Abnormality diagnosing device and abnormality diagnosing method for rotary electric machine

Abstract

According to the present invention, a motor abnormality monitoring device is a rotary electric machine abnormality monitoring device provided with an abnormality determining unit for determining whether there is an abnormality in a rotary electric machine on the basis of time-series current data indicating a change over time in an AC current in a rotary electric machine, the abnormality monitoring device comprising: a waveform determining unit which determines whether the waveform of the AC current indicated by the time-series current data includes a unique waveform indicating an operating state of the rotary electric machine that is not to be subjected to abnormality monitoring; and a data processing unit which, if it is determined that the unique waveform is included, executes data processing for excluding time-series data of a non-target AC current including an AC current corresponding to the unique waveform from the abnormality determination performed by the abnormality determining unit.

Description

Abnormality diagnosis device and abnormality diagnosis method for rotating electrical machine

[0001] This application claims priority to Japanese Patent Application No. 2024-002733, filed with the Japan Patent Office on January 11, 2024, the contents of which are incorporated herein by reference.

[0002] Patent Document 1 discloses a motor diagnostic device that includes a data acquisition unit that acquires target data related to current values ​​between a power conversion device and a motor, a detection unit that detects peak values ​​of the current in a time-series waveform of the target data, and a diagnosis unit that diagnoses motor abnormalities based on the detected peak values.

[0003] Patent No. 7151827

[0004] In the above-mentioned diagnostic device, the current value may be measured even during periods when the motor is in an operating state that is not favorable for abnormality diagnosis, and the measured current value may be reflected in the target data. As a more specific example, a current value that is not suitable for abnormality diagnosis may be measured during periods when the motor is in a running or stopped operating state. If the diagnostic unit performs a diagnosis based on the target data that reflects such a current value, an accurate abnormality diagnosis may not be possible.

[0005] An object of the present disclosure is to provide an abnormality diagnosis device and an abnormality diagnosis method for a rotating electric machine that can accurately determine whether or not there is an abnormality in the rotating electric machine.

[0006] An abnormality monitoring device for a rotating electric machine according to at least one embodiment of the present disclosure is an abnormality monitoring device for a rotating electric machine that includes an abnormality determination unit for determining whether there is an abnormality in the rotating electric machine based on current time series data that indicates changes in AC current in the rotating electric machine over time, and includes: a waveform determination unit for determining whether the waveform of the AC current indicated by the current time series data includes a peculiar waveform that indicates an operating state of the rotating electric machine that is not subject to abnormality monitoring; and a data processing unit for performing data processing such that, when it is determined that the peculiar waveform is included, time series data of non-target AC currents that include the AC current corresponding to the peculiar waveform are not subject to abnormality determination by the abnormality determination unit.

[0007] An abnormality monitoring method for a rotating electric machine according to at least one embodiment of the present disclosure is an abnormality monitoring method for a rotating electric machine, comprising: an abnormality determination step for determining whether or not there is an abnormality in the rotating electric machine based on current time series data indicating a change over time in AC current in the rotating electric machine, and comprising: a waveform determination step for determining whether or not a peculiar waveform indicating an operating state of the rotating electric machine that is not subject to abnormality monitoring is included in the waveform of the AC current indicated by the current time series data; and a data processing step for performing data processing to exclude time series data of a non-target AC current, which is the AC current that includes the peculiar waveform, from the abnormality determination step when it is determined that the peculiar waveform is included.

[0008] According to the present disclosure, it is possible to provide an abnormality diagnosis device and an abnormality diagnosis method for a rotating electric machine that can accurately determine whether or not there is an abnormality in the rotating electric machine.

[0009] FIG. 1 is a schematic diagram of an anomaly monitoring system according to an embodiment; FIG. 2 is a schematic diagram of a motor anomaly monitoring device according to an embodiment; FIG. 3 is a schematic diagram of current time series data generated by a time series data generating unit according to an embodiment; FIG. 4 is a schematic diagram of divided time series data including a peculiar waveform according to an embodiment; FIG. 5 is a schematic diagram of a waveform determining unit according to a first example; FIG. 6 is a schematic diagram of a peculiar waveform including a sudden waveform and a start / stop waveform according to an embodiment; FIG. 7 is a schematic diagram of a peculiar waveform including a start / stop waveform according to an embodiment; FIG. 8 is a schematic diagram of a waveform determining unit according to a second example; FIG. 9 is a schematic diagram of a peculiar waveform including a load decrease waveform according to an embodiment; FIG. 10 is a schematic diagram of a normally identified fundamental waveform according to an embodiment; FIG. 11 is a schematic diagram of a peculiar fundamental waveform according to an embodiment; FIG. 12 is a schematic diagram of a waveform determining unit according to a third example; FIG. 13 is a schematic diagram of a load increase waveform according to an embodiment; FIG. 14 is a flowchart of an anomaly monitoring process according to an embodiment; FIG. 15 is a flowchart of a waveform determining process according to an embodiment.

[0010] Several embodiments of the present disclosure will be described below with reference to the accompanying drawings. However, the dimensions, materials, shapes, relative arrangements, etc. of components described as embodiments or shown in the drawings are not intended to limit the scope of the present disclosure and are merely illustrative examples. For example, expressions expressing relative or absolute arrangements, such as "in a certain direction," "along a certain direction," "parallel," "orthogonal," "center," "concentric," or "coaxial," not only strictly express such arrangements, but also express relative displacements with a tolerance or angle or distance to the extent that the same function is achieved. For example, expressions expressing the equality of things, such as "same," "equal," and "homogeneous," not only express strict equality, but also express tolerance or differences to the extent that the same function is achieved. For example, expressions expressing shapes such as a square or cylindrical shape not only express shapes such as a square or cylindrical shape in the strict geometric sense, but also express shapes including concave and convex portions, chamfered portions, etc., to the extent that the same effect is achieved. On the other hand, the expressions "comprise," "include," or "have" one component are not exclusive expressions that exclude the existence of other components. Note that similar components may be assigned the same reference numerals and descriptions thereof may be omitted.

[0011] 1 is a schematic diagram of an abnormality monitoring system 10 according to an embodiment of the present disclosure. The abnormality monitoring system 10 includes a motor 7, power supply equipment 2 for supplying AC power to the motor 7, a mechanical device 8 driven by the motor 7, a measuring device 6 for measuring the power supply current of the motor 7, an A / D converter 3 for converting the measured analog value of the power supply current output from the measuring device 6 into a digital value, and a motor abnormality monitoring device 1 electrically connected to the A / D converter 3. The motor 7 is an example of a "rotating electric machine" in the present invention, and the motor abnormality monitoring device 1 is an example of an "abnormality monitoring device for a rotating electric machine" in the present invention.

[0012] The motor 7 is, for example, a three-phase AC motor (however, the present disclosure is not limited to this and the motor 7 may be a DC motor). The power supply equipment 2 includes, for example, a three-phase AC power supply (however, the present disclosure is not limited to this and the power supply equipment 2 may be an inverter power supply). The mechanical device 8 may be any movable device that is driven by the motor 7. Although merely an example, the mechanical device 8 may be a movable device incorporated into a gas turbine system, and as a more specific example, it may be a blower.

[0013] The measuring device 6 is configured to measure the power supply current of the motor 7. For example, the measuring device 6 may measure the current value flowing through the electric wire 4 connected to the power supply equipment 2 and the motor 7. The power supply current may be the excitation current in one of the coils of the motor 7. Hereinafter, the current measured by the measuring device 6 may be simply referred to as the "power supply current." The power supply current is an example of the "AC current" of the present invention. The A / D converter 3 is configured to convert the current value continuously measured by the measuring device 6 into a digital value. The measured time series data Dm generated by the A / D converter 3 indicates the change over time in the measured power supply current. The measured time series data Dm is sent to the motor abnormality monitoring device 1.

[0014] The motor abnormality monitoring device 1 is configured by a computer and includes a processor, memory (storage medium), and an external communication interface. The processor may be a CPU, GPU, MPU, DSP, or a combination thereof. In other embodiments, the processor may be implemented by an integrated circuit such as a PLD, ASIC, FPGA, or MCU. The memory is configured to temporarily or non-temporarily store various data and may be implemented by at least one of a RAM, a ROM, or a flash memory, for example. The processor executes various control processes according to instructions from a program loaded into the memory.

[0015] 2 is a schematic diagram of a motor abnormality monitoring device 1 according to an embodiment of the present disclosure. The motor abnormality monitoring device 1 includes a time-series data generator 15 for generating current time-series data Da based on measurement time-series data Dm sent from the A / D converter 3, and an abnormality determination unit 28 for determining whether an abnormality exists in the motor 7 based on the current time-series data Da.

[0016] 3 is a schematic diagram of current time series data Da generated by the time series data generator 15 according to an embodiment of the present disclosure. The current time series data Da includes multiple divided time series data Dv indicating changes over time in the power supply current in different time periods. The time width (dimension L) of each of the multiple divided time series data Dv is constant, and is, for example, greater than or equal to 5 seconds and less than or equal to 20 seconds. The time series data generator 15 performs data processing to cut out the measured time series data Dm generated by the A / D converter 3 at predetermined time widths, thereby generating current time series data Da from the measured time series data Dm.

[0017] 2 , for example, generates spectral data by performing a fast Fourier transform on each of the plurality of divided time series data Dv included in the current time series data Da, and determines whether or not there is an abnormality in the motor 7 based on the spectral data. As a more specific example, a plurality of spectral data corresponding to each of the plurality of divided time series data Dv is generated. If the spectral intensity corresponding to a specific frequency in any of the spectral data deviates from the allowable range, the abnormality determination unit 28 determines that there is an abnormality in the motor 7. Note that the present disclosure is not limited to performing a Fourier transform on the current time series data Da. As another example, the abnormality determination unit 28 may directly determine whether or not there is an abnormality in the motor 7 from the current time series data Da.

[0018] The inventors of the present application have discovered that if the multiple divided time-series data Dv include data that is not suitable for abnormality determination, the abnormality determination unit 28 may not accurately perform abnormality determination. For example, the motor load applied to the motor 7 may become temporarily or momentarily unstable due to rotational resistance of the mechanical device 8. This phenomenon can be observed while the motor 7 is in a start-up operating state, a stop-operating state, or a rated operating state under limited conditions. If the current value measured by the measuring device 6 fluctuates significantly temporarily or momentarily due to fluctuations in the motor load, the above-mentioned spectral intensity may exceed the allowable range. In this case, the abnormality determination unit 28 may erroneously determine that the motor 7 is abnormal even though there is actually no abnormality in the motor 7, resulting in an erroneous determination phenomenon.

[0019] Therefore, in the embodiment illustrated in Figure 2, the motor abnormality monitoring device 1 is equipped with a waveform judgment unit 30 and a data processing unit 29 so that data that may be contained in the multiple divided time series data Dv and is not suitable for abnormality judgment is not used as judgment material by the abnormality judgment unit 28.

[0020] The waveform determination unit 30 is configured to determine whether a peculiar waveform 90 is included in the power supply current waveform 9 (see FIG. 4 ) indicated by the current time-series data Da. The peculiar waveform 90 indicates an operating state of the motor 7 that is not subject to abnormality monitoring. The operating states that are not subject to monitoring are predetermined by a program stored in the memory of the motor abnormality monitoring device 1. The waveform determination unit 30 determines whether the power supply current waveform 9 includes the peculiar waveform 90 for each of the multiple divided time-series data Dv. For example, as shown in FIG. 4 , if the absolute value of the power supply current is extremely low over a certain period of time, the current waveform drawn by the power supply current is determined to be the peculiar waveform 90. Note that the peculiar waveform 90 shown in FIG. 4 is merely an example of the present disclosure.

[0021] Hereinafter, the power supply current that forms the specific waveform 90 may be referred to as an "asymmetric power supply current." The waveform 9 of the asymmetric power supply current may include a portion of the normal waveform 9 in addition to the specific waveform 90. The asymmetric power supply current is an example of the "asymmetric AC current" of the present invention.

[0022] 2 treats the divided time-series data Dv including the non-target power supply current among the plurality of divided time-series data Dv as time-series data D1, and sends data for identifying the time-series data D1 to the data processing unit 29. The data for identification may be the time-series data D1 itself.

[0023] The data processing unit 29 is configured to perform data processing on the current time-series data Da to exclude the time-series data D1 indicating the non-target power supply current from the target for abnormality determination by the abnormality determination unit 28. In one embodiment of the present disclosure, the time-series data D1 is deleted and is not input to the abnormality determination unit 28. Then, the remaining plurality of divided time-series data Dv included in the current time-series data Da are input to the abnormality determination unit 28.

[0024] According to the above configuration, when the AC current waveform 9 represented by the current time-series data Da includes a peculiar waveform 90 indicating an operating state of the motor 7 that is not subject to abnormality monitoring, the time-series data D1 of the non-target power supply current that depicts the peculiar waveform 90 is treated as not subject to abnormality determination by the abnormality determination unit 28. This allows the abnormality determination unit 28 to determine whether or not there is an abnormality in the motor 7 based on the current time-series data Da that is suitable for abnormality determination. Therefore, a motor abnormality monitoring device 1 that can accurately determine whether or not there is an abnormality in the motor 7 is realized.

[0025] Below, three detailed examples of the waveform determination unit 30 that detects the peculiar waveform 90 will be given. In each detailed example, the operating state of the motor 7 indicated by the peculiar waveform 90 will be explained.

[0026] <Details of the Waveform Determining Unit 30 (First Example)> Figure 5 is a schematic diagram of the waveform determining unit 30A (30) according to the first example. In the first example, the peculiar waveform 90 includes a sudden waveform 98 indicating the start-up operating state of the motor 7 or the operating state of the motor 7 in which a temporary load fluctuation (e.g., a load increase) has occurred (see Figure 6). The sudden current corresponds to an inrush current that momentarily assumes a very high absolute value after the motor 7 has started operating from a stopped state, or a first current that momentarily assumes a very high absolute value due to a temporary load fluctuation. In the first example, the current waveform formed by the sudden current is considered to be the sudden waveform 98.

[0027] A configuration for detecting a sudden waveform 98 will be described with reference to FIG. 5 . The waveform determination unit 30A according to the first example includes a sudden waveform determination unit 38 configured to determine whether a sudden waveform 98 is included in the waveform 9 of the power supply current. The sudden waveform determination unit 38 is configured to determine that the waveform 9 includes a sudden waveform 98 when any of the power supply currents in the plurality of divided time-series data Dv includes a sudden current whose absolute value exceeds a first specified threshold. The first specified threshold is a numerical value used to determine the sudden current. As merely an example, the first specified threshold is a numerical value that is 90% or more of the upper limit of the current value that can be measured by the measurement device 6. Note that in FIG. 6 , the first specified threshold is also set below the t-axis, but is not shown for the sake of simplicity.

[0028] The sudden waveform determination unit 38 treats the divided time-series data Dv determined to include the sudden waveform 98 as time-series data D1 of the non-target power supply current. The time-series data D1 is deleted by the data processing unit 29 (see FIG. 2).

[0029] According to the inventor's findings, the inrush current that can be measured while the motor 7 is in a start-up operating state, or the first current that changes suddenly due to load fluctuations on the motor 7, has a very high absolute value, and therefore the power supply current including such a sudden current can hinder accurate abnormality determination. In this regard, with the above-described configuration, the time-series data D1 of the non-target power supply current that forms the sudden waveform 98 is treated as not being subject to abnormality determination. This allows the abnormality determination unit 28 to accurately determine whether there is an abnormality in the motor 7.

[0030] In the first example, the peculiar waveform 90 further includes a start / stop waveform 97 indicating whether the motor 7 is in a start / stop state or a stop state (see FIGS. 6 and 7 ). For example, as shown in FIG. 6 , when the motor 7 is started, a weak current that follows the start / stop waveform 97 is generated before a sudden current occurs. The weak current is a power supply current whose absolute value remains below a second specified threshold for a predetermined period of time or more. As shown in FIG. 7 , the start / stop waveform 97 can also be generated when the motor 7 is in a stop state. When the motor 7 changes from a rated operation state to a stop state, the motor rotation speed gradually decreases from the rated rotation speed, generating a weak current that follows the start / stop waveform 97. Note that the second specified threshold can also be set below the t-axis in FIGS. 6 and 7 , but this is not shown for the sake of simplicity.

[0031] 5, a configuration for detecting the start / stop waveform 97 will be described. The waveform determination unit 30A further includes a start / stop waveform determination unit 37 configured to determine whether the power supply current includes a weak current. The start / stop waveform determination unit 37 is configured to determine that the waveform 9 includes the start / stop waveform 97 when the power supply current of any of the plurality of divided time-series data Dv includes a weak current whose absolute value is equal to or less than a second specified threshold.

[0032] The second specified threshold is a value used to determine whether or not a weak current is present. By way of example, the second specified threshold is a value that is 30% or less of the upper limit of the rated current value of the motor 7, and more specifically, a value that is 25% or less.

[0033] The start / stop waveform determination unit 37 treats the divided time-series data Dv determined to include the start / stop waveform 97 as time-series data D1 of the non-target power supply current. The time-series data D1 is deleted by the data processing unit 29 (see FIG. 2).

[0034] According to the inventor's findings, while the motor 7 is in a start-up or stop-up state, the weak current has an extremely small value for a certain period of time, and therefore the power supply current including the weak current can hinder accurate abnormality determination. In this regard, with the above-described configuration, the time-series data D1 of the non-target power supply current including the start-up / stop waveform 97 is treated as not being subject to abnormality determination. This allows the abnormality determination unit 28 to accurately determine whether there is an abnormality in the motor 7.

[0035] <Details of the Waveform Determining Unit 30 (Second Example)> FIG. 8 is a schematic diagram of the waveform determining unit 30B (30) according to a second example. In the second example, the peculiar waveform 90 includes a load reduction waveform 96 (see FIG. 9). The load reduction waveform 96 represents the operating state of the motor 7 when the motor load momentarily decreases during rated operation. The load reduction waveform 96 also occurs when the operating state of the motor 7 momentarily becomes unstable due to a power capacity shortage in the power supply equipment 2 (see FIG. 1). Taking the case where the motor load momentarily decreases during operation of the motor 7 (which may be during rated operation) as an example, the torque required to drive the motor 7 momentarily decreases. In this case, the maximum absolute value of the power supply current measured by the measuring device 6 momentarily falls below a first reference value serving as a threshold. Hereinafter, such a power supply current may be referred to as a "low current."

[0036] 9 , a low current forming a load decrease waveform 96 can occur instantaneously in a basic time period corresponding to one cycle of the power supply current (i.e., in one cycle of a basic waveform 95 in the waveform 9 of the power supply current). In this example, a plurality of basic waveforms 95 are identified from the waveform 9 of the current time-series data Da. Then, a determination is made for each of the plurality of basic waveforms 95 as to whether the power supply current includes a low current. If it is determined that a low current is included, it is determined that the basic waveform 95 includes a load decrease waveform 96. Note that the basic time period (dimension M0) corresponding to one cycle of the power supply current may vary slightly among the detected plurality of basic waveforms 95.

[0037] Here, the first reference value, which is a threshold for detecting low current, is a value obtained by multiplying a peak value corresponding to 1.41 times the effective value of the power supply current by a first coefficient smaller than 1. By way of example only, the first coefficient is a value greater than or equal to 0.7 and less than 1.0, more specifically, a value greater than or equal to 0.8 and less than or equal to 0.9. The peak values ​​of the power supply current may differ among the multiple divided time-series data Dv. Therefore, the first reference value is specified in association with each of the multiple divided time-series data Dv. Note that in FIG. 9, the first reference value is also set below the t-axis, but is not shown in the figure for the sake of simplicity.

[0038] A configuration for detecting a load decrease waveform 96 will be described with reference to FIG. 8 . The waveform determination unit 30B according to the second example includes a first acquisition unit 21 for acquiring a first reference value for each of the plurality of divided time series data Dv included in the current time series data Da. The first acquisition unit 21 acquires the first reference value for each of the divided time series data Dv by multiplying the peak value of the power supply current by a first coefficient. The first acquisition unit 21 also performs data processing to associate the plurality of first reference values ​​with the plurality of divided time series data Dv, respectively. The results of the data processing are output to a load decrease waveform determination unit 36 ​​(described below), which is a component of the waveform determination unit 30B.

[0039] The waveform determination unit 30B further includes an identification unit 23. The identification unit 23 is configured to identify a plurality of basic waveforms 95, each of which is regarded as one period of the power supply current, in the power supply current waveform 9 indicated by each of the plurality of divided time-series data Dv.

[0040] An example of a method for identifying the fundamental waveform 95 is as follows. As shown in FIG. 10 , the intersection of the current waveform and the time axis includes two types of points (zero-crossing points). The first point is a rising point P1 where the current value switches from a negative value to a positive value (including 0), and the second point is a falling point P2 where the current value switches from a positive value (including 0) to a negative value. The identification unit 23 identifies the fundamental waveform 95 by detecting the rising points P1 and falling points P2 that are alternately arranged on the time axis. However, the method for identifying the fundamental waveform 95 is not limited to this. Instead of the falling points P1 and rising points P2, the fundamental waveform 95 can also be identified by detecting the maximum and minimum values ​​of the power supply current. Regardless of which embodiment is adopted, the identification result of the identification unit 23 is output to a load reduction waveform determination unit 36 ​​(described below), which is a component of the waveform determination unit 30B.

[0041] The load decrease waveform determination unit 36 ​​illustrated in FIG. 8 is configured to determine whether a load decrease waveform 96 is included in the waveform 9 of the power supply current based on a first reference value and a fundamental waveform 95. More specifically, the load decrease waveform determination unit 36 ​​is configured to divide the time period indicated by each of the divided time series data Dv into a plurality of fundamental time periods corresponding to one cycle of the power supply current. For the division, the fundamental waveform 95 identified by the identification unit 23 is referenced. Furthermore, the load decrease waveform determination unit 36 ​​compares the maximum absolute value of the power supply current with the first reference value for each of the plurality of fundamental time periods. A power supply current whose maximum absolute value is smaller than the first reference value is determined to be a low current, and it is determined that the waveform 9 includes the load decrease waveform 96.

[0042] The load decrease waveform determination unit 36 ​​treats the divided time-series data Dv determined to include the load decrease waveform 96 as the time-series data D1 of the non-target power supply current. The time-series data D1 is deleted by the data processing unit 29 (see FIG. 2).

[0043] According to the inventor's findings, low currents that can be measured when the load momentarily decreases during operation have smaller absolute values ​​than normal power supply currents measured while the load is being applied. Therefore, power supply currents that include low currents can hinder accurate abnormality determination. In this regard, with the above configuration, the time-series data D1 of the non-target power supply currents that include the load decrease waveform 96 is treated as not being subject to abnormality determination. This allows the abnormality determination unit 28 to accurately determine whether there is an abnormality in the motor 7.

[0044] 2 , the time-series data generator 15 may perform approximation processing on the measured time-series data Dm generated by the A / D converter 3. The approximation processing is processing for setting the power supply current included in a dead-band region R (see FIG. 11 ), where the absolute value is equal to or less than a third specified threshold value, to zero. The dead-band region R is the region defined by two dashed lines Su and Sd in FIG. 11 (the dashed lines Su and Sd correspond to the third specified threshold value). The approximation processing is reflected in each of the multiple divided time-series data Dv.

[0045] The measured value of the power supply current that fluctuates around 0 is set to 0 in each divided time-series data Dv. This allows the load decrease waveform determination unit 36 ​​to more accurately identify the current (fundamental waveform 95) corresponding to one cycle included in the power supply current indicated by the current time-series data Da.

[0046] In order for the load decrease waveform determination unit 36 ​​to accurately determine whether or not a load decrease waveform 96 exists, it is preferable that the identification unit 23 accurately identify the fundamental waveform 95. However, a peculiar fundamental waveform 94 having only about half the cycle of the power supply current may be erroneously identified as the fundamental waveform 95. Two peculiar fundamental waveforms 94 are shown in FIG.

[0047] According to the inventor's findings, the period (dimension M1) of each peculiar fundamental waveform 94 is only about half the period (dimension M0 in FIG. 10) of the normal fundamental waveform 95. The reason for this is as follows: The power supply current before approximation processing may fluctuate near the dead band region R. In this case, the power supply current may cross the boundary (dashed lines Su, Sd) that defines the dead band region R multiple times within a short period of time. In particular, when the power supply is inverter-controlled, data in the above state often occurs.

[0048] In the enlarged view of FIG. 11 , the power supply current before approximation is indicated by a two-dot chain line N, and the power supply current after approximation is indicated by a solid line T. For example, if the power supply current before approximation (two-dot chain line N) crosses the dashed line Sd multiple times in a short period of time, the power supply current after approximation (solid line T) will assume a value of zero multiple times in a short period of time. The identifying unit 23 identifies the falling point P1 or the rising point P2 each time the power supply current indicated by the solid line T assumes a value of zero. As a result, the peculiar fundamental waveform 94 is identified. Such erroneous point identification is likely to occur when the power supply current decreases from a maximum value to the dead band region R or increases from a minimum value to the dead band region R. Therefore, the period of the peculiar fundamental waveform 94 tends to be approximately 50% of the dimension M0.

[0049] Based on this knowledge, the load decrease waveform determination unit 36 ​​of the present example is configured to exclude the peculiar fundamental waveform 94 identified as the fundamental waveform 95 by the identification unit 23 from the objects to be determined. More specifically, if the plurality of fundamental waveforms 95 identified by the identification unit 23 includes one having a period that is less than 55% of one period of the power supply current, the load decrease waveform determination unit 36 ​​treats the fundamental waveform 95 as the peculiar fundamental waveform 94. Then, the load decrease waveform determination unit 36 ​​excludes the divided time-series data Dv including the peculiar fundamental waveform 94 from the objects to be determined for determining whether a low current is included.

[0050] According to the above configuration, a fundamental waveform 95 having a period that is less than 55% of one period of the power supply current is determined to be an anomalous fundamental waveform 94, and is therefore excluded from the determination of whether a low current is included. This allows the load decrease waveform determination unit 36 ​​to accurately classify the time period indicated by the current time-series data Da into basic time periods that correspond to one period of the power supply current. Therefore, it is possible to accurately determine whether the waveform 9 of the power supply current includes a load decrease waveform 96.

[0051] 11, the solid line T and the two-dot chain line N outside the dead zone R actually coincide with each other. In the drawing, the two lines are shown separated from each other for ease of viewing.

[0052] <Details of the Waveform Determining Unit 30 (Third Example)> FIG. 12 is a schematic diagram of a waveform determining unit 30C (30) according to a third example. In the third example, the peculiar waveform 90 includes a load increase waveform 93 (see FIG. 13). The load increase waveform 93 indicates an operating state of the motor 7 in which the motor load momentarily increases during operation (which may be rated operation). In other words, the load increase waveform 93 can occur under limited conditions during operation of the motor 7. When the motor load momentarily increases during operation of the motor 7, the torque required to drive the motor 7 momentarily increases. In this case, the absolute value of the power supply current measured by the measuring device 6 momentarily exceeds a second reference value serving as a threshold value. Hereinafter, such a power supply current may be referred to as a "high current."

[0053] Here, the second reference value, which is a threshold for detecting high current, is a value obtained by multiplying a peak value corresponding to 1.41 times the effective value of the power supply current by a second coefficient greater than 1. As a mere example, the second coefficient is a value greater than or equal to 1.1 and less than or equal to 1.3. The peak values ​​of the power supply current may differ among the multiple divided time-series data Dv. Therefore, the second reference value is specified in association with each of the multiple divided time-series data Dv. Note that in FIG. 13, the second reference value is also set below the t-axis, but is not shown in the figure to simplify the drawing.

[0054] A configuration for detecting a load increase waveform 93 will be described with reference to FIG. 12 . The waveform determination unit 30C according to the third example includes a second acquisition unit 22 for acquiring a second reference value for each of the plurality of divided time series data Dv included in the current time series data Da. The second acquisition unit 22 acquires the second reference value for each of the divided time series data Dv by multiplying the peak value of the power supply current by a second coefficient. The second acquisition unit 22 also performs data processing to associate the plurality of second reference values ​​with the plurality of divided time series data Dv, respectively. The results of the data processing are output to a load increase waveform determination unit 35 (described below), which is a component of the waveform determination unit 30C.

[0055] The waveform determination unit 30C further includes a load increase waveform determination unit 35 for determining whether the waveform 9 of the power supply current includes a load increase waveform 93. The load increase waveform determination unit 35 is configured to determine that the waveform 9 includes the load increase waveform 93 when the power supply current indicated by each of the plurality of divided time series data Dv includes a high current whose absolute value exceeds a second reference value. The load increase waveform determination unit 35 treats the divided time series data Dv determined to include the load increase waveform 93 as time series data D1 of a non-target power supply current. The time series data D1 is deleted by the data processing unit 29 (see FIG. 2 ).

[0056] According to the inventor's findings, high currents that can be measured when the load momentarily increases during operation have a larger absolute value than normal AC currents measured while the load is being applied. Therefore, power supply currents that include high currents can hinder accurate abnormality determination. In this regard, with the above configuration, the time-series data D1 of the non-target power supply currents that include the load increase waveform 93 is treated as excluded from abnormality determination. This allows the abnormality determination unit 28 to accurately determine whether there is an abnormality in the motor 7.

[0057] The waveform determination units 30A, 30B, and 30C described above may be combined in any manner. For example, the configuration of waveform determination unit 30B for detecting load decrease waveform 96 may be additionally applied to waveform determination unit 30A, and the configuration of waveform determination unit 30C for detecting load increase waveform 93 may be additionally applied to waveform determination unit 30A.

[0058] <History Data Generator 25> Returning to Figure 2, the motor abnormality monitoring device 1 may further include a history data generator 25 configured to generate history data Dh. The history data Dh indicates the operating state of the motor 7 corresponding to the peculiar waveform 90 that has been excluded from the target of abnormality determination by the data processing unit 29. The history data generator 25 generates the history data Dh by referencing the peculiar waveform 90 detected by the waveform determination unit 30 and the time-series data D1 deleted by the data processing unit 29.

[0059] For example, if the waveform determination unit 30 has a function of detecting a load decrease waveform 96 and a load increase waveform 93, there are three types of time-series data D1 to be deleted by the data processing unit 29. First, time-series data D1 including the load decrease waveform 96, second, time-series data D1 including the load increase waveform 93, and third, time-series data D1 including the load decrease waveform 96 and the load increase waveform 93 are to be deleted by the data processing unit 29. The history data generation unit 25 identifies these three types of time-series data D1 to be deleted by the data processing unit 29 based on the determination result of the waveform determination unit 30. In this example, data indicating the identification result is added to each piece of time-series data D1, thereby generating history data Dh.

[0060] The historical data Dh is referenced by an operator. By referencing the historical data Dh, the operator can confirm the detection frequency of each of the load decrease waveform 96 and the load increase waveform 93. The operator may reset the first coefficient and the second coefficient based on the confirmation results in order to optimize the detection sensitivity of these peculiar waveforms 90. If the first coefficient increases, the detection sensitivity of the load decrease waveform 96 increases, and if the second coefficient increases, the detection sensitivity of the load increase waveform 93 decreases. The reset commands for each of the first coefficient and the second coefficient input by the operator are accepted by the motor abnormality monitoring device 1.

[0061] According to the above configuration, the operator can check whether the determination made by the waveform determination unit 30 regarding the presence or absence of the peculiar waveform 90 was appropriate by checking the history data Dh generated by the history data generation unit 25. This allows the operator to change the setting values ​​(first coefficient and second coefficient) for detecting the peculiar waveform 90 that are set in the motor abnormality monitoring device 1.

[0062] <Abnormality Monitoring Process for Motor 7> The abnormality monitoring process (abnormality monitoring method) for the motor 7 will be described with reference to Figures 14 and 15. This process is executed by the processor of the motor abnormality monitoring device 1. Below, an example of abnormality monitoring process that detects a load decrease waveform 96 (see Figure 9) and a load increase waveform 93 (see Figure 13) will be described. Also, below, "step" may be abbreviated as "S".

[0063] 14, first, the processor executes a current time-series data generating step of generating current time-series data Da (S11). The processor executing S11 is an example of the time-series data generating unit 15.

[0064] Next, the processor executes a waveform determination step of determining whether the waveform 9 of the power supply current indicated by the current time-series data Da includes a peculiar waveform 90 that is not subject to abnormality monitoring (S13). The processor executing S13 is an example of the waveform determination unit 30.

[0065] 15 , the waveform determination step will be described in detail. The processor executes a first acquisition step of acquiring first reference values ​​for the plurality of divided time series data Dv included in the current time series data Da (S31). The processor executing S31 is an example of the first acquisition unit 21. Next, the processor executes an identification step of identifying a plurality of fundamental waveforms 95, each of which is regarded as one cycle of the power supply current, for the plurality of divided time series data Dv included in the current time series data Da (S33). The processor executing S33 is an example of the identification unit 23.

[0066] Next, the processor determines whether the waveform 9 for each of the plurality of divided time-series data Dv includes a load decrease waveform 96 (S35). The processor executing S35 is an example of the load decrease waveform determination unit 36. If it is determined that the load decrease waveform determination unit 36 ​​is not included (S35: NO), the processor proceeds to S39. If it is determined that the load decrease waveform determination unit 36 ​​is included (S35: YES), the processor regards the divided time-series data Dv determined to include the load decrease waveform determination unit 36 ​​as time-series data D1 (S37). In other words, time-series data D1 is generated.

[0067] Next, the processor executes a second acquisition step of acquiring second reference values ​​for the plurality of divided time-series data Dv included in the current time-series data Da (S39). At this time, the second reference values ​​are also acquired for the divided time-series data Dv that were regarded as the time-series data D1 in S37. The processor executing S39 is an example of the second acquisition unit 22.

[0068] Next, the processor determines whether the waveform 9 of each of the multiple divided time-series data Dv includes a load increase waveform 93 (S41). The determination targets of S41 include the time-series data D1 generated in S37. The processor executing S41 is an example of the load increase waveform determination unit 35. If it is determined that the load increase waveform 93 is not included (S41: NO), the processor terminates the peculiar waveform determination process and proceeds to S15 of FIG. 14. If it is determined that the load increase waveform 93 is included (S41: YES), the processor regards the divided time-series data Dv determined to include the load increase waveform 93 as time-series data D1 (S43). In other words, the time-series data D1 is generated. Here, if the load increase waveform 93 is included in the time-series data D1 generated in S37, data processing indicating that two peculiar waveforms 90 are included is performed on the time-series data D1. Thereafter, the processor terminates the peculiar waveform determination process.

[0069] 14 , the processor executes a data processing step (S15). In S15, if it is determined in S13 that at least one of the load decrease waveform 96 and the load increase waveform 93 is included in the power supply current waveform 9, data processing is executed to exclude the time series data D1 generated in S37 and S43 from the abnormality determination. More specifically, the processor executes processing to delete the time series data D1. Note that if neither the load decrease waveform 96 nor the load increase waveform 93 is detected in S13, S15 is skipped. The processor executing S15 is an example of the data processing unit 29.

[0070] Next, the processor executes an abnormality determination step of determining whether or not there is an abnormality in the motor 7 based on the plurality of divided time-series data Dv (S17). If it is determined that there is an abnormality, the processor may execute a display control process of displaying an error message on a monitor provided in the motor abnormality monitoring device 1. The processor executing S17 is an example of the abnormality determination unit 28.

[0071] Next, the processor executes a history data generation step (S19) for generating history data Dh. In this example, the history data Dh indicates the operating state of the motor 7 corresponding to the load decrease waveform 96 or the load increase waveform 93 that was excluded from the abnormality determination target in S19. The generated history data Dh may be displayed on the monitor of the motor abnormality monitoring device 1. Note that if neither the load decrease waveform 96 nor the load increase waveform 93 is detected in S13, S19 is skipped. The processor that executes S19 is an example of the history data generation unit 25.

[0072] Next, the processor executes a first reception step (S21) of receiving an adjustment command for the first coefficient to be executed based on the generated history data Dh, and a second reception step (S23) of receiving an adjustment command for the second coefficient to be executed based on the generated history data Dh. In S21 and S23, an operator who has checked the history data Dh inputs an adjustment command to the motor abnormality monitoring device 1, and the adjustment command is accepted by the processor. The processor executes control to change the first coefficient and the second coefficient based on the accepted adjustment command. Thereafter, this control process ends.

[0073] According to the above configuration, the first and second coefficients as set values ​​can be adjusted based on the generated history data Dh, allowing the operator to change the first and second coefficients so that the load decrease waveform 96 and the load increase waveform 93 are more appropriately detected.

[0074] <Modification> The time-series data generator 15 does not have to generate multiple divided time-series data Dv from the current time-series data Da. In this case, the waveform determiner 30 may determine whether the waveform 9 includes the peculiar waveform 90 based on the current time-series data Da. Furthermore, the data processing performed by the data processor 29 to exclude data from the target of anomaly determination is not limited to deletion processing. For example, a data table indicating time periods in which the peculiar waveform 90 occurs among the time periods indicated by the current time-series data Da may be separately generated. The anomaly determiner 28 identifies normal time periods of the current time-series data Da excluding the time periods indicated in the data table. The anomaly determiner 28 may determine whether an anomaly exists in the motor 7 based on the spectrum data for the normal time periods.

[0075] Furthermore, in the above embodiment, an example has been described in which the motor 7 is the target of abnormality monitoring, but the generator may be the target of monitoring instead of the motor 7. If the waveform 9 drawn by the generator's AC current (i.e., the AC current generated by power generation) includes a peculiar waveform 90, the peculiar waveform 90 is treated as an exception to the target of abnormality determination. This makes it possible to accurately determine whether there is an abnormality in the generator. The motor 7 and the generator are examples of the "rotating electric machine" of the present invention.

[0076] <Summary> The contents described in the above-described embodiments can be understood, for example, as follows.

[0077] 1) An abnormality monitoring device for a rotating electric machine (motor abnormality monitoring device 1) according to one embodiment of the present disclosure is an abnormality monitoring device for a rotating electric machine including an abnormality determination unit (28) for determining whether or not an abnormality exists in the rotating electric machine based on current time series data (Da) indicating a change over time in an AC current (power supply current) in the rotating electric machine (motor 7 or generator), and including: a waveform determination unit (30) for determining whether or not a peculiar waveform (90) indicating an operating state of the rotating electric machine that is not subject to abnormality monitoring is included in a waveform (9) of the AC current indicated by the current time series data; and a data processing unit (29) for executing data processing such that, when it is determined that the peculiar waveform is included, time series data (D1) of a non-target AC current (non-target power supply current) including the AC current corresponding to the peculiar waveform is not subject to abnormality determination by the abnormality determination unit.

[0078] According to the configuration 1), when the AC current waveform indicated by the current time-series data contains a peculiar waveform indicating an operating state of the rotating electric machine that is not subject to abnormality monitoring, the time-series data of the non-target AC current, including the AC current corresponding to the peculiar waveform, is treated as not subject to abnormality detection by the abnormality determination unit. This allows the abnormality determination unit to determine the presence or absence of an abnormality based on the current time-series data suitable for abnormality detection. This realizes an abnormality monitoring device for a rotating electric machine that can accurately determine whether an abnormality exists in the rotating electric machine.

[0079] 2) In some embodiments, in the abnormality monitoring device for a rotating electric machine described in 1) above, the peculiar waveform has a sudden waveform (98) that indicates a startup operating state of a motor (7) as the rotating electric machine, or an operating state in which a load on the motor temporarily increases, and the waveform determination unit includes a sudden waveform determination unit (30) configured to determine that the waveform of the AC current includes the sudden waveform when the AC current indicated by the current time series data includes a sudden current whose absolute value exceeds a first specified threshold value.

[0080] According to the inventor's findings, inrush currents that can be measured while the motor is in a start-up operating state, or currents that change suddenly due to an increase in the motor load (load fluctuation), have very high absolute values, so AC currents that include sudden currents can hinder accurate abnormality detection. In this regard, with the configuration of 2) above, time-series data of asymmetric AC currents corresponding to sudden waveforms is treated as outside the scope of abnormality detection. This allows the abnormality detection unit to accurately determine whether there is an abnormality in the motor.

[0081] 3) In some embodiments, in the abnormality monitoring device for a rotating electric machine described in 1) or 2) above, the peculiar waveform has a start-stop waveform (97) that indicates either a started operating state or a stopped operating state of a motor (7) as the rotating electric machine, and the waveform determination unit includes a start-stop waveform determination unit (30) that determines that the waveform of the AC current includes the start-stop waveform when the AC current indicated by the current time series data includes a weak current whose absolute value is equal to or less than a second specified threshold value for a predetermined period of time or more.

[0082] According to the inventor's findings, weak currents that can be measured while the motor is in a start-up or stop-up state have very small values ​​for a predetermined period of time or more, so AC currents that include weak currents can hinder accurate abnormality detection. In this regard, with the configuration of 3) above, time-series data of asymmetric AC currents that include start-up and stop waveforms is treated as excluded from abnormality detection. This allows the abnormality detection unit to accurately determine whether there is an abnormality in the motor.

[0083] 4) In some embodiments, in the abnormality monitoring device for a rotating electric machine described in 1) or 3) above, the peculiar waveform has a load reduction waveform (96) that indicates the operating state of a motor (7) as the rotating electric machine in which a load is temporarily reduced during operation (which may be, but is not limited to, rated operation), and the waveform determination unit includes: a first acquisition unit (21) for acquiring a first reference value obtained by multiplying a peak value of the AC current indicated by the current time series data by a first coefficient that is smaller than 1; and a load reduction waveform determination unit (30) that is configured to divide a time period indicated by the current time series data into a plurality of basic time periods corresponding to one cycle of the AC current, and that is configured to determine that the waveform of the AC current includes the load reduction waveform when the AC current in any of the plurality of basic time periods includes a low current whose maximum absolute value is smaller than the first reference value.

[0084] According to the inventor's findings, low currents that can be measured when the load momentarily decreases during operation have a smaller absolute value than normal AC currents measured while a load (which may be, but is not limited to, a rated load) is being applied. Therefore, AC currents that include low currents can hinder accurate abnormality detection. In this regard, with the configuration of 4) above, time series data of asymmetric AC currents that include a load decrease waveform are treated as excluded from abnormality detection. This allows the abnormality detection unit to accurately determine whether there is an abnormality in the motor.

[0085] 5) In some embodiments, the abnormality monitoring device for a rotating electric machine described in 4) above further comprises a time series data generating unit (15) for generating the current time series data by performing an approximation process on the measurement time series data indicating the change over time in the measurement value of the AC current, in which the AC current included in a dead band region (R) where the absolute value is equal to or less than a third specified threshold is set to 0.

[0086] According to the configuration of 5) above, a measured value of AC current that fluctuates around 0 is set to 0 in the current time-series data. This allows the load decrease waveform determination unit to accurately identify the current equivalent to one cycle included in the AC current indicated by the current time-series data.

[0087] 6) In some embodiments, the abnormality monitoring device for a rotating electric machine described in 5) above further comprises an identification unit (23) for identifying a plurality of fundamental waveforms (95) each regarded as one cycle of the AC current in the waveform of the AC current indicated by the current time series data, and when the identified plurality of fundamental waveforms includes a peculiar fundamental waveform (94) having a period that is less than 55% of one cycle of the AC current, the load reduction waveform determination unit is configured to exclude the peculiar fundamental waveform from the targets for determining whether the low current is included.

[0088] According to the inventor's findings, even if the identification unit identifies multiple fundamental waveforms based on current time-series data that has undergone approximation processing, the fundamental waveforms may contain peculiar current waveforms that are inappropriate for determining whether a low current is present. This is because the measured AC current values ​​included in the measured time-series data fluctuate near the dead zone. Therefore, the period of the peculiar current waveform is approximately 50% of one AC current period. In this regard, according to the configuration of 6) above, fundamental waveforms having a period less than 55% of one AC current period are considered peculiar fundamental waveforms and are excluded from the determination of whether a low current is present. This allows the load reduction determination unit to accurately classify the time period indicated by the current time-series data into basic time periods corresponding to one AC current period. This allows accurate determination of whether the AC current waveform contains a load reduction waveform. This measure is particularly effective for inverter power supplies.

[0089] 7) In some embodiments, in the abnormality monitoring device for a rotating electric machine described in any one of 1) to 6) above, the peculiar waveform has a load increase waveform (93) that indicates the operating state of a motor (7) as the rotating electric machine in which a load temporarily increases during operation (which may be during rated operation, but is not limited to this), and the waveform determination unit includes: a second acquisition unit (22) that acquires a second reference value obtained by multiplying a peak value of the AC current indicated by the current time series data by a second coefficient that is greater than 1; and a load increase waveform determination unit (30) that determines that the waveform of the AC current includes the load increase waveform when the AC current indicated by the current time series data includes a high current whose absolute value exceeds the second reference value.

[0090] According to the inventor's findings, high currents that can be measured when the load momentarily increases during operation have a larger absolute value than normal AC currents measured while a load (which may be, but is not limited to, a rated load) is occurring. Therefore, AC currents that include high currents can hinder accurate abnormality detection. In this regard, with the configuration of 7) above, time series data of asymmetric AC currents that include a load increase waveform are treated as excluded from abnormality detection. This allows the abnormality detection unit to accurately determine whether there is an abnormality in the motor.

[0091] 8) In some embodiments, the abnormality monitoring device for a rotating electric machine described in any one of 1) to 7) above further includes a history data generation unit (25) for generating history data (Dh) indicating the operating state of the rotating electric machine corresponding to the peculiar waveform that has been excluded from the target of the abnormality determination by the data processing unit.

[0092] According to the configuration of 8) above, an operator of the rotating electric machine can check the history data generated by the history data generator to confirm whether the determination made by the waveform determiner regarding the presence or absence of a peculiar waveform was appropriate. This allows the operator to change the setting value for detecting a peculiar waveform that is set in the anomaly monitoring device for the rotating electric machine.

[0093] 9) An abnormality monitoring method for a rotating electric machine according to at least one embodiment of the present disclosure is an abnormality monitoring method for a rotating electric machine, comprising: an abnormality determination step (S17) for determining whether or not an abnormality exists in the rotating electric machine based on current time series data (Da) indicating a change over time in an AC current (power supply current) in the rotating electric machine (e.g., a motor 7), and comprising: a waveform determination step (S13) for determining whether or not a peculiar waveform (90) indicating an operating state of the rotating electric machine that is not subject to abnormality monitoring is included in a waveform (9) of the AC current indicated by the current time series data; and a data processing step (S15) for performing data processing to exclude time series data (D1) of a non-target AC current, which is the AC current that includes the peculiar waveform, from the abnormality determination in the abnormality determination step when it is determined that the peculiar waveform is included.

[0094] The configuration 9) above provides the same technical advantages as the configuration 1).

[0095] 10) In some embodiments, there is provided the abnormality monitoring method for a rotating electric machine described in 9) above, wherein the peculiar waveform has a load reduction waveform that indicates the operating state of the rotating electric machine as a motor (7) in which the load is temporarily reduced during rated operation, and the waveform determination step includes: a first acquisition step (S31) of acquiring a first reference value obtained by multiplying a peak value of the AC current indicated by the current time series data by a first coefficient that is smaller than 1; and a load reduction waveform determination step (S35) of dividing the time period indicated by the current time series data into a plurality of basic time periods corresponding to one cycle of the AC current, and determining that the waveform of the AC current includes the load reduction waveform when the AC current in any of the plurality of basic time periods includes a low current whose maximum absolute value is smaller than the first reference value.

[0096] The configuration 10) above provides the same technical advantages as the configuration 4).

[0097] 11) In some embodiments, the abnormality monitoring method for a rotating electric machine described in 10) above further includes a history data generation step (S19) for generating history data indicating the operating state of the motor (7) corresponding to the peculiar waveform that has been excluded from the target of the abnormality determination by the data processing step, and a first reception step (S21) for receiving an adjustment command for the first coefficient that is executed based on the generated history data.

[0098] According to the configuration of 11), the first coefficient as a set value can be adjusted based on the generated history data, and thus the motor operator can change the first coefficient so that the load reduction waveform is more appropriately detected.

[0099] 12) In some embodiments, in the method for monitoring an abnormality for a rotating electric machine described in any one of 9) to 11) above, the peculiar waveform has a load increase waveform that indicates the operating state of the rotating electric machine as a motor (7) in which a load has temporarily increased during rated operation, and the waveform determination step includes: a second acquisition step (S23) of acquiring a second reference value obtained by multiplying a peak value of the AC current indicated by the current time series data by a second coefficient greater than 1; and a load increase waveform determination step (S41) of determining that the waveform of the AC current includes the load increase waveform when the AC current indicated by the current time series data includes a high current whose absolute value exceeds the second reference value.

[0100] The configuration 12) above provides the same technical advantages as the configuration 7) above.

[0101] 13) In some embodiments, the abnormality monitoring method for a rotating electric machine described in 12) above further includes a history data generation step (S19) for generating history data (Dh) indicating the operating state of the motor (7) corresponding to the specific waveform that has been excluded from the abnormality determination by the data processing step, and a second reception step (S23) for receiving an adjustment command for the second coefficient that is executed based on the generated history data.

[0102] According to the configuration of 13), the second coefficient as a set value can be adjusted based on the generated history data, and thus the motor operator can change the second coefficient so that the load increase waveform is detected more appropriately.

[0103] DESCRIPTION OF SYMBOLS 1: Motor abnormality monitoring device 2: Power supply equipment 3: A / D converter 4: Electric wire 6: Measuring device 7: Motor 8: Mechanical device 9: Waveform 10: Abnormality monitoring system 15: Time series data generation unit 21: First acquisition unit 22: Second acquisition unit 23: Identification unit 25: History data generation unit 28: Abnormality determination unit 29: Data processing unit 30: Waveform determination unit 35: Load increase waveform determination unit 36: Load decrease waveform determination unit 37: Start / stop waveform determination unit 38: Sudden waveform determination unit 90: Peculiar waveform 93: Load increase waveform 94: Peculiar basic waveform 95: Basic waveform 96: Load decrease waveform 97: Start / stop waveform 98: Sudden waveform D1: Time series data Da: Current time series data Dh: History data Dm: Measurement time series data Dv: Divided time series data L, M0, M1: Dimensions N: Two-dot chain line P1: Rising point P2: Falling point R: Dead zone Sd, Su: Broken line T: Solid line

Claims

1. An abnormal monitoring device for a rotating electrical machine, comprising an abnormality determination unit for determining whether there is an abnormality in the rotating electrical machine based on current time-series data indicating a change over time of an alternating current in the rotating electrical machine, a waveform determination unit for determining whether a specific waveform indicating an operating state of the rotating electrical machine that is not an object of abnormal monitoring is included in the waveform of the alternating current indicated by the current time-series data, and a data processing unit for executing data processing for excluding time-series data of a non-target alternating current including the alternating current corresponding to the specific waveform from being an object of abnormality determination by the abnormality determination unit.

2. The specific waveform has a burst waveform indicating a starting operation state of a motor as the rotating electrical machine or an operating state in which a load temporarily increases on the motor, and the waveform determination unit includes a burst waveform determination unit configured to determine that the burst waveform is included in the waveform of the alternating current when the alternating current indicated by the current time-series data includes a burst current whose absolute value exceeds a first specified threshold value. The abnormal monitoring device for a rotating electrical machine according to claim 1.

3. The specific waveform has a start / stop waveform indicating either a starting operation state or a stopping operation state of a motor as the rotating electrical machine, and the waveform determination unit includes a start / stop waveform determination unit for determining that the start / stop waveform is included in the waveform of the alternating current when the alternating current indicated by the current time-series data includes a weak current whose absolute value is equal to or less than a second specified threshold value over a predetermined time or more. The abnormal monitoring device for a rotating electrical machine according to claim 1 or 2.

4. The specific waveform has a load reduction waveform indicating the operating state of the motor as the rotating electrical machine in which the load temporarily decreases during operation. The waveform determination unit includes: a first acquisition unit for acquiring a first reference value obtained by multiplying the peak value of the alternating current indicated by the current time series data by a first coefficient smaller than 1; and a load reduction waveform determination unit configured to divide the time period indicated by the current time series data into a plurality of basic time periods corresponding to one cycle of the alternating current, and to determine that the load reduction waveform is included in the waveform of the alternating current when a low current having an absolute value smaller than the first reference value is included in the alternating current in any of the plurality of basic time periods. The abnormal monitoring device for a rotating electrical machine according to claim 1 or 2.

5. The abnormal monitoring device for a rotating electrical machine according to claim 4 further includes a time series data generation unit configured to perform an approximation process of setting the alternating current included in the dead zone where the absolute value is equal to or less than a third specified threshold value to 0 with respect to the measurement time series data indicating the change over time of the measured value of the alternating current, thereby generating the current time series data.

6. The abnormal monitoring device for a rotating electrical machine according to claim 5 further includes a specifying unit for specifying a plurality of basic waveforms each regarded as one cycle of the alternating current in the waveform of the alternating current indicated by the current time series data. When the plurality of specified basic waveforms include a specific basic waveform having a period less than 55% with respect to one cycle of the alternating current, the load reduction waveform determination unit is configured to exclude the specific basic waveform from the determination target of whether the low current is included.

7. The specific waveform has a load increase waveform indicating the operating state of the motor as the rotating electrical machine in which the load temporarily increases during rated operation. The waveform determination unit includes: a second acquisition unit for acquiring a second reference value obtained by multiplying the peak value of the alternating current indicated by the current time series data by a second coefficient larger than 1; and a load increase waveform determination unit for determining that the load increase waveform is included in the waveform of the alternating current when a high current having an absolute value exceeding the second reference value is included in the alternating current indicated by the current time series data. The abnormal monitoring device for a rotating electrical machine according to claim 1 or 2.

8. The abnormal monitoring device for a rotating electrical machine according to claim 1 or 2, further comprising a history data generation unit for generating history data indicating the operating state of the rotating electrical machine corresponding to the specific waveform excluded from the target of the abnormal determination by the data processing unit.

9. An abnormal monitoring method for a rotating electrical machine, comprising an abnormal determination step for determining whether there is an abnormality in the rotating electrical machine based on current time-series data indicating the change over time of an alternating current in the rotating electrical machine, the method comprising: a waveform determination step of determining whether the waveform of the alternating current indicated by the current time-series data includes a specific waveform indicating the operating state of the rotating electrical machine that is excluded from the target of abnormal monitoring; and a data processing step of, when it is determined that the specific waveform is included, performing data processing to exclude the time-series data of the non-target alternating current, which is the alternating current including the specific waveform, from the target of the abnormal determination in the abnormal determination step.

10. The specific waveform has a load reduction waveform indicating the operating state of the motor as the rotating electrical machine in which the load has temporarily decreased during operation, and the waveform determination step includes: a first acquisition step of acquiring a first reference value obtained by multiplying the peak value of the alternating current indicated by the current time-series data by a first coefficient smaller than 1; and a load reduction waveform determination step of determining that the load reduction waveform is included in the waveform of the alternating current when a low current having a maximum absolute value smaller than the first reference value is included in any of the alternating currents in the plurality of basic time zones obtained by dividing the time zone indicated by the current time-series data into a plurality of basic time zones corresponding to one cycle of the alternating current. The abnormal monitoring method for a rotating electrical machine according to claim 9.

11. The abnormal monitoring method for a rotating electrical machine according to claim 10, further comprising: a history data generation step for generating history data indicating the operating state of the motor corresponding to the specific waveform excluded from the target of the abnormal determination by the data processing step; and a first reception step of receiving an adjustment command for the first coefficient executed based on the generated history data.

12. The specific waveform has a load increase waveform indicating the operating state of the motor as the rotating electrical machine when the load temporarily increases during operation. The waveform determination step includes: a second acquisition step of acquiring a second reference value obtained by multiplying the peak value of the alternating current indicated by the current time series data by a second coefficient greater than 1; and a load increase waveform determination step of determining that the load increase waveform is included in the waveform of the alternating current when the alternating current indicated by the current time series data includes a high current whose absolute value exceeds the second reference value. The abnormal monitoring method for a rotating electrical machine according to any one of claims 9 to 11.

13. The method further includes a history data generation step for generating history data indicating the operating state of the motor corresponding to the specific waveform excluded from the target of the abnormality determination by the data processing step, and a second reception step of receiving an adjustment command for the second coefficient executed based on the generated history data. The abnormal monitoring method for a rotating electrical machine according to claim 12.

Images