Abnormality diagnosing apparatus and abnormality diagnosing method

Abstract

An abnormality diagnosing apparatus (100) of the present application determines at least either one of an abnormality of a motor (15) driven by an inverter (81) driven at a predetermined operating frequency and an abnormality of a power transmission mechanism (16) that transmits power from the motor (15) to a load (30), performs FFT analysis on a detected current of the motor (15) to analyze extracted spectral peaks, acquires frequencies of spectral peaks caused by noise of the inverter in advance from an operating frequency and a frequency of a sideband wave with respect to the operating frequency, performs FFT analysis on the detected current of the motor at the time of abnormality diagnosis to extract spectral peaks caused by noise of the inverter from the extracted spectral peaks, and then performs abnormality determination.

Description

Technical Field

[0001] This application relates to anomaly diagnostic devices and anomaly diagnostic methods. Background Technology

[0002] Many mechanical devices in the factory are connected to electric motors through power transmission mechanisms. In order to perform maintenance, abnormal diagnoses are made on the electric motors and power transmission mechanisms.

[0003] In Patent Document 1, the applicant disclosed a technology that can detect abnormalities in the power transmission mechanism connected to the electric motor as early, easily, and at low cost by using a current detector without the need for special sensors.

[0004] Furthermore, it is known that in motors driven by commercial power supplies, frequency analysis can be performed on the measured current to diagnose motor abnormalities based on the spectral intensity of sideband waves caused by anomalies generated in frequencies near the power supply frequency (for example, see Patent Document 2).

[0005] Existing technical documents

[0006] Patent documents

[0007] Patent Document 1: Japanese Patent No. 6628905

[0008] Patent Document 2: Japanese Patent No. 6410572 Summary of the Invention

[0009] The technical problem that the invention aims to solve

[0010] In recent years, more and more motors have used inverters in drive control. However, when using the peak values ​​of the current flowing through the motor for anomaly diagnosis, there is a possibility that the peak values ​​caused by the inverter, the power transmission mechanism, and the motor anomaly may overlap. If these peak values ​​overlap, the anomaly diagnosis in existing patent documents 1 and 2 may not be able to be correctly identified.

[0011] This application discloses a technology for solving the above-mentioned problems. Its purpose is to provide an abnormality diagnosis device and an abnormality diagnosis method. When the inverter is used for the drive control of the motor, by extracting the spectral peak caused by the inverter, there is no possibility of incorrect judgment.

[0012] Technical means for solving technical problems

[0013] The anomaly diagnosis device disclosed herein is an anomaly diagnosis device for determining at least one of the following: anomalies in a motor driven by electricity converted by a motor control device, and anomalies in a power transmission mechanism that transmits power from the motor to a load. The motor control device includes an inverter. The anomaly diagnosis device includes: a current detection circuit that detects the current of the motor; and a monitoring and diagnosis unit that uses spectral peaks extracted by performing FFT analysis on the current detected by the current detection circuit to determine the anomaly. The monitoring and diagnosis unit includes: a peak analysis unit that uses the operating frequency of the motor control device used to drive the inverter and the frequency of a sideband wave relative to the operating frequency to... The system performs analysis to infer whether the extracted spectral peaks are caused by inverter noise; an inverter noise frequency determination unit determines the frequency of the spectral peaks caused by inverter noise based on the spectral peaks inferred by the peak analysis unit as being caused by inverter noise; an inverter noise frequency storage unit stores the frequencies of the spectral peaks caused by inverter noise determined by the inverter noise frequency determination unit in advance; and an anomaly determination unit extracts the spectral peaks caused by inverter noise from the spectral peaks extracted by performing FFT analysis on the current detected by the current detection circuit, and then determines an anomaly.

[0014] Invention Effects

[0015] According to the anomaly diagnosis device and anomaly diagnosis method disclosed herein, when an inverter is used for the drive control of a motor, by extracting the spectral peak caused by the noise of the inverter, an anomaly diagnosis device and anomaly diagnosis method that eliminates the possibility of erroneous judgment can be provided. Attached Figure Description

[0016] Figure 1 This is a diagram illustrating a simplified structure of the anomaly diagnosis device according to Embodiment 1.

[0017] Figure 2 This is a diagram showing an example of the structure of the electric motor control device according to Embodiment 1.

[0018] Figure 3 This is a block diagram showing the structure of the monitoring and diagnostic unit according to Embodiment 1.

[0019] Figure 4 This is a graph showing an example of the analysis results of the current FFT using the anomaly diagnostic device according to Embodiment 1, namely a spectrum waveform.

[0020] Figure 5AThis is a flowchart illustrating the overall steps of performing abnormality diagnosis of a power transmission mechanism using the abnormality diagnosis device according to Embodiment 1.

[0021] Figure 5B This is a flowchart illustrating the overall steps of performing abnormality diagnosis of the power transmission function using the abnormality diagnosis device according to Embodiment 1.

[0022] Figure 6 This is a flowchart illustrating the steps of performing anomaly diagnosis using the anomaly diagnosis device according to Embodiment 1.

[0023] Figure 7 This is a block diagram showing another structure of the monitoring and diagnostic unit involved in Embodiment 1.

[0024] Figure 8 This is a flowchart illustrating another step in performing anomaly diagnosis using the anomaly diagnosis device according to Embodiment 1.

[0025] Figure 9 This is a diagram illustrating a simplified structure of the anomaly diagnosis device according to Embodiment 2.

[0026] Figure 10 This is a block diagram showing the structure of the monitoring and diagnostic unit according to Embodiment 2.

[0027] Figure 11 This is a graph showing an example of the analysis results of the current FFT using the anomaly diagnosis device according to Embodiment 2, namely, a spectrum waveform.

[0028] Figure 12A This is a flowchart illustrating the overall steps of performing anomaly diagnosis on the mechanical system of an electric motor using the anomaly diagnosis device according to Embodiment 2.

[0029] Figure 12B This is a flowchart illustrating the overall steps of performing anomaly diagnosis on the mechanical system of an electric motor using the anomaly diagnosis device according to Embodiment 2.

[0030] Figure 13 This is a flowchart illustrating the steps of performing anomaly diagnosis using the anomaly diagnosis device according to Embodiment 2.

[0031] Figure 14 This is a block diagram showing the structure of the monitoring and diagnostic unit according to Embodiment 3.

[0032] Figure 15A This is a flowchart illustrating the overall steps of performing anomaly diagnosis using the anomaly diagnosis device according to Embodiment 3.

[0033] Figure 15BThis is a flowchart illustrating the overall steps of performing anomaly diagnosis using the anomaly diagnosis device according to Embodiment 3.

[0034] Figure 16 This is a diagram illustrating a simplified structure of the anomaly diagnosis device according to Embodiment 4.

[0035] Figure 17 This is a diagram illustrating a simplified structure of the anomaly diagnosis device according to Embodiment 5.

[0036] Figure 18 This is a hardware structure diagram of the anomaly diagnosis device involved in embodiments 1 to 5. Detailed Implementation

[0037] Hereinafter, this embodiment will be described with reference to the accompanying drawings. Furthermore, in each drawing, the same reference numerals denote the same or equivalent parts.

[0038] Implementation method 1.

[0039] Hereinafter, the diagnostic device for the electric motor according to Embodiment 1 will be described with reference to the accompanying drawings.

[0040] Figure 1 This is a diagram illustrating a simplified structure of the fault diagnosis device according to Embodiment 1. The fault diagnosis device 100 typically detects and diagnoses faults in the motor 15 and in the power transmission mechanism 16 that transmits power from the motor 15 to the mechanical equipment 30, which serves as a load. In Embodiment 1, as described later, an example of detecting and diagnosing faults in the power transmission mechanism 16 will be explained. In the diagram, the motor 15 is an example widely used in factories, for example, connected to a power line 11 for driving the motor via a motor control device 80. Circuit breakers 12a, 12b, and 12c are connected to each phase of the power line 11, and electromagnetic contactors 13a, 13b, and 13c are connected to each phase of the power line 11.

[0041] The fault diagnosis device 100 includes a motor control device 80, a monitoring and diagnostic unit 20, a display unit 40, an alarm unit 50, and a current detector 14, which is connected to any one phase of the three-phase power supply line 11 connected to the motor 15. The fault diagnosis device 100 can be mounted on a motor control center for managing multiple motors configured in a factory or similar facility, or it can be a separate motor diagnosis device from the motor control center. The motor control device 80 can be included in the fault diagnosis device 100 or can be installed independently of it.

[0042] The current detector 14 is sometimes installed on each phase of the three-phase power supply line 11. However, it is sufficient to measure only one phase. Furthermore, the installation location of the current detector 14 is not limited as long as it is possible to measure the drive current of the motor 15. This means that the detection accuracy does not change with the measurement location.

[0043] In this embodiment, each motor 15 is equipped with a monitoring and diagnostic unit 20.

[0044] The display unit 40 displays the diagnostic results of the monitoring and diagnostic unit 20. The alarm unit 50 outputs an alarm in an audible or visual manner based on the diagnostic results of the monitoring and diagnostic unit 20, and notifies the motor and power transmission mechanism of any abnormalities.

[0045] Figure 2 This diagram illustrates the structure of the motor control device 80. The motor control device 80 includes an inverter (power conversion device) 81 and a control unit 82 for driving the inverter 81. For example, when the inverter 81 is composed of semiconductor switching elements, the semiconductor switching elements of the inverter 81 are driven and controlled, causing power conversion from a carrier wave and a rectangular wave generated by the control unit 82 via PWM (Pulse Width Modulation) or similar methods. Here, the frequency of the carrier wave used to drive the inverter 81 is the operating frequency f of the motor control device 80. s The power converted by inverter 81 is supplied to motor 15. That is, motor 15 is driven and controlled by inverter 81.

[0046] The power transmission mechanism 16 is constructed by winding, for example, a belt 161, which serves as a power transmission component, onto a pulley Pu1 connected to the rotating shaft of the motor 15 and a pulley Pu2 connected to the drive shaft of the mechanical device 30. The power transmission component is not limited to a belt; it can also be a speed reducer, a chain, etc.

[0047] <Structure of Monitoring and Diagnostic Unit 20>

[0048] Next, the structure of the monitoring and diagnostic unit 20 will be explained. Figure 3 This is a block diagram showing the structure of the monitoring and diagnostic unit 20 of the abnormality diagnostic device 100 according to Embodiment 1. Figure 3 The monitoring and diagnostic unit 20 includes a motor setting unit 21, a memory unit 22, a storage unit 23, an arithmetic unit 25, an anomaly determination unit 27, and a diagnostic result storage unit 28.

[0049] Furthermore, in the abnormality diagnosis device 100 according to Embodiment 1, an example of the structure of the monitoring and diagnosis unit 20 for detecting abnormalities in the power transmission mechanism 16 will be described.

[0050] The motor setting unit 21 is used to set information about the power transmission mechanism 16 and the motor 15. If the power transmission mechanism 16 is a belt, a setting is made to identify that the belt is installed. Alternatively, if the power transmission mechanism 16 is a speed reducer, a setting is made to identify that the speed reducer is installed. If the power transmission mechanism 16 is not present, a setting is made to identify that it is absent.

[0051] Additionally, the motor setting unit 21 is used to obtain the specifications of the motor 15, such as power supply frequency, number of poles, and rated speed, from the information on the nameplate mounted on the motor 15. The rotational frequency of the motor 15 under no-load conditions can be determined using 2·f s / p(f s : The operating frequency of the motor control device, p: number of poles) is calculated. Therefore, the rotational frequency f of motor 15 is... r It must be a value between the no-load rotational frequency and the rated rotational frequency; therefore, the range of rotational frequencies is limited. Using this information, the rotational frequency of motor 15 is determined online in real time with high accuracy, and is used for the detection of mechanical system anomalies in the motor.

[0052] The specifications of these electric motors 15 and the information about their power transmission mechanisms are stored in the memory unit 22. Additionally, the drive current of the electric motors 15, obtained by the current detector 14, is also stored in the memory unit 22.

[0053] The storage unit 23 includes a determination reference storage unit 23a, an inverter noise frequency storage unit 23d, and a power transmission mechanism frequency storage unit 23e.

[0054] The judgment reference storage unit 23a is used to store thresholds, etc., for judging abnormalities of the power transmission mechanism 16.

[0055] The inverter noise frequency storage unit 23d stores the frequency values ​​of the spectral peaks caused by the inverter noise generated from the motor control device 80. In addition to storing the frequency values, it can also store the signal strength of the spectral peaks caused by the inverter noise, as well as the operating frequency of the motor control device 80 and the signal strength of the spectral peaks of the operating frequency. All of these are stored as results analyzed by the calculation unit 25, which will be described later. Alternatively, if the generated frequency is determined in advance, it can be acquired and set by the motor setting unit 21.

[0056] The power transmission mechanism frequency storage unit 23e stores the frequency value of the spectral peak caused by the power transmission mechanism. Preferably, it stores not only the frequency value, but also the signal strength of the spectral peak caused by the power transmission mechanism, the operating frequency of the motor control device 80 when the spectral peak caused by the power transmission mechanism is acquired, and the signal strength of the spectral peak of the operating frequency.

[0057] The computing unit 25 includes a spectrum analysis unit 25a, a sideband wave analysis unit 25b, a peak analysis unit 25c, an inverter noise frequency determination unit 25d, and a power transmission mechanism frequency determination unit 25e.

[0058] The spectrum analysis unit 25a uses the current detected by the current detector 14 to perform current FFT (Fast Fourier Transform) analysis (frequency analysis).

[0059] The sideband wave analysis unit 25b detects all spectral peaks from the spectral waveform analyzed by the spectrum analysis unit 25a. The frequency range detected is preferably between 0 and 1000 Hz. Next, the spectral peaks that satisfy the sideband wave condition are determined from the detected spectral peaks.

[0060] Peak analysis unit 25c analyzes the sideband waves extracted by sideband wave analysis unit 25b according to frequency.

[0061] Based on the results analyzed by the peak analysis unit 25c, the inverter noise frequency determination unit 25d determines whether the sideband wave is a spectral peak caused by inverter noise. If it is determined to be a frequency caused by inverter noise, the frequency value of the spectral peak is stored in the inverter noise frequency storage unit 23d of the storage unit 23. Preferably, the signal strength of the spectral peak, the operating frequency of the motor control device 80 when the spectral peak caused by inverter noise is acquired, and the signal strength of the spectral peak of the operating frequency are also stored simultaneously.

[0062] The power transmission mechanism frequency determination unit 25e determines whether the sideband wave is a spectral peak caused by the power transmission mechanism based on the analysis results from the peak analysis unit 25c. If it is determined that the spectral peak is caused by the power transmission mechanism, the frequency values ​​of these spectral peaks are stored in the power transmission mechanism frequency storage unit 23e. Preferably, the signal strength of the spectral peak, the operating frequency of the motor control device 80 when the spectral peak caused by the power transmission mechanism is acquired, and the signal strength of the spectral peak of the operating frequency are also stored simultaneously.

[0063] The anomaly determination unit 27 determines whether there is an anomaly in the power transmission mechanism 16. Based on the threshold value pre-stored in the determination reference storage unit 23a, it compares it with the spectral peak value determined by the power transmission mechanism frequency determination unit 25e, and makes a determination.

[0064] The diagnostic result storage unit 28 stores the results determined by the anomaly determination unit 27.

[0065] <Analysis of Spectral Peaks in Arithmetic Unit 25>

[0066] Next, the method for analyzing the spectral peaks in the arithmetic unit 25 will be explained.

[0067] Figure 4 This is a graph showing the spectral waveform of the current FFT analysis results. The upper part shows the normal state of the belt 161 of the power transmission mechanism 16, and the lower part shows the broken state of the belt 161.

[0068] exist Figure 4 In the spectrum waveform shown, the operating frequency f of the motor control device 80 is... s Centered on 60Hz, spectral peaks P appear at equal intervals on both the high-frequency and low-frequency sides. That is, from the power supply frequency to the high-frequency side +f... b +2f b Wait, low-frequency side -f b -2f b Spectral peaks P appear at equal intervals. These spectral peaks P are spectral peaks of a sideband wave. The signal strength and occurrence pattern of the spectral peaks P vary according to the rotational speed of the motor 15.

[0069] First, let's explain the reason for the appearance of the spectral peak P. For example, since belt 161 is connected to pulley Pu1, which is connected to the rotating shaft of motor 15, the speed variation of belt 161 causes a variation in the rotor speed of motor 15, which affects the drive current of motor 15. At this time, since the speed variation also occurs at the frequency of one rotation of belt 161, the spectral peak P of the frequency of one rotation of belt 161 and its higher harmonics appears. If D... r Let f be the radius of the pulley Pu1 connected to the rotating shaft of motor 15. r Let L be the rotational frequency of the rotating shaft of motor 15, and let L be the length of belt 161. Then the frequency band where the spectral peak P appears is f. b It is represented by the following formula (1).

[0070] f b =(2πD) r ·f r ) / L···Formula (1)

[0071] Therefore, frequency band f b The radius D of pulley Pu1 r The rotational frequency f of the rotating shaft of motor 15 r The length L of belt 161 is determined. Then, if frequency analysis of the current waveform is performed using FFT, the operating frequency f... s both sides f s ±f b Sideband waves were observed. Simultaneously, higher harmonic components f of the sideband waves were also observed. s ±2f b Based on the operating conditions of motor 15, f was also observed. s ±3f bHigher harmonics.

[0072] Next, in the peak analysis unit 25c, the analysis method for the sideband waves acquired by the sideband wave analysis unit 25b will be explained. The frequency value of the sideband wave acquired by the sideband wave analysis unit 25b, i.e., the value obtained from the operating frequency f, will be used as the peak value. s The frequency of the offset sideband wave is set to f. p Perform the following calculation (2).

[0073] Δ=f s / f p ...Formula (2)

[0074] When the calculated Δ is an integer, it is more likely that the spectral peak is caused by inverter noise. On the other hand, when Δ is not an integer, it is more likely that the spectral peak is caused by a sideband wave from the power transmission mechanism.

[0075] Here, we will explain the case where if Δ is an integer, it is highly likely that the spectral peak is caused by inverter noise.

[0076] In the inverter 81 of the motor control device 80, the frequency of the carrier wave used to determine the switching timing of the switching elements during DC-AC conversion is defined as the operating frequency f. s The timing used to determine the value of the temporarily stored modulated wave is defined by the sampling frequency compared with the carrier wave as f. sm The frequency of the modulating wave is defined as f0. At this time, at the carrier frequency f... s or sampling frequency f sm If the frequency is not a multiple of the modulation wave frequency f0, noise with frequencies having the greatest common divisor of these values ​​and multiples of those values ​​is generated at the spectral peaks. Therefore, at the operating frequency f... s or sampling frequency f sm The inverter noise spectral peaks are generated at the greatest common divisor and higher harmonics of the modulation wave frequency f0. In other words, the operating frequency f of the motor control device... s The frequency f of the spectral peak caused by the generated inverter noise inv Divisible by integer, Δ in equation (2) is an integer.

[0077] exist Figure 4 In the spectrum waveform shown, the operating frequency f of the motor control device 80 is... s Centered on (60Hz here), the spectral peak caused by inverter noise appears at f. s ±f inv f s ±2f inv place, f inv=20 (Hz). Therefore, according to equation (2), Δ = 60 / 20 = 3 is an integer.

[0078] Next, the frequency f of the spectral peak train caused by the power transmission mechanism is analyzed. b Please provide an explanation.

[0079] Frequency f b Calculated according to equation (1). Here, the rotational frequency f of the rotating shaft of motor 15 is... r Calculate according to the following formula (3).

[0080] f r =(2f s ·(1-s)) / p···Equation (3)

[0081] As a result, if the frequency of the spectral peak column caused by the power transmission mechanism is calculated using equation (2), then equation (4) is as shown below.

[0082] Δ=f s / f p =f s / f b

[0083] =f s / ((2πD r ·f r ) / L)

[0084] =pL / (4πD) r ·(1-s))···Equation (4)

[0085] Equation (4) contains the irrational number π (pi) in the denominator, so theoretically Δ will not be an integer.

[0086] Even observation Figure 4 The spectral waveform shown, Δ=f s / f p =f s / f b It is not an integer either.

[0087] As described above, if equation (2) is calculated in the peak analysis unit 25c, it can be determined whether the peak value is caused by inverter noise or by the power transmission mechanism.

[0088] Next, the operation of the inverter noise frequency determination unit 25d will be explained in detail. In the peak analysis unit 25c, for spectral peaks whose calculated Δ is an integer and are judged to be highly likely to be caused by inverter noise (spectral peaks speculated to be caused by inverter noise), the inverter noise frequency determination unit 25d detects whether they are part of a spectral peak sequence. A spectral peak sequence refers to spectral peaks generated at equally spaced frequencies. The condition is that these spectral peaks are sideband waves. If it is determined to be a spectral peak sequence, the frequency value of the spectral peak is stored in the inverter noise frequency storage unit 23d of the storage unit 23. As described above, it is preferable to also store the signal strength of the spectral peaks and the operating frequency of the motor control device 80, and the signal strength of the spectral peaks of the operating frequency.

[0089] Next, the operation of the power transmission mechanism frequency determination unit 25e will be explained in detail. In the peak analysis unit 25c, for spectral peaks where Δ is not an integer and is judged to be highly likely to be caused by the power transmission mechanism, the power transmission mechanism frequency determination unit 25e determines the frequency f of the motor control device 80 to be the same as the operating frequency f of the motor control device 80. s The difference in signal strength between the peak values ​​of the spectrum is checked to see if it is less than or equal to a constant A (dB). The constant A is preferably a value of around 50 (dB) or 60 (dB). That is, it indicates that the signal strength of the peak value caused by the power transmission mechanism is greater than that of its surrounding peak values. If the difference in signal strength is less than or equal to a constant A (dB), the presence of two or more peak value sequences is detected. When a peak value sequence is determined, the frequency values ​​of these peak values ​​are stored in the power transmission mechanism frequency storage unit 23e. As described above, it is preferable to also store the signal strength of the peak values ​​of the spectrum, the operating frequency of the motor control device 80, and the signal strength of the peak values ​​of the operating frequency.

[0090] In the anomaly determination unit 27, it determines whether there is an anomaly in the power transmission mechanism 16. The frequency peak sequence caused by the power transmission mechanism determined by the power transmission mechanism frequency determination unit 25e is compared with the threshold B (dB) pre-stored in the determination reference storage unit 23a, thereby determining that the belt is broken.

[0091] Furthermore, if D or more spectral peaks exceeding the threshold C (dB) pre-stored in the determination reference storage unit 23a are detected, an anomaly is determined. Additionally, the number D of spectral peaks used for anomaly determination is, for example, 2, but not limited to 2, but preferably 2 or more.

[0092] The threshold used for anomaly detection can be pre-stored in the judgment reference storage unit 23a, or it can be determined by storing data of the power transmission mechanism in the memory unit 22 when it is in a normal state and performing statistical processing.

[0093] One method for determining the threshold using statistical processing is to calculate the signal strength of the spectral peak caused by the power transmission mechanism and the operating frequency f of the motor control device 80 by performing learning over a certain period. s The difference in signal intensity between the peak and the peak values ​​of the spectrum is measured by the deviation σ under normal conditions, and a threshold of 3σ is set. In this case, if it exceeds ±3σ, it is judged as abnormal (belt breakage).

[0094] In addition to statistical processing, machine learning methods can also be used to determine the threshold. Alternatively, the behavior of data can be learned in advance from normal data of motors of the same type, and the threshold can be determined based on the learned data. Or, patterns in various data from multiple motors can be learned, these patterns can be classified, and the threshold can be determined based on the classified patterns.

[0095] <Step 1 for abnormal diagnosis of power transmission mechanism 16: Initial learning>

[0096] Next, the abnormality diagnosis method involved in Embodiment 1 will be described with reference to the accompanying drawings.

[0097] Figure 5A as well as Figure 5B This is a flowchart illustrating the steps for diagnosing an abnormality in the power transmission mechanism involved in Implementation 1.

[0098] First, in step S000, the specifications of the electric motor and the information of the power transmission mechanism are input into the electric motor setting unit 21.

[0099] Next, in step S001, the initial learning begins.

[0100] In step S002, the current flowing through the motor 15 is detected by the current detector 14.

[0101] In step S003, the spectrum analysis unit 25a performs current FFT analysis on the detected current to obtain the spectrum waveform.

[0102] In step S004, the sideband wave analysis unit 25b detects all spectral peaks from the spectral waveform and determines the spectral peaks that satisfy the sideband wave condition among the detected spectral peaks as sideband waves (step S005).

[0103] In step S006, the operating frequency f of the motor control device 80 is extracted. s The frequencies of nearby sideband waves. Preferably, for the spectral peaks determined by the sideband wave analysis unit 25b to be sideband waves, the sideband wave frequencies f are extracted sequentially starting from the spectral peaks closest to the operating frequency. p In addition, there are multiple frequencies f. p .

[0104] In step S007, the peak analysis unit 25c calculates the value by adjusting the operating frequency f. s Divide by the sideband frequency f p The obtained value f s / f p .

[0105] In step S008, the calculated f s / f p If the value is an integer ("Yes" in step S008), it is presumed to be a spectral peak caused by inverter noise, and the process proceeds to step S009, where the inverter noise frequency determination section 25d determines whether two or more spectral peak columns are detected.

[0106] If, in step S009, it is determined that no more than two spectral peak columns are detected (in step S009, it is "No"), then return to step S001.

[0107] If it is determined in step S009 that two or more spectral peak columns are detected (Yes in step S009), proceed to step S010, where the frequency f is... inv The frequency of the spectral peak, etc., is stored as the inverter noise frequency in the inverter noise frequency storage section 23d.

[0108] f calculated in step S008 s / f p If the value is not an integer ("No" in step S008), proceed to step S011, where the operating frequency f is calculated by the power transmission mechanism frequency determination unit 25e. s The signal strength I at the peak of the spectrum p (f s ) and sideband wave frequency f p The signal strength I at the peak of the spectrum p (f s ±f p The difference intensity between A and B is determined, and it is determined whether the difference intensity is below a constant A (dB).

[0109] In step S011, in I p (f s )-I p (f s ±f p If A ≤ A (Yes in step S011), proceed to step S012 to determine whether more than two spectral peak columns are detected.

[0110] If it is determined in step S012 that two or more spectral peak columns are detected (yes in step S012), proceed to step S013, where the frequency f is... bThe frequency of the spectral peak, etc., is stored as the frequency of the power transmission mechanism in the power transmission mechanism frequency storage unit 23e.

[0111] Next, if it is determined in step S014 that the inverter noise frequency has been stored in the inverter noise frequency storage unit 23d, then the inverter noise frequency and the power transmission mechanism frequency have been acquired, and the initial learning ends. Additionally, since there are typically multiple sideband wave frequencies f that are spectral peaks caused by the power transmission mechanism... b Therefore, it is best to obtain multiple sets of power transmission mechanism frequencies by repeating steps S011 to S013.

[0112] In step S011, I p (f s )-I p (f s ±f p If the result is greater than A (in step S011, it is "No"), and if in step S012 it is determined that no more than two spectral peak columns were detected (in step S012, it is "No"), then the process returns to step S001 because the frequency of the power transmission mechanism has not been obtained.

[0113] In addition, after obtaining the frequency of the power transmission mechanism, if it is determined in step S014 that the inverter noise frequency is not stored in the inverter noise frequency storage unit 23d (the value in step S014 is "No"), the process returns to step S001.

[0114] exist Figure 4 In the spectrum waveform shown, the upper part represents the normal condition of belt 161, where f appears. s ±f inv f s ±2·f inv f s ±f b and f s ±2·f b The peak value of the spectrum. Here, steps S007 to S014 are performed as initial learning. According to the operating frequency f... s =60 (Hz) sequence, for all sideband wave frequencies f p Calculate f s / f p Determine the frequency f p The frequency f of the spectral peaks caused by inverter noise inv Or is it the frequency f of the spectral peak series caused by the power transmission mechanism? b Specifically, f s 60Hz, f inv For 20Hz, f b It is 13.2Hz, fs / f inv When f = 3, it is an integer value. s / f b =4.55 is not an integer value.

[0115] f s and f inv It is a discrete value. Because it contains a decimal point, it is often not an integer value in calculations performed by microcomputers or other computing devices, but it takes a value very close to an integer. On the other hand, because f s / f b =4.55 is not an integer value, therefore it can be said that by calculating f s / f p It also performs integer value determination, and the accuracy of spectral peak identification is relatively high.

[0116] In f s / f b If the value is also an integer, the calculation can be retried at other operating frequencies, or the load can be changed at the same operating frequency to determine if the spectral peak has shifted. In the latter case, if the operating frequency of the motor control device is constant, theoretically the spectral peak caused by inverter noise will not shift. On the other hand, in the frequency caused by the power transmission mechanism, the slip s changes with the load variation, resulting in a shift in the spectral peak by the amount of slip. By detecting the difference, the spectral peak caused by inverter noise and the spectral peak caused by the power transmission mechanism can be more accurately distinguished.

[0117] In addition, the spectral peaks caused by inverter noise and the spectral peaks caused by the power transmission mechanism both have more than two spectral peak columns.

[0118] In the initial learning process from steps S001 to S014, the inverter noise frequency, power transmission mechanism frequency, and the intensity of their respective spectral peaks are obtained.

[0119] Return to Figure 5A and Figure 5B The flowchart is as follows. In step S014, if it is determined that the inverter noise frequency and power transmission mechanism frequency have been acquired, the initial learning ends, and the anomaly diagnosis begins in step S015. In the anomaly diagnosis, the same current FFT analysis is performed as in the initial learning to extract the spectral peaks caused by the inverter noise acquired in the initial learning, and anomaly diagnosis is performed based on the spectral peaks other than these.

[0120] In step S016, a determination is made as to whether there is an abnormality in the power transmission mechanism. If an abnormality is determined to exist, the result is stored in the diagnostic result storage unit 28, and in step S017, an alarm is triggered and the result is displayed.

[0121] <Step 2 of the abnormality diagnosis for power transmission mechanism 16: Abnormality diagnosis>

[0122] Next, use Figure 6 The flowchart illustrates the steps for abnormal diagnosis in step S015.

[0123] If the diagnosis begins in step S01501, then in step S01502, the current flowing through the motor 15 is detected by the current detector 14.

[0124] In step S01503, the spectrum analysis unit 25a performs current FFT analysis on the detected current to obtain the spectrum waveform.

[0125] In step S01504, the sideband wave analysis unit 25b detects all spectral peaks from the spectral waveform and determines the spectral peaks that satisfy the sideband wave condition among the detected spectral peaks as sideband waves (step S01505).

[0126] Here, steps S01502 to S01505 are the same as steps S002 to S005.

[0127] In step S01505, for the spectral peaks determined to be sideband waves, the frequency of the spectral peaks caused by inverter noise stored in the inverter noise frequency storage unit 23d is used to extract the frequency of the spectral peaks caused by inverter noise in the sideband waves (step S01506).

[0128] In step S01507, for sideband waves whose spectral peaks are not caused by inverter noise, the operating frequency f is determined using information such as the frequency of spectral peaks caused by the power transmission mechanism stored in the power transmission mechanism frequency storage unit 23e. s Is the difference between the signal strength and the signal strength at the location of the frequency peak caused by the power transmission mechanism greater than or equal to a threshold B (dB)?

[0129] Normally, assuming the belt is functioning correctly, as determined in step S011, the intensity I of the spectral peak caused by the power transmission mechanism would be... p (f s ±f p The value is relatively large, compared to the operating frequency f. s Signal strength I p (f s Within a certain range. Therefore, if f is determined based on the spectral waveform obtained during diagnosis... s ±f b Signal strength I at frequency p (f s ±f b ) and fs Signal strength I p (f s If the value is above a predetermined threshold B (dB), the abnormality of the power transmission mechanism can be determined by the abnormality determination unit 27.

[0130] As mentioned above, in Figure 4 In the lower part of the spectrum waveform, located at f s ±f b f s ±2·f b The spectral peak at the location disappears. Therefore, in step S01507, if it is determined to be I... p (f s )-I p (f s ±f b If the value is greater than or equal to B, proceed to step S016, where it is determined that there is an abnormality in the power transmission mechanism, i.e., the belt is broken. If the power transmission mechanism is determined to be abnormal, the result is stored in the diagnostic result storage unit 28, and an alarm is triggered and displayed in step S017.

[0131] The threshold B used in step S01507 can be stored in the determination reference storage unit 23a, and can be the same value as the constant A used in the determination in step S011. However, considering the change of the power transmission mechanism over time, the threshold B is preferably greater than the constant A.

[0132] In step S01507, it is determined to be I. p (f s )-I p (f s ±f b If B < B, the spectral peaks that are not sideband waves detected in step S01504 will be detected as spectral peaks of abnormal frequencies (step S01508).

[0133] If, in step S01509, D or more spectral peaks exceeding the threshold C (dB) pre-stored in the determination reference storage unit 23a are detected (marked as "Yes" in step S01509), an anomaly is determined to exist (step S016). Furthermore, as described above, the number D of spectral peaks used for anomaly determination is, for example, 2, but not limited to 2, but preferably 2 or more.

[0134] Anomaly diagnosis using this abnormal frequency peak can, for example, be performed using the method described in Patent Document 1.

[0135] As described above, steps S01501 to S01509 are repeated to perform anomaly diagnosis of the power transmission mechanism 16.

[0136] Next, using Figure 7 and Figure 8 Explain the operating frequency f of the motor control device 80 during diagnosis. s The situation changed from the time of learning. The operating frequency f of the motor control device 80 is appropriately adjusted according to the motor's driving conditions, etc. s The following includes determining the operating frequency f of the motor control device 80 during diagnosis. s The steps will be explained in detail below.

[0137] Figure 7 It is shown in Figure 3 The block diagram shows the structure of the monitoring and diagnostic unit 20, which also includes the frequency correction unit 25g in the arithmetic unit 25. Figure 8 This is a flowchart of the abnormal diagnosis in a variation of Implementation Method 1, which is a... Figure 6 The flowchart has undergone some changes.

[0138] If the diagnosis begins in step S01501, then in step S01502, the current flowing through the motor 15 is detected by the current detector 14.

[0139] In step S01503, the spectrum analysis unit 25a performs current FFT analysis on the detected current to obtain the spectrum waveform.

[0140] In step S01504, the sideband wave analysis unit 25b detects all spectral peaks from the spectral waveform and determines the spectral peaks that satisfy the sideband wave condition among the detected spectral peaks as sideband waves (step S01505).

[0141] In step S01504, the peak with the strongest signal among the detected spectral peaks is the operating frequency f of the motor control device 80. s The peak value of the spectrum. Therefore, the operating frequency f of the motor control device 80 acquired during learning will be... s Comparing the frequency of the spectrum with the highest signal strength acquired during diagnosis, if the frequencies differ, it is determined that the operating frequency of the motor control device 80 has changed. Conversely, if the operating frequency of the motor control device 80 remains unchanged, then... Figure 6 Diagnose using the steps outlined in the flowchart.

[0142] In step S01506a, the frequency of the frequency correction unit 25g with the highest frequency of the correction signal strength is used as the operating frequency f of the motor control device 80 during diagnosis. s And infer the frequency of the spectral peaks caused by the power transmission mechanism.

[0143] The frequency band f where the spectral peaks caused by the power transmission mechanism appear bEquations (1) and (3) can be used, and the following expression can be used.

[0144] f b =f s ·4πD r (1-s)p / L

[0145] In other words, it is known that f b At operating frequency f s The following changes occurred. Other D... r L and p are constant, only the value of s (sliding) changes slightly, but because its change is so small, it can be considered almost constant. That is, if the operating frequency f is obtained during learning... s If stored, the location of the power transmission mechanism's frequency (frequency band f) can be predicted when the operating frequency changes. b ).

[0146] For example, consider the change in operating frequency f due to variations in the operation of the inverter in the motor control unit 80. s The frequency changes from 60Hz to 30Hz. Operating frequency f s At 60Hz, f b The frequency is 13.2Hz. If the operating frequency f... s If changed to 30Hz, then f b The frequency changes from 13.2 Hz to approximately 6.6 Hz. Therefore, in step S01504, when the operating frequency f is detected... s After changing the frequency to 30Hz, the position of the power transmission mechanism frequency (frequency band f) can be inferred. b In step S01505, the peak value equivalent to f is extracted from the spectral peak value extracted as a sideband wave. b The frequency peak of the spectrum. Then, as performed during learning, the frequency of the power transmission mechanism is determined by determining the frequency of the power transmission mechanism (steps S007 to S012).

[0147] Next, in step S01506b, based on the corrected operating frequency f s (The operating frequency f of the motor control device 80 during diagnosis) s The peak frequency caused by inverter noise is determined, and the peak frequency is extracted from the peak frequency during diagnosis. The same steps as steps S007 to S010 during learning are used to determine the peak frequency based on the corrected operating frequency f. s The spectral peaks caused by inverter noise are sufficient.

[0148] The steps following step S01507 and their usage Figure 6 The steps described are the same.

[0149] Therefore, if the operating frequency f of the motor control device 80 during diagnosis... s In the event of changes occurring since the learning period, based on the corrected operating frequency f s The frequency band f of the power transmission mechanism b The ability to infer and determine spectral peaks caused by inverter noise enables abnormal diagnosis without false detection even when the operating frequency changes.

[0150] In addition, the operating frequency f of the so-called diagnostic motor control device 80 s The change that occurs from the time of learning refers to the operating frequency f that is stored in association with the frequencies stored in the inverter noise frequency storage unit 23d and the power transmission mechanism frequency storage unit 23e. s The operating frequency f of the motor control device 80 during diagnosis s The difference is that the operating frequency f is not present. s The learning results are as follows. Therefore, through repeated learning, the learning will be compared with multiple running frequencies f. s When the corresponding frequencies are stored in the inverter noise frequency storage unit 23d and the power transmission mechanism frequency storage unit 23e, these frequencies can be used. The above situation occurs when there is no learning result, and therefore no data is referenced from the frequencies stored in the inverter noise frequency storage unit 23d and the power transmission mechanism frequency storage unit 23e. For example, even the operating frequency f of the motor control device 80 during the most recent learning... s The operating frequency f of the motor control device 80 during diagnosis s Different, but with the operating frequency f during diagnosis. s If the corresponding frequencies are stored in the inverter noise frequency storage unit 23d and the power transmission mechanism frequency storage unit 23e, and there are already learning results, these frequencies can also be used.

[0151] As described above, according to Embodiment 1, for a motor driven and controlled by an inverter, during initial learning, the current flowing through the motor is detected, and the sideband wave extracted by FFT analysis of the detected current is used to determine the operating frequency f of the motor control device. s , has f s / f p When a sideband wave with an integer frequency forms two or more spectral peaks, it is identified as a sideband wave caused by inverter noise, and its frequency is stored. Therefore, in anomaly diagnosis, the spectral peaks caused by inverter noise can be extracted from the spectral peaks extracted by the FFT analysis of the current, eliminating false diagnoses caused by inverter noise and improving the accuracy of diagnosis.

[0152] Furthermore, in the initial learning, the sideband waves extracted through FFT analysis of the current flowing through the motor have a frequency f relative to the operating frequency f of the motor control device. s f s / f p If the difference between the signal strength of a sideband wave at a non-integer frequency and the signal strength of the operating frequency of the motor control device is less than a constant A, and forms two or more spectral peaks, then it is determined to be a sideband wave caused by the power transmission mechanism, and its frequency is stored. Therefore, in anomaly diagnosis, by comparing the signal strength of the frequency caused by the power transmission mechanism with the stored signal strength, anomalies in the power transmission mechanism, such as belt breakage, can be easily diagnosed.

[0153] Furthermore, during diagnosis, the operating frequency f of the motor control device is determined. s Prior to this, there was no such operation at that frequency f. s Given the learning results, by utilizing the already learned operating frequency f s The sideband wave caused by the power transmission mechanism is inferred, and the operating frequency f of the motor control device during diagnosis is extracted. s The frequency spectrum peaks caused by inverter noise are used as the operating frequency, thus enabling abnormal diagnosis without false diagnoses.

[0154] Implementation method 2.

[0155] Hereinafter, the anomaly diagnosis device according to Embodiment 2 will be described with reference to the accompanying drawings.

[0156] Figure 9 This is a diagram illustrating a simplified structure of the anomaly diagnosis device according to Embodiment 2. The anomaly diagnosis device 100 typically detects and diagnoses anomalies in the motor 15 and in the power transmission mechanism 17 that transmits power from the motor 15 to the mechanical equipment 30, which serves as a load. However, in Embodiment 2, as described below, an example of detecting and diagnosing anomalies in the mechanical system of the motor 15 will be explained. Therefore, the explanation will focus on the differences from Embodiment 1, and similarities will be omitted.

[0157] exist Figure 9 In Embodiment 2, the power transmission mechanism 17 that transmits power from the electric motor 15 to the mechanical device 30 differs from Embodiment 1, which uses belts, reducers, or chains, by using a coupling or similar device that directly transmits the rotation of the electric motor 15 to the mechanical device 30. Therefore, in Embodiment 2, any abnormalities in the power transmission mechanism 17 are considered negligible. Furthermore, the abnormality diagnosis device 100 includes a monitoring and diagnosis unit 20a. One monitoring and diagnosis unit 20a is provided for each electric motor 15.

[0158] <Structure of Monitoring and Diagnostic Unit 20a>

[0159] Next, the structure of the monitoring and diagnostic unit 20a will be described. Figure 10 This is a block diagram showing the structure of the monitoring and diagnostic unit 20a of the abnormality diagnostic device 100 according to Embodiment 2. Figure 10 In the monitoring and diagnostic unit 20a, there are motor setting unit 21, memory unit 22, storage unit 23, arithmetic unit 25, anomaly determination unit 27 and diagnostic result storage unit 28.

[0160] Furthermore, in the abnormality diagnosis device 100 according to Embodiment 2, an example of the structure of the monitoring and diagnosis unit 20a for detecting abnormalities in the motor 15 will be described. The motor setting unit 21 and the memory unit 22 have the same structure as in Embodiment 1, and their descriptions are omitted.

[0161] The storage unit 23 includes a judgment reference storage unit 23a, an inverter noise frequency storage unit 23d, and a mechanical system abnormal frequency storage unit 23f.

[0162] The judgment reference storage unit 23a is used to store thresholds, etc., for judging abnormalities of the mechanical system of the motor 15.

[0163] The inverter noise frequency storage unit 23d has the same structure as in Embodiment 1.

[0164] The mechanical system abnormality frequency storage unit 23f stores the frequency value of the spectral peak caused by the abnormality of the motor's mechanical system. Preferably, it stores not only the frequency value, but also the signal strength of the spectral peak, the operating frequency of the motor control device 80 when the spectral peak is acquired, and the signal strength of the spectral peak of the operating frequency.

[0165] The computing unit 25 includes a spectrum analysis unit 25a, a sideband wave analysis unit 25b, a peak analysis unit 25c, an inverter noise frequency determination unit 25d, and a mechanical system abnormal frequency determination unit 25f.

[0166] The spectrum analysis unit 25a, the sideband wave analysis unit 25b, and the peak analysis unit 25c have the same structure as in Embodiment 1.

[0167] Based on the results analyzed by the peak analysis unit 25c, the mechanical system abnormal frequency determination unit 25f determines whether the frequency of the sideband wave is caused by a mechanical system abnormality of the motor. If it is determined that the frequency is caused by a mechanical system abnormality, the frequency values ​​of these spectral peaks are stored in the mechanical system abnormal frequency storage unit 23f. Preferably, the signal strength of the spectral peaks and the operating frequency of the motor control device 80, as well as the signal strength of the spectral peaks of the operating frequency, are also stored simultaneously.

[0168] The anomaly determination unit 27 determines whether there is an anomaly in the mechanical system of the electric motor 15. Anomalies in the mechanical system of the electric motor 15 include bearing anomalies, eccentricity, misalignment, imbalance, loose bolts, shaking, and abnormal vibration. The determination is made by comparing the threshold value pre-stored in the determination reference storage unit 23a with the spectral peak value determined by the mechanical system anomaly frequency determination unit 25f.

[0169] The diagnostic result storage unit 28 stores the results determined by the anomaly determination unit 27.

[0170] <Analysis of Spectral Peaks in Arithmetic Unit 25>

[0171] The following section explains how to extract abnormal frequencies from the mechanical system of an electric motor.

[0172] Figure 11 This is a graph showing the spectrum waveform of the current FFT analysis result in the calculation unit 25 of the monitoring and diagnostic unit 20a in Embodiment 2. The upper part shows the motor in an unloaded state, and the lower part shows the operating state under rated load. In Embodiment 2, since the power transmission mechanism 17 is a coupling or the like that directly applies rotation to the load, the spectrum peaks caused by the power transmission mechanism are ignored.

[0173] exist Figure 11 In the spectrum waveform shown, both the upper and lower parts are based on the operating frequency f of the motor control device 80. s Centered on the inverter, sideband waves caused by inverter noise were observed at equal intervals on both the high-frequency and low-frequency sides. Furthermore, it is known that, as a sideband wave, at f... s ±f d Abnormal spectral peaks may also appear in the mechanical system of motor 15.

[0174] Next, for the determination Figure 11 The method for determining whether the peak values ​​shown in the spectrum indicate a mechanical system malfunction of the motor 15 will be explained.

[0175] Equation (2) is calculated in the peak analysis unit 25c, but it is known that the spectral peak caused by the mechanical system abnormality of the motor 15 occurs at the motor's rotational frequency f. r Nearby, here, if we focus on the rotational frequency f of the motor in equation (3) r Then equation (2) is transformed into equation (5) below.

[0176] Δ=f s / f p =p / (2·(1-s))···Equation (5)

[0177] At this time, under no-load conditions, the sliding s = 0, and equation (5) is an integer value. Therefore, it is difficult to distinguish it from the spectral peak caused by inverter noise. However, generally speaking, when a load is applied to motor 15, the sliding s is not 0, so when Δ is not an integer, it is highly likely that it is an abnormal frequency of the mechanical system. In addition, since it is known that the spectral peak caused by the mechanical system abnormality of motor 15 appears at the rotational frequency f of motor 15. r The frequency of the spectral peak caused by the mechanical system abnormality can be inferred from the rated information of the motor. From equation (3), the number of poles of motor 15 and the operating frequency f of motor control device 80 can be determined. s The value of the abnormal frequency of the mechanical system is derived from the rotational frequency f of motor 15. r The amount of offset sliding s, but the maximum offset is about 1 to 2 Hz.

[0178] That is, f is calculated by the peak analysis unit 25c according to equation (2). s / f p For sideband waves that are integers, it is determined whether the frequency of the spectral peak changes depending on the presence or absence of a load. If it does not change, the inverter noise frequency determination unit 25d determines whether the spectral peak is caused by inverter noise, similar to Embodiment 1.

[0179] In the abnormal frequency determination unit 25f of the mechanical system, for the f calculated by the peak analysis unit 25c s / f p The value of the spectral peak is not an integer, and even if f is calculated by the peak analysis unit 25c s / f p The value is an integer, and the frequency of the spectral peak varies depending on the presence or absence of a load. The spectral peak value is used to determine whether it is close to the motor's rotational frequency f. r Here, a threshold E (Hz) is preset, and it is determined whether the threshold E (Hz) is within this range. The threshold E can be around 1 to 2 (Hz). The threshold can be stored in the determination reference storage unit 23a, or it can be used as the frequency of the frequency peak calculated by the mechanical system abnormal frequency determination unit 25f and the rotation frequency f of the motor. r The comparison constants used to calculate the difference between the frequencies are stored. When the spectral peak is determined to be close to the motor's rotational frequency f... r At that time, the frequency of the spectrum is stored in the abnormal frequency storage unit 23f of the mechanical system. Preferably, the signal strength of the peak value of the spectrum, the operating frequency of the motor control device 80 when the peak value of the spectrum is acquired, and the signal strength of the peak value of the spectrum at the operating frequency are also stored.

[0180] In the anomaly determination unit 27, it determines whether there is an anomaly in the mechanical system of the motor 15. The signal strength of the spectral peak caused by the mechanical system anomaly determined by the mechanical system anomaly frequency determination unit 25f is compared with the operating frequency f of the motor control device 80. s A mechanical system malfunction is determined when the difference between the spectral intensities is less than F (dB). The threshold F (dB) is pre-stored in the determination reference storage unit 23a.

[0181] The spectral peak value caused by a mechanical system anomaly, as determined by the mechanical system anomaly frequency determination unit 25f, can be compared with the threshold value Fa (dB) pre-stored in the determination reference storage unit 23a. If it exceeds the threshold value Fa (dB), a mechanical system anomaly is determined to exist. Both the threshold value F (dB) and the threshold value Fa (dB) are used to determine if the spectral peak value caused by the mechanical system anomaly is sufficiently large.

[0182] The threshold used for anomaly detection can be pre-stored in the detection reference storage unit 23a, or it can be determined by storing data of the motor 15 in the memory unit 22 when it is in a normal state and performing statistical processing, etc.

[0183] In addition to statistical processing, machine learning methods can also be used to determine the threshold. Alternatively, the behavior of data can be learned in advance from normal data of motors of the same type, and the threshold can be determined based on the learned data. Or, patterns in various data from multiple motors can be learned, these patterns can be classified, and the threshold can be determined based on the classified patterns.

[0184] <Step 1 for diagnosing mechanical system malfunctions of motor 15: Initial learning>

[0185] Next, the anomaly diagnosis method involved in Embodiment 2 will be described with reference to the accompanying drawings.

[0186] Figure 12A as well as Figure 12B This is a flowchart illustrating the steps for diagnosing an abnormality in the mechanical system of the electric motor involved in Embodiment 2.

[0187] From step S000 of inputting the motor specifications to calculating f s / f p Step S007 is the same as in Implementation Method 1, and the description is omitted.

[0188] f calculated by peak analysis unit 25c s / f p If the value is an integer ("Yes" in step S008a), proceed to step S008b to determine the operating state of motor 15 under rated load and the no-load operating state. s / fp The value of has an integer value indicating whether the frequency of the sideband wave changes. Here, it is assumed that motor 15 is controlled at the same operating frequency f under rated load and no-load conditions. s Down.

[0189] In step S008b, if it is determined that there is no change (step S008b is "No"), proceed to step S009 and perform inverter noise frequency determination in the same way as in embodiment 1.

[0190] f calculated in step S008a s / f p If the value is not an integer ("No" in step S008a) and if it is determined in step S008b that there is a change ("Yes" in step S008b), proceed to step S021.

[0191] In step S021, the mechanical system abnormal frequency determination unit 25f determines whether the spectral peak is close to the rotational frequency f of the motor 15. r The frequency spectrum peak was determined to be close to the rotational frequency f of motor 15. r In the case of "Yes" in step S021, the frequency of the spectrum is stored as the abnormal frequency of the mechanical system in the abnormal frequency storage unit 23f of the mechanical system (step S022).

[0192] Next, if it is determined in step S023 that the inverter noise frequency has been stored in the inverter noise frequency storage unit 23d, then the inverter noise frequency and the mechanical system abnormal frequency have been acquired, and the initial learning ends.

[0193] Furthermore, in step S021, it was not determined that the spectral peak was close to the rotational frequency f of the motor 15. r If the condition is "No" in step S021, the process returns to step S001 because the abnormal frequency of the mechanical system was not obtained.

[0194] In addition, after obtaining the abnormal frequency of the mechanical system, if it is determined in step S023 that the inverter noise frequency is not stored in the inverter noise frequency storage unit 23d (the value in step S023 is "No"), the process returns to step S001.

[0195] Next, in step S025, abnormality diagnosis begins.

[0196] In step S026, a determination is made as to whether there are any abnormalities in the mechanical system of the motor. Here, the same current FFT analysis is performed as in the initial learning, but the spectral peaks caused by inverter noise that were acquired in the initial learning are extracted, and anomaly diagnosis is performed based on the spectral peaks other than those obtained in the initial learning.

[0197] If an abnormality is determined to exist, the result is stored in the diagnostic result storage unit 28, and an alarm is triggered and displayed in step S027.

[0198] <Step 2 of the abnormality diagnosis of the mechanical system of motor 15: Abnormality diagnosis>

[0199] Next, use Figure 13 The flowchart illustrates the steps for abnormal diagnosis in step S025.

[0200] Steps S02501 to S02505 are the same as steps S002 to S005.

[0201] In step S02506, the frequency of the spectral peak caused by inverter noise in the sideband wave is extracted from the spectral peak determined to be a sideband wave using information such as the frequency of the spectral peak caused by inverter noise stored in the inverter noise frequency storage unit 23d.

[0202] In step S02507, it is determined whether the spectral peak of the wave identified as a sideband wave is at the rotational frequency f of the motor 15. r The judgment is made based on the vicinity. The judgment is made based on the rotational frequency f of motor 15. r In the vicinity (step S02507 is "Yes"), the possibility of a sideband wave caused by an abnormality in the mechanical system of the electric motor is relatively high, and the frequency f of the spectral peak is detected in step S02508. p and signal strength I p (f s ±f p ).

[0203] In step S02509, the signal strength I of the spectral peak of the sideband wave caused by the mechanical system anomaly of the electric motor is calculated. p (f s ±f p The operating frequency f of the motor control device 80 s Signal strength I p (f s The difference between ) and ). If judged as I p (f s )-I p (f s ±f p If F ≤ F (step S02509 is "yes"), then the peak value of the sideband wave caused by the mechanical system abnormality of the motor is large enough, and it is determined that there is an abnormality in the mechanical system (step S026).

[0204] If it is determined that there is an abnormality in the mechanical system of the electric motor, the result is stored in the diagnostic result storage unit 28, and an alarm is triggered and displayed in step S027.

[0205] Here, in step S02507, the sideband wave also includes spectral peaks caused by inverter noise. Therefore, in the case of abnormal diagnosis, the motor 15 can be in both a loaded and an unloaded operating state, and the steps after step S02507 are only performed on spectral peaks whose frequency changes between the two.

[0206] In addition, the operating frequency f of the motor control device 80 is known in advance. s and the rotational frequency f of the electric motor r In the case where the frequency f of the spectral peak caused by inverter noise is known... inv and rotation frequency f r Whether they overlap can be used to perform steps after step S02507.

[0207] As described above, steps S02501 to S02509 are repeated to perform anomaly diagnosis of the mechanical system of the electric motor 15.

[0208] Next, the judgment Figure 11 Is the spectral peak shown a mechanical system malfunction of motor 15? In the figure, f s =60 (Hz), f r =30 (Hz). First, under no-load conditions at the top, sideband waves are detected from the detected spectral peaks. Next, if f is calculated... s / f p Then, due to f inv =20 (Hz), f d =30 (Hz), so f s / f inv =3,f s / f d =2.

[0209] Here, the rotational frequency f of the electric motor is... r The nearby frequency f has a sideband wave. d The spectral peak, and with and f inv The consistent spectral peaks do not overlap, this f inv With operating frequency f s They are stored together in the inverter noise frequency storage unit 23d. Additionally, they are stored together with the operating frequency f. s Rotation frequency f r f stored together in the abnormal frequency storage section of the mechanical system d Consistent. Therefore, at the motor rotation frequency fr The nearby spectral peaks were identified as abnormal frequencies of the mechanical system, and their signal strength was compared with the operating frequency f. s The signal strength of the peak spectrum is compared. If the difference between the two is below a threshold F, it is considered abnormal.

[0210] Alternatively, a spectrum analysis under rated load conditions can be performed during diagnostics to obtain... Figure 11 The lower part shows the spectrum waveform under rated load conditions. Under rated load conditions, the spectrum peak value changes due to mechanical system anomalies, and the frequency f of the sideband wave... d 28.8Hz, f s / f d =2.08 is not an integer value. Therefore, the value will be determined at the motor's rotational frequency f. r The nearby spectral peaks were identified as abnormal frequencies of the mechanical system, and their signal strength was compared with the operating frequency f. s The signal strength can be compared with the peak values ​​of the spectrum.

[0211] As described above, according to Embodiment 2, for a motor driven and controlled by an inverter, during initial learning, the current flowing through the motor is detected, and the sideband wave extracted by FFT analysis of the detected current is used to determine the operating frequency f of the motor control device. s , has f s / f p If a sideband wave with an integer frequency does not change with the presence or absence of a load and forms two or more spectral peaks, it is determined to be a sideband wave caused by inverter noise, and its frequency, etc., are stored. Therefore, in anomaly diagnosis, the spectral peaks caused by inverter noise can be extracted from the spectral peaks extracted by FFT analysis of the current, eliminating false diagnoses caused by inverter noise and improving the accuracy of diagnosis.

[0212] In addition, during the initial learning, the sideband waves extracted through FFT analysis of the current flowing through the motor are located at the motor's rotational frequency f. r The frequency of a spectral peak that is near the motor and whose frequency varies depending on the presence or absence of a load is identified as a sideband wave caused by a mechanical system malfunction, and its frequency is stored. Therefore, in malfunction diagnosis, it is possible to determine whether a frequency is caused by a mechanical system malfunction of the motor, and compare the signal strength of the spectral peak caused by the mechanical system malfunction with a preset threshold, thereby facilitating the diagnosis of mechanical system malfunctions of the motor.

[0213] Implementation method 3.

[0214] Hereinafter, the anomaly diagnosis device according to Embodiment 3 will be described with reference to the accompanying drawings.

[0215] Figure 14 This is a block diagram showing the structure of the monitoring and diagnostic unit 20b of the abnormality diagnostic device according to Embodiment 3. Figure 14 Shown in Figure 1 In the structure where a power transmission mechanism 16 such as a belt is installed between the electric motor 15 and the mechanical equipment 30, there is a monitoring and diagnostic unit 20b for diagnosing mechanical system abnormalities of both the power transmission mechanism 16 and the electric motor 15.

[0216] Therefore, the storage unit 23 is configured to have a power transmission mechanism frequency storage unit 23e and a mechanical system abnormality frequency storage unit 23f, and the arithmetic unit 25 has both a power transmission mechanism frequency determination unit 25e and a mechanical system abnormality frequency determination unit 25f. The other structures are the same as in embodiments 1 and 2.

[0217] Next, the abnormality diagnosis method involved in Embodiment 3 will be described with reference to the accompanying drawings.

[0218] Figure 15A as well as Figure 15B This is a flowchart illustrating the steps of diagnosing anomalies using the anomaly diagnosis device according to Embodiment 3.

[0219] <Anomaly Diagnosis Step 1: Initial Learning>

[0220] First, the initial learning steps will be explained.

[0221] In step S000, the specifications of the electric motor and the information of the power transmission mechanism are input into the electric motor setting unit 21.

[0222] Figure 15A Steps S001 to S010 are the same as in Embodiment 2.

[0223] In the sideband wave, it is determined to be f in step S008a. s / f p When the value is not an integer, and when the frequency changes due to the presence or absence of load relative to the motor's no-load condition and rated load operating state as determined in step S008b, proceed to step S011.

[0224] In step S011, in I p (f s )-I p (f s ±f p If A ≤ A (Yes in step S011), proceed to step S012 to determine whether more than two spectral peak columns are detected.

[0225] If it is determined in step S012 that two or more spectral peak columns are detected (yes in step S012), proceed to step S013, where the frequency f is... b The frequency of the spectral peak, etc., is stored as the frequency of the power transmission mechanism in the power transmission mechanism frequency storage unit 23e.

[0226] In step S011, I p (f s )-I p (f s ±f p If the frequency of the power transmission mechanism is not determined to be the frequency of the power transmission mechanism (if "No" is determined in step S011) and if it is determined in step S012 that no more than two spectral peak columns are detected (if "No" is determined in step S012), then proceed to step S021.

[0227] In step S021, the mechanical system abnormal frequency determination unit 25f determines whether the spectral peak is close to the rotational frequency f of the motor 15. r The frequency spectrum peak was determined to be close to the rotational frequency f of motor 15. r In the case of "Yes" in step S021, the frequency of the spectrum is stored in the mechanical system abnormal frequency storage unit 23f as the mechanical system abnormal frequency (step S022).

[0228] Next, in step S031, if it is determined that the inverter noise frequency has been stored in the inverter noise frequency storage unit 23d and the frequency caused by the power transmission mechanism has been stored in the power transmission mechanism frequency storage unit 23e, then the inverter noise frequency, the power transmission mechanism frequency and the mechanical system abnormal frequency have been acquired, and the initial learning ends.

[0229] Furthermore, in step S021, it was not determined that the spectral peak was close to the rotational frequency f of the motor 15. r If the condition is "No" in step S021, the process returns to step S001 because the abnormal frequency of the mechanical system was not obtained.

[0230] In addition, after obtaining the abnormal frequency of the mechanical system, in step S023, if it is determined that the inverter noise frequency is not stored in the inverter noise frequency storage unit 23d, or the frequency caused by the power transmission mechanism is not stored in the power transmission mechanism frequency storage unit 23e (in step S031, it is "No"), the process returns to step S001.

[0231] Next, abnormality diagnosis begins in step S035.

[0232] In step S036, a determination is made as to whether there are any abnormalities in the mechanical systems of the power transmission mechanism and the electric motor. Here, the same current FFT analysis is performed as in the initial learning, but the spectral peak value caused by inverter noise, which was obtained in the initial learning, is extracted, and the abnormality diagnosis is performed with reference to this spectral peak value.

[0233] If an abnormality is determined to exist, the result is stored in the diagnostic result storage unit 28, and an alarm is triggered and displayed in step S037.

[0234] In addition, detailed information on abnormal diagnosis is omitted; simply follow the steps for abnormal diagnosis in Implementation Methods 1 and 2.

[0235] As described above, according to Embodiment 3, the same effect is achieved as in Embodiments 1 and 2. That is, for a motor driven and controlled by an inverter, during initial learning, the current flowing through the motor is detected, and the sideband wave extracted through FFT analysis of the detected current is used to determine the operating frequency f of the motor control device. s , has f s / f p If a sideband wave with an integer frequency does not change with the presence or absence of a load and forms two or more spectral peaks, it is identified as a sideband wave caused by inverter noise, and its frequency is stored. Therefore, in anomaly diagnosis, spectral peaks caused by inverter noise can be extracted from the spectral peaks obtained through FFT analysis of the current, eliminating misdiagnosis caused by inverter noise and improving diagnostic accuracy.

[0236] Furthermore, in Embodiment 3, the frequency peaks caused by inverter noise, frequency peaks caused by power transmission mechanism, and frequency peaks of mechanical system abnormalities of the motor control device 80 can be identified respectively, thereby improving the accuracy of diagnosis.

[0237] Implementation method 4.

[0238] Hereinafter, the anomaly diagnosis device according to Embodiment 4 will be described with reference to the accompanying drawings.

[0239] Figure 16 This is a diagram illustrating a simplified structure of the anomaly diagnostic device according to Embodiment 4. In Embodiments 1 to 3, the monitoring and diagnostic units 20 and 20a are each configured to have one unit for each motor 15, but in this embodiment, one monitoring and diagnostic unit monitors multiple motors.

[0240] exist Figure 16In this configuration, multiple motors 15a, 15b, and 15c are connected to power lines 11a, 11b, and 11c for driving the motors via motor control devices 80a, 80b, and 80c, respectively. Current detectors 14a, 14b, and 14c for detecting the current flowing through each motor 15a, 15b, and 15c are configured on at least one power line. Furthermore, circuit breakers 12a, 12b, and 12c for wiring and electromagnetic contactors 13a, 13b, and 13c are respectively installed on each power line 11a, 11b, and 11c.

[0241] The monitoring and diagnostic unit 20 pre-stores specifications for each motor 15a, 15b, 15c, motor control devices 80a, 80b, 80c, and power transmission mechanisms 16a, 16b, 16c that transmit power from each motor 15a, 15b, 15c to mechanical devices 30a, 30b, 30c. Then, the current flowing through each motor 15a, 15b, 15c, detected by current detectors 14a, 14b, 14c, is input to the monitoring and diagnostic unit 20 for current FFT analysis and initial learning. Afterwards, anomaly diagnosis is performed.

[0242] In the structure of Embodiment 4, for example, multiple power transmission mechanisms 16a, 16b, and 16c can be diagnosed simultaneously using a single anomaly diagnostic device 100. Furthermore, for example, if the motors and load devices are of the same type, a comparison of the spectral waveforms can be performed for determination. When the motors and load devices are of the same type, similar waveforms are usually observed. Therefore, anomaly determination can be easily performed by comparing the spectral waveforms from the three current detectors 14a, 14b, and 14c. Additionally, if only a specific spectral waveform shows anomalies, precursors to an anomaly can be predicted. Predefined thresholds can also be preset when performing these comparisons.

[0243] As described above, according to Embodiment 4, since a structure in which one monitoring and diagnostic unit monitors multiple motors is adopted, it not only has the same effects as Embodiments 1 to 3, but also enables comparison of the spectral waveforms obtained from multiple motors and abnormality diagnosis. In addition, the number of monitoring and diagnostic devices is reduced, saving space.

[0244] In addition, three examples of electric motors are given, but not limited to three.

[0245] Implementation method 5.

[0246] Hereinafter, the anomaly diagnosis device according to Embodiment 5 will be described with reference to the accompanying drawings.

[0247] Figure 17This is a diagram illustrating a simplified structure of the anomaly diagnosis device according to Embodiment 5. In Embodiments 1 to 4, the anomaly diagnosis device and the motor control device are independent structures, but as... Figure 17 As shown, the motor control device 80 can also be configured within the fault diagnosis device 100.

[0248] Therefore, the fault diagnosis device and the motor control device are integrated into one unit, so that it is not necessary to ensure separate space for the motor control device and the fault diagnosis device.

[0249] As described above, according to Embodiment 5, while achieving the same effects as Embodiments 1 to 3, the integration of the fault diagnosis device and the motor control device helps to save configuration space.

[0250] Although an example has been shown that includes a motor control device inside the fault diagnosis device, the motor control device may also include a monitoring and diagnostic unit, a display unit, and an alarm unit, thus constituting a motor control device with fault diagnosis function.

[0251] Furthermore, in embodiments 1 to 5 described above, the operating frequency of the motor control device 80 at which the peak spectrum is obtained and the signal strength of the peak spectrum of the operating frequency are the same in the inverter noise frequency storage unit 23d, the power transmission mechanism frequency storage unit 23e, and the mechanical system abnormality frequency storage unit 23f, respectively. However, this is because the frequencies caused by inverter noise, the power transmission mechanism, and the mechanical system abnormalities of the motor will vary depending on the operating frequency of the motor control device 80. By storing the operating frequency of the motor control device 80 at which the peak spectrum is obtained and the signal strength of the peak spectrum of the operating frequency as a set, and using the learning data corresponding to the operating frequency of the motor control device 80 for diagnosis, the possibility of misdiagnosis is eliminated.

[0252] Furthermore, in embodiments 1 to 5 described above, the abnormality diagnosis device 100, as... Figure 18 As shown in the example hardware illustration, it comprises a processor 1000 and a storage device 2000. Although the storage device is not shown, it includes volatile storage devices such as random access memory (RAM) and non-volatile auxiliary storage devices such as flash memory. Alternatively, an auxiliary storage device such as a hard disk can be used instead of flash memory. The processor 1000 executes a program input from the storage device 2000. In this case, the program is input from the auxiliary storage device to the processor 1000 via the volatile storage device. Furthermore, the processor 1000 can output data such as calculation results to the volatile storage device of the storage device 2000, and can also save data to the auxiliary storage device via the volatile storage device.

[0253] This disclosure describes various exemplary embodiments and examples, but the various features, forms and functions described in one or more embodiments are not limited to the application of a specific embodiment, and can be applied to the embodiment alone or in various combinations.

[0254] Therefore, it can be assumed that numerous variations not illustrated are also included within the scope of the technology disclosed in this application. For example, this includes cases where at least one constituent element is modified, added to, or omitted, and cases where at least one constituent element is extracted and combined with constituent elements of other embodiments.

[0255] Label Explanation

[0256] 11, 11a, 11b, 11c: Power lines; 12a, 12b, 12c: Circuit breakers for wiring; 13a, 13b, 13c: Electromagnetic contactors; 14, 14a, 14b, 14c: Current detectors; 15, 15a, 15b, 15c: Motors; 16, 16a, 16b, 16c, 17: Power transmission mechanisms; 20, 20a, 20b: Monitoring and diagnostic units; 21: Motor setting unit; 22: Memory unit; 23: Storage unit; 23a: Judgment reference storage unit; 23d: Inverter noise frequency storage unit; 23e: Power transmission mechanism frequency storage unit; 23f: Mechanical system abnormality frequency storage unit. 25: Calculation unit; 25a: Spectrum analysis unit; 25b: Sideband wave analysis unit; 25c: Peak analysis unit; 25d: Inverter noise frequency determination unit; 25e: Power transmission mechanism frequency determination unit; 25f: Mechanical system abnormal frequency determination unit; 25g: Frequency correction unit; 27: Abnormal determination unit; 28: Diagnostic result storage unit; 30, 30a, 30b, 30c: Mechanical equipment; 40: Display unit; 50: Alarm unit; 80, 80a, 80b, 80c: Motor control device; 81: Inverter; 82: Control unit; 100: Abnormal diagnosis device; 161: Belt; 1000: Processor; 2000: Storage device.

Claims

1. An anomaly diagnostic device, which determines at least one of the following anomalies: an anomaly of a motor driven by electricity converted by a motor control device, and an anomaly of a power transmission mechanism that transmits power from the motor to a load; the anomaly diagnostic device is characterized in that... The motor control device includes an inverter. The abnormality diagnostic device includes: A current detection circuit is provided to detect the current of the motor. as well as The monitoring and diagnostic unit uses the spectral peaks extracted from the current detected by the current detection circuit through FFT analysis to determine anomalies. The monitoring and diagnostic unit includes: The peak analysis unit uses the operating frequency of the motor control device used to drive the inverter and the frequency of the sideband wave relative to the operating frequency to analyze and infer whether the extracted spectral peak is caused by noise from the inverter. The inverter noise frequency determination unit determines the frequency of the spectral peak caused by the inverter noise based on the spectral peak value that the peak analysis unit infers as being caused by the noise of the inverter. The inverter noise frequency storage unit stores in advance the frequencies of the spectral peaks caused by the noise of the inverter as determined by the inverter noise frequency determination unit. Anomaly Detection Department; Frequency determination unit of power transmission mechanism; and Frequency storage unit of power transmission mechanism If the value obtained by dividing the operating frequency of the motor control device by the frequency of the sideband wave relative to the operating frequency is an integer, the peak analysis unit infers that the extracted spectral peak is a first spectral peak caused by the noise of the inverter. If the peak analysis unit deduces that the first spectral peak caused by the inverter noise has two or more spectral peak sequences at the frequency of the sideband wave, the inverter noise frequency determination unit determines that the first spectral peak is a spectral peak caused by the inverter noise, and determines the frequency of the spectral peak caused by the inverter noise. For a second spectral peak whose value is not an integer obtained by the peak analysis unit dividing the operating frequency of the motor control device by the frequency of the sideband wave relative to the operating frequency, and where the difference between the signal strength of the second spectral peak and the signal strength of the spectral peak at the operating frequency is below a preset threshold, and there are two or more spectral peak columns at the frequency of the sideband wave of the second spectral peak, in the above case, the power transmission mechanism frequency determination unit determines that the second spectral peak is a spectral peak caused by the power transmission mechanism, and determines the frequency of the spectral peak caused by the power transmission mechanism. The frequency storage unit of the power transmission mechanism stores in advance the frequency of the spectral peak caused by the power transmission mechanism as determined by the frequency determination unit of the power transmission mechanism. The anomaly determination unit extracts the spectral peak value of the current detected by the current detection circuit by performing FFT analysis, extracts the spectral peak value caused by the inverter noise from the frequency of the sideband wave caused by the inverter noise stored in the inverter noise frequency storage unit, and then uses the frequency of the spectral peak value caused by the power transmission mechanism stored in the power transmission mechanism frequency storage unit to determine the anomaly.

2. An anomaly diagnostic device, which determines at least one of the following anomalies: an anomaly of a motor driven by electricity converted by a motor control device, and an anomaly of a power transmission mechanism that transmits power from the motor to a load; the anomaly diagnostic device is characterized in that... The motor control device includes an inverter. The abnormality diagnostic device includes: A current detection circuit is provided to detect the current of the motor. as well as The monitoring and diagnostic unit uses the spectral peaks extracted from the current detected by the current detection circuit through FFT analysis to determine anomalies. The monitoring and diagnostic unit includes: The peak analysis unit uses the operating frequency of the motor control device used to drive the inverter and the frequency of the sideband wave relative to the operating frequency to analyze and infer whether the extracted spectral peak is caused by noise from the inverter. The inverter noise frequency determination unit determines the frequency of the spectral peak caused by the inverter noise based on the spectral peak value that the peak analysis unit infers as being caused by the noise of the inverter. The inverter noise frequency storage unit stores in advance the frequencies of the spectral peaks caused by the noise of the inverter as determined by the inverter noise frequency determination unit. Anomaly Detection Department; A frequency determination unit for the power transmission mechanism; a frequency storage unit for the power transmission mechanism; and a frequency correction unit. If the value obtained by dividing the operating frequency of the motor control device by the frequency of the sideband wave relative to the operating frequency is an integer, the peak analysis unit infers that the extracted spectral peak is a first spectral peak caused by the noise of the inverter. If the peak analysis unit deduces that the first spectral peak caused by the inverter noise has two or more spectral peak sequences at the frequency of the sideband wave, the inverter noise frequency determination unit determines that the first spectral peak is a spectral peak caused by the inverter noise, and determines the frequency of the spectral peak caused by the inverter noise. For a second spectral peak whose value is not an integer obtained by the peak analysis unit dividing the operating frequency of the motor control device by the frequency of the sideband wave relative to the operating frequency, and where the difference between the signal strength of the second spectral peak and the signal strength of the spectral peak at the operating frequency is below a preset threshold, and there are two or more spectral peak columns at the frequency of the sideband wave of the second spectral peak, in the above case, the power transmission mechanism frequency determination unit determines that the second spectral peak is a spectral peak caused by the power transmission mechanism, and determines the frequency of the spectral peak caused by the power transmission mechanism. The frequency storage unit of the power transmission mechanism stores in advance the frequency of the spectral peak caused by the power transmission mechanism as determined by the frequency determination unit of the power transmission mechanism. The anomaly determination unit extracts the spectral peak value by performing FFT analysis on the current detected by the current detection circuit, and determines the frequency of the spectral peak value with the largest signal strength as the operating frequency of the motor control device during diagnosis. If the frequencies of the spectral peaks caused by the inverter noise and the spectral peaks caused by the power transmission mechanism, corresponding to the operating frequency during the diagnosis, are not stored in the inverter noise frequency storage unit and the power transmission mechanism frequency storage unit, respectively. The frequency correction unit corrects the diagnostic operating frequency to the operating frequency, and, based on the diagnostic operating frequency, infers the frequency of the spectral peaks caused by the power transmission mechanism from the frequency storage unit for the frequencies of the spectral peaks caused by the power transmission mechanism. The anomaly determination unit uses the frequency of the spectral peak caused by the power transmission mechanism, which is inferred from the operating frequency during diagnosis in the frequency correction unit, and the spectral peak caused by the noise of the inverter, which is determined based on the corrected operating frequency, to determine the anomaly.

3. An anomaly diagnostic device, which determines at least one of the following anomalies: an anomaly of a motor driven by electricity converted by a motor control device, and an anomaly of a power transmission mechanism that transmits power from the motor to a load; the anomaly diagnostic device is characterized in that... The motor control device includes an inverter. The abnormality diagnostic device includes: A current detection circuit is provided to detect the current of the motor. as well as The monitoring and diagnostic unit uses the spectral peaks extracted from the current detected by the current detection circuit through FFT analysis to determine anomalies. The monitoring and diagnostic unit includes: The peak analysis unit uses the operating frequency of the motor control device used to drive the inverter and the frequency of the sideband wave relative to the operating frequency to analyze and infer whether the extracted spectral peak is caused by noise from the inverter. The inverter noise frequency determination unit determines the frequency of the spectral peak caused by the inverter noise based on the spectral peak value that the peak analysis unit infers as being caused by the noise of the inverter. The inverter noise frequency storage unit stores in advance the frequencies of the spectral peaks caused by the noise of the inverter as determined by the inverter noise frequency determination unit. Anomaly Detection Department; The electric motor's mechanical system abnormal frequency determination unit; and the electric motor's mechanical system abnormal frequency storage unit. The peak analysis unit presumes that the third spectral peak is caused by noise from the inverter. Specifically, for the third spectral peak, the value obtained by dividing the operating frequency of the motor control device by the frequency of the sideband wave relative to the operating frequency is an integer, and the frequency of the sideband wave relative to the operating frequency does not change regardless of whether the motor is operating with or without the load. If the peak analysis unit deduces that the third spectral peak is caused by the inverter noise, and the third spectral peak has two or more peak sequences at the frequency of the sideband wave of the third spectral peak, the inverter noise frequency determination unit determines that the third spectral peak is a spectral peak caused by the inverter noise, and determines the frequency of the spectral peak caused by the inverter noise. In the mechanical system abnormal frequency determination unit, for the second and fourth spectral peaks, the difference between the frequencies of the second and fourth spectral peaks and the rotational frequency of the motor is compared. If the difference is below a preset threshold, it is determined that the second and fourth spectral peaks are spectral peaks caused by a mechanical system abnormality of the motor. The frequency of the spectral peaks caused by the mechanical system abnormality of the motor is determined. Specifically, for the second spectral peak, the peak analysis unit divides the operating frequency of the motor control device by the frequency of the sideband wave relative to the operating frequency, and the resulting value is not an integer. For the fourth spectral peak, the value obtained by dividing the operating frequency of the motor control device by the frequency of the sideband wave relative to the operating frequency is an integer. Furthermore, the frequency of the sideband wave relative to the operating frequency varies depending on whether the motor is under load. The mechanical system abnormal frequency storage unit stores in advance the frequencies of the spectral peaks caused by mechanical system abnormalities of the electric motor, as determined by the mechanical system abnormal frequency determination unit. The anomaly determination unit extracts the spectral peak value of the current detected by the current detection circuit by performing FFT analysis, and extracts the spectral peak value caused by the noise of the inverter from the frequency of the sideband wave caused by the noise of the inverter stored in the inverter noise frequency storage unit. Then, it uses the frequency of the spectral peak value caused by the mechanical system anomaly of the motor stored in the mechanical system anomaly frequency storage unit to determine the anomaly.

4. An anomaly diagnostic device, which determines at least one of the following anomalies: an anomaly of a motor driven by electricity converted by a motor control device, and an anomaly of a power transmission mechanism that transmits power from the motor to a load; the anomaly diagnostic device is characterized in that... The motor control device includes an inverter. The abnormality diagnostic device includes: A current detection circuit is provided to detect the current of the motor. as well as The monitoring and diagnostic unit uses the spectral peaks extracted from the current detected by the current detection circuit through FFT analysis to determine anomalies. The monitoring and diagnostic unit includes: The peak analysis unit uses the operating frequency of the motor control device used to drive the inverter and the frequency of the sideband wave relative to the operating frequency to analyze and infer whether the extracted spectral peak is caused by noise from the inverter. The inverter noise frequency determination unit determines the frequency of the spectral peak caused by the inverter noise based on the spectral peak value that the peak analysis unit infers as being caused by the noise of the inverter. The inverter noise frequency storage unit stores in advance the frequencies of the spectral peaks caused by the noise of the inverter as determined by the inverter noise frequency determination unit. Anomaly Detection Department; The power transmission mechanism frequency determination unit; the power transmission mechanism frequency storage unit; and the mechanical system abnormal frequency determination unit and the mechanical system abnormal frequency storage unit of the electric motor. The peak analysis unit presumes that the third spectral peak is caused by noise from the inverter. Specifically, for the third spectral peak, the value obtained by dividing the operating frequency of the motor control device by the frequency of the sideband wave relative to the operating frequency is an integer, and the frequency of the sideband wave relative to the operating frequency does not change regardless of whether the motor is operating with or without the load. If the peak analysis unit deduces that the third spectral peak is caused by the inverter noise, and the third spectral peak has two or more peak sequences at the frequency of the sideband wave of the third spectral peak, the inverter noise frequency determination unit determines that the third spectral peak is a spectral peak caused by the inverter noise, and determines the frequency of the spectral peak caused by the inverter noise. In the frequency determination unit of the power transmission mechanism, for the second and fourth spectral peaks, the signal strengths of the second and fourth spectral peaks are compared with the signal strength of the spectral peak at the operating frequency. If the difference is below a preset threshold, and the fifth spectral peak has two or more spectral peak columns at its sideband frequency, it is determined that the fifth spectral peak is a spectral peak caused by the power transmission mechanism. The frequency of the spectral peak caused by the power transmission mechanism is determined. For the second spectral peak, the peak analysis unit divides the operating frequency of the motor control device by the frequency of the sideband wave relative to the operating frequency, and the value obtained is not an integer. For the fourth spectral peak, the value obtained by dividing the operating frequency of the motor control device by the frequency of the sideband wave relative to the operating frequency is an integer. The frequency of the sideband wave relative to the operating frequency varies depending on whether the motor is under load. The frequency storage unit of the power transmission mechanism stores in advance the frequency of the spectral peak caused by the power transmission mechanism as determined by the frequency determination unit of the power transmission mechanism. The mechanical system abnormal frequency determination unit compares the signal strength of the second and fourth spectral peaks with the signal strength of the spectral peak of the operating frequency in the power transmission mechanism frequency determination unit. For the sixth spectral peak whose difference is greater than a preset threshold, and the seventh spectral peak among the fifth spectral peaks that is not determined to be caused by the power transmission mechanism, the mechanical system abnormal frequency determination unit compares the frequency of the sixth and seventh spectral peaks with the difference of the rotational frequency of the motor. If the difference is below the preset threshold, the unit determines that the sixth and seventh spectral peaks are spectral peaks caused by a mechanical system abnormality of the motor, and determines the frequency of the spectral peaks caused by the mechanical system abnormality of the motor. The mechanical system abnormal frequency storage unit stores in advance the frequencies of the spectral peaks caused by mechanical system abnormalities of the electric motor, as determined by the mechanical system abnormal frequency determination unit. The anomaly determination unit extracts the spectral peak value of the current detected by the current detection circuit by performing FFT analysis. It then extracts the spectral peak value caused by the inverter noise from the frequency of the sideband wave caused by the inverter noise stored in the inverter noise frequency storage unit. Finally, it uses the frequency of the spectral peak value caused by the power transmission mechanism stored in the power transmission mechanism frequency storage unit and the frequency of the spectral peak value caused by the mechanical system anomaly of the motor stored in the mechanical system anomaly frequency storage unit to determine the anomaly.

5. The abnormality diagnostic device as described in any one of claims 1 to 4, characterized in that, The power transmission mechanism includes any one of a belt, a reducer, and a chain.

6. The abnormality diagnostic device according to any one of claims 1 to 4, characterized in that, For each of the multiple motors driven by a separate motor control device, a current detection circuit is included to determine abnormalities for each of the multiple motors.

7. The abnormality diagnostic device according to any one of claims 1 to 4, characterized in that, The motor control device is included internally.

8. An anomaly diagnosis method, the method determining at least one of an anomaly in a motor driven by electricity converted by a motor control device having an inverter and driving the inverter, and an anomaly in a power transmission mechanism that transmits power from the motor to a load, the anomaly diagnosis method being characterized by comprising: The learning process consists of five steps: the first step, the second step, the third step, the fourth step, and the fifth step. as well as An abnormal diagnosis procedure includes a sixth step, a seventh step, an eighth step, and an abnormal diagnosis step performed after the eighth step. The first step involves detecting the current of the motor. The second step involves performing FFT analysis on the detected current to detect spectral peaks. The third step extracts the frequency of the sideband wave that is relative to the operating frequency of the motor control device used to drive the inverter from the spectral peak detected in the second step. The fourth step uses the operating frequency and the extracted sideband wave frequency to infer whether the spectral peak is caused by noise from the inverter. The fifth step determines and stores the frequency of the spectral peak caused by the inverter noise based on the spectral peak value that is presumed to be caused by the noise of the inverter. The sixth step involves detecting the current of the motor. The seventh step involves performing FFT analysis on the current detected in the sixth step to detect spectral peaks. The eighth step involves extracting the frequency of the spectral peak detected in the seventh step, which is caused by the noise of the inverter. The fourth step of the learning process includes the following steps: If the value obtained by dividing the operating frequency of the motor control device by the frequency of the sideband wave relative to the operating frequency is an integer, it is presumed that the extracted spectral peak is a first spectral peak caused by the noise of the inverter. as well as For a second spectral peak whose value obtained by dividing the operating frequency of the motor control device by the frequency of a sideband wave relative to the operating frequency is not an integer, and whose signal strength difference between the second spectral peak and the spectral peak at the operating frequency is below a preset threshold, and which has two or more spectral peak columns at the frequency of the sideband wave of the second spectral peak, it is determined that the second spectral peak is a spectral peak caused by the power transmission mechanism, and the frequency of the spectral peak caused by the power transmission mechanism is stored. In the fifth step, if the first spectral peak has two or more spectral peak sequences at the frequency of the sideband wave, it is determined that the first spectral peak is a spectral peak caused by the noise of the inverter, and the frequency of the spectral peak caused by the noise of the inverter is stored. The anomaly diagnosis step includes a step after the eighth step to determine the anomaly by using the frequency of the spectral peak caused by the power transmission mechanism.

9. An anomaly diagnosis method, the method determining at least one of an anomaly in a motor driven by electricity converted by a motor control device having an inverter and driving the inverter, and an anomaly in a power transmission mechanism that transmits power from the motor to a load, the anomaly diagnosis method being characterized by comprising: The learning process consists of five steps: the first step, the second step, the third step, the fourth step, and the fifth step. as well as An abnormal diagnosis procedure includes a sixth step, a seventh step, an eighth step, and an abnormal diagnosis step performed after the eighth step. The first step involves detecting the current of the motor. The second step involves performing FFT analysis on the detected current to detect spectral peaks. The third step extracts the frequency of the sideband wave that is relative to the operating frequency of the motor control device used to drive the inverter from the spectral peak detected in the second step. The fourth step uses the operating frequency and the extracted sideband wave frequency to infer whether the spectral peak is caused by noise from the inverter. The fifth step determines and stores the frequency of the spectral peak caused by the inverter noise based on the spectral peak value that is presumed to be caused by the noise of the inverter. The sixth step involves detecting the current of the motor. The seventh step involves performing FFT analysis on the current detected in the sixth step to detect spectral peaks. The eighth step involves extracting the frequency of the spectral peak detected in the seventh step, which is caused by the noise of the inverter. The fourth step of the learning process includes the following steps: If the value obtained by dividing the operating frequency of the motor control device by the frequency of the sideband wave relative to the operating frequency is an integer, it is presumed that the extracted spectral peak is a first spectral peak caused by the noise of the inverter; and For a second spectral peak whose value obtained by dividing the operating frequency of the motor control device by the frequency of a sideband wave relative to the operating frequency is not an integer, and whose signal strength difference between the second spectral peak and the spectral peak at the operating frequency is below a preset threshold, and which has two or more spectral peak columns at the frequency of the sideband wave of the second spectral peak, it is determined that the second spectral peak is a spectral peak caused by the power transmission mechanism, and the frequency of the spectral peak caused by the power transmission mechanism is stored. In the fifth step, if the first spectral peak has two or more spectral peak sequences at the frequency of the sideband wave, it is determined that the first spectral peak is a spectral peak caused by the noise of the inverter, and the frequency of the spectral peak caused by the noise of the inverter is stored. The eighth step of the abnormality diagnosis process includes the following steps: For the spectral peak detected in the seventh step, the frequency of the spectral peak with the largest signal strength is determined as the operating frequency of the motor control device during diagnosis; In the learning step, without storing the frequencies of the spectral peaks caused by the inverter noise and the spectral peaks caused by the power transmission mechanism corresponding to the operating frequency at the time of diagnosis, the operating frequency at the time of diagnosis is corrected to the operating frequency, and the frequency of the spectral peaks caused by the power transmission mechanism is inferred based on the operating frequency at the time of diagnosis. The operating frequency during the diagnostic process is used to determine the frequency of spectral peaks caused by noise from the inverter; and For the spectral peak detected in the seventh step, the frequency of the spectral peak caused by the inverter noise, determined by using the operating frequency during the diagnosis, is extracted. The abnormality diagnosis step includes a step of determining the abnormality after the eighth step.

10. An anomaly diagnosis method, the method determining at least one of an anomaly in a motor driven by electricity converted by a motor control device having an inverter and driving the inverter, and an anomaly in a power transmission mechanism that transmits power from the motor to a load, the anomaly diagnosis method being characterized by comprising: The learning process consists of five steps: the first step, the second step, the third step, the fourth step, and the fifth step. as well as An abnormal diagnosis procedure includes a sixth step, a seventh step, an eighth step, and an abnormal diagnosis step performed after the eighth step. The first step involves detecting the current of the motor. The second step involves performing FFT analysis on the detected current to detect spectral peaks. The third step extracts the frequency of the sideband wave that is relative to the operating frequency of the motor control device used to drive the inverter from the spectral peak detected in the second step. The fourth step uses the operating frequency and the extracted sideband wave frequency to infer whether the spectral peak is caused by noise from the inverter. The fifth step determines and stores the frequency of the spectral peak caused by the inverter noise based on the spectral peak value that is presumed to be caused by the noise of the inverter. The sixth step involves detecting the current of the motor. The seventh step involves performing FFT analysis on the current detected in the sixth step to detect spectral peaks. The eighth step involves extracting the frequency of the spectral peak detected in the seventh step, which is caused by the noise of the inverter. The fourth step of the learning process includes the following steps: It is presumed that the third spectral peak is caused by noise from the inverter, wherein, for the third spectral peak, the value obtained by dividing the operating frequency of the motor control device by the frequency of the sideband wave relative to the operating frequency is an integer, and the frequency of the sideband wave relative to the operating frequency does not change with or without the load on the motor; and For the second and fourth spectral peaks, the differences between their frequencies and the rotational frequency of the motor are compared. If this difference is below a preset threshold, it is determined that the second and fourth spectral peaks are caused by a mechanical system malfunction of the motor. The frequencies of the spectral peaks caused by the mechanical system malfunction are stored. Specifically, for the second spectral peak, the value obtained by dividing the operating frequency of the motor control device by the frequency of the sideband wave relative to the operating frequency is not an integer. For the fourth spectral peak, the value obtained by dividing the operating frequency of the motor control device by the frequency of the sideband wave relative to the operating frequency is an integer. Furthermore, the frequency of the sideband wave relative to the operating frequency varies depending on whether the motor is under load. In the fifth step, if the third spectral peak has two or more spectral peak sequences at the frequency of the sideband wave, it is determined that the third spectral peak is a spectral peak caused by the noise of the inverter, and the frequency of the spectral peak caused by the noise of the inverter is stored. The anomaly diagnosis step includes a step after the eighth step to determine the anomaly by using the frequency of the spectral peak caused by the mechanical system anomaly of the electric motor.

11. An anomaly diagnosis method, the method determining at least one of an anomaly in a motor driven by electricity converted by a motor control device having an inverter and driving the inverter, and an anomaly in a power transmission mechanism that transmits power from the motor to a load, the anomaly diagnosis method being characterized by comprising: The learning process consists of five steps: the first step, the second step, the third step, the fourth step, and the fifth step. as well as An abnormal diagnosis procedure includes a sixth step, a seventh step, an eighth step, and an abnormal diagnosis step performed after the eighth step. The first step involves detecting the current of the motor. The second step involves performing FFT analysis on the detected current to detect spectral peaks. The third step extracts the frequency of the sideband wave that is relative to the operating frequency of the motor control device used to drive the inverter from the spectral peak detected in the second step. The fourth step uses the operating frequency and the extracted sideband wave frequency to infer whether the spectral peak is caused by noise from the inverter. The fifth step determines and stores the frequency of the spectral peak caused by the inverter noise based on the spectral peak value that is presumed to be caused by the noise of the inverter. The sixth step involves detecting the current of the motor. The seventh step involves performing FFT analysis on the current detected in the sixth step to detect spectral peaks. The eighth step involves extracting the frequency of the spectral peak detected in the seventh step, which is caused by the noise of the inverter. The fourth step of the learning process includes the following steps: It is speculated that the third spectral peak is caused by the noise of the inverter. For the third spectral peak, the value obtained by dividing the operating frequency of the motor control device by the frequency of the sideband wave relative to the operating frequency is an integer, and the frequency of the sideband wave relative to the operating frequency does not change with or without the load of the motor. For the second and fourth spectral peaks, the signal strength of the frequencies of the second and fourth spectral peaks is compared with the signal strength of the spectral peak of the operating frequency. The second and fourth spectral peaks whose comparison results are below a preset threshold are taken as the fifth spectral peak. If there are two or more spectral peak columns at the frequency of the sideband wave of the fifth spectral peak, it is determined that the fifth spectral peak is a spectral peak caused by the power transmission mechanism. The frequency of the spectral peak caused by the power transmission mechanism is stored. For the second spectral peak, the value obtained by dividing the operating frequency of the motor control device by the frequency of the sideband wave relative to the operating frequency is not an integer. For the fourth spectral peak, the value obtained by dividing the operating frequency of the motor control device by the frequency of the sideband wave relative to the operating frequency is an integer. Furthermore, the frequency of the sideband wave relative to the operating frequency varies depending on whether the motor is under load. The signal strengths of the second and fourth spectral peaks are compared with the signal strength of the spectral peak at the operating frequency. For the sixth spectral peak whose difference is greater than a preset threshold, and the seventh spectral peak among the fifth spectral peaks that is not determined to be caused by the power transmission mechanism, the frequencies of the sixth and seventh spectral peaks are compared with the rotational frequency of the motor. If the difference is below the preset threshold, it is determined that the sixth and seventh spectral peaks are spectral peaks caused by a mechanical system malfunction of the motor, and the frequencies of spectral peaks caused by the mechanical system malfunction of the motor are stored. In the fifth step, if the third spectral peak has two or more spectral peak sequences at the frequency of the sideband wave, it is determined that the third spectral peak is a spectral peak caused by the noise of the inverter, and the frequency of the spectral peak caused by the noise of the inverter is stored. The abnormality diagnosis step includes a step after the eighth step to determine the abnormality by using the frequency of the spectral peak caused by the power transmission mechanism and the frequency of the spectral peak caused by the mechanical system abnormality of the electric motor.

Images