Diagnostic device for motor and diagnostic system for motor

Abstract

A diagnostic device for a motor according to the present invention comprises: a storage unit (51E) for storing a monitorable frequency band [F], a monitoring characteristic frequency sequence [G], and a signal strength thereof; and a determination unit (51F) that determines an abnormality of the motor (3) and / or an abnormality of the power transmission mechanism (60) by comparing the monitorable frequency band [F] and the monitoring characteristic frequency sequence [G] stored in the storage unit (51E) with a current spectrum newly acquired from the current detector (4).

Description

Technical Field

[0001] This application relates to a diagnostic device and a diagnostic system for electric motors. Background Technology

[0002] When diagnosing anomalies in motors driven by pulse-width modulation control of power converters based on current, a significant number of spectral peaks originating from the power converter are present in the spectral band used for anomaly diagnosis, compared to cases driven by commercial power supplies. If the frequencies of these spectral peaks overlap with characteristic frequencies that detect motor anomalies, it can lead to false detections by the motor diagnostic device.

[0003] To address this problem, for example, in Patent Document 1 below, the noise predicted to be generated based on the specifications of the motor and power conversion device is theoretically calculated and compared with the spectral peak extracted during anomaly diagnosis, thereby avoiding false detection.

[0004] Specifically, the diagnostic device of Patent Document 1 includes: a detection unit that detects current flowing in a motor; an analysis unit that performs frequency analysis on the current detected by the detection unit and outputs analysis results; a determination unit that determines an abnormality of the motor based on the spectral peak value of at least one sideband wave component in the modulation wave obtained from the analysis results; and a frequency setting unit that presets a noise frequency in the current, wherein the determination unit infers whether there is noise interference in the spectral peak value of the sideband wave component based on the frequency of the sideband wave component and the set noise frequency, and determines an abnormality of the motor. Existing technical documents Patent documents

[0005] Patent Document 1: Japanese Patent No. 6824494 Summary of the Invention The technical problem that the invention aims to solve

[0006] In existing technologies, noise frequencies are theoretically calculated using the modulation wave frequency, carrier frequency, sampling frequency, and power supply frequency in the frequency setting unit, and these frequencies are excluded from the frequencies used for anomaly diagnosis. However, in actual measurements, there is a problem of unpredictable noise occurring in addition to these theoretically calculated values.

[0007] This application discloses a technology for solving the above-mentioned problems, the purpose of which is to provide a diagnostic device and a diagnostic system for electric motors that can diagnose electric motor abnormalities with high accuracy even in the presence of theoretically unpredictable noise. Technical means for solving technical problems

[0008] The diagnostic device for electric motors disclosed in this application is a diagnostic device for diagnosing at least one of the following abnormalities in an electric motor driven by electricity converted by a power conversion device and an abnormality in the power transmission mechanism connected to the electric motor. The diagnostic device includes: A current detector that detects the current flowing from the power conversion device to the motor; and The monitoring and diagnostic unit performs spectrum analysis based on the current detected by the current detector and monitors for at least one of the abnormalities of the motor and the power transmission mechanism. The monitoring and diagnostic unit includes: The input section is for inputting at least the specifications of the power conversion device and the motor. The setting unit calculates, based on the information from the input unit, a characteristic frequency sequence for diagnosing at least one of the abnormalities of the motor and the power transmission mechanism, a monitoring frequency range for monitoring the abnormality, and a noise frequency sequence that can be derived from the specification information of the power conversion device. The analysis unit performs spectral analysis based on the current detected by the current detector and calculates the current spectrum. It extracts a peak frequency sequence with a spectral peak value above a predetermined signal strength within the monitoring frequency range from the current spectrum. It compares the peak frequency sequence with the characteristic frequency sequence and the noise frequency sequence to derive a decision exclusion frequency sequence and a monitoring characteristic frequency sequence. It also derives a monitorable video band obtained by subtracting the decision exclusion frequency sequence from the monitoring frequency range. A storage unit that stores the monitorable video tape, the monitoring characteristic frequency sequence, and its signal strength; and The determination unit determines at least one abnormality, either an abnormality of the motor or an abnormality of the power transmission mechanism, by comparing the monitorable video tape and the monitoring characteristic frequency sequence stored in the storage unit with a current spectrum newly acquired from the current detector. The diagnostic system for electric motors disclosed in this application Diagnostic devices for multiple of the aforementioned motors; It also includes: the database device, which is connected to diagnostic devices for each of the motors; and A diagnostic algorithm modification device is connected to the database device and the diagnostic devices of each of the motors, and modifies the diagnostic algorithm of each of the motor diagnostic devices. Invention Effects

[0009] According to the diagnostic device and system for electric motors disclosed in this application, abnormalities of electric motors can be diagnosed with high accuracy even in the presence of theoretically unpredictable noise. Attached Figure Description

[0010] Figure 1 This is a structural block diagram showing a simplified structure of the power conversion device and the diagnostic device for the electric motor according to Embodiment 1. Figure 2 This is a circuit structure diagram showing the main circuit section of the power conversion device according to Embodiment 1. Figure 3 This is a diagram illustrating the general operation of the control circuit section of the power conversion device according to Embodiment 1. Figure 4 This is a block diagram showing the structure of the monitoring and diagnostic section of the diagnostic device for an electric motor according to Embodiment 1. Figure 5 This is a diagram illustrating the learning process of the diagnostic device for the electric motor according to Embodiment 1. Figure 6 This is a diagram illustrating the learning process of the diagnostic device for the electric motor according to Embodiment 1. Figure 7 In Figure 7 A is a graph showing an example of the spectrum obtained when the electric motor of Embodiment 1 is driven. Figure 7 B is a schematic diagram showing the monitorable video band [F] obtained by subtracting the decision exclusion frequency sequence [E] from the monitor frequency range [B]. Figure 8 This is a diagram illustrating the monitoring and diagnostic process of the diagnostic device for the electric motor according to Embodiment 1. Figure 9 This is a diagram illustrating the monitoring and diagnostic process of the diagnostic device for the electric motor according to Embodiment 1. Figure 10 This is a block diagram showing a simplified structure of the power conversion device and the diagnostic device for the electric motor according to Embodiment 1. Figure 11 This is a block diagram showing the structure of the monitoring and diagnostic section of the diagnostic device for the electric motor according to Embodiment 2. Figure 12 This is a diagram illustrating the learning process of the diagnostic device for the electric motor according to Embodiment 2. Figure 13 This is a diagram illustrating the learning process of the diagnostic device for the electric motor according to Embodiment 2. Figure 14 In Figure 14 A is a graph showing an example of the spectrum obtained when the electric motor of Embodiment 1 is driven. Figure 14B is a diagram schematically showing the monitorable video band [F1] obtained by subtracting the decision exclusion frequency sequence [E1] from the monitor frequency range [B1]. Figure 15 This is a diagram illustrating the monitoring and diagnostic process of the diagnostic device for the electric motor according to Embodiment 2. Figure 16 This is a diagram illustrating the monitoring and diagnostic process of the diagnostic device for the electric motor according to Embodiment 2. Figure 17 This is a diagram illustrating the verification process of the monitoring characteristic frequency sequence [G] in Implementation 3. Figure 18 This is a diagram illustrating the verification process of the monitoring characteristic frequency sequence [G] in Implementation 3. Figure 19 This is a diagram illustrating the process of changing the setting conditions of the proposed power conversion device in Embodiment 4. Figure 20 This is a diagram illustrating the process of changing the setting conditions of the proposed power conversion device in Embodiment 4. Figure 21 This is a diagram illustrating the process of changing the setting conditions of the proposed power conversion device in Embodiment 4. Figure 22 In Figure 22 A is a diagram showing an example of the spectrum obtained when the setting conditions of the power conversion device in Embodiment 4 are changed before driving the motor. Figure 22 B is a schematic diagram showing the monitorable video band [F1] obtained by subtracting the decision exclusion frequency sequence [E1] from the monitor frequency range [B1]. Figure 23 In Figure 23 A is a diagram showing an example of the spectrum obtained when driving the motor after the setting conditions of the power conversion device in Embodiment 4 are changed. Figure 23 B is a schematic diagram showing the monitorable video band [F1] obtained by subtracting the decision exclusion frequency sequence [E1] from the monitor frequency range [B1]. Figure 24 This is a block diagram showing the structure of the monitoring and diagnostic section of the diagnostic device for the electric motor according to Embodiment 5. Figure 25 This is a block diagram showing the structure of the diagnostic system for the electric motor according to Embodiment 5. Figure 26 This is a diagram illustrating an example of the hardware structure of the monitoring and diagnostic unit of the diagnostic device for the electric motor in each embodiment. Detailed Implementation

[0011] Implementation method 1. Figure 1This is a block diagram showing a simplified structure of the power conversion device and the diagnostic device for the electric motor according to Embodiment 1. like Figure 1 As shown, the power conversion device 20 converts the frequency of the AC power from the AC power source 1, such as a commercial power source, and supplies this power to the motor 3. The diagnostic device 50 detects the current of at least one phase of the current supplied to the motor 3 from the power conversion device 20 using a current detector 4, and analyzes the detected current to detect abnormalities in the motor 3. Alternatively, the current detector 4 can be built into the power conversion device 20 or installed externally.

[0012] The power conversion device 20 includes a main circuit section 21 for converting the frequency of power, a control circuit section 22 for operating the main circuit section 21, and a power conversion device setting section 23 for determining the settings of the control circuit section 22.

[0013] The diagnostic device 50 includes: a monitoring and diagnostic unit 51 that monitors and diagnoses abnormalities in the motor 3 based on the current detected by the current detector 4; a display unit 52 that displays the results obtained by the monitoring and diagnostic unit 51; and an alarm unit 53 that issues an alarm when the monitoring and diagnostic unit 51 detects an abnormality. Furthermore, a network output unit can be provided in the display unit 52 and the alarm unit 53 to remotely alert the user to this information.

[0014] Figure 2 This is a circuit structure diagram showing the main circuit section of the power conversion device according to Embodiment 1. like Figure 2 As shown, the main circuit section 21 includes a converter 21A, a smoothing capacitor 21B, and an inverter 21C. The converter 21A converts the alternating current from the AC power source 1 into direct current and stores the direct current in the smoothing capacitor 21B. The inverter 21C converts the direct current stored in the smoothing capacitor 21B into alternating current and supplies the alternating current to the motor 3.

[0015] Converter 21A consists of a three-phase bridge circuit with six diodes Da, and the input lines of each phase are connected to AC power supply 1. Inverter 21C consists of a three-phase bridge circuit with six switching elements Q, each connected in reverse parallel with diodes Db, and the output lines of each phase are connected to motor 3. The switching elements Q may be, for example, IGBTs (Insulated Gate Bipolar Transistors) or MOSFETs (Metal-Oxide-Semiconductor Field Effect Transistors). The switching operation of inverter 21C is controlled by a signal generated by control circuit unit 22. The operation of control circuit unit 22 follows the operating conditions of power conversion device 20 set by the user in power conversion device setting unit 23.

[0016] Furthermore, the structures of converter 21A and inverter 21C are not limited to the structures shown in the figure. In addition, although the main circuit section 21 of the power conversion device 20 is shown to include converter 21A and be connected to AC power supply 1, converter 21A may not be required as long as it is an inverter 21C that converts DC power to AC power and supplies power to motor 3. Furthermore, in this example, the AC power supply 1, the power conversion device 20, and the motor 3 are shown to have, for example, a three-phase structure, but are not limited thereto.

[0017] Figure 3 This is a diagram illustrating the general operation of the control circuit section of the power conversion device according to Embodiment 1. The control circuit section 22 of the power conversion device 20 drives the main circuit section 21 of the power conversion device 20 via pulse width modulation (PWM). PWM modulates the frequency of the power supplied to the motor 3 by extracting the DC voltage of the smoothing capacitor 21B within a short time span. This short time span extraction is achieved by rapidly controlling the switching element Q of the inverter 21C to turn on and off. The control signal G for controlling the switching element Q is generated by the modulation wave generation section 22A, the signal discretization section 22B, and the signal extraction section 22C of the control circuit section 22. The modulation wave generation unit 22A causes the user-input modulation wave frequency f0 to oscillate sinusoidally. The signal discretization unit 22B samples the modulation wave output by the modulation wave generation unit 22A at a frequency fs and acquires discrete data. The control circuit unit 22 compares the discrete data with the carrier frequency fc (e.g., a triangular wave) generated by the signal extraction unit 22C and generates a control signal G.

[0018] exist Figure 3 During the generation of the control signal G shown, noise is generated at frequencies that are integer multiples of the greatest common divisor of the modulation wave frequency f0, the sampling frequency fs, and the carrier frequency fc, i.e., GCD(f0, fs, fc). As shown in equation (1), this noise is defined as Fpwm.

[0019] Fpwm=i·GCD(f0, fs, fc) (1)

[0020] Where i is a positive integer. For example, when f0 = 60Hz, fs = 4000Hz, and fc = 2000Hz, GCD(f0, fs, fc) = 20Hz, and Fpwm is 20Hz, 40Hz, 60Hz, ... Since such low-frequency noise covers the frequency spectrum monitored by the diagnostic device 50, this noise needs to be properly processed.

[0021] In addition to Fpwm mentioned above, noise covering the frequency band monitored by the diagnostic device 50 is also generated during the process of generating DC voltage by the converter 21A of the power conversion device 20. As shown in equation (2), this noise is defined as Fv.

[0022] Fv=|m·fac±n·f0| (2)

[0023] Where fac is the frequency of the power supplied to converter 21A, and m and n are positive integers. For example, when f0 = 60Hz and fac = 50Hz, fv is 10Hz, 20Hz, 30Hz, ... Since such low-frequency noise covers the frequency range monitored by diagnostic device 50, this noise needs to be properly handled.

[0024] Furthermore, as harmonics of the modulating wave, components that are integer multiples of the modulating wave are generated. As shown in equation (3), these are defined as F0.

[0025] F0=k·f0 (3) Where k is a positive integer.

[0026] Other noise generated by the power conversion device 20 includes noise caused by the dead time set to protect the switching elements from damage when generating the control signal G, noise caused by overmodulation due to the amplitude conditions of the modulation wave and carrier wave, and noise generated when extracting the control signal G. If these noises also cover the frequency band monitored by the diagnostic device 50, appropriate processing is required.

[0027] Current detector 4 detects the current generated in power conversion device 20. The detected current is processed by diagnostic device 50. Diagnostic device 50 first learns the normal operating state of motor 3. That is, it performs spectrum analysis on the current when motor 3 is operating normally and stores its characteristics. During diagnosis, it compares the current characteristics with the learned current characteristics.

[0028] When motor 3 malfunctions, the frequency corresponding to each malfunction mode becomes abnormal. Here, as an example, mechanical system malfunctions and rotor rod damage in motor 3 will be described.

[0029] [Mechanical system malfunction] When the mechanical system of motor 3 malfunctions, the rotor vibrates and becomes eccentric. This eccentricity causes the gap length between the rotor and stator to change periodically. The periodic change in gap length alters the electrical characteristics of motor 3 and generates a small disturbance in the current detected by current detector 4. By performing spectral analysis on this disturbance, the sideband wave f0±fm' of the modulation frequency f0 can be identified. Here, fm' is shown in equation (4).

[0030] fm'=((1-s) / p)f0 (4) Where p and s are the pole number and the slip, respectively.

[0031] [Damaged rotor rod] When the rotor rod of motor 3 is damaged, the current in the rotor rod generates an antiphase component (-s·f0). This antiphase component returns a current with a frequency of (1-s)·f0 in the stator. This current generates a torque oscillation with a frequency of 2s·f0 and induces a magnetic flux oscillation of (1±2·s)·f0. As a result, a small disturbance occurs in the current detected by current detector 4. By performing spectral analysis on this disturbance, the sideband wave f0±fr' of the modulation wave frequency f0 can be identified. Wherein, fr' is as shown in equation (5).

[0032] fr'=2·s·f0 (5)

[0033] In addition, as shown in equation (6), the slip s is defined by the rotational speed N0 of the rotating magnetic field and the rotational speed Nr of the rotor. s=(N0-Nr) / N0 (6)

[0034] Furthermore, the rotor speed Nr can be roughly estimated based on the rated speed of the motor 3, and the rotational speed N0 of the rotating magnetic field can be defined by using the number of pole pairs p and the modulation wave frequency f0 as in equation (7). Therefore, the slip s of the motor 3 can be roughly estimated.

[0035] N0=(60 / p)f0 (7)

[0036] The sideband waves of mechanical system abnormalities and rotor rod damage, represented by equations (4) and (5), exist with a certain signal strength in the spectrum detected by the current detector 4 even when the motor 3 is normal. However, when the motor 3 malfunctions, these signal strengths increase. The diagnostic device 50 diagnoses the abnormality of the motor 3 based on this increase in signal strength.

[0037] Figure 4 This is a block diagram showing the structure of the monitoring and diagnostic unit of the diagnostic device for the electric motor according to Embodiment 1. Figure 5 and Figure 6 A diagram showing the learning process of the diagnostic device for the electric motor according to Embodiment 1 is provided. like Figure 4 As shown, the monitoring and diagnostic unit 51 of the diagnostic device 50 includes a detection unit 51A, an analysis unit 51B, an input unit 51C, a setting unit 51D, a storage unit 51E, and a determination unit 51F, which will determine the results based on... Figure 5 and Figure 6 The learning process is used to illustrate the function of each component during the learning process.

[0038] like Figure 5 and Figure 6 As shown, the diagnostic device 50 performs the following steps to diagnose the abnormality of the motor 3. First, in step S10, the specifications of the motor 3 and the power conversion device 20 (e.g., the number of pole pairs of the motor 3, rated speed, input power frequency fac, modulation frequency f0, carrier frequency fc, sampling frequency fs, etc. of the power conversion device 20) are input to the input unit 51C. Alternatively, this specifications can be directly input to the input unit 51C of the diagnostic device 50, or remotely input to the input unit 51C via a network.

[0039] Next, in step S11, based on the input specification information, the setting unit 51D calculates the characteristic frequency sequence [A] used when diagnosing abnormalities of the motor 3 based on the above equations (4) and (5).

[0040] Furthermore, in step S12, the setting unit 51D calculates the monitoring frequency range [B] for monitoring the aforementioned characteristic frequency sequence [A]. The monitoring frequency range [B] is obtained using equations (4) and (5) above. For example, consider the following... Figure 7 A and Figure 7 Under conditions B, with a modulation frequency f0 = 40 Hz and a pole pair number p = 1, the slip s of motor 3 is monitored to be 2% to 10%. In this case, f0 + fm = 76 Hz to 79.2 Hz, f0 - fm = 0.8 Hz to 4 Hz, f0 + fr = 41.6 Hz to 48 Hz, and f0 - fr = 32 Hz to 38.4 Hz. The monitoring frequency range [B] is determined so that these frequencies can be monitored. Information about the range of slip s under this condition can be obtained or estimated from the datasheet of motor 3, or it can be specified as part of the specifications when the diagnostic device 50 is commercialized.

[0041] Next, in step S13, the setting unit 51D calculates the calculable noise frequency sequence [C] that is considered to be generated within the monitoring frequency range [B] based on the above equations (1) to (3). Then, the setting unit 51D sends the calculated characteristic frequency sequence [A], monitoring frequency range [B], and noise frequency sequence [C] to the analysis unit 51B.

[0042] On the other hand, in step S14, the current of the motor 3 is detected by the current detector 4. Then, the current of the motor 3 detected by the current detector 4 is digitized by the detection unit 51A and sent as a current signal to the analysis unit 51B. Next, in step S15, the analysis unit 51B performs spectrum analysis on the current signal sent from the detection unit 51A using methods such as current FFT (Fast Fourier Transform). Next, in step S16, the analysis unit 51B extracts the frequency and signal strength of the spectral peak. Then, in step S17, the analysis unit 51B stores the peak frequency sequence [D] of the spectral peaks within the monitored frequency range [B] in the storage unit 51E. Here, the peak frequency sequence [D] is a frequency sequence of spectral peaks having a preset signal strength, for example, -65dB or higher.

[0043] Next, in steps S18 to S22, the analysis unit 51B compares the peak frequency sequence [D] with the characteristic frequency sequence [A] (a frequency sent from the setting unit 51D), the monitoring frequency range [B], and the computable noise frequency sequence [C], and classifies the extracted peak frequency sequence [D] according to each cause. At this time, by excluding the frequencies of spectral peaks that cannot be classified as computable noise frequency sequences [C] from the monitoring frequency range [B] used when performing anomaly diagnosis by the diagnostic device 50, i.e., the unclassifiable frequency sequences, the monitorable video band [F] is determined and stored in the storage unit 51E. The width of the spectral peaks of the excluded frequency sequence [E] can be a statistical width obtained during the learning period, or it can uniformly have a width of, for example, a few Hz. In this embodiment, for example, a case with a width of ±1.5 Hz is shown.

[0044] Steps S18 to S22 will be explained in detail below. In step S18, the analysis unit 51B determines whether the peak frequency sequence [D] is included in the computable noise frequency sequence [C]. If the peak frequency sequence [D] is included in the computable noise frequency sequence [C] in step S18, the process proceeds to step S20 and stores it as a excluded frequency sequence [E] in the storage unit 51E. In step S18, if the peak frequency sequence [D] is not included in the computable noise frequency sequence [C], proceed to step S19 to determine whether the peak frequency sequence [D] is not included in the characteristic frequency sequence [A]. In step S19, if the peak frequency sequence [D] is not included in the characteristic frequency sequence [A], proceed to step S20 and store it as a excluded frequency sequence [E] in the storage unit 51E. In step S20, if the exclusion frequency sequence [E] is stored in the storage unit 51E, the process proceeds to step S21, where the monitorable video band [F] is obtained by subtracting the exclusion frequency sequence [E] from the monitoring frequency range [B], and is stored in the storage unit 51E. On the other hand, in step S19, if the peak frequency sequence [D] is included in the characteristic frequency sequence [A], the process proceeds to step S22 and stores it as a monitoring characteristic frequency sequence [G] in the storage unit 51E.

[0045] Figure 7 A is a graph showing an example of the spectrum obtained when the electric motor of Embodiment 1 is driven. Figure 7 B is a schematic diagram showing the monitorable video band [F] obtained by subtracting the decision exclusion frequency sequence [E] from the monitor frequency range [B].

[0046] Figure 7 A is a diagram showing an example of the spectrum obtained by the analysis unit 51B of the monitoring and diagnostic unit 51 when the motor 3 is driven under the conditions of f0 (modulation frequency) = 40Hz, fac (power supply frequency) = 60Hz, fc (carrier frequency) = 2000Hz, and fs (sampling frequency) = 4000Hz. For example, 0Hz to 100Hz can be specified as the monitoring frequency range for monitoring anomalies [B]. Furthermore, here, a frequency sequence with a pre-defined signal strength, such as a spectral peak of -65dB or higher, is considered as a peak frequency sequence [C]. The Fpwm obtained by equation (1) is an integer multiple of 40Hz, consistent with F0 obtained by equation (3).

[0047] exist Figure 7 In A, the spectral peaks can be confirmed at 40Hz and 80Hz respectively. The Fv obtained by equation (2) is 20Hz, 40Hz, 60Hz, 80Hz, and 100Hz. Figure 7The spectral peaks that can be confirmed in A are 20Hz, 40Hz, 60Hz, 80Hz, and 100Hz. The frequencies that can also be confirmed from the measurement results among the spectral peaks of the frequencies predicted according to these theoretical formulas are stored in the decision exclusion frequency sequence [E] as a calculable noise frequency sequence [C]. In addition, it can be confirmed that spectral peaks not mentioned in theoretical formulas such as equations (1) to (3) or characteristic frequency sequence [A] appear in 10-18Hz and 21-30Hz. When performing anomaly diagnosis, if these spectral peaks also overlap with the characteristic frequency sequence [A], it may lead to false detection, so they are stored as decision exclusion frequency sequence [E].

[0048] Figure 7 B is a schematic diagram illustrating the monitorable video band [F] obtained by subtracting the decision exclusion frequency sequence [E] from the monitored frequency range [B]. The diagnostic device 50 stores this monitorable video band [F] in the storage unit 51E during the learning process. Additionally, in Figure 7 In B, the monitorable video band [F] represents the blank area obtained by subtracting the decision exclusion frequency sequence [E] from the monitor frequency range [B]. In the monitorable video tape [F], for example, in the case of a mechanical system abnormality and rotor rod damage in a motor 3 with 1 pole pair, by... Figure 7 In B, [P1] and [P2], and [Q1] and [Q2], represent the frequency ranges for diagnosing various abnormalities. Specifically, [P1] and [P2] are the frequency ranges used to diagnose mechanical system abnormalities, and [Q1] and [Q2] are the frequency ranges used to diagnose rotor rod damage. Here, the slip range for mechanical system abnormalities is 4–20%, and the slip range for rotor rod damage is 2–12%. As can be seen from the above, under the usage conditions within this scope, the method disclosed herein can detect mechanical system abnormalities and rotor rod damage while avoiding the influence of noise.

[0049] Figure 8 and Figure 9 This is a diagram illustrating the monitoring and diagnostic process of the diagnostic device for the electric motor according to Embodiment 1. In step S30, the determination unit 51F of the diagnostic device 50, which has completed the above learning, reads the monitorable video tape [F] and the monitoring characteristic frequency sequence [G] stored in the storage unit 51E. Then, in step S31, the current of the motor 3 is detected by the current detector 4 in the same manner as during learning. The current of the motor 3 detected by the current detector 4 is then digitized by the detection unit 51A and sent as a current signal to the analysis unit 51B. Next, in step S32, the analysis unit 51B performs spectrum analysis on the current signal sent from the detection unit 51A using current FFT (Fast Fourier Transform) and the like. Next, in step S33, the analysis unit 51B extracts the frequency and signal strength of the spectral peak and sends them to the determination unit 51F.

[0050] Next, in step S34, the determination unit 51F extracts the spectral peaks located within the monitorable video band [F] from the spectral peaks sent from the analysis unit 51B. Next, in step S35, it is determined whether the change in the signal strength of the spectral peak in the monitorable video band [F] compared with the spectral peak during learning exceeds a preset strength range. If it is determined in step S35 that the change exceeds the above-mentioned intensity range, proceed to step S36 to determine whether the frequency of the spectral peak of the signal intensity change that exceeds the above-mentioned intensity range is consistent with the monitoring characteristic frequency sequence [G]. On the other hand, if it is determined in step S35 that the change does not exceed the above intensity range, in order to perform the next abnormal diagnosis, return to step S31, where the current detector 4 detects the current of the motor 3, or end the abnormal diagnosis of the motor. Then, if it is not determined in step S35 that the change exceeds the above-mentioned intensity range, the meaning of "normal" is stored in the storage unit 51E as a determination result, and the meaning of "normal" is displayed by the display unit 52.

[0051] If the frequency of the spectral peak determined in step S36 is consistent with the monitoring characteristic frequency sequence [G], proceed to step S37 to determine the abnormal location. Here, the frequency of the spectral peak is substituted into the above equations (4) and (5) to determine whether the cause of the abnormality is due to a mechanical system malfunction or rotor rod damage, and then proceed to step S38. In step S38, the determination result of the abnormal part in step S37 is stored in the storage unit 51E and displayed on the display unit 52 of the diagnostic device 50, or an alarm is issued by the alarm unit 53.

[0052] If the frequency of the spectrum peak determined in step S36 is inconsistent with the monitoring characteristic frequency sequence [G], proceed to step S38, determine that the change in the signal strength of the spectrum peak is not due to any of the mechanical system abnormalities or rotor rod damage, but rather that something different from the normal situation has occurred, and store the determination result in the storage unit 51E, and display it on the display unit 52 of the diagnostic device 50, or issue an alarm by the alarm unit 53. In addition, the network output unit can be installed in the display unit 52 and alarm unit 53 of the diagnostic device 50, so as to remotely prompt the user with this information.

[0053] As described above, according to Embodiment 1, A diagnostic device for an electric motor, used to diagnose abnormalities in an electric motor driven by electricity converted by a power conversion device. The diagnostic device includes: A current detector that detects the current flowing from the power conversion device to the motor; and The monitoring and diagnostic unit performs spectrum analysis based on the current detected by the current detector and monitors for abnormalities in the motor. The monitoring and diagnostic unit includes: The input section is for inputting at least the specifications of the power conversion device and the motor. The setting unit calculates, based on the information from the input unit, a characteristic frequency sequence for diagnosing abnormalities in the motor, a monitoring frequency range for monitoring the abnormalities, and a noise frequency sequence that can be derived from the specification information of the power conversion device. The analysis unit performs spectral analysis based on the current detected by the current detector and calculates the current spectrum. It extracts a peak frequency sequence with a spectral peak value above a predetermined signal strength within the monitoring frequency range from the current spectrum. It compares the peak frequency sequence with the characteristic frequency sequence and the noise frequency sequence to derive a decision exclusion frequency sequence and a monitoring characteristic frequency sequence. It also derives a monitorable video band obtained by subtracting the decision exclusion frequency sequence from the monitoring frequency range. A storage unit that stores the monitorable video tape, the monitoring characteristic frequency sequence, and its signal strength; and The determination unit determines the abnormality of the motor by comparing the monitorable video tape and the monitoring characteristic frequency sequence stored in the storage unit with the current spectrum newly acquired from the current detector.

[0054] Furthermore, before the diagnostic device diagnoses the abnormality of the motor, The analysis unit stores the monitorable video band, the characteristic frequency sequence, and their signal strength as normal data in the storage unit. When the diagnostic device diagnoses an abnormality in the motor, The determination unit reads the monitorable video band, the characteristic frequency sequence, and its intensity as normal data from the storage unit, and compares them with the current spectrum newly acquired from the current detector. When the change in signal strength of the monitoring feature frequency sequence contained in the monitorable video band compared to normal data exceeds a preset threshold, it is determined to be abnormal.

[0055] Furthermore, when the diagnostic device diagnoses an abnormality in the motor, The input section inputs the specifications of the power conversion device and the motor. The setting unit calculates the noise frequency sequence using the specification information of the power conversion device input to the input unit, and calculates the characteristic frequency sequence using the specification information of the power conversion device and the motor input to the input unit. The analysis unit classifies the various spectral peaks within the monitored frequency range into the noise frequency sequence, the characteristic frequency sequence, and unclassifiable sequences based on the current spectrum acquired by the current detector. The noise frequency sequence and the unclassifiable sequence are defined as decision exclusion frequency sequences, and the monitorable video tape is derived from the decision exclusion frequency sequences.

[0056] By configuring it as described above in Implementation 1, abnormalities of the motor can be diagnosed with high accuracy even in the presence of theoretically unpredictable noise.

[0057] Implementation method 2. Figure 10 This is a block diagram showing a simplified structure of the power conversion device and the diagnostic device for the electric motor according to Embodiment 2. like Figure 10 As shown, the power conversion device 20 converts the frequency of the AC power from the AC power source 1 and supplies it to the motor 3. The motor 3 is connected to the load 7 via the power transmission mechanism 60. The diagnostic device 50 detects the current of at least one phase of the current supplied from the power conversion device 20 to the motor 3 via the current detector 4, and analyzes the detected current to detect abnormalities in the motor 3 and the power transmission mechanism 60 of the motor 3. In Embodiment 1, the diagnostic device 50 diagnoses mechanical system abnormalities and rotor rod damage of the motor 3. In contrast, in Embodiment 2, the diagnostic device 50 also performs a diagnosis of the power transmission mechanism 60 connected to the motor 3. Furthermore, the diagnosis of mechanical system abnormalities and rotor rod damage of the electric motor 3 described in Embodiment 1 can be easily combined with the diagnosis of the power transmission mechanism 60 of the electric motor 3 described in Embodiment 2. The following description focuses on the differences from Embodiment 1, and the similarities are omitted.

[0058] like Figure 10As shown, the motor 3 is connected to the load 7 via a power transmission mechanism 60. In the power transmission mechanism 60, for example, a belt 61 connects a pulley Pu1 mounted on the rotating shaft of the motor 3 and a pulley Pu2 mounted on the rotating shaft of the load 7, transmitting power from the motor 3 to the load 7. Assuming the radius of pulley Pu1 is Dr, the length of the belt 61 is L, and the rotational speed of the motor 3 is Nr, then the length of the belt that pulley Pu1 rotates per unit time is 2πDrNr. The rotational frequency fb' of the belt can be calculated based on the ratio of this length to the total length L of the belt.

[0059] fb'=((2·π·Dr) / L)·fr (8)

[0060] Where fr is the rotational frequency of motor 3. When a belt rotating at frequency fb' is connected to the rotating shaft of a motor 3 rotating at frequency fr, its vibration is transmitted to the rotor through the rotating shaft of the motor 3, and manifests as a sideband wave of f0±fb' in the spectrum of the current detected by the current detector 4. In addition, harmonics such as f0±2fb', f0±3fb', ... also appear in this sideband.

[0061] When belt 61 is properly connected to pulley Pu1, the signal strength of these sideband waves is observed to be high; if damage such as cutting occurs, the signal strength decreases. Since this is the opposite of the mechanical system malfunction and rotor rod damage described in Embodiment 1, different steps are required when diagnosing the power transmission mechanism 60 compared to Embodiment 1.

[0062] Figure 11 This is a block diagram showing the structure of the monitoring and diagnostic unit of the diagnostic device for the electric motor according to Embodiment 2. Figure 12 and Figure 13 This is a diagram illustrating the learning process of the diagnostic device for the electric motor according to Embodiment 2. like Figure 11 As shown, the monitoring and diagnostic unit 51 of the diagnostic device 50 includes a detection unit 51A, an analysis unit 51B, an input unit 51C, a setting unit 51D, a storage unit 51E, and a determination unit 51F, which will determine the results based on... Figure 12 and Figure 13 The learning process is used to illustrate the function of each component during the learning process.

[0063] Figure 12 and Figure 13 This is a diagram illustrating the learning process of the diagnostic device for the electric motor according to Embodiment 2. First, in step S110, the specifications of the motor 3, the power conversion device 20, and the power transmission mechanism 60 are input to the input unit 51C. In Embodiment 1, the specifications of the motor 3 and the power conversion device 20 are input, but in Embodiment 2, the specifications of the power transmission mechanism 60 are added (e.g., the radius Dr of the pulley Pu1, the length L of the belt 61, the rotational speed Nr of the motor 3, etc.).

[0064] Next, in step S111, based on the input specification information, the setting unit 51D calculates the characteristic frequency sequence [A1] used when diagnosing abnormalities in the motor 3 and the power transmission mechanism 60 based on the above formulas (4), (5) and (8). Furthermore, in step S112, the setting unit 51D calculates the monitoring frequency range [B1] for monitoring the aforementioned characteristic frequency sequence [A1]. Additionally, the monitoring frequency range [B1] is obtained using the aforementioned equations (4), (5), and (8). Next, in step S113, the setting unit 51D calculates the calculable noise frequency sequence [C1] that is considered to be generated within the monitoring frequency range [B1] based on the above equations (1) to (3). Then, the setting unit 51D sends the calculated characteristic frequency sequence [A1], the monitoring frequency range [B1], and the calculable noise frequency sequence [C1] to the analysis unit 51B.

[0065] On the other hand, in step S114, the current of the motor 3 is detected by the current detector 4. Then, the current of the motor 3 detected by the current detector 4 is digitized by the detection unit 51A and sent as a current signal to the analysis unit 51B. Next, in step S115, the analysis unit 51B performs spectrum analysis on the current signal sent from the detection unit 51A using methods such as current FFT (Fast Fourier Transform). Next, in step S116, the analysis unit 51B extracts the frequency and signal strength of the spectral peak. Then, in step S117, the analysis unit 51B stores the peak frequency sequence [D1] of the spectral peaks within the monitored frequency range [B1] in the storage unit 51E.

[0066] Next, in steps S118 to S122, the analysis unit 51B compares the peak frequency sequence [D1] with the characteristic frequency sequence [A1] (which is a frequency sent from the setting unit 51D), the monitoring frequency range [B1], and the computable noise frequency sequence [C1], and classifies the extracted peak frequency sequence [D1] according to each cause. At this time, by excluding the frequencies of spectral peaks that cannot be classified as computable noise frequency sequences [C1] from the monitoring frequency range [B1] used when performing anomaly diagnosis by the diagnostic device 50, i.e., the unclassifiable frequency sequences, the monitorable video band [F1] is determined and stored in the storage unit 51E. The width of the spectral peaks of the exclusion decision frequency sequence [E1] used for exclusion can be a statistical width obtained during the learning period, or it can uniformly have a width of, for example, a few Hz. In this embodiment, for example, a case with a width of ±1.5 Hz is shown.

[0067] The following will describe steps S118 to S122 in detail. In step S118, the analysis unit 51B determines whether the peak frequency sequence [D1] is included in the computable noise frequency sequence [C1]. If the peak frequency sequence [D1] is included in the computable noise frequency sequence [C1] in step S118, the process proceeds to step S120 and stores it as a excluded frequency sequence [E1] in the storage unit 51E. In step S118, if the peak frequency sequence [D1] is not included in the computable noise frequency sequence [C1], proceed to step S119 and determine whether the peak frequency sequence [D1] is included in the characteristic frequency sequence [A1]. In step S119, if the peak frequency sequence [D1] is not included in the characteristic frequency sequence [A1], proceed to step S120 and store it as a excluded frequency sequence [E1] in the storage unit 51E. In step S120, if the exclusion frequency sequence [E1] is stored in the storage unit 51E, the process proceeds to step S121, whereby the monitorable video band [F1] obtained by subtracting the exclusion frequency sequence [E1] from the monitoring frequency range [B1] is calculated and stored in the storage unit 51E. On the other hand, in step S119, if the peak frequency sequence [D1] is included in the characteristic frequency sequence [A1], the process proceeds to step S122 and stores it as a monitoring characteristic frequency sequence [G1] in the storage unit 51E.

[0068] Figure 14 A is a graph showing an example of the spectrum obtained when the motor of Embodiment 2 is driven. Figure 14B is a schematic diagram illustrating the monitorable video band [F1] obtained by subtracting the exclusion frequency sequence [E1] from the monitored frequency range [B1]. Additionally, in Figure 14 In B, the monitorable video band [F1] represents the blank area obtained by subtracting the decision exclusion frequency sequence [E1] from the monitor frequency range [B1].

[0069] Figure 14 A is a diagram showing an example of the spectrum obtained by the current detector 4 when driving the motor 3 under the conditions of f0 (modulation frequency) = 40 Hz, fac (power supply frequency) = 60 Hz, fc (carrier frequency) = 2000 Hz, and fs (sampling frequency) = 4000 Hz. The following will describe the differences from Embodiment 1, and will omit the contents that are the same as those in Embodiment 1 as appropriate. Figure 7 A and Figure 7 (Explanation of B).

[0070] For example, 0–100 Hz is designated as the monitoring frequency range for monitoring anomalies [B1]. For example, under the above conditions, as a power transmission mechanism 60, a belt 61 with a length of 1 m is rotated by a pulley Pu1 with a radius of 10 cm. Assuming that the rotating motor 3 has 1 pole pair and the slippage is 5%, the sideband waves observed during the diagnosis of the power transmission mechanism 60 are 16.12 Hz, 63.88 Hz, 87.75 Hz, … from the low-frequency side.

[0071] When comparing Figure 14 The spectrum shown and the sideband waves observed during the diagnosis of the aforementioned power transmission mechanism 60 indicate that 87.75Hz is within the monitorable video band [F1]. That is, when the power transmission mechanism 60 is connected, Figure 14 The sideband waves at 16.12 Hz and 63.88 Hz in the spectrum shown in A are undiagnosable due to significant noise, but the 87.75 Hz frequency, lacking noise or having low intensity, can be used to diagnose the power transmission mechanism 60. Furthermore, although this frequency was determined by setting the slip to 5%, the frequency range when the slip is set to 1–16% is… Figure 14 The frequency range [R] shown in B. That is, Figure 14 In B, [R] represents the frequency range used for diagnosing the power transmission mechanism. Therefore, even under varying load conditions, the peak values ​​used for diagnosing the power transmission mechanism can be adequately monitored. In addition, Figure 14 A and Figure 14 In B, with Figure 7 A and Figure 7Similarly, frequency ranges [P1] and [P2] for diagnosing mechanical system malfunctions and frequency ranges [Q1] and [Q2] for diagnosing rotor rod damage are shown.

[0072] Figure 15 and Figure 16 This is a diagram illustrating the monitoring and diagnostic process of the diagnostic device for the electric motor according to Embodiment 2. In step S130, the determination unit 51F of the diagnostic device 50, which has completed the above learning, reads the monitorable video tape [F1] and the monitoring characteristic frequency sequence [G1] stored in the storage unit 51E. Then, in step S131, the current of the motor 3 is detected by the current detector 4 in the same manner as during learning. The current of the motor 3 detected by the current detector 4 is then digitized by the detection unit 51A and sent as a current signal to the analysis unit 51B. Next, in step S132, the analysis unit 51B performs spectrum analysis on the current signal sent from the detection unit 51A using methods such as current FFT (Fast Fourier Transform). Next, in step S133, the analysis unit 51B extracts the frequency and signal strength of the spectral peak and sends them to the determination unit 51F.

[0073] Next, in step S134, the determination unit 51F extracts the spectral peaks located in the monitorable video band [F1] from the spectral peaks sent from the analysis unit 51B. Next, in step S135, it is determined whether the change in the signal strength of the spectral peak in the monitorable video band [F1] compared with the spectral peak during learning exceeds a preset strength range. If it is determined in step S135 that the change exceeds the above-mentioned intensity range, proceed to step S136 to determine whether the frequency of the spectral peak of the signal intensity change that exceeds the above-mentioned intensity range is consistent with the monitoring characteristic frequency sequence [G1]. On the other hand, if it is determined in step S135 that the change does not exceed the above-mentioned intensity range, in order to perform the next abnormal diagnosis, return to step S131, where the current detector 4 detects the current of the motor 3, or end the abnormal diagnosis of the motor. Then, if it is determined in step S135 that the change does not exceed the above-mentioned intensity range, the meaning of "normal" is stored in the storage unit 51E as a determination result, and the meaning of "normal" is displayed by the display unit 52.

[0074] If the frequency of the spectral peak determined in step S136 matches the monitoring characteristic frequency sequence [G], proceed to step S137 to determine the abnormal location. Here, the frequency of the spectral peak is substituted into equations (4) and (5) above to determine whether the cause of the abnormality is due to a mechanical system malfunction or rotor rod damage. Furthermore, the frequency of the spectral peak is applied to equation (8) above to determine whether the cause of the abnormality is caused by the power transmission mechanism. Then, proceed to step S138. In step S138, the determination result of the abnormal part in step S137 is stored in the storage unit 51E and displayed on the display unit 52 of the diagnostic device 50, or an alarm is issued by the alarm unit 53.

[0075] If the frequency of the spectrum peak determined in step S136 is inconsistent with the monitoring characteristic frequency sequence [G1], proceed to step S138. Even if the change in the signal strength of the spectrum peak is not due to any of the mechanical system abnormality, rotor rod damage, or power transmission system abnormality, but rather a situation different from the normal one has occurred, store the determination result in the storage unit 51E, and display it on the display unit 52 of the diagnostic device 50, or issue an alarm by the alarm unit 53. In addition, the network output unit can be installed in the display unit 52 and alarm unit 53 of the diagnostic device 50, so as to remotely prompt the user with this information.

[0076] Furthermore, as mentioned at the beginning, in the diagnosis of the power transmission mechanism 60, an abnormality is diagnosed when the signal strength of the characteristic frequency sequence [G1] decreases, which differs from the diagnosis of mechanical system abnormalities and rotor rod damage described in Embodiment 1. Therefore, this embodiment and Embodiment 1 can be easily combined.

[0077] As described above, according to embodiment 2, A diagnostic device for an electric motor, diagnosing abnormalities in an electric motor driven by electricity converted by a power conversion device and abnormalities in the power transmission mechanism connected to the electric motor. The diagnostic device includes: A current detector that detects the current flowing from the power conversion device to the motor; and The monitoring and diagnostic unit performs spectrum analysis based on the current detected by the current detector and monitors for anomalies in the motor and the power transmission mechanism. The monitoring and diagnostic unit includes: The input section is for inputting at least the specifications of the power conversion device and the motor. The setting unit calculates, based on the information from the input unit, a characteristic frequency sequence for diagnosing abnormalities in the motor and the power transmission mechanism, a monitoring frequency range for monitoring the abnormalities, and a noise frequency sequence that can be derived from the specification information of the power conversion device. The analysis unit performs spectral analysis based on the current detected by the current detector and calculates the current spectrum. It extracts a peak frequency sequence from the current spectrum that has a spectral peak value above a predetermined signal strength within the monitoring frequency range. The analysis unit compares the peak frequency sequence with the characteristic frequency sequence and the noise frequency sequence to derive a decision exclusion frequency sequence and a monitoring characteristic frequency sequence. It also derives a monitorable video band obtained by subtracting the decision exclusion frequency sequence from the monitoring frequency range. A storage unit that stores the monitorable video tape, the monitoring characteristic frequency sequence, and its signal strength; and The determination unit determines the abnormality of the motor and the abnormality of the power transmission mechanism by comparing the monitorable video tape and the monitoring characteristic frequency sequence stored in the storage unit with the current spectrum newly acquired from the current detector.

[0078] Furthermore, before the diagnostic device diagnoses any abnormalities in the motor and the power transmission mechanism, The analysis unit stores the monitorable video band, the characteristic frequency sequence, and their signal strength as normal data in the storage unit. When the diagnostic device diagnoses abnormalities in the electric motor and the power transmission mechanism, The determination unit reads the monitorable video band, the characteristic frequency sequence, and its intensity as normal data from the storage unit, and compares them with the current spectrum newly acquired from the current detector. When the change in signal strength of the monitoring feature frequency sequence contained in the monitorable video band compared to normal data exceeds a preset threshold, it is determined to be abnormal.

[0079] Furthermore, when the diagnostic device diagnoses abnormalities in the motor and the power transmission mechanism, The input section inputs the specifications of the power conversion device, the electric motor, and the power transmission mechanism. The setting unit uses the specification information of the power conversion device input to the input unit to calculate the noise frequency, and uses the specification information of the power conversion device, the motor, and the power transmission mechanism input to the input unit to calculate the characteristic frequency sequence. The analysis unit classifies the various spectral peaks into the noise frequency sequence, the characteristic frequency sequence, and unclassifiable sequences based on the current spectrum acquired by the current detector. The noise frequency sequence and the unclassifiable sequence are defined as the exclusion frequency sequence, and the monitorable video band is calculated based on the exclusion frequency sequence.

[0080] By configuring it as described above in Embodiment 2, abnormalities in the motor and power transmission mechanism can be diagnosed with high accuracy even in the presence of theoretically unpredictable noise.

[0081] Implementation method 3. In Embodiment 3, in order to more clearly distinguish between the decision exclusion frequency sequence [E] and the monitoring characteristic frequency sequence [G] in Embodiment 1, and in order to more clearly distinguish between the decision exclusion frequency sequence [E1] and the monitoring characteristic frequency sequence [G1] in Embodiment 2, the diagnostic device 50 is trained while the load rate of the motor 3 is changed. Here, the mechanical system abnormality, rotor rod damage, and characteristic frequency of the power transmission mechanism, as shown in Equations (4), (5), and (8) above, depend on the slip s.

[0082] Figure 17 and Figure 18 The verification process of the monitoring feature frequency sequence [G] in Implementation 3 is shown. exist Figure 17 and Figure 18 In order to more clearly distinguish between the decision exclusion frequency sequence [E1] and the monitoring characteristic frequency sequence [G1] in Implementation 2, the process of learning the diagnostic device 50 while changing the load rate of the motor 3 is described. Changing the load rate of the motor 3 means, for example, when the load connected to the motor 3 is a pump, changing the load rate of the motor 3 by opening and closing the valve of the pump.

[0083] As in embodiment 2 Figure 13 As shown, the storage unit 51E stores the decision exclusion frequency sequence [E1] and the monitoring feature frequency sequence [G1]. In step S200, the analysis unit 51B reads the decision exclusion frequency sequence [E1] and the monitoring feature frequency sequence [G1] stored in the storage unit 51E. Next, in step S201, the load rate connected to the motor 3 is changed, and in step S202, the current is detected by the current detector 4. Next, in step S203, the analysis unit 51B performs spectrum analysis on the detected current signal using methods such as current FFT (Fast Fourier Transform). Next, in step S204, the analysis unit 51B extracts the frequency and signal strength of the spectral peak.

[0084] Next, in step S205, the analysis unit 51B compares the spectral peak obtained in step S204 with the frequency sequence [G1] of the monitored object. Here, since the monitored frequency sequence [G1] is a characteristic frequency calculated based on equations (4), (5), and (8), the change in load rate manifests as a slip, with each characteristic frequency shifting. Assuming that the shift of the characteristic frequency is not confirmed ("yes" in step S206), then at least these are not characteristic frequencies. Therefore, in step S207, the unshifted frequencies are added to the exclusion frequency sequence [E1] and omitted from the monitored characteristic frequency sequence [G1]. The above steps can more clearly distinguish between the exclusion frequency sequence [E1] and the monitoring feature frequency sequence [G1].

[0085] As described above, according to embodiment 3, When the analysis unit confirms the settings of the monitored feature frequency sequence and the decision exclusion frequency sequence, By changing the load rate of the motor, the current spectrum is obtained by the current detector. It was confirmed that the peak value of the current spectrum corresponding to each component of the monitored characteristic frequency sequence shifts according to the load rate of the motor. If the offset cannot be confirmed, the component of the monitoring characteristic frequency sequence corresponding to the peak value of the current spectrum for which the offset cannot be confirmed is excluded from the monitoring characteristic frequency sequence, and then added to the decision exclusion frequency sequence. Therefore, it can more clearly distinguish between the monitored characteristic frequency sequence and the exclusion frequency sequence.

[0086] Implementation method 4. In Embodiment 1, the motor 3 is diagnosed by inputting the specification information of the motor 3 and the power conversion device 20 into the diagnostic device 50. Furthermore, in Embodiment 2, the motor 3 and the power conversion device 20 and the power transmission mechanism 60 are diagnosed by inputting the specification information of the motor 3, the power conversion device 20, and the power transmission mechanism 60 into the diagnostic device 50. In each embodiment, a calculable noise frequency sequence generated by the power conversion device 20 is learned, but the diagnosis of the calculable noise frequency sequence is not performed during diagnosis. On the other hand, the noise generation status of the power conversion device 20 depends on the setting conditions of the power conversion device 20. Therefore, due to the large amount of noise generated, it is sometimes impossible to sufficiently ensure the range of the monitorable video bands [F] and [F1], and thus, adequate anomaly diagnosis cannot be performed. In this Embodiment 4, the solution to this problem will be explained.

[0087] Figure 19 , Figure 20 , Figure 21 The process of changing the setting conditions of the proposed power conversion device 20 in Embodiment 4 is shown. in addition, Figure 19 , Figure 20 , Figure 21 The process shown in the diagram illustrates the learning process in Implementation 2, which is based on... Figure 12 and Figure 13 The process.

[0088] exist Figure 19 , Figure 20 , Figure 21 In the process, the steps from S310 to S322 are the same as... Figure 12 and Figure 13 Steps S110 to S122 are the same, so their description is omitted. exist Figures 19-21 The process involves storing the monitorable video band [F1] and the monitoring characteristic frequency sequence [G1] in the storage unit 51E through steps S321 and S322. At this time, when the monitorable video band [F1] is within a narrower range than the monitoring frequency range [B1], it is difficult to perform anomaly diagnosis.

[0089] As an example, Figure 22 Figure A shows the spectrum of the current detected by the current detector 4 when f0 (modulation frequency) = 119 Hz, fac (power supply frequency) = 60 Hz, fc (carrier frequency) = 1000 Hz, and fs (sampling frequency) = 4000 Hz. The monitoring frequency range [B1] for monitoring anomalies is set to 0 Hz to 220 Hz. It can be confirmed that spectral peaks occur throughout the entire monitoring frequency range [B1]. Similar to the example of the above embodiment, spectral peaks with signal strength greater than -65 dB are set as the decision exclusion frequency sequence [E1], with a width of ±1.5 Hz. Furthermore, statistical values ​​obtained through learning can be used for these frequencies and widths.

[0090] Figure 22 B shows Figure 22 In case A, the decision excludes the frequency sequence [E1] and the monitorable video band [F1]. Furthermore, as mentioned above, in... Figure 22 In B, the monitorable video band [F1] is the frequency band represented by the blank spaces between the decision exclusion frequency sequences [E1]. from Figure 22 As shown in B, the monitorable video band [F1] is a relatively narrow range relative to the monitoring frequency range [B1], which is 29% when assessing the ratio of the monitorable video band [F1] to the monitoring frequency range [B1]. It is difficult to diagnose anomalies within such a narrow range.

[0091] Therefore, in this embodiment, a preset ratio threshold is set for the above ratio, and if the ratio is below the threshold, a prompt is made to change the setting conditions of the power conversion device 20. Set, for example, 50% as the threshold for the aforementioned ratio. Figure 22 Under the conditions shown in A, since the ratio of the monitorable video band [F1] to the monitor frequency range [B1] is 29%, which is lower than 50% as the ratio threshold, the diagnostic device 50 prompts the power conversion device 20 to change its settings. That is, in step S330, it is determined whether the ratio of the monitorable video band [F1] to the monitoring frequency range [B1] is below a preset ratio threshold. In step S330, when the ratio of the monitorable video band [F1] to the monitor frequency range [B1] is below a preset ratio threshold, the process proceeds to step S340 to change the setting conditions of the power conversion device 20. Then, with the settings of the power conversion device 20 changed, the processing steps S310 to S330 are repeated. Additionally, in step S330, the learning process ends when the ratio of the monitorable video band [F1] to the monitoring frequency range [B1] is greater than a preset ratio threshold.

[0092] When prompted to change the setting conditions of the power conversion device 20, it is suggested that the change be made with the smallest possible deviation from the original setting conditions to avoid deviating from the intended use of the motor 3. As an example, Fpwm is expressed by the above formula (1). Figure 22 A and Figure 22 Under the conditions shown in B, i is set to a positive integer, Fpwm = i. According to equation (1), Fpwm is generated every 1 Hz, but from Figure 22 Such a spectral peak cannot be confirmed in A. However, if GCD(f0, fc, fs) increases, Fpwm following equation (1) is generated on the spectrum, and the proportion of noise tends to decrease from the overall spectrum perspective. Therefore, the diagnostic device 50 proposes to shift f0 (modulation wave frequency) = 119Hz by 1Hz to reach 120Hz. In this case, Fpwm = 40i, and Fpwm is generated every 40Hz, suppressing the generation of noise.

[0093] Figure 23 A shows the spectrum of the current detected under the conditions of f0 (modulation frequency) = 120 Hz, fac (power supply frequency) = 60 Hz, fc (carrier frequency) = 1000 Hz, and fs = 4000 Hz. Figure 23 B shows the decision to exclude the frequency sequence [E1] and the monitorable video band [F1] in this case. As described above, the monitorable video band [F1] extends relative to the monitoring frequency range [B1], and when evaluating the ratio of the monitorable video band [F1] to the monitoring frequency range [B1], this ratio is 77%. Diagnosis can be performed within such a large range. Furthermore, the percentage change in f0 (modulation frequency) is approximately 1%, thus not deviating from the intended use of motor 3.

[0094] As described above, according to embodiment 4, When the ratio of the monitorable video band to the monitoring frequency range is below a preset ratio threshold, the analysis unit suggests changing the setting of the power conversion device's specification information. Therefore, it can expand the ratio of the monitorable video band to the monitoring frequency range, enabling high-precision diagnosis of motor anomalies and improving reliability.

[0095] Implementation method 5. In embodiments 1 to 4, the electric motor diagnostic device 50 diagnoses abnormalities in the electric motor 3 driven by the power conversion device 20 or the power transmission mechanism 60 connected to the electric motor 3. At this time, the noise generated from the power conversion device 20 is calculated according to the above equations (1), (2), and (3), and the noise generation status of each device is learned. Besides the operating principle of the power conversion device 20, it is believed that the generation of this noise also depends on the model of the power conversion device 20 or the electric motor 3, or the installation status of these devices, etc. In this embodiment 5, the handling of these problems will be explained.

[0096] Figure 24 This is a block diagram showing the structure of the monitoring and diagnostic unit of the diagnostic device for the electric motor according to Embodiment 5. Figure 25 This is a block diagram showing the structure of the diagnostic system for the electric motor according to Embodiment 5. like Figure 24 As shown, the monitoring and diagnostic unit 51 of the electric motor diagnostic device 50 in Embodiment 5 includes the detection unit 51A, analysis unit 51B, input unit 51C, setting unit 51D, storage unit 51E, and determination unit 51F described in the above embodiments, and also includes a network input / output unit 51G, which inputs and outputs data to and from an external device via a network. like Figure 25As shown, the motor diagnostic system of Embodiment 5 includes a database device 100, which stores various data sent from multiple motor diagnostic devices 50 (diagnostic device A, diagnostic device B, and diagnostic device C) via a network input / output unit 51G. Furthermore, the database device 100 is connected to a diagnostic algorithm modification device 110, which can access the data stored in the database device 100 via a network. Additionally, the diagnostic algorithm modification device 110 can send modified diagnostic algorithms to the multiple motor diagnostic devices 50 (diagnostic device A, diagnostic device B, and diagnostic device C) via a network.

[0097] As described in the above embodiments, the diagnostic devices 50 for each motor (diagnostic device A, diagnostic device B, diagnostic device C) are, for example, according to... Figure 5 and Figure 6 or Figure 12 and Figure 13 The learning process is used to acquire the specification information of the power conversion device 20, the motor 3, and the power transmission mechanism 60. Then, based on this specification information, the exclusion frequency sequence [E] or [E1], the monitorable video band [F] or [F1], and the monitoring characteristic frequency sequence [G] or [G1] are calculated. In addition, the diagnostic devices 50 of each motor (diagnostic device A, diagnostic device B, and diagnostic device C) output these calculated results and the acquired specification information of the power conversion device 20, the motor 3, or the power transmission mechanism 60 to the database device 100 via the network input / output unit 51G. At this time, the peak value of the current spectrum compared by the determination unit 51F, the determination result of the determination unit 51F, and the model of the power conversion device 20 and the motor 3 can be added as output information.

[0098] Regarding the configuration of the database device 100, although it can be located at a position where data can be input from diagnostic devices 50 of multiple motors, it is preferable to locate it at a position where data can be input from diagnostic devices 50 of even more motors. That is, although the database device 100 can be located within a facility and data can be collected from diagnostic devices 50 of motors within that facility, it is preferable to locate the database device 100 outside the facility and collect data from diagnostic devices 50 of multiple motors in multiple facilities.

[0099] The diagnostic algorithm modification device 110 of the manufacturer of the motor diagnostic device 50 analyzes the data stored in the database device. The analysis by the diagnostic algorithm modification device 110 uses the model or setting conditions of the power conversion device 20, motor 3, and power transmission mechanism 60 stored in the database device 100, the decision exclusion frequency sequence [E] or [E1] calculated by each diagnostic device 50, the monitorable video band [F] or [F1], the monitor characteristic frequency sequences [G] and [G1], the peak value of the current spectrum compared by the decision unit 51F, and the decision result of the decision unit 51F, etc. Based on this data, the diagnostic algorithm of each motor diagnostic device 50 is modified. After modification by the diagnostic algorithm modification device 110, the network input / output unit 51G provided in the monitoring diagnostic unit 51 of each diagnostic device 50 receives the modified data, and the diagnostic algorithm of each motor diagnostic device 50 is modified.

[0100] As one possible modification, there is a proposal to change the setting conditions of the power conversion device 20 described in Embodiment 4. As described in Embodiment 4, the characteristics of the noise generated in the power conversion device 20 vary depending on the magnitude of GCD(f0, fc, fs), for example, as expressed by Equation (1). Although the generated noise can be adjusted to some extent according to the magnitude of GCD(f0, fc, fs), it is difficult to grasp the generation status of subtle noises that depend on the model of the power conversion device 20 or the motor 3. In this regard, the diagnostic algorithm is modified so that the data of the database device 100 is applied when a change in setting conditions is proposed. Suppose that when a specific power conversion device 20 and motor 3 are driven under specific setting conditions, a large amount of noise is generated, which is difficult to diagnose. In this case, in order to fine-tune the setting of the power conversion device 20, a setting value that is considered to improve the noise generation can be searched from similar conditions in the data of the database device 100, and it is proposed as a diagnostic algorithm to the user. Alternatively, the manufacturer can analyze the data of the database device 100 and import newly generated diagnostic algorithms that improve the noise generation status into each diagnostic device 50. However, these modifications to the diagnostic algorithms are limited by the specifications of the devices mounted on each diagnostic device 50, such as computing speed or storage capacity.

[0101] As another modification, it involves... Figure 8 and Figure 9 ,or Figure 15 and Figure 16 The diagnostic device 50 shown in the figure diagnoses the electric motor 3 or the power transmission mechanism 60. As described in Embodiments 1 to 4, the electric motor diagnostic device 50 diagnoses abnormalities in the electric motor 3 or the power transmission mechanism 60 while avoiding noise generation from the power conversion device 20. However, many false detections may occur during diagnosis. For example, such false detections may exist in the method for setting the decision exclusion frequency sequences [E] and [E1] learned by the diagnostic device 50. In Embodiments 1 and 2, the spectral peak of the peak frequency sequence is defined as a peak value exceeding a predetermined signal strength. Furthermore, the frequency width in this case is either uniformly defined as a certain value or defined using a statistical width obtained through learning. False detections may occur because the decision exclusion frequency sequences [E] and [E1] cannot be accurately set within these definitions. Moreover, the criteria used by the diagnostic device 50 to diagnose an anomaly may also be problematic.

[0102] When diagnosing the motor 3 or the power transmission mechanism 60, the diagnostic device 50 monitors the signal strength of the characteristic frequency corresponding to the abnormal location. If the change in this signal strength compared to normal data exceeds a preset threshold, an abnormality is determined. This threshold can be a preset value or a statistical signal strength obtained through learning. False detections may occur because the threshold setting is inaccurate. Even in cases where false detections occur due to problems with the settings of the diagnostic device 50, a database can be applied. However, in such cases, the user needs to input the correctness and error of the diagnostic results into the input section 51C of the monitoring and diagnostic unit 51 of the diagnostic device 50.

[0103] As described above, based on the exclusion of frequency sequences [E], [E1], etc., each motor diagnostic device 50 also outputs to the database device 100 via the network input / output unit 51G a reference for the signal strength when extracting the spectral peak, or a reference for setting the spectral peak width, a reference for the signal strength used when an anomaly is determined, and correct / incorrect information of the diagnostic results. The diagnostic algorithm modification device 110 at the manufacturer can analyze this information stored in the database device 100 and modify the diagnostic algorithm to modify these references, thereby preventing false detections by each motor diagnostic device 50.

[0104] As described above, according to embodiment 5, The monitoring and diagnostics unit includes a network input / output unit. The network input / output unit will input the specification information of the power conversion device, the motor, and the power transmission mechanism into the input unit; The monitorable video band, the characteristic frequency sequence, and the exclusion frequency sequence calculated by the analysis unit; The peak value of the current spectrum compared by the determination unit; and At least one of the data from the determination result shown by the determination unit is output to an external database device.

[0105] In addition, it includes diagnostic devices for multiple electric motors; It also includes: the database device, which is connected to diagnostic devices for each of the motors; and A diagnostic algorithm modification device is connected to the database device and the diagnostic devices of each of the motors, and modifies the diagnostic algorithm of each of the motor diagnostic devices.

[0106] Furthermore, the diagnostic device for each of the motors receives modification data from the diagnostic algorithm modification device via the network input / output unit, and modifies the diagnostic algorithm of the diagnostic device for each of the motors.

[0107] Furthermore, the items input by the user into the input section are modified by the modification data sent from the diagnostic algorithm modification device.

[0108] Furthermore, the diagnostic devices for each of the motors modify the content output to the database device via the network input / output unit by modifying the data of the diagnostic algorithm modification device.

[0109] Furthermore, the correctness or error information of the determination result shown by the determination unit is input to the input unit. When the network input / output unit outputs to the outside the specifications of the power conversion device, the motor, the power transmission mechanism, the monitorable video band, the characteristic frequency sequence, and the determination and exclusion frequency sequence, Add correct error information for the judgment result.

[0110] In the above embodiments, such as Figure 26 As shown in the example of hardware, the monitoring and diagnostic unit 51 of the diagnostic device 50 includes a processor 1000 and a storage device 1010. Although the storage device 1010 is not shown, it includes volatile storage devices such as random access memory and non-volatile auxiliary storage devices such as flash memory. Alternatively, an auxiliary storage device such as a hard disk drive may be included instead of flash memory. The processor 1000 executes a program input from the storage device 1010. In this case, the program is input from the auxiliary storage device to the processor 1000 via the volatile storage device. In addition, the processor 1000 can output data such as calculation results to the volatile storage device of the storage device 1010, and can also save data to the auxiliary storage device via the volatile storage device.

[0111] Although this application describes various exemplary embodiments and examples, the various features, methods and functions described in one or more embodiments are not limited to the application of a particular embodiment and can be applied to the embodiment individually or in various combinations. Therefore, numerous variations not illustrated can be conceived within the scope of the technology disclosed in this application. For example, this could include variations, additions, or omissions of at least one structural element, or the extraction of at least one structural element and its combination with structural elements of other embodiments. Label Explanation

[0112] 3 electric motors, 20 power conversion devices, 50 diagnostic devices, 51 monitoring and diagnostic units, 51A detection units, 51B analysis units, 51C input units, 51D setting units, 51E storage units, 51F judgment units, 51G network input / output units, 60 power transmission mechanisms, 100 database devices, and 110 diagnostic algorithm modification devices.

Claims

1. A diagnostic device for an electric motor, diagnosing at least one of anomalies in an electric motor driven by electricity converted by a power conversion device and anomalies in a power transmission mechanism connected to said electric motor, characterized in that, The diagnostic device includes: A current detector that detects the current flowing from the power conversion device to the motor; and The monitoring and diagnostic unit performs spectrum analysis based on the current detected by the current detector and monitors for at least one of the abnormalities of the motor and the power transmission mechanism. The monitoring and diagnostic unit includes: The input section is for inputting at least the specifications of the power conversion device and the motor. The setting unit calculates, based on the information from the input unit, a characteristic frequency sequence for diagnosing at least one of the abnormalities of the motor and the power transmission mechanism, a monitoring frequency range for monitoring the abnormality, and a noise frequency sequence that can be derived from the specification information of the power conversion device. The analysis unit performs spectral analysis based on the current detected by the current detector and calculates the current spectrum. It extracts a peak frequency sequence with a spectral peak value above a predetermined signal strength within the monitoring frequency range from the current spectrum. It compares the peak frequency sequence with the characteristic frequency sequence and the noise frequency sequence to derive a decision exclusion frequency sequence and a monitoring characteristic frequency sequence. It also derives a monitorable video band obtained by subtracting the decision exclusion frequency sequence from the monitoring frequency range. A storage unit that stores the monitorable video tape, the monitoring characteristic frequency sequence, and its signal strength; and The determination unit determines at least one abnormality, either an abnormality of the motor or an abnormality of the power transmission mechanism, by comparing the monitorable video tape and the monitoring characteristic frequency sequence stored in the storage unit with a current spectrum newly acquired from the current detector.

2. The diagnostic device for an electric motor as described in claim 1, characterized in that, Before the diagnostic device diagnoses at least one of the abnormalities in the electric motor and the power transmission mechanism, The analysis unit stores the monitorable video band, the characteristic frequency sequence, and their signal strength as normal data in the storage unit. When the diagnostic device diagnoses at least one of the abnormalities in the electric motor and the power transmission mechanism, The determination unit reads the monitorable video band, the characteristic frequency sequence, and its intensity as normal data from the storage unit, and compares them with the current spectrum newly acquired from the current detector. When the change in signal strength of the monitoring feature frequency sequence contained in the monitorable video band compared to normal data exceeds a preset threshold, it is determined to be abnormal.

3. The diagnostic device for an electric motor as described in claim 2, characterized in that, When the diagnostic device diagnoses an abnormality in the motor, The input section inputs the specifications of the power conversion device and the motor. The setting unit uses the specification information of the power conversion device input to the input unit to calculate the noise frequency sequence, and uses the specification information of the power conversion device and the motor input to the input unit to calculate the characteristic frequency sequence. The analysis unit classifies the various spectral peaks within the monitored frequency range into the noise frequency sequence, the characteristic frequency sequence, and unclassifiable sequences based on the current spectrum acquired by the current detector. The noise frequency sequence and the unclassifiable sequence are defined as the decision exclusion frequency sequences. The monitorable video band is derived from the determined exclusion frequency sequence.

4. The diagnostic device for an electric motor as described in claim 2, characterized in that, When the diagnostic device diagnoses an abnormality in the power transmission mechanism, The input section inputs the specifications of the power conversion device, the electric motor, and the power transmission mechanism. The setting unit uses the specification information of the power conversion device input to the input unit to calculate the noise frequency, and uses the specification information of the power conversion device, the motor, and the power transmission mechanism input to the input unit to calculate the characteristic frequency sequence. The analysis unit classifies the various spectral peaks into the noise frequency sequence, the characteristic frequency sequence, and unclassifiable sequences based on the current spectrum acquired by the current detector. The noise frequency sequence and the unclassifiable sequence are defined as the decision exclusion frequency sequences. The monitorable video bands are calculated based on the frequency sequence excluded by the determination.

5. The diagnostic device for an electric motor as described in any one of claims 2 to 4, characterized in that, When the analysis unit confirms the settings of the monitored feature frequency sequence and the decision exclusion frequency sequence, By changing the load rate of the motor, the current spectrum is obtained by the current detector. It was confirmed that the peak value of the current spectrum corresponding to each component of the monitored characteristic frequency sequence shifts according to the load rate of the motor. If the offset cannot be confirmed, the component of the monitoring characteristic frequency sequence corresponding to the peak value of the current spectrum for which the offset cannot be confirmed is excluded from the monitoring characteristic frequency sequence, and added to the decision exclusion frequency sequence.

6. The diagnostic device for an electric motor as described in any one of claims 2 to 5, characterized in that, When the ratio of the monitorable video band to the monitoring frequency range is below a preset ratio threshold, the analysis unit proposes to change the setting of the power conversion device's specification information.

7. The diagnostic device for an electric motor as described in any one of claims 2 to 6, characterized in that, The monitoring and diagnostics unit includes a network input / output unit. The network input / output unit will input the specification information of the power conversion device, the motor, and the power transmission mechanism into the input unit; The monitorable video band, the characteristic frequency sequence, and the exclusion frequency sequence calculated by the analysis unit; The peak value of the current spectrum compared by the determination unit; and At least one of the data from the determination result shown by the determination unit is output to an external database device.

8. A diagnostic system for an electric motor, characterized in that, Includes multiple diagnostic devices for electric motors as described in claim 7; It also includes: the database device, which is connected to diagnostic devices for each of the motors; and A diagnostic algorithm modification device is connected to the database device and the diagnostic devices of each of the motors, and modifies the diagnostic algorithm of each of the motor diagnostic devices.

9. The diagnostic system for an electric motor as described in claim 8, characterized in that, Each of the motor diagnostic devices receives modification data from the diagnostic algorithm modification device via the network input / output unit, and modifies the diagnostic algorithm of each of the motor diagnostic devices.

10. The diagnostic system for an electric motor as described in claim 9, characterized in that, The user inputs items to the input unit are modified by the modification data sent from the diagnostic algorithm modification device.

11. The diagnostic system for an electric motor as described in claim 9, characterized in that, The diagnostic devices for each of the motors modify the content output to the database device via the network input / output unit by modifying the data of the diagnostic algorithm modification device.

12. The diagnostic system for an electric motor as described in claim 9, characterized in that, The correctness or error information of the determination result shown by the determination unit is input into the input unit. When the network input / output unit outputs to the outside the specifications of the power conversion device, the motor, the power transmission mechanism, the monitorable video band, the characteristic frequency sequence, and the determination and exclusion frequency sequence, Add correct error information for the judgment result.

Images