the monitored device 1.

[0011] The status monitoring device 10 comprises an operating status input unit 11, a current detection unit 12, an analog-to-digital (A / D) converter 13, an RMS value calculation unit 14, a waveform band generation unit 15, a memory 16, a status monitoring unit 17, and an error output unit 18.

[0012] The operating state input unit 11 is connected to the operating state output unit 2 of the monitored device 1 and receives the operating state information from the monitored device 1. An example of the operating state input unit 11 is an input port that receives the operating state information. The operating state input unit 11 outputs the received operating state information to the waveform band generation unit 15 and the state monitoring unit 17.

[0013] The current detection unit 12 converts an analog current signal indicating the current value from the current sensor 3 provided in the monitored device 1 into an analog voltage signal.

[0014] The A / D converter 13 converts the analog voltage signal output from the current detection unit 12 into a digital value. Every preset period T1, the current detection unit 12 converts the current value output from the current sensor 3 into an analog voltage signal, and the A / D converter 13 converts each of the plurality of analog voltage signals converted by the current detection unit 12 into a digital signal. This digital signal includes a digital value indicating the instantaneous value of the load current. The period T1 is, for example, 1 [ms]. The A / D converter 13 outputs the converted plurality of digital signals to the effective value calculation unit 14.

[0015] The effective value calculation unit 14 calculates the effective value of the current from the digital value of the current value output from the A / D converter 13. The effective value calculation unit 14 calculates the load current effective value Irms, which is the effective value of the load current flowing through the monitored device 1, based on the plurality of digital signals output from the A / D converter 13 every period T2. In one example, the effective value calculation unit 14 can obtain the load current effective value Irms by taking the moving average of the squared values of the instantaneous values of the load current included in the digital signals output from the A / D converter 13 and calculating the square root of the result of this moving average. The effective value calculation unit 14 outputs the calculated effective value of the current to the waveform band generation unit 15 and the state monitoring unit 17.

[0016] The waveform band generation unit 15 receives the operating status information of the monitored device 1 from the operating status input unit 11 and the effective current value calculated by the effective value calculation unit 14 as input, and generates a waveform band, which is a judgment criterion used when monitoring abnormalities in the monitored device 1. When the waveform band generation unit 15 detects an abnormal value in the waveform data acquired as a judgment criterion during waveform band generation, it notifies the error output unit 18 of the detection of the abnormal value in the waveform data. The waveform band generation unit 15 also outputs the generated waveform band to the memory 16. The waveform band generation unit 15 corresponds to the judgment criterion generation device.

[0017] Memory 16 stores data used by the waveform band generation unit 15. In this example, memory 16 stores waveform data acquisition information, reference waveform storage information, and band waveform storage information. Waveform data acquisition information is waveform data showing the waveform of the operation of the monitored device 1. As described later, multiple samples are taken in one cycle, and the effective current values ​​of the monitored device 1 for multiple cycles are sampled when the operation status information is active. Reference waveform storage information includes the average value of the effective current values ​​for multiple cycles at each sampling point. Band waveform storage information is the criterion for detecting abnormalities in the operation of the monitored device, and is information that defines the range, i.e., the width, of normal data at each sampling point.

[0018] The status monitoring unit 17 uses the band waveform storage information, which is the judgment criterion, stored in the memory 16 to monitor the operating status information from the monitored device 1, and when the operating status information of the monitored device 1 falls outside the waveform band, it notifies the error output unit 18 of the occurrence of an abnormality. In one example, when the operating status information indicates that the monitored device is operating, the status monitoring unit 17 acquires the effective current value of the monitored device at the sampling point and determines whether the effective current value is within the range of the judgment criterion at the sampling point generated by the waveform band generation unit 15, which is the judgment criterion generation device.

[0019] When the error output unit 18 receives notification from the waveform band generation unit 15 indicating the detection of an abnormal value in the waveform data acquired when generating the waveform band, it notifies the user of the status monitoring device 10 of a data invalidity error indicating that the waveform data acquired as the judgment criterion is invalid. When the error output unit 18 receives notification from the status monitoring unit 17 indicating the occurrence of an abnormality in the operating state of the monitored device 1, it notifies the user of the status monitoring device 10 of a device operation error indicating the occurrence of an abnormality in the operating state of the monitored device 1.

[0020] Here, we will describe in detail the waveform band generation unit 15, which is a judgment criterion generation device according to Embodiment 1. The waveform band generation unit 15 includes an operating state acquisition unit 15a, a waveform acquisition unit 15b, an outlier detection unit 15c, a reference waveform calculation unit 15d, and a band generation unit 15e.

[0021] The operating state acquisition unit 15a acquires operating state information input to the operating state input unit 11. The waveform acquisition unit 15b acquires the current waveform of the monitored device 1 while it is operating from the effective current value flowing through the monitored device 1 calculated by the effective value calculation unit 14 and the operating state information of the monitored device 1. Specifically, the waveform acquisition unit 15b performs multiple samplings in one cycle to obtain waveform data that shows the waveform of the operation of the monitored device 1, and samples the effective current value of the monitored device 1 for multiple cycles when the operating state information indicates that it is operating. In one example, when the operating state information of the monitored device 1 is ON, the waveform acquisition unit 15b takes a predetermined number of samplings as one cycle, acquires the effective current value over a predetermined number of cycles, and stores it in the memory 16 as waveform data acquisition information. The effective current value acquired at each sampling point in one cycle represents the current waveform. The number of samplings in one cycle is m, and the predetermined number of cycles is n. In one cycle, the first sample is counted as the 1st sample, the last sample as the mth sample, and i is an integer between 1 and m. Then, the effective current value obtained at the i-th sample is also called the data at sampling point "i".

[0022] The outlier detection unit 15c detects whether there is any invalid data mixed in with the current waveform data acquired by the waveform acquisition unit 15b due to errors in waveform acquisition timing, etc., i.e., data that is unsuitable to be used as a judgment criterion. Invalid data is also called an outlier. Specifically, the outlier detection unit 15c uses waveform data from multiple cycles to calculate an outlier calculation criterion range, which is the standard for determining whether the data is suitable to be used as a judgment criterion, based on a statistical method, and determines whether the waveform data is outside the outlier calculation criterion range.

[0023] Here, the outlier detection unit 15c calculates an outlier calculation reference range using the interquartile range or interquartile deviation obtained for the current RMS values ​​for multiple cycles at each sampling point, and determines whether there are outliers in the current RMS values ​​for multiple cycles at each sampling point that fall outside the outlier calculation reference range. If there is data outside the outlier calculation reference range, the outlier detection unit 15c discards the data acquired by the waveform acquisition unit 15b. If there is no data outside the outlier calculation reference range, the data acquired by the waveform acquisition unit 15b is treated as normal data. This process is executed sequentially from sampling point "1" and terminates when an outlier is found. If no outliers are found up to sampling point "m", the data acquired by the waveform acquisition unit 15b is considered normal data.

[0024] The reference waveform calculation unit 15d calculates a reference waveform from the waveform data if no outliers are found by the outlier detection unit 15c. Specifically, the reference waveform calculation unit 15d calculates the average value of the effective current at each sampling point and stores it in the memory 16 as reference waveform storage information.

[0025] The band generation unit 15e generates bands for the reference waveform based on the reference waveform calculated by the reference waveform calculation unit 15d. The band generation unit 15e calculates the standard deviation of the RMS current value at each sampling point, and calculates a positive band boundary value for each sampling point by adding the standard deviation to the average value of the RMS current value calculated by the reference waveform calculation unit 15d, and a negative band boundary value by subtracting the standard deviation from the average value of the RMS current value, and stores these in the memory 16 as band waveform storage information. The reference waveform calculation unit 15d and the band generation unit 15e correspond to the judgment criterion generation unit.

[0026] Next, the processing of the waveform band generation unit 15 will be described. Figure 2 is a flowchart showing an example of the processing procedure for the judgment criterion generation method in Embodiment 1. First, the waveform acquisition unit 15b performs waveform data acquisition processing (step S100). Here, the waveform acquisition unit 15b acquires waveform data for n cycles to be used as judgment criteria. Next, the outlier detection unit 15c performs outlier detection processing on the waveform data acquired in the waveform data acquisition processing (step S200).

[0027] Subsequently, the outlier detection unit 15c determines whether an outlier has been detected in the outlier detection process (step S300). If an outlier is detected (if Yes in step S300), the error output unit 18 outputs a data invalid error to the outside (step S400), and the process ends. If no outlier is detected (if No in step S300), the reference waveform calculation unit 15d performs a reference waveform calculation process to calculate a reference waveform (step S500). Subsequently, the band generation unit 15e performs a band generation process to generate bands based on the reference waveform (step S600), and the process ends.

[0028] The following describes in detail the waveform data acquisition process in step S100, the outlier detection process in step S200, the reference waveform calculation process in step S500, and the band generation process in step S600.

[0029] Figure 3 is a flowchart showing an example of the waveform data acquisition procedure in Figure 2. In step S100, the waveform acquisition process from step S102 to step S106, which is sandwiched between steps S101 and S107, is repeated until a predetermined number of cycles are reached.

[0030] First, the operating state acquisition unit 15a acquires operating state information indicating the operating state of the monitored device 1, which is input to the operating state input unit 11 (step S102). Next, the waveform acquisition unit 15b determines whether the operating state information acquired in step S102 is ON (step S103). If the operating state information is ON (Yes in step S103), the monitored device 1 is in the process of processing, so the waveform acquisition unit 15b acquires the effective current value of the current flowing through the monitored device 1 at a predetermined sampling period from the effective value calculation unit 14 (step S104), and stores the acquired effective current value in the waveform data acquisition information of the memory 16 (step S105).

[0031] Subsequently, or if the operating status information is not ON in step S103 (if No in step S103), the waveform acquisition unit 15b determines whether it has detected an OFF edge, which is the point at which the operating status information acquired in step S102 changes from ON to OFF (step S106). If an OFF edge has not been detected (if No in step S106), the acquisition of the effective current value for one cycle is not complete, so the process returns to step S102 to perform the next sampling. Note that if the operating status information is not ON, the monitored device 1 is not in the process of processing and does not need to be monitored, so the waveform acquisition unit 15b does not acquire the effective current value.

[0032] On the other hand, if an OFF edge is detected (Yes in step S106), it means that the acquisition of the RMS current value for one cycle has been completed. Then, the process proceeds to step S107, and the processes from step S102 to step S106 described above are repeated for all cycles. Once the acquisition of the RMS current value for all cycles is complete, the process returns to Figure 2.

[0033] Here, we will explain an example of waveform data acquisition processing. Figure 4 is a diagram illustrating an example of waveform data acquisition processing, and Figures 5 and 6 are diagrams illustrating an example of waveform data acquisition information obtained by the waveform data acquisition processing. In Figure 4, the vertical axis represents the RMS current value, and the horizontal axis represents the number of samples. In the example in Figure 4, we show an example where 20 samples are taken in one cycle, i.e., m is "20". In Figure 5, the number of samples in one cycle is represented by m, and the sampling point indicating which sampling it is in one cycle is represented by i. The sampling point i is represented using an integer from 1 to m. The number of cycles is represented by n, and the cycle value indicating which cycle it is is represented by j. As shown in Figures 5 and 6, the waveform data acquisition information may have a table format that records the RMS current value of the sampling point for each cycle value. In this case, the cycle value j is represented using an integer from 1 to n. The data for the sampling point i of cycle value j, i.e., the RMS current value, is represented by S[i,j].

[0034] In the example in Figure 4, sampling is performed 20 times in one cycle. Therefore, the 21st sampling process detects an OFF edge, and the process moves on to acquiring the first sample in the next cycle. Also, the sampling processes from the 2nd to the 20th sampling are part of one cycle and correspond to cases where no OFF edge was detected. In the example in Figure 5, it is shown that the process has been completed up to the 2nd sampling of the 4th cycle.

[0035] Once the sampling of waveform data for the specified number of cycles, i.e., n=10 times, is complete, the waveform data acquisition information is completed as shown in Figure 6. In the example in Figure 6, the waveform data acquisition information has a sampling rate of "20" per cycle and a cycle count of "10". Thus, the waveform data acquisition information is the data of the effective current obtained by performing a specified sampling rate m for a specified number of cycles n.

[0036] Next, the outlier detection process will be explained. Figure 7 is a flowchart showing an example of the procedure for the outlier detection process in Figure 2. In the outlier detection process of step S200, the outlier calculation process from step S202 to step S208, which is sandwiched between steps S201 and S209, is repeated according to the number of samples.

[0037] First, the outlier detection unit 15c acquires data for all cycles of sampling point i from the waveform data acquisition information (step S202). As described above, i is an integer from 1 to m. In one example, data is acquired for cycle values ​​j from "1" to "10" for sampling point i of the waveform data acquisition information shown in Figure 6. Then, the outlier detection unit 15c performs interquartile extraction processing using the acquired data (step S203).

[0038] Figure 8 is a flowchart showing an example of the procedure for the interquartile extraction process in Figure 7. First, the outlier detection unit 15c sorts the data for all cycles of sampling point i obtained in step S202 in ascending order (step S2031).

[0039] Next, the outlier detection unit 15c extracts the first quartile, which is the 1 / 4th data point in the total number of data points, from the minimum value of the sorted data (step S2032). The outlier detection unit 15c also extracts the second quartile, which is the 1 / 2th data point in the total number of data points, i.e., the median, from the minimum value of the sorted data (step S2033). Furthermore, the outlier detection unit 15c extracts the third quartile, which is the 3 / 4th data point in the total number of data points, from the minimum value of the sorted data (step S2034).

[0040] Next, the interquartile range is calculated by subtracting the data of the first quartile from the data of the third quartile (step S2035). This completes the interquartile extraction process in step S203, and the process returns to Figure 7. Note that although the above example uses the interquartile range, other methods such as the interquartile deviation obtained by dividing the interquartile range by 2 can also be used.

[0041] Returning to Figure 7, after the interquartile extraction process in step S203, the outlier detection unit 15c calculates an outlier calculation criterion range from the interquartile range calculated in the interquartile extraction process, which serves as the criterion for determining whether the data is suitable to be used as a judgment criterion (step S204). In one example, the outlier calculation criterion range can be calculated as 1.5 times the interquartile range.

[0042] Next, the outlier detection unit 15c determines whether there are any data points outside the outlier calculation reference range, i.e., outliers, in the data acquired in step S202 (step S205). If there are no outlier data points (no in step S205), the outlier detection unit 15c determines that the data is normal, i.e., there is no invalid data (step S206). After that, the process proceeds to step S209, and processing for the next sampling point i is carried out.

[0043] On the other hand, if there is incorrect data (if the answer is Yes in step S205), the outlier detection unit 15c determines that there is invalid data (step S207) and discards the acquired data (step S208). Then the process returns to Figure 2.

[0044] In this explanation, the outlier detection unit 15c determines whether an outlier exists in order from sampling point "1," and discards the acquired data if an outlier is found. In this case, all acquired data is discarded, and the data from cycles without outliers is wasted. Therefore, the outlier detection unit 15c may also be configured to discard the waveform data of the cycle in which an outlier was found if a sampling point is determined to be an outlier. In this case, the waveform data without outliers can be used without being wasted.

[0045] Figure 9 is a flowchart showing an example of the procedure for calculating the reference waveform shown in Figure 2. In the reference waveform calculation process of step S500, the reference waveform generation process from step S502 to step S504, which is sandwiched between steps S501 and S505, is repeated according to the number of samples m.

[0046] First, the reference waveform calculation unit 15d acquires data for all cycles of sampling point i from the waveform data acquisition information shown in Figure 6, which was acquired in the waveform data acquisition process in step S100 (step S502). Next, the reference waveform calculation unit 15d calculates the average value of the data for all cycles at sampling point i (step S503), and stores the calculated average value in the reference waveform storage information as the reference waveform data for sampling point i (step S504). In one example, the reference waveform storage information has a table format that records the average value of the current effective value at the sampling point for each cycle value, similar to the waveform data acquisition information. After that, the process proceeds to step S505, and the processes from steps S502 to S504 described above are repeated for the total number of samples. In this example, the average value of the data for all cycles is calculated sequentially from sampling point "1" to "20". After that, the process returns to Figure 2.

[0047] Figure 10 is a flowchart showing an example of the band generation process procedure in Figure 2. In the band generation process of step S600, the band generation processes from step S602 to step S606, which are sandwiched between steps S601 and S607, are repeated according to the number of samples m.

[0048] First, the band generation unit 15e acquires data for all cycles of sampling point i from the waveform data acquisition information shown in Figure 6, which was acquired in the waveform data acquisition process of step S100 (step S602). Next, the band generation unit 15e calculates the standard deviation of the acquired data for all cycles at sampling point i (step S603). The process from step S602 to step S603 described above constitutes the standard deviation calculation process for calculating the standard deviation.

[0049] Subsequently, the band generation unit 15e obtains the average value of the sampling point i from the reference waveform storage information in the memory 16 (step S604). Then, the band generation unit 15e calculates the band boundary value at the sampling point i using the obtained average value, i.e., the reference waveform data, and the calculated standard deviation (step S605). Here, the positive band boundary value and the negative band boundary value are calculated as shown in equations (1) and (2) below, respectively.

[0050] Positive band boundary value = reference waveform data + standard deviation ···(1) Negative band boundary value = reference waveform data - standard deviation ···(2)

[0051] The range between the positive band boundary value and the negative band boundary value constitutes the band. The band also includes the positive band and the negative band, with reference to the reference waveform data. The positive band is the range from the reference waveform data, i.e., the mean value at sampling point i, to the positive band boundary value. The negative band is the range from the reference waveform data, i.e., the mean value at sampling point i, to the negative band boundary value. In this way, with the mean value as the starting point, a positive band with a width equal to the standard deviation is formed on the positive side, and a negative band with a width equal to the standard deviation is formed on the negative side. The process from step S604 to step S605 described above constitutes the band generation process that generates the bands.

[0052] Subsequently, the band generation unit 15e stores the calculated positive and negative band boundary values ​​in the band waveform storage information provided in the memory 16 (step S606). In one example, the band waveform storage information may have a table format that associates the positive and negative band boundary values ​​for each sampling point.

[0053] Next, the process proceeds to step S607, where the processes described in steps S602 to S606 are repeated for all samples m. In this example, the positive and negative bands are generated sequentially from sampling point "1" to "20". As a result, the positive and negative bands at each sampling point from "1" to "20" are stored in the band waveform memory information. This completes the band generation process. After that, the process returns to Figure 2.

[0054] The process described above allows for the acquisition of a judgment criterion band using accurate normal waveform data, without the need for a display, and suppressing the inclusion of fraudulent data. Figure 11 shows an example of normal reference waveform data and bands. In this figure, the horizontal axis represents the number of samples, and the vertical axis represents the effective current value. Figure 11 connects the positive band boundary values ​​at each sampling point stored in the band waveform memory information, and connects the negative band boundary values ​​at each sampling point. The region between the positive and negative band boundary values ​​is the range in which the monitored device 1 is considered to be operating normally, and this becomes the judgment criterion band.

[0055] Next, the monitoring operation of the current RMS value, which is the value to be monitored by the state monitoring unit 17, will be described. Figure 12 is a flowchart showing an example of the processing procedure for the state monitoring method in the state monitoring device according to Embodiment 1. The state monitoring unit 17 receives the operating state information, which is the output of the operating state input unit 11, and the current RMS value calculated by the RMS value calculation unit 14 as input. The state monitoring unit 17 also reads the positive band boundary value and the negative band boundary value at each sampling point from the band waveform storage information in the memory 16 and monitors the operating state of the device to be monitored 1.

[0056] In the status monitoring process, the status monitoring processes from step S702 to step S711, which are sandwiched between step S701 and step S712, are repeatedly performed.

[0057] First, the status monitoring unit 17 acquires the operating status information of the monitored device 1 from the operating status input unit 11 (step S702). Next, the status monitoring unit 17 determines whether the operating status information acquired in step S702 is ON (step S703). If the operating status information is ON (Yes in step S703), the status monitoring unit 17 acquires the effective current value flowing through the monitored device 1 at the sampling point from the effective value calculation unit 14 (step S704). The status monitoring unit 17 also reads the positive and negative band boundary values ​​of the acquired effective current value at the sampling point from the band waveform storage information in the memory 16 (step S705).

[0058] Subsequently, the state monitoring unit 17 determines whether the effective current value obtained in step S704 is greater than the positive band boundary value (step S706). If the effective current value is less than or equal to the positive band boundary value (no in step S706), the state monitoring unit 17 determines whether the effective current value obtained in step S704 is less than the negative band boundary value (step S707). If the effective current value is greater than or equal to the negative band boundary value (no in step S707), the effective current value lies between the negative band boundary value and the positive band boundary value. In other words, the effective current value of the sampling point processed this time is within the band and is not abnormal, so the state monitoring unit 17 increments the sampling point by "1" in order to process the next sampling point (step S708). After that, the process returns from step S712 to step S701, at which point the state monitoring unit 17 determines whether it has detected an OFF edge, where the operation state information changes from ON to OFF, in the operation state information obtained in step S702 (step S709). If no OFF edge is detected (the answer in step S709 is No), the process returns to step S701. If an OFF edge is detected (the answer in step S709 is Yes), the operating status information of the monitored device 1 acquired in step S702 changes from ON to OFF, which means that one cycle has ended. Therefore, the sampling point is reset to "0" (step S710), and the process returns to step S701.

[0059] If, in step S706, the effective current value is greater than the positive band boundary value (if the answer is Yes in step S706), or if, in step S707, the effective current value is less than the negative band boundary value (if the answer is Yes in step S707), the effective current value is outside the band, and the status monitoring unit 17 detects that the device is out of band (step S711). In the case of an out-of-band operation, the status monitoring unit 17 outputs information indicating that an out-of-band operation has been detected to the error output unit 18, and the error output unit 18, for example, notifies an external party of a device operation error. After that, the process ends.

[0060] Furthermore, if the operating status information in step S703 is OFF (No in step S703), the process moves to step S709.

[0061] The state monitoring process described above will now be explained with a specific example. Figure 13 shows an example of the relationship between the RMS current value obtained during the state monitoring process and the band used as the judgment criterion. In this figure, the horizontal axis represents the number of samples, and the vertical axis represents the RMS current value. Figure 13 also shows the result of plotting the RMS current value at each sampling point on the band shown in Figure 11. As shown in Figure 13, the RMS current value is within the band range for sampling points "1" to "5" and "12" to "20", but it is outside the band range for sampling points "6" to "11".

[0062] According to the flowchart in Figure 12, the determination of whether the current RMS value at sampling point "1" is within the band is performed sequentially. As described above, sampling points "1" through "5" are within the band, but sampling point "6" is outside the band. The status monitoring unit 17 detects the out-of-band state at sampling point "6" and outputs a device operation error.

[0063] In the flowchart in Figure 12, a device operation error was immediately output when a band deviation was detected, but the status monitoring process is not limited to this. For example, the status monitoring unit 17 performs a comparison with the band for one cycle, sampling 20 times in the example in Figure 13, and calculates the deviation rate, which is the percentage of sampling points that deviated from the band in one cycle, from this result. When the deviation rate exceeds a predetermined standard value, the system may notify the user that a waveform different from the normal state has been detected. Examples of user notifications include contact output, communication output, and LCD (Liquid Crystal Display) blinking.

[0064] In the status monitoring device 10, the judgment criterion generation method is executed when an instruction to generate a band is given, and the status monitoring method is executed when an instruction to monitor the status is given. In other words, the status monitoring device 10 switches between executing the above methods.

[0065] As described above, the waveform band generation unit 15 of Embodiment 1 performs multiple samplings in one cycle, sampling the effective current value of the monitored device 1 for multiple cycles while the monitored device 1 is operating. The waveform band generation unit 15 uses the waveform data from multiple cycles to calculate an outlier calculation criterion range, which is a criterion for determining whether the data is suitable to be used as a judgment criterion, based on a statistical method, and determines whether the waveform data is outside the outlier calculation criterion range. If there is data that is outside the outlier calculation criterion range at a sampling point, the sampled data is discarded, and if there is no data that is outside the outlier calculation criterion range at any sampling point, the acquired data is used to generate a band that serves as a judgment criterion for monitoring abnormalities in the monitored device 1. Conventionally, it was difficult to prevent the inclusion of fraudulent data unless it was used in conjunction with a display, and therefore it was not possible to monitor for fraudulent data in the waveform data used to generate the band unless it was a product set with a display. However, Embodiment 1 has the effect of being able to generate the judgment criteria used in the status monitoring device without using a display and without introducing fraudulent data.

[0066] Next, the hardware configuration of the waveform band generation unit 15, which is a judgment criterion generation device according to Embodiment 1, will be described. In the judgment criterion generation device according to Embodiment 1, the computer system functions as a judgment criterion generation device when a program, which is a computer program describing the processing in the judgment criterion generation device, is executed on the computer system. Figure 14 is a block diagram showing an example of the configuration of a computer system that realizes the judgment criterion generation device according to Embodiment 1. As shown in Figure 14, this computer system includes a control unit 81, an input unit 82, a storage unit 83, a display unit 84, a communication unit 85, and an output unit 86, which are connected via a system bus 87.

[0067] In Figure 14, the control unit 81 is a processor such as a CPU (Central Processing Unit) and executes a program describing the processing in the judgment criterion generation device according to Embodiment 1. Note that a part of the control unit 81 may be implemented by dedicated hardware such as a GPU (Graphics Processing Unit) or FPGA (Field-Programmable Gate Array). The input unit 82 consists of a keyboard, mouse, etc., and is used by the user of the computer system to input various information. The storage unit 83 includes various types of memory such as RAM (Random Access Memory) and ROM (Read Only Memory), and storage devices such as a hard disk, and stores the program to be executed by the control unit 81, necessary data obtained during processing, etc. The storage unit 83 is also used as a temporary storage area for the program. The display unit 84 consists of a display, liquid crystal display panel, etc., and displays various screens to the user of the computer system. The communication unit 85 is a receiver and transmitter that perform communication processing. The output unit 86 is a printer, speaker, etc. Note that Figure 14 is an example, and the configuration of the computer system is not limited to the example in Figure 14.

[0068] Here, we will describe an example of the operation of the computer system until the above-mentioned program becomes executable. In a computer system with the above configuration, for example, a computer program is installed in the storage unit 83 from a CD-ROM or DVD-ROM set in a CD (Compact Disc)-ROM drive or DVD (Digital Versatile Disc)-ROM drive (not shown). When the program is executed, the program read from the storage unit 83 is stored in the main memory area of ​​the storage unit 83. In this state, the control unit 81 performs processing as a judgment criterion generation device according to Embodiment 1, according to the program stored in the storage unit 83.

[0069] In the above explanation, a program describing the processing in the judgment criteria generation device is provided using a CD-ROM or DVD-ROM as the recording medium. However, the explanation is not limited to this, and depending on the configuration of the computer system, the capacity of the program to be provided, a program provided via a transmission medium such as the Internet via the communication unit 85 may also be used.

[0070] This computer program, in one example, causes the computer system to execute the processing steps shown in Figures 2, 3, 7 to 10.

[0071] The waveform band generation unit 15 shown in Figure 1, including the operating state acquisition unit 15a, waveform acquisition unit 15b, outlier detection unit 15c, reference waveform calculation unit 15d, and band generation unit 15e, is realized by the control unit 81 executing a computer program stored in the storage unit 83. The storage unit 83 is also used to realize the waveform acquisition unit 15b, outlier detection unit 15c, reference waveform calculation unit 15d, and band generation unit 15e shown in Figure 1. The memory 16 shown in Figure 1 is realized by the storage unit 83. Furthermore, the judgment criterion generation device may be realized by multiple computer systems. In one example, the judgment criterion generation device may be realized by a cloud computer system.

[0072] Furthermore, the status monitoring device 10 may also be implemented by a computer system, similar to the judgment criterion generation device. In this case, the computer program, in one example, causes the computer system to execute the processing procedure shown in Figure 12.

[0073] The configurations shown in the above embodiments are merely examples, and it is possible to combine them with other known technologies, combine different embodiments, and omit or modify parts of the configuration without departing from the gist of the invention. [Explanation of Symbols]

[0074] 1. Device to be monitored, 2. Operating status output unit, 3. Current sensor, 10. Status monitoring device, 11. Operating status input unit, 12. Current detection unit, 13. A / D converter, 14. RMS value calculation unit, 15. Waveform band generation unit, 15a. Operating status acquisition unit, 15b. Waveform acquisition unit, 15c. Outlier detection unit, 15d. Reference waveform calculation unit, 15e. Band generation unit, 16. Memory, 17. Status monitoring unit, 18. Error output unit.

Claims

1. An operation status acquisition unit acquires operation status information indicating whether or not a monitored object that performs repetitive operations is currently operating, A waveform acquisition unit performs multiple samplings in one cycle to obtain waveform data showing the waveform of the operation of the monitored target, and samples the effective current value, which is the effective value of the load current of the monitored target for multiple cycles, when the operation status information indicates that it is in operation. An outlier detection unit that uses the waveform data for multiple cycles to calculate an outlier calculation criterion range that serves as a standard for determining whether the data is suitable to be used as a judgment criterion based on a statistical method, and determines whether the waveform data is an outlier that falls outside the outlier calculation criterion range. A judgment criterion generation unit generates a judgment criterion for detecting an abnormality in the operation of the monitored target using the waveform data for multiple cycles that the outlier detection unit has determined not to be outlier data, Equipped with, The judgment criterion generation device is characterized in that, when an outlier detection unit finds a sampling point that has been determined to be an outlier, it discards all acquired data, or the waveform data formed by data sampled in the cycle in which the sampling point determined to be an outlier was found.

2. The judgment criterion generating device according to claim 1, characterized in that the outlier detection unit calculates the outlier calculation criterion range using the interquartile range or interquartile deviation obtained for the current effective values ​​for multiple cycles at each sampling point, and determines whether there are outliers in the current effective values ​​for multiple cycles at each sampling point that are data outside the outlier calculation criterion range.

3. The aforementioned judgment criterion generation unit is A reference waveform calculation unit that calculates the average value of the current effective value at each sampling point, A band generation unit calculates the standard deviation of the effective current value at each sampling point, and calculates a positive band boundary value obtained by adding the standard deviation to the average value, and a negative band boundary value obtained by subtracting the standard deviation from the average value, for each sampling point. A judgment criterion generating device according to claim 1 or 2, characterized by having the following features.

4. A judgment criterion generating device according to claim 1 or 2, When the operating status information indicates that the monitored object is in operation, the status monitoring unit acquires the effective current value of the monitored object at the sampling point and determines whether the effective current value is within the range of the judgment criteria at the sampling point generated by the judgment criterion generation device. A condition monitoring device characterized by comprising the following features.

5. The judgment criteria generating device according to claim 3, When the operating status information indicates that the monitored object is in operation, the status monitoring unit acquires the effective current value of the monitored object at the sampling point and determines whether the effective current value is within the range of the judgment criteria at the sampling point generated by the judgment criterion generation device. A condition monitoring device characterized by comprising the following features.

Images