JP2017201859A - Control device for rotary electric machine - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a control device for a rotary electric machine capable of reducing a noise of the rotary electric machine caused by application of a high-frequency voltage for angle estimation.SOLUTION: A control device operates a power conversion circuit to apply high-frequency voltages to a plurality of coil groups. The control device estimates a rotational angle of a rotary electric machine on the basis of high-frequency currents flowing through the plurality of coil groups depending on the high-frequency voltages applied to the plurality of coil groups. The control device variably sets directions of high-frequency voltage vectors VVh1 and VVh2 to be applied to the plurality of coil groups on the basis of a drive state of the rotary electric machine. The control device applies the high-frequency voltages to the plurality of coil groups so as to satisfy a condition that the directions of the high-frequency voltage vectors VVh1 and VVh2 applied to the plurality of coil groups are set to the setting directions, and a condition that a magnitude of a resultant vector of the respective high-frequency voltage vectors VVh1 and VVh2 is set lower than magnitudes of the high-frequency voltage vectors VVh1 and VVh2 applied to the plurality of coil groups.SELECTED DRAWING: Figure 5

Description

  The present invention relates to a control device for a rotating electrical machine.

  Conventionally, there has been known a control device that estimates a rotation angle of a rotating electrical machine by applying a high-frequency voltage to a winding of the rotating electrical machine. For example, the control device described in Patent Document 1 below detects an inductance in each direction by applying a high-frequency voltage to each of a current vector direction of a rotating electrical machine having saliency and a direction orthogonal to the current vector. The rotation angle is estimated based on the inductance.

JP 2004-135425 A

  However, there is a concern that the noise of the rotating electrical machine increases as the high frequency voltage for angle estimation is applied to the winding.

  A main object of the present invention is to provide a control device for a rotating electrical machine that can reduce noise generated when a high-frequency voltage for angle estimation is applied.

  Hereinafter, means for solving the above-described problems and the operation and effects thereof will be described.

  The present invention relates to a multi-winding rotating electrical machine (10) having a plurality of winding groups (10A, 10B) wound around a stator (13), and a power conversion circuit for applying a voltage to the plurality of winding groups ( 20A, 20B), and a high frequency that operates the power conversion circuit to apply a high frequency voltage that fluctuates at an angular velocity higher than the electrical angular velocity of the rotating electrical machine to each of the plurality of winding groups. An operation unit (40) and an angle estimation unit (50) for estimating a rotation angle of the rotating electrical machine based on a high frequency current flowing in the plurality of winding groups according to a high frequency voltage applied to the plurality of winding groups. The high frequency operation unit includes a direction setting unit that variably sets the direction of the high frequency voltage vector applied to the plurality of winding groups based on the driving state of the rotating electrical machine, and each of the high frequency voltage vectors Direction The condition that the direction is set by the direction setting unit, and the condition that the magnitude of the combined vector of the high-frequency voltage vectors is smaller than the magnitude of each of the high-frequency voltage vectors applied to the plurality of winding groups. A high frequency voltage is applied to the plurality of winding groups so as to satisfy.

  A system to which the above invention is applied includes a multi-winding rotating electrical machine having a plurality of winding groups. In the said invention, the rotation angle of a rotary electric machine is estimated based on the high frequency current which flows into a some winding group according to the high frequency voltage applied to the some winding group. Here, in the above invention, the plurality of winding groups are arranged so as to satisfy the condition that the magnitude of the combined vector of each high-frequency voltage vector is smaller than the magnitude of each high-frequency voltage vector applied to the plurality of winding groups. A high frequency voltage is applied. For this reason, the noise of the rotary electric machine generated with the application of the high frequency voltage for angle estimation can be reduced.

  Further, in the above invention, the high frequency voltage is applied to the plurality of winding groups so as to satisfy the condition that the direction of the vector of each high frequency voltage is set to the direction set by the direction setting unit. Depending on the driving state, the influence of the applied high frequency voltage on the control of the rotating electrical machine is different. For this reason, according to the said invention, the influence which the high frequency voltage for angle estimation has on control of a rotary electric machine can be suppressed.

1 is an overall configuration diagram of a motor control system according to a first embodiment. The block diagram which shows a motor drive control process. The figure which shows the relationship of (alpha) (beta) coordinate system, dq coordinate system, and (gamma) (delta) coordinate system. The block diagram which shows the process of a high frequency voltage setting part. The figure which shows the superimposition aspect of the high frequency voltage in (gamma) delta coordinate system. The figure which shows a high frequency voltage waveform. The figure which shows the superimposition aspect of the high frequency voltage which concerns on the modification of 1st Embodiment. The block diagram which shows the process of the angle estimation part which concerns on 1st Embodiment. The figure which shows the high frequency voltage and high frequency current in an alpha beta coordinate system and an XY coordinate system. The figure which shows the relationship between the assumption variation | change_quantity of a high frequency current, and the magnitude | size of an electric current vector. The figure which shows the variation | change_quantity of the X-axis direction component of a high frequency current. The figure which shows the superimposition aspect of the high frequency voltage in the (gamma) (delta) coordinate system which concerns on 2nd Embodiment. The block diagram which shows the process of a high frequency voltage setting part. The figure which shows the relationship between the variation | change_quantity of the high frequency current based on 3rd Embodiment, the estimation error of a rotation angle, and the direction of a high frequency voltage vector. The figure which shows the superimposition aspect of the high frequency voltage in (gamma) delta coordinate system. The block diagram which shows the process of a high frequency voltage setting part. The flowchart which shows the process sequence of the high frequency voltage setting part which concerns on 4th Embodiment. The block diagram which shows the process of the high frequency voltage setting part which concerns on 5th Embodiment. The flowchart which shows the process sequence of a selection part. The figure which shows the superimposition aspect of the high frequency voltage which concerns on other embodiment. The figure which shows the superimposition aspect of the high frequency voltage which concerns on other embodiment. The figure which shows the superimposition aspect of the high frequency voltage which concerns on other embodiment.

(First embodiment)
Hereinafter, a first embodiment in which a control device according to the present invention is applied to a vehicle including an engine as an in-vehicle main machine will be described with reference to the drawings.

  As shown in FIG. 1, the motor 10 is a rotating electrical machine having a multiphase multiple winding, and in the present embodiment, is a synchronous machine having a three-phase double winding. In particular, in this embodiment, a non-salient pole machine is used as the motor 10. Incidentally, as the motor 10, for example, a permanent magnet field type or a wound field type can be employed.

  In the present embodiment, the motor 10 is an integrated starter generator (ISG) that integrates the functions of a starter and an alternator (generator). The rotor 12 constituting the motor 10 can transmit power to the crankshaft 14 a of the engine 14. In the present embodiment, the rotor 12 is mechanically connected to the crankshaft 14a via a belt, for example.

  In the present embodiment, in addition to starting the engine 14 for the first time, the engine 14 is automatically stopped when a predetermined automatic stop condition is satisfied, and then the engine 14 is automatically started when a predetermined restart condition is satisfied. The motor 10 also functions as a starter when executing the idling stop function for restarting.

  A first winding group 10A and a second winding group 10B, which are two armature winding groups, are wound around the stator 13 constituting the motor 10. The rotor 12 is made common to the first and second winding groups 10A and 10B. Each of the first and second winding groups 10A and 10B includes three-phase windings having different neutral points. The first winding group 10A has U, V, W-phase windings UA, VA, WA that are shifted from each other by 120 ° in electrical angle, and the second winding group 10B is shifted from each other by 120 ° in electrical angle. U, V, W phase windings UB, VB, WB. In the present embodiment, the angle formed by the first winding group 10A and the second winding group 10B is 0 ° in electrical angle. That is, the angle formed by the U-phase winding UA of the first winding group 10A and the U-phase winding UB of the second winding group 10B is set to 0 ° in electrical angle. In the present embodiment, the first winding group 10A and the second winding group 10B have the same configuration. Specifically, the number of turns of each of the U, V, W phase windings UA, VA, WA constituting the first winding group 10A and the U, V, W phase windings UB constituting the second winding group 10B. , VB, WB are set equal to the number of turns.

  Returning to FIG. 1, the motor 10 is electrically connected to the first and second inverters 20A and 20B corresponding to the first and second winding groups 10A and 10B. The first and second inverters 20A and 20B correspond to a power conversion circuit that converts an input DC voltage into an AC voltage and outputs the AC voltage. A battery 21 that is a common DC power source is connected in parallel to each of the first inverter 20A and the second inverter 20B. The output voltage of the battery 21 is 12V, for example. Note that a capacitor 22 is connected to the battery 21 in parallel.

  The first inverter 20A includes a series connection body of first U, V, W phase upper arm switches SUp1, SVp1, SWp1 and first U, V, W phase lower arm switches SUn1, SVn1, SWn1. U, V, and W phase windings UA, VA, and WA constituting the first winding group 10A are connected to connection points of the series connection bodies in the U, V, and W phases. In the present embodiment, IGBTs are used as the switches SUp1 to SWn1. The diodes DUp1, DUn1, DVp1, DVn1, DWp1, DWn1 are connected in antiparallel to the switches SUp1, SUn1, SVp1, SVn1, SWp1, SWn1. The switches SUp1 to SWn1 are not limited to IGBTs, but may be N-channel MOSFETs, for example.

  Similarly to the first inverter 20A, the second inverter 20B is connected in series with the second U, V, W phase upper arm switches SUp2, SVp2, SWp2 and the second U, V, W phase lower arm switches SUn2, SVn2, SWn2. Has a body. U, V, and W phase windings UB, VB, and WB constituting the second winding group 10B are connected to connection points of the series connection bodies in the U, V, and W phases. In the present embodiment, IGBTs are used as the switches SUp2 to SWn2. The diodes DUp2, DUn2, DVp2, DVn2, DWp2, and DWn2 are connected in antiparallel to the switches SUp2, SUn2, SVp2, SVn2, SWp2, and SWn2. Note that the switches SUp2 to SWn2 are not limited to IGBTs, and may be N-channel MOSFETs, for example.

  The positive terminal of the battery 21 is connected to the collectors of the upper arm switches of the first and second inverters 20A and 20B. The negative terminal of the battery 21 is connected to the emitters of the lower arm switches of the first and second inverters 20A and 20B. That is, in this embodiment, the battery 21 is shared by the inverters 20A and 20B.

  The control system according to the present embodiment includes a voltage detection unit 30, a first phase current detection unit 31A, and a second phase current detection unit 31B. The voltage detector 30 detects the voltage applied from the battery 21 to the first and second inverters 20A and 20B as the power supply voltage VDC. The first phase current detector 31A detects a current for at least two phases of the three phase currents flowing through the first winding group 10A. The second phase current detector 31B detects a current for at least two phases of the three phase currents flowing through the second winding group 10B. In the present embodiment, the first and second phase current detectors 31A and 31B detect V-phase current and W-phase current. In addition, as 1st, 2nd phase electric current detection part 31A, 31B, what is provided with a current transformer or a resistor can be used, for example.

  The detection values of the various detection units are taken into the control device 40 mainly composed of a microcomputer. The control device 40 includes a CPU and a memory, and executes a program stored in the memory by the CPU. The control device 40 generates an operation signal for turning on / off each switch of the first inverter 20A and the second inverter 20B based on the detection values of these various sensors in order to control the control amount of the motor 10 to the command value. . In the present embodiment, the control amount is a torque, and the command value of the control amount is a command torque Trq *. In FIG. 1, signals for operating the switches SUp1, SUn1, SVp1, SVn1, SWp1, SWn1 of the first inverter 20A are shown as first operation signals gUp1, gUn1, gVp1, gVn1, gWp1, gWn1, and the second inverter 20B. Signals for operating the switches SUp2, SUn2, SVp2, SVn2, SWp2, and SWn2 are shown as second operation signals gUp2, gUn2, gVp2, gVn2, gWp2, and gWn2.

  Next, torque control of the motor 10 will be described with reference to FIG. This control is position sensorless control that is a control that does not use an angle detector such as a resolver that directly detects the rotation angle that is the magnetic pole position of the motor 10. The control device 40 performs torque control based on the estimated rotation angle.

  The control device 40 includes a first processing unit 41 corresponding to the first inverter 20A. In the first processing unit 41, the first current conversion unit 41a includes an estimated angle θγ that is an electrical angle of the motor 10 estimated by the angle estimation unit 50 described later, and the V phase detected by the first phase current detection unit 31A. Based on the current IV1 and the W-phase current IW1, the U, V, and W-phase currents of the first winding group 10A in the UVW coordinate system are converted into the first γ-axis current Iγ1r and the first δ-axis current Iδ1r in the γδ coordinate system. . Here, the UVW coordinate system is a three-phase fixed coordinate system of the motor 10, and the γδ coordinate system is an estimated coordinate system of a dq coordinate system that is a two-phase rotational coordinate system of the motor 10. FIG. 3 shows a γδ coordinate system, a dq axis coordinate system, and an αβ coordinate system which is a two-phase fixed coordinate system. In FIG. 3, the angle formed between the α axis of the αβ coordinate system and the γ axis of the γδ coordinate system is shown as an estimated angle θγ, and the angle formed between the α axis and the d axis of the dq coordinate system is shown as an actual electrical angle θe. The angle formed between the d-axis and the γ-axis is indicated as an estimation error Δθ. The dq coordinate system is a coordinate system that rotates at the electrical angular velocity of the motor 10 with respect to the αβ coordinate system.

  Returning to the description of FIG. 2, the first command current setting unit 41b sets the first γ-axis command current Iγ1 * and the first δ-axis command current Iδ1 * based on the command torque Trq *. The first γ-axis deviation calculating unit 41c calculates the first γ-axis deviation ΔIγ1 as a value obtained by subtracting the first γ-axis current Iγ1r from the first γ-axis command current Iγ1 *. The first δ-axis deviation calculating unit 41d calculates the first δ-axis deviation ΔIδ1 as a value obtained by subtracting the first δ-axis current Iδ1r from the first δ-axis command current Iδ1 *.

  Based on the first γ-axis deviation ΔIγ1, the first command voltage setting unit 41e calculates the first γ-axis voltage Vγ1r as an operation amount for performing feedback control of the first γ-axis current Iγ1r to the first γ-axis command current Iγ1 *. Further, the first command voltage setting unit 41e calculates the first δ-axis voltage Vδ1r as an operation amount for performing feedback control of the first δ-axis current Iδ1r to the first δ-axis command current Iδ1 * based on the first δ-axis deviation ΔIδ1. To do. For example, proportional-integral control can be used as the feedback control.

  The first γ-axis superimposing unit 41f outputs an addition value of the first γ-axis voltage Vγ1r and a first γ-axis high-frequency voltage Vγ1h set by a high-frequency voltage setting unit 60 described later as a first γ-axis command voltage Vγ1 *. The first δ-axis superimposing unit 41g outputs an addition value of the first δ-axis voltage Vδ1r and the first δ-axis high-frequency voltage Vδ1h set by the high-frequency voltage setting unit 60 as the first δ-axis command voltage Vδ1 *. The first γ-axis high-frequency voltage Vγ1h and the first δ-axis high-frequency voltage Vδ1h are signals that fluctuate at an angular velocity sufficiently higher than the electrical angular velocity of the fundamental component of the first γ-axis command voltage Vγ1 * and the first δ-axis command voltage Vδ1 *.

  Based on the first γ-axis command voltage Vγ1 *, the first δ-axis command voltage Vδ1 *, the power supply voltage VDC detected by the voltage detection unit 30, and the estimated angle θγ, the first voltage conversion unit 41h , Δ-axis command voltages Vγ1 *, Vδ1 * are converted into first U, V, W-phase command voltages VU1, VV1, VW1 in the UVW coordinate system. In the present embodiment, the first U, V, and W phase command voltages VU1, VV1, and VW1 have waveforms that are 120 degrees out of phase with each other in electrical angle.

  The first generator 41i generates the first operation signals gUp1 to gWn1 based on the first U, V, and W phase command voltages VU1, VV1, and VW1 output from the first voltage converter 41h. The first generation unit 41i outputs the generated first operation signals gUp1 to gWn1 to the switches SUp1 to SWn1. Here, the first operation signal may be generated by, for example, PWM control based on a magnitude comparison between a carrier signal such as a triangular wave signal and the phase command voltages VU1, VV1, and VW1. The first operation signal on the upper arm side and the corresponding first operation signal on the lower arm side are complementary to each other. For this reason, the upper arm switch and the corresponding lower arm switch are alternately turned on.

  The control device 40 includes a second processing unit 42 corresponding to the second inverter 20B. Since the second processing unit 42 has the same configuration as the first processing unit 41, detailed description thereof is omitted as appropriate.

  In the second processing unit 42, the second current conversion unit 42a includes the second winding group based on the estimated angle θγ and the V-phase current IV2 and the W-phase current IW2 detected by the second-phase current detection unit 31B. The U, V, and W phase currents corresponding to 10B are converted into a second γ-axis current Iγ2r and a second δ-axis current Iδ2r in the γδ coordinate system.

  The second command current setting unit 42b sets the second γ-axis command current Iγ2 * and the second δ-axis command current Iδ2 * based on the command torque Trq *. The command currents Iγ2 *, Iδ2 *, Iγ1 *, Iδ1 * are set to values necessary for setting the torque of the motor 10 to the command torque Trq *. The second γ-axis command current Iγ2 * may be set to the same value as the first γ-axis command current Iγ1 *, or may be set to a different value. Further, the second δ-axis command current Iδ2 * may be set to the same value as the first δ-axis command current Iδ1 *, or may be set to a different value.

  The second γ-axis deviation calculating unit 42c calculates the second γ-axis deviation ΔIγ2 as a value obtained by subtracting the second γ-axis current Iγ2r from the second γ-axis command current Iγ2 *. The second δ-axis deviation calculating unit 42d calculates the second δ-axis deviation ΔIδ2 as a value obtained by subtracting the second δ-axis current Iδ2r from the second δ-axis command current Iδ2 *.

  The second command voltage setting unit 42e calculates a second γ-axis voltage Vγ2r based on the second γ-axis deviation ΔIγ2. The second command voltage setting unit 42e calculates the second δ-axis voltage Vδ2r based on the second δ-axis deviation ΔIδ2. For example, proportional-integral control can be used as feedback control used by the second command voltage setting unit 42e.

  The second γ-axis superimposing unit 42f outputs the added value of the second γ-axis voltage Vγ2r and the second γ-axis high-frequency voltage Vγ2h set by the high-frequency voltage setting unit 60 as the second γ-axis command voltage Vγ2 *. The second δ-axis superimposing unit 42g outputs the addition value of the second δ-axis voltage Vδ2r and the second δ-axis high-frequency voltage Vδ2h set by the high-frequency voltage setting unit 60 as the second δ-axis command voltage Vδ2 *. The second γ-axis high-frequency voltage Vγ2h and the second δ-axis high-frequency voltage Vδ2h are signals that fluctuate at an angular velocity sufficiently higher than the electrical angular velocity of the fundamental wave component of the second γ-axis command voltage Vγ2 * and the second δ-axis command voltage Vδ2 *.

  Based on the second γ-axis command voltage Vγ2 *, the second δ-axis command voltage Vδ2 *, the power supply voltage VDC, and the estimated angle θγ, the second voltage conversion unit 42h generates the second γ, δ-axis command voltage Vγ2 *, Vδ2 * is converted into second U, V, and W phase command voltages VU2, VV2, and VW2 in the UVW coordinate system.

  The second generation unit 42i generates second operation signals gUp2 to gWn2 based on the second U, V, and W phase command voltages VU2, VV2, and VW2 output from the second voltage conversion unit 42h. The second generation unit 42i outputs the generated second operation signals gUp2 to gWn2 to the switches SUp2 to SWn2.

  Next, the high frequency voltage setting unit 60 will be described.

  The high frequency voltage setting unit 60 is provided to generate the respective high frequency voltages Vγ1h, Vδ1h, Vγ2h, and Vδ2h necessary for calculating the estimated angle θγ during torque control of the motor 10. Here, the motor to which the angle estimation using high-frequency voltage superposition is applied is usually a salient pole machine. However, the motor 10 of this embodiment is a non-salient pole machine. Even in this case, for example, a large current flows through the motor 10 to start the engine 14, and magnetic saturation occurs in the motor 10. When magnetic saturation occurs, the d-axis inductance and the q-axis inductance of the motor 10 are different. For this reason, even when the motor 10 is a non-salient pole machine, the control device 40 including the high-frequency operation unit and the direction setting unit can perform angle estimation using high-frequency voltage superposition.

  In the high-frequency voltage setting unit 60, the signal generation unit 60a sets the estimated angle θγ, the first γ-axis command current Iγ1 *, the first δ-axis command current Iδ1 *, the second γ-axis command current Iγ2 *, and the second δ-axis command current Iδ2 *. Based on this, the first γ-axis high-frequency voltage Vγ1h and the first δ-axis high-frequency voltage Vδ1h are generated. Hereinafter, a method for generating the first γ-axis high-frequency voltage Vγ1h and the first δ-axis high-frequency voltage Vδ1h according to the present embodiment will be described with reference to FIG.

  FIG. 5 shows an equal torque curve Lc in the γδ coordinate system. The equal torque curve Lc is a curve defined by γ and δ axis currents for realizing the command torque Trq *. FIG. 5 shows a tangent line Lta of the equal torque curve Lc passing through the intersection point P1 between the equal torque curve Lc and the drive current vector VIr flowing through the motor 10. The drive current vector VIr is a drive current vector defined by the first γ-axis command current Iγ1 * and the first δ-axis command current Iδ1 *, and a drive defined by the second γ-axis command current Iγ2 * and the second δ-axis command current Iδ2 *. It is a combined vector with a current vector. In the present embodiment, the direction parallel to the tangent line Lta is defined as the torque ripple reduction direction. A first high-frequency voltage vector VVh1 applied to the first winding group 10A is set as a voltage vector extending from the origin O in the torque ripple reduction direction. A first γ-axis high-frequency voltage Vγ1h is generated based on the γ-axis direction component of the first high-frequency voltage vector VVh1, and a first δ-axis high-frequency voltage Vδ1h is generated based on the δ-axis direction component of the first high-frequency voltage vector VVh1. . In the present embodiment, each of the high-frequency voltages Vγ1h and Vδ1h is a rectangular wave pulse signal having the same period. As the magnitude of the first high-frequency voltage vector VVh1 is larger, the amplitudes of the high-frequency voltages Vγ1h and Vδ1h are set larger. In FIG. 6A, the amplitude of the first γ-axis high-frequency voltage Vγ1h is indicated by Va.

  Returning to the description of FIG. 4, the high-frequency voltage setting unit 60 includes a ripple reduction storage unit 60b. The ripple reduction storage unit 60b stores the torque ripple reduction direction associated with the magnitude Iamp and the phase θI of the drive current vector VIr as map information. The ripple reduction storage unit 60b is configured by a memory. In the present embodiment, the phase θI of the drive current vector VIr is defined in the counterclockwise direction with the γ axis as a reference, as shown in FIG. The map information is created by adapting in advance by experiments, calculations, and the like.

  The signal generator 60a generates a magnitude Iamp of the drive current vector VIr based on the first γ-axis command current Iγ1 *, the first δ-axis command current Iδ1 *, the second γ-axis command current Iγ2 *, and the second δ-axis command current Iδ2 *. And the phase θI are calculated. The signal generation unit 60a selects a direction corresponding to the magnitude Iamp and the phase θI of the calculated drive current vector VIr as the command direction θX from the plurality of torque ripple reduction directions stored in the ripple reduction storage unit 60b. As shown in FIG. 5, the signal generation unit 60a includes the amplitude of the first γ-axis high-frequency voltage Vγ1h and the amplitude of the first δ-axis high-frequency voltage Vδ1h required to set the direction of the first high-frequency voltage vector VVh1 as the command direction θX. Set the ratio. The signal generator 60a outputs the high-frequency voltages Vγ1h and Vδ1h according to the input estimated angle θγ. The rising timings of the high-frequency voltages Vγ1h and Vδ1h may be synchronized or may not be synchronized.

  The first γ-axis high-frequency voltage Vγ1h output from the signal generator 60a is input to the first sign inversion unit 60c. As shown in FIG. 6, the first code inverting unit 60 c generates the second γ-axis high-frequency voltage Vγ2 h by inverting the sign of the input first γ-axis high-frequency voltage Vγ1 h. The first δ-axis high frequency voltage Vδ1h output from the signal generation unit 60a is input to the second sign inversion unit 60d. The second sign inverting unit 60d generates the second δ-axis high-frequency voltage Vδ2h by inverting the sign of the input first δ-axis high-frequency voltage Vδ1h. As shown in FIG. 5, the second winding as a voltage vector extending from the origin O where the angle between the first γ-axis high-frequency voltage Vγ2h and the second δ-axis high-frequency voltage Vδ2h and the first high-frequency voltage vector VVh1 is 180 °. A second high-frequency voltage vector VVh2 applied to the group 10B is realized. In the present embodiment, the magnitude of the second high-frequency voltage vector VVh2 is equal to the magnitude of the first high-frequency voltage vector VVh1. The first and second high-frequency voltage vectors VVh1 and VVh2 are superimposed on the drive voltage vector VVr. The drive voltage vector VVr is a combined vector of the drive voltage vector defined by the first γ-axis voltage Vγ1r and the first δ-axis voltage Vδ1r and the drive voltage vector defined by the second γ-axis voltage Vγ2r and the second δ-axis voltage Vδ2r. It is.

  Even when the first and second high-frequency voltage vectors VVh1 and VVh2 described above for angle estimation are applied to the first and second winding groups 10A and 10B, noise generated with the application of the high-frequency voltage can be reduced. . That is, with the application of the high frequency voltage, the drive current vector VIr in the γδ coordinate system shown in FIG. 5 temporarily changes. In this case, if the deviation between the tip of the drive current vector VIr and the equal torque curve Lc is large, the torque fluctuation of the motor 10 becomes large, and the noise of the motor 10 increases. Here, when the torque ripple reduction direction, which is parallel to the direction of the tangent Lta of the equal torque curve Lc, is set to the command direction θX, the tip of the drive current vector VIr is equal to the equal torque curve Lc even if a high frequency voltage is applied. It moves along the tangent line Lta. For this reason, it is possible to suppress the deviation between the tip of the drive current vector VIr and the equal torque curve Lc. Thereby, the torque fluctuation of the motor 10 generated with the application of the high-frequency voltage can be reduced, and consequently the noise of the motor 10 caused by the torque fluctuation can be reduced.

  In FIG. 5, the first high-frequency voltage vector VVh1 is applied to the first winding group 10A and the second high-frequency voltage vector VVh2 is applied to the second winding group 10B. The second high-frequency current vector VIh2 is shown. FIG. 5 shows an example in which the phase shift between the high-frequency voltage vectors VVh1 and VVh2, the high-frequency current vectors, and VIh1 and VIh2 is substantially zero.

  Incidentally, the control device 40 may include a basic operation unit for operating each of the inverters 20A and 20B in order to control the torque of the motor 10 to the command torque Trq * by maximum torque per ampere control. . In this case, in the first command current setting unit 41b and the second command current setting unit 42b shown in FIG. 2, the command currents Iγ1 *, Iδ1 *, Iγ2 *, Iδ2 * are set by the minimum current / maximum torque control. The In this configuration, as shown in FIG. 7, the torque ripple reduction direction can be set to a direction orthogonal to the current drive current vector VIr. In this case, the torque ripple reduction direction can be easily determined.

  Next, the angle estimation unit 50 will be described with reference to FIG.

  In the angle estimation unit 50, the high frequency extraction unit 50a corresponds to the first γ-axis high-frequency voltage Vγ1h and the first δ-axis high-frequency voltage Vδ1h from the V-phase current IV1 and the W-phase current IW1 detected by the first phase current detection unit 31A. The first V-phase high-frequency current IVh1 and the first W-phase high-frequency current IWh1 that are high-frequency current components are extracted. The high frequency extraction unit 50a also uses a high frequency current component corresponding to the second γ-axis high-frequency voltage Vγ2h and the second δ-axis high-frequency voltage Vδ2h from the V-phase current IV2 and the W-phase current IW2 detected by the second phase current detection unit 31B. A certain second V-phase high-frequency current IVh2 and second W-phase high-frequency current IWh2 are extracted. Note that the high-frequency extraction unit 50a may be configured with, for example, a band-pass filter or a high-pass filter.

  The two-phase conversion unit 50b converts the first V-phase high-frequency current IVh1 and the first W-phase high-frequency current IWh1 extracted by the high-frequency extraction unit 50a into a first α-axis high-frequency current Ihα1 and a first β-axis high-frequency current Ihβ1 in the αβ coordinate system. . The two-phase conversion unit 50b also converts the second V-phase high-frequency current IVh2 and the second W-phase high-frequency current IWh2 extracted by the high-frequency extraction unit 50a into the second α-axis high-frequency current Ihα2 and the second β-axis high-frequency current Ihβ2 in the αβ coordinate system. Convert.

  The inner product calculation unit 50c receives the first α-axis high-frequency current Ihα1 and the first β-axis high-frequency current Ihβ1, and calculates a first inner product value K1 based on the following equation (eq1).

上式（ｅｑ１）において、Ｖｈα１は、図９に示すように、第１高周波電圧ベクトルＶＶｈ１のα軸方向成分である第１α軸高周波電圧を示し、Ｖｈβ１は、第１高周波電圧ベクトルＶＶｈ１のβ軸方向成分である第１β軸高周波電圧を示す。第１α軸高周波電圧Ｖｈα１及び第１β軸高周波電圧Ｖｈβ１は、第１γ軸高周波電圧Ｖγ１ｈ及び第１δ軸高周波電圧Ｖδ１ｈに基づいて設定される。

In the above equation (eq1), as shown in FIG. 9, Vhα1 represents a first α-axis high-frequency voltage that is an α-axis direction component of the first high-frequency voltage vector VVh1, and Vhβ1 represents the β-axis of the first high-frequency voltage vector VVh1. The 1st beta axis high frequency voltage which is a direction component is shown. The first α-axis high-frequency voltage Vhα1 and the first β-axis high-frequency voltage Vhβ1 are set based on the first γ-axis high-frequency voltage Vγ1h and the first δ-axis high-frequency voltage Vδ1h.

  The first inner product value K1 represented by the above equation (eq1) indicates the inner product value of the first high-frequency voltage vector VVh1 and the first high-frequency current vector VIh1. That is, the inner product value of the first high-frequency voltage vector VVh1 and the first high-frequency current vector VIh1 is expressed by the following equation (eq2), and the right side of the lower equation (eq2) matches the right side of the upper equation (eq1). .

内積算出部５０ｃは、第２α軸高周波電流Ｉｈα２，第２β軸高周波電流Ｉｈβ２を入力として、下式（ｅｑ３）に基づいて第２内積値Ｋ２を算出する。

The inner product calculation unit 50c receives the second α-axis high-frequency current Ihα2 and the second β-axis high-frequency current Ihβ2, and calculates the second inner product value K2 based on the following equation (eq3).

上式（ｅｑ２）において、Ｖｈα２は、第２高周波電圧ベクトルＶＶｈ２のα軸方向成分である第２α軸高周波電圧を示し、Ｖｈβ２は、第２高周波電圧ベクトルＶＶｈ２のβ軸方向成分である第２β軸高周波電圧を示す。第２α軸高周波電圧Ｖｈα２及び第２β軸高周波電圧Ｖｈβ２は、第２γ軸高周波電圧Ｖγ２ｈ及び第２δ軸高周波電圧Ｖδ２ｈに基づいて設定される。

In the above equation (eq2), Vhα2 represents the second α-axis high-frequency voltage that is the α-axis direction component of the second high-frequency voltage vector VVh2, and Vhβ2 is the second β-axis that is the β-axis direction component of the second high-frequency voltage vector VVh2. Indicates high frequency voltage. The second α-axis high-frequency voltage Vhα2 and the second β-axis high-frequency voltage Vhβ2 are set based on the second γ-axis high-frequency voltage Vγ2h and the second δ-axis high-frequency voltage Vδ2h.

  The reason why the currents in the α and β axis directions are used for calculating the inner product value is to reduce the calculation load of the control device 40 and to increase the calculation accuracy of the estimated angle θγ. That is, since the electrical angle information is not required for conversion from the UVW coordinate system to the αβ coordinate system, the calculation load of the control device 40 can be reduced. Further, since the estimated angle θγ as the electrical angle information is not used for coordinate conversion, the influence of the estimation error Δθ included in the estimated angle θγ on the calculation accuracy of the estimated angle θγ can be eliminated.

  The inner product calculation unit 50c calculates a total inner product value Kr by adding the first inner product value K1 and the second inner product value K2. The inner product deviation calculating unit 50d calculates the inner product deviation ΔK by subtracting the total inner product value Kr calculated by the inner product calculating unit 50c from the target inner product value Ktgt. In the present embodiment, the inner product deviation ΔK corresponds to the actual deviation.

  The target inner product value Ktgt is based on the estimated angular velocity ωγ, which is an electrical angular velocity estimated by the velocity estimation unit 50f described later, the high-frequency voltages Vγ1h, Vδ1h, Vγ2h, Vδ2h, and the currents Iγ1r, Iδ1r, Iγ2r, Iδ2r. It is set by the value setting unit 50e. The target inner product value Ktgt is determined as an addition value of a first reference value K1tgt that is an assumed value that the first inner product value K1 can take and a second reference value K2tgt that is an assumed value that the second inner product value K2 can take. The map information is stored in the control device 40 as map information preliminarily adapted by experiments and calculations. Since each of the reference values K1tgt and K2tgt depends on the electric angular velocity, the torque ripple reduction direction, and the magnitude Iamp of the drive current vector VIr, the target inner product value Ktgt also depends on these parameters. Therefore, the input parameters used for setting the target dot product value Ktgt include the high frequency voltages Vγ1h, Vδ1h, Vγ2h, Vδ2h for grasping the direction of torque ripple reduction, and the individual parameters for grasping the magnitude Iamp of the drive current vector VIr. Currents Iγ1r, Iδ1r, Iγ2r, and Iδ2r are included. FIG. 10 shows that the first reference value K1tgt depends on the magnitude Iamp of the drive current vector VIr.

  The speed estimation unit 50f calculates an estimated angular speed ωγ as an operation amount for performing feedback control of the inner product deviation ΔK to zero. In this embodiment, proportional-integral control is used as feedback control in the speed estimation unit 50f.

  The integrator 50g calculates the estimated angle θγ as a time integral value of the estimated angular velocity ωγ calculated by the speed estimation unit 50f.

  In the present embodiment, the inner product values represented by the above equations (eq1) and (eq3) are calculated, but instead, the X-axis direction components of the high-frequency current vectors VIh1 and VIh2 can be calculated. Here, the X-axis direction is a direction in which the high-frequency current vectors VIh1 and VIh2 extend from the origin O as shown in FIG. FIG. 9 also shows the Y axis orthogonal to the X axis. In FIG. 9, Ihx and Ihy indicate a first X-axis high-frequency current and a first Y-axis high-frequency current that are components in the X- and Y-axis directions of the first high-frequency current vector VIh1.

  A current calculation unit instead of the inner product calculation unit 50c calculates the first X-axis high-frequency current Ihx1 based on the first α-axis high-frequency current Ihα1, the first β-axis high-frequency current Ihβ1, and the command direction θX. Further, based on the second α-axis high-frequency current Ihα2, the second β-axis high-frequency current Ihβ2, and the command direction θX, the second X-axis high-frequency current Ihx2 that is the X-axis direction component of the second high-frequency current vector VIh2 is calculated. The current calculation unit calculates the total current value by adding the first X-axis high-frequency current Ihx1 and the second X-axis high-frequency current Ihx2.

  A current deviation calculation unit instead of the inner product deviation calculation unit 50d calculates a current deviation by subtracting the total current value from the target current value. The target current value is determined as an addition value of a first current reference value that is an assumed value that can be taken by the first X-axis high-frequency current Ihx1 and a second current reference value that is an assumed value that can be taken by the second X-axis high-frequency current Ihx2. ing. FIG. 11 shows the transition of the first X-axis high-frequency current Ihx1 that flows when the first X-axis high-frequency voltage Vhx1 that is the X-axis direction component of the first high-frequency voltage vector VVh1 is applied. The first current reference value corresponds to the amplitude of the waveform indicated by the broken line in FIG. The speed estimation unit 50f calculates an estimated angular speed ωγ as an operation amount for performing feedback control of the current deviation to zero.

  According to the embodiment described in detail above, the following effects can be obtained.

  The high-frequency voltages Vγ1h, Vδ1h, Vγ2h, and Vδ2h were applied to the winding groups 10A and 10B so that the first high-frequency voltage vector VVh1 and the second high-frequency voltage vector VVh2 cancel each other. For this reason, the noise generated by superimposing the high frequency voltage for estimating the angle of the motor 10 can be reduced.

  The directions of the first high-frequency voltage vector VVh1 and the second high-frequency voltage vector VVh2 were set in a direction parallel to the tangential direction of the equal torque curve Lc passing through the intersection point P1 between the equal torque curve Lc and the drive current vector VIr. For this reason, the torque fluctuation of the motor 10 accompanying application of a high frequency voltage can be reduced, and the noise of the motor 10 can be further reduced.

(Second Embodiment)
Hereinafter, the second embodiment will be described with reference to the drawings with a focus on differences from the first embodiment. In this embodiment, the setting method of the direction of the 1st, 2nd high frequency voltage vector VVh1, VVh2 superimposed on 1st, 2nd winding group 10A, 10B is changed.

  FIG. 12 shows a constant voltage circle Ca in the γδ coordinate system. The constant voltage circle Ca is a circle defined by γ and δ axis currents for realizing a predetermined voltage value. The predetermined voltage value is the magnitude of the current drive voltage vector VVr. FIG. 12 shows a tangent line Ltb of the constant voltage circle Ca passing through the intersection P2 between the constant voltage circle Ca and the drive voltage vector VVr.

  In the present embodiment, the direction parallel to the tangent line Ltb is defined as the voltage reduction direction. A first high-frequency voltage vector VVh1 is set as a voltage vector extending from the origin O in the voltage reduction direction. According to this setting method, fluctuations in the output voltage of the inverters 20A and 20B due to application of the high frequency voltage can be reduced. That is, if the tip of the voltage vector extending from the origin O of the γδ coordinate system has a large degree of deviation from the constant voltage circle Ca with the application of the high-frequency voltage, the fluctuation of the output voltage of each inverter 20A, 20B increases, and the output voltage May exceed the allowable upper limit. Here, the direction of the high-frequency voltage vectors VVh1 and VVh2 is set to the voltage fluctuation reduction direction, so that even if a high-frequency voltage is applied, the tip of the voltage vector moves along the tangent line Ltb of the constant voltage circle Ca. For this reason, the shift | offset | difference of the front-end | tip of the drive voltage vector VVr and the constant voltage circle Ca can be suppressed. Thereby, the fluctuation | variation of the output voltage of inverter 20A, 20B accompanying the application of a high frequency voltage can be reduced, and the fall of the reliability of a motor control system can be avoided by extension. Note that the setting method described above is an effective method when, for example, the motor 10 generates maximum torque, the output voltages of the inverters 20A and 20B do not have a sufficient margin with respect to the allowable upper limit value.

  FIG. 13 shows a block diagram of processing of the high-frequency voltage setting unit 60 according to the present embodiment. In FIG. 13, the same processes as those shown in FIG. 4 are given the same reference numerals for the sake of convenience.

  In the high-frequency voltage setting unit 60, the voltage reduction storage unit 60f stores, as map information, the voltage fluctuation reduction direction associated with the magnitude Iamp, phase θI, and estimated angular velocity ωγ of the drive current vector VIr. The signal generation unit 60e designates a direction corresponding to the magnitude Iamp and phase θI of the calculated drive current vector VIr and the estimated angular velocity ωγ from the plurality of voltage reduction directions stored in the voltage reduction storage unit 60f as a command direction. Select as θX. The reason why the magnitude Iamp and the phase θI of the drive current vector VIr are used is that information on the drive voltage vector VVr can be grasped from these pieces of information. However, in order to grasp the information of the drive voltage vector VVr, the first γ-axis voltage Vγ1r, the first δ-axis voltage Vδ1r, the second γ-axis voltage Vγ2r, and the second δ-axis voltage Vδ2r may be used.

  The signal generator 60e sets the amplitude of each high-frequency voltage Vγ1h, Vδ1h required to set the direction of the first high-frequency voltage vector VVh1 as the command direction θX, and according to the input estimated angle θγ, each high-frequency voltage Vγ1h, Vδ1h is output.

  In the present embodiment, the signal generation unit 60e sets the command direction θX in a direction orthogonal to the drive voltage vector VVr when it is determined that the estimated angular velocity ωγ is equal to or lower than the low speed determination value ωthL. This is because when the electrical angular velocity is 0 or the electrical angular velocity is low, the voltage drop due to the resistance component is dominant, and the constant voltage circle Ca is close to a perfect circle centered on the origin O. On the other hand, when the electrical angular velocity becomes high, the constant voltage circle Ca becomes an ellipse.

  Incidentally, the signal generator 60e may set the direction orthogonal to the drive current vector VIr as the voltage fluctuation reduction direction when it is determined that the estimated angular velocity ωγ is equal to or lower than the low speed determination value ωthL. This setting is based on the fact that the phase difference between the drive voltage vector VVr and the drive current vector VIr becomes small.

  As described above, in the present embodiment, the command direction θX is set to the voltage fluctuation reduction direction. For this reason, even if the output voltages of the inverters 20A and 20B have no margin with respect to the allowable upper limit value, the high-frequency voltage for angle estimation can be applied. Therefore, it is possible to prevent the calculation opportunity of the estimated angle θγ from being restricted, and the rotation speed range in which the motor 10 can be driven can be expanded.

(Third embodiment)
Hereinafter, the third embodiment will be described with reference to the drawings with a focus on differences from the first embodiment. In the present embodiment, the method for setting the direction of the first and second high-frequency voltage vectors VVh1 and VVh2 is changed. Hereinafter, this setting method will be described.

  In the present embodiment, as shown in FIG. 11B, the first X-axis high-frequency current Ihx1 that flows through the first winding group 10A according to the first high-frequency voltage vector VVh1 applied to the first winding group 10A. And the first current reference value is defined as a first current change amount Δamp1.

  FIG. 14 shows the relationship between the estimation error Δθ and the first current change amount Δamp1 when the direction of the first high-frequency voltage vector VVh1 is 0, 30, 60, 90, 120, 150 °. As shown in FIG. 14, when the estimation error Δθ is 0, the first X-axis high-frequency current Ihx1 matches the first current reference value, so the first current change amount Δamp1 becomes 0. In the present embodiment, when the direction of the first high-frequency voltage vector VVh1 is set to a plurality of different directions in a state where the estimation error Δθ is a value other than 0, the first corresponding to each of the plurality of directions. Of the current change amount Δamp1, the direction corresponding to the first current change amount other than the first current change amount having the minimum value is defined as the sensitivity increasing direction. The sensitivity increasing direction is set to the command direction θX. The direction in which the sensitivity increases varies depending on the driving state of the motor 10.

  In particular, in this embodiment, when the estimation error Δθ is 0 °, the direction in which the slope of the first current change amount Δamp1 when the estimation error Δθ is 0 is set as the sensitivity increasing direction. In the example shown in FIG. 14, 150 ° is set as the sensitivity increasing direction.

  According to the setting method described above, the deviation between the total inner product value Kr and the target value Ktgt is increased when the first current change amount Δamp1 with respect to the high-frequency voltage that becomes a disturbance increases and the estimation error Δθ is a value other than zero. It becomes easy to grasp. As a result, it is possible to improve the responsiveness when feedback control of the total inner product value Kr to the target value Ktgt is achieved, and as a result, the calculation accuracy of the estimated angle θγ can be increased. FIG. 15 shows the first high-frequency voltage vector VVh1 when the command direction θX is set to 150 ° and the second high-frequency voltage vector VVh2 whose angle formed by the first high-frequency voltage vector VVh1 is 180 °. . Further, CI shown in FIG. 15 is an ellipse showing the locus of each high-frequency current vector VIh1, VIh2 when the direction of each high-frequency voltage vector VVh1, VVh2 is set in various directions.

  Incidentally, as a method for setting the sensitivity increasing direction, for example, a method described below can be adopted. Specifically, for example, a specific error is defined as the first error in the range of the estimation error Δθ of 0 to + 90 °. Further, for example, an estimation error having the same absolute value as the first error is defined as the second error in the range of the estimation error Δθ of 0 to −90 °.

  Of the characteristic curves showing the relationship between the estimation error Δθ and the first current change amount Δamp1 corresponding to each of the plurality of directions shown in FIG. 14, the curve having the maximum absolute value of the first current change amount Δamp1 in the first error is shown. Selected as the first characteristic curve. When the first error is set to + 60 °, for example, a characteristic curve of 150 ° is selected as the first characteristic curve. Further, among the characteristic curves, a curve having the maximum absolute value of the first current change amount Δamp1 in the second error is selected as the second characteristic curve. When the second error is set to, for example, −60 °, the 0 ° characteristic curve is selected as the second characteristic curve.

  In the first characteristic curve, the smaller one of the absolute values of the first current change amount Δamp1 in each of the first error and the second error is set as the first determination value, and the first error and the second error in the second characteristic curve. The smaller one of the absolute values of the first current change amount Δamp1 in each of them is set as the second determination value. Then, the direction of the characteristic curve corresponding to the larger one of the first determination value and the second determination value is set as the sensitivity increasing direction. In FIG. 14, when the first and second errors are set to + 60 ° and −60 °, for example, 150 ° is set as the direction of increasing sensitivity.

  FIG. 16 shows a block diagram of processing of the high-frequency voltage setting unit 60 according to the present embodiment. In FIG. 16, the same processes as those shown in FIG. 4 are given the same reference numerals for the sake of convenience.

  In the high frequency voltage setting unit 60, the sensitivity storage unit 60h stores, as map information, the sensitivity increasing direction associated with the magnitude Iamp, phase θI, and estimated angular velocity ωγ of the drive current vector VIr. The signal generation unit 60g selects a direction corresponding to the calculated magnitude Iamp and phase θI of the drive current vector VIr and the estimated angular velocity ωγ from the plurality of sensitivity increasing directions stored in the sensitivity storage unit 60h. Select as. The signal generator 60g sets the amplitude of each high-frequency voltage Vγ1h, Vδ1h required to set the direction of the first high-frequency voltage vector VVh1 as the command direction θX, and according to the input estimated angle θγ, each high-frequency voltage Vγ1h, Vδ1h is output.

  According to this embodiment described above, the calculation accuracy of the estimated angle θγ can be increased.

(Fourth embodiment)
Hereinafter, the fourth embodiment will be described with reference to the drawings with a focus on differences from the third embodiment. In the present embodiment, the sensitivity increasing direction that changes according to the driving state of the motor 10 is calculated without using map information that has been adapted in advance.

  FIG. 17 shows a procedure of high-frequency voltage setting processing according to the present embodiment. This process is repeatedly executed by the high frequency voltage setting unit 60, for example, at a predetermined cycle. In the present embodiment, this process includes a direction operation unit and an actual deviation calculation unit.

  In this series of processes, first, in step S10, it is determined whether or not the driving state of the motor 10 has changed from the previous steady state to a new steady state.

  If an affirmative determination is made in step S10, in steps S12 to S16, the direction θc of the first high-frequency voltage vector VVh1 is set to each of a plurality of different directions, and the high-frequency voltages Vγ1h and Vδ1h are applied to the first winding group 10A. Is applied, and the first current change amount Δamp1 is calculated for each of the plurality of directions.

  When the calculation of the first current change amount Δamp1 is completed for all of the plurality of directions, the process proceeds to step S18. In step S18, a command direction θX as a sensitivity increasing direction is set based on the first current change amount Δamp1. Here, as the method for setting the sensitivity increasing direction, the method described in the third embodiment can be used. For example, the direction corresponding to the first current change amount having the largest absolute value among the first current change amounts Δamp1 corresponding to each of the plurality of directions can be selected as the sensitivity increasing direction. Note that the direction of increasing sensitivity may be learned in association with the driving state of the motor 10.

  In the subsequent step S20, the high-frequency voltages Vγ1h, Vδ1h, Vγ2h, Vδ2h are generated and output based on the command direction θX.

  According to the present embodiment described above, the direction of increasing sensitivity can be set without creating map information. For this reason, the man-hour at the time of design of the control apparatus 40 can be reduced. Moreover, according to this embodiment, even if the magnetic saturation state changes each time, an appropriate high-frequency voltage vector can be set according to the state.

(Fifth embodiment)
Hereinafter, the fifth embodiment will be described with reference to the drawings with a focus on differences from the first to third embodiments. In the present embodiment, a direction selected from the torque ripple reduction direction, the voltage fluctuation reduction direction, and the sensitivity increase direction is set as the command direction θX according to the driving state of the motor 10.

  FIG. 18 shows a block diagram of processing of the high-frequency voltage setting unit 60 according to the present embodiment. In FIG. 18, the same processes as those shown in FIGS. 4, 13, and 16 are given the same reference numerals for the sake of convenience.

  As illustrated, in the present embodiment, the signal generation unit of the first embodiment is a first signal generation unit 60a, the signal generation unit of the second embodiment is a second signal generation unit 60e, and The signal generation unit of the third embodiment is a third signal generation unit 60g. Based on the estimated angular velocity ωγ and the command voltages Vγ1 *, Vδ1 *, Vγ2 *, and Vδ2 *, the selection unit 60i includes the first signal generation unit 60a, the second signal generation unit 60e, and the third signal generation unit 60g. Among them, which signal generation unit is used to select the first γ-axis high-frequency voltage Vγ1h and the first δ-axis high-frequency voltage Vδ1h.

  FIG. 19 shows a procedure of processing executed by the selection unit 60i. This process is repeatedly executed at a predetermined cycle, for example.

  In this series of processing, first, in step S30, based on the command voltages Vγ1 *, Vδ1 *, Vγ2 *, Vδ2 *, a combined vector of the drive voltage vector VVr, the first high-frequency voltage vector VVh1, and the second high-frequency voltage vector VVh2. The magnitude of Vamp is calculated. In this embodiment, the magnitude of the combined vector is obtained by adding a specified value (> 0) to the magnitude of the drive voltage vector VVr calculated based on the command voltages Vγ1 *, Vδ1 *, Vγ2 *, Vδ2 *. Vamp is calculated. Here, the specified value is, for example, the maximum increase in the magnitude of the drive voltage vector VVr assumed when the direction of each of the high-frequency voltage vectors VVh1 and VVh2 is any one of the torque ripple reduction direction, the voltage fluctuation reduction direction, and the sensitivity increase direction. Is set to The specified value is a value determined in advance by experiment, calculation, or the like.

  In a succeeding step S32, it is determined whether or not the calculated combined vector size Vamp is larger than the limit voltage Vmax. When an affirmative determination is made in step S32, the process proceeds to step S34, and the second signal generation unit 60e is selected as a signal generation unit that generates the first γ-axis high-frequency voltage Vγ1h and the first δ-axis high-frequency voltage Vδ1h.

  On the other hand, if a negative determination is made in step S32, the process proceeds to step S36, where the actual torque Trq or the command torque Trq * of the motor 10 is equal to or higher than the high torque determination value TrqthH, and the drive current vector VIr flowing in the motor 10 Or the command current vector magnitude Iamp *, which is the command value of the drive current vector VIr, is determined as to whether or not the logical sum of the conditions is true. Here, the actual torque may be calculated based on the detection values of the phase current detection units 31A and 31B, for example. When an affirmative determination is made in step S36, the process proceeds to step S38, and the first signal generator 60a is selected as a signal generator that generates the first γ-axis high-frequency voltage Vγ1h and the first δ-axis high-frequency voltage Vδ1h.

  On the other hand, if a negative determination is made in step S36, the process proceeds to step S40, and the third signal generator 60g is selected as a signal generator that generates the first γ-axis high-frequency voltage Vγ1h and the first δ-axis high-frequency voltage Vδ1h.

  According to the present embodiment described above, the direction of the high-frequency voltage vector for angle estimation can be appropriately set according to the driving state of the motor 10.

(Other embodiments)
Each of the above embodiments may be modified as follows.

  In the signal generator 60a of FIG. 4 of the first embodiment, the magnitude of the drive current vector VIr is based on the first γ-axis current Iγ1r, the first δ-axis current Iδ1r, the second γ-axis current Iγ2r, and the second δ-axis current Iδ2r. The length Iamp and the phase θI may be calculated.

  The motor is not limited to one having two winding groups, and may be one having three or more winding groups. In this case, an inverter may be provided individually corresponding to each winding group. FIG. 20 shows a method for setting three high-frequency voltage vectors VVh1, VVh2, and VVh3 corresponding to the three inverters. In FIG. 20, the second high-frequency voltage vector VVh2 is canceled by the first high-frequency voltage vector VVh1 and the third high-frequency voltage vector VVh3.

  FIG. 21 shows a method for setting four high-frequency voltage vectors VVh1, VVh2, VVh3, and VVh4 corresponding to the four inverters. In FIG. 21, the second and fourth high-frequency voltage vectors VVh2 and VVh4 are canceled by the combined vector of the first and third high-frequency voltage vectors VVh1 and VVh3. In FIG. 21, the magnitude of the first high-frequency voltage vector VVh1 is equal to the magnitude of the second high-frequency voltage vector VVh2, and the magnitude of the third high-frequency voltage vector VVh3 is set equal to the magnitude of the fourth high-frequency voltage vector VVh4. .

  In the above embodiments, the high-frequency voltage is applied so that the combined vector of the first and second high-frequency voltage vectors VVh1 and VVh2 is 0, but the present invention is not limited to this. For example, as shown in FIG. 22, the high frequency voltage is applied so that the combined vector Vt of the high frequency voltage vectors VVh1 and VVh2 is smaller than the respective sizes of the first and second high frequency voltage vectors VVh1 and VVh2. May be. Even in this case, a noise reduction effect can be obtained.

  In each of the above embodiments, the angle formed by the first high-frequency voltage vector VVh1 and the second high-frequency voltage vector VVh2 is 180 °, but is not limited thereto. The first high-frequency voltage vector VVh1 and the second high-frequency voltage are satisfied if the condition that the combined vector of the high-frequency voltage vectors VVh1 and VVh2 is smaller than the first and second high-frequency voltage vectors VVh1 and VVh2 is satisfied. The angle formed with the vector VVh2 may be an angle other than 180 °.

  In each of the above embodiments, the angle formed by the first winding group 10A and the second winding group 10B is 0, but is not limited thereto, and may be a predetermined angle other than 0. In this case, the angle formed between the γ axis of the γδ coordinate system corresponding to the first winding group 10A and the γ axis of the γδ coordinate system corresponding to the second winding group 10B is the predetermined angle. For this reason, for example, in the first embodiment, in any one coordinate system selected from the γδ coordinate system corresponding to the first winding group 10A and the γδ coordinate system corresponding to the second winding group 10B. The high frequency voltage vectors VVh1 and VVh2 may be set so that the direction of the first high frequency voltage vector VVh1 and the direction of the second high frequency voltage vector VVh2 coincide with the direction in which the tangent line Lta extends.

  In the inner product calculation unit 50c of FIG. 8 of the first embodiment, the inner product value of the first high-frequency voltage vector VVh1 and the first high-frequency current vector VIh1 may be calculated based on the following equation (eq4).

上式（ｅｑ４）において、Ｉｈｘ，Ｉｈｙは第１高周波電流ベクトルＶＩｈ１のＸ，Ｙ軸方向成分である第１Ｘ軸高周波電流，第１Ｙ軸高周波電流を示し、Ｖｈｘ，Ｖｈｙは第１高周波電圧ベクトルＶＶｈ１のＸ，Ｙ軸方向成分である第１Ｘ軸高周波電圧，第１Ｙ軸高周波電圧を示す。なお、第２高周波電圧ベクトルＶＶｈ２についても同様である。

In the above equation (eq4), Ihx and Ihy indicate the first X-axis high-frequency current and the first Y-axis high-frequency current that are the X and Y-axis direction components of the first high-frequency current vector VIh1, and Vhx and Vhy are the first high-frequency voltage vector VVh1. The 1st X-axis high frequency voltage and 1st Y-axis high frequency voltage which are the X and Y-axis direction components of are shown. The same applies to the second high-frequency voltage vector VVh2.

  The high-frequency voltage is not limited to a rectangular wave pulse signal, and may be, for example, a sinusoidal high-frequency signal.

  -As a motor, not only a non-salient pole machine but a salient pole machine may be sufficient.

  DESCRIPTION OF SYMBOLS 10 ... Motor, 10A, 10B ... 1st, 2nd winding group, 20A, 20B ... 1st, 2nd inverter, 40 ... Control apparatus.

Claims (15)

A multi-winding rotating electrical machine (10) having a plurality of winding groups (10A, 10B) wound around a stator (13), and a power conversion circuit (20A, 20B) for applying a voltage to the plurality of winding groups And applied to a system comprising:
A high-frequency operation unit (40) for operating the power conversion circuit to apply a high-frequency voltage that fluctuates at an angular velocity higher than the electrical angular velocity of the rotating electrical machine to each of the plurality of winding groups;
An angle estimator (50) for estimating a rotation angle of the rotating electrical machine based on a high-frequency current flowing in the plurality of winding groups according to a high-frequency voltage applied to the plurality of winding groups,
The high-frequency operation unit includes a direction setting unit that variably sets a direction of a high-frequency voltage vector applied to the plurality of winding groups based on a driving state of the rotating electrical machine, and sets the direction of each high-frequency voltage vector as the direction setting So as to satisfy the condition of the direction set by the unit and the condition that the combined vector of the high-frequency voltage vectors is smaller than the magnitudes of the high-frequency voltage vectors applied to the plurality of winding groups. A control device for a rotating electrical machine that applies a high-frequency voltage to the plurality of winding groups. In the γδ coordinate system that is the estimated two-phase rotational coordinate system of the rotating electrical machine, a curve defined by the current for realizing the command torque of the rotating electrical machine is defined as an equal torque curve,
In the γδ coordinate system, a direction parallel to the tangential direction of the equal torque curve passing through the intersection of the equal torque curve and the current vector flowing through the rotating electrical machine is defined as a torque ripple reduction direction.
2. The control device for a rotating electrical machine according to claim 1, wherein the direction setting unit sets the direction of each high-frequency voltage vector in the torque ripple reduction direction. A ripple reduction storage unit (60b) for storing a plurality of torque ripple reduction directions associated with information relating to the current vector;
The direction setting unit selects a direction corresponding to the current current vector from the plurality of torque ripple reduction directions stored in the ripple reduction storage unit, and selects the direction of each high-frequency voltage vector as the selected direction. The control device for a rotating electrical machine according to claim 2, wherein A basic operation unit (40) for operating the power conversion circuit to control the torque of the rotating electrical machine to the command torque by minimum current maximum torque control;
The control device for a rotating electrical machine according to claim 2, wherein the direction setting unit sets the torque ripple reduction direction to a direction orthogonal to the current current vector. In the γδ coordinate system that is the estimated two-phase rotational coordinate system of the rotating electrical machine, a circle defined by a current for realizing a predetermined voltage value is defined as a constant voltage circle,
In the γδ coordinate system, a direction parallel to the tangential direction of the constant voltage circle passing through the intersection of the drive voltage vector of the rotating electrical machine and the constant voltage circle required to control the torque of the rotating electrical machine to a command torque Is defined as the voltage fluctuation reduction direction,
5. The control device for a rotating electrical machine according to claim 1, wherein the direction setting unit sets the direction of each of the high-frequency voltage vectors in the voltage fluctuation reduction direction. A voltage reduction storage unit (60f) that stores a plurality of the voltage fluctuation reduction directions in association with information related to the drive voltage vector;
The direction setting unit selects a direction corresponding to the current drive voltage vector from the plurality of voltage fluctuation reduction directions stored in the voltage reduction storage unit, and selects the direction of each high-frequency voltage vector. The control device for a rotating electrical machine according to claim 5, wherein the control device is set in the direction.   The said direction setting part sets the said voltage fluctuation reduction direction to the direction orthogonal to the said current drive voltage vector, when it determines with the electrical angular velocity of the said rotary electric machine being below a low speed determination value. Control device for rotating electrical machines. An angle formed by the d axis of the dq coordinate system that is the two-phase rotational coordinate system of the rotating electrical machine and the estimated γ axis of the γδ coordinate system that is the dq coordinate system is defined as an estimation error,
The amount of deviation between the high-frequency current flowing through the winding group in accordance with the high-frequency voltage applied to the winding group and its reference value is defined as an actual deviation,
In the state where the estimation error is a value other than 0, when the direction of the high-frequency voltage vector applied to the winding group is each of a plurality of different directions, the actual corresponding to each of the plurality of directions. Of the deviations, the direction corresponding to the actual deviation other than the actual deviation that minimizes the value is defined as the direction of increasing sensitivity,
The said direction setting part is a control apparatus of the rotary electric machine of any one of Claims 1-7 which sets the direction of each said high frequency voltage vector to the said sensitivity increase direction. A sensitivity storage unit (60h) for storing a plurality of sensitivity increasing directions associated with information relating to a current vector flowing through the rotating electrical machine;
The direction setting unit selects a direction corresponding to the current current vector from the plurality of sensitivity increasing directions stored in the sensitivity storage unit, and sets the direction of each high-frequency voltage vector to the selected direction. The control device for a rotating electrical machine according to claim 8 to be set. A direction operation unit (40) for operating the power conversion circuit to set the direction of the high-frequency voltage vector applied to the winding group in each of a plurality of different directions and to apply the high-frequency voltage;
An actual deviation calculating unit (40) for calculating the actual deviation when a high frequency voltage is applied to the winding group by the direction operation unit for each of the plurality of directions;
The direction setting unit determines a direction corresponding to an actual deviation other than the actual deviation having a minimum value among the actual deviations corresponding to each of the plurality of directions calculated by the actual deviation calculating unit as the sensitivity increasing direction. The control device for a rotating electrical machine according to claim 8, wherein A two-phase rotational coordinate system of the rotating electrical machine is defined as a dq coordinate system, and the estimated dq coordinate system is defined as a γδ coordinate system;
In the γδ coordinate system, a curve defined by a current for realizing the command torque of the rotating electrical machine is defined as an equal torque curve,
In the γδ coordinate system, a direction parallel to the tangential direction of the equal torque curve passing through the intersection of the equal torque curve and the current vector flowing through the rotating electrical machine is defined as a torque ripple reduction direction.
In the γδ coordinate system, a circle defined by a current for realizing a predetermined voltage value is defined as a constant voltage circle,
In the γδ coordinate system, parallel to the tangential direction of the constant voltage circle passing through the intersection of the drive voltage vector of the rotating electrical machine and the constant voltage circle required to control the torque of the rotating electrical machine to the command torque. The direction is defined as the voltage fluctuation reduction direction,
An angle formed by the d-axis of the dq coordinate system and the γ-axis of the γδ coordinate system is defined as an estimation error,
The amount of deviation between the high-frequency current flowing through the winding group in accordance with the high-frequency voltage applied to the winding group and its reference value is defined as an actual deviation,
In the state where the estimation error is a value other than 0, when the direction of the high-frequency voltage vector applied to the winding group is each of a plurality of different directions, the actual corresponding to each of the plurality of directions. Of the deviations, the direction corresponding to the actual deviation other than the actual deviation that minimizes the value is defined as the direction of increasing sensitivity,
A composite value calculation unit (40) that calculates the magnitude of a composite vector of the drive voltage vector and each high-frequency voltage vector;
The direction setting unit
When it is determined that the magnitude calculated by the composite value calculation unit is larger than the limit voltage, the direction of each high-frequency voltage vector is set to the voltage fluctuation reduction direction,
It is determined that the magnitude calculated by the composite value calculation unit is less than or equal to the limit voltage, and the actual torque or command torque of the rotating electrical machine is greater than or equal to a high torque determination value, or the current that flows to the rotating electrical machine When it is determined that the magnitude of the vector or the magnitude of the command current vector is greater than or equal to the large current determination value, the direction of each high-frequency voltage vector is set to the torque ripple reduction direction,
It is determined that the magnitude calculated by the composite value calculation unit is equal to or less than the limit voltage, and the actual torque or command torque of the rotating electrical machine is smaller than the high torque determination value, or the current vector that flows to the rotating electrical machine 11. The direction of each of the high-frequency voltage vectors is set to the direction of increasing sensitivity when it is determined that the magnitude of the signal current or the magnitude of the command current vector is smaller than the large current determination value. Rotating electrical machine control device. The amount of deviation between the high-frequency current flowing through the winding group in accordance with the high-frequency voltage applied to the winding group and its reference value is defined as an actual deviation,
The angle estimation unit calculates the actual deviation, calculates an electrical angular velocity of the rotating electrical machine as an operation amount for feedback control of the calculated actual deviation to the reference value, and integrates the calculated electrical angular velocity. The control device for a rotating electrical machine according to any one of claims 1 to 11, wherein the rotation angle is estimated.   The angle estimator includes a shift amount between an inner product value of the high-frequency voltage vector and a current vector flowing in the rotary electric machine in the αβ coordinate system which is a two-phase fixed coordinate system of the rotary electric machine, and the high-frequency voltage vector. The control apparatus for a rotating electrical machine according to claim 12, wherein a deviation amount between the change amount of the high-frequency current and the reference value in the direction is calculated as the actual deviation.   The condition that the magnitude of the combined vector of each high-frequency voltage vector is smaller than the magnitude of each of the high-frequency voltage vectors applied to the plurality of winding groups is that the magnitude of the combined vector of the respective high-frequency voltage vectors is 0. The control device for a rotating electrical machine according to any one of claims 1 to 13, wherein the control device is a condition to be used. The plurality of winding groups are two winding groups;
The control device for a rotating electrical machine according to any one of claims 1 to 14, wherein the power conversion circuit is individually provided corresponding to each of the two winding groups.

Images