Motor control device, motor control system, and motor control method

Abstract

This electric motor control device controls an electric motor that operates by power supplied from a power source and converted using a power conversion device having an inverter. The electric motor control device is provided with: a deterioration estimation unit for calculating an estimated deterioration degree obtained by estimating the degree of deterioration of an apparatus on which the electric motor is mounted or the inverter; an operation determination unit for comparing the estimated deterioration degree with a prescribed reference deterioration degree, setting the operation mode of the electric motor to normal operation if the estimated deterioration degree is less than the reference deterioration degree, and setting the operation mode to low-noise pulse operation if the estimated deterioration degree is equal to or greater than the reference deterioration degree; and, a control unit for controlling the inverter on the basis of the operation mode. The low-noise pulse operation reduces switching loss more than the normal operation does.

Description

[Technical field]

[0001] The present disclosure relates to an electric motor control device that controls an electric motor, an electric motor control system, and a method for controlling an electric motor. [Background technology]

[0002] Conventionally, there is known a motor control device that controls a motor operated by power supplied from a power source and converted by a power conversion device having an inverter. When the equipment in which the motor is mounted is a compressor, the failure of the compressor is caused, for example, by deterioration or damage of the inverter, the compressor bearings, and the motor. Even if the compressor has not yet failed, if the compressor is continued to be driven in a state in which the inverter, the compressor bearings, and the motor are deteriorated or damaged, the compressor will eventually fail, causing an unexpected downtime for the user. Patent Document 1 discloses a device that operates the compressor when the air conditioner is not operating, and diagnoses whether the compressor has a failure or not. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2004-92976 A Summary of the Invention [Problem to be solved by the invention]

[0004] The fault diagnosis device of Patent Document 1 diagnoses whether or not there is a fault in the compressor when the air conditioner is not operating. In other words, the fault diagnosis device of Patent Document 1 is for performing planned maintenance and inspection of the compressor. However, the occurrence of downtime that is unexpected for the user generally requires unplanned repair arrangements, etc. For this reason, the unexpected downtime for the user may be longer than the predictable downtime such as planned maintenance and inspection. Therefore, in order to improve the overall operating rate of equipment equipped with an electric motor, it is also necessary to suppress the occurrence of unexpected downtime for the user.

[0005] The present disclosure has been made to solve the above-mentioned problems, and aims to provide a motor control device, a motor control system, and a motor control method that prevent unexpected downtime from occurring for a user in equipment equipped with an electric motor. [Means for solving the problem]

[0006] The motor control device according to the present disclosure is a motor control device that controls a motor that operates with power supplied from a power source and converted by a power conversion device having an inverter, and includes a deterioration estimation unit that calculates an estimated deterioration level that estimates the degree of deterioration of a device in which the motor is mounted or the inverter, an operation determination unit that compares the estimated deterioration level with a predetermined standard deterioration level, and sets the operation mode of the motor to normal operation when the estimated deterioration level is less than the standard deterioration level, and sets the operation mode of the motor to low-noise pulse operation when the estimated deterioration level is equal to or greater than the standard deterioration level, and a control unit that controls the inverter based on the operation mode, and the low-noise pulse operation reduces switching loss more than the normal operation. The deterioration estimation unit obtains an estimated deterioration degree based on the intensity of sideband waves appearing in two phases of the three-phase AC current flowing from the inverter to the motor. . Effect of the Invention

[0007] The motor control device of the present disclosure performs low-noise pulse operation that reduces switching loss more than in normal operation. This suppresses the progression of deterioration of the inverter and the motor. This suppresses the occurrence of unexpected downtime for the user in the equipment equipped with the motor until the scheduled maintenance time. [Brief description of the drawings]

[0008] [Figure 1] 1 is a schematic configuration diagram showing a compressor, a power source, and a power conversion device according to a first embodiment. [Diagram 2] 1 is a circuit diagram showing an inverter according to a first embodiment. [Diagram 3] 1 is a schematic configuration diagram showing a compressor according to a first embodiment. [Figure 4] 4 is a diagram for explaining the positional relationship between a main shaft and a main bearing according to the first embodiment. FIG. [Diagram 5] 4 is a diagram for explaining the positional relationship between a main shaft and a main bearing according to the first embodiment. FIG. [Figure 6] FIG. 2 is a functional block diagram showing a control unit according to the first embodiment. [Figure 7] FIG. 2 is a diagram for explaining a method of generating a carrier mode according to the first embodiment. [Figure 8] 5 is a flowchart showing an operation of a correction amount calculation unit according to the first embodiment. [Figure 9] FIG. 4 is a diagram for explaining an estimated deterioration degree of a compressor according to the first embodiment. [Figure 10] FIG. 4 is a diagram for explaining the degree of imbalance of three-phase currents according to the first embodiment. [Figure 11] FIG. 4 is a diagram showing a notification screen according to the first embodiment. [Figure 12] FIG. 4 is a diagram for explaining an output voltage during normal operation according to the first embodiment. [Figure 13] FIG. 4 is a diagram for explaining an output voltage in a low-noise pulse operation according to the first embodiment. [Figure 14] 5 is a flowchart showing the operation of a driving determination unit according to the first embodiment. [Figure 15] FIG. 11 is a functional block diagram showing a control unit according to the second embodiment. [Figure 16] 10 is a flowchart showing an operation of a learning unit according to the second embodiment. [Figure 17] 13 is a flowchart showing the operation of a gate pulse generating section 136 according to the second embodiment. [Figure 18] FIG. 11 is a functional block diagram showing a control unit according to the third embodiment. [Figure 19] FIG. 11 is a spectrum diagram showing the frequency characteristics of the current output to the electric motor in a normal state according to the third embodiment. [Figure 20] FIG. 11 is a spectral diagram showing frequency characteristics of a current output to a deteriorated electric motor according to the third embodiment. [Figure 21] 13 is a flowchart showing the operations of a learning unit and a state observing unit according to the third embodiment. [Figure 22] FIG. 11 is a schematic configuration diagram showing a compressor, a power source, and a power conversion device according to a fourth embodiment. [Diagram 23] FIG. 11 is a spectral diagram showing frequency characteristics of a current output to a deteriorated electric motor according to the fourth embodiment. [Figure 24] 13 is a flowchart showing the operation of a driving determination unit according to the fourth embodiment. [Diagram 25] FIG. 11 is a schematic configuration diagram showing a compressor, a power source, and a power conversion device according to a fifth embodiment. [Figure 26] 4 is a schematic configuration diagram showing a compressor, a power supply, and a power conversion device according to a modified example of the first embodiment. FIG. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0009] The following describes the motor control device 1, the motor control system 1A, the compressor 2, and the control method of the motor 21 of the present disclosure with reference to the drawings. Here, in each drawing, the configurations with the same reference numerals are the same or equivalent configurations, and are common to all the embodiments described below. In addition, the forms of the components shown in all the embodiments described below are merely examples, and are not limited to the forms described below. In particular, the combination of the components is not limited to only the combinations in each embodiment, and the components described in other embodiments can be applied to other embodiments. Furthermore, the high and low of parameters such as pressure and temperature are not determined in relation to absolute values, but are determined relatively in the state and operation of the device, etc.

[0010] Embodiment 1 FIG. 1 is a schematic diagram showing a compressor 2, a power source 3, and a power conversion device 4 according to the first embodiment. As shown in FIG. 1, the compressor 2 has an electric motor 21 and an electric motor control device 1 that controls the electric motor 21. The compressor 2 compresses a fluid such as a refrigerant used in a refrigeration cycle device such as an air conditioner by rotating the electric motor 21 using power supplied from a power source 3 and converted by the power conversion device 4. In the following, a case in which the electric motor 21 is mounted on the compressor 2 and the compressor 2 is applied to an air conditioner will be described as an example. Note that the electric motor 21 may be mounted on another device, and the compressor 2 may be applied to another device. The power source 3 is, for example, a commercial three-phase AC power source having U phase, V phase, and W phase. The electric motor 21 is, for example, a three-phase permanent magnet type synchronous motor having U phase, V phase, and W phase. The power conversion device 4 is connected to the power source 3 and the electric motor 21, which is a load.

[0011] The power conversion device 4 has a rectifier circuit 41, an electrolytic capacitor 42, and an inverter 43. The power source 3, the rectifier circuit 41, the electrolytic capacitor 42, the inverter 43, and the electric motor 21 are connected by wiring 44. The rectifier circuit 41 converts three-phase AC power from the power source 3 into DC power. The electrolytic capacitor 42 smoothes the DC power from the rectifier circuit 41. The inverter 43 converts the DC power from the rectifier circuit 41 into three-phase AC power, and outputs the three-phase AC power to the compressor 2.

[0012] The power conversion device 4 is provided with a current sensor 5. The current sensor 5 is provided on a wiring 44a. The wiring 44a is a portion of the wiring 44 that is midway from the inverter 43 to the motor 21. The current sensor 5 detects currents of two phases of the three-phase AC current that flows from the inverter 43 to the compressor 2. Here, the current sensor 5 detects a current Iu flowing through the U phase and a current Iv flowing through the V phase. In the following, the U phase current Iu and the V phase current Iv may be collectively referred to as a current Iuv. The current sensor 5 transmits the detected current Iuv to the motor control device 1.

[0013] The motor control device 1 executes a method for controlling the motor 21 by adjusting a switching pattern of a gate pulse GP output to the inverter 43. The switching pattern is a combination of switching states for one period of a voltage command value. The voltage command value is a signal that commands the output voltage of the inverter 43. The motor control device 1 generates a gate pulse GP for each period of the voltage command value.

[0014] The motor control device 1 is configured with dedicated hardware, or a storage unit 180 and a CPU (Central Processing Unit) that executes a program stored in the storage unit 180. When the motor control device 1 is dedicated hardware, the motor control device 1 is configured with, for example, an ASIC (Application Specific Integrated Circuit) or an FPGA (Field-Programmable Gate Array).

[0015] When the motor control device 1 is composed of a storage unit 180 and a CPU, each function executed by the motor control device 1 is realized by software, firmware, or a combination of software and firmware. The software and firmware are written as programs and stored in the storage unit 180. Here, the storage unit 180 is, for example, a non-volatile or volatile semiconductor memory such as a RAM, a ROM, a flash memory, an EPROM, or an EEPROM.

[0016] The motor control device 1 has a control unit 130, a deterioration estimation unit 150, a deterioration notification unit 160, and an operation determination unit 170. The control unit 130, the deterioration estimation unit 150, the deterioration notification unit 160, and the operation determination unit 170 will be described later. Note that it is sufficient for the motor control device 1 to have at least the control unit 130, and the deterioration estimation unit 150, the deterioration notification unit 160, and the operation determination unit 170 may be implemented in other hardware or the like.

[0017] (Inverter 43) Fig. 2 is a circuit diagram showing the inverter 43 according to the first embodiment. As shown in Fig. 2, the inverter 43 has six switching elements 43a arranged in upper and lower pairs corresponding to each phase, and six backflow prevention elements 43b provided in parallel to each switching element 43a. The inverter 43 supplies electric power obtained by converting a DC voltage into a three-phase AC voltage to the motor 21 of the compressor 2 in response to a gate pulse GP generated by the control unit 130.

[0018] The substrate material for the switching element 43a and the backflow prevention element 43b may be a wide bandgap semiconductor such as silicon carbide (SiC), gallium nitride (GaN), or diamond. The switching element 43a and the backflow prevention element 43b using wide bandgap semiconductors have high voltage resistance and allowable current, and can be made smaller. By using the miniaturized switching element 43a and the backflow prevention element 43b, the semiconductor module incorporating these elements can be made smaller.

[0019] Furthermore, the switching element 43a and the backflow prevention element 43b using wide band gap semiconductors have high heat resistance, making it possible to miniaturize a cooling mechanism (not shown) required for dissipating heat from the inverter 43. The cooling mechanism is, for example, a heat dissipation fin, a water-cooling mechanism, or an air-cooling mechanism. Note that, for example, by adopting an air-cooling method with a simple structure for the cooling mechanism, the cooling method can be simplified and the semiconductor module incorporating the switching element 43a and the backflow prevention element 43b can be further miniaturized.

[0020] Furthermore, the switching element 43a and the backflow prevention element 43b using wide band gap semiconductors have low power loss and improve power conversion efficiency, so that the compressor 2 can be driven with high conversion efficiency.

[0021] It is preferable that both the switching element 43a and the backflow prevention element 43b are formed using wide band gap semiconductors, but either one of the elements may be formed using wide band gap semiconductors. However, both elements may be formed using materials other than wide band gap semiconductors. Silicon (Si), which is commonly used, is used as a material other than wide band gap semiconductors.

[0022] When the switching element 43a of the inverter 43 is deteriorated, the power supplied from the inverter 43 to the electric motor 21 is unstable. Therefore, if the compressor 2 continues to be driven when the switching element 43a of the inverter 43 is deteriorated, the rotation of the electric motor 21 becomes unstable, which may lead to a breakdown of the compressor 2. Therefore, deterioration of the switching element 43a of the inverter 43 is a factor that causes unexpected downtime for the user.

[0023] (Compressor 2) Fig. 3 is a schematic configuration diagram showing the compressor 2 according to the first embodiment. Fig. 3 shows a cross-sectional view of the compressor 2 cut in the vertical direction. As shown in Fig. 3, the compressor 2 has an electric motor 21, a suction pipe 22, a main shaft 23, an oil pump 24, an auxiliary bearing 25, a main bearing 26, a compression mechanism 27, and a discharge pipe 28. The compressor 2 is connected to a condenser (not shown), an expansion valve (not shown), an evaporator (not shown), and the like by refrigerant piping (not shown).

[0024] The electric motor 21 is connected to three-phase AC wiring 44a (see FIG. 1) and is driven according to a voltage applied from the inverter 43. When the electric motor 21 is in a deteriorated state, the rotation of the electric motor 21 becomes unstable, which may lead to a breakdown of the compressor 2. For this reason, deterioration of the electric motor 21 is a factor that causes unexpected downtime for the user.

[0025] The suction pipe 22 is a pipe for sucking low-temperature, low-pressure refrigerant into the compressor 2. The main shaft 23 is connected to the electric motor 21 and transmits rotational energy to the compression mechanism 27. The oil pump 24 supplies the lubricating oil L accumulated at the bottom of the compressor 2 to the main shaft 23 and the auxiliary bearing 25 to lubricate the main shaft 23 and the auxiliary bearing 25. Note that, as a means for checking the amount of the lubricating oil L, a liquid level sensor capable of detecting the oil level of the lubricating oil L may be attached to measure the amount of the lubricating oil L.

[0026] The auxiliary bearing 25 rotatably supports a lower portion of the main shaft 23. The main bearing 26 rotatably supports an upper portion of the main shaft 23. The compression mechanism 27 compresses the refrigerant supplied from the suction pipe 22, and sends it to the discharge pipe 28. The discharge pipe 28 is a pipe for discharging the high-temperature, high-pressure refrigerant compressed by the compression mechanism 27 to the outside of the compressor 2.

[0027] The compressor 2 has a pressure sensor 71, a flow rate sensor 72, a temperature sensor 73, and a humidity sensor 74. The pressure sensor 71 is attached inside the compressor 2 and measures the pressure of the refrigerant inside the compressor 2. The flow rate sensor 72 is attached to the suction pipe 22 and measures the flow rate of the refrigerant flowing inside the pipe. The temperature sensor 73 is provided outside the housing of the compressor 2 and measures the temperature around the compressor 2. The humidity sensor 74 is provided outside the housing of the compressor 2 and measures the humidity around the compressor 2. Note that a sensor for measuring the pressure, temperature, or humidity of the refrigerant flowing inside the pipe may be attached to the suction pipe 22 or the discharge pipe 28. Measurement information indicating physical quantities such as the refrigerant pressure, refrigerant flow rate, ambient temperature, and ambient humidity measured by the pressure sensor 71, the flow rate sensor 72, the temperature sensor 73, and the humidity sensor 74 is transmitted to a control device (not shown) of the air conditioner, etc., and is used to control each device of the air conditioner.

[0028] Here, the positional relationship between the main shaft 23 and the main bearing 26 when the compressor 2 is driven will be described with reference to Figs. 4 and 5. Figs. 4 and 5 are diagrams for explaining the positional relationship between the main shaft 23 and the main bearing 26 according to the first embodiment. Fig. 4 shows a cross-sectional view of the main shaft 23 and the main bearing 26 when the compressor 2 is cut in a direction perpendicular to the central axis of the main shaft 23. Fig. 4 shows a cross-sectional view when the main shaft 23 and the main bearing 26 are kept in a lubricated state and the compressor 2 is driven normally. Fig. 5 shows a cross-sectional view when the main shaft 23 and the main bearing 26 are not kept in a lubricated state and the compressor 2 is driven abnormally.

[0029] As shown in FIG. 4, when the compressor 2 is operating normally, sufficient lubricating oil L is filled between the main shaft 23 and the main bearing 26, and the main shaft 23 rotates smoothly with a certain clearance secured between the main bearing 26. On the other hand, as shown in FIG. 5, when the compressor 2 is operating abnormally, the main shaft 23 and the main bearing 26 are in contact with each other in part. This is because the viscosity of the lubricating oil L is reduced due to the effects of temperature and aging, and an oil film between the main shaft 23 and the main bearing 26 is not secured. The main bearing 26 wears at the contact portion between the main shaft 23 and the main bearing 26. If the main shaft 23 continues to rotate in a worn state, the wear of the main bearing 26 will worsen, and the compressor 2 will eventually be unable to operate. As described above, wear of the main bearing 26 is a factor that causes unexpected downtime for the user.

[0030] (Control unit 130) Returning to FIG. 1, the control unit 130 of the motor control device 1 will be described. The control unit 130 outputs a gate pulse GP to the inverter 43 based on the current Iuv, the speed command value ω_ref, and the operation mode switching signal O_s. The current Iuv is a current flowing through the wiring 44a detected by the current sensor 5. The speed command value ω_ref is a command value for the rotation speed of the motor 21 according to the load state of the compressor 2. The speed command value ω_ref is calculated comprehensively based on the operation state of the compressor 2 in the air conditioner or an operation instruction from a user. The operation mode switching signal O_s is received from the operation determination unit 170, and is a signal indicating either "ON" or "OFF", indicating the type of operation mode of the motor 21.

[0031] The operation modes of the motor 21 include normal operation and low noise pulse operation. "OFF" of the operation mode switching signal O_s indicates normal operation, and "ON" of the operation mode switching signal O_s indicates low noise pulse operation. The normal operation is an operation in which asynchronous PWM control is performed in which the frequency of the carrier signal is not synchronized with the frequency of the output voltage of the inverter 43. The low noise pulse operation is an operation in which synchronous PWM control is performed in which the frequency of the carrier signal in the PWM control is an integer multiple of the frequency of the output voltage of the inverter 43. In the following, the frequency of the carrier signal in the PWM control may be simply described as "carrier frequency". In this way, the control unit 130 causes the motor 21 to perform either normal operation or low noise pulse operation based on the operation mode set by the operation determination unit 170. It is to be noted that how the operation determination unit 170 determines whether the operation mode switching signal O_s is "ON" or "OFF" will be described later.

[0032] The control unit 130 will be described in detail with reference to Fig. 6. Fig. 6 is a functional block diagram showing the control unit 130 according to the embodiment 1. The control unit 130 has a vector control unit 131, a synchronization pattern selection unit 132, a voltage phase calculation unit 133, a correction amount calculation unit 134, a carrier wave generation unit 135, and a gate pulse generation unit 136.

[0033] A vector control unit 131 performs vector control, which is a known technique, based on a speed command value ω_ref and currents Iuv of two phases of the three-phase AC, and outputs a dq-axis voltage command value Vdq_ref and a reference voltage phase θv subjected to control delay correction to a voltage phase calculation unit 133. Note that the control delay correction is a process for adjusting the timing between the speed command value and the feedback detection value, and generally, taking into account dead time, a value obtained by multiplying the integrated value of the speed command value ω_ref by 1.5 is set to θv.

[0034] The synchronization pattern selection unit 132 generates a carrier mode ptn based on the speed command value ω_ref, and transmits the carrier mode ptn to the correction amount calculation unit 134 and the carrier wave generation unit 135. The carrier mode ptn indicates a PWM control pattern related to motor control. Specifically, the carrier mode ptn has two patterns depending on the difference in the carrier frequency generated by the carrier wave generation unit 135, that is, whether the operation mode is asynchronous PWM control or synchronous PWM control.

[0035] The first pattern is a carrier mode ptn that indicates asynchronous PWM control in which the carrier frequency is set regardless of the frequency of the output voltage of the inverter 43. As described above, since asynchronous PWM control is used in normal operation, the carrier mode ptn of the first pattern is generated when the electric motor 21 is caused to perform normal operation.

[0036] The second pattern is a carrier mode ptn indicating synchronous PWM control in which the carrier frequency is set to be an integer multiple of the frequency of the output voltage of the inverter 43. As described above, synchronous PWM control is used in low noise pulse operation, so the carrier mode ptn of the second pattern is generated when the motor 21 is made to perform low noise pulse operation. Furthermore, as the synchronous PWM control, a plurality of synchronous patterns are used, such as a synchronous 9-pulse mode in which the carrier frequency is nine times the frequency of the output voltage of the inverter 43, and a synchronous 3-pulse mode in which the carrier frequency is three times the frequency. Hereinafter, the synchronous 9-pulse mode may be simply described as "synchronous 9 pulses", and the synchronous 3-pulse mode may be simply described as "synchronous 3 pulses".

[0037] FIG. 7 is a diagram for explaining a method of generating a carrier mode ptn. A method of generating a carrier mode ptn by the synchronization pattern selection unit 132 will be explained with reference to FIG. 7. As shown in FIG. 7, the synchronization pattern selection unit 132 compares the speed at which the electric motor 21 is driven, that is, the speed command value ω_ref, with the switching speeds ω_ref1 [rps] and ω_ref2 [rps] to determine the carrier mode ptn. The switching speeds ω_ref1 [rps] and ω_ref2 [rps] are preset thresholds. The synchronization pattern selection unit 132 generates a carrier mode ptn corresponding to the synchronization mode in a range equal to or greater than the switching speed ω_ref1. In addition, the synchronization pattern selection unit 132 generates the carrier mode ptn so that the frequency ratio between the carrier signal and the output voltage of the inverter 43 is sequentially and stepwise reduced as the speed command value ω_ref increases.

[0038] Specifically, the synchronization pattern selection unit 132 generates the carrier mode ptn as follows. When the speed command value ω_ref is equal to or greater than 0 [rps] and less than ω_ref1 [rps], the synchronization pattern selection unit 132 generates 0, which indicates the asynchronous mode, as the carrier mode ptn. When the speed command value ω_ref is equal to or greater than ω_ref1 [rps] and less than ω_ref2 [rps], the synchronization pattern selection unit 132 generates 9, which indicates the synchronous 9 pulse of the synchronous mode, as the carrier mode ptn. Then, when the speed command value ω_ref is equal to or greater than ω_ref2 [rps], the synchronization pattern selection unit 132 generates 3, which indicates the synchronous 3 pulse of the synchronous mode, as the carrier mode ptn. Therefore, the carrier mode ptn is 0 in the asynchronous mode and is either 3 or 9 in the synchronous mode. The synchronization pattern selection unit 132 transmits the carrier mode ptn to the correction amount calculation unit 134 and the carrier wave generation unit 135 every time the carrier mode ptn is switched.

[0039] The voltage phase calculation unit 133 calculates a voltage phase θv2 that has been subjected to phase adjustment, using the dq-axis voltage command value Vdq_ref and the reference voltage phase θv received from the vector control unit 131. The voltage phase θv2 is, for example, a phase that is advanced by 90 [deg] relative to the reference voltage phase θv. The voltage phase calculation unit 133 transmits the voltage phase θv2 to the correction amount calculation unit 134.

[0040] Furthermore, the voltage phase calculation unit 133 converts the dq-axis voltage command value Vdq_ref into three-phase coordinates using the voltage phase θv2 to calculate the three-phase voltage command value Vuvw_ref. The voltage phase calculation unit 133 transmits the three-phase voltage command value Vuvw_ref to the gate pulse generation unit 136.

[0041] The correction amount calculation unit 134 calculates a carrier cycle correction amount Δtc based on the voltage phase θv2 received from the voltage phase calculation unit 133 and the carrier mode ptn received from the synchronization pattern selection unit 132. The carrier cycle correction amount Δtc is used to correct the difference between the voltage phase θv2 and the phase command value θv2_ref corresponding to the carrier mode ptn. The correction amount calculation unit 134 transmits the calculated carrier cycle correction amount Δtc to the carrier wave generation unit 135.

[0042] The calculation of the carrier cycle correction amount Δtc will be specifically described. The correction amount calculation unit 134 generates a phase command value θv2_ref for each synchronization pulse to generate a carrier wave. That is, the correction amount calculation unit 134 generates a phase command value θv2_ref corresponding to the switched carrier mode ptn every time the carrier mode ptn is switched. For example, when controlling with 9 synchronization pulses, the voltage phase θv2 is preset every 40=360 / 9 [deg] for each cycle of the carrier wave. This ensures continuity when switching the carrier mode.

[0043] Here, a difference may occur between the phase command θv2_ref and the voltage phase θv2. The correction amount calculation unit 134 performs the following process to quickly eliminate the difference. First, for example, using the following formula (1), a phase difference value ΔP corresponding to the difference between the phase command θv2_ref and the voltage phase θv2 is calculated.

[0044] ΔP = θv2_ref - θv2 (1)

[0045] Next, the correction amount calculation unit 134 calculates the carrier period correction amount Δtc using, for example, the following formula (2). In formula (2), GAIN is the carrier period gain. The correction amount calculation unit 134 converts the phase difference value by GAIN to a period difference amount. The carrier period gain GAIN may be set to a fixed value or a variable value as long as it is within a range in which the phase difference value ΔP converges in the entire operating region. For example, when the carrier period gain GAIN is set to a variable value, the carrier period gain GAIN may be set to be adjusted according to the speed command value ω_ref. The correction amount calculation unit 134 also transmits the carrier period correction amount Δtc to the carrier wave generation unit 135.

[0046] Δtc = ΔP × GAIN (2)

[0047] Here, a method in which the correction amount calculation unit 134 generates the carrier cycle correction amount Δtc will be described with reference to Fig. 8. Fig. 8 is a flowchart showing the operation of the correction amount calculation unit 134 according to the first embodiment. First, the correction amount calculation unit 134 generates a phase command value θv2_ref (step S1). Next, the correction amount calculation unit 134 calculates a phase difference value ΔP (step S2). Then, the correction amount calculation unit 134 calculates the carrier cycle correction amount Δtc (step S3).

[0048] The carrier wave generating unit 135 generates a carrier wave Carrier based on the carrier mode ptn received from the synchronization pattern selecting unit 132, the carrier period correction amount Δtc received from the correction amount calculating unit 134, and the speed command value ω_ref. Specifically, the carrier wave generating unit 135 calculates the carrier period tc by equation (3).

[0049] tc=1 / (ptn×ω_ref)+Δtc···(3)

[0050] Then, the carrier wave generating unit 135 outputs the carrier wave Carrier based on the carrier period tc such that the carrier period correction amount Δtc converges to 0. As a result, the frequency of the carrier wave Carrier output by the carrier wave generating unit 135 quickly converges to the accurate frequency ptn×ω_ref of the synchronization pulse after the carrier mode is switched. The carrier wave generating unit 135 transmits the carrier wave Carrier to the gate pulse generating unit 136. Furthermore, when the carrier mode ptn is 0 indicating the asynchronous mode, the carrier wave generating unit 135 transmits a fixed carrier frequency Carrier to the gate pulse generating unit 136.

[0051] The gate pulse generating unit 136 compares the three-phase voltage command value Vuvw_ref received from the voltage phase calculation unit 133 with the carrier wave Carrier received from the carrier wave generating unit 135 to output a gate pulse GP. Here, the gate pulse generating unit 136 uses either the fixed carrier frequency during normal operation or the carrier frequency during low-noise pulse operation as a comparison target according to the operation mode switching signal O_s received from the operation determination unit 170. Specifically, when the operation mode switching signal O_s is "OFF", the gate pulse generating unit 136 outputs a gate pulse GP with the fixed carrier frequency Carrier during normal operation as a comparison target. At this time, the electric motor 21 performs normal operation. Also, when the operation mode switching signal O_s is "ON", the gate pulse generating unit 136 outputs a gate pulse GP with the carrier frequency Carrier during low-noise pulse operation as a comparison target. At this time, the electric motor 21 performs low-noise pulse operation.

[0052] (Deterioration estimation unit 150) Returning to Fig. 1, the deterioration estimation unit 150 of the motor control device 1 will be described. The deterioration estimation unit 150 obtains an estimated deterioration degree W_est indicating the degree of deterioration estimated for the inverter 43, the main bearing 26 of the compressor 2, and the electric motor 21, which are factors that cause a failure of the compressor 2, based on the current Iuv of two phases of the three-phase AC. The inverter 43, the main bearing 26 of the compressor 2, and the electric motor 21, which are factors that cause a failure of the compressor 2, are referred to as estimation targets. However, any one or two of the inverter 43, the main bearing 26 of the compressor 2, and the electric motor 21 may be the estimation targets.

[0053] Specifically, sidebands appear in the current Iuv of two phases of the three-phase AC, with a specific frequency as the center, for each estimation target. The deterioration estimation unit 150 obtains the estimated deterioration degree W_est by analyzing the intensity of the sidebands appearing in the current Iuv. Note that, when there are multiple types of estimation targets, the estimated deterioration degree W_est may be calculated for each type of estimation target, and the highest estimated deterioration degree W_est may be calculated.

[0054] Furthermore, the deterioration estimation unit 150 may calculate the estimated deterioration degree W_est based on the degree of unbalance of the three-phase current Iuvw. The degree of unbalance of the three-phase current Iuvw is calculated based on the U-phase and V-phase currents Iuv and the W-phase current Iw obtained using equation (4). The deterioration estimation unit 150 may also perform a frequency domain transformation of the three-phase currents and calculate the estimated deterioration degree W_est based on the transformed values.

[0055] Iw = -Iu -Iv (4)

[0056] The degree of imbalance of the three-phase current Iuvw occurs as follows. FIG. 10 is a diagram for explaining the degree of imbalance of the three-phase current Iuvw according to the first embodiment. First, in the compressor 2, when the main bearing 26 deteriorates due to various factors such as poor lubrication and foreign matter contamination, the gap between the main shaft 23 and the main bearing 26 widens as shown in FIG. 5 compared to a state in which the main bearing 26 is not deteriorated (FIG. 4). At this time, the main shaft 23 whirls, and a sideband wave of a fundamental wave frequency is generated. Then, as shown in FIG. 3, the main bearing 26 supports the radial movement of the main shaft 23, so that when the gap between the main shaft 23 and the main bearing 26 widens, the main shaft 23 is displaced in a tilting direction. Accordingly, as shown in FIG. 10, in the electric motor 21, the rotor core 21a is attracted to the stator core 21b by a magnetic attraction force and rotates, and the gap between the inner circumference of the stator core 21b and the rotor core 21a becomes uneven. 10, the movable range of rotor core 21a of motor 21 when main bearing 26 is deteriorated is shown by a dashed line. This non-uniformity in the gap causes the impedance of motor 21 to be unbalanced, resulting in a state in which the impedance differs between phases. As a result, the current waveform contains an unbalanced component.

[0057] Here, the estimated deterioration degree W_est and the reference deterioration degree x will be described in detail with reference to FIG. 9. FIG. 9 is a diagram for explaining the estimated deterioration degree W_est and the reference deterioration degree x of the compressor 2 according to the first embodiment. In FIG. 9, the horizontal axis represents time, and the vertical axis represents the estimated deterioration degree W_est of the inverter 43 and the compressor 2. As shown in FIG. 9, the estimated deterioration degree W_est gradually increases with time. The larger the estimated deterioration degree W_est, the more deteriorated the estimation target is. The reference deterioration degree x is determined in advance by an experiment or the like so that the time until the compressor breaks down when the estimation target reaches the reference deterioration degree x is a predetermined time. The predetermined time is, for example, a time that is considered to be sufficient for a user to arrange for repairs or the like.

[0058] (Deterioration notification section 160) The deterioration notification unit 160 displays a notification screen on the user terminal 200, indicating that deterioration has occurred in the equipment in which the electric motor 21 is mounted, according to the estimated deterioration degree W_est received from the deterioration estimation unit 150. Specifically, the deterioration notification unit 160 determines whether the estimated deterioration degree W_est is equal to or greater than a predetermined reference deterioration degree x. If the estimated deterioration degree W_est is equal to or greater than the predetermined reference deterioration degree x, the deterioration notification unit 160 displays a notification screen on the user terminal 200 (see FIG. 1). The user terminal 200 is, for example, a setting monitor or a remote control provided in an indoor unit of an air conditioner and having a display and buttons. The setting monitor or the remote control may have a touch-type display instead of a combination of a display and buttons. The user terminal 200 is an example of a "display device" in the present disclosure.

[0059] FIG. 11 is a diagram illustrating a notification screen G according to the first embodiment. In the deterioration notification unit 160, as shown in FIG. 11, the notification screen G includes a statement M1 notifying that deterioration has occurred in the device in which the motor 21 is mounted, and a statement M2 asking whether or not to accept the start of low-noise pulse operation. The "life extension operation" included in the statement M1 is a representation of low-noise pulse operation that is easy for the user to understand. As described above, the objects of estimated deterioration are the inverter 43, the main bearing 26 of the compressor 2, and the motor 21, but in FIG. 11, the object of failure is simply expressed as "device". By displaying the notification screen G by the deterioration notification unit 160, the user is made to recognize that the device is about to fail, and it is possible to prompt the user to secure maintenance resources, such as arranging for repairs.

[0060] The user terminal 200 accepts an operation of accepting or rejecting the start of low-noise pulse operation by the user who has confirmed the notification screen G. If the user accepts the life-extension operation, that is, selects "Yes" in FIG. 11, the operation determination unit 170 outputs "ON" of the low-noise pulse operation start signal L_s to the control unit 130. On the other hand, if the user rejects the life-extension operation, that is, selects "No" in FIG. 11, the operation determination unit 170 outputs "OFF" of the low-noise pulse operation start signal L_s. In addition to the above, the user terminal 200 is also used when the user inputs a set temperature or a set airflow, for example.

[0061] (Driving decision unit 170) The operation determination unit 170 transmits an "ON" signal of the operation mode switching signal O_s to the control unit 130 when the estimated deterioration level W_est is equal to or greater than the reference deterioration level x, the speed command value ω_ref is equal to or greater than the switching speed ω_ref1, and the low-noise pulse operation start signal L_s is output as "ON." When the above conditions are not met, the operation determination unit 170 transmits an "OFF" signal of the operation mode switching signal O_s to the control unit 130. When the "ON" signal of the operation mode switching signal O_s is input, the control unit 130 executes low-noise pulse operation. Also, when the "OFF" signal of the operation mode switching signal O_s is input, the control unit 130 executes normal operation.

[0062] Furthermore, when the estimated deterioration level W_est is equal to or greater than the reference deterioration level x, the operation determination unit 170 displays on a management terminal 300 (see FIG. 1) or the like provided outside the compressor 2 that the equipment in which the motor 21 is provided is deteriorated. This notifies the manager who manages the equipment that the equipment is deteriorated, and allows the manager to prepare for maintenance work. The management terminal 300 is a dedicated hardware device for managing the compressor 2, or a general-purpose computer.

[0063] Here, the output voltage in normal operation and low noise pulse operation will be described with reference to Figs. 12 and 12. Fig. 12 is a diagram for explaining the output voltage in normal operation according to the first embodiment. Fig. 13 is a diagram for explaining the output voltage in low noise pulse operation according to the first embodiment. In both diagrams, the solid line represents the voltage applied from the inverter 43 to the compressor 2, and the dotted line represents the voltage command value of any one phase among the three-phase voltage command values ​​Vuvw_ref. Here, the description will be made assuming that it represents one cycle of the U-phase voltage command value Vu_ref.

[0064] In normal operation with asynchronous PWM control, when the rotation speed of the motor 21 increases and the number of output voltage pulses decreases, the positive and negative pulses of the output voltage become irregular with respect to one period of the U-phase voltage command value Vu_ref, as shown in Fig. 12. This causes the current supplied to the motor 21 to become disrupted, which ultimately leads to a breakdown of the compressor 2.

[0065] On the other hand, in low-noise pulse operation that performs synchronous PWM control, as shown in FIG. 13, the positive and negative pulse outputs of the output voltage are equal for one period of the U-phase voltage command value Vu_ref. In other words, the output voltage pulse waveform repeats positive and negative waveforms of the same waveform in each period. This prevents the current supplied to the motor 21 from being disturbed, and the compressor 2 operates stably. This suppresses the progression of deterioration of each component of the compressor 2 compared to normal operation.

[0066] In addition, in low-noise pulse operation, the carrier mode ptn is generated so that the frequency ratio between the carrier signal and the output voltage of the inverter 43 is gradually reduced as the speed command value ω_ref increases. Therefore, the frequency of the carrier signal decreases as the speed command value ω_ref increases. As the carrier signal decreases, the number of switching operations of the inverter 43 also decreases, and the switching loss is reduced. In this way, in low-noise pulse operation, the switching pattern of the inverter 43 is adjusted so that the switching loss is reduced more than in normal operation.

[0067] The reduction in switching loss also suppresses the rise in temperature of the switching element 43a and the backflow prevention element 43b of the inverter 43. This suppresses the progression of deterioration of the inverter 43. Furthermore, the reduction in switching loss also suppresses switching noise that occurs as a transient phenomenon during the same switching, so the progression of deterioration of the electric motor 21 and the main bearing 26 is also suppressed.

[0068] In addition, in the synchronous PWM operation, the pulse output of the output voltage is equivalent between the upper and lower switching elements 43a of the inverter 43, so it is possible to suppress uneven progression of deterioration of the upper and lower switching elements 43a. As described above, the low-noise pulse operation can suppress the progression of deterioration of the inverter 43, the main bearing 26 of the compressor 2, and the electric motor 21, which are the subjects of the estimation.

[0069] The operation of the operation determination unit 170 will be described with reference to FIG. 14. FIG. 14 is a flowchart showing the operation of the operation determination unit 170 according to the first embodiment. First, the operation determination unit 170 judges whether the estimated deterioration level W_est is equal to or greater than the reference deterioration level x (step S11). If the estimated deterioration level W_est is less than the reference deterioration level x (step S11: NO), the operation determination unit 170 transmits an "OFF" signal of the operation mode switching signal O_s to the control unit 130 (step S16). This causes the electric motor 21 to perform normal operation. If the estimated deterioration level W_est is equal to or greater than the reference deterioration level x (step S11: YES), the operation determination unit 170 notifies the management terminal 300 that the equipment in which the electric motor 21 is provided is deteriorated (step S12).

[0070] Next, the operation determination unit 170 determines whether or not the speed command value ω_ref is equal to or greater than the switching speed ω_ref1 (step S13). If the speed command value ω_ref is less than the switching speed ω_ref1 (step S13: NO), the operation determination unit 170 transmits an "OFF" signal of the operation mode switching signal O_s to the control unit 130 (step S16). This causes the electric motor 21 to perform normal operation. If the speed command value ω_ref is equal to or greater than the switching speed ω_ref1 (step S13: YES), the operation determination unit 170 determines whether or not the low-noise pulse operation start signal L_s received from the deterioration notification unit 160 is "ON" (step S14).

[0071] When the low-noise pulse operation start signal L_s received from the deterioration notification unit 160 is "OFF" (step S14: NO), the operation determination unit 170 transmits an "OFF" signal of the operation mode switching signal O_s to the control unit 130 (step S16). As a result, the electric motor 21 performs normal operation. When the low-noise pulse operation start signal L_s received from the deterioration notification unit 160 is "ON" (step S14: YES), the operation determination unit 170 transmits an "ON" signal of the operation mode switching signal O_s to the control unit 130 (step S15). As a result, the electric motor 21 performs low-noise pulse operation.

[0072] As described above, the motor control device 1 of the first embodiment performs low-noise pulse operation in which the switching pattern of the inverter 43 is adjusted so as to reduce switching loss compared to normal operation. This suppresses deterioration of the inverter 43 and the motor 21. This suppresses the occurrence of unanticipated downtime for the user in the device equipped with the motor 21 until the scheduled maintenance time.

[0073] Furthermore, in the case where the equipment in which the electric motor 21 is mounted is the compressor 2, deterioration of the bearings of the compressor 2 is suppressed. Therefore, in the compressor 2, the occurrence of unforeseen downtime for the user until the scheduled maintenance time is reached is suppressed.

[0074] As described above, the motor control device 1 of the first embodiment executes a method for controlling the motor 21, which controls the motor 21 operated by the power supplied from the power source 3 and converted by the power conversion device 4 having the inverter 43. Specifically, in the method for controlling the motor 21 of the first embodiment, first, an estimated deterioration degree is calculated by estimating the degree of deterioration of the device in which the motor 21 is mounted or the inverter 43. Next, the estimated deterioration degree is compared with a predetermined reference deterioration degree, and the operation mode of the motor 21 is set to normal operation when the estimated deterioration degree is less than the reference deterioration degree, and is set to low-noise pulse operation when the estimated deterioration degree is equal to or greater than the reference deterioration degree. Then, the inverter 43 is controlled based on the set operation mode. Here, the low-noise pulse operation reduces switching loss more than the normal operation. Therefore, deterioration of the inverter 43 and the motor 21 is suppressed. Therefore, according to the method for controlling the motor 21 of the first embodiment, it is possible to suppress the occurrence of unexpected downtime for the user in the device in which the motor 21 is mounted until the planned maintenance time.

[0075] Embodiment 2 Fig. 15 is a functional block diagram showing a control unit 130 according to embodiment 2. As shown in Fig. 15, embodiment 2 differs from embodiment 1 in that a control unit 130 of a motor control device 1 has a learning unit 141 and an inference unit 142. In embodiment 2, the same parts as those in embodiment 1 are given the same reference numerals and their description is omitted, and the description will focus on the differences from embodiment 1.

[0076] The motor control device 1 according to the second embodiment uses so-called artificial intelligence (AI) to perform low-noise pulse operation of the gate pulse GP transmitted to the inverter 43, which reduces switching loss more than asynchronous PWM control. A GPU (Graphics Processing Unit) may be added to the hardware configuration of the motor control device 1 in addition to those described in the first embodiment.

[0077] (Voltage phase calculation unit 133) In the second embodiment, unlike the first embodiment, the voltage phase calculation unit 133 does not calculate the voltage phase θv2.

[0078] (Gate pulse generating unit 136) The gate pulse generating unit 136 has an inference unit 142. The gate pulse generating unit 136 outputs different gate pulses GP depending on whether the operation mode is normal operation or low-noise pulse operation, based on the operation mode switching signal O_s. In normal operation, the gate pulses GP that perform the asynchronous PWM control described in the first embodiment are output. In low-noise pulse operation, the gate pulses GP based on the processing performed by the inference unit 142 are output. The inference unit 142 will be described later.

[0079] (Learning Section 141) The learning unit 141 uses a supervised learning method to generate a trained model PGF that generates a switching pattern of the gate pulse GP. The trained model PGF is a program including a function that generates a switching pattern using parameters adjusted by learning. The trained model PGF is generated by the learning unit 141 when the motor 21 is operating stably, in other words, when the load and speed of the motor 21 fluctuate little. Here, it is assumed that if the fluctuation of the three-phase voltage command value Vuvw_ref is within a predetermined value, the fluctuation of the load and speed of the motor 21 is small, and the trained model PGF is generated by the learning unit 141.

[0080] A method for generating the trained model PGF will be specifically described below. The learning unit 141 uses, as learning data, a set of data D_set obtained by collecting a plurality of periods of the switching patterns of the three-phase voltage command value Vuvw_ref and the gate pulse GP. The control methods used for the set of data D_set include model predictive control, low-order harmonic elimination, and optimized pulse patterns. All of these are control methods that reduce the switching loss of the inverter 43 compared to asynchronous PWM control. Note that a switching pattern that has been reproduced in advance in a simulation environment that simulates deterioration of the device may be used as the set of data D_set. The set of data D_set is stored in the storage unit 180 in advance.

[0081] Specifically, the data set D_set is a record of multiple switching patterns for each period, for an arbitrary period t, of the gate pulse GP_tr(t-1) generated in the previous period t-1, and the switching pattern of the gate pulse GP_tr(t) generated in an arbitrary period t when the three-phase voltage command value Vuvw_ref_tr(t) at the arbitrary period t is determined.

[0082] The learning unit 141 includes a data acquiring unit 141a and a model generating unit 141b. The data acquiring unit 141a acquires, as input data, a switching pattern of a gate pulse GP_tr(t-1) generated in a period t-1 preceding an arbitrary period, and a three-phase voltage command value Vuvw_ref_tr(t) in an arbitrary period t, from a data set D_set. The data acquiring unit 141a also acquires, as a label, a switching pattern of a gate pulse GP_tr(t) generated in an arbitrary period t. The data acquiring unit 141a acquires a plurality of teacher data sets consisting of the input data and the labels from a data set D_set in which a plurality of switching patterns of the gate pulse GP are recorded for each period.

[0083] The model generating unit 141b performs learning so that the switching pattern of the gate pulse GP(t) approaches the switching pattern of the gate pulse GP_tr(t) of the teacher data set in a model that outputs a switching pattern of the gate pulse GP(t) from the switching pattern of the gate pulse GP_tr(t-1) generated in the period t-1 before an arbitrary period of the teacher data set and the three-phase voltage command value Vuvw_ref_tr(t) of an arbitrary period t. The model generating unit 141b performs learning of the model based on teacher data sets of multiple periods.

[0084] Specifically, the model consists of a neural network composed of multiple perceptrons, and each perceptron is set with a bias value and a weighting coefficient. The bias value and weighting coefficient are parameters that are adjusted by learning. In learning, multiple pieces of training data are provided to the neural network, and the bias value and weighting coefficient of each perceptron are adjusted so that the output of the neural network approaches the label. Back-propagation and other methods are used to adjust the bias value and weighting coefficient of the perceptron. In back-propagation, the bypass value and weighting coefficient are adjusted to reduce the error between the output of the neural network and the label.

[0085] Note that the neural network used by the model generating unit 141b for learning may be multi-layered, so that learning may be performed by so-called deep learning.

[0086] By the above-mentioned method, the model generating unit 141b generates a trained model PGF for inferring a switching pattern of an arbitrary cycle by using learning data including a voltage command value in an arbitrary cycle and a switching pattern of a cycle one cycle before the arbitrary cycle. That is, the trained model PGF learns the characteristics of the teacher data set, and outputs a switching pattern of the gate pulse GP(t) from the switching pattern of the gate pulse GP(t-1) determined one cycle before and the three-phase voltage command value Vuvw_ref(t) during control. The trained model PGF is stored in the storage unit 180. The trained model PGF is used by the inference unit 142 to generate the gate pulse GP. Note that, for simplification, FIG. 15 does not show the storage unit 180, and illustrates that the trained model PGF is acquired from the learning unit 141 by the inference unit 142.

[0087] (Inference part 142) The inference unit 142 inputs, to the learned model PGF, the switching pattern of the gate pulse GP(t-1) determined one period earlier and the three-phase voltage command value Vuvw_ref(t) during control, to determine the switching pattern of the gate pulse GP(t) to be output.

[0088] As described above, the data set D_set used in the learning is a control pattern in a control method that outputs a gate pulse GP that reduces the switching loss of the inverter 43 compared to asynchronous PWM control. Therefore, the learned model PGF can also output a gate pulse GP that reduces the switching loss of the inverter 43 compared to asynchronous PWM control. Hereinafter, the control of the inverter 43 using the learned model PGF performed by the gate pulse generating unit 138 having the inference unit 142 may be referred to as AI switching control.

[0089] The learning unit 141 may construct multiple trained models PGF depending on the type of control method used during learning. For example, in addition to a trained model PGF intended only to reduce switching loss, a trained model PGF may be constructed to achieve other purposes according to user requests, such as reducing vibration of the compressor 2. When multiple trained models PGF are constructed, a pattern table in which the trained models PGF correspond to individual numbers assigned to each trained model PGF is stored in the storage unit 180.

[0090] When multiple trained models PGF are constructed, the inference unit 142 generates a type selection command TSC for the trained model PGF and outputs it to the storage unit 180, thereby acquiring the trained model PGF corresponding to the type selection command TSC from the storage unit 180. The type selection command TSC includes an individual number assigned to each trained model PGF in the pattern table, and the storage unit 180 can read out the corresponding trained model PGF. Note that, for the sake of simplicity, FIG. 15 does not show the storage unit 180, and illustrates the type selection command TSC being output from the learning unit 141 to the inference unit 142.

[0091] Fig. 16 is a flowchart showing the operation of the learning section 141 according to the second embodiment. Fig. 17 is a flowchart showing the operation of the gate pulse generating section 136 according to the second embodiment. Note that the processing procedure described below is an example of a learning method by the learning section 141. Therefore, each process may be changed as much as possible, and processes may be omitted, replaced, or added as appropriate depending on the embodiment.

[0092] First, as shown in FIG. 16, the learning unit 141 determines whether or not to perform learning (step S21). If learning is to be performed (step S21: YES), the data acquisition unit 141a of the learning unit 141 acquires input data from the teacher data (step S22). Next, the data acquisition unit 141a of the learning unit 141 acquires labels from the teacher data (step S23). Then, the model generation unit 141b of the learning unit 141 performs learning with teacher data using a teacher data set consisting of these input data and labels, and acquires a trained model PGF (step S24). The learning unit 141 stores the acquired trained model PGF in the storage unit 180 (step S25).

[0093] When learning is not performed (step S21: NO), as shown in Fig. 17, the gate pulse generating unit 136 determines whether or not the operation mode switching signal O_s is "ON" (step S31). When the operation mode switching signal O_s is "OFF" (step S31: NO), the gate pulse generating unit 136 generates a gate pulse based on asynchronous PWM control (step S32). Then, the gate pulse generating unit 136 controls the inverter 43 with the gate pulse based on asynchronous PWM control (step S36), thereby causing the inverter 43 to perform normal operation.

[0094] When the operation mode switching signal O_s is "ON" (step S31: YES), the gate pulse generating unit 136 acquires the learned model PGF stored in the storage unit 180 (step S33). Next, the gate pulse generating unit 136 acquires the switching pattern of the gate pulse GP(t-1) generated one period before the control time point, which is input data, and the three-phase voltage command value Vuvw_ref(t) at the control time point (step S34). The gate pulse generating unit 136 inputs the input data to the learned model PGF and generates a switching pattern of the gate pulse based on the AI ​​switching control (step S35). Then, the gate pulse generating unit 136 controls the inverter 43 with the gate pulse based on the AI ​​switching control (step S36), thereby causing the inverter 43 to perform low-noise pulse operation.

[0095] Similarly to the first embodiment, the motor control device 1 of the second embodiment also performs low-noise pulse operation in which the switching pattern of the inverter 43 is adjusted so as to reduce switching loss compared to the normal operation. This suppresses the progression of deterioration of the inverter 43 and the motor 21. This suppresses the occurrence of unanticipated downtime for the user in the device equipped with the motor 21 until the scheduled maintenance time.

[0096] Furthermore, in the second embodiment, by generating multiple trained models PGF for each purpose, various effects such as reducing vibration of the compressor 2 can be obtained according to the user's request.

[0097] Embodiment 3 FIG. 18 is a functional block diagram showing a control unit 130 according to the third embodiment. As shown in FIG. 18, the third embodiment differs from the second embodiment in the following two points. The first difference is that the motor control device 1 has a state observation unit 137. The second difference is that the learning unit 141 has a reward calculation unit 141c and a function update unit 141d. In the third embodiment, the same parts as those in the first and second embodiments are given the same reference numerals and their explanations are omitted, and the differences from the first and second embodiments will be mainly explained.

[0098] The motor control device 1 of the third embodiment performs reinforcement learning in addition to the motor control device 1 of the second embodiment. Reinforcement learning is a learning method for learning an agent to maximize value in a given environment. Q-learning and TD-learning are known as representative methods of reinforcement learning. Here, the given environment corresponds to the current flowing through the motor 21, that is, the state of the compressor 2, the value corresponds to the reduction in switching loss of the inverter 43, and the agent corresponds to the learned model PGF. That is, in the third embodiment, while controlling the motor 21 based on the learned model PGF generated by learning with teacher data, reinforcement learning of low-noise pulse operation based on the learned model PGF is repeatedly performed so as to further reduce switching loss.

[0099] (Gate pulse generating unit 136) The gate pulse generating unit 136 acquires a trained model PGF in the initial state of the motor control device 1. The trained model PGF in the initial state is, for example, a trained model PGF acquired first after the operation of the motor 21 becomes stable.

[0100] The gate pulse generating unit 136 transmits the gate pulse GP in asynchronous PWM control or the gate pulse GP based on the learned model PGF in the initial state of the motor control device 1 to the inverter 43 and the state observing unit 137 based on the operation mode switching signal O_s. When the learned model PGF is used, as described above, during control, the switching pattern of the gate pulse GP(t) is output from the switching pattern of the gate pulse GP(t-1) determined one period before and the three-phase voltage command value Vuvw_ref(t) during control. By outputting the gate pulse GP to the inverter 43, the motor 21 of the compressor 2 is driven at a speed corresponding to the three-phase voltage command value Vuvw_ref(t).

[0101] (Status Observation Unit 137) The state observing unit 137 counts the number of switching operations Sw_count from the gate pulse GP. The state observing unit 137 also acquires a sideband intensity Sw_level indicating the intensity of the sideband a of the fundamental frequency fa based on the result of frequency conversion of the current of any one of the three-phase AC currents Iuvw. Note that, although the example shown here is one in which the target of frequency conversion is the U-phase current Iu, the target of frequency conversion may be a current of another phase.

[0102] Here, the deterioration state of the motor 21 will be described with reference to Fig. 19 and Fig. 20. Fig. 19 is a spectrum diagram showing the frequency characteristics of the current Iu output to the motor 21 in a normal state. Fig. 20 is a spectrum diagram showing the frequency characteristics of the current Iu output to the motor 21 in which deterioration has progressed. Figs. 19 and 20 show the case where the load conditions of the motor 21 are the same. In Figs. 19 and 20, the vertical axis represents the "intensity" of each frequency, and the horizontal axis represents the "frequency."

[0103] As shown in FIG. 19 and FIG. 20, even if the motor 21 is normal or if the motor 21 is deteriorating, the intensity of the fundamental frequency fa is higher than the other frequencies. The intensity of the fundamental frequency fa when the motor 21 is normal is equal to the intensity of the fundamental frequency fa when the motor 21 is deteriorating. However, as shown in FIG. 20, when the motor 21 is deteriorating, a sideband wave a is generated with the fundamental frequency fa as the center. The intensity of this sideband wave a increases as the deterioration of the motor 21 progresses. Therefore, the degree of deterioration of the motor 21 can be determined by the degree of the sideband wave a with the fundamental frequency fa as the center, that is, the sideband wave intensity Sw_level. Note that the deterioration of the motor 21 here means the early deterioration, which is classified into so-called early deterioration and later deterioration. The later deterioration is an irreversible change. On the other hand, the early deterioration is a change in the presence or absence of the occurrence of the whirling phenomenon depending on the influence of the load on the motor 21, and therefore the sideband wave intensity Sw_level is changed.

[0104] (Remuneration Calculation Department 141c) The reward calculation unit 141c determines the reward based on the sideband wave strength Sw_level calculated by the state observation unit 137. Specifically, the reward calculation unit 141c determines whether the sideband wave strength Sw_level calculated by the state observation unit 137 is within a preset value. When the reward calculation unit 141c determines that the sideband wave strength Sw_level is within the preset value, it increases a change amount Δ1 that is preset as a reward. On the other hand, when the reward calculation unit 141c determines that the sideband wave strength Sw_level exceeds the preset value, it decreases the change amount Δ1 that is preset.

[0105] Similarly, the reward calculation unit 141c determines the reward based on the switching count Sw_count calculated by the state observation unit 137. Specifically, the reward calculation unit 141c determines whether the switching count Sw_count calculated by the state observation unit 137 has decreased from the switching count Sw_count based on the trained model PGF before the update. If the switching count Sw_count has decreased from the switching count Sw_count based on the trained model PGF before the update, the reward calculation unit 141c increases a change amount Δ2 preset as a reward. If the switching count Sw_count has increased from the switching count based on the trained model PGF before the update, the reward calculation unit 141c decreases the change amount Δ2 preset as a reward. Note that the reward based on the switching count Sw_count may be determined by comparison with a predetermined specified value.

[0106] (Function update unit 141d) The function update unit 141d updates the value function based on the change amount Δ1 and the change amount Δ2 obtained as the reward by the reward calculation unit 141c. Specifically, the weighting coefficient and the bias are adjusted to maximize the change amount Δ1 and the change amount Δ2, and the value function is updated so that the number of switching times is reduced while maintaining the intensity of the sideband wave a within a range of a preset value. In addition, by reducing the number of switching times, the switching loss is also reduced. Then, the trained model is updated based on the updated value function. The value function is, for example, one used in known Q-learning. Updating the trained model means that the weighting coefficient and bias of each perceptron of the neural network that constitutes the trained model are adjusted.

[0107] 21 is a flowchart showing the operations of the learning unit 141 and the state observing unit 137 according to the third embodiment. Note that the processing procedure described below is an example of a learning method by the learning unit 141. Each process may be changed as much as possible, and processes may be omitted, replaced, or added as appropriate.

[0108] First, the learning unit 141 acquires a trained model PGF in the initial state of the motor control device 1 (step S41). Next, the learning unit 141 determines a switching pattern based on the acquired trained model PGF (step S42), and the state observation unit 137 acquires the intensity of sideband waves in the current flowing through the motor 21 at that time (step S43).

[0109] The reward calculation unit 141c of the learning unit 141 determines whether the strength of the sideband wave of the current flowing through the motor 21 is equal to or greater than a specified value (step S44). If the strength of the sideband wave of the current flowing through the motor 21 is within the specified value (step S44: YES), the reward calculation unit 141c increases the reward (step S45). If the strength of the sideband wave of the current flowing through the motor 21 is less than the specified value (step S44: NO), the reward calculation unit 141c decreases the reward (step S46).

[0110] The reward calculation unit 141c determines whether the number of switching times for the voltage command has decreased from the number of switching times Sw_count based on the trained model PGF before the update (step S47). If the number of switching times for the voltage command has decreased from the number of switching times Sw_count based on the trained model PGF before the update (step S47: YES), the reward calculation unit 141c increases the reward (step S48). If the number of switching times for the voltage command has increased from the number of switching times Sw_count based on the trained model PGF before the update (step S47: NO), the reward calculation unit 141c decreases the reward (step S49). The function update unit 141d of the learning unit 141 updates the trained model PGF based on the reward given based on the strength of the sideband wave of the current flowing through the motor 21 and the reward given based on the number of switching times for the voltage command. Then, the processes of S42 to S50 are repeated to obtain a trained model PGF that generates a switching pattern in which switching loss is further reduced.

[0111] Similarly to the first and second embodiments, the motor control device 1 of the third embodiment also performs low-noise pulse operation in which the switching pattern of the inverter 43 is adjusted so as to reduce switching loss compared to the normal operation. This suppresses the progression of deterioration of the inverter 43 and the motor 21. This suppresses the occurrence of unanticipated downtime for the user in the device equipped with the motor 21 until the scheduled maintenance time.

[0112] Moreover, according to the third embodiment, the trained model generated in the second embodiment is updated. Therefore, in the third embodiment, every time the trained model PGF is updated, the switching loss of the inverter 43 can be further reduced compared to the trained model in the second embodiment.

[0113] Embodiment 4 FIG. 22 is a schematic diagram showing a compressor 2, a power source 3, and a power conversion device 4 according to the fourth embodiment. The fourth embodiment differs from the first to third embodiments in the following three points. The first difference is that, as shown in FIG. 22, the deterioration estimation unit 150 transmits the deterioration point W_point to the operation determination unit 170. The second difference is that the control unit 130 transmits the drive time D_time to the operation determination unit 170. The third difference is that, as low noise pulse operation, the asynchronous PWM control described in the first embodiment and the control using the AI ​​switching control described in the second and third embodiments are selectively used. In the fourth embodiment, the same parts as those in the first to third embodiments are denoted by the same reference numerals and the description is omitted, and the differences from the first to third embodiments are mainly described.

[0114] (Deterioration estimation unit 150) The deterioration estimation unit 150 identifies the deterioration point W_point based on the result of frequency conversion of the U-phase current Iu, which is one phase of the three-phase AC current. Note that, although the case where the target of the frequency conversion is the U-phase current Iu is taken as an example here, the target of the frequency conversion may be a current of another phase. The deterioration estimation unit 150 outputs the identified deterioration point W_point to the operation determination unit 170. The deterioration point W_point includes information indicating either the motor 21 or the inverter 43.

[0115] A method of determining a deteriorated portion will be described with reference to FIG. 23. FIG. 23 is a spectrum diagram showing the frequency characteristics of a current output to the deteriorated motor 21 according to the fourth embodiment. In FIG. 23, the vertical axis indicates the "intensity" of each frequency, and the horizontal axis indicates the "frequency." As shown in FIG. 23, when the motor 21 is deteriorated, a sideband wave a is generated centered on the fundamental frequency fa. Also, when the inverter 43 is deteriorated, a sideband wave b is generated centered on the switching frequency fc. For this reason, if the frequency region near the fundamental frequency fa is set as the low frequency region and the frequency region near the switching frequency fc is set as the high frequency region, it can be determined that deterioration has occurred in the motor 21 if the sideband wave a exists in the low frequency region. Similarly, it can be determined that deterioration has occurred in the inverter 43 if the sideband wave b exists in the high frequency region. Based on such a determination method, the deterioration estimation unit 150 sets information indicating either the motor 21 or the inverter 43 to the deteriorated portion W_point, and outputs the information to the deterioration notification unit 160 and the control unit 130.

[0116] (Deterioration notification section 160) The deterioration notification unit 160 causes the user terminal 200 to display a notification screen indicating whether deterioration has occurred in the electric motor 21 or the inverter 43, based on the deterioration point W_point.

[0117] (Control unit 130) The control unit 130 accumulates the drive time D_time of the compressor 2. The drive time D_time is obtained by counting the actual drive time of the compressor 2 based on, for example, the gate pulse GP generated by the control unit 130 or the current Iuv of two phases of the three-phase AC. The control unit 130 outputs the drive time D_time to the operation determination unit 170.

[0118] Furthermore, the control unit 130 executes normal operation, or low-noise pulse operation based on the synchronous PWM control described in the first embodiment, or low-noise pulse operation based on the AI ​​switching control described in the second and third embodiments. The operation mode is determined by an operation mode switching signal O_s received from the operation determination unit 170. The operation mode switching signal O_s indicates any one of "1" indicating normal operation, "2" indicating low-noise pulse operation based on synchronous PWM control, and "3" indicating low-noise pulse operation based on AI switching control.

[0119] (Driving decision unit 170) The operation determination unit 170 switches the operation mode switching signal O_s to one of "1", "2" or "3" based on the estimated deterioration degree W_est, the drive time D_time, the low noise pulse operation start signal L_s, and the deterioration point W_point, and outputs the signal to the control unit 130. Specifically, when the following deterioration estimation conditions are satisfied and the "ON" signal of the low noise pulse operation start signal L_s is output, the operation determination unit 170 transmits the "2" signal or the "3" signal of the operation mode switching signal O_s to the control unit 130. The first deterioration estimation condition is when the estimated deterioration degree W_est is equal to or greater than the reference deterioration degree x. The second deterioration estimation condition is when the drive time D_time is the reference drive time y. The reference drive time y is determined in advance by experiments or the like so that the time until the compressor 2, whose cumulative drive time is the reference drive time y, breaks down is a predetermined time. The predetermined time is, for example, a time that is considered to be sufficient for a user to arrange for repairs or the like. When the deterioration estimation condition is satisfied, the operation determination unit 170 notifies the manager, regardless of the low-noise pulse operation start signal L_s.

[0120] Furthermore, if the estimated object indicated by the deterioration point W_point is the electric motor 21, the operation determination unit 170 transmits a signal of "2" of the operation mode switching signal O_s to the control unit 130. In response to the input of the signal of "2" of the operation mode switching signal O_s, the control unit 130 executes low-noise pulse operation based on synchronous PWM control. On the other hand, if the estimated object indicated by the deterioration point W_point is the inverter 43, the operation determination unit 170 transmits a signal of "3" of the operation mode switching signal O_s to the control unit 130. In response to the input of the signal of "3" of the operation mode switching signal O_s, the control unit 130 executes low-noise pulse operation based on AI switching control.

[0121] When the deterioration estimation condition is not satisfied or the "OFF" signal of the low-noise pulse operation start signal L_s is output, the operation determination unit 170 transmits a "1" signal of the operation mode switching signal O_s to the control unit 130. When the "1" signal of the operation mode switching signal O_s is input, the control unit 130 executes normal operation.

[0122] 24 is a flowchart showing the operation of the driving determination unit 170 according to the fourth embodiment. The process procedure described below is an example of a driving mode determination method. Each process may be changed as much as possible, and processes may be omitted, replaced, or added as appropriate.

[0123] First, the operation determination unit 170 determines whether the deterioration estimation condition is satisfied (step S51). If the deterioration estimation condition is not satisfied (step S51: NO), the operation determination unit 170 transmits a signal of "1" of the operation mode switching signal O_s indicating normal operation to the control unit 130 (step S52). If the deterioration estimation condition is satisfied (step S51: YES), the operation determination unit 170 notifies the administrator (step S53) and determines whether the low noise pulse operation start signal L_s is "ON" (step S54). If the low noise pulse operation start signal L_s is "OFF" (step S54: NO), the operation determination unit 170 transmits a signal of "1" of the operation mode switching signal O_s indicating normal operation to the control unit 130 (step S52).

[0124] When the low-noise pulse operation start signal L_s is "ON" (step S54: YES), the operation determination unit 170 determines the estimation target indicated by the deterioration point W_point (step S55). When the deterioration point W_point indicates the motor 21 (step S55: "motor 21"), a signal of "2" of the operation mode switching signal O_s indicating low-noise pulse operation based on synchronous PWM control is sent to the control unit 130 (step S56). When the deterioration point W_point indicates the inverter 43 (step S55: "inverter 43"), a signal of "3" of the operation mode switching signal O_s indicating low-noise pulse operation based on AI switching control is sent to the control unit 130 (step S57).

[0125] Similarly to the first to third embodiments, the motor control device 1 of the fourth embodiment also performs a low-noise pulse operation in which the switching pattern of the inverter 43 is adjusted so that the switching loss is reduced compared to the normal operation. This suppresses the progression of deterioration of the inverter 43 and the motor 21. This suppresses the occurrence of unanticipated downtime for the user in the device in which the motor 21 is mounted until the scheduled maintenance time.

[0126] Moreover, the motor control device 1 of the fourth embodiment can determine the deteriorated part based on the current output to the motor 21. Furthermore, the motor control device 1 of the fourth embodiment switches between low noise pulse operation based on synchronous PWM control and low noise pulse operation based on AI switching control for each deteriorated part. Here, the synchronous PWM control determines the switching according to the fundamental wave frequency (driving rotation speed), so it is suitable for reducing noise related to the driving of the motor 21. On the other hand, the AI ​​switching control also takes into account the carrier frequency, that is, the sideband wave of a frequency higher than the fundamental wave frequency (driving rotation speed), so it is suitable for reducing noise related to the inverter 43. In this way, the low noise pulse operation suitable for the deteriorated part is executed, so that the progress of deterioration of the inverter 43 and the motor 21 is further suppressed. Therefore, the occurrence of unexpected downtime for the user until the planned maintenance time is reached in the equipment equipped with the motor 21 is further suppressed.

[0127] Embodiment 5. Fig. 25 is a schematic configuration diagram showing a compressor 2, a power source 3, and a power conversion device 4 according to a fifth embodiment. The fifth embodiment differs from the first to fourth embodiments in that an operation determination unit 170 and a deterioration notification unit 160 are provided outside a control unit 10 consisting of a control unit 130, a deterioration estimation unit 150, a power source 3, and a power conversion device 4, and an electric motor control system 1A is configured to execute the method for controlling the electric motor 21 described in the first to fourth embodiments. In the fifth embodiment, the same parts as those in the first to fourth embodiments are denoted by the same reference numerals and description thereof is omitted, and the differences from the first to fourth embodiments will be mainly described.

[0128] (Control unit 10) The control unit 130 and the deterioration estimation unit 150 of the motor control device 1, the power source 3, and the power conversion device 4 are installed in a control box as the control unit 10. In the fifth embodiment, an air conditioner in which the compressor 2 is provided is described as being installed in a facility such as a building. The control unit 10 is installed outdoors of the facility, such as on the roof of a building. The deterioration estimation unit 150 of the control unit 10 transmits the estimated deterioration degree W_est to the deterioration notification unit 160 and the operation determination unit 170. In addition, the control unit 130 of the control unit 10 receives an operation mode switching signal O_s from the operation determination unit 170.

[0129] (Deterioration notification section 160) The deterioration notification unit 160 is configured by software installed in a control device of the management panel 400 as one functional unit of the management panel 400 installed indoors in a facility such as a management room. The management panel 400 is provided with a display and operation buttons, and a user can check and input settings for the operation of the air conditioner. The management panel 400 may have a touch-type display instead of a combination of a display and operation buttons. The deterioration notification unit 160 communicates with the control unit 10 wirelessly or wired and receives the estimated deterioration degree W_est from the deterioration estimation unit 150 of the control unit 10. The deterioration notification unit 160 displays a notification screen on the display indicating that deterioration has occurred in the equipment equipped with the electric motor 21 based on the received estimated deterioration degree W_est. The display of the management panel 400 is an example of a "display device" in the present disclosure.

[0130] (Driving decision unit 170) The driving determination unit 170 is realized as one functional unit of the server device by software or the like installed in the server device. The driving determination unit 170 communicates with the control unit 10 using the Internet or the like, and receives the estimated deterioration degree W_est from the deterioration estimation unit 150 of the control unit 10. The driving determination unit 170 also transmits an driving mode switching signal O_s to the control unit 10. The specific operation is similar to that described in the first to fourth embodiments. The driving determination unit 170 may be realized as a functional unit on a cloud server, rather than as a functional unit on a physical server.

[0131] Similarly to the first to fourth embodiments, the motor control system 1A of the fifth embodiment performs a low-noise pulse operation in which the switching pattern of the inverter 43 is adjusted so that the switching loss is reduced as compared to the normal operation. This suppresses the progression of deterioration of the inverter 43 and the motor 21. This suppresses the occurrence of unanticipated downtime for the user in the device in which the motor 21 is mounted until the scheduled maintenance time.

[0132] Each embodiment disclosed herein should be considered to be illustrative and not restrictive in all respects. The scope of the present disclosure is indicated by the claims, not by the above description, and is intended to include all modifications within the meaning and scope of the claims. In addition, the embodiments and modifications described herein are intended to be implemented as far as possible, either alone or in combination. For example, each embodiment may indicate a general replacement period for the inverter 43 or the compressor 2, or a recommended replacement time based on the estimated deterioration degree W_est and the reference deterioration degree x, or a recommended replacement time based on the drive time D_time and the reference drive time y.

[0133] In each embodiment, a notification screen may be displayed on a mobile terminal carried by the user. In this case, the deterioration notification unit 160 is realized by dedicated software installed on the mobile terminal carried by the user. In this case, the user can know the deterioration of the inverter 43 and the electric motor 21 even from a remote location. This makes it possible to speed up maintenance response. The display of the mobile terminal corresponds to the "display device" of this disclosure.

[0134] Also, the method of obtaining the estimated deterioration degree W_est in the first embodiment may be modified as follows. FIG. 26 is a schematic configuration diagram showing the compressor 2, the power source 3, and the power conversion device 4 according to a modification of the first embodiment. As shown in FIG. 26, the power conversion device 4 is provided with a voltage sensor 6 that detects a bus voltage Vdc. The deterioration estimation unit 150 acquires the bus voltage detected by the voltage sensor 6. The deterioration estimation unit 150 acquires the refrigerant pressure Rp, the refrigerant flow rate Rfr, the ambient temperature Tmp, and the ambient humidity Hud from the pressure sensor 71, the flow rate sensor 72, the temperature sensor 73, and the humidity sensor 74 described in the first embodiment. The deterioration estimation unit 150 may obtain the estimated deterioration degree W_est based on the bus voltage Vdc, the refrigerant pressure Rp, the refrigerant flow rate Rfr, the ambient temperature Tmp, and the ambient humidity Hud. Although not shown in the figure, the same applies not only to the first embodiment but also to the second to fifth embodiments.

[0135] Furthermore, in the second embodiment, a trained model PGF that achieves another objective in addition to reducing switching loss may be generated based on the current Iuv, the bus voltage Vdc, the refrigerant pressure Rp, the refrigerant flow rate Rfr, the ambient temperature Tmp, and the ambient humidity Hud. The other objective may be, for example, any one of the driving noise, vibration, and current harmonics of the compressor 2, and the voltage applied to the inverter 43. In this case, the current Iuv, the bus voltage Vdc, the refrigerant pressure Rp, the refrigerant flow rate Rfr, the ambient temperature Tmp, and the ambient humidity Hud are used as input data to the trained model PGF.

[0136] In the fifth embodiment, the configuration in which the operation determination unit 170 and the deterioration notification unit 160 are provided outside the control unit 10 including the control unit 130, the deterioration estimation unit 150, the power source 3, and the power conversion device 4 is referred to as the motor control system 1A. However, the motor control system is not limited to the above configuration, and any of the deterioration estimation unit 150, the deterioration notification unit 160, and the operation determination unit 170 may be provided outside the control unit 10 including the control unit 130, the power source 3, and the power conversion device 4. In this case, the deterioration estimation unit 150, the deterioration notification unit 160, and the operation determination unit 170 are realized by software or the like as hardware or a CPU different from the motor control device in which the control unit 130 is implemented, or as one functional unit of a server device such as a cloud server. In addition, the features of the motor control device 1 described in the first to fourth embodiments can be combined with the motor control system.

[0137] In each embodiment, the operation determination unit 170 may change the content of the notification to the management terminal 300 according to the degree to which the estimated deterioration degree exceeds the reference deterioration degree. Similarly, the deterioration notification unit 160 may change the notification screen to be displayed on the display device. For example, the more the estimated deterioration degree exceeds the reference deterioration degree, the more frequent the notification may be, or the notification interval may be shortened. Furthermore, the wording to be displayed on the management terminal 300 or the display device may be changed to create a sense of crisis about the deterioration of the equipment in which the electric motor 21 is mounted.

[0138] In addition, in the second to fifth embodiments, the learning unit 141 of the control unit 130 includes a data acquisition unit 141a that acquires learning data including a voltage command value in an arbitrary cycle and the switching pattern in a cycle one cycle before the arbitrary cycle, and a model generation unit 141b that generates a learned model for inferring the switching pattern in the arbitrary cycle using the learning data. However, the data acquisition unit 141a and the model generation unit 141b may be omitted from the learning unit 141. Specifically, the motor control device 1 or the motor control system 1A may be a model in which a learned model is stored in advance in the storage unit 180. The learned model stored in the storage unit 180 is generated using learning data including the voltage command value in an arbitrary cycle and the switching pattern in a cycle one cycle before the arbitrary cycle, as in the second to fifth embodiments, and is for inferring the switching pattern in the arbitrary cycle. The trained model stored in the storage unit 180 is trained, for example, by a computer provided outside the motor control device 1 or the motor control system 1A. [Explanation of symbols]

[0139] 1 motor control device, 1A motor control system, 2 compressor, 3 power supply, 4 power conversion device, 5 current sensor, 6 voltage sensor, 10 control unit, 21 motor, 21a rotor core, 21b stator core, 22 suction pipe, 23 main shaft, 24 oil pump, 25 auxiliary bearing, 26 main bearing, 27 compression mechanism, 28 discharge pipe, 41 rectifier circuit, 42 electrolytic capacitor, 43 inverter, 43a switching element, 43b backflow prevention element, 44 wiring, 71 pressure sensor, 72 flow rate sensor, 73 temperature sensor, 74 humidity sensor, 80 bus voltage sensor, 130 control unit, 131 vector control unit, 132 synchronization pattern selection unit, 133 voltage phase calculation unit, 134 correction amount calculation unit, 135 carrier wave generation unit, 136 gate pulse generation unit, 137 State observation unit, 141 learning unit, 141a data acquisition unit, 141b model generation unit, 141c reward calculation unit, 141d function update unit, 142 inference unit, 150 deterioration estimation unit, 160 deterioration notification unit, 170 operation judgment unit, 180 memory unit, 200 user terminal, 300 management terminal, 400 management panel.

Claims

1. A motor control device that controls a motor that operates using power supplied from a power source and converted using a power conversion device having an inverter, a deterioration estimation unit that calculates an estimated deterioration degree by estimating a degree of deterioration of a device in which the electric motor is mounted or the inverter; an operation determination unit that compares the estimated deterioration degree with a predetermined reference deterioration degree, and sets an operation mode of the electric motor to a normal operation when the estimated deterioration degree is less than the reference deterioration degree, and sets an operation mode of the electric motor to a low noise pulse operation when the estimated deterioration degree is equal to or greater than the reference deterioration degree; A control unit that controls the inverter based on the operation mode, The low noise pulse operation is The switching loss is reduced compared to the normal operation, The deterioration estimation unit is The estimated deterioration degree is calculated based on the intensity of sideband waves appearing in two phases of the three-phase AC current flowing from the inverter to the motor. Motor control device.

2. The control unit controls the inverter by changing a switching pattern, which is a combination of switching states for one cycle of a voltage command value that commands an output voltage of the inverter; The low noise pulse operation is and adjusting the switching pattern. The motor control device according to claim 1 .

3. The control unit performs the low noise pulse operation based on synchronous PWM control in which a frequency of a carrier signal is an integer multiple of a frequency of an output voltage of the inverter, In the synchronous PWM control, the greater the speed command value that commands the rotation speed of the motor, the less the number of switching operations. The motor control device according to claim 2.

4. The control unit is a data acquisition unit that acquires learning data including the voltage command value in an arbitrary cycle and the switching pattern in a cycle one cycle before the arbitrary cycle; A model generation unit that generates a trained model for inferring the switching pattern of the arbitrary cycle using the training data. The motor control device according to claim 2 or 3.

5. The switching pattern control circuit further includes a storage unit that stores a learned model for inferring the switching pattern of an arbitrary cycle, the learned model being generated using learning data including the voltage command value in the arbitrary cycle and the switching pattern in a cycle one cycle before the arbitrary cycle. The motor control device according to claim 2 or 3.

6. The control unit is a gate pulse generating unit that generates the switching pattern at a control time point based on the learned model, using as input the switching pattern generated one cycle before the control time point and the voltage command value at the control time point. The motor control device according to claim 4 or 5.

7. The learned model determines the switching pattern adjusted so that, in the adjustment to reduce switching loss compared to the normal operation, at least one of driving noise, vibration, current harmonics, and voltage applied to the inverter of the device in which the electric motor is mounted is reduced compared to the normal operation. The motor control device according to any one of claims 4 to 6.

8. The control unit is A reward calculation unit that calculates a reward based on the intensity of the sideband wave centered on the fundamental frequency of the current flowing through the motor or the number of switching operations of the inverter; A function update unit that updates the trained model based on the reward input from the reward calculation unit. The motor control device according to any one of claims 4 to 7.

9. The reward calculation unit increases the reward when the number of switching counts in the switching pattern is decreased compared to the switching pattern based on the trained model before the update, and decreases the reward when the number of switching counts is increased compared to the switching pattern based on the trained model before the update. The motor control device according to claim 8.

10. The reward calculation unit increases the reward when the strength of the sideband wave is equal to or less than a predetermined value in the switching pattern, and decreases the reward when the strength of the sideband wave exceeds a predetermined value. The motor control device according to claim 8 or 9.

11. The deterioration estimation unit is The estimated deterioration degree is calculated based on a current flowing between the inverter and the device in which the electric motor is mounted. The motor control device according to any one of claims 1 to 10.

12. a degradation notification unit that displays, on a display device, a notification screen indicating that degradation has occurred in the device equipped with the electric motor when the estimated degradation level reaches the reference degradation level or when a drive time of the device equipped with the electric motor reaches a predetermined reference drive time; The operation determination unit sets the low-noise pulse operation based on a low-noise pulse operation start signal received from the deterioration notification unit and corresponding to a user's consent to the start of the low-noise pulse operation. The motor control device according to any one of claims 1 to 11.

13. The deterioration estimation unit is estimating a deterioration location indicating whether the motor or the inverter is deteriorated based on a current flowing between the inverter and the device in which the motor is mounted; The degradation notification unit displays the deteriorated portion on the display device. The motor control device according to claim 12.

14. The deterioration estimation unit is estimating a deterioration location indicating whether the motor or the inverter is deteriorated based on a current flowing between the inverter and the device in which the motor is mounted; The operation determination unit sets the operation mode to the low-noise pulse operation based on the synchronous PWM control or the low-noise pulse operation based on the AI ​​switching control using the learned model, based on the deteriorated location.

6. The motor control device according to claim 4 or 5 dependent on claim 3.

15. The device is a compressor having a main shaft connected to the electric motor and transmitting rotational energy to a compression mechanism, the main shaft being rotatably supported by a main bearing, The deterioration estimation unit obtains the estimated deterioration degree by analyzing an intensity of the sideband wave of a fundamental frequency that is generated due to deterioration of the main bearing. The motor control device according to any one of claims 1 to 14.

16. A motor control device that controls a motor operated by power supplied from a power source and converted using a power conversion device having an inverter, a deterioration estimation unit that calculates an estimated deterioration degree by estimating a degree of deterioration of a device in which the electric motor is mounted or the inverter; an operation determination unit that compares the estimated deterioration degree with a predetermined reference deterioration degree, and sets an operation mode of the electric motor to a normal operation when the estimated deterioration degree is less than the reference deterioration degree, and sets an operation mode of the electric motor to a low noise pulse operation when the estimated deterioration degree is equal to or greater than the reference deterioration degree; a control unit that controls the inverter by changing a switching pattern, which is a combination of switching states for one cycle of a voltage command value that commands an output voltage of the inverter, based on the operation mode; The low noise pulse operation is By adjusting the switching pattern, switching loss is reduced compared to the normal operation, The control unit is a data acquisition unit that acquires learning data including the voltage command value in an arbitrary cycle and the switching pattern in a cycle one cycle before the arbitrary cycle; a model generation unit that generates a trained model for inferring the switching pattern of the arbitrary cycle by using the training data; A reward calculation unit that calculates a reward based on the intensity of a sideband wave centered on a fundamental frequency of the current flowing through the motor or the number of switching operations of the inverter; a function update unit that updates the trained model based on the reward input from the reward calculation unit; The reward calculation unit increases the reward when the number of switching counts in the switching pattern is decreased compared to the switching pattern based on the trained model before the update, and decreases the reward when the number of switching counts is increased compared to the switching pattern based on the trained model before the update. Motor control device.

17. A motor control device that controls a motor operated by power supplied from a power source and converted using a power conversion device having an inverter, a deterioration estimation unit that calculates an estimated deterioration degree by estimating a degree of deterioration of a device in which the electric motor is mounted or the inverter; an operation determination unit that compares the estimated deterioration degree with a predetermined reference deterioration degree, and sets an operation mode of the electric motor to a normal operation when the estimated deterioration degree is less than the reference deterioration degree, and sets an operation mode of the electric motor to a low noise pulse operation when the estimated deterioration degree is equal to or greater than the reference deterioration degree; a control unit that controls the inverter by changing a switching pattern, which is a combination of switching states for one cycle of a voltage command value that commands an output voltage of the inverter, based on the operation mode; The low noise pulse operation is By adjusting the switching pattern, switching loss is reduced compared to the normal operation, The control unit is a data acquisition unit that acquires learning data including the voltage command value in an arbitrary cycle and the switching pattern in a cycle one cycle before the arbitrary cycle; a model generation unit that generates a trained model for inferring the switching pattern of the arbitrary cycle by using the training data; A reward calculation unit that calculates a reward based on the intensity of a sideband wave centered on a fundamental frequency of the current flowing through the motor or the number of switching operations of the inverter; a function update unit that updates the trained model based on the reward input from the reward calculation unit; The reward calculation unit increases the reward when the strength of the sideband wave is equal to or less than a predetermined value in the switching pattern, and decreases the reward when the strength of the sideband wave exceeds a predetermined value. Motor control device.

18. A motor control system for controlling a motor operated by power supplied from a power source and converted using a power conversion device having an inverter, a deterioration estimation unit that calculates an estimated deterioration degree by estimating a degree of deterioration of a device in which the electric motor is mounted or the inverter; an operation determination unit that compares the estimated deterioration degree with a predetermined reference deterioration degree, and sets an operation mode of the electric motor to a normal operation when the estimated deterioration degree is less than the reference deterioration degree, and sets an operation mode of the electric motor to a low noise pulse operation when the estimated deterioration degree is equal to or greater than the reference deterioration degree; A control unit that controls the inverter based on the operation mode, The low noise pulse operation is The switching loss is reduced compared to the normal operation, The deterioration estimation unit is The estimated deterioration degree is calculated based on the intensity of sideband waves appearing in two phases of the three-phase AC current flowing from the inverter to the motor. Motor control system.

19. A method for controlling an electric motor that operates using power supplied from a power source and converted using a power conversion device having an inverter, comprising: The method for controlling the electric motor comprises: Calculating an estimated deterioration degree by estimating a degree of deterioration of a device in which the electric motor is mounted or of the inverter; comparing the estimated deterioration degree with a predetermined reference deterioration degree, and setting an operation mode of the electric motor to a normal operation when the estimated deterioration degree is less than the reference deterioration degree, and to a low noise pulse operation when the estimated deterioration degree is equal to or greater than the reference deterioration degree; The inverter is controlled based on the operation mode. The low noise pulse operation is The switching loss is reduced compared to the normal operation, The estimated deterioration degree is calculated based on the intensity of sideband waves appearing in two phases of the three-phase AC current flowing from the inverter to the motor. A method for controlling an electric motor.

Images