Electric motor diagnostic device, electric motor diagnostic method, and electric motor abnormality prediction inference device

Abstract

A device (100) for diagnosing an electric motor that is driven by an inverter (17), the device (100) comprising an electric current detection circuit (7) for detecting an electric current of the electric motor (5), and a computation processing unit (10) for receiving an output of the electric current detection circuit (7) and determining a winding short-circuit abnormality in the electric motor (5). The computation processing unit (10) has an operation state determination unit (41) for determining an operation state of the electric motor (5), an initial reverse-phase electric current analysis unit (31) for analyzing an initial reverse-phase electric current during a normal state, and an inverter-driving frequency calculation unit (22). The computation processing unit (10) is configured to determine a winding short-circuit in the electric motor in accordance with the inverter-driving frequency by using, as an evaluation value, the difference of the reverse-phase electric current calculated from the electric current of the electric motor during operation from the initial reverse-phase electric current, and therefore a short-circuit malfunction of stator windings of the electric motor can be detected with high accuracy.

Description

[Technical Field]

[0001] The present application relates to a diagnostic device for an electric motor, a diagnostic method for an electric motor, and an abnormality sign inference device for an electric motor. [Background technology]

[0002] Plants are equipped with numerous electric motors, and the maintenance department diagnoses the equipment using the five senses. Highly important motors require regular diagnosis, which increases costs. Furthermore, once motor degradation begins, it accelerates. In the case of AC motors, gaps and damaged areas in the insulation caused by mechanical stress and thermal degradation can induce layer shorts (interlayer shorts) through electrical discharges, leading to sudden insulation breakdown. Therefore, once a motor begins to deteriorate, it only continues to deteriorate.

[0003] As a result, interest in continuous motor monitoring technology is growing. However, most continuous motor monitoring methods require the installation of various sensors and other measuring instruments on each motor. Examples of measuring instruments include torque meters, encoders, and acceleration sensors. However, applying such instruments to motor control centers that centrally manage hundreds to thousands of motors would require a large number of wiring, making such applications impractical. Therefore, a device is needed that can easily diagnose the condition of motors from current and voltage information measured at motor control centers without using special sensors, thereby improving reliability, productivity, and safety.

[0004] In response to this, the applicant has proposed a diagnosis device for a motor that includes a current detection circuit that detects the current of the motor and a processing unit that receives the output of the current detection circuit and judges whether the motor has a winding short circuit (see Patent Document 1). The processing unit includes an operating state judgment unit that judges the operating state of the motor, an initial negative-phase-sequence current analysis unit that analyzes the initial negative-phase-sequence current in a normal state, and a unit that judges whether the motor has a winding short circuit by using the difference between the initial negative-phase-sequence current and the negative-phase-sequence current calculated from the current of the motor during operation as an evaluation value. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] International Publication No. 2019 / 202651 Summary of the Invention [Problem to be solved by the invention]

[0006] The motor diagnostic device disclosed in the above-mentioned Patent Document 1 is a method for detecting winding short circuits in motors driven by commercial power. Recently, the number of motors driven by inverters has increased, so there is a demand for detecting winding short circuits in motors driven by inverters. However, the motor diagnostic device in Patent Document 1 does not take into account the characteristics of inverter drive, in which the inverter drive frequency is changed to drive the motor.

[0007] The present application discloses technology for solving the above-mentioned problems, and aims to provide a motor diagnostic device, a motor diagnostic method, and a motor abnormality prediction inference device that can detect winding short circuits in a motor driven by an inverter. [Means for solving the problem]

[0008] The diagnostic device for an electric motor disclosed in the present application is a diagnostic device for an electric motor including a current detection circuit that detects a current of an electric motor driven by an inverter, and an arithmetic processing unit that receives an output of the current detection circuit and determines a winding short-circuit abnormality of the electric motor, wherein the arithmetic processing unit includes an operating state determination unit that calculates an effective value from the current of the electric motor to determine an operating state, an inverter driving frequency calculation unit that calculates an inverter driving frequency, an initial negative-phase-sequence current analysis unit that analyzes an initial negative-phase-sequence current in a normal state in accordance with the calculated inverter driving frequency, an evaluation value analysis unit that calculates an evaluation value for a winding short-circuit based on the difference between the negative-phase-sequence current calculated from the current of the electric motor during operation and the initial negative-phase-sequence current corresponding to the inverter driving frequency calculated during operation, and a winding short-circuit determination unit that determines a winding short-circuit of the electric motor by comparing the calculated evaluation value with a preset evaluation threshold. The winding short circuit determination unit changes the evaluation threshold value in accordance with the inverter drive frequency calculated during operation. It is something.

[0009] A diagnostic method for an electric motor disclosed in the present application includes the steps of: detecting a current of an electric motor driven by an inverter; calculating an effective value from the current of the electric motor to determine an operating state of the electric motor; calculating an inverter drive frequency for driving the electric motor; analyzing an initial negative-phase-sequence current in a normal state according to the calculated inverter drive frequency; calculating an evaluation value for a winding short circuit based on the difference between the negative-phase-sequence current calculated from the current of the electric motor during operation and the initial negative-phase-sequence current corresponding to the inverter drive frequency calculated during operation; and determining a winding short circuit of the electric motor by comparing the calculated evaluation value with a preset evaluation threshold. In the step of determining whether a winding short circuit has occurred, the evaluation threshold value is changed in accordance with the inverter drive frequency during operation.

[0010] The electric motor abnormality prediction inference device disclosed in the present application is used together with the above-mentioned electric motor diagnostic device, and includes a learning device having a data acquisition unit that acquires learning data from the electric motor diagnostic device, including the evaluation value and the winding short circuit judgment result corresponding to the evaluation value, and a model generation unit that uses the learning data to generate a trained model for inferring an abnormality prediction inference result of the electric motor from the evaluation value data of the electric motor diagnostic device, and an inference device having an inference unit that uses the trained model to output an abnormality prediction inference result of the electric motor from the evaluation value data of the electric motor diagnostic device. [Effects of the Invention]

[0011] According to the present invention, a winding short circuit is determined using the difference between the negative-phase sequence current calculated from the current during operation of the inverter-driven motor and the initial negative-phase sequence current as an evaluation value, so there is no need to calculate negative-phase sequence admittance. Furthermore, the inverter drive frequency is taken into consideration when determining whether a winding short circuit has occurred, so short-circuit faults in the stator winding of the motor can be detected with high accuracy. Furthermore, according to the electric motor abnormality prediction inference device disclosed in the present application, learning data is generated from an electric motor diagnostic device and a learned model is generated, and by using this learned model, it is possible to easily infer abnormality predictions for an electric motor. [Brief explanation of the drawings]

[0012] [Figure 1] 1 is a circuit configuration diagram showing an electric motor diagnostic device according to a first embodiment. [Figure 2] 1 is a hardware configuration diagram of a diagnostic device for an electric motor according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram of a winding short circuit. [Figure 4] 2 is a configuration diagram of a calculation processing unit of the electric motor diagnostic device according to the first embodiment. FIG. [Figure 5] FIG. 10 is a diagram showing the relationship between the inverter drive frequency and the initial negative-phase current. [Figure 6]FIG. 6A is a diagram illustrating an example in which an abnormality in the motor cannot be detected when the inverter drive frequency changes, and FIG. 6B is a diagram illustrating an example in which an abnormality in the motor can be detected when the inverter drive frequency changes. [Figure 7] 5 is a flowchart for analyzing an initial negative-phase sequence current using the motor diagnostic device according to the first embodiment. [Figure 8] 4 is a flowchart for determining whether a winding is short-circuited using the motor diagnostic device according to the first embodiment. [Figure 9] FIG. 10 is a circuit configuration diagram showing a diagnostic device for an electric motor according to a second embodiment. [Figure 10] FIG. 10 is a configuration diagram of a calculation processing unit of an electric motor diagnostic device according to a second embodiment. [Figure 11] FIG. 10 is a diagram showing the relationship between the inverter drive frequency and the initial negative-phase sequence current at each voltage unbalance rate. [Figure 12] FIG. 10 is a diagram showing the relationship between the effective voltage value and the initial negative-phase sequence current at each voltage unbalance rate. [Figure 13] 10 is a flowchart for analyzing an initial negative-phase sequence current using the motor diagnostic device according to the second embodiment. [Figure 14] 10 is a flowchart for determining whether a winding is short-circuited using the motor diagnostic device according to the second embodiment. [Figure 15] FIG. 10 is a circuit configuration diagram showing a diagnostic device for an electric motor according to a third embodiment. [Figure 16] FIG. 10 is a circuit configuration diagram showing another motor diagnostic device according to the third embodiment. [Figure 17] FIG. 10 is a circuit configuration diagram showing a diagnostic device for an electric motor according to a fourth embodiment. [Figure 18] FIG. 10 is a circuit configuration diagram showing another motor diagnostic device according to the fourth embodiment. [Figure 19] FIG. 10 is a diagram showing the configuration of an abnormality sign inference device for an electric motor according to a fifth embodiment. [Figure 20] FIG. 11 is a diagram showing the configuration of a learning device of an abnormality sign inference device for an electric motor according to a fifth embodiment. [Figure 21] 21 is a flowchart showing a learning process using the learning device of FIG. 20. [Figure 22] FIG. 10 is a diagram showing the configuration of an inference device in an abnormality sign inference device for an electric motor according to a fifth embodiment. [Figure 23] 23 is a flowchart for performing abnormality sign inference for an electric motor using the inference device of FIG. 22. [Figure 24] FIG. 10 is a hardware configuration diagram of an abnormality sign inference device for an electric motor according to a fifth embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0013] The present embodiment will be described below with reference to the drawings. In the drawings, the same reference numerals indicate the same or corresponding parts.

[0014] Embodiment 1 The motor diagnostic device according to the first embodiment will be described below with reference to FIGS. Figure 1 is a circuit diagram showing a motor diagnostic device according to embodiment 1, which is primarily used in a control center that is a closed switchboard. In the diagram, a main circuit 1 of the power supply drawn in from the power system is equipped with a molded case circuit breaker 2, an electromagnetic contactor 3, a drive control device 17, and an instrument transformer 4 that detects the load current of the main circuit 1. Furthermore, an electric motor 5, which is a load, is connected, and this electric motor 5 drives mechanical equipment 6.

[0015] The drive control device 17 converts the voltage of the main power supply circuit 1 and supplies power to drive the electric motor 5. The drive control device 17 is configured as a so-called inverter, and includes a circuit that converts the voltage and a control unit that controls this circuit.

[0016] The motor diagnostic device 100 includes a current detection circuit 7 connected to an instrument transformer 4, an arithmetic processing unit 10, a memory unit 11, a setting circuit 12, a display unit 13, a drive circuit 14, an external output unit 15, and a communication circuit 16.

[0017] The current detection circuit 7 converts the load current of the main circuit 1 detected by the instrument transformer 4 into a predetermined signal such as a phase current of the motor 5 to detect the current of the motor 5, and outputs the signal to the calculation processing unit 10 and the memory unit 11. In other words, the current of the motor 5 is detected from the current flowing in the main circuit 1 of the power supply connected to the motor 5.

[0018] The output of the current detection circuit 7 is input to the calculation processing unit 10, which calculates the negative-phase current and the like by current analysis of the motor, and determines whether a winding short circuit has occurred while the motor is in operation.

[0019] The storage unit 11 is connected to the arithmetic processing unit 10 and the setting circuit 12, and exchanges data with the arithmetic processing unit 10. The setting circuit 12 connected to the storage unit 11 has a set key, and by pressing this set key (for example, by pressing and holding it), data in the initial normal state is stored and held in the storage unit 11. Data can also be stored until the set key is released. The display unit 13 is connected to the arithmetic processing unit 10, and displays detected physical quantities such as load current and, when the arithmetic processing unit 10 detects an abnormality in the electric motor 5, an abnormal state, a warning, or the like. The drive circuit 14 is connected to the arithmetic processing unit 10, and outputs a control signal for opening and closing the electromagnetic contactor 3 based on the results of calculations performed by the arithmetic processing unit 10 on the basis of the current signal detected by the instrument transformer 4. The external output unit 15 outputs signals such as abnormal states and warnings from the arithmetic processing unit 10 to the outside.

[0020] The external monitoring device 200 is comprised of a PC (personal computer) or the like, and is connected to one or more electric motor diagnostic devices 100, appropriately receiving information from the arithmetic processing unit 10 via the communication circuit 16 and monitoring the operating status of the electric motor diagnostic devices 100. The connection between this external monitoring device 200 and the communication circuit 16 of the electric motor diagnostic devices 100 may be by cable or wirelessly. A network may be configured between the plurality of electric motor diagnostic devices 100, and the connection may be via the Internet.

[0021] The electric motor diagnostic device 100 includes a processor 1001 and a storage device 1002, as shown in FIG. 2, which is an example of hardware. The storage device 1002 includes a volatile storage device such as a random access memory (RAM) and a non-volatile auxiliary storage device such as a flash memory (for example, an electrically erasable programmable read-only memory (EEPROM)). A hard disk auxiliary storage device may be used instead of the flash memory. The processor 1001 executes a program input from the storage device 1002. In this case, the program is input to the processor 1001 from the auxiliary storage device via the volatile storage device. The processor 1001 may output data such as calculation results to the volatile storage device of the storage device 1002, or may store the data in the auxiliary storage device via the volatile storage device.

[0022] By executing a program, the processor 1001 realizes the functions of the calculation processing unit 10, memory unit 11, setting circuit 12, display unit 13, drive circuit 14, external output unit 15 and communication circuit 16 of the electric motor diagnostic device 100. The processor 1001 includes a CPU (Central Processing Unit) as an arithmetic processing device. The arithmetic processing device may include various logic circuits such as an ASIC (Application Specific Integrated Circuit), an IC (Integrated Circuit), and a DSP (Digital Signal Processor), and various signal processing circuits.

[0023] The drive control device 17 also has the hardware configuration shown in FIG. 2 and operates as an inverter.

[0024] 3 is a conceptual diagram of a winding short circuit diagnosed by the present disclosure. In the figure, when a short circuit occurs in phase a and a short-circuit current If flows, the number of short-circuited turns is Nf. The short-circuit ratio μ as a ratio to the total number of turns N is μ=Nf / N Furthermore, both a same-layer short circuit and an inter-layer short circuit may occur in the stator winding of the electric motor 5, and in this embodiment, both types of short circuits are diagnosed as a winding short circuit. In the present disclosure, it is not necessary to input the rating information of the electric motor 5 before starting operation.

[0025] 4 is a configuration diagram showing an overview of the arithmetic processing unit 10 in the motor diagnostic device according to embodiment 1. The arithmetic processing unit 10 includes a current conversion unit 20, an initial analysis unit 30, a determination unit 40, an analysis unit 50, and an abnormality determination unit 60.

[0026] The current conversion unit 20 includes an effective value calculation unit 21, an inverter drive frequency calculation unit 22, and a negative-phase current calculation unit 23, and calculates the negative-phase current Isn from the three-phase current detected by the current detection circuit 7 using symmetric coordinate transformation processing according to the following equation (1).

[0027]

number

[0028] The initial analysis unit 30 includes an initial negative-sequence current analysis unit 31. The initial negative-sequence current analysis unit 31 analyzes the negative-sequence current value Isn under normal conditions (initial state of the motor or a state where no fault has occurred) before determining whether a winding is short-circuited. For example, the negative-sequence current values for one month are calculated, and the average value is set as the initial negative-sequence current value Isn0. At the same time, a standard deviation σ may be calculated to check for variations.

[0029] At this time, the initial negative-phase-sequence current value Isn0 is stored in the storage unit 11 together with the inverter driving frequency calculated by the inverter driving frequency calculation unit 22. The inverter driving frequency is determined by receiving a signal from inside the inverter, that is, calculated by the inverter driving frequency calculation unit 22. Alternatively, the inverter driving frequency calculation unit 22 calculates the inverter driving frequency from the current waveform. 。The specific method for calculating the inverter drive frequency from the current waveform is as follows: The AC waveform of the current detected by the current detection circuit 7 is sampled. To measure the time between zero-crossing points of the AC waveform over multiple cycles, the number of samples is counted and the time between the zero-crossing points of the AC waveform is calculated. The frequency is then calculated from the calculated time. Here, the reason for measuring over multiple cycles is that, although the sampling interval can be a source of error, by measuring multiple times the error can be divided by the number of times and reduced.

[0030] FIG. 5 shows the relationship between the inverter drive frequency and the initial negative-phase-sequence current. The initial negative-phase-sequence current value is stored for each range of inverter drive frequencies. That is, the inverter frequency and the initial negative-phase-sequence current are mapped and stored. This is because the negative-phase-sequence current varies depending on the inverter drive frequency. In addition to storing the values for each inverter drive frequency, it is also possible to store the values for each range of effective current values. This is because it is preferable to perform winding short-circuit detection under the same inverter drive frequency or effective current value. The inverter drive frequency range in FIG. 5 can be read as, for example, 5 Hz or higher and lower than 10 Hz.

[0031] The determination unit 40 includes an operating state determination unit 41. The operating state determination unit 41 determines the operating state of the motor based on the effective value of the current calculated by the effective value calculation unit 21. The operating state of the motor may be determined not only based on the effective value of the current but also based on the instantaneous value of the current or on / off signals of the electromagnetic contactor 3.

[0032] The analysis unit 50 includes an evaluation value analysis unit 51, which performs analysis for determination by the winding short-circuit determination unit 61 of the abnormality determination unit 60. The evaluation value analysis unit 51 calculates the evaluation value A = |Isn - Isn0|. At this time, it is preferable that the negative-phase-sequence current Isn and the initial negative-phase-sequence current Isn0 have the same inverter drive frequency. In other words, the negative-phase-sequence current Isn and the initial negative-phase-sequence current Isn0 calculated at the same inverter drive frequency are used.

[0033] Next, the calculation of the evaluation value A will be described. A winding short circuit is a short circuit between coil wires, and when a winding short circuit occurs, the three-phase stator current becomes asymmetric, so it can be detected by the negative-sequence component.If we assume that the short-circuit rate when part of the stator winding of a three-phase induction motor is short-circuited is μ (μ = Nf / N), and μ << 1, the following relationship can be derived between the positive-sequence voltage Vsp and the negative-sequence voltage Vsn, and the positive-sequence current Isp and the negative-sequence current Isn.

[0034]

number

[0035]

number

[0036]

number

[0037]

number

[0038] where Ypp: admittance of positive-positive sequence components, Ynn: admittance of negative-negative sequence components, Ypn: admittance of positive-negative sequence components, Ynp: admittance of negative-positive sequence components, Yn: negative-sequence admittance, ω: power supply angular velocity, rs: stator resistance, rr: rotor resistance, rf: short-circuit resistance, Ls: stator leakage inductance, Lr: rotor leakage inductance, Lm: excitation inductance, μ: short-circuit ratio.

[0039] The off-diagonal component Ypn of admittance Y can be used as an indicator of winding short circuit, but it is not easy to calculate the off-diagonal component Ypn in an actual device. Therefore, here we adopt a method of measuring and monitoring only the negative-sequence current Isn by analyzing normal data for the initial negative-sequence current Isn0.

[0040] When no winding short circuit occurs (μ=0), the off-diagonal element Ypn of admittance Y is zero, so Isn=Yn·Vsn=Isn0 ···(6) When a winding short occurs, Isn=Yn·Vsn+Ypn·Vsp=Isn0+Ypn·Vsp···(7) and Isn changes.

[0041] That is, by analyzing the initial negative-phase current Isn0, only Isn and Isn0 are measured, Evaluation value A = |Isn-Isn0| (8) It can be seen that the occurrence of a winding short circuit can be detected by using this as an index. When the motor is first installed, it is initialized assuming that no winding short circuit has occurred (the negative-phase sequence admittance Yn is calculated), and then the occurrence of a winding short circuit is determined by monitoring the evaluation value A of equation (8).

[0042] The abnormality determination unit 60 includes a winding short-circuit determination unit 61, which determines whether or not a winding short-circuit exists based on whether or not the evaluation value A calculated by the evaluation value analysis unit 51 exceeds a preset threshold δ1. Note that the threshold δ1 is a value that differs depending on the rating of the electric motor 5. The threshold δ1 is an evaluation threshold.

[0043] Furthermore, the threshold value δ1 is corrected depending on the inverter drive frequency. Fig. 6A is a diagram illustrating an example in which an abnormality in the electric motor cannot be detected when the threshold value δ1 is kept constant even when the inverter drive frequency changes, and Fig. 6B is a diagram illustrating an example in which an abnormality in the electric motor can be detected by changing the threshold value δ1 when the inverter drive frequency changes.

[0044] As shown in Figure 6A, if threshold value δ1 is kept constant regardless of the inverter drive frequency, a situation may arise in which a layer short cannot be detected when the inverter drive frequency changes, for example, from 60 Hz to 40 Hz. This is because the change in negative-phase current due to a layer short is small when the inverter drive frequency is low. Therefore, as shown in Figure 6B, by changing threshold value δ1a, which was set when the inverter drive frequency was 60 Hz, to threshold value δ1b when the inverter drive frequency changed to 40 Hz, it is possible to prevent missed detection of a winding short circuit.

[0045] The following equation (9) shows that the negative-phase current when a winding is short-circuited varies depending on the inverter drive frequency. fs is the inverter drive frequency calculation is the inverter drive frequency calculated in the previous section, and fb is the reference power supply frequency (commercial frequency). Even with the same short circuit rate, a lower inverter drive frequency results in a smaller change in negative-phase current. This characteristic is related to V / f control. Generally, with V / f control, the voltage value changes when the inverter drive frequency fs is below the power supply frequency fb (60 Hz or below, or 50 Hz or below). On the other hand, when the power supply frequency is above fb (60 Hz or above, or 50 Hz or above), the voltage value remains constant, so there are cases where it is not necessary to correct the threshold value δ1 within this range. It is preferable to change whether or not to perform correction depending on the inverter drive method.

[0046]

number

[0047] Since the average value Isn0av and standard deviation σ of the initial negative-phase-sequence current are calculated during the initial negative-phase-sequence current analysis, the threshold δ1 may be simply set to Isn0av+3σ or Isn0av+4σ, for example.

[0048] Next, the process of diagnosis using the motor diagnosis device 100 will be described with reference to FIGS. FIG. 7 is a flowchart for analyzing an initial negative-phase sequence current using the motor diagnostic device according to the first embodiment. The current detection circuit 7 acquires the current (current of each phase) of the motor 5 (step S11), and the effective value calculation unit 21 calculates the effective value of the current (step S12). The operating state determination unit 41 determines whether the motor 5 is operating based on the effective current value. If the motor 5 is determined to be operating (Yes in step S13), the inverter drive frequency calculation unit 22 calculates the inverter drive frequency (step S14). The negative-phase sequence current calculation unit 23 then calculates the negative-phase sequence current (step S15). It is then determined whether the number of calculations of the negative-phase sequence current exceeds a predetermined number. If the number of calculations exceeds the predetermined number (Yes in step S16), the negative-phase sequence current values for the predetermined number of calculations are averaged to obtain the initial normal negative-phase sequence current value Isn0 (step S17). At the same time, the standard deviation σ may be calculated. This initial negative-phase sequence current value Isn0 is stored in the storage unit 11. If the number of times the negative-phase-sequence current has been calculated does not reach the predetermined number (No in step S16), the current of the motor is acquired again, and steps S11 to S16 are repeated until the predetermined number of times is reached.

[0049] After calculating the initial negative-phase sequence current value Isn0, a winding short circuit diagnosis is performed while the motor is running. 8 is a flowchart for determining whether a winding is short-circuited. The current (current of each phase) of the motor 5 is obtained from the current detection circuit 7 (step S21), and the effective value calculation unit 21 calculates the effective value of the current (step S22). The operating state determination unit 41 determines whether the motor 5 is operating based on the effective value of the current. If the motor 5 is determined to be operating (Yes in step S23), the inverter driving frequency calculation unit 22 calculates the inverter driving frequency (step S24). The preset threshold value δ1 is corrected based on the calculated inverter driving frequency (step S25). Because the threshold value δ1 varies depending on the inverter driving frequency, threshold values for each inverter driving frequency may be mapped and stored in a database in advance. Next, the negative-phase sequence current calculation unit 23 calculates the negative-phase sequence current Isn (step S26).

[0050] 7 and stored in the storage unit 11 and the negative-phase-sequence current Isn calculated in step S26, the evaluation value analysis unit 51 calculates the evaluation value A of equation (8) (step S27). The winding short-circuit determination unit 61 compares the evaluation value A with the threshold value δ1 corrected in step S25, and if A≧δ1 is satisfied (Yes in step S28), it determines that a winding short has occurred and outputs this to the outside (step S29).

[0051] If A<δ1 in step S28, the process returns to step S21 to again acquire the current (current of each phase) of the electric motor 5. If the inverter drive frequency calculated in step S24 has not changed, the threshold value δ1 is not corrected in step S25, and the process proceeds to the next step S26.

[0052] As described above, according to the first embodiment, the difference from the initial value of the negative-phase-sequence current is used as evaluation value A of a winding short circuit, and evaluation value A is compared with a preset threshold value δ1 to determine whether a winding short circuit has occurred. In this case, the difference between the calculated negative-phase-sequence current and the initial negative-phase-sequence current corresponding to the inverter drive frequency during operation is used, so the initial state can be offset and there is no need to calculate negative-phase-sequence admittance, making it possible to accurately detect short-circuit faults in the stator winding of an inverter-driven motor. Furthermore, since a voltage detection circuit is not required, it is possible to provide a motor diagnostic device that has a simple configuration, reduces power consumption, and is able to accurately detect short-circuit faults in the stator winding of an inverter-driven motor.

[0053] Furthermore, if the inverter drive frequency changes during operation, a winding short circuit is determined by comparing the evaluation value A with a preset threshold value δ1 corrected according to the inverter drive frequency or the effective current value, thereby reducing the risk of missing a short circuit in the stator winding of an inverter-driven motor.

[0054] In this embodiment, it is assumed that the voltage unbalance rate is small. This is because when the voltage unbalance rate is large, fluctuations in load torque change the negative-phase current value, increasing the possibility of erroneous detection when determining whether a winding is short-circuited. In a system used in a control center, which is an enclosed switchboard as exemplified in this embodiment, the magnitude of the voltage unbalance rate can be selected in advance based on the load balance. In this embodiment, it is sufficient to target a system with a selected small voltage unbalance rate. Alternatively, the voltage unbalance rate, which will be described later, can be measured in parallel or obtained in advance to determine whether this embodiment is applicable.

[0055] Embodiment 2 The motor diagnostic device according to the second embodiment will be described below with reference to FIGS. 9 is a circuit configuration diagram showing a motor diagnostic device according to embodiment 2, which differs from embodiment 1 in that a voltage transformer 8 that detects the voltage of the main circuit 1 is provided in the main circuit 1, and a voltage detection circuit 9 connected to the voltage transformer 8 is provided in the motor diagnostic device 100. The rest of the configuration is the same as in embodiment 1.

[0056] The voltage detection circuit 9 detects the line voltage of the main circuit 1 of the power supply connected to the motor 5, converts it into a predetermined signal such as the phase voltage of the motor 5, detects the voltage of the motor 5, and outputs it to the calculation processing unit 10 and the memory unit 11. The outputs of the current detection circuit 7 and the voltage detection circuit 9 are input to the calculation processing unit 10, which calculates the negative-phase current, voltage unbalance rate, etc. by analyzing the voltage and current of the motor 5, and determines and detects winding short circuits while the motor is operating. The motor diagnostic device 100 according to the second embodiment also includes the hardware shown in FIG.

[0057] 10 is a configuration diagram showing an overview of a calculation processing unit 10 in an electric motor diagnosis device according to embodiment 2. The calculation processing unit 10 includes a current-voltage conversion unit 20a, an initial analysis unit 30, a determination unit 40a, an analysis unit 50, and an abnormality determination unit 60.

[0058] The current-voltage conversion unit 20a includes an effective value calculation unit 21, an inverter drive frequency calculation unit 22, a negative-phase current calculation unit 23, and a voltage unbalance rate calculation unit 24, and calculates the negative-phase current Isn from the three-phase current detected by the current detection circuit 7 using symmetric coordinate transformation processing according to equation (1) described in embodiment 1.

[0059] The voltage unbalance rate calculation unit 24 calculates the voltage unbalance rate Vunbal using the phase voltages of each phase or the line voltages. When calculating the voltage unbalance rate Vunbal from the line voltages, for example, the following formula is used. Vunbal = ((maximum difference between each line voltage and the average voltage) / average voltage) x 100% That is, (Vuv-Vavg) / Vavg×100% (Vvw-Vavg) / Vavg×100% (Vwu-Vavg) / Vavg×100% maximum However, the average voltage Vavg = (Vuv + Vvw + Vwu) / 3 Here, Vuv is the line voltage between the u phase and the v phase, Vvw is the line voltage between the v phase and the w phase, and Vwu is the line voltage between the w phase and the u phase.

[0060] It is preferable that the voltage unbalance rate calculation unit 24 calculates the negative-phase-sequence current only when the voltage unbalance rate Vunbal is 1% or less, and does not calculate the negative-phase-sequence current when the voltage unbalance rate exceeds 1%. This is because, when the voltage unbalance rate is large, the value of the negative-phase-sequence current changes due to fluctuations in the load torque, increasing the possibility of erroneous detection when determining whether a winding short circuit has occurred. When determining whether a winding short circuit has occurred, the accuracy of winding short-circuit detection can be improved by limiting the voltage unbalance rate to 1% or less, for example.

[0061] The initial analysis unit 30 includes an initial negative-sequence current analysis unit 31. The initial negative-sequence current analysis unit 31 analyzes the normal negative-sequence current value Isn before determining whether a winding is short-circuited. As in the first embodiment, the negative-sequence current values for one month are calculated, and the average value is set as the initial negative-sequence current value Isn0. At the same time, a standard deviation σ may be calculated to check for variations.

[0062] At this time, the initial negative-phase-sequence current value Isn0 is stored in the memory unit 11 along with the inverter driving frequency and the voltage unbalance ratio. For example, as shown in FIG. 11, the initial negative-phase-sequence current value at a certain range of inverter driving frequencies is mapped and stored for each range of voltage unbalance ratios. This is because the negative-phase-sequence current varies depending on the inverter driving frequency and voltage unbalance ratio. The effective voltage value acquired by the voltage detector, or the positive-phase-sequence voltage and the voltage unbalance ratio may also be recorded. This is because the amount of change in the negative-phase-sequence current varies depending on the effective voltage value and the positive-phase-sequence voltage. FIG. 12 shows the relationship between the initial negative-phase-sequence current value at a certain range of effective voltage values and each range of voltage unbalance ratios. This relationship is stored in the memory unit 11. In FIG. 11, the range of inverter driving frequencies is interpreted as, for example, 5 Hz or more and less than 10 Hz, and in FIG. 12, the range of effective voltages is interpreted as, for example, 100 V or more and less than 110 V.

[0063] The determination unit 40a includes an operating state determination unit 41 and a voltage imbalance determination unit 42. The operating state determination unit 41 determines the operating state of the motor based on the effective values of the current and the voltage calculated by the effective value calculation unit 21. The operating state of the motor can be determined not only based on the effective values of the current and the effective values of the voltage, but also based on the instantaneous values of the current or the voltage or on / off signals of the electromagnetic contactor 3. The voltage unbalance determination unit 42 determines whether the voltage unbalance rate Vunbal is greater than a preset threshold value δ2. As described above, the threshold value δ2 is set to, for example, 1%. The threshold value δ2 is a voltage unbalance rate threshold value.

[0064] The analysis unit 50 includes an evaluation value analysis unit 51, which performs analysis for determination in the winding short-circuit determination unit 61 of the abnormality determination unit 60. The evaluation value analysis unit 51 calculates the value of the evaluation value A = |Isn - Isn0| in equation (8). The method of calculating the evaluation value A is the same as in the first embodiment, and uses the negative-phase-sequence current Isn and the initial negative-phase-sequence current Isn0 calculated at the same inverter drive frequency. That is, by analyzing the initial negative-phase sequence current Isn0, it is possible to detect the occurrence of a winding short circuit by measuring only Isn and Isn0 and using the evaluation value A as an index. At this time, the voltage unbalance rate Vunbal can be detected with high accuracy if it is limited to, for example, 1% or less. Also, at the beginning of installation of the motor 5, it is initialized (the negative-phase sequence admittance Yn is calculated) assuming that no winding short circuit has occurred, and then the winding short circuit is judged by monitoring the evaluation value A of equation (8).

[0065] The configuration of the abnormality determination unit 60 is the same as in the first embodiment, and includes a winding short-circuit determination unit 61, which determines whether or not a winding short-circuit has occurred based on whether or not the evaluation value A calculated by the evaluation value analysis unit 51 exceeds a preset threshold value δ1. The threshold value δ1 is corrected by the inverter drive frequency, the effective voltage, or the positive-sequence voltage. The reason for the correction is that, as shown in equation (9) and as described above, the negative-sequence current when a winding is short-circuited varies depending on the inverter drive frequency, the effective voltage, and the positive-sequence voltage.

[0066] As in the first embodiment, the average value Isn0av and standard deviation σ of the initial negative-phase-sequence current are calculated during the initial negative-phase-sequence current analysis, so the threshold δ1 may simply be set to Isn0av+3σ or Isn0av+4σ, for example.

[0067] Next, the process of diagnosis using the motor diagnosis device 100 will be described with reference to FIGS. 13 is a flowchart for analyzing an initial negative-phase sequence current using the motor diagnostic device according to embodiment 2. The current (current of each phase) of motor 5 is obtained from current detection circuit 7, and the voltage (line voltage or phase voltage) of motor 5 is obtained from voltage detection circuit 9 (step S31), and effective value calculation unit 21 calculates the effective values of the current and voltage (step S32). Operating state determination unit 41 determines whether motor 5 is in an operating state based on the effective values of the current and voltage, and if it is determined to be in an operating state (Yes in step S33), voltage unbalance rate calculation unit 24 calculates the voltage unbalance rate Vunbal (step S34).

[0068] The voltage unbalance rate Vunbal is compared with a preset threshold value δ2. If the voltage unbalance rate Vunbal≦δ2 is satisfied (Yes in step S35), the inverter driving frequency calculation unit 22 calculates the inverter driving frequency (step S36). Next, the negative-phase-sequence current calculation unit 23 calculates the negative-phase-sequence current (step S37). It is determined whether the number of calculations of the negative-phase-sequence current exceeds a predetermined number. If it is determined that the number of calculations exceeds the predetermined number (Yes in step S38), the negative-phase-sequence current values for the predetermined number of calculations are averaged to set the initial negative-phase-sequence current value Isn0 (step S39). At the same time, it is advisable to calculate the standard deviation σ. This initial negative-phase-sequence current value Isn0 is stored in the storage unit 11. If the number of calculations of the negative-phase-sequence current does not reach the predetermined number of times (No in step S38), the motor current is acquired again, and steps S31 to S38 are repeated until the predetermined number of calculations is reached.

[0069] After calculating the initial negative-phase sequence current value Isn0, a diagnosis of winding short circuit is performed. 14 is a flowchart for determining whether a winding is short-circuited. The current (current of each phase) of the motor 5 is obtained from the current detection circuit 7, and the voltage (line voltage or phase voltage) of the motor 5 is obtained from the voltage detection circuit 9 (step S41). The effective value calculation unit 21 calculates the effective value of the current and the effective value of the voltage (step S42). The operating state determination unit 41 determines whether the motor 5 is operating based on the effective value of the current and the effective value of the voltage. If the motor 5 is determined to be operating (Yes in step S43), the voltage unbalance rate calculation unit 24 calculates the voltage unbalance rate Vunbal (step S44).

[0070] The voltage unbalance rate Vunbal is compared with a preset threshold value δ2, and if the voltage unbalance rate Vunbal≦δ2 is satisfied (Yes in step S45), the inverter drive frequency calculation unit 22 calculates the inverter drive frequency (step S46). The preset threshold value δ1 is corrected based on the inverter drive frequency calculated in step S46 and the effective voltage or positive-phase voltage calculated in step S42 (step S47). Next, the negative-phase sequence current calculation unit 23 calculates the negative-phase sequence current Isn (step S48).

[0071] 13 and stored in the storage unit 11 and the negative-phase-sequence current Isn calculated in step S48, the evaluation value analysis unit 51 calculates the evaluation value A of equation (8) (step S49). The winding short-circuit determination unit 61 compares the evaluation value A with the threshold value δ1 corrected in step S47, and if A≧δ1 is satisfied (Yes in step S50), it determines that a winding short has occurred and outputs this to the outside (step S51).

[0072] If A<δ1 in step S50, the process returns to step S41, where the current (current of each phase) and voltage (line voltage or phase voltage) of the electric motor 5 are acquired again. If the inverter drive frequency calculated in step S46 has not changed, the threshold value δ1 is not corrected in step S47, and the process proceeds to the next step S48. Alternatively, if the inverter drive frequency calculated in step S46 has not changed, the same threshold value δ1 is set again in step S47, and the process proceeds to the next step S48.

[0073] As described above, the second embodiment achieves the same effects as the first embodiment. Furthermore, for those in which the voltage unbalance rate Vunbal is unknown or is estimated to be large, the voltage unbalance rate Vunbal is calculated, and if the voltage unbalance rate Vunbal is greater than a preset threshold δ2, for example, 1%, it is specified that a short-circuit determination is not made for the negative-phase-sequence current, thereby reducing errors in the short-circuit determination. Needless to say, if the calculated voltage unbalance rate Vunbal is smaller than the preset threshold δ2, a winding short-circuit determination is made according to the flowcharts of FIGS. 13 and 14.

[0074] Embodiment 3 The motor diagnostic device according to the third embodiment will be described below. FIG. 15 is a circuit diagram showing a motor diagnostic device 100a according to embodiment 3. The difference from embodiment 1 is that while the configuration shown in embodiment 1 is in the form of a control center, embodiment 3 uses an instrument transformer 4 that detects the current in the main circuit 1, which is, for example, a clamp type and is configured to be attached to the main circuit as appropriate. With this configuration, it becomes possible to attach the motor diagnostic device 100a to each distribution board. In other words, it can be configured as an external device.

[0075] Also, Fig. 16 is a circuit diagram showing another motor diagnostic device 100a according to embodiment 3. Fig. 16 differs from embodiment 2 in that both the potential transformer 4 that detects the current in the main circuit 1 and the potential transformer 8 that detects the voltage are, for example, clamp-type and are configured to be attached to the main circuit as appropriate. With this configuration, the motor diagnostic device 100a can be attached to each distribution board. In other words, it can be configured as an external device.

[0076] As described above, according to the third embodiment, in addition to the effects of the first embodiment, the motor diagnostic device 100a can be configured independently and can be attached to the main circuit to which the motor is connected as needed.

[0077] Embodiment 4 The motor diagnostic device according to the fourth embodiment will be described below. 17 is a circuit diagram showing a motor diagnostic device 100b according to embodiment 4. In embodiment 1, motor diagnostic device 100 is provided independently of drive control device 17, but motor diagnostic device 100b according to embodiment 4 has a configuration in which a motor diagnostic function is incorporated into the drive control device. For example, the configuration is such that a diagnostic function is provided in the microcomputer of the drive control device, and the diagnostic device is built into the drive control device.

[0078] 18 is a circuit diagram showing another electric motor diagnostic device 100b according to the fourth embodiment. FIG. 18 shows an example in which the functions of the electric motor diagnostic device according to the second embodiment are incorporated into a drive control device. In this case as well, the diagnostic function is provided in the microcomputer of the drive control device, and the diagnostic device is built into the drive control device.

[0079] As described above, according to the fourth embodiment, in addition to the effects of the first embodiment, it is possible to configure the device integrally with the drive device, thereby enabling the device to be made smaller.

[0080] Embodiment 5. The following describes an abnormality sign diagnostic device for an electric motor according to the fifth embodiment. The abnormality sign inference device shown in the fifth embodiment may be added to the electric motor diagnosis devices 100, 100a, and 100b of the first to fourth embodiments described above. The abnormality sign inference device may be built in, built in the external monitoring device 200, or further, may be external to the monitoring device 200. Fig. 19 is a diagram showing the configuration of an electric motor abnormality sign diagnosis device according to the fifth embodiment, and is an example of a configuration added to electric motor diagnosis devices 100, 100a, and 100b. In Fig. 19, electric motor abnormality sign inference device 300 includes a learning device 310 and an inference device 320. Below, the procedure for inferring an abnormality sign of an electric motor will be explained, divided into a "learning phase" and an "utilization phase" in which inference is actually performed.

[0081] <Learning Phase> 20 is a diagram showing the configuration of the learning device 310. The learning device 310 includes a data acquisition unit 311, a model generation unit 312, and a trained model storage unit 313. FIG. 21 is a flowchart showing the processing steps for executing the learning phase using the learning device 310.

[0082] The data acquisition unit 311 acquires, as input data, time-series data b1 of the evaluation value A and the winding short circuit determination result b2 from the motor diagnostic device, and associates the combination of both to create learning data (step S101).

[0083] The model generation unit 312 learns signs of an abnormality in the motor based on the learning data output from the data acquisition unit 311 (step S102). That is, the model generation unit 312 learns a time series pattern common to abnormality determinations from time series data of multiple evaluation values A over a certain period before a winding short-circuit is determined to be an abnormality, and generates a trained model 314. For the learning used to generate the trained model 314, a method can be used in which, from time series data of multiple evaluation values A over a certain period before a winding short-circuit is determined to be an abnormality, the time series pattern is inferred by deep learning to infer a time series pattern common to abnormality determinations but not included in the time series pattern for normality determinations from the time series data of multiple evaluation values A over a certain period before a winding short-circuit is determined to be an abnormality. Also, well-known machine learning techniques such as genetic programming, functional logic programming, and support vector machines may be performed.

[0084] The model generation unit 312 generates and outputs a trained model 314 by performing the above-described learning, and the trained model storage unit 313 stores the trained model 314 output from the model generation unit 312 (step S103).

[0085] <Utilization phase> 22 is a diagram showing the configuration of the inference device 320. The inference device 320 includes a data acquisition unit 321 and an inference unit 322. FIG. 23 is a flowchart showing the processing steps for executing the utilization phase in which an abnormality sign of the motor is inferred using the inference device 320.

[0086] The data acquisition unit 321 acquires, as input data, time-series data b1 of the evaluation value A from a diagnostic device for the electric motor (step S111).

[0087] The inference unit 322 infers an abnormality sign of the motor using the trained model 314. That is, by inputting the time series data of the evaluation value A acquired by the data acquisition unit into this trained model 314 (step S112), it is possible to output an abnormality sign inference result 323 of the motor inferred from the time series data of the evaluation value A (step S113).

[0088] The motor abnormality sign inference result 323 is output from the motor diagnostic devices 100, 100a, 100b to, for example, the monitoring device 200 (step S114). For example, if it is inferred that there is a sign of an abnormality, that information is transmitted to the monitoring device 200. By obtaining the sign of an abnormality early, it is possible to perform maintenance of the electric motor 5 in a planned manner and to adjust the stop period of the machinery and equipment 6 connected to the electric motor 5. Furthermore, in step S114, the information is not limited to being transmitted to the monitoring device 200. For example, the information that there is a sign of an abnormality can be displayed on the display unit 13 via the calculation processing unit 10, a signal such as a sign of an abnormality or a warning can be output from the external output unit 15 to the outside, or the information can be reflected in the drive control device 17 so that the electric motor 5 is driven to reduce the load on the electric motor 5. In this way, by obtaining a sign of an abnormality of the electric motor 5 by inference, it is possible to take various measures before the electric motor becomes abnormal.

[0089] Furthermore, if it is inferred that no signs of motor abnormality are observed, there is no need for urgent maintenance of the motor in question, which has the effect of enabling the smooth execution of maintenance plans for many motors 5 arranged in the plant. In this way, by aggregating information on the presence or absence of signs of motor abnormality in the monitoring device 200, it becomes possible to execute or quickly change maintenance plans for many motors 5 arranged in the plant.

[0090] In this fifth embodiment, it has been described that the motor abnormality sign inference result 323 is output using the trained model 314 trained by the model generation unit 312, but it is also possible to acquire a trained model from another external source and output the abnormality sign inference result based on this trained model. In addition, although the fifth embodiment has been described as being divided into a learning phase and an utilization phase, the learning phase may be performed first, followed by the utilization phase, or both may be performed in parallel. When both are performed in parallel, some kind of learning completion threshold, such as the number of acquired data, is set, and only learning is performed while learning is incomplete, and after learning is completed, both are performed in parallel.

[0091] The motor abnormality sign inference device 300 is configured from a processor 350 and a storage device 360, as shown in Fig. 24, as an example of hardware. This configuration is similar to the hardware described in Fig. 2, so a description thereof will be omitted. The processor 350 executes a program to realize the functions of the learning device 310 and the inference device 320 of the motor abnormality sign inference device 300.

[0092] As described above, according to embodiment 5, in addition to the effects of embodiment 1, a trained model is generated using a combination of time series data b1 of evaluation value A obtained by the motor diagnostic device and the winding short circuit judgment result b2 as training data, and by inputting the time series data b1 of evaluation value A obtained from the motor diagnostic device into this trained model, signs of abnormality in the motor are inferred, making it possible to carry out sound operation and maintenance plans for the motor based on the results.

[0093] Although the present disclosure describes various exemplary embodiments and examples, the various features, aspects, and functions described in one or more embodiments are not limited to application to a particular embodiment, but may be applied to the embodiments alone or in various combinations. Therefore, countless variations not illustrated are conceivable within the scope of the technology disclosed in the present specification, including, for example, cases where at least one component is modified, added, or omitted, and cases where at least one component is extracted and combined with components of another embodiment. [Explanation of symbols]

[0094] 1: Main circuit, 2: Molded case circuit breaker, 3: Electromagnetic contactor, 4: Instrument transformer, 5: Electric motor, 6: Mechanical equipment, 7: Current detection circuit, 8: Instrument transformer, 9: Voltage detection circuit, 10: Arithmetic processing unit, 11: Memory unit, 12: Setting circuit, 13: Display unit, 14: Drive circuit, 15: External output unit, 16: Communication circuit, 17: Drive control device (inverter), 20: Current conversion unit, 20a: Current-voltage conversion unit, 21: Effective value calculation unit, 22: Inverter drive frequency calculation unit, 23: Negative-phase sequence current calculation unit, 24: Voltage unbalance rate calculation unit, 30: Initial analysis unit, 31: Initial negative-phase sequence current analysis unit, 40, 40a: Judgment unit, 41: Operation state judgment unit, 42: Voltage unbalance judgment unit, 50: Analysis unit, 51: Evaluation value analysis unit, 60: Abnormality determination unit, 61: Winding short circuit determination unit, 100, 100a, 100b: Motor diagnosis device, 200: Monitoring device, 300: Motor abnormality prediction inference device, 310: Learning device, 311: Data acquisition unit, 312: Model generation unit, 313: Learned model storage unit, 314: Learned model, 320: Inference device, 321: Data acquisition unit, 322: Inference unit, 323: Motor abnormality prediction inference result, 350, 1001: Processor, 360, 1002: Storage device.

Claims

1. a current detection circuit for detecting a current of an electric motor driven by an inverter; an arithmetic processing unit that receives the output of the current detection circuit and determines whether there is a winding short circuit abnormality in the electric motor; A diagnostic device for an electric motor comprising: The arithmetic processing unit an operating state determination unit that calculates an effective current value from the current of the motor and determines the operating state; an inverter drive frequency calculation unit that calculates an inverter drive frequency; an initial negative-phase-sequence current analysis unit that analyzes an initial negative-phase-sequence current in a normal state in accordance with the calculated inverter drive frequency; an evaluation value analysis unit that calculates an evaluation value of a winding short circuit based on the difference between a negative-phase-sequence current calculated from the current of the motor during operation and the initial negative-phase-sequence current corresponding to the inverter drive frequency calculated during operation; a winding short circuit determination unit that determines whether a winding short circuit has occurred in the electric motor by comparing the calculated evaluation value with a preset evaluation threshold value; and The winding short circuit determination unit changes the evaluation threshold value in accordance with the inverter drive frequency calculated during operation.

2. 2. The motor diagnostic device according to claim 1, wherein the inverter drive frequency calculation unit calculates the inverter drive frequency from a waveform of the current detected by the current detection circuit, or calculates the inverter drive frequency using a signal received from the inverter.

3. 3. The motor diagnostic device according to claim 1, wherein the initial negative-phase-sequence current analysis unit stores a map of the initial negative-phase-sequence current corresponding to the inverter drive frequency.

4. further comprising a voltage detection circuit for detecting a voltage of the electric motor; The arithmetic processing unit the operating state determination unit that calculates a current effective value and a voltage effective value from the current and voltage of the motor, respectively, to determine the operating state; a voltage unbalance rate determination unit that calculates a voltage unbalance rate from the voltage of the motor, compares the calculated voltage unbalance rate with a set voltage unbalance rate threshold, and determines whether to perform a winding short circuit determination; an initial negative-phase-sequence current analysis unit that analyzes an initial negative-phase-sequence current in a normal state in accordance with the inverter drive frequency calculated by the inverter drive frequency calculation unit when the voltage unbalance rate determination unit determines that a winding short circuit should be determined; an evaluation value analysis unit that calculates, when the voltage unbalance rate determination unit determines that a winding short circuit should be determined, an evaluation value of a winding short circuit based on the difference between a negative-phase sequence current calculated from the current of the motor during operation and the initial negative-phase sequence current corresponding to the inverter drive frequency; the winding short-circuit determination unit that determines a winding short-circuit in the motor by comparing the evaluation value with a preset evaluation threshold, and changes the evaluation threshold in accordance with the inverter drive frequency calculated during operation; 2. The diagnostic device for an electric motor according to claim 1, further comprising:

5. 5. The motor diagnostic device according to claim 4, wherein the inverter drive frequency calculation unit calculates the inverter drive frequency from a waveform of the current detected by the current detection circuit, or calculates the inverter drive frequency using a signal received from the inverter.

6. 6. The electric motor diagnosis device according to claim 4, wherein the initial negative-sequence current analysis unit stores at least one of a map correlating the voltage unbalance rate, the inverter drive frequency, and the initial negative-sequence current, a map correlating the voltage unbalance rate, the effective voltage value, and the initial negative-sequence current, and a map correlating the voltage unbalance rate, the positive-sequence voltage of the electric motor, and the initial negative-sequence current.

7. 7. The motor diagnostic device according to claim 1, wherein the diagnostic device is integrated with an inverter that drives the motor.

8. detecting a current of an electric motor driven by the inverter; calculating an effective value from the current of the motor to determine the operating state of the motor; calculating an inverter drive frequency for driving the electric motor; analyzing an initial negative-phase sequence current during normal operation according to the calculated inverter drive frequency; calculating an evaluation value of a winding short circuit based on a difference between a negative-phase-sequence current calculated from the current of the motor during operation and the initial negative-phase-sequence current corresponding to the inverter drive frequency calculated during operation; and determining whether a winding short circuit has occurred in the electric motor by comparing the calculated evaluation value with a preset evaluation threshold value, A diagnostic method for an electric motor, wherein in the step of determining whether a winding is short-circuited, the evaluation threshold value is changed depending on the inverter drive frequency during operation.

9. 8. An electric motor abnormality sign inference device used together with the electric motor diagnostic device according to claim 1, the learning device including: a data acquisition unit that acquires learning data from the electric motor diagnostic device, the learning data including the evaluation value and a winding short circuit determination result corresponding to the evaluation value; and a model generation unit that uses the learning data to generate a trained model for inferring an abnormality sign inference result of the electric motor from the evaluation value data of the electric motor diagnostic device; an inference device having an inference unit that uses the trained model to output an abnormality sign inference result for the electric motor from the evaluation value data of the electric motor diagnostic device; An abnormality prediction inference device for an electric motor equipped with the device.

Images