JP2016005368A - Control device for rotary machine - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a control device for a rotary machine which reduces a battery DC current pulsation or a smoothing capacitor ripple current, the control device for the rotary machine being configured to operate a switching state of an inverter while using a pulse pattern.SOLUTION: In processing for evaluating the pulse pattern and storing optimal solution candidates, a pulse pattern setting section estimates a motor phase current from candidate pulse patterns (S22 and S23), estimates an inverter input current (S24) and further performs at least either "higher harmonic wave analysis of a battery DC current" or "estimation of the smoothing capacitor ripple current" (S25). An amplitude of a sixth-order multiple component closest to an LC resonant frequency fc is derived as an evaluation function Hyo (S26) and such a pulse pattern as to minimize the evaluation function Hyo is stored as the optimal solution candidate. Thus, the pulsation of the battery DC current or the smoothing capacitor ripple current can be reduced. Therefore, heating in a smoothing capacitor or a battery can be prevented.

Description

  The present invention relates to a rotating machine control device that controls driving of a rotating machine using an output voltage of a pulse waveform.

  Conventionally, in the control of a rotating machine such as a motor, PWM control has been widely used in which the fundamental wave voltage output from the controller is PWM modulated to manipulate the switching state of the inverter. However, the PWM control has problems such as a large switching loss and a decrease in the number of times of switching with respect to one electrical cycle at the time of high rotation due to a restriction on the number of times of switching, thereby increasing current distortion.

Therefore, for example, depending on the voltage modulation rate obtained from the required torque of the motor and the number of times of switching in the electrical (1/4) cycle, an output voltage pattern (hereinafter referred to as an optimal pulse waveform synchronized with the electrical angle of the motor) In other words, a technique is known in which loss is reduced by creating a “pulse pattern” and operating the switching state of an inverter using the pulse pattern.
In this technique, there is a problem in what index the created pulse pattern is evaluated to select an optimal pulse pattern. In the technique disclosed in Patent Document 1, a pulse pattern that minimizes motor power loss, power distortion, or torque ripple is selected.

JP 2013-162660 A

  With the prior art of Patent Document 1, it is possible to reduce power loss in the inverter and the electric motor. However, Patent Document 1 does not show a specific configuration of the inverter (power conversion device), and a resistance component and a reactance component of a smoothing capacitor that is usually provided at an input portion of the inverter and a wiring that connects the battery and the inverter. No mention is made.

  Therefore, a circuit model in which a smoothing capacitor is provided between the battery and the inverter is assumed (see FIG. 3). This model further takes into account the internal resistance of the battery, the resistance component and reactance component of the wiring connecting the battery and the inverter, and the resistance component of the wiring connecting the smoothing capacitor. Then, the pulsation due to LC resonance between the battery and the smoothing capacitor occurs in the battery DC current flowing in the wiring from the battery to the inverter as the motor is energized, and the ripple current accompanying the charging and discharging of the smoothing capacitor It turns out that occurs. Battery direct current pulsation and smoothing capacitor ripple current may cause heat generation of the smoothing capacitor and the battery.

Therefore, in determining the optimum pulse pattern in motor control using a pulse pattern, not only the power loss, power distortion, and torque ripple of the motor as in the prior art of Patent Document 1, but also between the battery and the inverter. It is also required to pay attention to reduction of battery direct current pulsation and smoothing capacitor ripple current at the inverter input section.
The present invention was created in view of the above points, and the object of the present invention is to control battery direct current pulsation or smoothing capacitor ripple current in a control device for a rotating machine that operates a switching state of an inverter using a pulse pattern. It is an object of the present invention to provide a control device for a rotating machine that can be reduced.

The present invention includes a pulse pattern setting unit that sets a pulse pattern of an output voltage that is synchronized with an electrical angle of a multiphase rotating machine having three or more phases, and a plurality of inverters based on the pulse pattern set by the pulse pattern setting unit. It is a control device for a rotating machine that controls the driving of the rotating machine by operating on / off of the switching element, converting the DC power of the battery to AC power, and outputting the AC power to the rotating machine.
The pulse pattern setting unit includes at least one of a candidate pulse pattern generation unit, a pulse pattern harmonic analysis unit, a phase current estimation unit, an inverter input current estimation unit, a battery DC current estimation unit, and a smoothing capacitor ripple current estimation unit. On the other hand, a pulse pattern determination unit is provided.

The candidate pulse pattern generation unit generates a plurality of candidate pulse patterns based on the modulation rate required for driving the rotating machine and the number of pulses in a predetermined electrical angle section.
The pulse pattern harmonic analysis unit analyzes the harmonic component of the candidate pulse pattern.
The phase current estimation unit estimates a phase current that flows in each phase of the rotating machine when the candidate pulse pattern is applied.
The inverter input current estimation unit estimates an inverter input current (Iinv_in) that is a current that flows through the input unit of the inverter when the candidate pulse pattern is applied.

The battery direct current estimation unit uses a transfer function between the inverter input current and the battery direct current for the battery direct current (Ibatt) that flows between the battery and a smoothing capacitor provided in the input unit of the inverter. DC current is estimated and harmonic components are analyzed.
The smoothing capacitor ripple current estimator estimates the smoothing capacitor ripple current using a transfer function between the inverter input current and the smoothing capacitor ripple current for the smoothing capacitor ripple current (Ic) that is the ripple current flowing through the smoothing capacitor, and generates a harmonic component. Is analyzed.
The pulse pattern determination unit determines an optimal pulse pattern using an evaluation function based on information including at least one of the harmonic of the battery direct current or the harmonic of the smoothing capacitor ripple current.

  In the controller for a rotating machine according to the present invention, the pulse pattern setting unit estimates the motor phase current from the candidate pulse pattern, estimates the inverter input current from the motor phase current and the pulse pattern, and further receives the inverter input current as an input. Using the transfer function, at least one of the battery direct current and the smoothing capacitor ripple current is estimated, and its harmonic component is analyzed. Then, an optimum pulse pattern is determined using an evaluation function based on information including at least one of the battery direct current or the harmonics of the smoothing capacitor ripple current.

  As a result, the rotating machine control device of the present invention can reduce battery direct current pulsation or smoothing capacitor ripple current in the rotating machine control device that operates the switching state of the inverter using a pulse pattern. Therefore, heat generation of the smoothing capacitor and the battery can be prevented. Further, there is a possibility that the effect of reducing the power loss, current distortion, and torque ripple of the rotating machine is obtained as a secondary effect.

Assuming that the multi-phase rotating machine is a three-phase rotating machine, the pulse pattern determining unit specifically determines the pulse pattern using the following evaluation function.
<1> Of the “sixth multiple component of the electric frequency of the motor” (hereinafter simply referred to as “sixth multiple component”) in harmonic analysis of the battery direct current or smoothing capacitor ripple current, The pulse pattern that minimizes the “amplitude of the sixth-order multiple component appearing in the vicinity of the resonance frequency” of the LC resonance that occurs in between is determined as the optimum pulse pattern.
<2> The pulse pattern that minimizes the “sum of the amplitudes of the sixth-order multiple components” in the harmonic analysis of the battery DC current or the smoothing capacitor ripple current is determined as the optimum pulse pattern.

  Alternatively, if the sum of the amplitudes of the sixth-order multiple components in the harmonic analysis of the <3> battery direct current or smoothing capacitor ripple current is less than a predetermined threshold, any one of the phase current, the inverter input current, or the pulse pattern The pulse pattern that minimizes the sum of the amplitudes of the sixth-order multiple component or the component corresponding to the sixth-order multiple in the above harmonic analysis may be determined as the optimum pulse pattern.

The schematic block diagram in the electric current feedback control system of the motor control apparatus by one Embodiment of this invention. The schematic block diagram in the torque feedback control system of the motor control apparatus by one Embodiment of this invention. FIG. 3 is a detailed diagram of a pulse pattern setting unit in FIGS. 1 and 2 and a circuit model diagram. (A) The figure explaining the pulse number of a line voltage pulse pattern. (B) The figure explaining the pulse parameter of each phase pulse pattern. The main flowchart of the pulse pattern setting by one Embodiment of this invention. FIG. 6 is a sub-flowchart of the first embodiment of S20 of FIG. (A) Waveform diagram and (b) FFT analysis diagram for explaining a model of an induced voltage between lines. The frequency characteristic figure of the transfer function of (a) inverter input current-battery DC current, (b) inverter input current-smoothing capacitor ripple current. FIG. 6 is a sub-flowchart of a second embodiment of S20 in FIG. FIG. 6 is a sub-flowchart of the third embodiment of S20 of FIG. The figure which shows the FFT result of the inverter input current, etc. at the time of using the line voltage pulse pattern by the prior art, and the pulse pattern. The figure which shows the FFT result of a pulse pattern same as the above. The figure which shows the FFT result of the inverter input current, etc. when the line voltage pulse pattern by one Embodiment of this invention and the pulse pattern are used. The figure which shows the FFT result of a pulse pattern same as the above. The figure explaining the pulsation reduction effect of battery direct current. The figure explaining the reduction effect of a smoothing capacitor ripple current. The figure explaining the temperature reduction effect of a smoothing capacitor. The figure which shows another example of the pulse pattern which can suppress LC resonance. The figure which shows the example of the pulse pattern which LC resonance suppression cannot be performed. The figure which shows the example of the pulse pattern which LC resonance suppression cannot be performed.

Embodiments of a control device for a rotating machine according to the present invention will be described below with reference to the drawings.
(One embodiment)
A schematic configuration of a synchronous motor control apparatus according to an embodiment of the present invention will be described with reference to FIGS. As shown in FIG. 1 and FIG. 2, the motor control device 60 as the “rotary machine control device” converts the DC power of the battery 11 into three-phase AC power by operating the switching state of the inverter 40. It outputs to the motor generator 50 as a “rotor”.

The motor generator 50 is a synchronous three-phase AC motor such as an IPMSM (embedded permanent magnet synchronous motor), and is used, for example, as a power source for a hybrid vehicle or an electric vehicle. The motor generator 50 has a function as an electric motor that generates torque by consuming electric power by a power running operation, and a function as a generator that generates regenerative electric power by receiving torque.
In the following description, the case where the motor generator 50 functions as an electric motor is mainly assumed, and is simply referred to as “motor 50”. In the vicinity of the rotor of the motor 50, a position sensor 55 such as a resolver is provided for detecting the rotor position (electrical angle) θ.

The battery 11 is a chargeable / dischargeable DC power source such as a nickel metal hydride battery or a lithium ion battery. An electric double layer capacitor and the like are also included in the “battery” of the present invention.
The inverter 40 includes switching elements 41 to 46 of three-phase upper and lower arms that are bridge-connected. The upper arm switching elements 41, 42, 43 and the lower arm switching elements 44, 45, 46 correspond to the U phase, the V phase, and the W phase, respectively. As the switching element, for example, an IGBT or the like is used.

A smoothing capacitor 31 that smoothes the input voltage is provided on the input side of the inverter 40. A voltage sensor 35 that detects the inverter input voltage VH is provided in parallel with the smoothing capacitor 31.
A known boost converter may be provided between the battery 11 and the inverter 40, and the DC voltage boosted by the boost converter may be regarded as the inverter input voltage VH.

  A model diagram corresponding to the circuit from the battery 11 to the inverter 40 is shown on the lower side of FIG. In FIG. 3, the part including the battery 11 is represented as the battery part 10, the wiring between the battery 11 and the inverter 40 input part is represented as the DC wiring part 20, and the part including the smoothing capacitor 31 is represented as the smoothing capacitor part 30. The battery unit 10 includes a battery 11 and an internal resistor 12. The DC wiring unit 20 includes a resistance component 22 and a reactance component 23. The smoothing capacitor unit 30 includes a smoothing capacitor 31, a resistance component 32, and a reactance component 33. In the following description, LC resonance characteristics are analyzed using this model.

Returning to FIG. 1 and FIG. 2, the configuration of the motor control device 60 will be described.
The motor control device 60 switches between the current feedback control method shown in FIG. 1 and the torque feedback control method shown in FIG. 2 according to the modulation rate, for example.
First, the configuration of the current feedback control method will be described with reference to FIG. The motor control device 60 includes a current command calculation unit 61, a current subtractor 62, a controller 63, and a modulation factor / voltage phase calculation unit 64 as a configuration of current feedback control. Further, as a common configuration, a dq conversion unit 65, a differentiator 66, a pulse number setting unit 67, and a pulse pattern setting unit 70 are provided.

The current command calculation unit 61 calculates dq-axis current command values Id ^* and Iq ^* using a map, a mathematical formula, or the like based on the torque command value trq ^* acquired from the host controller.
The current subtractor 62 subtracts the dq axis current detection values Id and Iq fed back from the dq converter 65 from the dq axis current command values Id ^* and Iq ^* to calculate a dq axis current deviation.
The controller 63 calculates the dq axis voltage command values Vd ^* and Vq ^* by PI control calculation or the like so that the dq axis current deviation converges to zero.

また、（Ｖｑ^*／Ｖｄ^*）のアークタンジェントにより、電圧位相φを演算する。 The modulation factor / voltage phase calculator 64 calculates the modulation factor m using Equation 1 based on the inverter input voltage VH and the dq axis voltage command values Vd ^* and Vq ^* .

Further, the voltage phase φ is calculated by the arc tangent of (Vq ^* / Vd ^* ).

The dq converter 65 receives the phase current detection value from the current sensors 51 and 52 provided on the power line connected from the inverter 40 to the MG 50. In this embodiment, detected values of the V-phase current Iv and the W-phase current Iw are input from the current sensors 51 and 52 provided in the V-phase and the W-phase, and the remaining U-phase current Iu is estimated based on Kirchhoff's law. doing. In other embodiments, any two-phase current may be detected, and a three-phase current may be detected. Or you may employ | adopt the technique which estimates the other two-phase electric current based on the electric current detection value of one phase.
The dq converter 65 uses the electrical angle θ acquired from the position sensor 55 to dq convert the three-phase current detection values Iu, Iv, and Iw into dq-axis current detection values Id and Iq, and feeds back to the current subtractor 62. .
The differentiator 66, the pulse number setting unit 67, and the pulse pattern setting unit 70 will be described together after the description of the torque feedback control method.

  Next, the configuration of the torque feedback control system will be described with reference to FIG. The motor control device 60 includes a command voltage setting unit 81, a multiplier 82, a modulation factor calculation unit 83, a torque estimation unit 84, a filter 85, a torque subtractor 86, and a controller 87 as a configuration of torque feedback control. . Further, as a common configuration, a dq conversion unit 65, a differentiator 66, a pulse number setting unit 67, and a pulse pattern setting unit 70 are provided.

The command voltage setting unit 81 can realize the most efficient control for obtaining the maximum torque with the minimum current based on the inverter input voltage VH and the torque command value trq ^* or the torque estimated value trq_est (shown by a broken line). The ratio (Va / ωb) between the voltage amplitude Va and the speed ωb is determined by a map. Here, ωb is the speed when the map is measured.
The multiplier 82 multiplies the ratio (Va / ωb) determined by the command voltage setting unit 81 by the electrical angular velocity ω calculated by the differentiator 66.
The modulation factor calculator 83 calculates the modulation factor m based on the inverter input voltage VH and the voltage amplitude (Va × ω / ωb) = Va ′.

Based on the dq axis currents Id and Iq converted by the dq conversion unit 65, the torque estimation unit 84 estimates a torque estimation value trq_est using a map, a known torque equation, or the like.
The filter 85 is a filter for attenuating a ripple component that is synchronized with the electrical cycle generated in the overmodulation region. For example, “1 / {(1 / kω ₁ ) s + 1}” corresponding to the electrical frequency ω ₁ is used. A primary low-pass filter (LPF).

The torque subtractor 86 subtracts the estimated torque value trq_est from the torque command value trq ^* to calculate a torque deviation.
The controller 87 calculates the voltage phase φ by PI control calculation or the like so that the torque deviation converges to zero.

Subsequently, a configuration common to the current feedback control method (FIG. 1) and the torque feedback control method (FIG. 2) will be described.
The pulse number setting unit 67 sets the number of pulses n (described later) of the line voltage pulse pattern according to the electrical angular velocity ω calculated by the differentiator 66 by time differentiation of the electrical angle θ.
The pulse pattern setting unit 70 sets a pulse pattern for driving the inverter 40 based on the modulation factor m, the voltage phase φ, and the number of pulses n. The detailed configuration of the pulse pattern setting unit 70 will be described later.

  Here, an example of the pulse pattern is shown in FIG. The term “pulse pattern” used in the present embodiment includes both “each phase pulse pattern” and “line voltage pulse pattern”. FIG. 4A shows the respective phase pulse patterns of the switching element 41 of the U-phase upper arm and the switching element 42 of the V-phase upper arm, and the U-V phase line voltage pulse pattern. Each phase pulse pattern and the line voltage pulse pattern correspond to each other.

  The pulse pattern of the switching element 42 of the V-phase upper arm is shifted from the pulse pattern of the switching element 41 of the U-phase upper arm by 120 electrical degrees. In addition, if the dead time is ignored, the pulse pattern of the switching element in the lower arm of each phase has a shape that complements the pulse pattern of the switching element in the upper arm of the same phase, that is, a vertically inverted form.

  When the line voltage pulse pattern is shifted 60 deg in the negative direction so that the electrical angle 60 deg on the horizontal axis in FIG. 4A is 0 deg, the voltage in the interval between 0 and 180 deg. The region where the voltage in the 180 to 360 deg interval is negative is a relationship in which the sign is inverted, that is, a “half-wave symmetric” relationship. The section of electrical angle 0-90 deg and the section of electrical angle 90-180 deg are centered on the electrical angle 90 deg, the section of electrical angle 180-270 deg and the section of the electrical angle 270-360 deg are centered on the electrical angle 270 deg, Each has a line-symmetric relationship. Thus, when considering the symmetry of the pulse pattern, the electrical angle reference (0 deg) is appropriately shifted.

In addition, the “number of pulses n” of the line voltage pulse pattern means the number of pulses in the electrical (1/2) period, that is, the electrical angle of 180 degrees, and “n = 7” in the example of FIG. Become.
In the following description of the pulse pattern, if distinction is necessary, it is clearly indicated as “each phase pulse pattern” or “inter-line voltage pulse pattern”, and it is obvious whether it is pointing or can be interpreted in any way. When the meaning is established, it is simply written as “pulse pattern”.

The pulse pattern is output as switching signals UU, UL, VU, VL, WU, WL for the switching elements 41 to 46 of the inverter 40. The switching elements 41 to 46 of the inverter 40 are operated in accordance with the switching signals UU, UL, VU, VL, WU, WL, so that the driving of the MG 50 is controlled so that torque according to the torque command value trq ^* is output. The
The switching frequency can be reduced by operating the inverter 40 using a pulse pattern instead of the PWM signal widely used in motor control. Therefore, especially in a high rotation overmodulation region, switching loss is reduced. Is effective.

Next, a detailed configuration of the pulse pattern setting unit 70 will be described with reference to FIG. The detailed configuration of the pulse pattern setting unit 70 and the specific procedure of pulse pattern setting described later correspond to the gist of the present invention, that is, the characteristic part.
The pulse pattern setting unit 70 of this embodiment includes a candidate pulse pattern generation unit 71, a pulse pattern harmonic analysis unit 72, a phase current estimation unit 73, an inverter input current estimation unit 74, a battery DC current estimation unit 75, and a smoothing capacitor ripple current. An estimation unit 76 and a pulse pattern determination unit 77 are included. Here, the battery direct current estimation unit 75 and the smoothing capacitor ripple current estimation unit 76 may have only one of them.

The role of each part will be described together with the pulse pattern setting procedure. Here, the current estimated by the inverter input current estimation unit 74, the battery DC current estimation unit 75, and the smoothing capacitor ripple current estimation unit 76, and the circuit model diagram, respectively. Will be described.
In the circuit model diagram, a direct current flowing between the smoothing capacitor unit 30 and the inverter 40 is defined as “inverter input current Iinv_in”, and the inverter input current estimation unit 74 estimates this. The direct current flowing between the battery 11 and the smoothing capacitor 31 is defined as “battery direct current Ibatt”, and the battery direct current estimation unit 75 estimates this. The ripple current flowing through the smoothing capacitor 31 is defined as “smoothing capacitor ripple current Ic”, and the smoothing capacitor ripple current estimation unit 76 estimates this.

Here, the notation of symbols in the present specification, drawings, and mathematical expressions in the specification will be omitted. For example, regarding the current, in the drawings and the mathematical formulas below, the “inv_in”, “batt”, and “c” portions are described by subscripts. However, in the specification text, it may be difficult to see if all of these are written in subscripts, so they may be written in normal letters as appropriate depending on the connection with the main symbol (in this case, “I”). .
In addition, for current “I” and voltage “V”, lowercase letters “i” and “v” may be used in combination for convenience of mathematical expressions.

For example, in the prior art described in Patent Document 1, the pulse pattern is set so as to minimize power loss, current distortion, or torque ripple in the motor 50. However, regarding the ripple current of the smoothing capacitor 31 provided at the input part of the inverter 40 and the DC current pulsation caused by the LC resonance in the DC wiring part 20 connecting the battery 11 and the inverter 40, they are the smoothing capacitor 31 and the battery. No mention is made in spite of the possibility of 11 heat generation.
Therefore, the present invention is characterized by focusing on the battery DC current Ibatt and the smoothing capacitor ripple current Ic, and setting the pulse pattern so that the pulsation of the battery DC current Ibatt and the smoothing capacitor ripple current Ic can be reduced. .

Next, each phase pulse pattern setting procedure by the pulse pattern setting unit 70 will be described with reference to FIGS.
As shown in FIG. 4B, each phase pulse pattern according to the present embodiment is set by a model in which a section of an electrical angle of 0 to 90 deg is a unit. In the example of FIG. 4B, three off periods and three on periods are set in the section of electrical angle 0 to 90 deg. The angle width from the electrical angle 0 deg to the first on-period is defined as a first angle width δ1. In the second and third off periods, the center positions are θ1 and θ2, and (½) of the width of the off period is the angular widths δ2 and δ3. These center positions θ1, θ2 and angular widths δ1, δ2, δ3 are referred to as “pulse parameters”. If each pulse parameter is determined, the pulse pattern in the section of electrical angle 0 to 90 deg is determined.

Then, a pulse pattern with an electrical angle of 90 to 180 deg is set by reversing the pulse pattern with an electrical angle of 0 to 90 deg about line symmetric about the electrical angle of 90 deg. In addition, a pulse pattern having an electrical angle of 180 to 360 deg is set by inverting a pulse pattern having an electrical angle of 0 to 180 deg in a point-symmetric manner with the electrical angle of 180 deg as the center. “Point symmetry” means to invert line symmetry and to invert the on and off sides of the pulse.
As described above, each phase pulse pattern of one electrical cycle for one phase (for example, U phase) is determined. The other two phases (V phase and W phase) are set by shifting the one-phase pulse pattern by an electrical angle of ± 120 deg.

By the way, when the number of pulses is set to be larger, parameters after the center position θ3 and after the angular width δ4 are added. However, in the following description regarding FIG. 4 to FIG. 10, the description will be made on the premise of “pulse pattern having five pulse parameters” shown in FIG.
That is, in the present embodiment, “setting an optimal pulse pattern” specifically means obtaining optimum values of the pulse parameters δ1, δ2, δ3, θ1, and θ2.

Next, the main flow of setting each phase pulse pattern will be described with reference to FIG. In the following description of the flowchart, the symbol “S” means a step.
S11 to S14 are executed by the candidate pulse pattern generation unit 71. In S11, “the desired modulation rate m of the pulse pattern to be designed” calculated by the modulation rate / voltage phase calculation unit 64 or the modulation rate calculation unit 83 is set. In S12, initial values of the center positions θ1, θ2 in the candidate pulse pattern are set, and in S13, initial values of the angular widths δ2, δ3 in the candidate pulse pattern are set. If the number of pulses of the line voltage pulse pattern is greater than 5, and there is a center position θ3 or later and an angle width δ4 or later, these are set including them.

In S14, the first angular width δ1 is calculated using Equation 2 based on the desired modulation factor m, the center positions θ1, θ2, and the angular widths δ2, δ3.

In S20, the candidate pulse pattern which becomes an optimal solution by pulse pattern evaluation is stored. Details of this step will be described later.
After the end of S20, it is determined in order whether the search range has ended for the parameters of the angular widths δ3 and δ2 and the center positions θ2 and θ1 (S31, S33, S35, and S37). If the search range of any parameter has not ended, the parameter is changed by the minute values dδ and dθ (S32, S34, S36, S38), the process returns to S14, and the first angular width δ1 is recalculated.
When all the search ranges of the angular widths δ3 and δ2 and the center positions θ2 and θ1 are completed (S37: YES), the pulse pattern determination unit 77 outputs the stored candidate pulse pattern of the optimal solution as the optimal solution.
While driving the motor 50, the pulse pattern setting unit 70 may calculate an optimal solution online each time, or store the previously calculated optimal solution in a map using the modulation factor m and the number of pulses n as arguments. Good.

Next, details of the “pulse pattern optimum solution candidate storage” process of S20 will be described with reference to the sub-flowcharts of FIGS. 6, 9, and 10.
First, S21 to S25 common to the three embodiments will be described with reference to FIG.
In S21, the pulse pattern harmonic analysis unit 72 performs harmonic analysis on the pulse pattern. The following harmonic analysis is performed using FFT (Fast Fourier Transform). The (6k ± 1) -order, that is, the fifth-order, seventh-order, eleventh-order, thirteenth-order, etc. harmonic amplitude of each phase pulse pattern is expressed by Equation 3. In Equation 3, information on the modulation factor m is reflected in the first angular width δ1, and information on the number of pulses n is reflected in the number of terms “l” in the Σ calculation.

In S22 and S23, the phase current estimation unit 73 estimates the motor phase current Iuvw that flows in each phase of the motor 50 when applying the candidate pulse pattern, and performs harmonic analysis.
In S22, the dq-axis voltage applied to the motor 50 when the pulse pattern is applied at the inverter input voltage VH is calculated by Equation 4. Formula 4 indicates that the pulse pattern is applied with a fundamental phase φ _v1 of the voltage phase with respect to the electrical angle θ for one phase, and is applied with a phase shifted by ± 120 deg from the other two phases. This is an assumed formula.

In S23, harmonic analysis is performed on the harmonics of the sixth-order multiple component (shown as “6k” in the equation) of the motor phase current Iuvw. First, the voltage equation of the dq axis voltage is expressed by Equation 5 when the flux linkage is described as a two-variable function of id and iq.

However,
R: Armature resistance id, iq: d-axis current, q-axis current s: differential operator (see Equation 6)
Ld, Lq: Inductance (see Equation 6)
Ld ′, Lq ′, Ldq ′, Lqd ′: differential inductance (see Equation 6)
ω ₁ : Electrical angular velocity (rotation speed)
λd, λq: d-axis component of magnetic flux, q-axis component ed, eq: d-axis harmonic component of induced voltage, q-axis harmonic component Ke: induced voltage coefficient

ここで、高回転領域の動作に注目すると、「Ｒ＜＜ωＬ」であるため「Ｒ≒０」とみなす。また、ｄｑ軸相互間での鎖交磁束の影響を無視し「Ｌｄｑ’＝Ｌｑｄ’≒０」とみなすと、６次倍数成分高調波についての電圧方程式は、簡略化した数式７で表される。

Here, when attention is paid to the operation in the high rotation region, since “R << ωL”, it is regarded that “R≈0”. When the influence of the interlinkage magnetic flux between the dq axes is ignored and “Ldq ′ = Lqd′≈0” is assumed, the voltage equation for the sixth-order multiple component harmonic is expressed by the simplified expression 7. .

数式７中のｄｑ軸電圧（ｖ）、電流（ｉ）、誘起電圧（ｅ）の末尾の「ｈ（６ｋ）」は、６ｋ（ｋは自然数）次の値であることを示す。

“H (6k)” at the end of the dq-axis voltage (v), current (i), and induced voltage (e) in Expression 7 indicates a 6k (k is a natural number) order value.

When “s = jω ₁ ” is substituted in Equation 7 to calculate the inverse matrix, Equation 8 indicating the dq-axis current of the sixth-order multiple component is obtained.

Here, with reference to FIG. 7, it is considered to model the induced voltage of the last term on the right side of Expression 8.
As shown in FIG. 7A, when the line voltage between the U and V phases is measured with an actual machine, for example, a waveform in which the distortion component of the induced voltage is superimposed is obtained. When this line voltage is subjected to FFT analysis, as shown in FIG. 7B, the main component of the distortion is (6k ± 1) order (5th, 7th, 11th, 13th, 17th, 19th,. ..) Components are extracted. The extracted component is dq-axis coordinate transformed, and the d-axis component and the q-axis component of the (6k ± 1) -th line induced voltage vector at the line induced voltage phase φ _e1 with respect to the electrical angle θ are expressed by Equation 9.

Next, when the dq-axis current in Expression 8 is converted into a phase current, the (6k ± 1) -order component of the U-phase current iu is expressed by Expression 10. The second equation of Equation 10 is a Fourier series expansion of the first equation. The (6k ± 1) -order component of the V-phase current iv and the W-phase current iw is obtained by shifting the phase of Equation 10 by the electrical angle ± 120 deg.

The amplitude iu _amp (6k ± 1) of the (6k ± 1) -order component of the U-phase current is expressed by Equation 11.

The Fourier coefficients A _{d — k} , A _{q — k} , B _{d — k} , and B _{q — k} in Equations 10 and 11 are expressed by Equation 12, and the coefficient Δ _k based on the determinant is expressed by Equation 13.

In Equation 12, the voltage coefficients V _dc , V _ds , V _qc , V _qs corresponding to the cos term and the sin term, and the line induced voltage coefficients E _dc , E _ds , corresponding to the cos term and the sin term, E _qc and E _qs are expressed by Equations 14 and 15, respectively.

Further, in Equation 14,15 (6k ± 1) the voltage amplitude v _{6k ± 1,} and voltage phase _φ v_6k _{± 1} of the following components expressed in Equation 16, (6k ± 1) induced between the lines of the next component voltage amplitude e _{6k ± 1} and the line induced voltage phase φ _{e — 6k ± 1} are expressed by Expression 17. The phases of Equations 16 and 17 are in rad units.

The symbols in Equations 16 and 17 have the following meanings.
M _{6k ± 1} : Amplitude of the (6k ± 1) order harmonic of the pulse pattern (see Equation 3)
φ _v1 : fundamental phase of voltage phase (see Equation 4)
K _{e — 6k ± 1} : (6k ± 1) order harmonic amplitude of line induced voltage coefficient φ _{e_6k ± 1} : (6k ± 1) order harmonic phase of line induced voltage coefficient

Next, in S24, the inverter input current estimation unit 74 estimates the inverter input current Iinv_in for one electrical cycle.
As shown in Table 1, in the three-phase AC inverter 40, two patterns of zero voltage vectors V0 and V7 and six patterns of effective voltage vectors V1 to V7 are generated corresponding to the switching pattern. In the “switching state” column of Table 1, “1” indicates that the switching element of the upper arm is ON, and “0” indicates that the switching element of the upper arm is OFF. Further, an inverter input current Iinv_in as shown in the right column of the table flows corresponding to the voltage vectors V0 to V7.
An FFT analysis is performed on the inverter input current Iinv_in estimated based on Table 1, and the amplitude and phase of the inverter input current harmonic are calculated.

In S25, at least one of “estimation and harmonic analysis of battery direct current Ibatt” by battery direct current estimation unit 75 or “estimation and harmonic analysis of smoothing capacitor ripple current Ic” by smoothing capacitor ripple current estimation unit 76 is performed. . For this estimation calculation, a transfer function between the inverter input current Iinv_in and the battery DC current Ibatt and a transfer function between the inverter input current Iinv_in and the smoothing capacitor ripple current Ic are used.
In the circuit model considering the resistance component and reactance component of the battery unit 10, the DC wiring unit 20, and the smoothing capacitor unit 30 shown in FIG. 3, the wiring impedance Zdc of the DC wiring unit 20 and the impedance Zc of the smoothing capacitor 31 are , Expressed by Equation 18.

From the internal resistance Rbatt of the battery 11, the wiring impedance Zdc of the DC wiring unit 20, and the impedance Zc of the smoothing capacitor 31, the transfer function of the inverter input current Iinv_in and the battery DC current Ibatt, and the inverter input current Iinv_in and the smoothing capacitor ripple The transfer functions with the current Ic are expressed by Equations 19 and 20, respectively.

共振周波数ｆｃ付近では、伝達関数の入力が共振により増幅されるため、バッテリ直流電流Ｉｂａｔｔの脈動や平滑コンデンサリプル電流Ｉｃが増大する。 When the transfer characteristics of Equations 19 and 20 are represented by frequency characteristic diagrams, they are as shown in FIGS. 8A and 8B, and both exhibit LC resonance characteristics. For simplicity, when the entire reactance component of the circuit portion from the battery 11 to the smoothing capacitor 31 is represented by L, the LC resonance frequency fc is represented by Equation 21.

In the vicinity of the resonance frequency fc, the input of the transfer function is amplified by resonance, so that the pulsation of the battery DC current Ibatt and the smoothing capacitor ripple current Ic increase.

Hereinafter, the steps after S25 will be described for each example of the “pulse pattern optimum solution candidate storage” process. In the process up to S25, the amplitude “I _{batt — 6f} , I _{batt — 12f} ,... I _{batt — (6k) f} ” of the battery DC current Ibatt, or the amplitude “6 b of the smoothing capacitor ripple current Ic” _{I c_6f, I c_12f, ··· I} c_ (6k) f "is calculated.

In the first embodiment shown in FIG. 6, “the amplitude of the sixth-order multiple component closest to the LC resonance frequency fc” is derived as the evaluation function Hyo in S26 after S25. For example, if the frequency of the 12th-order component is the closest to the LC resonance frequency fc among the 6th-order, 12th-order, 18th-order, etc., pay attention only to the amplitude of the 12th-order component, and other 6th-order components and 18th-order components Etc. are not considered.
In S28, it is determined whether or not the evaluation function Hyo of the candidate pulse pattern derived this time is smaller than the minimum value Hyo_min of the evaluation function of the candidate pulse pattern derived so far.

  In the case of YES in S28, in S29, the evaluation function Hyo of the candidate pulse pattern derived this time is set to “minimum value Hyo_min of evaluation function”, and is stored as the pulse pattern of the optimal solution candidate. That is, the current parameters δ1, δ2, δ3, θ1, and θ2 are set to optimum values δ1_opt, δ2_opt,. On the other hand, in the case of NO in S28, the subflow of S20 is terminated as it is, and the process returns to the main flow of FIG.

  In the second embodiment shown in FIG. 9, “the sum of the amplitudes of the sixth-order multiple components” is derived as the evaluation function Hyo in S27 after S25. Similarly to the first embodiment, if the evaluation function Hyo of the candidate pulse pattern derived this time is smaller than the minimum value Hyo_min of the evaluation function of the candidate pulse pattern derived so far, the candidate pulse pattern derived this time is optimized. Store as a pulse pattern of the solution candidate.

The third embodiment shown in FIG. 10 is common to the second embodiment in that the candidate pulse pattern is evaluated based on “the sum of the amplitudes of the sixth-order multiple components”. However, two types of evaluation functions Hyo1 and Hyo2 are used, and the processing is different in that it includes cases.
In S271, as in S27 of the second embodiment, “the sum of the amplitudes of the sixth-order multiple components” is derived as the first evaluation function Hyo1. In S272, the sum of the amplitudes of the (6k ± 1) -order components of the phase current (for example, the U-phase current Iu), the sum of the amplitudes of the sixth-order multiple components of the inverter input current Iinv_in, and (6k ± 1) of the pulse pattern M At least one of the total sums of the amplitudes of the next components is derived as the second evaluation function Hyo2.

When two or three of these three are used, the second evaluation function Hyo2 may be overlapped by multiplying by weighting factors A, B, and C as shown in Equation 22.

  In S281, it is determined whether or not the first evaluation function Hyo1 of the candidate pulse pattern derived this time is less than a predetermined threshold value Hyo_th. If YES in S281, it is determined in S282 whether or not the second evaluation function Hyo2 of the candidate pulse pattern derived this time is smaller than the minimum value Hyo_min of the evaluation function of the candidate pulse pattern derived so far.

In the case of YES in S282, in S292, the second evaluation function Hyo of the candidate pulse pattern derived this time is set as the “minimum value Hyo_min of the evaluation function” and stored as the pulse pattern of the optimal solution candidate.
That is, in the third embodiment, when the sum of the amplitudes of the sixth-order multiple components in the harmonic analysis of the battery DC current Ibatt or the smoothing capacitor ripple current Ic is less than a predetermined threshold, the phase current Iuvw, the inverter input current Iinv_in, or the pulse An optimal pulse pattern is determined based on the pattern information.

  Next, a comparison between an example of a pulse pattern according to the prior art and an example of a pulse pattern according to the present embodiment will be described with reference to FIGS. 11 and 12 show (a) a line voltage pulse pattern according to the prior art, (b) a phase current waveform, (c) transfer characteristics between the inverter input current Iinv_in and the battery DC current Ibatt, and the inverter input current Iinv_in. (D) shows the FFT result of the pulse pattern. 13 and 14 show similar views according to the present embodiment.

  11 (c) and 13 (c), the transfer function between the inverter input current Iinv_in and the battery direct current Ibatt has a maximum gain at the LC resonance frequency fc. The gain is substantially constant on the lower frequency side than the frequency fc, and the gain decreases as the frequency increases on the higher frequency side than the frequency fc.

Regarding the FFT spectrum of the inverter input current Iinv_in, in the present embodiment, the spectrum moves to the high frequency side as compared with the prior art. When attention is paid to the sixth-order multiple component in the vicinity of the LC resonance frequency fc, the current amplitude is lower in the present embodiment than in the prior art.
This means that the LC resonance can be reduced by driving the motor 50 using the pulse pattern of the present embodiment. Further, since the gain is reduced on the higher frequency side than the frequency fc, the influence of the spectrum moved to the higher frequency side is reduced.
The same can be said about this tendency even if the FFT results of the pulse patterns in FIGS. 12D and 14D are compared. Alternatively, the same tendency can be obtained by comparing the FFT results of the phase currents.

The effect of this embodiment will be described with reference to FIGS. FIG. 15 shows the pulsation reduction effect of the battery DC current Ibatt, and FIGS. 16 and 17 show the reduction of the smoothing capacitor ripple current Ic and the temperature reduction effect of the smoothing capacitor 31.
In FIG. 15, the motor 50 is used under the conditions of (a) a pulse pattern according to the prior art and (b) a pulse pattern according to the present embodiment, an inverter input voltage (VH) of 120 V, a modulation factor of 111%, and the number of pulses of 15 times. The waveforms of the U-phase current Iu, the V-phase current Iv, and the battery DC current Ibatt when driving is shown. In the conventional pattern, the distortion of the phase current is large and the pulsation of the battery DC current Ibatt is large, whereas in the pattern according to the present embodiment, the distortion of the phase current is small and the pulsation of the battery DC current Ibatt is reduced.

  FIGS. 16A and 16B show the smoothing capacitor ripple current Ic when the motor 50 is driven using the pulse pattern according to the prior art and the pulse pattern according to the present embodiment. FIG. 16C shows transfer characteristics between the inverter input current Iinv_in and the smoothing capacitor ripple current Ic, and the FFT result of the smoothing capacitor ripple current Ic.

In FIG. 16C, the transfer function between the inverter input current Iinv_in and the smoothing capacitor ripple current Ic has a maximum gain at the LC resonance frequency fc. The gain decreases as the frequency increases on the lower frequency side than the frequency fc, and the gain is substantially constant on the higher frequency side than the frequency fc. That is, the transfer function between the inverter input current Iinv_in and the battery direct current Ibatt tends to be opposite to the frequency.
As shown in FIGS. 16A and 16B, in the pattern according to the present embodiment, the smoothing capacitor ripple current Ic is reduced as compared with the conventional pattern. Further, from the FFT result shown in FIG. 16C, it can be seen that the sixth-order multiple component near the LC resonance frequency fc is significantly reduced.

  FIG. 17 shows the temperature rise of the smoothing capacitor 31 over time when the motor 50 is driven using the pulse pattern according to the prior art set to minimize the voltage distortion and the pulse pattern according to the present embodiment. Show. In this embodiment, the temperature can be reduced by 5.6 ° C. compared to the prior art after 40 minutes from the start of driving.

Next, another example of a pulse pattern that can suppress LC resonance generated according to the present embodiment and an example (two) of pulse patterns that cannot suppress LC resonance are shown in FIGS. Shown in In these three examples, the line voltage pulse pattern is set with 7 pulses.
18 to 20, as in FIGS. 11 and 13, (a) line voltage pulse pattern, (b) phase current waveform, (c) transfer characteristics between inverter input current Iinv_in and battery DC current Ibatt, and The FFT result of the inverter input current Iinv_in is shown. Focusing on the sixth-order multiple component in the vicinity of the LC resonance frequency fc in (c), the pattern of FIG. 18 has a relatively small current amplitude and can suppress LC resonance. On the other hand, the patterns of FIGS. 19 and 20 have a relatively large current amplitude and cannot suppress LC resonance.

  As described above through the plurality of examples, in the motor control device 60 of the present embodiment, the pulse pattern setting unit 70 estimates the motor phase current Iuvw from the candidate pulse pattern, and the inverter input current from the motor phase current Iuvw and the pulse pattern. After estimating Iinv_in, further, at least one of the battery DC current Ibatt and the smoothing capacitor ripple current Ic is estimated using a transfer function having the inverter input current Iinv_in as an input, and the harmonic component thereof is analyzed. Then, based on information including at least one of the battery direct current Ibatt and the harmonics of the smoothing capacitor ripple current Ic, an optimum pulse pattern is determined using an evaluation function.

  Thereby, the motor control device 60 of the present embodiment can reduce the pulsation of the battery DC current Ibatt or the smoothing capacitor ripple current Ic. Therefore, heat generation of the smoothing capacitor 31 and the battery 11 can be prevented. Further, there is a possibility that a secondary power loss, current distortion, and torque ripple reduction effect can be obtained.

(Other embodiments)
(A) In the harmonic analysis of the motor phase current in the above-described embodiment (S23 in FIG. 6 and the like), Equation 7 is derived by assuming that “R≈0” and “Ldq ′ = Lqd′≈0” in Equation 5. . On the other hand, when Expression 5 is strictly used, an expression indicating the dq-axis current of the sixth-order multiple component is expressed by Expression 23.

Further, when substituting “s = jω ₁ ” in Equation 23 and calculating the inverse matrix, the exact Equation 24 corresponding to Equation 8 is obtained. By using Equation 24, more accurate analysis can be performed when the R term cannot be ignored in the low rotation region or when the influence of the flux linkage between the dq axes is taken into consideration.

(A) The method of harmonic analysis is not limited to FFT (Fast Fourier Transform), and any method may be used.
Further, each current may be analyzed by electromagnetic field analysis instead of harmonic analysis. Specifically, commercially available electromagnetic field analysis software (JMAG (registered trademark) or the like) can be used.

  (C) In the above embodiment, attention is paid to the harmonics of the sixth-order multiple component on the assumption that the number of phases of the multiphase rotating machine is three phases, but in the case where the number of phases is other than three phases, The order of the harmonics of interest may be set as appropriate.

(D) The definition of the “pulse number n” of the pulse pattern is not limited to that illustrated in FIG. 4A, and may be, for example, the number of pulses in one cycle of the line voltage pulse pattern.
(E) The evaluation function for obtaining the optimal solution of the pulse pattern is not limited to that exemplified in the above embodiment. For example, instead of calculating the sum of squares in S27 of FIG. 9, S271, S272 of FIG. 10, etc., other functions may be used. Further, a value obtained by combining two or more of the sums of the 6th multiple components of the battery direct current Ibatt, the smoothing capacitor ripple current Ic, and the inverter input current Iinv_in may be used as the evaluation function.

  (F) In the description of FIG. 1 in the above embodiment, it is stated that a known boost converter may be provided between the battery 11 and the inverter 40, and the DC voltage boosted by the boost converter may be regarded as the inverter input voltage VH. It was. In that case, the mathematical formula of the estimation calculation may be changed as appropriate, for example, by estimating the smoothing capacitor ripple current and the inverter DC current pulsation in consideration of the electrical transfer characteristics of the boost converter.

  (G) The rotating machine to be controlled in the present invention is not limited to synchronous motors such as IPMSM and SPMSM, but can be widely applied to all three-phase or more multi-phase rotating machines including induction motors. Furthermore, a rotating machine such as a motor generator is not limited to one used as a power source for a vehicle, but may be used for an auxiliary machine of a vehicle, a train other than a vehicle, an elevator, a general machine, or the like.

(H) The present invention focuses on the LC resonance between the battery and the smoothing capacitor in the control device for the rotating machine that operates the switching state of the inverter using the pulse pattern, and determines the battery DC current pulsation and the smoothing capacitor ripple current. The purpose is to set the pulse pattern so as to reduce. However, this concept of the present invention can be applied not only to pulse pattern control but also to PWM control.
For example, regarding the setting of the carrier frequency for PWM control, it has been conventionally known that the controllability increases as the frequency is increased, but the efficiency decreases due to the increase in switching loss, and the optimum value is selected according to the required characteristics. The In addition, since the harmonic components of the battery DC current and the smoothing capacitor ripple current change according to the carrier frequency, the optimum carrier frequency for avoiding LC resonance is selected by applying the technical idea of the present invention. be able to.

  As mentioned above, this invention is not limited to the said embodiment at all, In the range which does not deviate from the meaning of invention, it can implement with a various form.

11 ... Battery,
31 ... smoothing capacitor,
40: Inverter,
41-46 ... switching elements,
50 ... Motor (motor generator, rotating machine)
60: Motor control device (rotary machine control device),
70: Pulse pattern setting section,
71 ... Candidate pulse pattern generation unit,
72 ... pulse pattern harmonic analysis unit,
73: Phase current estimation unit,
74: Inverter input current estimation unit,
75... Battery direct current estimation unit,
76 ... smoothing capacitor ripple current estimation unit,
77: A pulse pattern determination unit.

Claims (13)

A pulse pattern setting unit (70) for setting a pulse pattern of an output voltage synchronized with an electrical angle of a multiphase rotating machine (50) of three or more phases, and an inverter based on the pulse pattern set by the pulse pattern setting unit The rotation which controls the drive of the said rotary machine by operating on / off of the some switching element (41-46) of (40), converting the direct current power of a battery (11) into alternating current power, and outputting it to the said rotary machine Machine control device (60),
The pulse pattern setting unit
A candidate pulse pattern generation unit (71) for generating a plurality of candidate pulse patterns based on the modulation rate required for driving the rotating machine and the number of pulses in a predetermined electrical angle section;
A pulse pattern harmonic analysis unit (72) for analyzing a harmonic component of the candidate pulse pattern;
A phase current estimation unit (73) for estimating a phase current (Iuvw) flowing in each phase of the rotating machine when the candidate pulse pattern is applied;
An inverter input current estimation unit (74) that estimates an inverter input current (Iinv_in) that is a current that flows through the input unit of the inverter when the candidate pulse pattern is applied based on the phase current;
The battery direct current (Ibatt), which is a current flowing between the battery and a smoothing capacitor (31) provided at the input portion of the inverter, is calculated using a transfer function between the inverter input current and the battery direct current. A battery direct current estimation unit (75) for estimating a battery direct current and analyzing a harmonic component; or
A smoothing capacitor ripple current (Ic) that is a ripple current flowing through the smoothing capacitor is estimated by using the transfer function between the inverter input current and the smoothing capacitor ripple current to estimate the smoothing capacitor ripple current and analyze a harmonic component. At least one of the capacitor ripple current estimator (76);
A pulse pattern determination unit (77) for determining an optimal pulse pattern using an evaluation function based on information including at least one of the harmonics of the battery direct current or the harmonics of the smoothing capacitor ripple current;
A control device for a rotating machine, comprising:   2. The rotating machine control device according to claim 1, wherein the multi-phase rotating machine is a three-phase rotating machine. Of the 6th multiple component of the electrical frequency of the rotating machine in the harmonic analysis of the battery direct current or the smoothing capacitor ripple current, the 6 appearing in the vicinity of the resonance frequency of the LC resonance generated between the battery and the smoothing capacitor. Let the amplitude of the second multiple component be the evaluation function,
The said pulse pattern determination part determines the pulse pattern which makes the said evaluation function the minimum as an optimal pulse pattern, The control apparatus of the rotary machine of Claim 2 characterized by the above-mentioned. The sum of the amplitudes of the sixth multiple components of the electrical frequency of the rotating machine in the harmonic analysis of the battery direct current is the evaluation function,
The said pulse pattern determination part determines the pulse pattern which makes the said evaluation function the minimum as an optimal pulse pattern, The control apparatus of the rotary machine of Claim 2 characterized by the above-mentioned. The sum of the amplitudes of the sixth multiple components of the electrical frequency of the rotating machine in the harmonic analysis of the smoothing capacitor ripple current is used as the evaluation function,
The said pulse pattern determination part determines the pulse pattern which makes the said evaluation function the minimum as an optimal pulse pattern, The control apparatus of the rotary machine of Claim 2 characterized by the above-mentioned. When the total sum of the amplitudes of the sixth multiple components of the electrical frequency of the rotating machine in the harmonic analysis of the battery direct current or the smoothing capacitor ripple current is less than a predetermined threshold,
The evaluation function is the sum of the amplitudes of the 6th multiple component or the component corresponding to the 6th multiple in the harmonic analysis of any one or more of the phase current, the inverter input current, or the pulse pattern,
The said pulse pattern determination part determines the pulse pattern which makes the said evaluation function the minimum as an optimal pulse pattern, The control apparatus of the rotary machine of Claim 2 characterized by the above-mentioned. A value obtained by combining two or more of the sum totals of amplitudes of sixth-order multiple components of the electrical frequency of the rotating machine in harmonic analysis of the battery direct current, the smoothing capacitor ripple current, and the inverter input current, As an evaluation function,
The said pulse pattern determination part determines a pulse pattern based on the said evaluation function, The control apparatus of the rotary machine of Claim 2 characterized by the above-mentioned.   8. The harmonic analysis is performed by Fourier series expansion for the pulse pattern, the phase current, the inverter input current, the battery DC current, or the smoothing capacitor ripple current. The control apparatus of the rotary machine as described in 2.   The said phase current estimation part estimates the said phase current using a voltage equation, The control apparatus of the rotary machine as described in any one of Claims 2-8 characterized by the above-mentioned.   10. The control device for a rotating machine according to claim 9, wherein a differential inductance (Ld ′, Lq ′) obtained by partial differentiation of the magnetic flux (λd, λq) with respect to the current (id, iq) is used in the voltage equation.   The controller for a rotating machine according to claim 9, wherein the induced voltage (e) used in the voltage equation is calculated as a sum of distortion components with respect to a rotational position.   The control apparatus for a rotating machine according to any one of claims 2 to 8, wherein the phase current estimation unit estimates the phase current using electromagnetic field analysis. The pulse pattern calculator is
About one-phase pulse pattern among the three phases,
A section of electrical angle 0-90 deg is set as one unit, and a pulse pattern of electrical angle 90-180 deg is set by reversing the pulse pattern of electrical angle 0-90 deg about the electrical angle 90 deg. A pulse pattern with an electrical angle of 180 to 360 deg is set by reversing the pulse pattern with an electrical angle of 0 to 180 deg around the 180 deg in a point symmetry,
For other two-phase pulse patterns,
The control device for a rotating machine according to any one of claims 2 to 12, wherein the one-phase pulse pattern is set by shifting the electrical angle by ± 120 deg.

Images