Power conversion device and motor unit

Abstract

One embodiment of this power conversion device is provided with: a bridge circuit that performs interconversion of direct-current power and N-phase alternating-current power (where N is an integer 3 or greater); and a control unit that individually controls the duty ratio of pulses of each phase output from the bridge circuit by controlling the bridge circuit through pulse width modulation. Within one cycle of the pulse width modulation, the control unit, with the requirement that the duty ratio of the pulses of each phase is at least 0% and no more than 100%, shifts the timing of the leading edges of the pulses of each phase by a first duration and shifts the timing of the trailing edges of the pulses of each phase by a second duration. The first duration and the second duration change randomly every one or more cycles of pulse width modulation.

Description

Power conversion device and motor unit

[0001] This application claims priority to Japanese Patent Application No. 2024-108327, filed on July 4, 2024, the contents of which are incorporated herein by reference.

[0002] Pulse width modulation (PWM) has traditionally been used to generate drive signals for motors. In systems using PWM, noise caused by the switching frequency of switching elements poses an obstacle to EMC (Electro Magnetic Compatibility) measures. Furthermore, when the switching frequency is within the audible range, the noise caused by the switching frequency is perceived as noise.

[0003] In recent years, it has become common for microprocessors such as MCUs (Microcontroller Units) to have a function for performing pulse width modulation. For example, the microprocessor compares a carrier wave with a threshold value, referred to herein as a duty command value, to generate a pulse having a duty ratio corresponding to the duty command value. The carrier wave is generated by a counter built into the microprocessor. In other words, the instantaneous value of the carrier wave is the output value (count value) of the counter. A commonly known method for reducing noise is to vary the frequency of the clock signal that operates the counter or the frequency of the carrier wave.

[0004] Patent Document 1 discloses a noise reduction technique in which a PWM control signal is shifted so that the rising and falling edges of any two of the three phases of the PWM control signal are close to each other, so that the three phases of the PWM control signal are not simultaneously turned on.

[0005] Non-Patent Document 1 discloses a technique for reducing noise without increasing the switching frequency even when the motor speed is almost zero, in which a zero-phase sequence voltage generated by an M-sequence signal is used to diffuse the frequency components of ripples contained in the current.

[0006] Japanese Patent Application Publication No. 2021-64989

[0007] Ayano, et al., "Carrier Noise Diffusion Method Using Zero-Sequence Voltage During Zero-Speed ​​Operation," IEEJ Transactions on Electrical Engines, Vol. 137, No. 4, 2017, pp. 302-309

[0008] When using a method that varies the frequency of a clock signal, the microprocessor must be equipped with a special function for varying the frequency of the clock signal.When using a method that varies the frequency of a carrier wave, complex calculations must be performed to vary the frequency of the carrier wave, which increases the calculation load on the microprocessor.

[0009] When the technology of Patent Document 1 is adopted, complex calculations must be performed to determine the amount and timing of shifting the PWM control signal, which increases the calculation load on the microprocessor. When the technology of Non-Patent Document 1 is adopted, there is a risk of noise occurring due to the periodicity of the M-sequence signal.

[0010] One aspect of the power conversion device of the present invention includes a bridge circuit that performs mutual conversion between DC power and N-phase AC power (N is an integer of 3 or greater), and a control unit that controls the bridge circuit by pulse width modulation to individually control the duty ratio of a pulse of each phase output from the bridge circuit, wherein the control unit shifts the rising timing of the pulse of each phase by a first time and shifts the falling timing of the pulse of each phase by a second time, under the condition that the duty ratio of the pulse of each phase is greater than or equal to 0% and less than 100% within one cycle of the pulse width modulation, and the first time and the second time change randomly for one or more cycles of the pulse width modulation.

[0011] One aspect of the motor unit of the present invention includes an N-phase motor and the power conversion device of the above aspect that supplies N-phase AC power to the N-phase motor.

[0012] According to the above aspects of the present invention, a power conversion device and a motor unit are provided that can reduce noise caused by the switching frequency of pulse width modulation while using an inexpensive general-purpose microprocessor as a control unit.

[0013] FIG. 1 is a diagram schematically illustrating the overall configuration of a motor unit according to this embodiment. FIG. 2 is a timing chart illustrating a carrier wave and a U-phase pulse, a V-phase pulse, and a differential pulse generated by general pulse width modulation. FIG. 3 is a flowchart illustrating an example of the operation of a control unit. FIG. 4 is a flowchart illustrating a first offset calculation process performed by the control unit. FIG. 5 is a first timing chart illustrating a carrier wave and a U-phase pulse, a V-phase pulse, and a differential pulse generated by the operation of the control unit according to this embodiment. FIG. 6 is a graph illustrating an FFT analysis result of a phase current generated by general pulse width modulation and an FFT analysis result of a phase current generated by the operation of the control unit according to this embodiment. FIG. 7 is a second timing chart illustrating a carrier wave and a U-phase pulse, a V-phase pulse, and a differential pulse generated by the operation of the control unit according to this embodiment. FIG. 8 is a flowchart illustrating a second offset calculation process performed by the control unit. FIG. 9 is a timing chart illustrating a carrier wave and a U-phase pulse, a V-phase pulse, and a W-phase pulse generated when an offset amount of 0 is determined. Fig. 10 is a timing chart showing a carrier wave and U-phase pulses, V-phase pulses, and W-phase pulses generated when "1-dmax" is determined as the offset amount. Fig. 11 is a timing chart showing a carrier wave and U-phase pulses, V-phase pulses, and W-phase pulses generated when "-dmin" is determined as the offset amount.

[0014] An embodiment of the present invention will be described in detail below with reference to the drawings. FIG. 1 is a diagram schematically illustrating the overall configuration of a motor unit 1 according to this embodiment. As shown in FIG. 1, the motor unit 1 includes a power conversion device 10 and a three-phase motor 20. The power conversion device 10 supplies three-phase AC power to the three-phase motor 20. For example, the three-phase motor 20 is an inner rotor type three-phase brushless DC motor. The three-phase motor 20 is, for example, a drive motor (traction motor) mounted on an electric vehicle. The three-phase motor 20 is an example of an N-phase motor (N is an integer equal to or greater than 3).

[0015] The three-phase motor 20 has a U-phase terminal 21u, a V-phase terminal 21v, a W-phase terminal 21w, a U-phase coil 22u, a V-phase coil 22v, and a W-phase coil 22w. Although not shown in FIG. 1 , the three-phase motor 20 also has a motor case, a rotor, and a stator housed in the motor case. The rotor is a rotating body rotatably supported by bearing components such as rotor bearings inside the motor case. The rotor has an output shaft that axially penetrates the radially inner side of the rotor and is coaxially joined to the rotor. The stator is set inside the motor case, surrounding the outer circumferential surface of the rotor, and generates the electromagnetic force required to rotate the rotor.

[0016] The U-phase terminal 21u, the V-phase terminal 21v, and the W-phase terminal 21w are each metal terminals exposed from the surface of the motor case. The U-phase terminal 21u is electrically connected to the U-phase connecting terminal 13u of the power conversion device 10. The V-phase terminal 21v is electrically connected to the V-phase connecting terminal 13v of the power conversion device 10. The W-phase terminal 21w is electrically connected to the W-phase connecting terminal 13w of the power conversion device 10. The U-phase coil 22u, the V-phase coil 22v, and the W-phase coil 22w are each excitation coils provided in the stator. As an example, the U-phase coil 22u, the V-phase coil 22v, and the W-phase coil 22w are star-connected inside the three-phase motor 20.

[0017] The U-phase coil 22u is electrically connected between the U-phase terminal 21u and the neutral point N. The V-phase coil 22v is electrically connected between the V-phase terminal 21v and the neutral point N. The W-phase coil 22w is electrically connected between the W-phase terminal 21w and the neutral point N. The power conversion device 10 controls the energization states of the U-phase coil 22u, the V-phase coil 22v, and the W-phase coil 22w, thereby generating an electromagnetic force required to rotate the rotor. As the rotor rotates, the output shaft also rotates in synchronization with the rotor.

[0018] The power conversion device 10 includes a bridge circuit 11 and a control unit 12. The bridge circuit 11 is electrically connected to a three-phase motor 20 and a DC power supply 30, and performs mutual conversion between DC power and three-phase AC power. For example, when the bridge circuit 11 functions as an inverter, the bridge circuit 11 converts DC power supplied from the DC power supply 30 into three-phase AC power and outputs it to the three-phase motor 20. As an example, the DC power supply 30 is one of multiple batteries mounted on an electric vehicle.

[0019] The bridge circuit 11 includes 2N switches. As described above, in this embodiment, the value of N is 3, so the bridge circuit 11 includes six switches. The bridge circuit 11 includes a U-phase high-side switch Q ＵＨ and V-phase high-side switch Q ＶＨ and W-phase high-side switch Q ＷＨ and U-phase low-side switch Q ＵＬ and V-phase low-side switch Q ＶＬ and W-phase low-side switch Q ＷＬ In this embodiment, each switch is, for example, an IGBT (Insulated Gate Bipolar Transistor).

[0020] U-phase high-side switch Q ＵＨ , V-phase high-side switch Q ＶＨ , and W-phase high-side switch Q ＷＨ The collector terminals of the U-phase low-side switch Q are electrically connected to the positive terminal of the DC power supply 30. ＵＬ , V-phase low-side switch Q ＶＬ , and W-phase low-side switch QＷＬ The emitter terminals of the respective elements are electrically connected to the negative terminal of the DC power supply 30 .

[0021] U-phase high-side switch Q ＵＨ The emitter terminal of the U-phase connection terminal 13u is connected to the U-phase low-side switch Q ＵＬ That is, the U-phase high-side switch Q ＵＨ The emitter terminal of the inverter 13 is electrically connected to a U-phase terminal 21u of the three-phase motor 20 via a U-phase connecting terminal 13u.

[0022] V-phase high-side switch Q ＶＨ The emitter terminal of the V-phase connection terminal 13v and the V-phase low-side switch Q ＶＬ That is, the V-phase high-side switch Q ＶＨ The emitter terminal of is electrically connected to a V-phase terminal 21v of a three-phase motor 20 via a V-phase connecting terminal 13v.

[0023] W-phase high-side switch Q ＷＨ The emitter terminal of the W-phase connecting terminal 13w is connected to the W-phase low-side switch Q ＷＬ That is, the W-phase high-side switch Q ＷＨ The emitter terminal of the inverter 13 is electrically connected to a W-phase terminal 21w of the three-phase motor 20 via a W-phase connecting terminal 13w.

[0024] U-phase high-side switch Q ＵＨ , V-phase high-side switch Q ＶＨ , and W-phase high-side switch Q ＷＨ The gate terminals of the U-phase low-side switch Q are electrically connected to the control unit 12. ＵＬ , V-phase low-side switch Q ＶＬ , and W-phase low-side switch Q ＷＬ The gate terminals of the transistors 11 and 12 are also electrically connected to the control unit 12 .

[0025] As described above, the bridge circuit 11 in this embodiment is a three-phase full-bridge circuit having three high-side switches and three low-side switches. The bridge circuit 11 configured in this manner converts DC power and three-phase AC power mutually by controlling the switching of each switch by the control unit 12. The U-phase connection terminal 13u, the V-phase connection terminal 13v, and the W-phase connection terminal 13w are connection terminals of the bridge circuit 11.

[0026] The control unit 12 controls the bridge circuit 11 by pulse width modulation, thereby individually controlling the duty ratio of the pulse of each phase output from the bridge circuit 11. For example, the control unit 12 is a microprocessor such as an MCU. The control unit 12 generates a gate signal required to control the bridge circuit 11 by pulse width modulation by comparing the duty command value of each phase with a carrier wave. As an example, the carrier wave is a triangular wave. The carrier wave is generated by a counter built into the control unit 12. In other words, the instantaneous value of the carrier wave is the output value (count value) of the counter.

[0027] In the following description, the duty command value for the U phase may be referred to as the U-phase duty command value, the duty command value for the V phase may be referred to as the V-phase duty command value, and the duty command value for the W phase may be referred to as the W-phase duty command value.

[0028] The control unit 12 controls the U-phase high-side switch Q ＵＨ A U-phase high-side gate signal G1 for controlling the U-phase high-side switch Q ＵＨ The control unit 12 sets the level of the U-phase high-side gate signal G1 to a high level when the instantaneous value of the carrier wave is equal to or less than the U-phase duty command value. The control unit 12 sets the level of the U-phase high-side gate signal G1 to a low level when the instantaneous value of the carrier wave is greater than the U-phase duty command value.

[0029] The control unit 12 controls the U-phase low-side switch Q ＵＬ The U-phase low-side gate signal G2 for controlling the U-phase low-side switch Q ＵＬThe U-phase low-side gate signal G2 is a complementary signal of the U-phase high-side gate signal G1.

[0030] The control unit 12 controls the V-phase high-side switch Q ＶＨ A V-phase high-side gate signal G3 for controlling the V-phase high-side switch Q ＶＨ The control unit 12 sets the level of the V-phase high-side gate signal G3 to a high level when the instantaneous value of the carrier wave is equal to or less than the V-phase duty command value. The control unit 12 sets the level of the V-phase high-side gate signal G3 to a low level when the instantaneous value of the carrier wave is greater than the V-phase duty command value.

[0031] The control unit 12 controls the V-phase low-side switch Q ＶＬ A V-phase low-side gate signal G4 for controlling the V-phase low-side switch Q ＶＬ The V-phase low-side gate signal G4 is a complementary signal of the V-phase high-side gate signal G3.

[0032] The control unit 12 controls the W-phase high-side switch Q ＷＨ A W-phase high-side gate signal G5 for controlling the W-phase high-side switch Q ＷＨ The control unit 12 sets the level of the W-phase high-side gate signal G5 to a high level when the instantaneous value of the carrier wave is equal to or less than the W-phase duty command value. The control unit 12 sets the level of the W-phase high-side gate signal G5 to a low level when the instantaneous value of the carrier wave is greater than the W-phase duty command value.

[0033] The control unit 12 controls the W-phase low-side switch Q ＷＬ The W-phase low-side gate signal G6 for controlling the W-phase low-side switch Q ＷＬ The W-phase low-side gate signal G6 is a complementary signal of the W-phase high-side gate signal G5. A dead time is inserted in each gate signal to prevent the high-side switch and low-side switch of the same phase from being switched on at the same time.

[0034] This completes the description of the configuration of the motor unit 1. Before describing the operation of the control unit 12 included in the power conversion device 10, commonly known pulse width modulation will be described below with reference to FIG. 2 in order to facilitate understanding of the operation of the control unit 12. For ease of explanation, the following description will be made of common pulse width modulation using the components shown in FIG. 1.

[0035] 2 is a timing chart showing a carrier wave CW and a U-phase pulse Pu, a V-phase pulse Pv, and a differential pulse Puv generated by general pulse width modulation. The carrier wave CW is a triangular wave having a period Tc. One period Tc of the carrier wave CW corresponds to one period of pulse width modulation. Therefore, in the following description, "one period Tc of the carrier wave CW" may be rephrased as "one period Tc of pulse width modulation."

[0036] The U-phase pulse Pu is a pulse output from the U-phase connection terminal 13u of the bridge circuit 11. The V-phase pulse Pv is a pulse output from the V-phase connection terminal 13v of the bridge circuit 11. The differential pulse Puv is a pulse indicating the difference between the U-phase pulse Pu and the V-phase pulse Pv. In other words, the differential pulse Puv is the line voltage between the U phase and the V phase.

[0037] In general pulse width modulation, the duty command value of each phase is compared with the carrier wave CW for each pulse width modulation period Tc. In the example shown in Fig. 2, the U-phase duty command value du is compared with the carrier wave CW for each pulse width modulation period Tc. Similarly, the V-phase duty command value dv is compared with the carrier wave CW for each pulse width modulation period Tc. When a duty ratio of 100% is 1, the duty command value of each phase is a value greater than or equal to 0 and less than or equal to 1. Fig. 2 shows an example in which the U-phase duty command value du is greater than the V-phase duty command value dv.

[0038] 2 shows an example in which the U-phase duty command value du and the V-phase duty command value dv are constant over a first period T1, which corresponds to one cycle Tc from time t1 to time t2, and a second period T2, which corresponds to one cycle Tc from time t2 to time t3. However, the duty command values ​​for each phase dynamically change depending on the state (rotation speed, current, etc.) of the three-phase motor 20. Therefore, the duty command values ​​for each phase in the second period T2 may differ from the duty command values ​​for each phase in the first period T1. However, the duty command values ​​for each phase do not change within one cycle Tc of the pulse width modulation.

[0039] When the instantaneous value of the carrier wave CW is equal to or less than the U-phase duty command value du, the level of the U-phase pulse Pu is high. On the other hand, when the instantaneous value of the carrier wave CW is greater than the U-phase duty command value du, the level of the U-phase pulse Pu is low. As a result, a U-phase pulse Pu having a duty ratio corresponding to the U-phase duty command value du is generated.

[0040] When the instantaneous value of the carrier wave CW is equal to or less than the V-phase duty command value dv, the level of the V-phase pulse Pv is high. On the other hand, when the instantaneous value of the carrier wave CW is greater than the V-phase duty command value dv, the level of the V-phase pulse Pv is low. As a result, a V-phase pulse Pv having a duty ratio corresponding to the V-phase duty command value dv is generated.

[0041] When the U-phase pulse Pu is at a high level and the V-phase pulse Pv is at a low level, the differential pulse Puv is at a high level. On the other hand, when the U-phase pulse Pu and the V-phase pulse Pv are at the same level, the differential pulse Puv is at a low level. The high level of each pulse corresponds to the positive electrode potential of the DC power supply 30. The low level of each pulse corresponds to the negative electrode potential of the DC power supply 30, i.e., the ground potential.

[0042] The above is a description of commonly known pulse width modulation. Note that while the above description focuses on the relationship between the U phase and the V phase, the same applies to the relationship between the V phase and the W phase, and the relationship between the W phase and the U phase.

[0043] Many of the conventional techniques that address the technical problem of reducing noise or sounds caused by the switching frequency of pulse width modulation attempt to solve the technical problem by focusing on individual pulses of each phase output from the bridge circuit 11, such as the U-phase pulse Pu and the V-phase pulse Pv. However, the inventors of the present application have focused on the fact that line voltages, such as the differential pulse Puv, actually affect the drive of the three-phase motor 20, and have proposed a technique for reducing noise caused by the switching frequency by spreading the frequency components of the line voltage without changing the pulse width of the line voltage. The following describes the operation of the control unit 12 that realizes the above-mentioned frequency spreading of the line voltage.

[0044] The control unit 12 shifts the rising timing of the pulse of each phase by a first time Δt1 and the falling timing of the pulse of each phase by a second time Δt2, under the condition that the duty ratio of the pulse of each phase output from the bridge circuit 11 is greater than or equal to 0% and less than or equal to 100% within one cycle Tc of the pulse width modulation. The first time Δt1 and the second time Δt2 change randomly for one or more cycles of the pulse width modulation. In other words, the first time Δt1 and the second time Δt2 change randomly for each period represented by N×Tc, where N is an integer greater than or equal to 1.

[0045] Specifically, the control unit 12 determines the rising timing and falling timing of the pulse of each phase based on the duty command value of each phase during one cycle Tc of the pulse width modulation. Here, the control unit 12 shifts the rising timing of the pulse of each phase by a first time Δt1 and shifts the falling timing of the pulse of each phase by a second time Δt2 by adding an offset amount, which is randomly determined for one or more cycles of the pulse width modulation, to the duty command value of each phase.

[0046] As already explained, when a duty ratio of 100% is 1, the duty command value of each phase is a value greater than or equal to 0 and less than or equal to 1. When the maximum duty command value among the duty command values ​​of each phase is dmax, the minimum duty command value among the duty command values ​​of each phase is dmin, and the offset amount is Δd, the following equation (1) holds: -dmin≦Δd≦1−dmax (1)

[0047] 3 is a flowchart showing an example of the operation of the control unit 12. As shown in FIG. 3, the control unit 12 acquires three-phase duty command values ​​(step S1). The three-phase duty command values ​​include a U-phase duty command value, a V-phase duty command value, and a W-phase duty command value. The control unit 12 acquires the maximum duty command value of the three phases as dmax and the minimum duty command value as dmin (step S2). The control unit 12 calculates an offset amount Δd based on dmax and dmin (step S3). The control unit 12 adds the offset amount Δd to the three-phase duty command values ​​(step S4).

[0048] 4 is a flowchart showing the first offset calculation process executed by the control unit 12 in step S3 of FIG. 3. As shown in FIG. 4, the control unit 12 generates a random number RND (step S11). For example, the control unit 12 generates the random number RND by using the "BSD formula." The "BSD formula" is a method for obtaining uniformly distributed random numbers using the following equation (2). Note that the method for generating the random number RND is not limited to the "BSD formula."

[0049]

[0050] The control unit 12 determines whether the random number RND is greater than 1-dmax (step S12). If the random number RND is greater than 1-dmax (step S12: YES), the control unit 12 determines 1-dmax as the offset amount Δd (step S13). On the other hand, if the random number RND is equal to or less than 1-dmax (step S12: NO), the control unit 12 determines whether the random number RND is smaller than -dmin (step S14).

[0051] If the random number RND is smaller than -dmin (step S14: YES), the control unit 12 determines -dmin as the offset amount Δd (step S15). On the other hand, if the random number RND is equal to or larger than -dmin (step S14: NO), the control unit 12 determines the random number RND as the offset amount Δd (step S16).

[0052] The operation of the control unit 12 will be described in more detail below with reference to Fig. 5. For ease of explanation, the U-phase pulse Pu shown in Fig. 2, i.e., the U-phase pulse Pu generated based on the U-phase duty command value du, may be referred to as the "original U-phase pulse Pu." Similarly, in the following explanation, the V-phase pulse Pv shown in Fig. 2, i.e., the V-phase pulse Pv generated based on the V-phase duty command value dv, may be referred to as the "original V-phase pulse Pv."

[0053] 2, the pulse waveform located on the left side of the time axis may be referred to as the “original first differential pulse Puv1,” and the pulse waveform located on the right side of the time axis may be referred to as the “original second differential pulse Puv2.” Furthermore, in the following description, the pulse waveform located on the left side of the time axis of the differential pulse Puv included in the second period T2 shown in FIG. 2 may be referred to as the “original third differential pulse Puv3,” and the pulse waveform located on the right side of the time axis may be referred to as the “original fourth differential pulse Puv4.”

[0054] 5 is a first timing chart showing a carrier wave CW and a U-phase pulse Pu, a V-phase pulse Pv, and a differential pulse Puv generated by the operation of the control unit 12. The control unit 12 randomly determines the offset amount Δd that satisfies the above equation (1) based on the duty command values ​​of the phases in a first period T1 shown in FIG. 5. For example, if in the first period T1 the U-phase duty command value du is 0.7, the V-phase duty command value dv is 0.4, and the W-phase duty command value (not shown) is 0.2, the control unit 12 randomly determines the offset amount Δd to be a value included in the range from −0.2 to 0.3.

[0055] Here, the control unit 12 may determine a value randomly extracted from the range of -0.2 to 0.3 as the offset amount Δd. Alternatively, if a value randomly extracted from the range of -1 to 1 is -0.2 or more and 0.3 or less, the control unit 12 may determine the extracted value as the offset amount Δd, if the extracted value is greater than 0.3, the control unit 12 may determine 0.3 as the offset amount Δd, and if the extracted value is smaller than -0.2, the control unit 12 may determine -0.2 as the offset amount Δd.

[0056] After determining the offset amount Δd for the first period T1 as described above, the control unit 12 adds the offset amount Δd to the duty command value of each phase during the first period T1. Here, it is assumed that the offset amount Δd for the first period T1 is a value having a positive sign. The first new U-phase duty command value du1 shown in FIG. 5 is a new U-phase duty command value obtained by adding the offset amount Δd having a positive sign to the original U-phase duty command value du. The first new V-phase duty command value dv1 shown in FIG. 5 is a new V-phase duty command value obtained by adding the offset amount Δd having a positive sign to the original V-phase duty command value dv.

[0057] During the first period T1, the control unit 12 compares the first new U-phase duty command value du1 with the carrier wave CW. When the instantaneous value of the carrier wave CW is equal to or less than the first new U-phase duty command value du1, the control unit 12 sets the level of the U-phase pulse Pu to a high level. On the other hand, when the instantaneous value of the carrier wave CW is greater than the first new U-phase duty command value du1, the control unit 12 sets the level of the U-phase pulse Pu to a low level. As a result, during the first period T1, a U-phase pulse Pu having a duty ratio corresponding to the first new U-phase duty command value du1 is generated.

[0058] During the first period T1, the control unit 12 compares the first new V-phase duty command value dv1 with the carrier wave CW. When the instantaneous value of the carrier wave CW is equal to or less than the first new V-phase duty command value dv1, the control unit 12 sets the level of the V-phase pulse Pv to a high level. On the other hand, when the instantaneous value of the carrier wave CW is greater than the first new V-phase duty command value dv1, the control unit 12 sets the level of the V-phase pulse Pv to a low level. As a result, during the first period T1, a V-phase pulse Pv having a duty ratio corresponding to the first new V-phase duty command value dv1 is generated.

[0059] 5, the rising timing of the U-phase pulse Pu generated based on the first new U-phase duty command value du1 in the first period T1 is shifted leftward on the time axis by a first time Δt1 from the rising timing of the original U-phase pulse Pu. Meanwhile, the falling timing of the U-phase pulse Pu generated based on the first new U-phase duty command value du1 in the first period T1 is shifted rightward on the time axis by a second time Δt2 from the falling timing of the original U-phase pulse Pu.

[0060] Similarly, the rising timing of the V-phase pulse Pv generated based on the first new V-phase duty command value dv1 in the first period T1 is shifted by a first time Δt1 to the left on the time axis from the rising timing of the original V-phase pulse Pv. On the other hand, the falling timing of the V-phase pulse Pv generated based on the first new V-phase duty command value dv1 in the first period T1 is shifted by a second time Δt2 to the right on the time axis from the falling timing of the original V-phase pulse Pv.

[0061] 5, the rise and fall timings of the first differential pulse Puv1 located on the left side of the time axis among the differential pulses Puv included in the first period T1 are shifted leftward on the time axis by a first time Δt1 from the rise and fall timings of the original first differential pulse Puv1. The pulse width of the first differential pulse Puv1 shown in FIG. 5 is the same as the pulse width of the original first differential pulse Puv1.

[0062] 5, the rise and fall timings of the second differential pulse Puv2 located on the right side of the time axis among the differential pulses Puv included in the first period T1 are shifted by a second time Δt2 toward the right on the time axis from the rise and fall timings of the original second differential pulse Puv2. The pulse width of the second differential pulse Puv2 shown in FIG. 5 is the same as the pulse width of the original second differential pulse Puv2.

[0063] The lengths of the first time Δt1 and the second time Δt2 in the first period T1 are determined by the offset amount Δd determined in the first period T1 and the slope of the carrier wave CW. In the example shown in Fig. 5, one offset amount Δd is used in the first period T1, so the length of the first time Δt1 is equal to the length of the second time Δt2 in the first period T1.

[0064] Next, the control unit 12 randomly determines the offset amount Δd that satisfies the above formula (1) based on the duty command value of each phase in the second period T2. The method of randomly determining the offset amount Δd in the second period T2 is the same as that in the first period T1.

[0065] After determining the offset amount Δd for the second period T2, the control unit 12 adds the offset amount Δd to the duty command value of each phase during the second period T2. Here, it is assumed that the offset amount Δd for the second period T2 is a value having a negative sign. The second new U-phase duty command value du2 shown in FIG. 5 is a new U-phase duty command value obtained by adding the offset amount Δd having a negative sign to the original U-phase duty command value du. The second new V-phase duty command value dv2 shown in FIG. 5 is a new V-phase duty command value obtained by adding the offset amount Δd having a negative sign to the original V-phase duty command value dv.

[0066] During the second period T2, the control unit 12 compares the second new U-phase duty command value du2 with the carrier wave CW. If the instantaneous value of the carrier wave CW is equal to or less than the second new U-phase duty command value du2, the control unit 12 sets the level of the U-phase pulse Pu to a high level. On the other hand, if the instantaneous value of the carrier wave CW is greater than the second new U-phase duty command value du2, the control unit 12 sets the level of the U-phase pulse Pu to a low level. As a result, during the second period T2, a U-phase pulse Pu having a duty ratio corresponding to the second new U-phase duty command value du2 is generated.

[0067] During the second period T2, the control unit 12 compares the second new V-phase duty command value dv2 with the carrier wave CW. When the instantaneous value of the carrier wave CW is equal to or less than the second new V-phase duty command value dv2, the control unit 12 sets the level of the V-phase pulse Pv to a high level. On the other hand, when the instantaneous value of the carrier wave CW is greater than the second new V-phase duty command value dv2, the control unit 12 sets the level of the V-phase pulse Pv to a low level. As a result, during the second period T2, a V-phase pulse Pv having a duty ratio corresponding to the second new V-phase duty command value dv2 is generated.

[0068] 5, the rising timing of the U-phase pulse Pu generated based on the second new U-phase duty command value du2 in the second period T2 is shifted by a first time Δt1 toward the right on the time axis from the rising timing of the original U-phase pulse Pu. On the other hand, the falling timing of the U-phase pulse Pu generated based on the second new U-phase duty command value du2 in the second period T2 is shifted by a second time Δt2 toward the left on the time axis from the falling timing of the original U-phase pulse Pu.

[0069] Similarly, the rising timing of the V-phase pulse Pv generated based on the second new V-phase duty command value dv2 in the second period T2 is shifted by a first time Δt1 to the right on the time axis from the rising timing of the original V-phase pulse Pv. On the other hand, the falling timing of the V-phase pulse Pv generated based on the second new V-phase duty command value dv2 in the second period T2 is shifted by a second time Δt2 to the left on the time axis from the falling timing of the original V-phase pulse Pv.

[0070] 5, among the differential pulses Puv included in the second period T2, the rising and falling timings of the third differential pulse Puv3 located on the left side of the time axis are shifted by the first time Δt1 toward the right side of the time axis from the rising and falling timings of the original third differential pulse Puv3. The pulse width of the third differential pulse Puv3 shown in FIG. 5 is the same as the pulse width of the original third differential pulse Puv3.

[0071] 5, among the differential pulses Puv included in the second period T2, the rising and falling timings of the fourth differential pulse Puv4 located on the right side of the time axis are shifted by a second time Δt2 toward the left on the time axis from the rising and falling timings of the original fourth differential pulse Puv4. The pulse width of the fourth differential pulse Puv4 shown in FIG. 5 is the same as the pulse width of the original fourth differential pulse Puv4.

[0072] The lengths of the first time Δt1 and the second time Δt2 in the second period T2 are determined by the offset amount Δd determined in the second period T2 and the slope of the carrier wave CW. In the example shown in Fig. 5, one offset amount Δd is used in the second period T2, so the length of the first time Δt1 is equal to the length of the second time Δt2 in the second period T2.

[0073] As described above, the control unit 12 of this embodiment shifts the rising timing of the pulse of each phase by the first time Δt1 and the falling timing of the pulse of each phase by the second time Δt2, under the condition that the duty ratio of the pulse of each phase output from the bridge circuit 11 is greater than or equal to 0% and less than or equal to 100% within one cycle Tc of the pulse width modulation. The first time Δt1 and the second time Δt2 change randomly for one or more cycles of the pulse width modulation. As a result, the frequency components of the line voltage, such as the differential pulse Puv, can be spread without changing the pulse width of the line voltage.

[0074] 6 is a graph showing the results of FFT analysis of the phase current generated by typical pulse width modulation and the results of FFT analysis of the phase current generated by the operation of the control unit 12 of this embodiment. Both analysis results were obtained by performing simulations assuming a duty ratio of 3%. In FIG. 6 , the dotted line L1 indicates the peak level of the noise component contained in the phase current generated by typical pulse width modulation, and the dashed-dotted line L2 indicates the peak level of the noise component contained in the phase current generated by the operation of the control unit 12. As shown in FIG. 6 , this embodiment can suppress the peak level of noise compared to typical pulse width modulation.

[0075] Furthermore, according to this embodiment, unlike the prior art, it is not necessary to provide the microprocessor (controller 12) with a special function for varying the frequency of the clock signal, nor is it necessary for the control unit 12 to perform complex calculations to determine the shift amounts (first time Δt1 and second time Δt2) of the rise and fall timings of the pulses of each phase output from the bridge circuit 11. Therefore, according to this embodiment, it is possible to reduce noise caused by the switching frequency of pulse width modulation while using an inexpensive general-purpose microprocessor as the control unit 12.

[0076] In the above embodiment, an example has been described in which one offset amount Δd is used in one cycle Tc of pulse width modulation, but two offset amounts may be used in one cycle Tc of pulse width modulation. In this case, the control unit 12 shifts the rising timing of the pulse of each phase by a first time Δt1 by adding a first offset amount that is randomly determined for one or more cycles of pulse width modulation to the duty command value for each phase, and shifts the falling timing of the pulse of each phase by a second time Δt2 by adding a second offset amount that is randomly determined for one or more cycles of pulse width modulation to the duty command value for each phase.

[0077] Here, when the maximum duty command value among the duty command values ​​of the phases is dmax, the minimum duty command value among the duty command values ​​of the phases is dmin, the first offset amount is Δd1, and the second offset amount is Δd2, the following equations (3) and (4) hold: -dmin≦Δd1≦1−dmax (3) -dmin≦Δd2≦1−dmax (4)

[0078] Hereinafter, the operation of the control unit 12 when the first offset amount Δd1 and the second offset amount Δd2 are used within one cycle Tc of pulse width modulation will be described with reference to Fig. 7. Fig. 7 is a second timing chart showing the carrier wave CW and the U-phase pulse Pu, V-phase pulse Pv, and differential pulse Puv generated by the operation of the control unit 12. The control unit 12 randomly determines the first offset amount Δd1 that satisfies the above formula (3) and the second offset amount Δd2 that satisfies the above formula (4) based on the duty command values ​​of each phase during the first period T1 shown in Fig. 7.

[0079] For example, if the U-phase duty command value du is 0.7, the V-phase duty command value dv is 0.4, and the W-phase duty command value (not shown) is 0.2 during the first period T1, the control unit 12 randomly determines a value within the range from −0.2 to 0.3 as the first offset amount Δd1. The control unit 12 also randomly determines a value within the range from −0.2 to 0.3 as the second offset amount Δd2. Therefore, the first offset amount Δd1 and the second offset amount Δd2 may be the same or may be different.

[0080] After determining the first offset amount Δd1 and the second offset amount Δd2 for the first period T1 as described above, the control unit 12 adds the first offset amount Δd1 to the duty command value of each phase in the first half of the first period T1. Here, it is assumed that the first offset amount Δd1 is a value having a positive sign. The third new U-phase duty command value du3 shown in FIG. 7 is a new U-phase duty command value obtained by adding the first offset amount Δd1 to the original U-phase duty command value du. The third new V-phase duty command value dv3 shown in FIG. 7 is a new V-phase duty command value obtained by adding the first offset amount Δd1 to the original V-phase duty command value dv.

[0081] In the first half of the first period T1, the control unit 12 compares the third new U-phase duty command value du3 with the carrier wave CW. If the instantaneous value of the carrier wave CW is equal to or less than the third new U-phase duty command value du3, the control unit 12 sets the level of the U-phase pulse Pu to high. On the other hand, if the instantaneous value of the carrier wave CW is greater than the third new U-phase duty command value du3, the control unit 12 sets the level of the U-phase pulse Pu to low.

[0082] In the first half of the first period T1, the control unit 12 compares the third new V-phase duty command value dv3 with the carrier wave CW. If the instantaneous value of the carrier wave CW is equal to or less than the third new V-phase duty command value dv3, the control unit 12 sets the level of the V-phase pulse Pv to a high level. On the other hand, if the instantaneous value of the carrier wave CW is greater than the third new V-phase duty command value dv3, the control unit 12 sets the level of the V-phase pulse Pv to a low level.

[0083] The control unit 12 adds a second offset amount Δd2 to the duty command value of each phase in the second half of the first period T1. Here, it is assumed that the second offset amount Δd2 has a positive sign and is greater than the first offset amount Δd1. The fourth new U-phase duty command value du4 shown in FIG. 7 is a new U-phase duty command value obtained by adding the second offset amount Δd2 to the original U-phase duty command value du. The fourth new V-phase duty command value dv4 shown in FIG. 7 is a new V-phase duty command value obtained by adding the second offset amount Δd2 to the original V-phase duty command value dv.

[0084] In the second half of the first period T1, the control unit 12 compares the fourth new U-phase duty command value du4 with the carrier wave CW. If the instantaneous value of the carrier wave CW is equal to or less than the fourth new U-phase duty command value du4, the control unit 12 sets the level of the U-phase pulse Pu to high. On the other hand, if the instantaneous value of the carrier wave CW is greater than the fourth new U-phase duty command value du4, the control unit 12 sets the level of the U-phase pulse Pu to low.

[0085] In the second half of the first period T1, the control unit 12 compares the fourth new V-phase duty command value dv4 with the carrier wave CW. If the instantaneous value of the carrier wave CW is equal to or less than the fourth new V-phase duty command value dv4, the control unit 12 sets the level of the V-phase pulse Pv to a high level. On the other hand, if the instantaneous value of the carrier wave CW is greater than the fourth new V-phase duty command value dv4, the control unit 12 sets the level of the V-phase pulse Pv to a low level.

[0086] 7, in the first half of the first period T1, the rising timing of the U-phase pulse Pu generated based on the third new U-phase duty command value du3 is shifted leftward on the time axis by a first time Δt1 from the rising timing of the original U-phase pulse Pu. Meanwhile, in the second half of the first period T1, the falling timing of the U-phase pulse Pu generated based on the fourth new U-phase duty command value du4 is shifted rightward on the time axis by a second time Δt2 from the falling timing of the original U-phase pulse Pu.

[0087] Similarly, the rising timing of the V-phase pulse Pv generated based on the third new V-phase duty command value dv3 in the first half of the first period T1 is shifted leftward on the time axis by a first time Δt1 from the rising timing of the original V-phase pulse Pv, while the falling timing of the V-phase pulse Pv generated based on the fourth new V-phase duty command value dv4 in the second half of the first period T1 is shifted rightward on the time axis by a second time Δt2 from the falling timing of the original V-phase pulse Pv.

[0088] 7, of the differential pulses Puv included in the first period T1, the rise and fall timings of the first differential pulse Puv1 located on the left side on the time axis (i.e., in the first half of the first period T1) are shifted leftward on the time axis by a first time Δt1 from the rise and fall timings of the original first differential pulse Puv1. The pulse width of the first differential pulse Puv1 shown in FIG. 7 is the same as the pulse width of the original first differential pulse Puv1.

[0089] 7, of the differential pulse Puv included in the first period T1, the rise and fall timings of the second differential pulse Puv2 located on the right side on the time axis (i.e., in the second half of the first period T1) are shifted by a second time Δt2 toward the right on the time axis from the rise and fall timings of the original second differential pulse Puv2. The pulse width of the second differential pulse Puv2 shown in FIG. 7 is the same as the pulse width of the original second differential pulse Puv2.

[0090] The length of the first time Δt1 in the first half of the first period T1 is determined by the first offset amount Δd1 and the slope of the carrier wave CW. The length of the second time Δt2 in the second half of the first period T1 is determined by the second offset amount Δd2 and the slope of the carrier wave CW. As shown in Figure 7, when the second offset amount Δd2 is greater than the first offset amount Δd1, the length of the second time Δt2 is longer than the first time Δt1. When the first offset amount Δd1 and the second offset amount Δd2 are the same, the length of the second time Δt2 is equal to the length of the first time Δt1.

[0091] As described above, by adding the first offset amount Δd1 to the duty command value of each phase in the first half of one cycle Tc of the pulse width modulation and by adding the second offset amount Δd2 to the duty command value of each phase in the second half of one cycle Tc of the pulse width modulation, it is possible to spread the frequency components of the line voltages without changing the pulse widths of the line voltages such as the differential pulse Puv.

[0092] In the above embodiment, an example has been described in which the offset amount Δd that satisfies the above formula (1) is randomly determined. However, when the maximum duty command value among the duty command values ​​of each phase is dmax and the minimum duty command value among the duty command values ​​of each phase is dmin, the offset amount Δd may be randomly determined from three values: 0, 1-dmax, and -dmin.

[0093] Even when the offset amount Δd is randomly determined from the three values ​​of 0, 1-dmax, and -dmin, the control unit 12 performs the process shown in Fig. 3. In this case, however, the control unit 12 executes the second offset calculation process in step S3 of Fig. 3 according to the flowchart shown in Fig. 8.

[0094] 8 is a flowchart showing the second offset calculation process executed by the control unit 12 in step S3 of Fig. 3 when the offset amount Δd is randomly determined from three values, 0, 1-dmax, and -dmin. As shown in Fig. 8, the control unit 12 randomly selects an integer from 0, 1, and 2 (step S21).

[0095] For example, by using the "BSD formula" ３２ A random number within the range from 0 to -1 is generated. Therefore, for example, the control unit 12 determines whether the random number obtained by the "BSD formula" is within the range from 0 to (2 ３２ If the random number obtained by the "BSD formula" is within the range of (2 -1) / 3, the control unit 12 obtains 0 as the integer. ３２ -1) / 3 to (2 ３２ If the random number obtained by the "BSD formula" is within the range of (2 -1) × 2 / 3, the control unit 12 obtains 1 as an integer. ３２ -1) x 2 / 3 to (2 ３２ If the value is within the range of -1, the integer 2 is obtained.

[0096] The control unit 12 determines whether the acquired integer is 0 (step S22). If the acquired integer is 0 (step S22: YES), the control unit 12 determines -dmin as the offset amount Δd (step S23). On the other hand, if the acquired integer is not 0 (step S22: NO), the control unit 12 determines whether the acquired integer is 1 (step S24).

[0097] If the acquired integer is 1 (step S24: YES), the control unit 12 determines the offset amount Δd to be 0 (step S25). On the other hand, if the acquired integer is not 1 (step S24: NO), that is, if the acquired integer is 2, the control unit 12 determines 1-dmax as the offset amount Δd (step S26).

[0098] 9 is a timing chart showing a carrier wave CW and a U-phase pulse Pu, a V-phase pulse Pv, and a W-phase pulse Pw that are generated when the offset amount Δd is determined to be 0. As shown in FIG. 9, when the offset amount Δd is determined to be 0, the U-phase pulse Pu is generated based on the original U-phase duty command value du, the V-phase pulse Pv is generated based on the original V-phase duty command value dv, and the W-phase pulse Pw is generated based on the original W-phase duty command value dw. In other words, when the offset amount Δd is determined to be 0, typical pulse width modulation (three-phase modulation) is performed.

[0099] 10 is a timing chart showing a carrier wave CW and a U-phase pulse Pu, a V-phase pulse Pv, and a W-phase pulse Pw that are generated when the offset amount Δd is determined to be 1−dmax. As shown in FIG. 10, it is assumed that, among the duty command values ​​of the phases in a first period T1, the U-phase duty command value du is the largest, the V-phase duty command value dv is the second largest, and the W-phase duty command value dw is the smallest.

[0100] The fifth new U-phase duty command value du5 shown in FIG. 10 is a new U-phase duty command value obtained by adding an offset amount Δd, which is "1-dmax," to the original U-phase duty command value du. Since the U-phase duty command value du is the largest, the fifth new U-phase duty command value du5 is 1. The fifth new V-phase duty command value dv5 shown in FIG. 10 is a new V-phase duty command value obtained by adding an offset amount Δd, which is "1-dmax," to the original V-phase duty command value dv. The fifth new W-phase duty command value dw5 shown in FIG. 10 is a new W-phase duty command value obtained by adding an offset amount Δd, which is "1-dmax," to the original W-phase duty command value dw.

[0101] 10 , the rising timing of the U-phase pulse Pu generated based on the fifth new U-phase duty command value du5 is shifted by a first time Δt1 to the left on the time axis from the rising timing of the original U-phase pulse Pu, while the falling timing of the U-phase pulse Pu generated based on the fifth new U-phase duty command value du5 is shifted by a second time Δt2 to the right on the time axis from the falling timing of the original U-phase pulse Pu.

[0102] The rising timing of the V-phase pulse Pv generated based on the fifth new V-phase duty command value dv5 is shifted by a first time Δt1 to the left on the time axis from the rising timing of the original V-phase pulse Pv, while the falling timing of the V-phase pulse Pv generated based on the fifth new V-phase duty command value dv5 is shifted by a second time Δt2 to the right on the time axis from the falling timing of the original V-phase pulse Pv.

[0103] The rising timing of the W-phase pulse Pw generated based on the fifth new W-phase duty command value dw5 is shifted by a first time Δt1 to the left on the time axis from the rising timing of the original W-phase pulse Pw, while the falling timing of the W-phase pulse Pw generated based on the fifth new W-phase duty command value dw5 is shifted by a second time Δt2 to the right on the time axis from the falling timing of the original W-phase pulse Pw.

[0104] As described above, when "1-dmax" is determined as the offset amount Δd, the duty ratio of the pulse of the phase having the largest original duty command value (the U phase in the example shown in FIG. 10) becomes 100%. In other words, when "1-dmax" is determined as the offset amount Δd, so-called max-type two-phase modulation is performed. Max-type two-phase modulation is a two-phase modulation method in which, of the six switches included in the bridge circuit 11, the high-side switch of one phase is set to on and the switches of the remaining phases are controlled by pulse width modulation. In max-type two-phase modulation, the switching period of the phases in which the high-side switches are set to on corresponds to one-third of one electrical angle period (i.e., 120 electrical degrees).

[0105] 11 is a timing chart showing a carrier wave CW and a U-phase pulse Pu, a V-phase pulse Pv, and a W-phase pulse Pw that are generated when "-dmin" is determined as the offset amount Δd. As shown in FIG. 11, it is assumed that, among the duty command values ​​of each phase in a first period T1, the U-phase duty command value du is the largest, the V-phase duty command value dv is the second largest, and the W-phase duty command value dw is the smallest.

[0106] The sixth new U-phase duty command value du6 shown in FIG. 11 is a new U-phase duty command value obtained by adding an offset amount Δd of "-dmin" to the original U-phase duty command value du. The sixth new V-phase duty command value dv6 shown in FIG. 11 is a new V-phase duty command value obtained by adding an offset amount Δd of "-dmin" to the original V-phase duty command value dv. The sixth new W-phase duty command value dw6 shown in FIG. 11 is a new W-phase duty command value obtained by adding an offset amount Δd of "-dmin" to the original W-phase duty command value dw. Because the W-phase duty command value dw is the smallest, the sixth new W-phase duty command value dw6 is 0.

[0107] 11 , the rising timing of the U-phase pulse Pu generated based on the sixth new U-phase duty command value du6 is shifted by a first time Δt1 toward the right on the time axis from the rising timing of the original U-phase pulse Pu, while the falling timing of the U-phase pulse Pu generated based on the sixth new U-phase duty command value du6 is shifted by a second time Δt2 toward the left on the time axis from the falling timing of the original U-phase pulse Pu.

[0108] The rising timing of the V-phase pulse Pv generated based on the sixth new V-phase duty command value dv6 is shifted by a first time Δt1 to the right on the time axis from the rising timing of the original V-phase pulse Pv, while the falling timing of the V-phase pulse Pv generated based on the sixth new V-phase duty command value dv6 is shifted by a second time Δt2 to the left on the time axis from the falling timing of the original V-phase pulse Pv.

[0109] The rising timing of the W-phase pulse Pw generated based on the sixth new W-phase duty command value dw6 is shifted by a first time Δt1 to the right on the time axis from the rising timing of the original W-phase pulse Pw, while the falling timing of the W-phase pulse Pw generated based on the sixth new W-phase duty command value dw6 is shifted by a second time Δt2 to the left on the time axis from the falling timing of the original W-phase pulse Pw.

[0110] As described above, when "-dmin" is determined as the offset amount Δd, the duty ratio of the pulse of the phase with the smallest original duty command value (W-phase in the example shown in FIG. 11 ) becomes 0%. In other words, when "-dmin" is determined as the offset amount Δd, so-called Min-type two-phase modulation is performed. Min-type two-phase modulation is a two-phase modulation method in which, of the six switches included in the bridge circuit 11, the low-side switch of one phase is set to ON and the switches of the remaining phases are controlled by pulse width modulation. In Min-type two-phase modulation, the switching period of the phase in which the low-side switch is set to ON corresponds to one-third of one electrical angle period (i.e., 120 electrical degrees).

[0111] As described above, when the offset amount Δd is randomly determined from three values, 0, 1−dmax, and −dmin, for one or more pulse width modulation cycles, the modulation method is randomly determined from three-phase modulation, maximum two-phase modulation, and minimum two-phase modulation for one or more pulse width modulation cycles. Even in this case, the frequency components of the line voltage can be spread without changing the pulse width of the line voltage. Furthermore, when the modulation method switches from minimum two-phase modulation to minimum two-phase modulation, switching is not required for the phase with the smallest duty command value, and when the modulation method switches from maximum two-phase modulation to maximum two-phase modulation, switching is not required for the phase with the largest duty command value. Therefore, when the offset amount Δd is randomly determined from three values, 0, 1−dmax, and −dmin, the number of switching operations can be reduced, thereby reducing switching loss.

[0112] The above describes a case where the offset amount Δd is randomly determined from three values, 0, 1-dmax, and -dmin, for one or more cycles of pulse width modulation. However, the offset amount Δd may also be randomly determined from two values, 0 and -dmin, or two values, 0 and 1-dmax, depending on the electrical angle of the three-phase motor 20.

[0113] For example, when the electrical angle of the three-phase motor 20 is greater than 0 degrees and falls within a range of 60 degrees or less, the offset amount Δd is randomly determined from two values, 0 and -dmin. When the electrical angle of the three-phase motor 20 is greater than 60 degrees and falls within a range of 120 degrees or less, the offset amount Δd is randomly determined from two values, 0 and 1-dmax.

[0114] When the electrical angle of the three-phase motor 20 is greater than 120 degrees and falls within a range of 180 degrees or less, the offset amount Δd is randomly determined from two values, 0 and -dmin. When the electrical angle of the three-phase motor 20 is greater than 180 degrees and falls within a range of 240 degrees or less, the offset amount Δd is randomly determined from two values, 0 and 1-dmax.

[0115] When the electrical angle of the three-phase motor 20 is greater than 240 degrees and falls within a range of 300 degrees or less, the offset amount Δd is randomly determined from two values, 0 and -dmin. When the electrical angle of the three-phase motor 20 is greater than 300 degrees and falls within a range of 360 degrees or less, the offset amount Δd is randomly determined from two values, 0 and 1-dmax.

[0116] As described above, when the offset amount Δd is randomly determined from two values, 0 and -dmin, or two values, 0 and 1-dmax, depending on the electrical angle of the three-phase motor 20 for one or more pulse width modulation cycles, the modulation method is randomly determined from two, three-phase modulation and min-max two-phase modulation, for one or more pulse width modulation cycles. Even in this case, the frequency components of the line voltage can be spread without changing the pulse width of the line voltage. Note that min-max two-phase modulation is a modulation method that switches between min-type two-phase modulation and max-type two-phase modulation for every 60 electrical degrees.

[0117] The present invention is not limited to the above-described embodiment, and the configurations described in this specification can be combined as appropriate within a range that does not contradict each other. For example, in the above-described embodiment, the power conversion device 10 controlling the three-phase motor 20 is illustrated as an example, but the controlled object of the power conversion device of the present invention is not limited to the three-phase motor 20, and may be any N-phase motor (N is an integer of 3 or more). Furthermore, in the above-described embodiment, IGBTs are illustrated as the arm switches included in the bridge circuit 11, but the arm switches may be high-power switching elements other than IGBTs, such as MOS-FETs.

[0118] The present technology can be configured as follows: (1) A power conversion device including: a bridge circuit that performs mutual conversion between DC power and N-phase AC power (N is an integer of 3 or greater); and a control unit that controls the bridge circuit by pulse width modulation to individually control the duty ratio of a pulse of each phase output from the bridge circuit, wherein the control unit shifts the rising timing of the pulse of each phase by a first time and shifts the falling timing of the pulse of each phase by a second time within one cycle of the pulse width modulation, under the condition that the duty ratio of the pulse of each phase is greater than or equal to 0% and less than 100%, and the first time and the second time are changed randomly for one or more cycles of the pulse width modulation. (2) The power conversion device according to (1), wherein the control unit determines rise timings and fall timings of the pulses of each phase based on a duty command value of each phase during one cycle of the pulse width modulation, and the control unit adds an offset amount randomly determined for one or more cycles of the pulse width modulation to the duty command value of each phase, thereby shifting the rise timings of the pulses of each phase by the first time and shifting the fall timings of the pulses of each phase by the second time. (3) The power conversion device according to (2), wherein the duty command values ​​of each phase are greater than or equal to 0 and less than or equal to 1 when a duty ratio of 100% is 1, and the maximum duty command value of the duty command values ​​of each phase is dmax, the minimum duty command value of the duty command values ​​of each phase is dmin, and the offset amount is Δd, and the following equation (1) holds: -dmin≦Δd≦1−dmax (1) (4) The power conversion device according to (2), wherein, when a duty ratio of 100% is 1, the duty command value of each phase is a value greater than or equal to 0 and less than or equal to 1, and when the maximum duty command value of the duty command values ​​of each phase is dmax and the minimum duty command value of the duty command values ​​of each phase is dmin, the offset amount is randomly determined from three values ​​of 0, 1−dmax, and −dmin.(5) The power conversion device according to (2), wherein, when a duty ratio of 100% is 1, the duty command value of each phase is a value greater than or equal to 0 and less than or equal to 1, and when the maximum duty command value of the duty command values ​​of each phase is dmax and the minimum duty command value of the duty command values ​​of each phase is dmin, the offset amount is randomly determined from two values, 0 and -dmin, or two values, 0 and 1-dmax, depending on the electrical angle of the N-phase motor to which the N-phase AC power is supplied. (6) The power conversion device according to (1), wherein the control unit determines, in one cycle of the pulse width modulation, a rise timing and a fall timing of the pulse of each phase based on a duty command value of each phase, and the control unit shifts the rise timing of the pulse of each phase by the first time by adding a first offset amount that is randomly determined for one or more cycles of the pulse width modulation to the duty command value of each phase, and shifts the fall timing of the pulse of each phase by the second time by adding a second offset amount that is randomly determined for one or more cycles of the pulse width modulation to the duty command value of each phase. (7) The power conversion device according to (6), wherein when a duty ratio of 100% is 1, the duty command value of each phase is a value greater than or equal to 0 and less than or equal to 1, the maximum duty command value of the duty command values ​​of each phase is dmax, the minimum duty command value of the duty command values ​​of each phase is dmin, the first offset amount is Δd1, and the second offset amount is Δd2, the following equations (3) and (4) hold: -dmin≦Δd1≦1−dmax ... (3) -dmin≦Δd2≦1−dmax ... (4) (8) A motor unit comprising: an N-phase motor; and the power conversion device according to any one of (1) to (7), which supplies N-phase AC power to the N-phase motor.

[0119] According to aspects of the present invention, a power conversion device and a motor unit are provided that can reduce noise caused by the switching frequency of pulse width modulation while using an inexpensive general-purpose microprocessor as a control unit. Therefore, the present invention has industrial applicability.

[0120] REFERENCE SIGNS LIST 1 motor unit 10 power conversion device 11 bridge circuit 12 control unit 13u U-phase connection terminal 13v V-phase connection terminal 13w W-phase connection terminal 20 three-phase motor 21u U-phase terminal 21v V-phase terminal 21w W-phase terminal 22u U-phase coil 22v V-phase coil 22w W-phase coil 30 DC power supply Q ＵＨ U-phase high-side switch Q ＶＨ V-phase high-side switch Q ＷＨ W-phase high-side switch Q ＵＬ U-phase low-side switch Q ＶＬ V-phase low-side switch Q ＷＬ W-phase low-side switch

Claims

1. A power conversion device comprising: a bridge circuit that performs mutual conversion between DC power and N-phase AC power (N is an integer of 3 or greater); and a control unit that controls the bridge circuit with pulse width modulation to individually control the duty ratio of pulses of each phase output from the bridge circuit, wherein the control unit shifts the rising timing of the pulses of each phase by a first time and shifts the falling timing of the pulses of each phase by a second time, under the condition that the duty ratio of the pulses of each phase is between 0% and 100% within one cycle of the pulse width modulation, and the first time and the second time change randomly for one or more cycles of the pulse width modulation.

2. The power conversion device according to claim 1, wherein the control unit determines the rise timing and fall timing of the pulse of each phase based on the duty command value of each phase during one cycle of the pulse width modulation, and the control unit shifts the rise timing of the pulse of each phase by the first time and shifts the fall timing of the pulse of each phase by the second time by adding an offset amount that is randomly determined for one or more cycles of the pulse width modulation to the duty command value of each phase.

3. The power conversion device according to claim 2, wherein when a duty ratio of 100% is 1, the duty command value of each phase is a value greater than or equal to 0 and less than or equal to 1, and the following formula (1) holds when the maximum duty command value of each phase is dmax, the minimum duty command value of each phase is dmin, and the offset amount is Δd: -dmin≦Δd≦1−dmax (1) 4. The power conversion device according to claim 2, wherein, when a duty ratio of 100% is taken as 1, the duty command value of each phase is a value greater than or equal to 0 and less than or equal to 1, and when the maximum duty command value of the duty command values ​​of each phase is taken as dmax and the minimum duty command value of the duty command values ​​of each phase is taken as dmin, the offset amount is randomly determined from three values: 0, 1-dmax, and -dmin.

5. The power conversion device according to claim 2, wherein when a duty ratio of 100% is taken as 1, the duty command value of each phase is a value greater than or equal to 0 and less than or equal to 1, and when the maximum duty command value of the duty command values ​​of each phase is taken as dmax and the minimum duty command value of the duty command values ​​of each phase is taken as dmin, the offset amount is randomly determined from two values, 0 and -dmin, or two values, 0 and 1-dmax, depending on the electrical angle of the N-phase motor to which the N-phase AC power is supplied.

6. The power conversion device according to claim 1, wherein the control unit determines the rise timing and fall timing of the pulse of each phase based on the duty command value of each phase during one period of the pulse width modulation, and the control unit shifts the rise timing of the pulse of each phase by the first time by adding a first offset amount that is randomly determined for one or more periods of the pulse width modulation to the duty command value of each phase, and shifts the fall timing of the pulse of each phase by the second time by adding a second offset amount that is randomly determined for one or more periods of the pulse width modulation to the duty command value of each phase.

7. The power conversion device according to claim 6, wherein when a duty ratio of 100% is 1, the duty command value of each phase is a value greater than or equal to 0 and less than or equal to 1, and when the maximum duty command value of the duty command values ​​of each phase is dmax, the minimum duty command value of the duty command values ​​of each phase is dmin, the first offset amount is Δd1, and the second offset amount is Δd2, the following equations (3) and (4) hold: -dmin≦Δd1≦1-dmax (3) -dmin≦Δd2≦1-dmax (4) 8. A motor unit comprising: an N-phase motor; and a power conversion device according to any one of claims 1 to 7, which supplies N-phase AC power to the N-phase motor.

Images