Power converter

Abstract

To provide a power conversion apparatus in which common mode noise can be more preferably suppressed.SOLUTION: In a power conversion apparatus 1, a control device 21 includes a phase voltage command calculation unit 33 that calculates three-phase modulation command values cu1, cv1, cw1 for generating phase voltages, a line modulation calculation unit 34 that calculates two-phase modulation command values cu, cv, cw on the basis of the three-phase command values, and a PWM signal generation unit 36 that generates a PWM signal for performing PWM control on an inverter circuit 28 on the basis of a triangular carrier wave CA and the two-phase modulation command values. Regarding phase voltages of two phases outputted on the basis of the PWM signal, the rising timing of the phase voltage of one of the phases and the falling timing of the phase voltage of the other phase are made coincident with each other while the respective pulse widths based on the two-phase modulation command values are maintained during the first half period and / or the second half period of the carrier period of the triangular carrier wave CA.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] This invention relates to a power conversion device that drives a motor by applying a three-phase AC output to it using an inverter circuit. [Background technology]

[0002] Conventionally, power conversion devices are known that are based on a two-phase modulation scheme and can reduce common-mode noise by adjusting the rising or falling timing of pulsed PWM signals (PWM pulses) to reduce fluctuations in zero-sequence voltage (see, for example, Patent Document 1).

[0003] The power conversion device described in Patent Document 1 includes a phase voltage command calculation unit that calculates and outputs three-phase modulation command values ​​for generating voltages to be applied to each phase of a motor; a line-to-line modulation calculation unit that calculates two-phase modulation command values ​​based on the three-phase modulation command values ​​to fix the ON / OFF state of the upper and lower arm switching elements of one phase of the inverter circuit and modulate the ON / OFF states of the upper and lower arm switching elements of the other two phases; and a control device that generates a PWM signal for PWM control of the inverter circuit based on the two-phase modulation command values. The control device is configured to synchronize the switching timing of the two phases of upper and lower arm switching elements whose ON / OFF states are modulated, and to perform control that cancels out changes in the phase voltage applied to the motor with changes in the other phase voltages (control that cancels out fluctuations in zero-sequence voltage). As a result, fluctuations in the neutral point potential of the motor are small, and the generation of common-mode noise is suppressed. In this case, a sawtooth carrier wave is used in the PWM control. Since the sawtooth carrier wave can be switched at any timing, it is relatively easy to generate a PWM signal that performs the above operation. [Prior art documents] [Patent Documents]

[0004] [Patent Document 1] Japanese Patent Publication No. 2022-033476 [Overview of the Initiative] [Problems that the invention aims to solve]

[0005] Incidentally, isosceles triangular waves (hereinafter simply referred to as "triangular carrier waves") are relatively common as carrier waves. For this reason, it is desirable to use triangular carrier waves in the control that cancels out the zero-sequence voltage fluctuations mentioned above.

[0006] However, triangular carrier waves are based on the assumption of outputting symmetrical pulse widths, making it difficult to arbitrarily set the switching timing compared to sawtooth carrier waves. Specifically, depending on the algorithm used to generate the PWM signal, distortion may occur in the output current, leading to problems such as torque pulsation and noise generation in the motor.

[0007] The present invention has been made in view of the above-mentioned conventional situation, and provides a power converter that allows arbitrary setting of the rising and falling timings of the two-phase voltages when performing two-phase modulation control using a triangular carrier wave, thereby suppressing the generation of common-mode noise more effectively. [Means for solving the problem]

[0008] The present invention relates to a power conversion device comprising an inverter circuit that supplies phase voltages to a three-phase load, and a control device for the inverter circuit, wherein the control device includes a phase voltage command calculation unit that calculates a three-phase modulation command value for generating a phase voltage command value, a line-to-line modulation calculation unit that calculates a two-phase modulation command value based on the three-phase modulation command value, and a PWM signal generation unit that generates a PWM signal for PWM control of the inverter circuit based on a triangular carrier wave and the two-phase modulation command value, wherein the control device matches the rising timing of the phase voltage of one phase and the falling timing of the phase voltage of the other phase in at least one of the first half and second half periods within the carrier period of the triangular carrier wave, while maintaining the pulse width based on the two-phase modulation command value for each of the two phases of the phase voltages output based on the PWM signal,The command value inversion unit inverts one of the two-phase modulation command values; the command value division unit divides the two-phase modulation command value into a forward portion command value and a backward portion command value within the carrier cycle; and the command value shift unit matches the forward portion command values ​​of the two phases in the first half of the cycle, or matches the backward portion command values ​​of the two phases in the second half of the cycle. The present invention relates to a power conversion device characterized by the above. Effects of the Invention

[0009] According to the present invention, when two-phase modulation control is performed using a triangular carrier wave, an excellent effect is obtained in that the rising timing and falling timing of two-phase phase voltages can be arbitrarily set, and a power conversion device capable of more favorably suppressing the generation of common-mode noise is provided. Brief Description of the Drawings

[0010] [Figure 1] It is a vertical cross-sectional side view of an electric compressor according to an embodiment provided with a power conversion device. [Figure 2] It is a side view, with the cover removed, of the electric compressor of FIG. 1 viewed from the inverter accommodating portion side. [Figure 3] It is a diagram showing the power conversion device of the present embodiment, where (A) is a block diagram explaining an electric circuit, and (B) is a block diagram showing a partially extracted view. [Figure 4] (A) is a graph showing the calculation result of a three-phase modulation command value, (B) is a graph showing the calculation result of a modulation value cmod, and (C) is a graph showing the calculation result of inter-line modulation. [Figure 5] It is a diagram showing an example of a two-phase modulation command value and PWM pulses of each phase generated based thereon. [Figure 6] It is a diagram explaining shifting of PWM pulses. [Figure 7] It is a diagram explaining shifting of PWM pulses. [Figure 8] It is a flowchart explaining an algorithm for PWM control. [Figure 9] It is a diagram explaining shifting of PWM pulses. [Figure 10] It is a diagram explaining shifting of PWM pulses. [Figure 11] It is a diagram explaining shifting of PWM pulses. [Figure 12](A) Simulation results of the output voltage of the power converter (U-phase voltage Vu, V-phase voltage Vv, W-phase voltage Vw). (B) Current waveforms for each phase. [Figure 13] This graph shows the state of triangular carrier waves CA with electrical phase angles from 0 to 50 degrees, the two-phase modulation command values ​​cu, cv, cw for each phase, the phase voltages Vu, Vv, Vw, and the U-phase current iu, V-phase current iv, and W-phase current iw. [Figure 14] This diagram illustrates the shifting of PWM pulses. [Figure 15] This is a flowchart illustrating the PWM control algorithm. [Figure 16] This diagram illustrates the shifting of PWM pulses. [Figure 17] This diagram illustrates the sequence switching control of this embodiment. [Figure 18] This is the simulation result of the output voltage (U-phase voltage Vu, V-phase voltage Vv, W-phase voltage Vw) of the power converter. [Figure 19] This diagram illustrates the shifting of PWM pulses. [Figure 20] (A) Simulation results of the output voltage of the power converter (U-phase voltage Vu, V-phase voltage Vv, W-phase voltage Vw). (B) Current waveforms for each phase. [Figure 21] This figure illustrates another example of sequence switching control according to this embodiment. [Modes for carrying out the invention]

[0011] Embodiments of the present invention will be described in detail below with reference to the accompanying drawings. In the accompanying drawings, parts denoted by the same reference numerals represent the same components. Furthermore, some components are omitted in each drawing as appropriate to simplify the drawings. In addition, the shape and dimensions of some components are exaggerated in each drawing as appropriate.

[0012] <Configuration of an electric compressor> First, an example of an electric compressor 16 equipped with the power converter 1 of this embodiment will be described with reference to Figure 1. The electric compressor 16 of this embodiment is a so-called inverter-integrated electric compressor that has the power converter 1 integrated into it, and constitutes part of the refrigerant circuit of a vehicle air conditioning system installed in electric vehicles such as hybrid vehicles and electric vehicles.

[0013] The metallic cylindrical housing 2 of the electric compressor 16 is divided into a compression mechanism housing 4 and an inverter housing 6 by partition walls 3 that intersect the housing 2 in the axial direction. The compression mechanism housing 4 houses, for example, a scroll-type compression mechanism 7 and a motor 8 that drives this compression mechanism 7. In this case, the motor 8 is an IPMSM (Interior Permanent Magnet Synchronous Motor) consisting of a stator 9 fixed to the housing 2 and a rotor 11 that rotates inside the stator 9.

[0014] A bearing section 12 is formed in the center of the partition wall 3 on the side of the compression mechanism housing 4. One end of the drive shaft 13 of the rotor 11 is supported by this bearing section 12, and the other end of the drive shaft 13 is connected to the compression mechanism 7. An intake port 14 is formed near the partition wall 3 at the position corresponding to the compression mechanism housing 4 of the housing 2. When the rotor 11 (drive shaft 13) of the motor 8 rotates and the compression mechanism 7 is driven, a low-temperature refrigerant, which is the working fluid, flows into the compression mechanism housing 4 of the housing 2 from this intake port 14, is drawn into the compression mechanism 7, and is compressed.

[0015] The refrigerant, compressed by the compression mechanism 7 and becoming high temperature and pressure, is then discharged from a discharge port (not shown) into the refrigerant circuit outside the housing 2. In addition, the low-temperature refrigerant flowing in from the intake port 14 passes near the partition wall 3, around the motor 8, and is drawn into the compression mechanism 7, thus cooling the partition wall 3 as well.

[0016] Then, within the inverter housing 6, which is separated from the compression mechanism housing 4 by this partition wall 3, is a power conversion device 1 that drives and controls the motor 8. The power conversion device 1 supplies power to the motor 8 via sealed terminals and lead wires that penetrate the partition wall 3.

[0017] <Structure of a power converter> Figure 2 is a side view of the electric compressor 16 shown in Figure 1, excluding the cover 23, as seen from the inverter housing 6 side. The power converter 1 comprises a three-phase inverter circuit 28 that supplies phase voltage to a three-phase load (motor 8 in this embodiment), and a control device 21 for the inverter circuit 28. Specifically, the power converter 1 consists of a substrate 17, a total of six switching elements (upper arm switching elements 18A, 18B, 18C and lower arm switching elements 18D, 18E, 18F) wired on one side of the substrate 17, a control device 21 wired on the other side of the substrate 17, and HV connectors, LV connectors, etc. (not shown). Each upper and lower arm switching element 18A to 18F is a power switching element, and in this embodiment, as an example, it is composed of an insulated gate bipolar transistor (IGBT) with a MOS (metal oxide semiconductor) structure incorporated into the gate portion.

[0018] In this example, the upper arm switching elements 18A and lower arm switching elements 18D of the U-phase half-bridge circuit (inverter) 19U, the upper arm switching elements 18B and lower arm switching elements 18E of the V-phase half-bridge circuit (inverter) 19V, and the upper arm switching elements 18C and lower arm switching elements 18F of the W-phase half-bridge circuit (inverter) 19W, which constitute the three-phase inverter circuit 28 described later with reference to Figure 3, are arranged side by side in pairs. A pair of upper and lower arm switching elements 18A and 18D, upper and lower arm switching elements 18B and 18E, and upper and lower arm switching elements 18C and 18F are arranged radially around the center of the substrate 17 as shown in Figure 2.

[0019] In this example, the upper and lower arm switching elements 18A and 18D of the U-phase half-bridge circuit 19U are located on the intake port 14 side, the upper and lower arm switching elements 18B and 18E of the V-phase half-bridge circuit 19V are positioned 90° counterclockwise relative to them in Figure 2, and the upper and lower arm switching elements 18C and 18F of the W-phase half-bridge circuit 19W are positioned on the opposite side from the intake port 14. The refrigerant drawn in from the intake port 14 rotates counterclockwise around the axis of the housing 2, as indicated by the dashed arrows in Figure 2. Therefore, the upper and lower arm switching elements 18A and 18D of the U-phase half-bridge circuit 19U are located furthest upstream (the point where the electric compressor 16 is coldest) relative to the flow of the drawn refrigerant, the upper and lower arm switching elements 18B and 18E of the V-phase half-bridge circuit 19V are located downstream of them, and the switching elements 18C and 18F of the W-phase half-bridge circuit 19W are located furthest downstream. Furthermore, the terminals 22 of each upper and lower arm switching element 18A to 18F are connected to the circuit board 17 with the terminals facing the center of the circuit board 17.

[0020] Then, as shown in Figure 1, the power converter 1 assembled in this manner is housed in the inverter housing 6 with one side of each upper and lower arm switching element 18A to 18F facing the partition wall 3, and is attached to the partition wall 3, and then closed with a cover 23. The circuit board 17 is fixed to the partition wall 3, for example, via boss portions 24 that rise from the partition wall 3.

[0021] With the power converter 1 attached to the partition wall 3 in this manner, each switching element 18A to 18F is in close contact with the partition wall 3, either directly or via a predetermined insulating thermal conductive material, and is in a heat exchange relationship with the partition wall 3 of the housing 2. At this time, each upper and lower arm switching element 18A to 18F is positioned to avoid the areas corresponding to the bearing portion 12 and the drive shaft 13, and is arranged to surround them (Figure 2).

[0022] As mentioned above, the partition wall 3 is cooled by the refrigerant drawn into the compression mechanism housing 4. Therefore, each upper and lower arm switching element 18A to 18F exchanges heat with the drawn refrigerant via the partition wall 3, is cooled by the refrigerant drawn into the compression mechanism housing 4 through the thickness of the partition wall 3, and each upper and lower arm switching element 18A to 18F itself dissipates heat to the refrigerant via the partition wall 3.

[0023] <Circuit configuration of power converter> The power converter 1 will be described with reference to Figure 3. Figure 3(A) is a circuit block diagram showing the electrical circuit configuration of the power converter 1, and Figure 3(B) is a block diagram showing a part of the control device 21. The power converter 1 is equipped with the aforementioned three-phase inverter circuit 28 and the control device 21. The inverter circuit 28 is a circuit that converts the DC voltage of a DC power supply (vehicle battery: for example, 350V) 29 into a three-phase AC voltage (three-phase AC output) using a PWM (Pulse Width Modulation) control method and applies it to the motor 8. This inverter circuit 28 has a U-phase half-bridge circuit 19U, a V-phase half-bridge circuit 19V, and a W-phase half-bridge circuit 19W, and each phase half-bridge circuit 19U to 19W has upper arm switching elements 18A to 18C and lower arm switching elements 18D to 18F individually. Furthermore, each upper and lower arm switching element 18A to 18F has a flywheel diode 31 connected in antiparallel.

[0024] The upper ends of the upper arm switching elements 18A to 18C of the inverter circuit 28 are connected to the upper arm power line (positive bus) 10 of the DC power supply 29 and the smoothing capacitor 32. On the other hand, the lower ends of the lower arm switching elements 18D to 18F of the inverter circuit 28 are connected to the lower arm power line (negative bus) 15 of the DC power supply 29 and the smoothing capacitor 32. The voltage (DC voltage) of the DC power supply 29 is Vdc.

[0025] In this case, the upper arm switching element 18A and the lower arm switching element 18D of the U-phase half-bridge circuit 19U are connected in series, the upper arm switching element 18B and the lower arm switching element 18E of the V-phase half-bridge circuit 19V are connected in series, and the upper arm switching element 18C and the lower arm switching element 18F of the W-phase half-bridge circuit 19W are connected in series.

[0026] Furthermore, the connection point (U-phase voltage Vu) between the upper arm switching element 18A and the lower arm switching element 18D of the U-phase half-bridge circuit 19U is connected to the U-phase armature coil 41 of the motor 8, the connection point (V-phase voltage Vv) between the upper arm switching element 18B and the lower arm switching element 18E of the V-phase half-bridge circuit 19V is connected to the V-phase armature coil 42 of the motor 8, and the connection point (W-phase voltage Vw) between the upper arm switching element 18C and the lower arm switching element 18F of the W-phase half-bridge circuit 19W is connected to the W-phase armature coil 43 of the motor 8.

[0027] <Control device configuration> The control device 21 consists of a microcomputer with a processor. In this example, it receives rotational speed command values ​​from the vehicle ECU and motor current (phase current) from the motor 8, and controls the ON / OFF state (switching) of each upper and lower arm switching element 18A to 18F of the inverter circuit 28 based on these values. Specifically, it controls the gate voltage applied to the gate terminals of each upper and lower arm switching element 18A to 18F.

[0028] The control device 21 includes a phase voltage command calculation unit 33, a line-to-line modulation calculation unit 34, a PWM signal generation unit 36, a gate driver 37, and current sensors 26A, 26B, and 26C, which consist of current transformers for measuring the motor currents (phase currents) of each phase flowing through the motor 8: U-phase current iu, V-phase current iv, and W-phase current iw. Each of the current sensors 26A, 26B, and 26C is connected to the line-to-line modulation calculation unit 34.

[0029] The phase voltage command calculation unit 33 calculates the phase voltage command value (U phase voltage command value Vu * V-phase voltage command value Vv * W-phase voltage command value Vw * The three-phase modulation command values ​​cu1, cv1, and cw1 for generating the above are calculated and output.

[0030] The line-to-line modulation calculation unit 34 calculates the two-phase modulation command values ​​cu, cv, and cw for each phase based on the three-phase modulation command values ​​cu1, cv1, and cw1 calculated by the phase voltage command calculation unit 33.

[0031] The PWM signal generation unit 36 ​​generates and outputs PWM signals that serve as drive command signals for the U-phase half-bridge circuit 19U, V-phase half-bridge circuit 19V, and W-phase half-bridge circuit 19W of the inverter circuit 28 by comparing the magnitudes of the two-phase modulation command values ​​cu, cv, and cw for each phase (U-phase, V-phase, and W-phase) with the triangular carrier wave (carrier signal).

[0032] The gate driver 37 generates gate voltages for the switching elements 18A and 18D of the U-phase half-bridge circuit 19U, the gate voltages for the switching elements 18B and 18E of the V-phase half-bridge circuit 19V, and the gate voltages for the switching elements 18C and 18F of the W-phase half-bridge circuit 19W, based on the PWM signal output from the PWM signal generation unit 36.

[0033] Then, each of the upper and lower arm switching elements 18A to 18F of the inverter circuit 28 is driven ON / OFF based on the gate voltage output from the gate driver 37. That is, when the gate voltage is ON (a predetermined voltage value), the switching element operates ON, and when the gate voltage is OFF (zero), the switching element operates OFF.

[0034] This gate driver 37 is a circuit for applying a gate voltage to the IGBTs based on a PWM signal, when the switching elements 18A to 18F are the IGBTs mentioned above, and is composed of photocouplers, logic ICs, transistors, etc.

[0035] When an ON signal is input to the gate driver 37 as a PWM signal for each phase (U phase, V phase, W phase), the upper arm switching elements 18A to 18C of the corresponding phase turn ON, and the phase voltage of that phase becomes "H" level. Conversely, when an OFF signal is input to each phase (U phase, V phase, W phase) as a PWM signal, the upper arm switching elements 18A to 18C of the corresponding phase turn OFF, and the phase voltage of that phase becomes "L" level. The inverted signals of the upper arm switching elements 18A to 18C are input to the lower arm switching elements 18D to 18F. In practice, when switching the upper and lower arm switching elements 18A to 18F, a so-called dead time is introduced for each phase, during which both the upper arm switching element 18A and the lower arm switching element 18D of the U-phase half-bridge circuit 19U, both the upper arm switching element 18B and the lower arm switching element 18E of the V-phase half-bridge circuit 19V, and both the upper arm switching element 18C and the lower arm switching element 18F of the W-phase half-bridge circuit 19W are in the OFF state. This dead time is implemented by the gate driver 37 and the PWM signal generation unit 36. This dead time is introduced and operated in all operations of the present invention, but in this embodiment, the description of the dead time is omitted for the sake of simplicity.

[0036] The voltage at the connection point between the upper arm switching element 18A and the lower arm switching element 18D of the U-phase half-bridge circuit 19U is applied (output) to the U-phase armature coil 41 of the motor 8 as the U-phase voltage Vu (phase voltage). The voltage at the connection point between the upper arm switching element 18B and the lower arm switching element 18E of the V-phase half-bridge circuit 19V is applied (output) to the V-phase armature coil 42 of the motor 8 as the V-phase voltage Vv (phase voltage). The voltage at the connection point between the upper arm switching element 18C and the lower arm switching element 18F of the W-phase half-bridge circuit 19W is applied (output) to the W-phase armature coil 43 of the motor 8 as the W-phase voltage Vw (phase voltage).

[0037] Note that, in addition to measuring the motor current of each phase with current sensors 26A, 26B, and 26C as in the present embodiment, there are also methods such as detecting the current value on the lower arm power supply line 15 and estimating the same by the phase voltage command calculation unit 33 from said current value and the operating state of the motor 8. Therefore, the method for detecting and estimating each phase current is not limited to this example.

[0038] <Phase Voltage Command Calculation Unit> The phase voltage command calculation unit 33 generates U-phase voltage Vu, V-phase voltage Vv, and W-phase voltage Vw to be applied to the armature coils 41 to 43 of each phase of the motor 8 through vector control based on the electrical angle of the motor 8, the current command value, and the d-axis current and q-axis current obtained from the phase currents, and thereby calculates and generates the phase voltage command values Vu * , Vv * , Vw * . These phase voltage command values Vu * , Vv * , Vw * (U-phase voltage command value Vu * , V-phase voltage command value Vv * , and W-phase voltage command value Vw * ) are voltage command values used when performing three-phase modulation control for the motor 8.

[0039] The phase voltage command calculation unit 33 calculates the phase voltage command values Vu * , Vv * , Vw * for each phase by using the following formula (1) based on the d-axis voltage Vd and q-axis voltage Vq obtained from the d-axis current and q-axis current. Here, Vm and θm in formula (1) are respectively obtained from formula (2). Vm is the amplitude of the phase voltage command value. θ is the magnetic pole position referenced to the U phase, and θm is the voltage phase difference with respect to the magnetic pole position. [Formula] [Formula]

[0040] <Line-to-Line Modulation Calculation Unit> The line modulation calculation unit 34 calculates and outputs the pulse width command values ​​for each phase, cu1 (U-phase pulse width command value), cv1 (V-phase pulse width command value), and cw1 (W-phase pulse width command value), using the following equation (3).

number

[0041] In the case of a DC voltage Vdc, in the non-overmodulation region, the phase voltage command value Vu * , Vv * VW * Since the voltage range is -Vdc / 2 to Vdc / 2, the pulse width command values ​​cu1, cv1, and cw1 for each phase will be between 0 and 1. That is, the pulse width command values ​​cu1, cv1, and cw1 for each phase will be equal to the phase voltage command value Vu * , Vv * VW * This value is normalized (corrected to 0-1) by the DC voltage Vdc. The pulse width command values ​​cu1, cv1, and cw1 for each phase are command values ​​that cause a PWM signal of a certain pulse width to be output in PWM control. If the value is "1", it is a command to be always ON in the specified section (in this embodiment, the section is set based on the carrier period of the triangular carrier wave). If the value is "0", it is a command to be always OFF in the specified section. If the value is "0.5", it is a command to be ON in half of the specified section and OFF in the other half.

[0042] The pulse width command values ​​cu1, cv1, and cw1 for each phase are the values ​​before line-to-line modulation (values ​​without line-to-line modulation), and are hereafter referred to as the three-phase modulation command values ​​cu1, cv1, and cw1 for each phase (U-phase modulation command value cu1, V-phase modulation command value cv1, and W-phase modulation command value cw1). The pulse width of the PWM signal is also referred to as the "PWM pulse width".

[0043] Next, the line-to-line modulation calculation unit 34 uses equation (4) to perform line-to-line modulation (two-phase modulation in this embodiment) using the modulation value cmod from the three-phase modulation command values ​​cu1, cv1, and cw1 of each phase, and calculates and outputs the U-phase two-phase modulation command value cu, the V-phase two-phase modulation command value cv, and the W-phase two-phase modulation command value cw. The U-phase two-phase modulation command value cu, the V-phase two-phase modulation command value cv, and the W-phase two-phase modulation command value cw are the pulse width command values ​​of each phase after line-to-line modulation, respectively.

number

[0044] In equation (4), the modulation value Cmod is determined based on the following equation (5).

number

[0045] First, as shown in equation (5-1), c in equations (5-2) to (5-4) half is 1, and c quarter It is 0.5.

[0046] Equation (5-2) represents the minimum phase value among the U-phase three-phase modulation command value cu1, V-phase three-phase modulation command value cv1, and W-phase three-phase modulation command value cw1 in this embodiment. Equation (5-3) means that when the U-phase is the minimum phase, the three-phase modulation command value of the larger of the V-phase and W-phase is set to 1 (matching the maximum value of the triangular carrier wave). This performs two-phase modulation by fixing the upper arm switching element of the larger of the V-phase and W-phase to the ON state (phase voltage "H").

[0047] Furthermore, equation (5-4) means that if the V-phase or W-phase is the smallest phase, the three-phase modulation command value for that smallest phase is set to 0 (matching the minimum value of the triangular carrier wave). This performs two-phase modulation, fixing the upper arm switching element of the smaller of the V-phase and W-phase to the OFF state (phase voltage "L"). In this way, two-phase modulation command values ​​cu, cv, and cw are calculated to fix the switching of either the V-phase or W-phase to the ON state or the OFF state.

[0048] Figure 4 is a graph showing the results of calculations performed using equations (3) to (5) above for one rotation (360 degrees) of the electrical angular phase. Figure 4(A) shows the calculation result of equation (3), with the vertical axis representing the three-phase modulation command values ​​cu1, cv1, and cw1 for each phase, and the horizontal axis representing the electrical angular phase [deg]. Figure 4(B) shows the calculation result of equation (5), with the vertical axis representing the modulation value cmod and the horizontal axis representing the electrical angular phase [deg]. Figure 4(C) shows the calculation result of equation (4) using the modulation value cmod shown in Figure 4(B), with the vertical axis representing the two-phase modulation command values ​​cu, cv, and cw for each phase, and the horizontal axis representing the electrical angular phase [deg]. In this embodiment, the carrier frequency of the triangular carrier wave is 10 kHz and the modulation index kHz is 0.65 (the same applies in the following explanation). The modulation index kH represents the amplitude of the sinusoidal output voltage and is the value obtained by dividing the phase voltage command value amplitude Vm in equation (1) by Vdc / 2 (kH = Vm / (Vdc / 2)). Thus, the three-phase modulation command values ​​cu1, cv1, cw1 for each phase and the two-phase modulation command values ​​cu, cv, cw for each phase all have periodicity, and the phase voltage command value (U phase voltage command value Vu) generated by these has periodicity. * V-phase voltage command value Vv * W-phase voltage command value Vw * ) also exhibits periodicity.

[0049] As shown in Figure 4(C), in this embodiment, when the electrical angular phase is not between 210 and 330 degrees (the period when either the V-phase two-phase modulation command value cv or the W-phase two-phase modulation command value cw is the smallest), the PWM operation of the phase with the smaller two-phase modulation command value between the V-phase and the W-phase is stopped, and the upper arm switching element of the phase with the smaller two-phase modulation command value is set to the OFF state (phase voltage is at the "L" level). Hereinafter, this state is referred to as "lower fixed operation" in this embodiment, and the region where the electrical angular phase is not between 210 and 330 degrees is referred to as the "lower fixed region". The phase that performs the lower fixed operation is also referred to as the lower fixed phase. In Figure 4(C), when the electrical angular phase is between 0 and 90 degrees and between 330 and 360 degrees, the lower fixed phase is the V-phase, and when the electrical angular phase is between 90 and 210 degrees, the lower fixed phase is the W-phase.

[0050] In the above explanation, the range of electrical angular phase was described as "except for ~," but at the limit values ​​(210deg and 330deg), either upper fixed operation or lower fixed operation may be selected. That is, at 210deg, the W phase may be set as the lower fixed phase, or the V phase may be set as the upper fixed phase as described later, and in either case, this embodiment can obtain the same effect. The same applies to 330deg, where the V phase may be set as the lower fixed phase, or the W phase may be set as the upper fixed phase as described later.

[0051] Furthermore, when describing the range of electrical angular phase, there are cases where the boundary values ​​overlap for adjacent phase ranges (for example, "30deg to 90deg (range)" and "90deg to 150deg (range)"). In this case, the overlapping boundary value (90deg in this example) is considered to be included in either range, and it may be included in either range. For example, it is clear from the equations and Figure 4(C) that at 90deg, both the V phase and the W phase are lower fixed phases.

[0052] Furthermore, to make the phase relationships in the figures easier to understand, we have deliberately shown the phase relationship when θm = -90 degrees in equation (1), using a sinusoidal phase with respect to 0 degrees of the U phase as the reference. We will assume the same conditions apply to all subsequent figures.

[0053] Furthermore, during the period when the electrical angular phase is between 210 and 330 degrees (the period when the U-phase two-phase modulation command value cu is at its minimum), the PWM operation of the larger of the V-phase and W-phase is stopped, and the upper arm switching element of the larger phase is turned ON (phase voltage is at the "H" level). Hereinafter, this state is referred to as "upper fixed operation" in this embodiment, and the region where the electrical angular phase is between 210 and 330 degrees is referred to as the "upper fixed region." The phase that performs upper fixed operation is also referred to as the upper fixed phase. When the electrical angular phase is between 210 and 270 degrees, the upper fixed phase is the V-phase, and when the electrical angular phase is between 270 and 330 degrees, the upper fixed phase is the W-phase.

[0054] As described above, the three-phase modulation command values cu1, cv1, cw1 of each phase are modulated by the same modulation value cmod, enabling two-phase modulation that maintains the line-to-line voltage. Note that, as a method for setting the modulation value cmod, this example adopts a two-phase modulation scheme in which only the V phase and W phase are selected as phases to stop PWM operation (PWM operation stop phases), but the method for setting the modulation value is not limited thereto.

[0055] <PWM signal generator> The PWM signal generator 36, for example, generates a triangular carrier wave (carrier signal), and compares the magnitude of the two-phase modulation command values cu, cv, cw of each phase with the triangular carrier wave, thereby generating and outputting PWM signals that serve as drive command signals for the U-phase half-bridge circuit 19U, V-phase half-bridge circuit 19V, and W-phase half-bridge circuit 19W of the inverter circuit 28. Note that the triangular carrier wave may be input from outside the PWM signal generator 36.

[0056] Here, the PWM signal generator 36 of the present embodiment generates PWM pulses based on the two-phase modulation command values for the two phases that perform PWM operation. Then, while maintaining the PWM pulse widths of the two-phase PWM pulses respectively, in one of the first half period (the period before the peak (1 / 2) of the carrier period) and the second half period (the period after the peak (1 / 2) of the carrier period) within one period of the triangular carrier wave (carrier period), the PWM signal is generated such that the rising timing of the PWM pulse of one phase matches the falling timing of the PWM pulse of the other phase. The PWM pulse width refers to the ratio of the ON time of the PWM signal to the carrier period (one period) (PWM duty ratio), and may be simply referred to as "duty ratio" hereinafter.

[0057] The control device 21 then controls the application of phase voltages to the motor 8 based on the PWM signal. This ensures that the rising edge timing of one phase's phase voltage and the falling edge timing of the other phase's phase voltage coincide within the carrier period of the triangular carrier wave, either in the first or second half of the period. In this case, the two-phase PWM pulse widths (the pulse width at the "H" level of the phase voltage, i.e., the ON time) within the carrier period are maintained. As a result, fluctuations in the zero-sequence voltage during the rising or falling edge of the PWM pulse can be canceled out in PWM control using a triangular carrier wave. This control will be described in detail below.

[0058] Referring to Figure 3(B), the control device 21 (PWM signal generation unit 36) includes a command value inversion unit 51, a command value division unit 52, and a command value shift unit 53. The command value inversion unit 51 inverts one of the two-phase modulation command values, specifically, for example, the U-phase two-phase modulation command value cu, to generate an inverted U-phase two-phase modulation command value 1-cu. Here, "inversion" means subtracting the value of the U-phase two-phase modulation command value cu, which is normalized to between 0 and 1, from its maximum value of 1.

[0059] Then, by comparing the inverted U-phase two-phase modulation command value 1-cu, the V-phase two-phase modulation command value cv, the W-phase two-phase modulation command value cw with the triangular carrier wave, PWM pulses are generated to apply the phase voltages Vu, Vv, and Vw for each phase.

[0060] The inverted U-phase two-phase modulation command value 1-cu is the same as the U-phase two-phase modulation command value cu, but with a different value. In the PWM control of this embodiment, it is used in place of the U-phase two-phase modulation command value cu. In other words, in this embodiment, when we refer to the "two-phase modulation command values ​​cu, cv, cw for each phase" collectively, the inverted U-phase two-phase modulation command value 1-cu is also included.

[0061] The command value division unit 52 divides (bisects) the two-phase modulation command values ​​cu, cv, and cw for each phase, i.e., the inverted U-phase two-phase modulation command value 1-cu, the V-phase two-phase modulation command value cv, and the W-phase two-phase modulation command value cw, into a forward portion command value and a backward portion command value within each carrier period of the triangular carrier wave.

[0062] The command value shift unit 53 shifts the divided forward and backward command values ​​within each carrier cycle so that their values ​​increase or decrease. As a result, the PWM pulses of each phase are shifted in the forward / backward (forward / backward) direction within the carrier cycle while maintaining their respective PWM pulse widths Pu, Pv, and Pw.

[0063] Specifically, the operation of the command value inversion unit 51, the command value division unit 52, and the command value shift unit 53 (command value inversion control, command value division control, and command value shift control) will be explained with reference to Figures 5 to 7. First, Figure 5 shows one period of the triangular carrier wave CA when the electrical angular phase is 0 (deg), and an example of the PWM pulses for each phase generated based on the two-phase modulation command values ​​cu, cv, and cw for each phase.

[0064] First, referring to Figure 5(A), the right-hand figure of Figure 5(A) is a reproduction of the graphs of the two-phase modulation command values ​​cu, cv, and cw for each phase shown in Figure 4(C). In this case, from the right-hand figure of Figure 5(A), the U-phase two-phase modulation command value cu at an electrical angular phase of 0 (deg) is approximately 0.25, the W-phase two-phase modulation command value cw is approximately 0.57, and the V-phase two-phase modulation command value cv is 0. The left-hand figure of Figure 5(A) shows a comparison of the two-phase modulation command values ​​cu, cv, and cw for each phase with the triangular carrier wave CA, with one carrier period extracted near an electrical angular phase of 0 (deg). In this figure, one period of the triangular wave carrier CA (carrier period) and the two-phase modulation command values ​​cu (solid line), cv (dotted line), and cw (dashed line) for each phase are superimposed at the top. The lower part of the diagram illustrates the PWM pulses for each phase (the pulses of the applied phase voltage) generated by comparing the two-phase modulation command values ​​cu, cv, and cw for each phase with the triangular carrier wave CA. On the vertical axis, "Vu" represents the U-phase PWM pulse (U-phase voltage Vu), "Vv" represents the V-phase PWM pulse (V-phase voltage Vv), and "Vw" represents the W-phase PWM pulse (W-phase voltage Vw).

[0065] In the left diagram of Figure 5(A), it is shown that when the two-phase modulation command values ​​cu, cv, and cw for each phase are greater than the triangular carrier wave CA, the phase voltages Vu, Vv, and Vw for each phase become "H" level (ON), and the voltage of the DC power supply 29 is applied to the motor 8. For example, when the U-phase voltage Vu is "H", the upper arm switching element 18A of the U-phase half-bridge circuit 19U becomes ON, and the lower arm switching element 18D becomes OFF, and the voltage applied to the U-phase becomes the voltage of the DC power supply 29. The same applies to the V-phase voltage Vv and the W-phase voltage Vw. In this case, the PWM pulse widths Pu, Pv, and Pw for each phase (U-phase PWM pulse width Pu, V-phase PWM pulse width Pv, W-phase PWM pulse width Pw) are the ON times of the upper arm switching elements 18A, 18C, and 18E, respectively.

[0066] Furthermore, if the two-phase modulation command values ​​cu, cv, and cw for each phase are smaller than the triangular carrier wave CA, the phase voltages Vu, Vv, and Vw for each phase will be at the "L" level (OFF), indicating that 0V will be applied to the motor 8. For example, when the U-phase voltage Vu is "L", the upper arm switching element 18A of the U-phase half-bridge circuit 19U is OFF, the lower arm switching element 18D is ON, and the voltage applied to the U-phase becomes 0V. The same applies to the V-phase voltage Vv and the W-phase voltage Vw.

[0067] As described above, when the phase voltages Vu, Vv, and Vw of each phase are "H" or "L", they are linked to the ON / OFF state of the upper arm switching elements 18A, 18C, and 18E of each phase. Therefore, in this embodiment, the operation of the upper arm switching elements 18A, 18C, and 18E of each phase is described.

[0068] Note that the V phase remains fixed at the lower position for a period up to 90 degrees, and acts as the stop phase for PWM control. In this embodiment, the diagram shown in Figure 5(A) (left) is called the "carrier period PWM pulse diagram". The interpretation of the carrier period PWM pulse diagram is as described above.

[0069] Figure 5(B) is a PWM pulse diagram within the carrier period at an electrical phase angle of 0 (deg), showing an example of PWM pulses for each phase based on the inverted U-phase two-phase modulation command value 1-cu, the V-phase two-phase modulation command value cv, and the W-phase two-phase modulation command value cw, when the command value inversion unit 51 generates the inverted U-phase two-phase modulation command value 1-cu.

[0070] The command value inversion unit 51 performs command value inversion control. The right-hand figure of Figure 5(B) is a graph in which only the U-phase two-phase modulation command value cu is inverted vertically in the graph of the two-phase modulation command values ​​cu, cv, and c for each phase shown in the right-hand figure of Figure 5(A), and this represents the inverted U-phase two-phase modulation command value 1-cu. In this embodiment, the PWM control described below is performed based on the right-hand figure of Figure 5(B) in which the U-phase two-phase modulation command value cu is inverted. Therefore, the graph in the right-hand figure of Figure 5(B) after inversion of the U-phase two-phase modulation command value cu will be referred to as the "two-phase modulation command waveform diagram" and explained accordingly. Note that the two-phase modulation command waveform diagram and the waveform diagram shown in Figure 4(C) differ only in whether or not the U-phase is inverted. In other words, the two-phase modulation command waveform diagram in Figure 5(B) also has a lower fixed region, an upper fixed region, a lower fixed phase, and an upper fixed phase, and these are the same as those explained with reference to Figure 4(B).

[0071] In Figure 5(B), the two-phase modulation command value cu for the U-phase is inverted, so the determination of the magnitude relationship with the triangular carrier wave CA is also the opposite of that for the other phases. That is, in the PWM pulse diagram within the carrier period in the left diagram of Figure 5(B), for the U-phase, if the inverted U-phase two-phase modulation command value 1-cu (solid line) is smaller than the triangular carrier wave CA, the U-phase voltage Vu becomes "H" level (ON), and if the inverted U-phase two-phase modulation command value 1-cu is larger than the triangular carrier wave CA, the U-phase voltage Vu becomes "L" level (OFF). In order to invert the PWM order only for the U-phase, methods such as using a triangular carrier wave that is inverted only for the U-phase, or inverting the gate driver output only for the U-phase, can be considered, but any method can be used.

[0072] For the V-phase and W-phase, the PWM pulse diagrams within the carrier period are the same as in Figure 5(A). When the V-phase two-phase modulation command value cv (dotted line) and the W-phase two-phase modulation command value cw (dashed line) are greater than the triangular carrier wave CA, the phase voltages Vv and Vw of each phase are at the "H" level (ON). When they are smaller than the triangular carrier wave CA, the phase voltages Vv and Vw of each phase are at the "L" level (OFF, 0). In this case as well, the PWM pulse widths Pu, Pv, and Pw of each phase represent the ON time of the switch.

[0073] As shown in Figure 5(B), inverting the U-phase two-phase modulation command value cu causes the timing at which the U-phase voltage Vu reaches the "H" level (ON) to differ from that in Figure 5(A) before inversion (it shifts with respect to the forward / backward direction (left / right direction in the diagram) of the triangular carrier wave CA). That is, in the left diagram of Figure 5(A), the U-phase PWM pulses are generated at both ends of the carrier period with U-phase PWM pulse widths Pu1 and Pu2, while in Figure 5(B), they are generated near the center of the carrier period with a U-phase PWM pulse width Pu. However, in both cases, the total U-phase PWM pulse width Pu is the same (Pu = Pu1 + Pu2), and the total time at which the U-phase voltage Vu reaches "H" remains unchanged.

[0074] Figure 6 is a PWM pulse diagram within the carrier period to illustrate the operation of the command value division unit 52 and the command value shift unit 53. The command value division unit 52 performs command value division control, and the command value shift unit 53 performs command value shift control. Figure 6(A) is a reproduction of the left diagram of Figure 5(B). Figure 6(B) is a diagram illustrating the forward and backward portion command values ​​divided by the command value division unit 52, and Figure 6(C) is a diagram illustrating the shift of the forward and backward portion command values. For the sake of clarity, Figure 6 omits the description of the V phase (V phase voltage Vv), which is the stop phase of the PWM operation.

[0075] As shown in Figures 6(A) and 6(B), the command value division unit 52 divides the command values ​​of two phases that have a value greater than 0 among the inverted U-phase two-phase modulation command value 1-cu, the V-phase two-phase modulation command value cv, and the W-phase two-phase modulation command value cw (in this example, the inverted U-phase two-phase modulation command value 1-cu (solid line, similar in the following carrier period PWM pulse diagram) and the W-phase two-phase modulation command value cw (dashed line, similar in the following carrier period PWM pulse diagram)) into a forward portion command value and a backward portion command value within the carrier period. Here, the forward portion command value refers to a command value that can intersect with the slope (forward side slope, upward slope from 0 to 1 (peak)) in at least the first half of the triangular carrier wave CA, and the backward portion command value refers to a command value that can intersect with the slope (backward side slope, downward slope from 1 (peak) to 0) in at least the second half of the triangular carrier wave CA.

[0076] Furthermore, "division (bipartition)" refers to the ability to switch between two values ​​for each phase's two-phase modulation command values ​​1-cu, cv, and cw, which are used to compare the magnitude relationship with the triangular carrier wave CA within the carrier period. In other words, the command value used for comparison with the forward side slope (upward slope) is the forward part command value, and the command value used for comparison with the backward side slope (downward slope) is the backward part command value.

[0077] Specifically, the command value division unit 52 divides the inverted U-phase two-phase modulation command value 1-cu into a U-phase forward portion command value cuf and a U-phase backward portion command value cub within the carrier period (making it switchable). In the example in Figure 6(B), the division is made by a dividing line (switching line) DL that passes through the center (vertex) of the carrier period. The portion in front of the dividing line DL (left side in the figure) is the U-phase forward portion command value cuf, and the portion behind the dividing line DL (right side in the figure) is the backward portion command value cub. The U-phase forward portion command value cuf intersects the uphill slope of the triangular carrier wave CA, and the U-phase backward portion command value cub intersects the downhill slope of the triangular carrier wave CA.

[0078] Furthermore, the command value division unit 52 divides the W-phase two-phase modulation command value cw into two parts within the carrier period: the W-phase forward portion command value cwf (forward of the dividing line DL (left side in the figure)) and the W-phase backward portion command value cwb (backward of the dividing line DL (right side in the figure)). The forward portion command value cwf intersects the upward slope of the triangular carrier wave CA, and the W-phase backward portion command value cwb intersects the downward slope of the triangular carrier wave CA.

[0079] As shown in Figure 6(C), the command value shift unit 53 independently shifts the forward portion command values ​​and backward portion command values ​​of the two phases (U-phase forward portion command value cuf and U-phase backward portion command value cub, W-phase forward portion command value cwf and W-phase backward portion command value cwb) in a direction that increases or decreases the value within the range of 0 to 1 of the command value. However, in this case, the command value shift unit 53 shifts the forward portion command values ​​of the two phases (U-phase forward portion command value cuf and W-phase forward portion command value cwf) or the backward portion command values ​​of the two phases (U-phase backward portion command value cub and W-phase backward portion command value cwb) so that they match (become the same value).

[0080] In the example in Figure 6(C), the U-phase forward portion command value cuf and the W-phase forward portion command value cwf are shifted so that their command values ​​coincide around 0.5. Note that in Figure 6(C), for the sake of explanation, the U-phase forward portion command value cuf and the W-phase forward portion command value cwf are shown with a slight difference, but in this embodiment, when it is stated that "the command values ​​(for example, the U-phase forward portion command value cuf and the W-phase forward portion command value cwf) coincide," it is assumed that they are the same value (the same applies in the following drawings). In other words, in this case, both the inverted U-phase two-phase modulation command value 1-cu (shown by a small dashed line) and the W-phase two-phase modulation command value cw (shown by a small dashed line) before splitting are shifted in the direction of decrease. As a result, the timing of the rising edge command of the U-phase PWM pulse and the falling edge command of the W-phase PWM pulse coincide (the U-phase rising edge command value cuu and the W-phase falling edge command value cwd coincide). As a result, the timing of the rising edge of the U-phase voltage Vu and the falling edge of the W-phase voltage Vw can be synchronized, canceling out two zero-sequence voltage fluctuations.

[0081] Furthermore, the U-phase rear portion command value cub is shifted to maintain the U-phase PWM pulse width Pu before and after the shift of the U-phase front portion command value cuf. That is, the difference between the inverted U-phase two-phase modulation command value 1-cu before splitting and the U-phase front portion command value cuf after the shift (U-phase shift amount c) ushift ) is shifted in the opposite direction of increase / decrease. In other words, the U-phase backward portion command value cub is shifted from the inverted U-phase two-phase modulation command value 1-cu by a shift amount c ushift The value increases. If the U-phase trailing portion command value cub is greater than the triangular carrier wave CA, the U-phase PWM pulse falls (U-phase falling edge command value cud). In this example, the U-phase PWM pulse falls at the peak of the carrier period.

[0082] Similarly, the W-phase backward portion command value cwb is also shifted independently of the U-phase backward portion command value cub so as to maintain the W-phase PWM pulse width Pw before and after the shift of the W-phase forward portion command value cwf. That is, the difference between the W-phase two-phase modulation command value cw before splitting and the W-phase forward portion command value cwf after the shift (W-phase shift amount c) wshift The same amount is shifted in the opposite direction of increase or decrease (in this case, from the W-phase two-phase modulation command value cw to the increasing direction). If the W-phase trailing portion command value cwb is greater than the triangular carrier wave CA, the W-phase PWM pulse rises (W-phase rising edge command value cwu).

[0083] In this embodiment, the two-phase modulation command value of the two-phase PWM operation performed within the carrier period is divided into a forward portion command value and a backward portion command value. This makes it possible to match the two-phase forward portion command values ​​on the uphill slope of the triangular carrier wave CA, and to shift the two-phase backward portion command values ​​independently (to different values) on the downhill slope. Furthermore, it makes it possible to match the two-phase backward portion command values ​​on the downhill slope of the triangular carrier wave CA, and to shift the two-phase forward portion command values ​​independently (to different values) on the uphill slope.

[0084] By doing this, as shown in Figure (C), the timing of the rising edge command for the U-phase PWM pulse and the falling edge command for the W-phase PWM pulse can be synchronized (the U-phase rising edge command value cuu and the W-phase falling edge command value cwd are synchronized) while maintaining the U-phase PWM pulse width Pu and the W-phase PWM pulse width Pw (=Pw1 + Pw2). Alternatively, although not shown in the figure, the timing of the falling edge command for the U-phase PWM pulse and the rising edge command for the W-phase PWM pulse can be synchronized (the U-phase falling edge command value cud and the W-phase rising edge command value cwu are synchronized) while maintaining the U-phase PWM pulse width Pu and the W-phase PWM pulse width Pw.

[0085] Furthermore, the U-phase shift amount c in the case of Figure 6(C) ushift W-phase shift amount c wshift The U-phase rising edge command value cuu, the U-phase falling edge command value cud, the W-phase rising edge command value cwu, and the W-phase falling edge command value cwd are determined by the following equation (Equation 6).

number

[0086] Figure 7 shows another example of the PWM control method (command value inversion control, command value division control, and command value shift control) of this embodiment, and is a PWM pulse diagram within a carrier period, showing a carrier period (1 period) around 60 degrees in the two-phase modulation command waveform diagram (Figure 5(B), right diagram). In the PWM pulse diagram within the carrier period of Figure 7, some illustrations and explanations regarding the stop phase of the PWM operation (V phase in this example) are omitted (the same applies hereafter).

[0087] Figure 7(A) shows the U-phase PWM pulse and W-phase PWM pulse based on the inverted U-phase two-phase modulation command value 1-cu and the W-phase two-phase modulation command value cw in the two-phase modulation command waveform diagram (same as the right figure). In other words, it represents the state before the PWM pulse shift.

[0088] Figure 7(B) shows the state after command value division control and command value shift control have been performed. As already mentioned, the forward portion command value is a command value that can intersect at least the uphill slope of the triangular carrier wave CA, and the backward portion command value is a command value that can intersect at least the downhill slope of the triangular carrier wave CA. In other words, as shown in Figure 7(B), the division line DL in command value division control (the timing for switching between the forward portion command value and the backward portion command value) does not have to be limited to the center line of the carrier period passing through the peak of the triangular carrier wave CA. For example, the division line DL for the two-phase forward portion command values ​​(in this example, U-phase forward portion command value cuf, W-phase forward portion command value cwf) and backward portion command values ​​(in this example, U-phase backward portion command value cub, W-phase backward portion command value cwb) may be set behind the peak of the triangular carrier wave CA (to the right in the figure). Although not shown in the figure, the division line DL may also be set ahead of the peak of the triangular carrier wave CA (to the left in the figure).

[0089] Other control is the same as in Figure 6(C). That is, in this example, the U-phase forward portion command value cuf and the W-phase forward portion command value cwf are shifted so that they match on the uphill slope of the triangular carrier wave CA. The W-phase backward portion command value cwb and the U-phase backward portion command value cub are shifted independently (in the opposite direction to the forward portion) to maintain the U-phase PWM pulse width Pu and the W-phase PWM pulse width Pw, respectively. Here, in the example shown in Figure 7(B), although the W-phase two-phase modulation command value cw is divided into forward and backward portions, the W-phase rising edge command value cwu and the W-phase falling edge command value cwd are the same value. In other words, in this case, no shift is performed in the forward / backward (left / right in the diagram) direction for the W-phase PWM pulse. Thus, it is not necessary to shift the PWM pulse of one of the phases.

[0090] In contrast, in the example shown in Fig. 7(C), shifting is performed such that the U-phase front part command value cuf matches the W-phase front part command value cwf on the rising slope of the triangular carrier wave CA, and the W-phase rear part command value cwb and the U-phase rear part command value cub are each shifted independently (in the direction opposite to that of the front part) so as to maintain the U-phase PWM pulse width Pu and the W-phase PWM pulse width Pw (=Pw1+Pw2), respectively. This may in some cases expand the range in which the rising and falling timings of two-phase PWM pulses can be matched.

[0091] Note that when the dividing line DL is moved from the center of the carrier period as shown in Figs. 7(B) and 7(C), either the front part command value or the rear part command value becomes a command value straddling the first half period and the second half period. To explain with the examples of Figs. 7(B) and 7(C), since the front half command value (for example, the U-phase front part command value cuf) straddles the first half period and a part of the second half period, it may intersect both the first half period (rising slope) and the second half period (falling slope) (in this example, near the command value "1"). In such a case, the value on the second half period side (the value intersecting the falling slope) is ignored for the front half command value. Further, when the rear half command value straddles the second half period and a part of the first half period, and intersects both the first half period (rising slope) and the second half period (falling slope), the value on the first half period side (the value intersecting the rising slope) is ignored.

[0092] Note that the previous examples show the case where the rising edge of the U-phase PWM pulse is matched with the falling edge of the W-phase PWM pulse, but the present invention is not limited thereto. That is, the two-phase rear part command values may be matched on the falling slope, so that the falling edge of the U-phase PWM pulse is matched with the rising edge of the W-phase PWM pulse. Further, the above-described control may be performed with a combination of a U-phase PWM pulse and a V-phase PWM pulse, where the W-phase is set as a phase that stops PWM operation, and the U-phase and V-phase are set as phases that perform PWM operation.

[0093] <When simplifying the PWM control algorithm> As shown in Figure 7(C), if the U-phase forward command value cuf, the W-phase forward command value cwf, the U-phase backward command value cub, and the W-phase backward command value cwb are all shiftable, the flexibility of PWM control increases, but the PWM control algorithm becomes more complex.

[0094] Since the PWM pulse width directly affects the driving of the motor 8, avoiding errors and mistakes in the PWM control program is essential. For this reason, it is desirable to adopt the simplest possible PWM control algorithm for implementation in the product.

[0095] Figure 8 shows an example flowchart (Sequence 1) for implementing the PWM control of this embodiment using a relatively simple algorithm. This algorithm performs sequence switching control, which switches between multiple sequences (in this example, sequences 1A to 1F) with different PWM pulse shifting methods. The control device 21 determines a phase range (for example, 60 degrees) corresponding to the electrical angular phase of the motor 8 and performs sequence switching control for each phase range. Furthermore, it performs sequence switching control with a predetermined regularity across multiple phase ranges.

[0096] In Sequence 1, when shifting the PWM pulse, the forward partial command values ​​of the two phases are matched during the first half of the carrier cycle (uphill slope) (limiting the slope on which the partial command values ​​are matched to the uphill slope). Furthermore, during the second half of the carrier cycle (downhill slope), when the partial command values ​​(backward partial command values) are not matched, control is performed to set the backward partial command value to a predetermined value so as to stop the PWM operation of the power switching element of one of the two phase inverter circuits 28 (hereinafter referred to as "command value fixing"). The value set by command value fixing is 0 or 1.

[0097] First, in step S101, the two-phase modulation command values ​​cu, cv, and cw for each phase shown in Figure 4(C) are determined using equations (1) to (5) above. In the following step S102, the inverted U-phase two-phase modulation command value 1-cu is determined by command value inversion control. In step S103, it is determined whether the current electrical phase angle is in the lower fixed region (see Figure 5(B)). If it is in the lower fixed region, the process proceeds to step S104; otherwise, the process proceeds to step S113.

[0098] In step S104, it is determined whether the V phase is the lower fixed phase (see Figure 5(B)). If it is the lower fixed phase, the process proceeds to step S105; otherwise, it proceeds to step S109. In step S105, it is determined whether the U phase PWM pulse width Pu (duty cycle) and the W phase PWM pulse width Pw (duty cycle) are greater. If the U phase PWM pulse width Pu is greater than the W phase PWM pulse width Pw, the process proceeds to step S106; otherwise, the process proceeds to step S108.

[0099] In step S106, the W-phase PWM pulse is shifted so that the duty cycle of the W-phase PWM pulse in the latter half of the carrier cycle becomes 0%. Then, in step S107, the forward part command values ​​of both phases (U-phase forward part command value cuf and W-phase forward part command value cwf) are matched so that the rising timing of the U-phase PWM pulse and the falling timing of the W-phase PWM pulse coincide in the first half of the carrier cycle. In addition, the U-phase backward part command value cub and / or W-phase backward part command value cvb are shifted so that the U-phase PWM pulse width Pu and W-phase PWM pulse width Pw are maintained before and after the shift.

[0100] Furthermore, in step S108, which proceeds if the determination in step S105 is "No", the U-phase PWM pulse is shifted so that the duty cycle of the U-phase PWM pulse in the latter half of the carrier cycle becomes 0%, and then the process proceeds to step S107.

[0101] If the determination in step S104 is "No", step S109 proceeds to determine the magnitude of the U-phase PWM pulse width Pu (duty cycle) and the V-phase PWM pulse width Pv (duty cycle). If the U-phase PWM pulse width Pu is greater than the V-phase PWM pulse width Pv, proceed to step S110; otherwise, proceed to step S112.

[0102] In step S110, the V-phase PWM pulse is shifted so that the duty cycle of the V-phase PWM pulse in the latter half of the carrier cycle becomes 0%. Then, in step S111, the forward part command values ​​of both phases (U-phase forward part command value cuf and V-phase forward part command value cvf) are matched so that the rising timing of the U-phase PWM pulse and the falling timing of the V-phase PWM pulse coincide in the first half of the carrier cycle. In addition, the U-phase backward part command value cub and / or V-phase backward part command value cvb are shifted so that the U-phase PWM pulse width Pu and V-phase PWM pulse width Pv are maintained before and after the shift.

[0103] Furthermore, in step S112, which proceeds if the determination in step S109 is "No", the U-phase PWM pulse is shifted so that the duty cycle of the U-phase PWM pulse in the latter half of the carrier cycle becomes 0%, and then the process proceeds to step S111.

[0104] If the determination in step S103 is "No", step S113 proceeds to determine whether the V phase is the upper fixed phase. If the V phase is the upper fixed phase, proceed to step S114; otherwise, proceed to step S116. In step S114, the W phase PWM pulse is shifted so that the duty cycle of the W phase PWM pulse in the latter half of the carrier cycle becomes 100%. Then, in step S115, the forward part command values ​​of both phases (U phase forward part command value cuf and W phase forward part command value cwf) are matched so that the rising timing of the U phase PWM pulse and the falling timing of the W phase PWM pulse coincide in the first half of the carrier cycle. In addition, the U phase backward part command value cub and / or W phase backward part command value cwb are shifted so that the U phase PWM pulse width Pu and W phase PWM pulse width Pw are maintained before and after the shift.

[0105] Furthermore, in step S116, which proceeds if the determination in step S113 is "No", the V-phase PWM pulse is shifted so that the duty cycle of the V-phase PWM pulse in the latter half of the carrier cycle becomes 100%, and then the process proceeds to step S117. In step S117, the forward portion command values ​​of both phases (U-phase forward portion command value cuf and V-phase forward portion command value cvf) are matched so that the rising timing of the U-phase PWM pulse and the falling timing of the V-phase PWM pulse coincide in the first half of the carrier cycle. In addition, the U-phase backward portion command value cub and / or V-phase backward portion command value cvb are shifted so that the U-phase PWM pulse width Pu and V-phase PWM pulse width Pv are maintained before and after the shift.

[0106] Sequence 1 includes six types of control, from Sequence 1A to 1F, and this sequence switching control is executed within a phase range (for example, every 60 degrees). Each control in Sequence 1 will be explained in detail. Figure 9 is an example of control by Sequence 1A, Figure 10 is an example of control by Sequence 1B, and Figure 11 is an example of control by Sequence 1E. Figures 9 to 11 are PWM pulse diagrams within the carrier period, but for the sake of explanation, only the necessary waveforms are extracted and shown. Figures 9(A), 10(A), and 11(A) show the U-phase PWM pulse and W-phase PWM pulse before the shift based on the two-phase modulation command waveform diagram (Figure 5(B)), and Figures 9(B), 10(B), and 11(B) show the U-phase PWM pulse and W-phase PWM pulse after the PWM control (end of Sequence 1) in this embodiment.

[0107] Referring to Figure 9, sequence 1A is executed when the electrical angular phase is in the range of, for example, 30deg to 90deg, as shown by the dashed frame in the right diagram of Figure 9(A). Specifically, the process proceeds through steps S101 to S103 in the flowchart shown in Figure 8, where the determination in step S103 is "Yes" (lower fixed region), the determination in step S104 is "Yes" (V phase is the lower fixed phase), and the determination in step S105 is "Yes" (U phase PWM pulse width (duty cycle) > W phase PWM pulse width (duty cycle)) as shown in the left diagram of Figure 9(A).

[0108] In this case, the inverted U-phase two-phase modulation command value 1-cu is divided into the U-phase forward portion command value cuf and the U-phase backward portion command value cub, and the W-phase two-phase modulation command value cw is divided into the W-phase forward portion command value cwf and the W-phase backward portion command value cwb. Then, in step S106, the duty cycle of the W-phase PWM pulse is set to 0% in the latter half of the carrier cycle. That is, the W-phase PWM pulse that is in the latter half of the carrier cycle (right end) in the left diagram of Figure 9(A) is moved to the first half of the cycle as shown in Figure 9(B) (the W-phase PWM pulse width Pw in the first half of the cycle widens towards the center of the carrier cycle). Then, in step S107, the timing of the rising edge of the U-phase PWM pulse and the falling edge of the W-phase PWM pulse are matched in the first half of the carrier cycle. The shifted pulse width of the W-phase PWM pulse (W-phase PWM pulse width Pw) is maintained from the pulse width before the shift (the sum of W-phase PWM pulse widths Pw1 and Pw2) (Pw = Pw1 + Pw2). Furthermore, the U-phase PWM pulse shifts the U-phase trailing command value cub so that its pulse width (U-phase PWM pulse width Pu) does not change.

[0109] Shifting the W-phase PWM pulse to the earlier period in the latter half of the carrier cycle so that the duty cycle becomes 0% means setting the W-phase trailing command value cwb to 0 (fixing the command value for the W-phase). In program calculations and other control operations, the W-phase trailing command value cwb becomes 0 (the reference value), so the W-phase shift amount c wshift (and U-phase shift amount c) ushift This simplifies the calculation (control) of other numerical values.

[0110] Referring to Figure 10, Sequence 1B is executed when the electrical angular phase is in the range of, for example, 330deg to 30deg, as shown by the dashed frame in the right-hand diagram of Figure 10(A). Specifically, the process proceeds through steps S101 to S105 of the flowchart shown in Figure 8, and in step S108, the duty cycle of the U-phase PWM pulse is set to 0% in the latter half of the carrier cycle. In this case, the duty cycle of the U-phase PWM pulse is set to 0% in the latter half of the cycle by shifting it forward so that the trailing end of the U-phase PWM pulse (the rightmost end in the diagram), i.e., the falling edge timing, is located at the center of the carrier cycle. Then, in step S107, the timing of the rising edge of the U-phase PWM pulse and the falling edge of the W-phase PWM pulse are synchronized in the first half of the carrier cycle. The U-phase PWM pulse width Pu and the W-phase PWM pulse width Pw (=Pw1+Pw2) are maintained before and after the shift of the PWM pulses.

[0111] Shifting the U-phase PWM pulse to the earlier period in the latter half of the carrier cycle so that the duty cycle becomes 0% means setting the U-phase trailing command value cub to 1 (fixing the command value for the U-phase). This results in a U-phase shift amount c ushift (and W-phase shift amount c) wshift This simplifies the calculation (control) of other numerical values.

[0112] Referring to Figure 11, sequence 1E is executed when the electrical angular phase is in the range of, for example, 210deg to 270deg, as shown by the dashed frame in the right-hand diagram of Figure 11(A). Specifically, the sequence proceeds through steps S101 to S103 and S113 of the flowchart shown in Figure 8, and in step S114, the W-phase PWM pulse is set to a duty cycle of 100% in the latter half of the carrier cycle. In this case, the W-phase PWM pulse in the latter half of the cycle has its leading edge (rising edge timing) shifted toward the center of the carrier cycle, and the duty cycle is set to 100% in the latter half of the cycle. Then, in step S115, the timing of the rising edge of the U-phase PWM pulse and the falling edge of the W-phase PWM pulse are matched in the first half of the carrier cycle, and the U-phase trailing portion command value cuf is shifted so that the U-phase PWM pulse width Pu is maintained. The U-phase PWM pulse width Pu and the W-phase PWM pulse width Pw (=Pw1 + Pw2) are maintained before and after the shift of the PWM pulses.

[0113] Shifting the W-phase PWM pulse to the earlier period in the latter half of the carrier cycle so that the duty cycle is 100% means setting the W-phase trailing command value cwb to 1 (fixing the command value for the W-phase). This results in a U-phase shift amount c ushift (and W-phase shift amount c) wshift This simplifies the control of calculations involving other numerical values.

[0114] Figure (C) shows the case where the W-phase PWM pulse has a duty cycle of 0% in the latter half of the carrier cycle in step S114. When the electrical angular phase is around 240 degrees, the W-phase PWM pulse is larger than the U-phase PWM pulse, and the duty cycle is 50% or more (approximately 0.75 in command value), so it is not possible to set the duty cycle to 0% in the latter half of the carrier cycle (the hatched area does not fit within the first half of the cycle). Therefore, in sequence 1E, the W-phase PWM pulse has a duty cycle of 100% in the latter half of the carrier cycle.

[0115] In other words, as described in this embodiment, if the sequence is simplified on the premise that the duty cycle of the latter half of the period is output as either 0% or 100%, the determination for assigning the two phases performing PWM operation to sequences 1A to 1F (determination in steps S105 and S109) may be made not by comparing the duty cycle of the W-phase PWM pulse with that of the V-phase PWM pulse, but by determining whether the duty cycle of the U-phase PWM pulse exceeds 50% (if the duty cycle of the U-phase PWM > 50%, then "Yes"). This is because if the duty cycle of the U-phase exceeds 50%, the PWM pulse will not fit within the first half (or second half) of the carrier period, as shown in Figure 11(C).

[0116] Although not shown in the diagram, sequence 1C is a different pattern from sequence 1A in terms of phase. In other words, in sequence 1A, the phases performing PWM operation were the U phase and the W phase, but in sequence 1C, the W phase is the stop phase for PWM operation, and the phases performing PWM operation are the U phase and the V phase. However, the waveform is the same as in sequence 1A, and the control to match the timing of the rising edge of the U phase PWM pulse and the falling edge of the V phase PWM pulse is the same as in sequence 1A.

[0117] Furthermore, sequence 1D is a pattern with a different phase than sequence 1B (the phases performing PWM operation are the U phase and the V phase), but otherwise it is the same as sequence 1B.

[0118] Furthermore, sequence 1F is a pattern with a different phase from sequence 1E (the phases performing PWM operation are the U phase and the V phase), but otherwise it is the same as sequence 1E.

[0119] Thus, in Sequence 1, it is sufficient to prepare six different processes (controls) from Sequence 1A to Sequence 1F. Also, as shown in steps S106, S108, S110, S112, S114, and S116, a common flow is to set the duty cycle of the PWM pulse in the latter half of a specific phase to 0% or 100% depending on the conditions, and then synchronize the rising and falling timings of the two-phase PWM pulses in the first half of the cycle. Therefore, the algorithm can be made relatively simple, and implementation in products is easy.

[0120] However, when PWM control is performed using this sequence 1, periodic current distortion may occur.

[0121] Figure 12 shows the results of PWM control using sequence 1, with Figure 12(A) being the simulation result of the output voltage of power converter 1 (U-phase voltage Vu, V-phase voltage Vv, W-phase voltage Vw). Figure 12(B) shows the current waveforms for each phase. The dashed-dotted lines in Figure 12 indicate the switching timing of sequences 1A to 1F, which occur every 1 / 6 period. This periodic current distortion may cause torque pulsation and noise in the motor 8.

[0122] Figure 13 is a graph showing the state of the triangular carrier wave CA with an electrical phase angle from 0 to 50 degrees, the two-phase modulation command values ​​cu (solid line), cv (dotted line), cw (dashed line) for each phase, the phase voltages Vu, Vv, Vw, and the U-phase current iu (thick line), V-phase current iv (medium-thick line), and W-phase current iw (thin line) when PWM control is performed according to sequence 1. The two-phase modulation command values ​​cu, cv, and cw for each phase are, strictly speaking, the waveforms shown as the forward and backward portion command values ​​for each phase in Figures 6, 7, 9 to 11, etc.

[0123] These results clearly show that a positional shift occurs in the pulses of the two-phase modulation command values ​​cu, cv, and cw around the timing of 30 degrees (the dashed rectangle in Figure 13). As a result, the pulse width (ON time) of the U-phase voltage Vu becomes exceptionally long, and the U-phase current iu operates linearly without fluctuation. This pulse positional shift is due to the fact that in sequence 1, in order to simplify control, the duty cycle of one phase is always set to 0% or 100% (the command value is fixed) in the latter half of the carrier cycle (in any of sequences 1A to 1F).

[0124] Specifically, Figure 14 is a PWM pulse diagram within the carrier period showing the state where the sequence switches around 30 degrees. Figure 14(A) shows the state before the PWM pulse shift, Figure 14(B) shows the state when sequence 1B is being executed (near the end), Figure 14(C) shows the state when sequence 1A is being executed (immediately after the start), and Figure 14(D) shows the transition from Figure 15(B) to Figure 15(C).

[0125] In sequence 1, around 30 degrees, a sequence switch occurs from sequence 1B (step S108) to sequence 1A (step S106), depending on the duty cycle of the U-phase PWM pulse (and its relationship to the duty cycle of the W-phase PWM pulse).

[0126] In this case, in sequence 1B (step S108), the U-phase rear portion command value cub is fixed at 1, as shown in Figure 14(B). When the duty cycles of the U-phase PWM pulse and the W-phase PWM pulse are swapped from this state, the system proceeds to step S106 based on the determination in step S105 of sequence 1, and switches to sequence 1A. In other words, the PWM pulse state changes from the state shown in Figure 14(B) to the state shown in Figure 14(C). As a result, Figure 14(B) is output in the first half of the period, and Figure 14(C) is output in the second half of the period, resulting in a PWM pulse like that shown in Figure 14(D). In the case of Figure 14(D), the first half of the period corresponds to the PWM pulse of the first half of the period in Figure 14(B), and the second half of the period in Figure 14(D) corresponds to the PWM pulse of the second half of the period in Figure 14(C) (corresponding to the dashed box in Figure 13). In this case, although the zero-sequence voltage fluctuations are canceled out, when viewed over the carrier period, the U-phase is output with a duty cycle of almost 100%, and the W-phase is output with a duty cycle of almost 0%, which differs from the originally targeted values.

[0127] In other words, when controlling the switching of sequences 1A to 1F (sequence switching control), it is desirable to avoid fixing the command value in the same half-period (second half-period) of the carrier cycle before and after the switching, especially when the relationship between the duty cycles of the U-phase PWM pulse and the V-phase or W-phase PWM pulse is reversed (or when the duty cycle of one phase PWM pulse exceeds 50%).

[0128] Therefore, in this embodiment, in addition to sequence 1, sequence 2 shown in Figure 15 is also adopted, and the system switches from sequence 1 to sequence 2 at timings where there is a risk of pulse misalignment.

[0129] Referring to Figure 15, Sequence 2 includes multiple control methods with different PWM pulse shifting methods (six types in this example, Sequences 2A to 2F), and this sequence switching control is executed according to the phase range. Sequence 2 also fixes the duty cycle of one phase's PWM pulse to 0% or 100% in the first half of the carrier cycle (i.e., the command value is fixed in the first half of the cycle), and synchronizes the rising and falling timings of the two phases' PWM pulses in the second half of the cycle. In other words, the difference from Sequence 1 is that the half-cycle in which the duty cycle of the PWM pulse is fixed to 0% or 100% (command value fixed) is the first half of the cycle.

[0130] Specifically, referring to Figure 15, we will explain the steps that differ from Sequence 1. Steps S201 to S205 are the same as in Sequence 1. In step S205, the duty cycle of the W-phase PWM pulse is compared with that of the U-phase PWM pulse. If the duty cycle of the U-phase PWM pulse is larger, the process proceeds to step S206, where the W-phase PWM pulse is shifted so that its duty cycle in the first half of the carrier cycle becomes 0%. Then, in step S207, the backward portion command values ​​of both phases (U-phase backward portion command value cub and W-phase backward portion command value cwb) are matched so that the falling timing of the U-phase PWM pulse and the rising timing of the W-phase PWM pulse coincide in the second half of the carrier cycle. Additionally, the forward portion command value cuf and / or W-phase forward portion command value cwf are shifted so that the duty cycles of the U-phase PWM pulse and W-phase PWM are maintained before and after the shift.

[0131] Furthermore, in step S208, which proceeds if the determination in step S205 is "No", the U-phase PWM pulses are shifted so that the duty cycle of the U-phase PWM pulses in the first half of the carrier cycle becomes 0%, and then the process proceeds to step S207.

[0132] In step S209, the duty cycle of the V-phase PWM pulse is compared with that of the U-phase PWM pulse. If the duty cycle of the U-phase PWM pulse is large, in step S210, the V-phase PWM pulse is shifted so that the duty cycle of the V-phase PWM pulse in the first half of the carrier cycle becomes 0%. Then, in step S211, the backward portion command values ​​of both phases (U-phase backward portion command value cub and V-phase backward portion command value cvb) are matched so that the falling timing of the U-phase PWM pulse and the rising timing of the V-phase PWM pulse coincide in the second half of the carrier cycle. In addition, the forward portion command value of the U-phase PWM pulse cuf and / or the forward portion command value of the V-phase PWM pulse cvf are shifted so that the duty cycles of the U-phase PWM pulse and the V-phase PWM pulse are maintained before and after the shift.

[0133] Furthermore, in step S212, which proceeds if the determination in step S209 is "No", the U-phase PWM pulses are shifted so that the duty cycle of the U-phase PWM pulses in the first half of the carrier cycle becomes 0%, and then the process proceeds to step S211.

[0134] In step S214, the W-phase PWM pulse is shifted so that the duty cycle of the W-phase PWM pulse in the first half of the carrier cycle becomes 100%. Then, in step S215, the backward portion command values ​​of both phases (U-phase backward portion command value cub and W-phase backward portion command value cwb) are matched so that the falling timing of the U-phase PWM pulse and the rising timing of the W-phase PWM pulse coincide in the second half of the carrier cycle. In addition, the forward portion command value of the U-phase PWM pulse cuf and / or the forward portion command value of the W-phase PWM pulse cwf are shifted so that the duty cycles of the U-phase PWM pulse and the W-phase PWM are maintained before and after the shift.

[0135] In step S216, the V-phase PWM pulse is shifted so that the duty cycle of the V-phase PWM pulse in the first half of the carrier cycle becomes 100%. In step S217, the backward portion command values ​​of both phases (U-phase backward portion command value cub and W-phase backward portion command value cwb) are matched so that the falling timing of the U-phase PWM pulse and the rising timing of the W-phase PWM pulse coincide in the second half of the carrier cycle. The forward portion command value of the U-phase PWM pulse cuf and / or the forward portion command value of the V-phase PWM pulse cvf are also shifted so that the duty cycles of the U-phase PWM pulse and the V-phase PWM are maintained before and after the shift.

[0136] Thus, in the control of sequence 2, the control device 21 fixes the command value (controls to fix the command value to 0 or 1) for the phase with the smaller duty cycle among the two phase PWM pulses. Note that, as described in this embodiment, if the sequence is simplified on the premise that the duty cycle of the first half of the cycle is output as either 0% or 100%, then in sequence 2, the determination in steps S205 and S209 may be made not by comparing with the duty cycle of the W-phase PWM pulse or the V-phase PWM pulse, but by determining whether the duty cycle of the U-phase PWM pulse exceeds 50% (if the duty cycle of the U-phase PWM > 50%, the determination is "Yes").

[0137] Figure 16 is a carrier-period PWM pulse diagram showing an example of control of sequence 2. Figures (A) and (B) are examples of control by sequence 2A, with (A) showing the state before shifting the PWM pulse and (B) showing the state after shifting the PWM pulse. Figures (C) and (D) are examples of control by sequence 2B, with (C) showing the state before shifting the PWM pulse and (D) showing the state after shifting the PWM pulse. Figures (E) and (F) are examples of control by sequence 2E, with (E) showing the state before shifting the PWM pulse and (F) showing the state after shifting the PWM pulse. Figure 16(B) corresponds to Figure 9(B), Figure 16(D) corresponds to Figure 10(B), and Figure 16(F) corresponds to Figure 11(B).

[0138] Referring to Figure 16(A), sequence 2A is executed when the electrical angular phase is in the range of, for example, 30deg to 90deg. Specifically, the process proceeds through steps S201 to S205 of the flowchart shown in Figure 15, and in step S206, the duty cycle of the W-phase PWM pulse is set to 0% in the first half of the carrier cycle. That is, the W-phase PWM pulse present in the first half of the carrier cycle (left end) in Figure 16(A) is moved to the second half of the cycle as shown in Figure 16(B). Then, in step S207, the falling timing of the U-phase PWM pulse and the rising timing of the W-phase PWM pulse are synchronized in the second half of the carrier cycle.

[0139] Shifting the W-phase PWM pulse to the second half of the carrier cycle so that its duty cycle becomes 0% in the first half of the cycle means fixing the command value (value "0") for the W-phase (W-phase forward command value cwf). By shifting the U-phase forward command value cuf and / or the V-phase forward command value cvf, the duty cycles of the U-phase PWM pulse and W-phase PWM pulse are maintained before and after the PWM pulse shift (the same applies below).

[0140] Referring to Figure 16(D), sequence 2B is executed when the electrical angular phase is, for example, in the range of 330deg to 30deg. Specifically, the process proceeds through steps S201 to S205 of the flowchart shown in Figure 15, and in step S208, the duty cycle of the U-phase PWM pulse is set to 0% in the first half of the carrier cycle. In this case, the U-phase PWM pulse is shifted backward so that the leading edge of the U-phase PWM pulse (the left end in the figure), i.e., the rising edge timing, is located at the center of the carrier cycle, and the duty cycle is set to 0% in the first half of the cycle. Then, in step S207, the timing of the falling edge of the U-phase PWM pulse and the rising edge of the W-phase PWM pulse are synchronized in the second half of the carrier cycle.

[0141] Shifting the U-phase PWM pulse to the latter half of the carrier cycle so that the duty cycle becomes 0% in the first half of the cycle means fixing the command value (value "1") for the U-phase (U-phase forward portion command value cuf).

[0142] Referring to Figure 16(F), sequence 2E is executed when the electrical angular phase is in the range of, for example, 210deg to 270deg. Specifically, the process proceeds through steps S201 to S203 and step S213 of the flowchart shown in Figure 15, and in step S214, the duty cycle of the W-phase PWM pulse is set to 100% in the first half of the carrier cycle. In this case, the W-phase PWM pulse in the first half of the cycle has its trailing end (falling edge timing) shifted toward the center of the carrier cycle, and the duty cycle is set to 100% in the first half of the cycle. Then, in step S215, the timing of the falling edge of the U-phase PWM pulse and the rising edge of the W-phase PWM pulse are matched in the second half of the carrier cycle.

[0143] Shifting the W-phase PWM pulse to the latter half of the carrier cycle so that the duty cycle is 100% in the first half of the cycle means fixing the command value (value "1") for the W-phase (W-phase forward command value cwf).

[0144] Although not shown in the diagram, sequence 2C has a different phase pattern than sequence 2A. In other words, while sequence 2A uses the U and W phases for PWM operation, in sequence 2C the W phase is the stop phase for PWM operation, and the phases used for PWM operation are the U and V phases. However, the waveform is the same as sequence 2A, and the control to synchronize the rising edge of the U phase PWM pulse with the falling edge of the V phase PWM pulse is the same as in sequence 2A.

[0145] Furthermore, sequence 2D is a pattern with a different phase than sequence 2B (the phases performing PWM operation are the U phase and the V phase), but otherwise it is the same as sequence 2B.

[0146] Furthermore, sequence 2F is a pattern with a different phase from sequence 2E (the phases performing PWM operation are the U phase and the V phase), but otherwise it is the same as sequence 2E.

[0147] As shown in Figures 9(B), 10(B), and 11(B), Sequence 1 matches the forward portion command values ​​of the two phases in the first half of the period and compares the upward slope of the triangular carrier wave CA with the forward portion command value. On the other hand, the second half of the period is treated as a command value mismatch half-period (the backward portion command values ​​are not matched in the second half of the period), and one of the two phase command values ​​is fixed to 0 or 1 (command value fixed). This control method, which matches the forward portion command values ​​of the two phases in the first half of the carrier period (upward slope) and fixes the command value in the second half of the carrier period, is hereafter referred to as "upward matching control".

[0148] On the other hand, as shown in Figures 16(B), 16(D), and 16(E), sequence 2 matches the trailing portion command values ​​of the two phases in the latter half of the period and compares the descending slope of the triangular carrier wave CA with the trailing portion command values. On the other hand, the first half of the period is treated as a half-period of mismatched command values ​​(the trailing portion command values ​​are not matched in the first half of the period), and one of the two phase command values ​​is fixed to 0 or 1 (fixed command value). This type of control, in which the trailing portion command values ​​of the two phases are matched in the latter half of the carrier period (descending slope) and the command value is fixed in the first half of the carrier period, is called "descending match control".

[0149] The control device 21 of this embodiment performs sequence switching control to switch between multiple sequences with different PWM pulse shifting methods. Sequence switching control is performed, for example, at phase ranges (e.g., 60 degrees) corresponding to the electrical angular phase of the motor 8. Furthermore, sequence switching control is performed with a predetermined regularity within multiple phase ranges.

[0150] As sequence switching control, there are two types of control: one in which the command value is fixed in the latter half of the cycle (upward matching control, sequence 1 above), and another in which the command value is fixed in the first half of the cycle (downward matching control; sequence 2 above). Upward matching control includes multiple sequences, and downward matching control also includes multiple sequences. In other words, in the example above, 6 sequences of upward matching control and 6 sequences of downward matching control (12 sequences in total) are selected and switched appropriately for each phase range in order to ensure good continuity of switching.

[0151] Furthermore, the control device 21 determines whether the duty cycle of one phase exceeds 50% during the carrier cycle, and if it is 50% or more (or exceeds 50%), it performs a control that switches the period for fixing the command value between the first half of the cycle and the second half of the cycle (i.e., switches between uplink matching control and downlink matching control) (hereinafter referred to as "command value fixing period switching control").

[0152] Alternatively, the control device 21 determines whether the phase with the short pulse width changes during the carrier cycle, and if the phase with the short pulse width changes during the carrier cycle, it performs command value fixed switching control.

[0153] Specifically, if the duty cycle of a PWM pulse in a certain phase exceeds 50%, or if the phase with a short PWM pulse width changes, the system switches between uplink matching control (one of the sequences in Sequence 1) and downlink matching control (one of the sequences in Sequence 2). In other words, if uplink matching control (e.g., Sequence 1B) was performed in the carrier cycle before the switch, the same control will not be performed after the switch, and downlink matching control (e.g., Sequence 2B) will be performed instead. Also, if downlink matching control (e.g., Sequence 2E) was performed in the carrier cycle before the switch, the same control will not be performed after the switch, and uplink matching control (e.g., Sequence 1F) will be performed instead.

[0154] The control device 21 performs command value fixed switching control when at least the phase with the shortest PWM pulse width changes, or when the duty cycle of one of the phases (the phase with a duty cycle of less than 50% (50% or less)) exceeds 50% (becomes 50% or more) during the carrier cycle.

[0155] Figure 17 is a schematic diagram showing the phase range for switching control. In this example, the phase range is set to every 60 degrees of electrical phase angle, and each time the phase range changes, up-shift matching control (one of sequences 1A to 1F) and down-shift matching control (one of sequences 2A to 2F) are performed alternately.

[0156] In this example, the electrical angular phase range of 0 to 30 degrees is defined as up-phase matching control (e.g., sequence 1B), the range of 30 to 90 degrees is defined as down-phase matching control (e.g., sequence 2A), the range of 90 to 150 degrees is defined as up-phase matching control (e.g., sequence 1C), the range of 150 to 210 degrees is defined as down-phase matching control (e.g., sequence 2D), the range of 210 to 270 degrees is defined as up-phase matching control (e.g., sequence 1E), the range of 270 to 330 degrees is defined as down-phase matching control (e.g., sequence 2F), and the range of 330 to 30 degrees is defined as up-phase matching control (e.g., sequence 1B).

[0157] For example, if we focus on the area around the peak of the sinusoidal command value waveform, since all waveforms are symmetrical before and after switching, it can be assumed that the pulse position shift is also symmetrical. Therefore, according to the above embodiment, accurate correction can be achieved. Note that the phase range is set to 60 degrees in this example, but it is not limited to this range.

[0158] Figure 18 is a graph showing the state of the triangular carrier wave CA with an electrical phase angle from 0 to 50 degrees, the two-phase modulation command values ​​cu (solid line), cv (dotted line), cw (dashed line) for each phase, the phase voltages Vu, Vv, Vw, and the U-phase current iu (thick line), V-phase current iv (medium-thick line), and W-phase current iw (thin line) when the sequence switching control shown in Figure 17 is performed, and corresponds to the results in Figure 13.

[0159] Furthermore, Figure 19 is a diagram corresponding to Figures 14(B) to 14(D), showing the state of sequence switching control around 30 degrees.

[0160] In the switching control of this embodiment, around 30 degrees, the system switches from sequence 1B (step S108) to sequence 2A (step S206) according to the duty cycle value of the U-phase PWM pulse (its relationship to the duty cycle of the W-phase PWM pulse). As a result, Figure 19(A) is output in the first half of the cycle, and Figure 19(B) is output in the second half of the cycle, resulting in the output of a PWM pulse like that shown in Figure 19(C). In Figure 19(C), the first half of the cycle corresponds to the PWM pulse of the first half of the cycle shown in Figure 19(A), and the second half of the cycle shown in Figure 19(C) corresponds to the PWM pulse of the second half of the cycle shown in Figure 19(B) (corresponding to the dashed box in Figure 18). In Figure 19(C), the zero-sequence voltage fluctuations are canceled out in both the first and second half of the cycle, and the duty cycles of each output phase are as initially intended. This makes it possible to suppress the current fluctuations that occurred in Figure 14.

[0161] Furthermore, Figure 20 shows the execution result of PWM control when sequence switching control including the command value fixed switching control shown in Figure 17 is performed, and Figure 20(A) is the simulation result of the output voltage of power converter 1 (U-phase voltage Vu, V-phase voltage Vv, W-phase voltage Vw). Figure 20(B) shows the current waveform of each phase. In this case, it can be clearly seen that the periodically occurring current distortion has been reduced compared to Figure 12(B).

[0162] The control device 21 performs command value fixed switching control when at least the phase with the shortest PWM pulse width changes, or when the duty cycle of one phase exceeds 50% (becomes 50% or more) during the carrier cycle.

[0163] In other words, the switching of command value fixed switching control (upstream matching control and downstream matching control) only needs to be performed when the duty cycle of one phase is 50% or more (or exceeds 50%) during the carrier cycle, or when the phase with the shorter pulse width changes during the carrier cycle.

[0164] Furthermore, in the above embodiment, the control device 21 determines a phase range (for example, 60 degrees) corresponding to the electrical angular phase of the motor 8, and executes sequence switching control for each phase range. In addition, sequence switching control is executed with a predetermined regularity in multiple phase ranges. Then, command value fixed switching control (upward matching control and downward matching control) is performed in synchronization with the timing of a certain sequence switching control. For example, in the example shown in Figure 17 above, command value fixed switching control (upward matching control and downward matching control) is switched every predetermined phase range (for example, 60 degrees).

[0165] However, the fixed command value switching control (switching between up-match control and down-match control) may be performed at intervals corresponding to the electrical angular phase of the motor 8 (e.g., 60 degrees), or it may not be performed at intervals corresponding to each phase range. For example, as shown in Figure 21, the fixed command value switching control (up-match control and down-match control) may be performed only when the duty cycle of one phase is 50% or more (or exceeds 50%) in the carrier cycle, or when the phase with the shorter pulse width changes in the carrier cycle. In the example shown in Figure 21, the phase ranges 30ded to 90deg and 90deg to 150deg do not change the relative duty cycles of the two-phase PWM pulses (the phase with the shorter PWM pulse width does not change), so the fixed command value switching control is not performed, and only sequence switching control is performed. The same applies to the phase ranges 210ded to 270deg and 270deg to 330deg.

[0166] As described above, according to this embodiment, in PWM control using a triangular carrier wave CA, the PWM operation of one phase is first stopped by two-phase modulation, and the switching count of the upper and lower arm switching elements 18A to 18F is reduced. This reduces switching loss and heat generation, and also reduces fluctuations in the neutral point potential of the motor 8.

[0167] In addition, the control device 21 inverts the two-phase modulation command value of one of the two phases performing PWM operation (the U phase in this example) (command value inversion control), and divides the two-phase modulation command values ​​of the two phases performing PWM operation (for example, the inverted U-phase two-phase modulation command value 1-cu and the W-phase two-phase modulation command value cw) into two parts, one before and one after the carrier period (command value division control), and sets the two-phase modulation command values ​​independently for the uphill and downhill sides (of each carrier period) of the triangular carrier wave CA. Then, in either the uphill side (first half of the period) or the downhill side (second half of the period) of the carrier period, the divided two-phase command values ​​are made to match. That is, in the above example, the U-phase forward portion command value cuf and the V-phase forward portion command value cvf are made to match on the uphill side of the carrier period (command value shift control). Furthermore, in the half-period where the command values ​​are not matched, the PWM pulses are shifted so that the two-phase PWM pulse width does not change from before the shift.

[0168] This allows the rising edge timing of one phase's PWM pulse to be synchronized with the falling edge timing of the other phase's PWM pulse while maintaining the PWM pulse width (duty cycle) of each of the two phases performing PWM operation.

[0169] As a result, the rising timing of one phase voltage and the falling timing of the other phase are synchronized for the two phases that modulate the ON / OFF state (the two phases that perform PWM operation), so that changes in the phase voltage applied to the motor 8 can be canceled out by changes in the other phase voltage. Specifically, two zero-sequence voltage fluctuations can be canceled out in one carrier cycle. In normal three-phase modulation, six zero-sequence voltage fluctuations occur in one carrier cycle, but according to this embodiment, this can be reduced to 1 / 3 (2 times). Therefore, fluctuations in the neutral point potential of the motor can be further suppressed by the switching timing of the switching elements, and the generation of common-mode noise can be significantly suppressed.

[0170] Furthermore, when implementing this in a product, it is desirable to adopt a relatively simple algorithm to avoid errors in program control. In this embodiment, 12 sequence patterns are prepared, and the 12 sequence switching control patterns are performed for each phase range of the electrical phase angle. The 12 sequence patterns all share a unified method (algorithm) in which one of the two phases is commanded to 0 or 1 for half a period of the triangular carrier wave, and the rising and falling timings of the two-phase PWM pulses are matched for the remaining half a period. This significantly reduces errors in program control and facilitates implementation in products.

[0171] Furthermore, in particular, if the duty cycle of one phase is 50% or more (or exceeds 50%) during the carrier cycle, or if the phase with the shortest pulse width changes during the carrier cycle, command value fixed switching control (switching between uplink matching control and downlink matching control) is performed. This avoids current distortion that occurred with sequence switching. As a result, torque pulsation and noise generation of motor 8 can be avoided.

[0172] In the above embodiment, the command value inversion unit 51, the command value division unit 52, and the command value shift unit 53 may be composed of either software or hardware, or both software and hardware. Furthermore, these are not limited to configurations included in the PWM signal generation unit 36, but may be configurations included in the control device 21.

[0173] Furthermore, the configuration of part or all of the control device 21 may consist of either software or hardware, or both software and hardware.

[0174] Alternatively, the W and V phases may be inverted instead of the U phase. Furthermore, the phase to be inverted does not necessarily have to be the U phase.

[0175] Furthermore, in the sequences 1 and 2 shown in this embodiment, the configuration prioritizes ease of implementation as described above. To further simplify the configuration, the forward or backward command value is set to a predetermined value (fixed command value) so that the power switching element of the inverter circuit 28 of one of the two phases performing PWM operation stops PWM operation. Specifically, the command value is fixed so that the duty cycle of the PWM pulse in the first or second half of the carrier cycle is 0% or 100%.

[0176] Here, the command value fixing may be controlled so that in one of the two phases performing PWM operation, the duty cycle of the power switching element of the inverter circuit 28 is set to an arbitrary width according to the phase voltage command value amplitude modulation rate kH or the phase voltage command value amplitude Vm. Specifically, the command value may be fixed so that the duty cycle of the PWM pulse in the first or second half of the carrier cycle is an arbitrary width from 0 to 100%. By changing this arbitrary width with respect to the electrical angular phase and modulation rate, the output modulation rate can be improved. On the other hand, in that case, the probability of a PWM pulse misalignment occurring as shown in Figure 14(D) above naturally increases. Therefore, if the duty cycle of the PWM pulse in the first or second half of the carrier cycle is fixed to an arbitrary width, there is still a possibility of current distortion occurring. In other words, a configuration in which the command value is fixed so that the duty cycle of the PWM pulse in the first or second half of the carrier cycle is 0% or 100%, as in this embodiment, is more preferable.

[0177] Furthermore, in the above embodiment, the control device 21 is shown as performing command value fixed switching control six times (every 60 degrees) in one cycle of the periodic phase voltage command value, but it is not limited to this. That is, the control device 21 may perform command value fixed switching control at least once in one cycle of the periodic phase voltage command value, or it may perform it at least two or more times.

[0178] Furthermore, although the above embodiment illustrates the case where the triangular carrier wave is an isosceles triangle, the triangular carrier wave may also be an unequal triangle with no right angles (a signal where the first half and second half of the carrier period have different durations). In that case, the configuration may also be such that the command value for the first half and the command value for the second half are divided and switchable by a dividing line DL passing through the vertex of the unequal triangle carrier period.

[0179] It should be noted that the present invention is not limited to the embodiments described above, and various modifications can be made without departing from the spirit of the invention. [Explanation of Symbols]

[0180] 1. Power converter 1-cu Inverted U-phase two-phase modulation command value 2 Housing 3 Partition wall 4 Compression mechanism housing 6. Inverter housing 7 Compression mechanism 8 motors 10. Upper arm power line (positive terminal busbar) 15. Lower arm power line (negative side busbar) 16 Electric compressor 19U U-phase half-bridge circuit (inverter) 19V V-phase half-bridge circuit 19W W-phase half-bridge circuit 21 Control device 26A, 26B Current Sensor 28 Inverter Circuit 29 DC power supply 33 Phase voltage command calculation unit 34 Line-to-line modulation calculation unit 36 PWM signal generation section 37 Gate Driver 41-43 Armature coil 51 Command value inversion unit 52 Command value division unit 53 Command value shift section CA triangular carrier wave Vu* , Vv * VW * Phase voltage command value cmod modulation value cu1 U-phase modulation command value cv1 V-phase modulation command value cw1 W-phase modulation command value cu U-phase two-phase modulation command value cv V-phase two-phase modulation command value cw W-phase two-phase modulation command value cuf U-phase forward partial command value cub U-phase backward portion command value cwf W-phase forward partial command value cwb W-phase backward partial command value

Claims

1. An inverter circuit that supplies phase voltage to a three-phase load, A power conversion device comprising a control device for the inverter circuit, The control device is A phase voltage command calculation unit that calculates three-phase modulation command values ​​for generating phase voltage command values, A line-to-line modulation calculation unit calculates a two-phase modulation command value based on the three-phase modulation command value, It includes a PWM signal generation unit that generates a PWM signal for PWM control of the inverter circuit based on a triangular carrier wave and the two-phase modulation command value, The control device, with respect to the two phase voltages output based on the PWM signal, maintains the pulse width based on the two-phase modulation command value for each phase, and synchronizes the rising timing of the phase voltage of one phase with the falling timing of the phase voltage of the other phase in at least one of the first half and second half periods within the carrier period of the triangular carrier wave. A command value inversion unit that inverts one of the two-phase modulation command values, A command value division unit that divides the two-phase modulation command value into a forward portion command value and a backward portion command value within the carrier period, Includes a command value shifting unit that matches the forward portion command values ​​of the two phases in the first half of the cycle, or matches the backward portion command values ​​of the two phases in the second half of the cycle. A power conversion device characterized by the following features.

2. The control device is In the first half of the cycle, the forward portion command values ​​of the two phases are made to match, or in the second half of the cycle, the backward portion command values ​​of the two phases are made to match. The first half or second half of the cycle in which the forward portion command value of the two phases or the backward portion command value of the two phases is inconsistent (hereinafter referred to as the "command value mismatch half-cycle"). In the case of the above, control is performed to set the forward portion command value or the backward portion command value to a predetermined value such that the duty cycle of the power switching element of the inverter circuit in one of the two phases becomes an arbitrary width set in advance according to the modulation rate or the amplitude of the phase voltage command value (hereinafter referred to as "command value fixing"). The power conversion device according to feature 1.

3. The control device fixes the command value for the phase among the two phases whose duty cycle is less than 50%. The power conversion device according to feature 2.

4. The control device fixes the command value for the phase with the shorter pulse width among the two phases. The power conversion device according to feature 2.

5. The control device is In the carrier cycle, switching control is performed when the duty cycle of one of the two phases whose duty cycle is less than 50% exceeds 50%. The switching control is a control that switches to fixing the command value in a half-cycle that is different from the half-cycle in which the command value mismatch occurred in the carrier cycle immediately before the switching. The power conversion device according to feature 3.

6. The control device is Switching control is performed when the phase with the shorter pulse width among the two phases changes during the carrier cycle. The switching control is a control that switches to fixing the command value in a half-cycle that is different from the half-cycle in which the command value mismatch occurred in the carrier cycle immediately before the switching. The power conversion device according to feature 4.

7. The phase range for performing the switching control is determined according to the electrical angular phase of the three-phase load. The power conversion device according to claim 5 or 6.

8. The switching control is performed in a plurality of phase ranges such that there is a predetermined regularity. The power conversion device according to feature 7.

9. The control device performs the switching control at least twice in one period of the periodic phase voltage command value. The power conversion device according to feature 8.

10. The aforementioned three-phase load is a motor. A power conversion device according to any one of claims 1 to 6.

11. The motor is used in an electric compressor for vehicle air conditioning. The power conversion device according to feature 10.

Images