JP7467301B2 - Open winding motor drive device and refrigeration cycle device - Google Patents

Description

An embodiment of the present invention relates to a device for driving a motor with an open winding structure, and a refrigeration cycle device equipped with the device.

For example, when driving an AC motor such as a permanent magnet synchronous motor, it is necessary to use an inverter to convert the DC power source into three-phase AC power. However, as the capacity of the motor increases, the current flowing through the inverter also increases, causing problems such as heat generation in the power devices that make up the inverter.

To address this issue, for example, Patent Document 1 proposes a system in which the windings of a three-phase motor are left open without being star-connected, and inverters are connected to both ends of the three-phase windings to drive them. With this system, by using two inverters, the voltage that can be applied to both ends of the three-phase windings can be expanded by about twice, allowing the motor to be driven at higher speeds. Alternatively, by increasing the number of turns in the windings, a motor that outputs high torque with less current can be driven.

Patent Document 1 also proposes a technique for suppressing the zero-axis current that flows in common to the three phases of the motor, which occurs when a DC link voltage is shared between two inverters.

JP 2020-31458 A

However, as will be described in detail later, in the configuration of Patent Document 1, when the output voltage of a normal inverter is taken as "1", the output can only be up to √3 times that value.
Therefore, the present invention provides an open winding motor drive device capable of suppressing the generation of zero axis current and increasing the output voltage, and a refrigeration cycle device including the same.

The open winding motor drive device of the embodiment includes a primary side inverter connected to three of six terminals of a motor having an open winding structure in which three phase windings are independent from each other and the motor has six terminals ;
a secondary side inverter connected to the remaining three terminals of the motor;
a control unit that controls a current supplied to the motor and a rotation speed based on a line duty ratio of each of the primary side and secondary side inverters in PWM control;
a current detector for detecting a current flowing through the motor,
the control unit controls a current and a rotation speed of the motor based on the line-to-line duties of the primary side and secondary side inverters, and has a zero-axis current suppression unit that suppresses a zero-axis current flowing in the same direction through three phases between the primary side and secondary side inverters,
The zero-axis current suppression unit is configured to suppress the zero-axis current from being applied to a space voltage vector consisting of 64 voltage vectors which are combinations of on/off patterns of the primary side and secondary side inverters.
The six regions are divided into twelve sectors, each of which is defined by a point at which two second switching patterns that do not generate a zero-axis voltage and do not generate a voltage acting between the phases of the motor are located at the center and a point at which two first switching patterns that do not generate a zero-axis voltage that acts equally on the three phases of the motor and generate a voltage to be applied to the motor are located at the apex, and the six regions are further divided into twelve sectors, each of which is defined by a point at which two first switching patterns that do not generate a zero-axis voltage that acts equally on the three phases of the motor and generate a voltage to be applied to the motor are located at the center. Only one first switching pattern is selected for use in each sector during a control period.
In addition, the refrigeration cycle device of the embodiment includes a motor having an open winding structure in which three-phase windings are independent of each other and have six winding terminals;
The present invention also includes an open winding motor drive device according to an embodiment.

FIG. 1 is a diagram illustrating a configuration of a motor drive system according to a first embodiment. Functional block diagram showing the configuration of the current control unit Functional block diagram showing the configuration of an air conditioner A diagram showing the waveforms of the motor U-phase current and zero-axis current. FIG. 5 is an enlarged view of a portion of FIG. FIG. 1 shows the change in zero-axis voltage caused by switching of the primary and secondary inverters. A diagram showing space voltage vectors corresponding to a configuration for driving a typical three-phase motor. FIG. 1 shows space voltage vectors corresponding to a configuration for driving an open-winding motor. FIG. 1 is a diagram showing the first and second vector patterns used in each sector, with the space voltage vectors divided into six sectors. A diagram showing a first vector pattern used in each sector by dividing the space voltage vector into 12 sectors. A time series diagram of the first vector patterns used in each sector (part 1) A time series of the first vector patterns used in each sector (part 2) FIG. 13 is a diagram showing a configuration of a motor drive system according to a second embodiment. Functional block diagram showing the configuration of the zero-axis voltage generating unit FIG. 13 shows a vector pattern used to reduce the zero-axis current in sector 6. FIG. 13 shows a vector pattern used when increasing the zero-axis current in sector 6. FIG. 13 is a diagram showing a vector pattern used to reduce the zero-axis current in the sector 7. FIG. 13 is a diagram showing a vector pattern used when increasing the zero-axis current in the sector 7. Functional block diagram showing the configuration of a space voltage vector modulation unit FIG. 13 is a diagram showing waveforms of each phase current and a zero-axis current of a motor under control of the first embodiment; FIG. 13 is a diagram showing waveforms of each phase current and a zero-axis current of a motor in the control of the second embodiment. FIG. 13 is a diagram showing a configuration of a motor drive system according to a third embodiment. Functional block diagram showing the configuration of a dq current control unit Flowchart showing the process of switching from normal pulse control to synchronous pulse control Flowchart showing switching process from synchronous pulse control to normal pulse control A diagram conceptually illustrating the processing in steps S5 to S7. A diagram conceptually illustrating the process of increasing the voltage amplitude from an initial value to a target value. A diagram conceptually illustrating a process of switching from a space voltage vector used for normal pulse control to a space voltage vector used for synchronous pulse control. A diagram showing the simulation results when transitioning from normal pulse control to synchronous pulse control to normal pulse control.

First Embodiment
A first embodiment will be described below with reference to Figs. 1 to 12. This embodiment is based on the technology disclosed in Patent Document 1, with some improvements thereon. Fig. 1 is a diagram showing the circuit configuration of a motor drive system of this embodiment. The motor M is assumed to be a three-phase permanent magnet synchronous motor or an induction machine, but in this embodiment, a permanent magnet synchronous motor is used. The three-phase windings of the motor M are not connected to each other, and both terminals are in an open state. In other words, the motor M has six winding terminals Ua, Va, Wa, Ub, Vb, and Wb.

The primary inverter 1 and the secondary inverter 2 are each configured with a three-phase bridge connection of N-channel MOSFETs 3, which are switching elements, and are connected in parallel to a DC power supply 4. The DC power supply 4 may be an AC power supply converted to DC. The output terminals of each phase of the inverter 1 are connected to the winding terminals Ua, Va, and Wa of the motor M, respectively, and the output terminals of each phase of the inverter 2 are connected to the winding terminals Ub, Vb, and Wb of the motor M, respectively.

The position sensor 6 is a sensor that detects the rotor rotation position and rotation speed of the motor M, and the current sensors 7 (U, V, W) are sensors that detect the respective phase currents Iu, Iv, Iw of the motor M and correspond to current detectors. The voltage sensor 8 detects the voltage _VDC of the DC power supply 4.

The control device 11 is provided with a speed command value ω _Ref from a higher-level control device in the system that drives the motor, and performs control so that the detected motor speed ω coincides with the speed command value ω _Ref . The control device 11 generates switching signals to be provided to the gates of the FETs 3 that constitute the inverters 1 and 2, based on the phase currents Iu, Iv, Iw detected by the current sensor 7 and the DC voltage V _DC detected by the voltage sensor 8. The control device 11 corresponds to a control unit.

The current detection and coordinate conversion unit 12 converts the detected phase currents Iu, Iv, and Iw into currents Id and Iq in the d- and q-axis coordinates used for vector control using equation (1). Note that the conversion of the zero-axis current I0 shown in equation (1) is performed by the current detection and coordinate conversion unit 43 in the second embodiment.

The speed/position detection unit 13 detects the motor speed ω and the rotor rotational position θ from the signal detected by the position sensor 6. The rotational position θ is input to the current detection/coordinate conversion unit 12 and the dq/αβ conversion unit 17. The speed/position detection unit 13 may also be configured to estimate the speed and position from the voltage and current of the motor M. The speed control unit 14 generates and outputs a q-axis current command _IqRef by, for example, performing a PI calculation on the difference between the input speed command _ωRef and the speed ω.

The current control unit 16 generates and outputs d, q-axis voltage commands Vq, Vd from the input q-axis current command _IqRef , _DC voltage VDC, and the detected currents Id, Iq and motor speed ω. The dq/αβ conversion unit 17 converts the dq-axis voltage commands Vq, Vd into αβ-axis voltages Vα, Vβ using equation (2).

The space vector modulation unit 18 performs space vector calculations from the αβ-axis voltages Vα and Vβ to generate respective phase duties _Du1 , _Dv1 , and _Dw1 of inverter 1 and respective phase duties _Du2 , _Dv2 , and _Dw2 of inverter 2, and inputs them to the PWM signal generation unit 19. The PWM signal generation unit 19 generates and outputs switching signals, PWM signals U1±, V1±, W1±, U2±, V2±, and W2± to be given to the gates of the FETs 3 constituting the inverters 1 and 2, from the respective phase duties input thereto.

In this embodiment, the modulation factor of the inverters 1 and 2 is 1.0. Fig. 2 shows a detailed configuration of the current control unit 16. A subtractor 21 obtains the difference between the q-axis current command _IqRef obtained as a result of the speed control and the q-axis current Iq, and a PI control unit 22 performs a PI control calculation on the difference. The calculation result is output as a voltage phase _{θV_PI} . A non-interference term calculation unit 23 calculates a non-interference control term _{θV_FF} of the voltage phase from the motor constant, the d-axis current command _IdRef set equal to the d-axis current Id, the q-axis current command _IqRef , and the speed ω.

Furthermore, an adder 24 adds a non-interference control term θ _{V_FF} to the voltage phase θ _{V_PI} to generate a final voltage phase θ _V. The d, q-axis voltage commands Vd, Vq are generated by a dq-axis voltage calculation unit 25 from an output voltage amplitude V _amp corresponding to a DC voltage V _DC and a voltage phase θ _V. This makes it possible to directly apply a DC voltage to the motor M, and when the load changes, the motor M can be controlled at high speed by adjusting the voltage phase.

Figure 3 shows the configuration of an air conditioner 30 to which the motor drive system of this embodiment is applied. The compressor 32 constituting the heat pump system 31 is configured by housing the compression section 33 and the motor M in the same iron sealed container 35, and the rotor shaft of the motor M is connected to the compression section 33. The compressor 32, four-way valve 36, indoor heat exchanger 37, pressure reducing device 38, and outdoor heat exchanger 39 are connected to form a closed loop by pipes that serve as a heat transfer medium flow path. The compressor 32 is, for example, a rotary type compressor. The air conditioner 30 is configured with the above-mentioned heat pump system 31.

During heating, the four-way valve 36 is in the state shown by the solid line, and the high-temperature refrigerant compressed by the compression section 33 of the compressor 32 is supplied from the four-way valve 36 to the indoor heat exchanger 37, where it condenses, and then is decompressed by the pressure reducing device 38, becomes low temperature, flows to the outdoor heat exchanger 39, where it evaporates and returns to the compressor 32. On the other hand, during cooling, the four-way valve 36 is switched to the state shown by the dashed line. Therefore, the high-temperature refrigerant compressed by the compression section 33 of the compressor 32 is supplied from the four-way valve 6 to the outdoor heat exchanger 39, where it condenses, and then is decompressed by the pressure reducing device 38, becomes low temperature, flows to the indoor heat exchanger 37, where it evaporates and returns to the compressor 32. Then, fans 40 and 41 blow air to the indoor and outdoor heat exchangers 37 and 39, respectively, and the air blown is configured to efficiently exchange heat between the heat exchangers 37 and 39 and the indoor air and the outdoor air.

Next, the operation of this embodiment will be described with reference to Figures 4 to 12. To operate the open winding motor M, voltages are applied to the terminals Ua, Va, Wa, Ub, Vb, and Wb by the two inverters 1 and 2. The voltages obtained as a result of the speed control and current control are divided into voltage commands for the inverters 1 and 2 by the dq/αβ conversion unit 17, the space voltage vector modulation unit 18, and the PWM signal generation unit 19. The phase duties D _u1 , D _v1 , and D _w1 of the inverter 1 and the phase duties D _u2 , D _v2 , and D _w2 of the inverter 2 are energized as PWM signals having a phase difference of 180°. In this way, by applying voltages of opposite phases to the motor M by the two inverters 1 and 2, the voltage amplitude per phase can be increased, and the motor can be rotated at a higher speed.

As mentioned above, in a configuration in which inverters 1 and 2 share a DC link, the zero-axis current that flows in the same direction through the three phases becomes an issue. The zero-axis current is divided into a low-frequency current that flows with a frequency component three times the fundamental frequency of the phase current passed through motor M, and a current with a carrier frequency component that flows in synchronization with the switching of inverters 1 and 2. Figure 4 shows the U-phase current Iu and zero-axis current I0 of an open-winding motor that flows when there is no suppression control for the zero-axis current, and Figure 5 shows the current waveforms shown by enlarging the time axis of Figure 4. Ripples that change at the same timing can be confirmed in the U-phase current Iu and zero-axis current I0, which are the zero-axis current of the carrier frequency component. In addition, the zero-axis current I0 pulsates with a frequency component three times the fundamental frequency of the phase current, so the distortion of the U-phase current Iu is large.

V0_ripple is calculated by subtracting the average of the three-phase voltages of inverter 2 from the average of the three-phase voltages of inverter 1, as shown in equation (3). Note that each of the phase voltages Vu1, Vv1, Vw1, Vu2, Vv2, and Vw2 is V _DC when FET3 is on, and is 0 when FET3 is off.

As shown in Figure 6, the waveform of V0_ripple fluctuates between positive and negative depending on the switching state, with the zero-axis current I0 increasing during the period when it occurs on the positive side and decreasing during the period when it occurs on the negative side. Therefore, when V0_ripple becomes zero, the ripple of the zero-axis current I0, i.e., the fluctuation of the carrier frequency component, also disappears. Furthermore, the state in which V0_ripple occurs depends on the number of phases in which FET3 of inverters 1 and 2 is on, and when the number of on-phases of inverters 1 and 2 differs, it occurs in positive and negative directions according to the difference. In other words, if the number of on-phases of inverters 1 and 2 is the same, V0_ripple will not occur.

Here, we consider the space voltage vectors in order to consider the switching patterns of inverters 1 and 2 to achieve the above-mentioned objective. Figure 7 shows the space vectors when a typical motor is energized by a three-phase inverter. For example, V1 (100) indicates a state in which the upper arm of the U phase is on and the upper arms of the V and W phases are off, and there are eight vectors, V0 to V7.

In contrast, Figure 8 shows the spatial voltage vectors of an open winding motor, and since there are two inverters, there are 8 x 8 = 64 switching patterns. For example, the combination where inverter 1 is V1 and inverter 2 is V4 is represented as "V14". In the spatial vectors of an open winding motor, there are countless patterns of voltage vectors for outputting a certain command voltage. For example, to output the vector shown by the arrow in Figure 8, it is possible to output it by adjusting the energization time of any of the voltage vectors V06, V21, V30, V37, V45, and V76, and the energization time of any of the voltage vectors V01, V32, V40, V47, V56, and V71.

Now, considering the relationship between the zero-axis voltage and the space vector, out of the 64 patterns, there are 12 vector patterns that generate a voltage to be applied to the motor M and do not generate a zero-axis voltage that acts equally on the three phases, that is, patterns in which the number of on-phases is the same but at least two of the on-phases do not match: V15, V24, V26, V35, V31, V46, V42, V51, V53, V62, V64, and V13. These patterns are represented as space vectors in Figure 9.

In Fig. 9, the PWM waveforms corresponding to each voltage vector are also shown. The 12 patterns are arranged in pairs of two at the vertices to draw a regular hexagon, which is then divided into six sectors. For example, to output a vector belonging to sector 4, as indicated by the arrow in Fig. 9, the energization time of each of the voltage vectors V42 and V31 is adjusted. The PWM waveforms of each voltage vector are as follows:
V42: Inverter 1 (U, V, W) = (off, on, on)
Inverter 2 (U, V, W) = (on, on, off)
V31: Inverter 1 (U, V, W) = (off, on, off)
Inverter 2 (U, V, W) = (on, off, off)
To these are added V77, which turns on all phases of inverters 1 and 2, and V00, which turns off all phases. As can be seen from the PWM waveforms of each vector, the number of on-phases of inverters 1 and 2 is the same, so no zero-axis voltage V0 is generated. In other words, if current is applied using this PWM switching pattern, the ripple of the carrier component of the zero-axis current shown in Figures 4 and 5 can be suppressed.

9, the following first and second vector patterns are used: The first and second vector patterns correspond to the first and second switching patterns.
<First Vector Pattern>
A pattern that generates a voltage to be applied to the motor M and does not generate a zero-axis voltage that acts equally on the three phases. The 12 patterns of V15, V24, V26, V35, V31, V46, V42, V51, V53, V62, V64, and V13 above. <Second vector pattern>
A pattern in which no voltage acting between the phases of the motor M is generated, and no zero-axis voltage acting equally on the three phases is generated. V77 and V00 are second vector patterns in all sectors.

Figure 9 corresponds to Figure 8 of Patent Document 1, and the output sequence of the vector patterns in one control period is "second, first, first, second." However, in the PWM switching pattern shown in Figure 9, if the output of a normal inverter is 1, the driving range is within the inscribed circle of a regular hexagon connecting the points where each of the first switching patterns in Figure 9 is located, two by two. This means that the voltage can only be output up to √3 times that of a normal inverter.

Therefore, in this embodiment, the space voltage vector modulation unit 18 generates only one of the 12 first vector patterns during a control period. These patterns are represented as space vectors in FIG. 10. In FIG. 10, the six sectors shown in FIG. 9 are further divided into two equal parts, resulting in 12 sectors, and the voltage vector is switched every 30 electrical degrees depending on each sector. Note that the control periods shown in FIG. 9 and FIG. 10 are of equal length.

For example, to output a vector belonging to sector 7 indicated by an arrow in FIG. 10, only the voltage vector V42 is selected. The PWM waveform of the voltage vector is
V42: Inverter 1 (U, V, W) = (off, on, on)
Inverter 2 (U, V, W) = (on, on, off)
As a result, six points on the circumscribing circle of the regular hexagon shown in Fig. 10 become driving points, and since √3 x 2/√3 = 2, the output voltage can be doubled compared to a normal inverter. Furthermore, since the number of on-phases of inverters 1 and 2 is the same, the zero-axis voltage V0 is not generated, and the ripple of the carrier component of the zero-axis current can also be suppressed. The space voltage vector modulation unit 18 corresponds to the zero-axis current suppression unit.

Figure 11 shows the motor three-phase voltage waveforms corresponding to each voltage vector for one electrical angle revolution. The motor phase voltage waveform obtained by the switching pattern of inverters 1 and 2 is a 120-degree square wave voltage, which allows a DC voltage to be applied directly to the motor to achieve higher output compared to conventional pseudo-sine wave voltages. In addition, because the number of switching operations is small, the switching loss of FET3 can be reduced.

The switching pattern can also be simplified by partially changing it. For example, in Figure 12, the voltage vector of an odd sector is replaced with the voltage vector of the next even sector.

As described above, according to this embodiment, in a configuration in which a motor M with an open winding structure in which the three-phase windings are independent and which has six output terminals Ua to Wb is driven by a primary side inverter 1 and a secondary side inverter 2, a control device 11 controls the current and rotation speed of the motor M based on the line-to-line duties of the inverters 1 and 2, and is also provided with a space voltage vector modulation unit 18.

The spatial voltage vector modulation unit 18 divides the six regions, which are formed by combining the on-off patterns of inverters 1 and 2 and consist of 64 voltage vectors, into 12 sectors, each of which is divided by a point at which two second switching patterns are located, and a point at which two first switching patterns are located, into two equal halves, and selects only one first switching pattern to be used for each sector during the control period.

This suppresses the zero-axis current of the carrier component that flows in synchronization with the switching of the inverters 1 and 2, and doubles the output voltage to the motor M compared to a configuration in which a normal three-phase motor is driven by a single inverter. Furthermore, the space voltage vector modulation unit 18 selects the same first switching pattern for two adjacent sectors, making control easier. In addition, by applying the open winding motor drive device of this embodiment to the air conditioner 30, air conditioning operation can be performed with high output and high efficiency.

Second Embodiment
Hereinafter, the same parts as those in the first embodiment will be denoted by the same reference numerals and their explanation will be omitted, and only the different parts will be explained. First, the suppression of the zero-axis current that flows at the triple component of the fundamental frequency will be explained. Equation (4) is the relational expression between the dq0-axis voltage and the current of the open winding motor.

Here, we can see that when the d-axis and q-axis currents Id and Iq flow, a zero-axis voltage V0 is generated due to the influence of the diagonal elements shown in equation (4). This is interference from the d-axis and q-axis to the zero-axis, and as a result of the zero-axis voltage V0, a zero-axis current I0 flows.

The control device 42 of the second embodiment shown in FIG. 13 includes a current detection and coordinate conversion unit 43 replacing the current detection and coordinate conversion unit 12, a zero-axis voltage generation unit 44, and a space voltage vector modulation unit 45 replacing the space voltage vector modulation unit 18. As described above, the current detection and coordinate conversion unit 43 converts the detected phase currents Iu, Iv, and Iw into currents Id, Iq, and I0 of the d, q, and 0 axis coordinates used for vector control using equation (1).

14 shows the detailed configuration of the zero-axis voltage generator 44. The zero-axis voltage generator 44 includes a P controller 44A and a resonance controller 44B. The P controller 44A multiplies the difference between the zero-axis current command _I0Ref and the detected current I0 by a proportional gain _Kp0 . The resonance controller 44B is configured to improve the follow-up to the square of 3ω, which is three times the fundamental frequency ω of the phase current. The difference is multiplied by a resonance gain Kr, and a subtractor 50 takes the difference from the integration result of the integrator 46 and inputs it to the integrator 47.

The integration result of the integrator 47 is input to an adder 48 and a multiplier 49. The multiplier 49 multiplies the result by the frequency (3ω) ² , and the result is input to the integrator 46. The adder 48 adds the result of multiplying the difference value by a proportional gain _Kp0 , and outputs a value that is (1/√2) times the zero-phase voltage V0.

However, if only the spatial voltage vector pattern that does not generate the zero-axis voltage V0 shown in the first embodiment is used, it is not possible to generate the zero-axis voltage V0 required for control to suppress the zero-axis current I0 that flows due to interference, as described in relation to formula (4). Therefore, in the second embodiment, the following third vector pattern is additionally used. The third vector pattern corresponds to a third switching pattern.
<Third Vector Pattern>
A pattern that generates a voltage acting between the phases of the motor M and generates a zero-axis voltage that acts equally on all three phases.

Then, the third vector pattern is inserted before or after the first vector pattern used for each sector is output only once during the control period. This controls the zero-axis voltage V0 on average.

Here, a method of controlling the zero-axis voltage V0 in the case of an even-numbered sector will be described with reference to Fig. 15 and Fig. 16. For example, in sector 6, V31 is the first vector pattern. Then, as shown in Fig. 15, when reducing the zero-axis current I0 in sector 6, V01, which is a vector pattern for turning off the FET3 constituting the primary side inverter 1, is selected as the third vector pattern from the first vector pattern V31. When the third vector pattern is inserted after the first vector pattern, the switching pattern changes as follows.
Primary side inverter (UVW) (010) → (000)
Secondary side inverter (UVW) (100) → (100)
That is, in the third vector pattern V01, the number of phases in which the FET3 of the inverter 2 is turned on is one more than that of the inverter 1. This reduces the zero-axis current I0.

When the zero-axis voltage V0 is negative, six times the duty Vv calculated by equation (5) is added to the V-phase duty of the inverter 1. The reason for using six times the duty is to equalize the amount of control when switching the drive method from asynchronous PWM drive, as will be described in the third embodiment below. During asynchronous PWM drive, a zero-axis voltage of ±1/3 V _DC is applied three times. During synchronous PWM drive, a zero-axis voltage of ±1/3 V _DC is applied once in the even sectors. The double duty value is then tripled, resulting in six times the duty.
Here, the "double duty" refers to the method of generating PWM signal pulses. In asynchronous PWM drive, pulses are generated so that they expand and contract on both sides with the middle phase of the PWM cycle as the reference. In contrast, in synchronous PWM drive, pulses are generated so that they expand and contract on one side with one end of the PWM cycle as the reference. Therefore, the latter pulse width value to be added or subtracted is twice that of the former.

16, when the zero-axis current I0 is increased in the sector 6, the vector pattern V30 for turning off the FET 3 constituting the secondary-side inverter 2 is selected as the third vector pattern rather than the second vector pattern V31. When the third vector pattern is inserted after the first vector pattern, the switching pattern changes as follows.
Primary side inverter (UVW) (010) → (010)
Secondary side inverter (UVW) (100) → (000)
That is, in the third vector pattern V30, the number of phases in which the FET3 of the inverter 1 is turned on is one more than that of the inverter 2. This increases the zero-axis current I0. Note that, when the zero-axis voltage V0 is positive, a value six times the duty Vu calculated by the formula (5) is subtracted from the U-phase duty of the inverter 2.

Next, a method of controlling the zero-axis voltage V0 in the case of an odd-numbered sector will be described with reference to Fig. 17 and Fig. 18. For example, in sector 7, V42 is the first vector pattern. As shown in Fig. 17, when reducing the zero-axis current I0 in sector 7, vector pattern V02, which turns off FET3 constituting the primary side inverter 1, is selected as the third vector pattern from the first vector pattern V42. When the third vector pattern is inserted after the first vector pattern, the switching pattern changes as follows.
Primary side inverter (UVW) (011) → (000)
Secondary side inverter (UVW) (110) → (110)
That is, in the third vector pattern V02, the number of phases in which the FET3 of inverter 2 is turned on is two more than that of inverter 1. This reduces the zero-axis current I0. When the zero-axis voltage V0 is negative, six times the duty Vv, Vw calculated by equation (5) is added to the V- and W-phase duties of inverter 1. During asynchronous PWM driving, a zero-axis voltage of ±1/3 V _DC is applied three times. During synchronous PWM driving, a zero-axis voltage of ±2/3 V _DC is applied once in the odd sector. The magnitude is made the same as that of the even sector by V0/2, which will be described later.

18, when the zero-axis current I0 is increased in the sector 7, the vector pattern V40 for turning off the FET3 constituting the secondary-side inverter 2 is selected as the third vector pattern rather than the second vector pattern V42. When the third vector pattern is inserted after the first vector pattern, the switching pattern changes as follows.
Primary side inverter (UVW) (011) → (011)
Secondary side inverter (UVW) (110) → (000)

In this case, the third vector pattern V40 has two more phases in which FET3 of inverter 1 is turned on than inverter 2. This increases the zero-axis current I0. If the zero-axis voltage V0 is positive, the U- and V-phase duties of inverter 2 are reduced by six times the duties Vu and Vv calculated by equation (5).

By controlling in this way, the zero-axis voltage V0_ripple is generated continuously only on the positive and negative sides. Therefore, as shown in FIG. 6, ripples in the zero-axis current I0 caused by the positive and negative fluctuations of V0_ripple are not generated, and the triple frequency component can be suppressed. Note that a third vector pattern may be inserted before the first vector pattern during the control period.

Figure 19 shows the internal configuration of the space voltage vector modulation unit 45 based on the above-mentioned control principle, and includes a space vector calculation unit 45A and a zero-axis voltage synthesis unit 45B. The space vector calculation unit 45A determines which of the 12 sectors the input voltage commands Vα and Vβ belong to depending on their magnitude, and selects the first vector pattern depending on the sector. For example, if it is sector 7, it is V42.

The voltage value of the selected vector pattern and the zero-axis voltage V0 are input to the zero-axis voltage synthesis unit 45B together with the _DC voltage VDC. The zero-axis voltage synthesis unit 45B selects and inserts a third vector pattern according to the sector and the increase or decrease in the zero-axis current I0, as shown in Figs. 15 to 18. In the case of sector 7, the first vector pattern is V42, the third vector pattern to be inserted when the zero-axis current is decreased is V02, and the third vector pattern to be inserted when the zero-axis current is increased is V40.

In addition, in Figures 15 to 18, the voltage vector of the third vector pattern is inserted in only one place, but the magnitude of the zero-axis voltage generated in the even sector differs by two times from that in the odd sector. The zero-axis voltage generated by the third vector pattern needs to match the zero-axis voltage command V0 output from the zero-axis voltage generator 44. Therefore, in order to align the control amount for the zero-axis voltage command V0 for each sector, for example, the magnitude of the voltage vector of the third vector pattern in the odd sector is set to V0/2.

The magnitudes of the three-phase voltages of inverters 1 and 2 are obtained by the above calculations, and the duties Du1, Dv1, Dw1, Du2, Dv2, and Dw2 of each phase are determined by dividing them by the _DC voltage VDC and output.

Figures 20 and 21 show waveforms simulating the results of suppressing the zero-axis current by the control of the first and second embodiments, respectively. In Figure 21, it can be seen that the three-fold component of the fundamental frequency of the phase current is suppressed, compared to Figure 20.

As described above, according to the second embodiment, the space voltage vector modulation unit 45 inserts the third vector pattern before or after outputting only one vector pattern used for each sector during a control period. When generating a negative zero-axis voltage, a vector pattern that turns off all FETs 3 that constitute the inverter 1 based on the first vector pattern is selected as the third vector pattern, and when generating a positive zero-axis voltage, a vector pattern that turns off all FETs 3 that constitute the inverter 2 based on the first vector pattern is selected as the third vector pattern.

This makes it possible to suppress both the low-frequency zero-axis current that flows at three times the fundamental frequency of the phase current, and the zero-axis current of the carrier component that flows in synchronization with the switching of inverters 1 and 2, thereby realizing low current and low loss in inverters 1 and 2 and motor M.

Third Embodiment
The third embodiment relates to the switching control of the drive method. The first and second embodiments are particularly effective when the modulation factor is set to be greater than 0.866 in the region where the motor M is operated at high speed, and the output voltage is increased. Therefore, in the region where the motor M is operated at low speed, a different drive method, for example, the same drive method as that in Patent Document 1, is adopted. Hereinafter, the drive method of the first embodiment adopted in the high speed operation region is referred to as "synchronous pulse control", and the drive method adopted in the low speed operation region is referred to as "normal pulse control".

22 shows the configuration of a control device 51 when performing normal pulse control. Based on the configuration of the second embodiment, it includes a d-axis current command generating unit 52, a dq-axis current control unit 53 replacing the current control unit 16, and a space voltage vector modulation unit 54 replacing the space voltage vector modulation unit 45. The d-axis current command generating unit 52 generates and outputs a d-axis current command value for field weakening control by performing PI calculation on the difference between the DC voltage V _DC and the voltage amplitude V _dq of the dq axes.

23 shows a detailed configuration of the dq-axis current control unit 53. The dq-axis current control unit 53 includes PI control units 55d and 55q and a non-interference term calculation unit 56. The PI control unit 55d calculates a d-axis voltage _{Vd_PI} by PI control calculation from the difference between the d-axis current command value _IdRef and the d-axis current Id. Similarly, the PI control unit 55q calculates a q-axis voltage _{Vq_PI} by PI control calculation from the difference between the q-axis current command value _IqRef and the q-axis current Iq.

In order to prevent interference between the d-axis and q-axis, the non-interference term calculation unit 56 calculates non-interference terms _{Vd_FF} and _{Vq_FF} using the equations shown in the figure, and adds them to the d-axis voltage _{Vd_PI} and the q-axis voltage _{Vq_PI} by adders 58d and 58q, respectively. The sums of the adders 58d and 58q become the final d-axis voltage command value Vd and the q-axis voltage command value Vq.

The reason for separating the current control section configurations used for normal pulse control and synchronous pulse control will now be explained. Normal pulse control is a general vector control that directly generates the d-axis voltage Vd and q-axis voltage Vq by current control, and makes the current follow the current command value obtained from the results of field weakening control and speed control. This realizes motor control with good responsiveness in current phase lead control and motor output torque control. In other words, it applies current control that emphasizes controllability.

On the other hand, synchronous pulse control is used in the high output range of the motor, so compared to normal pulse control, it is necessary to advance the current phase further by setting the output voltage to the maximum DC voltage. In other words, it is control that emphasizes output. In this case, it is easier to control the motor by calculating the d-axis and q-axis voltages Vd and Vq from the voltage amplitude and voltage phase to command the output voltage. If the modulation rate is 1.0, the voltage amplitude will be input at its maximum value, and only the voltage phase will be controlled to control the q-axis current, which in turn controls the speed. As a result, field-weakening control is performed.

Next, the operation of the third embodiment will be described with reference to Fig. 24 to Fig. 29. As shown in Fig. 24, when a signal instructing switching from normal pulse control to synchronous pulse control, input from a higher-level control device, is turned ON (S1; YES), it is confirmed that the current control unit 16 in the synchronous pulse control is stopped (S2; YES), and then the voltage amplitude _Vdq and voltage phase θ in the normal pulse control at the present time are calculated (S3, S4). The voltage amplitude _Vdq is the square root of the sum of the squares of the d-axis voltage Vd and the q-axis voltage Vq, and the voltage phase θ is Atan (Vd/Vq).

Here, when switching from normal pulse control to synchronous pulse control, a modulation rate of 0.866 is set as the threshold, for example, and when it becomes necessary to set the modulation rate to a value greater than 0.866 due to the operating state of motor M, the above switching instruction signal is turned ON.

Next, the initial values V _amp and θv of the voltage amplitude and voltage phase when starting the synchronous pulse control are set (S5). The initial value V _amp of the voltage amplitude is set to the voltage amplitude V _dq , and the initial value θv of the voltage phase is set to the integral term of θ _{V_PI} output by the PI control unit 22 of the current control unit 16, which is a value obtained by subtracting the non-interference term θ _{V_FF} from the above-mentioned voltage phase θ. Note that the calculation of the non-interference term θ _{V_FF} starts at the time of switching. As a result, the initial value of the voltage phase θv is the voltage phase θ calculated from the voltages Vd and Vq before switching. This makes it possible to match the magnitudes of the voltages Vd and Vq before and after switching of the current control unit.

Next, the operation of the current control unit 16 used for the synchronous pulse control is started (S6), and the operation of the dq axis current control unit 53 is stopped only for the PI control unit 55 as shown in Fig. 26 (S7). Fig. 26 conceptually shows the processing of steps S5 to S7. The non-interference term calculation unit 56 is allowed to continue to operate because the calculation result is used when switching to the reverse direction, which will be described later. Then, it is determined whether the voltage amplitude has increased from the initial value _Vdq to ( _VDC × 2/√3), that is, to a phase voltage equivalent value of twice _{the DC} voltage VDC or more (S8, see Fig. 27).

If _Vamp has not reached ( _VDC x 2/√3) (S8; NO), the voltage amplitude is increased (S11). If _Vamp reaches ( _VDC x 2/√3) (S8; YES→S9), synchronous pulse control is performed using the space voltage vector as in the first embodiment (S10, see FIG. 28). That is, the space voltage vector modulation unit 54 is switched to the space voltage vector modulation unit 18.

Fig. 25 shows a process of switching from synchronous pulse control to normal pulse control, which is the opposite of Fig. 24. When a signal instructing switching from synchronous pulse control to normal pulse control is turned ON (S21; YES), the space voltage vector modulation unit 18 is switched to the space voltage vector modulation unit 54 (S22; YES -> S23). Then, a target value of the voltage amplitude is calculated (S24). The target value _Vdq is the square root of the sum of the squares of _{Vd_FF} and _{Vq_FF} calculated by the non-interference term calculation unit 56 in the normal pulse control.

Then, the output voltage _Vamp is gradually decreased from the maximum value ( _VDC x 2/√3) to the target value _Vdq (S25, S30). When the output voltage _Vamp reaches the target value _Vdq (S25; YES→S26), the operation of the dq-axis current control unit 53 is started (S28) and the operation of the current control unit 16 is stopped (S29). Prior to this, the initial values of the d-axis voltage Vd, the q-axis voltage Vq, and the d-axis current command _Idref are set (S27). As described above, since the non-interference term is input as the target value and the initial value, the initial value of the integral term of the PI control unit 55 is set to 0. In addition, since the d-axis current during the synchronous pulse control is overwritten as the d-axis current command _Idref , it is used as the initial value of the integral term of the field-weakening control unit when switching to the normal pulse control.

Figure 29 shows the simulation results when switching from normal pulse control to synchronous pulse control to normal pulse control. For a 6-pole motor, the PWM control carrier frequency is 5 kHz, the rotation speed is 80 rps, the output torque is 3 Nm, and the rate of change of the output voltage _Vamp is 0.7 V/ms. When switching the control, the motor rotation speed is less than 1 rps, and the switching is smooth with almost no fluctuation.

In the third embodiment, the drive method is switched by switching the function blocks used by the control devices 11, 42, and 51, but when actually configuring these control devices, software for a microcomputer or DSP (Digital Signal Processor) or a FPGA (Field Programmable Gate Array) can be used. Therefore, the switching of function blocks can be flexibly performed even in real-time control.

As described above, according to the third embodiment, normal pulse control is performed in which the first switching pattern is output twice following the output of the second switching pattern during a control period until the modulation rate of the signal driving inverters 1 and 2 reaches a threshold value, and then the second switching pattern is output again. Then, when the modulation rate reaches the threshold value, the control is switched to synchronous pulse control. This makes it possible to selectively use a drive method suited to the operating state of motor M. Furthermore, by appropriately setting the initial values of the parameters used for control when switching the drive method, fluctuations in the rotation speed of motor M can be suppressed, and switching can be performed smoothly.

Other embodiments
The current sensor 7 may be a shunt resistor or a CT.
The AC power supply may be single phase.
The switching elements are not limited to MOSFETs, but may also be other elements such as IGBTs, power transistors, and wide band gap semiconductors such as SiC and GaN.
The present invention is not limited to being applied to air conditioners, but may also be applied to other products.

Although several embodiments of the present invention have been described, these embodiments are presented as examples and are not intended to limit the scope of the invention. These novel embodiments can be embodied in various other forms, and various omissions, substitutions, and modifications can be made without departing from the gist of the invention. These embodiments and their modifications are included in the scope and gist of the invention, and are included in the scope of the invention and its equivalents described in the claims.

In the drawing, M indicates an open-structure winding motor, 1 indicates a primary inverter 1, 2 indicates a secondary inverter, 11 indicates a control device, 18 indicates a space voltage vector modulation unit, and 30 indicates an air conditioner.

Claims (5)

a primary-side inverter connected to three of six terminals of a motor having an open winding structure in which three phase windings are independent from each other and the motor has six terminals ;
a secondary side inverter connected to the remaining three terminals of the motor;
a control unit that controls a current supplied to the motor and a rotation speed based on a line duty ratio of each of the primary side and secondary side inverters in PWM control;
a current detector for detecting a current flowing through the motor;
the control unit controls a current and a rotation speed of the motor based on the line-to-line duties of the primary side and secondary side inverters, and has a zero-axis current suppression unit that suppresses a zero-axis current flowing in the same direction through three phases between the primary side and secondary side inverters,
The zero-axis current suppression unit is configured to suppress the zero-axis current from being applied to a space voltage vector consisting of 64 voltage vectors which are combinations of on/off patterns of the primary side and secondary side inverters.
an open winding motor drive device which divides six regions into 12 sectors by dividing the six regions into two equal parts, the six regions being centered at a point where two second switching patterns that do not generate a zero axis voltage and do not generate a voltage acting between the phases of the motor are located, and the six regions being centered at points where two first switching patterns that do not generate a zero axis voltage acting equally on the three phases of the motor and generate a voltage to be applied to the motor are located, and the six regions being centered at a point ... The open winding motor drive device according to claim 1, wherein the zero-axis current suppression unit selects the same first switching pattern for two adjacent sectors. the zero-axis current suppression unit inserts a third switching pattern which generates a zero-axis voltage and generates a voltage acting between phases of the motor before or after outputting only one pattern of the first switching pattern used according to each sector during a control period;
When generating a negative zero-axis voltage, only a switching pattern in which all switching elements constituting the primary side inverter are turned off is selected as the third switching pattern based on the first switching pattern;
3. The open winding motor drive device according to claim 1, wherein, when a positive zero-axis voltage is generated, only a switching pattern in which all switching elements constituting the secondary side inverter are turned off based on the first switching pattern is selected as the third switching pattern. If the control for selecting only one first switching pattern during a control period is called synchronous pulse control,
the zero-axis current suppression unit performs normal pulse control in which, during the control period, the first switching pattern is output twice following the output of the second switching pattern, and then the second switching pattern is output again, until a modulation rate of a signal that drives the inverter reaches a threshold value;
4. The open winding motor drive device according to claim 3, wherein the control is switched to the synchronous pulse control when the modulation rate reaches a threshold value. a motor having an open winding structure in which three phase windings are independent of each other and which has six winding terminals;
A refrigeration cycle device comprising: the open winding motor drive device according to any one of claims 1 to 4.

Images