Abnormality diagnosing device, power conversion device, and abnormality diagnosing method

Abstract

An abnormality diagnosing device (30) detects a current flowing through a motor (2) driven by a pulse width modulation control of a power conversion device (100) and performs frequency analysis, and a determination section (34) determines an abnormality of the motor (2) based on a spectrum peak value of at least one sideband component of a modulation wave obtained from the analysis result. The abnormality diagnosing device (30) includes a frequency setting section (33) that sets a noise frequency (fnα) in the current in advance, and the determination section (34) estimates whether or not there is noise interference in the spectrum peak value of the sideband component based on a frequency of the sideband component and the set noise frequency (fnα), thereby performing abnormality determination.

Description

Technical Field

[0001] The present application relates to an abnormality diagnosis device for diagnosing abnormality of a motor, a power conversion device including the abnormality diagnosis device and driving the motor, and a method for diagnosing abnormality of the motor. Background Art

[0002] To diagnose motor anomalies during operation, a conventional method, such as that described in Patent Document 1, performs frequency analysis on the current flowing through the motor and diagnoses anomalies based on frequency components that appear as sidebands of the power supply frequency component. This method then subtracts two cycles of the current flowing through the motor with the same phase from each other to cancel out noise components and extract pulsation components that appear when a rotor anomaly occurs, thereby enabling anomaly diagnosis.

[0003] In addition, in the existing method described in Patent Document 2, when the induction motor is driven by PWM (pulse width modulation) control of the inverter, the noise component generated in the vibration spectrum is removed, the spectrum without the noise component is inverse Fourier transformed, and the vibration acceleration waveform collected by the induction motor is obtained in the form of the noise component removed.

[0004] Prior art literature

[0005] Patent Literature

[0006] Patent Document 1: Japanese Patent Application Laid-Open No. 2003-274691

[0007] Patent Document 2: Japanese Patent Application Laid-Open No. 2016-116251 Summary of the Invention

[0008] Technical problem to be solved by the invention

[0009] In the conventional abnormality diagnosis method described in Patent Document 1, only noise components of the same phase and magnitude are canceled out within each cycle of the current flowing through the motor. However, noise components vary depending on the motor's driving conditions or abnormal conditions, and some noise components remain that cannot be reduced, making it difficult to reliably extract the frequency components used for abnormality diagnosis.

[0010] In the conventional abnormality diagnosis described in Patent Document 2, the acceleration component caused by the carrier frequency, ie, the noise signal, is avoided. Therefore, other noise components, particularly low-frequency noise components, remain, making it difficult to perform abnormality diagnosis with high reliability.

[0011] The present application discloses a technology for solving the above-mentioned problem, and aims to provide an abnormality diagnosis device that can diagnose abnormalities of a motor driven by pulse width modulation control of a power conversion device with high reliability by preventing the influence of noise including low-frequency domains.

[0012] Furthermore, an object of the present application is to provide a power conversion device including such an abnormality diagnosis device, which diagnoses abnormality of a motor with high reliability and drives the motor.

[0013] Furthermore, an object of the present application is to provide an abnormality diagnosis method for diagnosing abnormalities of a motor driven by pulse width modulation control of a power conversion device with high reliability by preventing the influence of noise including low-frequency noise.

[0014] Technical means for solving technical problems

[0015] The abnormality diagnosis device disclosed in the present application diagnoses abnormalities in a motor driven by pulse width modulation control of a power conversion device. The abnormality diagnosis device includes: a detection unit for detecting current flowing through the motor; an analysis unit for performing frequency analysis on the current detected by the detection unit and outputting the analysis result;

[0016] A determination unit is configured to determine an abnormality in the motor based on a spectrum peak of at least one sideband component of the modulated wave obtained from the analysis result; and a frequency setting unit is configured to pre-set a noise frequency in the current. The determination unit then estimates whether or not noise interference is present in the spectrum peak of the sideband component based on the frequency of the sideband component and the set noise frequency, and determines an abnormality in the motor.

[0017] In addition, the power conversion device disclosed in the present application includes a power conversion unit that converts direct current into alternating current and supplies power to the motor, and a control device that controls the output of the power conversion unit through the pulse width modulation control, and the control device includes the abnormality diagnosis device to diagnose abnormalities of the motor.

[0018] Furthermore, the abnormality diagnosis method disclosed in this application is a method for diagnosing abnormalities in a motor driven by pulse width modulation control of a power conversion device, comprising: a first step of calculating the greatest common divisor of two or more frequencies including the modulation wave frequency, among the three frequencies used for pulse width modulation control: the modulation wave frequency, the carrier frequency, and the sampling frequency for sampling the modulation wave; and setting a frequency that is an integer multiple of the greatest common divisor as a noise frequency; a second step of detecting the current flowing through the motor and performing frequency analysis; and a third step of determining abnormalities in the motor based on the spectral peaks of the modulation wave sideband components obtained from the analysis results of the second step. Then, in the third step, the presence or absence of noise interference in the spectral peaks of the sideband components is estimated based on the frequencies of the sideband components and the noise frequency set in the first step.

[0019] Furthermore, the abnormality diagnosis method disclosed in this application is a method for diagnosing abnormalities in an electric motor driven by pulse width modulation control of a power conversion device. The method includes: a first step of setting a frequency that deviates from the frequency of a modulation wave used for pulse width modulation control by an integer multiple of the frequency of an AC power source connected to the power conversion device as a noise frequency; a second step of detecting the current flowing through the electric motor and performing frequency analysis; and a third step of determining an abnormality in the electric motor based on the spectral peaks of the sideband components of the modulation wave obtained from the analysis results of the second step. Then, in the third step, based on the frequencies of the sideband components and the noise frequency set in the first step, it is estimated whether there is noise interference in the spectral peaks of the sideband components.

[0020] Effects of the Invention

[0021] According to the abnormality diagnosis device disclosed in the present application, it is possible to diagnose abnormalities of a motor driven by pulse width modulation control of a power conversion device with high reliability, while preventing the influence of noise including low-frequency noise.

[0022] Furthermore, according to the power conversion device disclosed in the present application, it is possible to prevent the influence of noise including low-frequency noise and diagnose abnormalities of a motor driven by pulse width modulation control of the power conversion device with high reliability.

[0023] Furthermore, according to the abnormality diagnosis method disclosed in the present application, it is possible to prevent the influence of noise including low-frequency noise and diagnose abnormalities of a motor driven by pulse width modulation control of a power conversion device with high reliability. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 This is a diagram showing the configuration of a power conversion device and an abnormality diagnosis device according to the first embodiment.

[0025] Figure 2 This is a block diagram showing a schematic configuration of the abnormality diagnosis device according to the first embodiment.

[0026] Figure 3 This is a block diagram showing a partial hardware configuration of the abnormality diagnosis device according to the first embodiment.

[0027] Figure 4 This is a diagram illustrating a spectrum waveform of a current in the abnormality diagnosis device according to the first embodiment.

[0028] Figure 5 This is a waveform diagram illustrating pulse width modulation control of the power conversion device according to the first embodiment.

[0029] Figure 6 This is a flowchart illustrating the operation of the abnormality diagnosis device according to the first embodiment.

[0030] Figure 7 is a block diagram showing a brief structure of an abnormality diagnosing apparatus according to Embodiment 2.

[0031] Figure 8 is a block diagram showing a brief structure of an abnormality diagnosing apparatus according to Embodiment 3.

[0032] Figure 9 is a diagram showing a structure of a power conversion apparatus and an abnormality diagnosing apparatus according to Embodiment 4.

[0033] Figure 10 is a block diagram showing a brief structure of an abnormality diagnosing apparatus according to Embodiment 4.

[0034] Figure 11 is a schematic diagram of a frequency spectrum waveform of a current for explaining an effect according to Embodiment 6.

[0035] Figure 12 is a flowchart showing an operation of an abnormality diagnosing apparatus according to Embodiment 4.

[0036] Figure 13 is a diagram showing a structure of a power conversion apparatus and an abnormality diagnosing apparatus according to Embodiment 5.

[0037] Figure 14 is a diagram showing a carrier according to another example of Embodiment 5.

[0038] Figure 15 is a schematic diagram of a frequency spectrum waveform of a current for explaining an effect according to Embodiment 6. DETAILED DESCRIPTION

[0039] Embodiment 1.

[0040] Figure 1 is a diagram showing a structure of a power conversion apparatus and an abnormality diagnosing apparatus according to Embodiment 1.

[0041] As shown in Figure 1 , a power conversion apparatus 100 is connected between an alternating current power supply 1 constituted by, for example, a commercial power supply and a motor 2, and drives and controls the motor 2. The power conversion apparatus 100 includes a power conversion section 10 and a control apparatus 20 that outputs-controls the power conversion section 10.

[0042] Further, a current i flowing from the power conversion section 10 to the motor 2 is detected by a current sensor 3, and an abnormality diagnosing apparatus 30 diagnoses an abnormality of the motor 2 on the basis of the current i. In addition, the current sensor 3 can be built in the power conversion apparatus 100, or can be externally attached, and the number and position of the current sensor 3 are not limited to those shown in the diagram.

[0043] The power conversion unit 10 includes a converter unit 10A, an inverter unit 10B, and a filter capacitor 10C, which are connected via a DC bus. The converter unit 10A converts AC power from the AC power supply 1 into DC power and outputs it to the filter capacitor 10C. The inverter unit 10B converts the DC power from the filter capacitor 10C into AC power and supplies it to the motor 2.

[0044] In this case, the AC power supply 1 , the motor 2 , and the power conversion device 100 are shown as a three-phase structure, but are not limited thereto.

[0045] Converter unit 10A is composed of a three-phase bridge circuit with six diodes Da, and the input and output lines of each phase are connected to AC power supply 1. Inverter unit 10B is composed of a three-phase bridge circuit with six switching elements Q, each of which has a diode Db connected in antiparallel. The input and output lines of each phase are connected to motor 2. Switching elements Q are, for example, IGBTs (Insulated Gate Bipolar Transistors) or MOSFETs (Metal-oxide-semiconductor Field Effect Transistors).

[0046] AC power from AC power source 1 is rectified by converter unit 10A, converted into DC power, and output to filter capacitor 10C. Control device 20 generates gate signals G for each switching element Q of inverter unit 10B through pulse width modulation control (PWM control), and controls the switching elements Q to switch on and off, thereby outputting desired power from power conversion unit 10 to motor 2. Thus, power conversion device 100 drives motor 2.

[0047] The configuration of the converter unit 10A and the inverter unit 10B is not limited to that shown in the figure. In this case, the power conversion unit 10 is shown as including the converter unit 10A and connected to the AC power supply 1. However, as long as the inverter unit 10B is provided to convert DC power into AC power and supply power to the motor 2, the converter unit 10A may be omitted.

[0048] Abnormality diagnosis device 30 obtains the frequencies of the modulation wave (fundamental wave), carrier wave, and clock signal (CLK) used for sampling by control device 20 for PWM control of power conversion unit 10, namely, modulation wave frequency f0, carrier wave frequency fc, and sampling frequency fs. Abnormality diagnosis device 30 then performs frequency analysis on current i flowing from power conversion unit 10 through motor 2 to diagnose abnormalities in motor 2.

[0049] Figure 2 is a block diagram showing a schematic structure of the abnormality diagnosis device 30. Figure 2As shown, the abnormality diagnosis device 30 includes a detection unit 31 for detecting the current i flowing through the motor 2, an analysis unit 32 for performing frequency analysis on the current i, a frequency setting unit 33 for presetting the frequency of the noise in the current i (noise frequency fnα), and a determination unit 34 for determining the abnormality of the motor 2.

[0050] The detection unit 31 acquires the output of the current sensor 3 and detects the current waveform of at least one phase current i flowing through the motor 2. The analysis unit 32 performs frequency analysis based on the detected current i and derives analysis results 32a including a spectrum waveform. The frequency setting unit 33 acquires the modulation wave frequency f0, the carrier frequency fc, and the sampling frequency fs, calculates their greatest common divisor GCD, and sets the GCD and its integer multiples as the noise frequency fnα.

[0051] Determination unit 34 obtains the spectrum peaks of the sideband components of the modulated wave from analysis result 32a of analysis unit 32, determines an abnormality in motor 2 based on the spectrum peaks, and outputs determination result 34a. At this time, based on noise frequency fnα, it estimates whether or not there is noise interference in the spectrum peaks of the sideband components of the modulated wave, and excludes sideband components estimated to be subject to noise interference from the abnormality determination.

[0052] In addition, the hardware constituting the abnormality diagnosis device 30 may be a combination of a known dedicated device for frequency analysis and a hardware component such as a controller. Figure 3 A processor 5 and a storage device 6 are shown.

[0053] Processor 5 executes a control program input from storage device 6. Storage device 6 includes an auxiliary storage device and a volatile storage device. The control program is input from the auxiliary storage device to processor 5 via the volatile storage device. Processor 5 outputs data such as calculation results to the volatile storage device of storage device 6 and, as needed, stores this data in the auxiliary storage device via the volatile storage device.

[0054] When motor 2 shows signs of an abnormality, a modulated wave with a specific frequency component, called a sideband component, increases in current i. For example, if the rotor's rotational frequency is fr due to dynamic eccentricity or vibration caused by an abnormality, then a sideband component of |k1·f0±k2·fr| increases based on the frequency (f0±fr). Here, k1 and k2 are positive integers.

[0055] Furthermore, if the conductor bars of the squirrel-cage rotor are damaged, when the slip is s, the sideband component of the frequency ((1±2s)·f0) increases.

[0056] Furthermore, when a bearing has a scratch, the sideband components that deviate from the modulating wave frequency f0 only increase in the characteristic frequency determined by the scratch location and bearing shape. For example, when the outer race of a bearing is damaged, the characteristic frequency is N·fr(1-dcosθ / D) / 2.

[0057] Where N, d, D, and θ are the number of balls in the bearing, the ball diameter, the pitch diameter, and the contact angle, respectively.

[0058] In the following, when simply referring to a sideband or a sideband component, it means a sideband of a modulated wave or a sideband component of a modulated wave.

[0059] Figure 4 1 is a diagram illustrating a spectrum waveform of the current i in the abnormality diagnosis device 30 when an abnormality occurs in the electric motor 2 .

[0060] like Figure 4 As shown, multiple spectra 41 and 42 appear on both sides of the spectrum 40 of the modulation wave frequency f0. In this case, at frequencies (f0±fr) on both sides of the modulation wave frequency f0 that deviate from the rotation frequency fr, the spectrum 41 of the sideband component of the modulation wave appears, and the spectrum 42 of the noise component caused by the switching operation of the inverter unit 10B also appears.

[0061] exist Figure 4 In the example, the spectrum 41 of the sideband component and the spectrum 42 of the noise component are neither close to nor overlapping, so the spectrum 41 of the sideband component indicating an abnormality can be distinguished from the spectrum 42 of the noise component and detected. Furthermore, when conditions change, the spectrum 42 of the noise component may approach the spectrum 41 of the sideband component indicating an abnormality, sometimes causing noise interference at the spectrum peak (not shown).

[0062] In this case, only the spectrum 41 when k1=k2=1 in the frequency |k1·f0±k2·fr| is shown. However, if a spectrum with a small spectrum peak is included, a spectrum 41 with a combination other than k1=k2=1 will generally appear.

[0063] Figure 5 It is a waveform diagram for explaining the PWM control of the power conversion device 100 .

[0064] like Figure 5 As shown, in PWM control, the modulation wave M is compared with the carrier wave Cr to generate the gate signal G. At this time, the modulation wave M is sampled at the timing of the clock signal (CLK), the value of the modulation wave M is temporarily stored, and compared with the carrier wave Cr.

[0065] If the carrier frequency fc or the sampling frequency fs is not a multiple of the modulation wave frequency f0, a spectrum 42 of a noise component caused by the switching operation of the inverter unit 10B is generated at a frequency of the greatest common divisor of their values ​​and an integral multiple thereof.

[0066] Therefore, by calculating the greatest common divisor of the carrier frequency fc or the sampling frequency fs and the modulation wave frequency f0, or all three frequencies, it is possible to grasp in advance which frequency noise may occur.

[0067] In this case, the frequency setting unit 33 obtains the modulation wave frequency f0, the carrier frequency fc, and the sampling frequency fs, calculates their greatest common divisor GCD, and sets the greatest common divisor GCD and its integer multiples as the noise frequency fnα. If the greatest common divisor GCD is the modulation wave frequency f0, the noise frequency fnα is not set.

[0068] Then, based on Figure 6 The operation of the abnormality diagnosis device 30 will be described with reference to the flowchart shown.

[0069] First, the abnormality diagnosis device 30 uses the detection unit 31 to detect the current waveform of at least one phase of the current i flowing from the power conversion unit 10 of the power conversion device 100 to the motor 2. In this case, the detection unit 31 detects the current waveform of the three phases. The current sensor 3 can detect the current i of each of the three phases, or it can detect the current i of two phases and obtain the current of the remaining phases through calculation (step S1).

[0070] Next, the analysis unit 32 performs frequency analysis based on the detected current i and derives an analysis result 32 a including a spectrum waveform (step S2 ).

[0071] On the other hand, the frequency setting unit 33 obtains the modulation wave frequency f0, the carrier frequency fc, and the sampling frequency fs from the control device 20 of the power conversion device 100 (step S3). The frequency setting unit 33 then calculates the greatest common divisor GCD of the modulation wave frequency f0, the carrier frequency fc, and the sampling frequency fs, and further calculates an integer multiple of the greatest common divisor GCD (step S4).

[0072] If the greatest common divisor GCD is not the modulation wave frequency f0, the calculated greatest common divisor GCD and its integer multiples are set as the noise frequency fnα. In addition, the noise frequency fnα is set within the range that does not exceed the measurable area.

[0073] The greatest common denominator GCD is a value lower than fc / 2. Furthermore, under normal control conditions of the power conversion device 100, the greatest common denominator GCD becomes a value lower than (fc-4f0). Therefore, the set noise frequency fnα includes frequencies in the frequency domain lower than fc / 2 and is typically set to include frequencies lower than (fc-4f0) (step S5).

[0074] The determination unit 34 estimates whether there is noise interference in the spectrum peak of the sideband component of the modulated wave based on the analysis result 32a derived in step S2 and the noise frequency fnα set in step S5. Specifically, the determination unit 34 determines whether the frequency of the sideband component of the modulated wave (spectrum 41) overlaps with or is close to the noise frequency fnα to estimate the presence of noise interference.

[0075] The sideband component (spectrum 41) of the modulated wave that increases as a sign of an abnormality in motor 2 is a specific frequency component as described above. Therefore, determination unit 34 monitors this specific frequency component and compares it with the noise frequency fnα. If the difference is less than a set value, it determines that the frequencies overlap or are close to each other. The set value is set to a few Hz, for example, 2 Hz. If the difference is greater than the set value, the peak of spectrum 41 is not affected by the noise component, and noise interference is eliminated (step S6).

[0076] If a sideband component estimated to be noise interference is present in step S6, determination unit 34 excludes that sideband component from the target of abnormality diagnosis (step S7), determines an abnormality in motor 2 based on the remaining sideband components, and outputs determination result 34a. At this time, an abnormality is determined if the spectral peak of the sideband component exceeds a preset reference value. The reference value is set, for example, based on the spectral peak of the modulation wave frequency f0 (step S8).

[0077] If the greatest common divisor GCD is the modulation wave frequency f0 in step S5, the frequency setting unit 33 does not set the noise frequency fnα and moves to step S8. The determination unit 34 then determines whether the motor 2 is abnormal based on the spectrum peak of the sideband component.

[0078] As described above, the abnormality diagnosis device 30 of this embodiment predetermines the frequency of the noise component (noise frequency fnα) in the current i flowing through the motor 2 and performs abnormality diagnosis on the sideband components of the modulated wave obtained by frequency analysis of the current i. During abnormality diagnosis, the device estimates the presence of noise interference in the spectral peaks of the sideband components based on the sideband component frequencies and the noise frequency fnα. Sideband components estimated to contain noise interference are excluded, and abnormality is determined based on the spectral peaks of the remaining sideband components.

[0079] Therefore, it is possible to prevent erroneous diagnosis caused by the influence of noise in the low-frequency region, and to perform abnormality diagnosis of the electric motor 2 with high reliability.

[0080] Furthermore, since the noise frequency fnα is set to the greatest common divisor GCD of the modulation wave frequency f0, the carrier frequency fc, and the sampling frequency fs for PWM control, and its integer multiples, the influence of noise components including low-frequency regions can be reliably prevented.

[0081] Furthermore, when the difference between the frequency of the sideband component and the set noise frequency fnα is smaller than the set value and the two are close to each other, the sideband component is estimated to be noise interference, so that noise interference can be estimated with high reliability.

[0082] The noise frequency fnα is set within a range that does not exceed the measurable region, but may be set only within a frequency range lower than 1 / 2 of the carrier frequency fc.

[0083] In addition, when the greatest common divisor GCD calculated by the frequency setting unit 33 is the modulation wave frequency f0, the noise frequency fnα is not set, but there is no problem even if the modulation wave frequency f0 and its integer multiples are directly set as the noise frequency fnα, because it is not a frequency component close to the sideband component of the modulation wave.

[0084] Furthermore, regarding the PWM control of the power conversion device 100 , the carrier wave Cr is illustrated as being based on a triangular wave. However, the carrier wave Cr is not limited to a triangular wave, and a sine wave may also be used.

[0085] Furthermore, in order to improve the voltage utilization rate, a third harmonic may be superimposed on the modulated wave M. In this case, the value of the greatest common divisor GCD does not change, and the noise frequency fnα can be set in the same manner.

[0086] Implementation method 2.

[0087] Figure 7 This is a block diagram showing a schematic configuration of an abnormality diagnosis device 30A according to the second embodiment.

[0088] like Figure 7 As shown, the abnormality diagnosis device 30A includes a detection unit 31 , an analysis unit 32 , a frequency setting unit 33 , and a determination unit 34 , similarly to the first embodiment, and further includes a notification unit 35 .

[0089] If there is a sideband component estimated to be noise interference, the determination unit 34 excludes the sideband component from the target of abnormality diagnosis (see Figure 6 Step S7) and outputs a notification instruction 34b to the notification unit 35. Then, the notification unit 35 outputs a notification signal 35a to notify the outside that there is noise interference. Other structures and operations are the same as those in the above-mentioned embodiment 1.

[0090] In this embodiment, as in the first embodiment, erroneous diagnosis due to the influence of noise components including low-frequency components can be prevented, thereby reliably performing abnormality diagnosis of the motor 2. Furthermore, during diagnosis, the user is informed of the estimated presence of noise interference, thereby improving convenience.

[0091] Furthermore, even if there is a sideband component estimated to be noise interference, the sideband component may not be excluded from the abnormality diagnosis target, and only the notification signal 35a may be output from the notification unit 35. In this case, the user is notified to draw attention, and the user can consider the influence of the noise component on the determination result 34a from the abnormality diagnosis device 30A, thereby preventing an erroneous diagnosis.

[0092] Implementation method 3.

[0093] Figure 8 This is a block diagram showing a schematic configuration of an abnormality diagnosis device 30B according to the third embodiment.

[0094] like Figure 8 As shown, abnormality diagnosis device 30B, like the first embodiment, includes a detection unit 31, an analysis unit 32, and a frequency setting unit 33. It also includes a determination unit 36, a noise detection unit 37, a storage unit 38, and a switch 39. The structure and operation other than determination unit 36, noise detection unit 37, storage unit 38, and switch 39 are the same as those of the first embodiment.

[0095] Similar to the first embodiment, the detection unit 31 acquires the output of the current sensor 3 and detects the current waveform of the current i flowing through at least one phase of the motor 2. The analysis unit 32 performs frequency analysis based on the detected current i and derives analysis results 32a including a spectral waveform. The frequency setting unit 33 acquires the modulation wave frequency f0, the carrier frequency fc, and the sampling frequency fs, calculates their greatest common divisor (GCD), and sets the GCD and its integer multiples as the noise frequency fnα within the range outside the measurable range.

[0096] Then, when the motor 2 is operating normally, the noise detection unit 37 detects the magnitude of the noise at the noise frequency fnα of the current i, for example, the value of the spectrum peak of the noise component, from the analysis result 32a of the analysis unit 32, and stores the detection result in the storage unit 38. Prior to diagnosing an abnormality in the motor 2, noise detection by the noise detection unit 37 is performed in advance during normal operation of the motor 2.

[0097] The switch 39 selectively switches the output destination of the analysis result 32a of the analysis unit 32 to one of the noise detection unit 37 and the determination unit 36. When the motor 2 is undergoing abnormality diagnosis, the determination unit 36 ​​is selected, and when noise detection is performed in advance during normal operation of the motor 2, the noise detection unit 37 is selected.

[0098] Determination unit 36 ​​obtains the spectrum peak of the sideband component of the modulated wave from analysis result 32a of analysis unit 32, determines an abnormality in motor 2 based on the spectrum peak, and outputs determination result 36a. At this time, based on noise frequency fnα, it estimates whether there is noise interference in the spectrum peak of the sideband component of the modulated wave. Specifically, as in the first embodiment described above, when the difference between the frequency of the sideband component of the modulated wave and the noise frequency fnα is less than a set value, it is determined that the frequency of the sideband component overlaps or is close to the noise frequency fnα, and noise interference is estimated.

[0099] Next, determination unit 36 ​​extracts the magnitude of the noise at frequency fnα, the source of the noise interference, from storage unit 38. It then determines an abnormality in motor 2 based on the peak value of the sideband component's spectrum and the magnitude of the extracted noise. Specifically, if the value obtained by subtracting the peak value of the noise component's spectrum from the peak value of the sideband component's spectrum exceeds a predetermined reference value, an abnormality is determined. The reference value is set, for example, based on the peak value of the spectrum of the modulation wave frequency f0.

[0100] As described above, abnormality diagnosis device 30B according to this embodiment predetermines the frequency of the noise component (noise frequency fnα) in current i flowing through motor 2 and performs abnormality diagnosis on the sideband components of the modulated wave obtained by frequency analysis of current i. Furthermore, prior to abnormality diagnosis, during normal operation of motor 2, abnormality diagnosis device 30B detects the magnitude of noise at noise frequency fnα in current i based on analysis results 32a from analysis unit 32 and stores this detection result. Then, during abnormality diagnosis, the presence of noise interference in the spectral peaks of the sideband components is estimated based on the frequency of the sideband components and the noise frequency fnα. The spectral peaks of the sideband components estimated to be subject to noise interference, taking into account the magnitude of the noise, are used for abnormality determination.

[0101] Therefore, similar to the first embodiment described above, it is possible to prevent erroneous diagnosis caused by the influence of noise components including low-frequency components, thereby reliably performing abnormality diagnosis of the motor 2. Furthermore, since sideband components estimated to be affected by noise interference are also used for abnormality diagnosis without being excluded, the sideband components of the monitoring target for abnormality diagnosis can be reliably monitored, thereby reliably performing abnormality diagnosis of the motor 2.

[0102] While noise detection by noise detection unit 37 is based on analysis results 32a of current i during normal operation of motor 2, it can also be based on the results of frequency analysis of the output voltage supplied to motor 2. In this case, detection during normal operation of motor 2 is not necessary; the magnitude of the noise equivalent to that during normal operation at the noise frequency fnα of current i can be calculated based on the frequency analysis results of the detected voltage. This calculation result can then be used by determination unit 36 ​​instead of being stored in storage unit 38, or storage unit 38 can be omitted.

[0103] Furthermore, the determination unit 36 ​​can use both the frequency analysis result of the detected voltage and the noise detection result obtained based on the analysis result 32 a of the current i when the motor 2 is operating normally, thereby improving the accuracy of abnormality determination.

[0104] Furthermore, in this third embodiment, the above-mentioned second embodiment may be applied to provide a notification unit 35 to notify the user of the estimation of the presence of noise interference.

[0105] Implementation method 4.

[0106] Figure 9 1 is a diagram showing the configuration of a power conversion device 100 and an abnormality diagnosis device 30C according to a fourth embodiment.

[0107] like Figure 9 As shown, power conversion device 100 is configured similarly to that of the first embodiment, and includes a power conversion unit 10 and a control device 20 for outputting and controlling power conversion unit 10. Furthermore, current i flowing from power conversion unit 10 to motor 2 is detected by current sensor 3, and abnormality diagnosis device 30 diagnoses abnormalities in motor 2 based on current i.

[0108] The power conversion unit 10 includes a converter unit 10A, an inverter unit 10B, and a filter capacitor 10C, which are connected via a DC bus. In this embodiment, the converter unit 10A is essential and converts AC power from the AC power source 1 into DC power, which is then output to the filter capacitor 10C. The inverter unit 10B converts the DC power from the filter capacitor 10C into AC power, which is then supplied to the motor 2.

[0109] In this case, in the power conversion unit 10 , the AC power source 1 , the motor 2 , and the power conversion device 100 are shown as a three-phase structure, but the present invention is not limited thereto.

[0110] AC power from AC power source 1 is rectified by converter unit 10A, converted into DC power, and output to filter capacitor 10C. Control device 20 generates gate signals G for each switching element Q of inverter unit 10B through PWM control, and controls the switching elements Q to switch on and off, thereby outputting desired power from power conversion unit 10 to motor 2.

[0111] Thus, the power conversion device 100 drives the motor 2. Then, the DC voltage of the filter capacitor 10C and the AC voltage output to the motor 2 slightly fluctuate at the frequency of the AC power supply 1 and its integral multiples, and a sideband component (noise component) is generated in the current i that deviates from the modulation wave frequency f0 by this value.

[0112] Abnormality diagnosis device 30C obtains the frequency of the modulation wave (modulation wave frequency f0) used by control device 20 in PWM control of power conversion unit 10 and the frequency of AC power supply 1 (AC power supply frequency fac). Abnormality diagnosis device 30C then performs frequency analysis on current i flowing from power conversion unit 10 through motor 2 to diagnose abnormalities in motor 2.

[0113] Figure 10 FIG is a block diagram showing a schematic structure of the abnormality diagnosis device 30C. Figure 10 As shown, the abnormality diagnosis device 30C includes a detection unit 31 for detecting the current i flowing through the motor 2, an analysis unit 32 for performing frequency analysis on the current i, a frequency setting unit 33A for presetting the frequency of the noise in the current i (noise frequency fnβ), and a determination unit 34 for determining the abnormality of the motor 2.

[0114] The detection unit 31 and the analysis unit 32 have the same structure as those in the first embodiment described above and operate in the same manner.

[0115] The frequency setting unit 33A obtains the modulation wave frequency f0 and the AC power supply frequency fac, and sets the noise frequency fnβ by calculating the following frequency: where m and n are positive integers.

[0116] |m·fac±n·f0|

[0117] That is, the noise frequency fnβ is the absolute value of a value obtained by deviating from an integral multiple of the modulation wave frequency f0 by an integral multiple of the AC power supply frequency fac.

[0118] Determination unit 34 obtains the spectrum peaks of the sideband components of the modulated wave from analysis result 32a of analysis unit 32, determines an abnormality in motor 2 based on the spectrum peaks, and outputs determination result 34a. At this time, based on noise frequency fnβ, it estimates whether or not there is noise interference in the spectrum peaks of the sideband components of the modulated wave, and excludes sideband components estimated to be subject to noise interference from the abnormality determination.

[0119] Figure 11 This is the spectrum waveform of the current i used to illustrate the noise frequency.

[0120] like Figure 11 As shown in the figure, when the modulation wave frequency f0 and the AC power supply frequency fac are 50Hz and 60Hz respectively, the frequency spectrum of the noise component appears at multiples of the modulation wave frequency f0 (100Hz, 150Hz, 200Hz) and other frequencies 10Hz, 70Hz, 110Hz, 170Hz. If the frequencies of the noise components other than multiples of the modulation wave frequency f0 are expressed by the modulation wave frequency f0 (50Hz) and the AC power supply frequency fac (60Hz), then

[0121] 10Hz=fac-f0

[0122] 70Hz=2·fac-f0

[0123] 110Hz=fac+f0

[0124] 170Hz=2·fac+f0

[0125] , and satisfy the above-mentioned noise frequency fnβ calculation formula.

[0126] Then, based on Figure 12 The operation of the abnormality diagnosis device 30C will be described with reference to the flowchart shown.

[0127] First, the abnormality diagnosis device 30C is similar to the above-mentioned embodiment 1, and the detection unit 31 detects the current waveform of at least one phase of the current i among the phase currents i flowing from the power conversion unit 10 of the power conversion device 100 to the motor 2 (step S1). The analysis unit 32 performs frequency analysis based on the detected current i and derives the analysis result 32a including the spectrum waveform (step S2).

[0128] On the other hand, the frequency setting unit 33A acquires the modulation wave frequency f0 and the AC power supply frequency fac (step SS3 ).

[0129] Then, the frequency setting unit 33A performs the following calculation as described above (step SS4):

[0130] |m·fac±f0|

[0131] This is set as the noise frequency fnβ. Note that the noise frequency fnβ includes the case where m=n=1 and is set within a range that does not exceed the measurable region.

[0132] When m=n=1, that is, (fac±f0) is a value lower than fc / 2. Furthermore, under normal control conditions of the power conversion device 100, (fac±f0) becomes a value lower than (fc-4f0). Therefore, the set noise frequency fnβ includes frequencies in the frequency domain lower than fc / 2 and is typically set to include frequencies lower than (fc-4f0) (step S5).

[0133] Based on the analysis result 32a derived in step S2 and the noise frequency fnβ set in step S5, the determination unit 34 estimates whether there is noise interference in the spectrum peak of the sideband component of the modulated wave. Specifically, it determines whether the frequency of the sideband component (spectrum 41) of the modulated wave overlaps with or is close to the noise frequency fnβ, and infers the presence of noise interference. In this case, as in the first embodiment described above, the determination unit 34 monitors a specific frequency component (sideband component) that increases due to an abnormality sign, compares this frequency with the noise frequency fnβ, and if the difference is less than a set value, determines that there is overlap or close proximity, thereby inferring the presence of noise interference. In this case, the set value is set to a few Hz, for example, 2 Hz (step S6).

[0134] If a sideband component estimated to be noise interference is present in step S6, determination unit 34 excludes that sideband component from the target of abnormality diagnosis (step S7) and determines an abnormality in motor 2 based on the remaining sideband components. At this time, an abnormality is determined if the spectral peak of the sideband component exceeds a preset reference value. The reference value is set, for example, based on the spectral peak of the modulation wave frequency f0 (step S8).

[0135] As described above, abnormality diagnosis device 30C according to this embodiment predetermines the frequency of the noise component (noise frequency fnβ) in current i flowing through motor 2, and performs abnormality diagnosis on the sideband components of the modulated wave obtained by frequency analysis of current i. During abnormality diagnosis, the system estimates the presence of noise interference in the spectral peaks of the sideband components based on the sideband component frequencies and noise frequency fnβ. Sideband components estimated to contain noise interference are eliminated, and abnormality is determined based on the spectral peaks of the remaining sideband components.

[0136] Therefore, it is possible to prevent erroneous diagnosis caused by the influence of noise components including low-frequency noise components, that is, the modulation wave frequency f0 and the AC power supply frequency fac in this case, and perform abnormality diagnosis of the motor 2 with high reliability.

[0137] Furthermore, when the difference between the frequency of the sideband component and the set noise frequency fnβ is smaller than the set value and the two are close to each other, it is estimated that the sideband component has noise interference, and thus noise interference can be estimated with high reliability.

[0138] The noise frequency fnα is set within a range that does not exceed the measurable region, but may be set only within a frequency range lower than 1 / 2 of the carrier frequency fc.

[0139] Furthermore, in this fourth embodiment, the above-mentioned second embodiment may be applied to provide a notification unit 35 to notify the user of the estimation of noise interference.

[0140] Furthermore, the third embodiment described above can also be applied to this fourth embodiment. In this case, a noise detector 37, a storage unit 38, and a switch 39 are provided. Prior to abnormality diagnosis and during normal operation of the motor 2, the noise level at the noise frequency fnβ of the current i is detected and stored in advance. Then, during abnormality diagnosis, the presence of noise interference in the spectral peaks of the sideband components is estimated based on the sideband component frequencies and the noise frequency fnβ. The spectral peaks of the sideband components estimated to be subject to noise interference, taking into account the noise level, are used to determine abnormality.

[0141] This makes it possible to reliably monitor the sideband component that is the monitoring target for abnormality diagnosis, and reliably perform abnormality diagnosis of the electric motor 2 .

[0142] Furthermore, when applying the third embodiment, noise detection by noise detection unit 37 can also be performed based on the results of frequency analysis of the line voltage output to motor 2 or the DC voltage across filter capacitor 10C. In this case, detection does not need to be performed during normal operation of motor 2. The magnitude of the noise equivalent to that during normal operation at the noise frequency fnβ of current i can be calculated based on the frequency analysis results of the detected voltage. The calculation results do not need to be stored in storage unit 38, and storage unit 38 can also be omitted.

[0143] Furthermore, the accuracy of abnormality determination may be improved by using both the frequency analysis result of the detected voltage and the noise detection result obtained based on the analysis result 32 a of the current i when the motor 2 is operating normally.

[0144] Furthermore, in the fourth embodiment, frequency setting unit 33A acquires modulation wave frequency f0 and AC power supply frequency fac and sets the noise frequency fnβ. However, the noise frequency fnα described in the first embodiment can also be set simultaneously. In this case, frequency setting unit 33A acquires modulation wave frequency f0, carrier frequency fc, sampling frequency fs, and AC power supply frequency fac, calculates and sets noise frequencies fnα and fnβ. This can widely suppress the influence of noise components, prevent misdiagnosis, and enable more reliable abnormality diagnosis of motor 2.

[0145] Furthermore, although the abnormality diagnosis device 30, 30A to 30C in the above-mentioned embodiments 1 to 4 is shown as being located outside the power conversion device 100, it can also be located within the control device 20 of the power conversion device 100, and the same effect can be obtained. In addition, the transmission and reception of information required for setting the noise frequencies fnα and fnβ becomes simple.

[0146] Implementation method 5.

[0147] Figure 13 This is a diagram showing the configuration of a power conversion device 100A according to the fifth embodiment.

[0148] like Figure 13 As shown, power conversion device 100A is configured similarly to Embodiment 1, and includes power conversion unit 10 and control device 20A for output control of power conversion unit 10. Control device 20A includes inverter control unit 21 for output control of power conversion unit 10 and abnormality diagnosis device 30D.

[0149] Furthermore, the current i flowing from the power conversion unit 10 to the motor 2 is detected by the current sensor 3 , and the abnormality diagnosis device 30 diagnoses abnormality of the motor 2 based on the current i.

[0150] In control device 20A, inverter control unit 21 generates gate signals G for each switching element Q of inverter unit 10B through PWM control, and controls the switching elements Q to be turned on and off, thereby outputting desired power from power conversion unit 10 to motor 2. Power conversion device 100A thus drives motor 2.

[0151] Abnormality diagnosis device 30D obtains the frequencies of the modulation wave (fundamental wave), carrier wave, and clock signal (CLK) used for sampling in PWM control by inverter control unit 21, namely, modulation wave frequency f0, carrier wave frequency fc, and sampling frequency fs. Abnormality diagnosis device 30D then performs frequency analysis on current i flowing from power conversion unit 10 through motor 2 to diagnose abnormalities in motor 2.

[0152] Abnormality diagnosis device 30D, similar to abnormality diagnosis device 30 shown in Embodiment 1, includes a detection unit 31, an analysis unit 32, a frequency setting unit 33, and a determination unit 34. Detection unit 31, analysis unit 32, and frequency setting unit 33 operate similarly to Embodiment 1. Determination unit 34, similar to Embodiment 1, estimates the presence or absence of noise interference in the spectrum peaks of the sideband components of the modulated wave based on the frequencies of the sideband components and the noise frequency fnα.

[0153] When there is no noise interference, the determination unit 34 performs abnormality diagnosis of the motor 2 based on the spectrum peak of the sideband component, similarly to the first embodiment. However, when there is noise interference, the determination unit 34 interrupts the abnormality diagnosis and sends a notification signal SS1 to the inverter control unit 21 .

[0154] When receiving the notification signal SS1 notifying the abnormality diagnosis interruption from the abnormality diagnosis device 30D, the inverter control unit 21 changes the carrier frequency fc and uses the changed carrier frequency fc to control the output of the power conversion unit 10 through PWM control, thereby driving the motor 2. The carrier frequency fc can be easily changed without directly affecting the output of the power conversion unit 10.

[0155] In abnormality diagnosis device 30D, each component resumes operation to continue abnormality diagnosis. Changing carrier frequency fc changes noise frequency fnα, and therefore the presence or absence of noise interference at the sideband component's spectral peak also changes. This allows determination unit 34 to derive an estimate of the absence of noise interference and perform abnormality diagnosis of motor 2 based on the sideband component's spectral peak.

[0156] Regarding the change of the carrier frequency fc, it is preferable to derive an estimate of the absence of noise interference in the determination unit 34 by changing the carrier frequency fc once, but it may be changed a plurality of times.

[0157] As described above, in the power conversion device 100A according to this embodiment, the abnormality diagnosis device 30D within the control device 20A estimates the presence of noise interference at the peak of the sideband component spectrum based on the sideband component frequency and the noise frequency fnα during abnormality diagnosis. If noise interference is estimated, the carrier frequency fc is changed. This changes the greatest common divisor (GCD) of the modulation wave frequency f0, the carrier frequency fc, and the sampling frequency fs, and also changes the frequency of the noise component caused by PWM control. This eliminates the noise interference at the peak of the sideband component spectrum, allowing for reliable abnormality diagnosis. This prevents misdiagnosis caused by the influence of noise components, allowing for highly reliable abnormality diagnosis of the motor 2.

[0158] In the fifth embodiment, the carrier frequency fc is changed. However, at least one of the frequencies used to calculate the greatest common divisor GCD among the modulation wave frequency f0, the carrier frequency fc, and the sampling frequency fs may be changed.

[0159] In addition, when changing the carrier frequency fc, as Figure 14As shown, the carrier frequency fc can be varied over time. In this case, the carrier frequency Cr alternately and repeatedly varies at two frequencies (1 / t1) and (1 / t2) based on two different periods t1 and t2. This is not limited to variations within a single period; three or more frequencies can be varied over time. Furthermore, the frequency can be varied discretely and non-continuously.

[0160] When the carrier frequency fc or sampling frequency fs varies over time as described above, the spectra of the greatest common denominator (GCD) and its integer multiple frequency components are dispersed across multiple frequency domains. This reduces the spectral peaks of noise components and eliminates or suppresses noise interference in the spectral peaks of sideband components, enabling highly reliable abnormality diagnosis.

[0161] Implementation method 6.

[0162] In the fifth embodiment, when abnormality diagnosis is performed by the abnormality diagnosis device 30D, if noise interference is estimated at the spectral peak of the sideband component, at least one of the frequencies used in calculating the greatest common divisor GCD is changed.

[0163] In this embodiment, in the case of the above-mentioned embodiment 5, at least one of the frequencies used in the calculation of the greatest common divisor GCD is further changed so that the greatest common divisor GCD is consistent with the modulation wave frequency f0, or the greatest common divisor GCD is below 10 Hz, preferably below several Hz.

[0164] Figure 15 It is a schematic diagram of a spectrum waveform of current for explaining the effects according to the sixth embodiment. Figure 15 The figure shows noise components in two cases: a case where the greatest common divisor GCD of the modulation wave frequency f0, the carrier frequency fc, and the sampling frequency fs is several Hz; and a case where the greatest common divisor GCD exceeds 10 Hz as a comparative example.

[0165] like Figure 15 As shown, in addition to spectrum 40 of modulation wave frequency f0, spectrums 42A and 42B of noise components caused by the switching operation of inverter unit 10B appear. Spectrum 42A represents a comparative example in which the greatest common divisor GCD exceeds 10 Hz, while spectrum 42B represents a comparative example in which the greatest common divisor GCD is several Hz. Spectrum 42B has a higher number of occurrences than spectrum 42A, but has a lower spectral peak.

[0166] By reducing the GCD to a few Hz, the frequency spectrum of the noise component increases, but the spectrum is dispersed across multiple frequency domains, reducing the spectral peaks. This eliminates or suppresses noise interference at the spectral peaks of the modulated wave's sideband components, enabling highly reliable abnormality diagnosis.

[0167] In addition, in the above-mentioned embodiment 6, when at least one of the frequencies used in the calculation of the greatest common divisor GCD is changed so that the greatest common divisor GCD is consistent with the modulation wave frequency f0, since the noise component assumed by the abnormality diagnosis device 30D is eliminated, abnormality diagnosis can be reliably performed with high reliability.

[0168] Implementation method 7.

[0169] This embodiment shows an example in which the abnormality diagnosis device 30C shown in Embodiment 4 is applied to the abnormality diagnosis device 30D in the power conversion device 100A shown in Embodiment 5. In this case, the abnormality diagnosis device 30C is provided in the control device 20A of the power conversion device 100A.

[0170] Abnormality diagnosis device 30C includes detection unit 31 , analysis unit 32 , frequency setting unit 33A, and determination unit 34 , similarly to the fourth embodiment. Detection unit 31 , analysis unit 32 , and frequency setting unit 33A operate similarly to the fourth embodiment.

[0171] Similar to the fourth embodiment, the determination unit 34 estimates whether or not there is noise interference in the spectrum peak of the sideband component of the modulated wave based on the frequency of the sideband component and the noise frequency fnβ.

[0172] When there is no noise interference, the determination unit 34 performs abnormality diagnosis of the motor 2 based on the spectrum peak of the sideband component, similarly to the fourth embodiment. However, when there is noise interference, the determination unit 34 interrupts the abnormality diagnosis and sends a notification signal SS1 to the inverter control unit 21.

[0173] When receiving the notification signal SS1 notifying the abnormality diagnosis interruption from the abnormality diagnosis device 30C, the inverter control unit 21 changes the modulation wave frequency f0 and controls the output of the power conversion unit 10 by PWM control using the changed modulation wave frequency f0 to drive the motor 2.

[0174] In abnormality diagnosis device 30C, each component resumes operation to continue abnormality diagnosis. Changing carrier frequency fc changes noise frequency fnβ, and therefore the presence or absence of noise interference at the sideband component's spectral peak also changes. This allows determination unit 34 to derive an estimate of the absence of noise interference and perform abnormality diagnosis of motor 2 based on the sideband component's spectral peak.

[0175] As described above, in the power conversion device 100A according to the present embodiment, when the abnormality diagnosing device 30C in the control device 20A performs abnormality diagnosis, it is determined whether or not there is noise interference at the frequency spectrum peak of the sideband component based on the frequency of the sideband component and the noise frequency fnβ, and if it is determined that there is noise interference, the modulation wave frequency f0 is changed.

[0176] Thus, the frequency of the noise component itself changes due to the variation in the voltage corresponding to the AC power supply frequency fac (the DC voltage of the filter capacitor 10C and the AC voltage output to the motor 2). Therefore, it is possible to eliminate noise interference at the frequency spectrum peak of the sideband component, and thus reliably perform abnormality diagnosis. Thus, it is possible to prevent erroneous diagnosis due to the influence of the noise component, and thus perform abnormality diagnosis of the motor 2 with high reliability.

[0177] In addition, in Embodiments 5 to 7 described above, the frequency related to the noise frequency is changed in the case where it is determined that there is noise interference, but it is also possible to remove or suppress the assumed noise interference from the beginning and operate the power conversion device 100A.

[0178] In this case, the modulation wave frequency f0, the carrier wave frequency fc, and the sampling frequency fs are determined so that the difference between the frequency of the sideband component that is the monitoring target and the assumed noise frequency is above a set value, and thus the power conversion device 100A is operated. Alternatively, the modulation wave frequency f0, the carrier wave frequency fc, and the sampling frequency fs are determined so that the greatest common divisor GCD is reduced to several Hz, and thus the power conversion device 100A is operated.

[0179] Although various exemplary embodiments and examples are described in the present application, the various features, modes, and functions described in one or more embodiments are not limited to the application of the specific embodiments, and can be applied to the embodiments individually or in various combinations.

[0180] Therefore, innumerable modifications not exemplified can be assumed within the technical scope disclosed in the present application. For example, cases where at least one constituent element is modified, added, or omitted, and cases where at least one constituent element is extracted and combined with the constituent elements of other embodiments are assumed.

[0181] Explanation of Reference Numerals

[0182] 1 AC power supply, 2 electric motor, 10 power conversion unit, 10A converter unit, 10B inverter unit, 10C filter capacitor, 20, 20A control device, 30, 30A to 30D abnormality diagnosis device, 31 detection unit, 32 analysis unit, 32a analysis result, 33, 33A frequency setting unit, 34 determination unit, 35 notification unit, 36 determination unit, 37 noise detection unit, 38 storage unit, 100, 100A power conversion device, f0 modulation wave frequency, fac AC power supply frequency, fc carrier frequency, fs sampling frequency, fnα and fnβ noise frequencies, M modulation wave.

Claims

1. An abnormality diagnosis device for diagnosing abnormalities of a motor driven by pulse width modulation control of a power conversion device, characterized in that: include: a detection unit configured to detect a current flowing through the motor; an analyzing unit that performs frequency analysis on the current detected by the detecting unit and outputs an analysis result; a determination unit that determines an abnormality of the motor based on a spectrum peak of at least one sideband component of the modulated wave obtained from the analysis result; as well as a frequency setting unit for presetting a noise frequency in the current; The frequency setting unit calculates the greatest common divisor of two or more frequencies including the modulation wave frequency among three frequencies used for the pulse width modulation control, namely, the modulation wave frequency, the carrier frequency, and the sampling frequency for sampling the modulation wave, and sets the greatest common divisor and its integer multiples as the noise frequency. The determination unit estimates the presence or absence of noise interference at the spectrum peak of the sideband component based on the frequency of the sideband component and the set noise frequency, thereby determining an abnormality in the motor.

2. An abnormality diagnosis device for diagnosing abnormalities of a motor driven by pulse width modulation control of a power conversion device, characterized in that: include: a detection unit configured to detect a current flowing through the motor; an analyzing unit that performs frequency analysis on the current detected by the detecting unit and outputs an analysis result; a determination unit that determines an abnormality of the motor based on a spectrum peak of at least one sideband component of the modulated wave obtained from the analysis result; as well as a frequency setting unit for presetting a noise frequency in the current; The frequency setting unit sets the noise frequency as an absolute value of a value obtained by deviating an integral multiple of the modulation wave frequency used for the pulse width modulation control from an integral multiple of the frequency of the AC power source connected to the power conversion device. The determination unit estimates the presence or absence of noise interference at the spectrum peak of the sideband component based on the frequency of the sideband component and the set noise frequency, thereby determining an abnormality in the motor.

3. The abnormality diagnosis device according to claim 1 or 2, wherein: The noise frequency set by the frequency setting unit includes a frequency lower than 1 / 2 of a carrier frequency.

4. The abnormality diagnosis device according to any one of claims 1 to 3, wherein: When the difference between the frequency of the sideband component and the noise frequency is smaller than a set value and the two are close to each other, the determination unit estimates that the noise interference exists with respect to the sideband component.

5. The abnormality diagnosis device according to claim 4, wherein: The determination unit removes the sideband component estimated to be the noise interference and determines the abnormality of the motor.

6. The abnormality diagnosis device according to claim 4, wherein: The present invention comprises a noise detection unit for detecting the noise level of the current at the noise frequency when the motor is operating normally; and a storage unit for storing the detection results of the noise detection unit. The determination unit determines the abnormality of the motor based on the spectrum peak of the sideband component and the detection result in the storage unit for the sideband component estimated to have the noise interference.

7. The abnormality diagnosis device according to any one of claims 1 to 6, wherein: A notification unit is included for notifying the outside whether the noise interference exists.

8. A power conversion device, characterized in that: include: a power conversion unit for converting direct current into alternating current and supplying power to the electric motor; as well as A control device for outputting control of the power conversion unit by the pulse width modulation control, The control device includes the abnormality diagnosis device according to any one of claims 1 to 7 to diagnose abnormality of the electric motor.

9. A power conversion device, characterized in that: include: a power conversion unit for converting direct current into alternating current and supplying power to the electric motor; as well as A control device for outputting control of the power conversion unit by the pulse width modulation control, The control device includes the abnormality diagnosis device according to claim 1 to diagnose abnormality of the motor, When the difference between the frequency of the sideband component and the noise frequency is smaller than a set value and the two frequencies are close to each other, the control device changes at least one of the two or more frequencies used to calculate the greatest common divisor to perform the pulse width modulation control. The abnormality diagnosis device diagnoses abnormality of the electric motor based on the changed frequency.

10. The power conversion device according to claim 9, wherein: The frequency change refers to changing the frequency over time.

11. The power conversion device according to claim 9, wherein: When the difference between the frequency of the sideband component and the noise frequency is less than a set value and the two are close to each other, the control device changes at least one of the two or more frequencies so that the greatest common divisor is consistent with the modulation wave frequency or becomes below 10 Hz.

12. The power conversion device according to any one of claims 9 to 11, characterized in that: The frequency change refers to changing the carrier frequency to the frequency.

13. A power conversion device, characterized in that: include: a power conversion unit including a converter unit for converting AC power from an AC power source into DC power, a filter capacitor, and an inverter unit for converting the DC power of the filter capacitor into AC power and supplying the power to the electric motor; and a control device for outputting control of the power conversion unit by the pulse width modulation control, The control device includes the abnormality diagnosis device as claimed in claim 2 to diagnose abnormality of the electric motor.

14. The power conversion device according to claim 13, wherein: When the difference between the frequency of the sideband component and the noise frequency is smaller than a set value and the two are close to each other, the control device changes the modulation wave frequency to perform the pulse width modulation control. The abnormality diagnosis device diagnoses abnormality of the electric motor based on the changed modulation wave frequency.

15. A method for diagnosing abnormality of a motor driven by pulse width modulation control of a power conversion device, characterized in that: include: The first step comprises calculating a greatest common divisor of two or more frequencies including the modulation wave frequency among three frequencies used for the pulse width modulation control, namely, the modulation wave frequency, the carrier frequency, and the sampling frequency for sampling the modulation wave, and setting a frequency that is an integer multiple of the greatest common divisor as a noise frequency. a second step of detecting the current flowing through the motor and performing frequency analysis; as well as a third step of determining abnormality of the motor based on a spectrum peak of a sideband component of the modulation wave obtained from the analysis result in the second step, In the third step, based on the frequency of the sideband component and the noise frequency set in the first step, it is estimated whether there is noise interference in the spectrum peak of the sideband component.

16. A method for diagnosing abnormality of a motor driven by pulse width modulation control of a power conversion device, characterized in that: include: A first step of setting a frequency obtained by deviating from a modulation wave frequency used for the pulse width modulation control by an integral multiple of a frequency of an AC power source connected to the power conversion device as a noise frequency; a second step of detecting the current flowing through the motor and performing frequency analysis; as well as a third step of determining abnormality of the motor based on a spectrum peak of a sideband component of the modulation wave obtained from the analysis result in the second step, In the third step, based on the frequency of the sideband component and the noise frequency set in the first step, it is estimated whether there is noise interference in the spectrum peak of the sideband component.

Images