JP5402106B2 - Electric motor control device and electric motor state estimation method - Google Patents

Description

  The present invention relates to a motor control device that drives a motor by supplying electric power and a motor state estimation method.

  2. Description of the Related Art Conventionally, motor control devices that include a position sensor such as an encoder or a resolver are known in order to detect the position and rotation speed of a rotor. The electric motor control device controls the electric motor based on the position and the rotational speed of the rotor detected by the position sensor.

  However, such a motor control device has a problem in that it is difficult to reduce the size in order to increase the cost associated with the attachment of the position sensor and to secure the installation space and the wiring space.

  On the other hand, the motor control device disclosed in Patent Document 1 uses the property of d-axis (excitation direction component) inductance <q-axis (torque direction component) inductance as the saliency of the embedded magnet type synchronous motor. Due to this property, when a high-frequency voltage vector that draws a perfect circle with a frequency different from the drive frequency of the motor is applied to the motor, the vector of the high-frequency component of the current flowing through the motor swells in the d-axis direction (magnetic pole position). Draw an elliptical trajectory. The motor controller detects the major axis direction of the ellipse by approximating the high-frequency current vector locus to an ellipse, estimates the rotor pole position and the number of rotations, assuming that the major axis direction of the ellipse is the pole position, and these estimated values The motor is controlled based on the above.

JP 2003-219682 A

  The technique described above is based on the premise that the inductance distribution between the rotor and the stator is determined by the rotor structure, and the high-frequency current in the d-axis direction takes the maximum value and the high-frequency current in the q-axis direction takes the minimum value.

  However, the inductance distribution due to the structure of the stator is not uniform in the d-axis and q-axis, and the error of the high-frequency current based on the inductance distribution may become a size that cannot be ignored. Therefore, when the vector locus of the high-frequency current is approximated by an ellipse, the approximated ellipse includes an error that varies depending on the position of the rotor. There is a problem that cannot be estimated well.

  Therefore, the present invention has been proposed in view of the above-described circumstances, and an object of the present invention is to provide an electric motor control device and an electric motor state estimation method capable of accurately estimating a rotor state.

  The present invention provides a voltage command value calculating means for calculating a voltage command value of an excitation direction component and a voltage command value of a torque direction component of the motor as a voltage command value for controlling the motor to a desired state, and a voltage command value High frequency voltage generating means for superimposing a high frequency voltage having a frequency different from the drive frequency of the motor on the voltage command value calculated by the calculating means, and an inverter based on the voltage command value on which the high frequency voltage is superimposed by the high frequency voltage generating means A drive signal for driving the switching element, a drive means for driving a motor by converting a DC voltage into an AC voltage by the inverter, a current detection means for detecting a current supplied from the drive means to the motor, A rotor state estimating means for estimating the state of the rotor in the electric motor based on the current detected by the current detecting means; and the rotor state Based on the rotor position estimated by the constant means, and control means for controlling the driving means to drive the electric motor.

In order to solve the above-described problem, the rotor state estimation means in the present invention calculates the excitation direction component current and the torque direction component current of the fundamental wave component having the same frequency as the high frequency voltage included in the current detected by the current detection means. In addition to extraction, the excitation direction component current and the torque direction component current of the harmonic component for the same frequency as the high frequency voltage included in the current detected by the current detection means are extracted, and the excitation direction component current of the extracted fundamental wave component Further, the high frequency current vector locus is approximated by an ellipse using the torque direction component current to obtain a first ellipse major axis direction that is a deviation of the ellipse major axis direction from the excitation direction, and the excitation direction of the extracted harmonic component A high-frequency current vector locus is approximated by an ellipse using the component current and the torque direction component current, and is the deviation of the ellipse major axis direction from the excitation direction. Obtains an ellipse long axis direction, based on the difference between the first ellipse long axis direction and the second ellipse major axis, and calculating a correction amount of the first ellipse long axis direction, based on the correction amount, first ellipse The major axis direction is corrected, and the rotor position is estimated based on the corrected first ellipse major axis direction.

  According to the present invention, the motor is driven by superimposing a high-frequency component on the voltage command value, and the ellipse major axis direction of the high-frequency current vector locus based on the fundamental wave component of the current actually supplied to the motor is The elliptical long axis direction of the high frequency current vector locus based on the fundamental wave component can be corrected using the elliptical long axis direction of the high frequency current vector locus based on the harmonic component. Thus, the state of the rotor can be accurately estimated by utilizing the fact that the current of the harmonic component has a correlation with the rotation of the rotor.

It is a block diagram which shows the structure of the control apparatus of the electric motor shown as 1st Embodiment of this invention. It is a figure which shows the locus | trajectory of a high frequency voltage vector in the control apparatus of the electric motor shown as 1st Embodiment of this invention. It is a block diagram which shows the structure of the magnetic pole position and rotation speed estimation part in the control apparatus of the electric motor shown as 1st Embodiment of this invention. It is a top view which shows the structure of the electric motor in the control apparatus of the electric motor shown as 1st Embodiment of this invention. It is a figure which shows the relationship between the position of the rotor and inductance in the control apparatus of the electric motor shown as 1st Embodiment of this invention. It is a figure which shows the relationship between the position of a stator and inductance in the control apparatus of the electric motor shown as 1st Embodiment of this invention. It is a figure explaining the ellipse of the high frequency current vector locus in case a stator is a distributed winding structure. In the motor control apparatus shown as the first embodiment of the present invention, when the stator has a concentrated winding structure, the high-frequency current vector locus becomes an ellipse. It is a block diagram which shows the structure of the magnetic pole position correction amount calculation part in the control apparatus of the electric motor shown as 1st Embodiment of this invention. It is a figure which shows the relationship between the difference of 1st ellipse major axis direction (deviation) and 2nd ellipse major axis direction (deviation), and a magnetic pole position correction amount. In the motor control apparatus shown as the first embodiment of the present invention, it is a diagram showing the relationship between the rotational operation of the motor and the high-frequency (fundamental) current vector locus and the high-frequency (harmonic) current vector locus. In the motor control apparatus shown as the first embodiment of the present invention, it is a diagram showing the relationship between the rotational operation of the motor and the high-frequency (fundamental) current vector locus and the high-frequency (harmonic) current vector locus. In the motor control apparatus shown as the first embodiment of the present invention, it is a diagram showing the relationship between the rotational operation of the motor and the high-frequency (fundamental) current vector locus and the high-frequency (harmonic) current vector locus. In the motor control apparatus shown as the first embodiment of the present invention, it is a diagram showing the relationship between the rotational operation of the motor and the high-frequency (fundamental) current vector locus and the high-frequency (harmonic) current vector locus. In the motor control apparatus shown as the first embodiment of the present invention, it is a diagram showing the relationship between the rotational operation of the motor and the high-frequency (fundamental) current vector locus and the high-frequency (harmonic) current vector locus. It is a flowchart which shows the operation | movement procedure in the control apparatus of the electric motor shown as 1st Embodiment of this invention. It is a block diagram which shows the structure of the magnetic pole position correction amount calculation part in the control apparatus of the electric motor shown as 2nd Embodiment of this invention. It is a block diagram which shows the structure of the magnetic pole position and rotation speed estimation part in the control apparatus of the electric motor shown as 3rd Embodiment of this invention.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

[First Embodiment]
"Configuration of motor control device"
The motor control device shown as the first embodiment of the present invention is configured as shown in FIG. 1, for example. The electric motor control device controls, for example, the operation of the electric motor 9 that drives the vehicle. In the present embodiment, the motor control device includes a command value generator 1, a current controller 2, a high frequency voltage generator 3, a coordinate converter 4, a PWM converter 5, a DC power supply 6, an inverter 7, current sensors 8u, 8v, A coordinate conversion unit 10 and a magnetic pole position / rotation number estimation unit 11 are included.

  The command value generation unit 1 and the current control unit 2 use the voltage command value and the torque direction component (δ axis) of the excitation direction component (γ axis) of the motor 9 as a voltage command value for controlling the motor 9 to a desired state. ) Functions as voltage command value calculation means for calculating the voltage command value.

The command value generation unit 1 includes a torque command value T ^* , an estimated rotational speed value ω ^ of the electric motor 9, a DC voltage V _dc of the DC power source 6, and dp-axis current command values i _d ^* , i _q ^* in the current control mode, and a table showing a correspondence relationship between the dq-axis interference voltage command values v _{d_dcpl} ^* and v _{q_dcpl} ^* ; The command value generator 1 is a dq-axis current corresponding to the torque command value T ^* , the estimated rotational speed value ω ^ of the motor 9 output from the magnetic pole position / rotational speed estimator 11, and the DC voltage V _dc of the DC power supply 6. The command values i _d ^* , i _q ^* and dq axis interference voltage command values v _{d_dcpl} ^* , v _{q_dcpl} ^* are retrieved from the table, and the retrieved dq axis current command values i _d ^* , i _q ^* and dq axis interference voltage commands The values v _{d_dcpl} ^* and v _{q_dcpl} ^* are output. In this way, the command value generation unit 1 controls the dq axis current command values i _d ^* and i _q ^* and the dq axis interference voltage command values v _{d_dcpl} ^* and v _{q_dcpl} ^* by an open loop method.

The current control unit 2 outputs the dq axis current command values i _d ^* and i _q ^* and the dq axis interference voltage command values v _{d_dcpl} ^* and v _{q_dcpl} ^* output from the command value generation unit 1 and the coordinate conversion unit 10. A general current vector control calculation including current deviation proportional integration (PI) amplification and non-interference control is performed using the γδ axis current detection values i _γ and i _δ . As a result, the current control unit 2 generates and outputs the γδ-axis voltage command values v _γ ^* ′ and v _δ ^* ′.

The motor control device performs current vector control for performing current control by separating the current supplied to the motor 9 into a q-axis current component orthogonal to the secondary magnetic flux of the motor 9 and a d-axis current component parallel to the secondary magnetic flux. Do. “Non-interference control” means that in this current vector control, the d-axis current component and the q-axis current component are changed to q-axis voltage and This means control that cancels the influence of interference with the other current component as the d-axis voltage. Specifically, it means control for correcting the voltage command value using the interference voltage command value. Further, the γδ axis means an estimated dq axis obtained from the estimated magnetic pole position θ ^ of the rotor estimated by the magnetic pole position / rotational speed estimation unit 11. The current deviation is a deviation between the d-axis current command value i _d ^* and the γ-axis current detection value i _{γ, and} a deviation between the q-axis current command value i _q ^* and the δ-axis current detection value i _δ .

  The high-frequency voltage generator 3 functions as a high-frequency voltage generator that superimposes a high-frequency voltage having a frequency different from the drive frequency of the motor 9 on the voltage command value calculated by the command value generator 1 and the current controller 2.

The high frequency voltage generator 3 generates high frequency voltages v _dh ^* and v _qh ^* different from the drive frequency of the electric motor 9. The high-frequency voltages v _dh ^* and v _qh ^* are superimposed on the γδ axis voltage command values v _γ ^* and v _δ ^* generated by the current control unit 2. As a result, the coordinate conversion unit 4 is supplied with γδ-axis voltage command values v _γ ^* and v _δ ^* including the high-frequency voltages v _dh ^* and v _qh ^* .

Further, the motor control device generates a drive signal for driving the switching element of the inverter 7 based on the voltage command value on which the high-frequency voltages v _dh ^* and v _qh ^* are superimposed by the high-frequency voltage generator 3, and the inverter 7 It functions as a driving means for driving the electric motor 9 by converting the DC voltage into an AC voltage. This configuration corresponds to the coordinate conversion unit 4, the PWM conversion unit 5, and the inverter 7.

The coordinate conversion unit 4 converts the γδ-axis voltage into a UVW phase voltage. The coordinate conversion unit 4 uses the following Equation 1 to input the γδ-axis voltage command values v _γ ^* and v _δ ^* and the magnetic pole position of the rotor of the motor 9 estimated by the magnetic pole position / rotation number estimation unit 11. The U-phase, V-phase, and W-phase voltage command values v _u ^* , v _v ^* , v _w ^* are calculated from the estimated value θ ^ and output.

The PWM converter 5 is a drive signal D for the inverter 7 corresponding to the three-phase voltage command values v _u ^* , v _v ^* , v _w ^* output from the coordinate converter 4. ^{_{uu *, D ul *, D}} vu *, D vl *, D wu *, to generate a _{D wl} ^*, to output. The drive signals D _uu ^* and D _ul ^* indicate signals for the upper and lower switching elements corresponding to the U phase, respectively. The drive signals D _vu ^* and D _vl ^* indicate signals for the upper and lower switching elements corresponding to the V phase, respectively. The driving signals D _wu ^* and D _wl ^* indicate signals for the upper and lower switching elements corresponding to the W phase, respectively.

The inverter 7 turns on / off the corresponding switching element according to the drive signals D _uu ^* , D _ul ^* , D _vu ^* , D _vl ^* , D _wu ^* , and D _wl ^* output from the PWM converter 5. As a result, the inverter 7 converts the DC voltage V _dc of the DC power supply 6 into three-phase AC voltages v _u , v _v , v _w and outputs the converted voltage to the motor 9.

The current sensors 8u and 8v function as current detection means for detecting the excitation direction component current and the torque direction component current supplied from the inverter 7 to the electric motor 9. Specifically, the current sensors 8 _u and 8 _v detect U-phase and V-phase current values i _u and iv and output them to the coordinate conversion unit 10. When the current sensor is attached to only two phases as in the present embodiment, the current value of the remaining one phase (W phase in the present embodiment) that is not detected by the current sensors 8u and 8v is calculated by the coordinate conversion unit 10 as follows: It can be calculated from Equation 2.

The coordinate conversion unit 10 converts current from the UVW phase to the γδ axis. The coordinate conversion unit 10 uses the following equation 3 to calculate the U-phase and V-phase current values i _u and iv detected by the current sensors 8 _u and 8 _v and the motor estimated by the magnetic pole position / rotation number estimation unit 11. The γδ axis current detection values i _γ and i _δ are calculated from the estimated magnetic pole position θ ^ of the 9 rotor.

The magnetic pole position / rotation number estimation unit 11 functions as a rotor state estimation unit that estimates the state of the rotor in the electric motor 9 based on the current detected by the current sensors 8 u and 8 v and the coordinate conversion unit 10. Pole position and rotational speed estimating unit 11, the ??-axis current detection value output from the coordinate conversion unit 10 i _gamma, type the i _[delta], these the ??-axis current detection value i _gamma, based on the i _[delta], the motor 9 The rotor magnetic pole position estimated value θ ^ and the rotational speed estimated value ω ^ are calculated and output.

  The magnetic pole position estimated value θ ^ and the rotational speed estimated value ω ^ are used for control of the electric motor 9 by the command value generation unit 1 and the coordinate conversion unit 10 (control means).

"Description of operation of high-frequency voltage generator 3"
As shown in FIG. 2, the high-frequency voltage generator 3 generates high-frequency voltages v _dh ^* and v _qh ^* that draw a perfect circle vector locus. The frequencies of the high-frequency voltages v _dh ^* and v _qh ^* are sufficiently higher than the drive frequency of the electric motor 9. For example, the vector size of the high frequency voltages v _dh ^* and v _qh ^* is 85 V and the frequency is 625 Hz.

"Detailed configuration of magnetic pole position / rotation speed estimation unit 11"
Next, the configuration of the magnetic pole position / rotation number estimation unit 11 in the above-described electric motor control device will be described with reference to FIG.

  The magnetic pole position / rotation speed estimation unit 11 includes a first bandpass filter 12 (BPF), a first digital Fourier transform processing unit 13 (DFT), a first ellipse analysis unit 14, a PI amplifier 15, an integrator 16, and magnetic pole position correction. An amount calculation unit 17 is provided. The magnetic pole position / rotation speed estimation unit 11 includes a second band pass filter 18, a second digital Fourier transform processing unit 19, and a second ellipse analysis unit 20.

The first band-pass filter 12 is included in the detected γδ-axis current values i _γ and i _δ output from the coordinate conversion unit 10, and the γδ-axis high-frequency current i _γh1 having the same frequency component as the high-frequency voltages v _dh ^* and v _qh ^*. , I _δh1 is extracted and output.

The first digital Fourier transform processing unit 13 uses the γδ-axis high-frequency currents i _γh1 and i _δh1 extracted by the first bandpass filter 12 as real-frequency components and imaginary-axis components having the same frequency components as the high-frequency voltages v _dh ^* and v _qh ^*. The components i _{γh1_Re} , i _{γh1_Im} , i _{δh1_Re} , and i _{δh1_Im} are decomposed and output.

The first ellipse analysis unit 14 uses the imaginary axis components i _{γh1_Re} , i _{γh1_Im} , i _{δh1_Re} , and i _{δh1_Im} output from the first digital Fourier transform processing unit 13 so that the γδ axis high-frequency currents i _γh1 and i _δh1 are γδ axes. The locus drawn above, that is, the γδ axis current detection values i _γ and i _δ are approximated to an ellipse (the ellipse of the high-frequency current vector locus). Then, the first ellipse analysis unit 14 calculates and outputs the first ellipse major axis direction (deviation) θ _ζ ′ of the high-frequency current vector locus with respect to the γ-axis by the following equation (4).

In the following description, an elliptical long axis in a high-frequency current vector locus obtained by using γδ-axis high-frequency currents i _γh1 and i _δh1 having the same frequency component (fundamental wave component) as the high-frequency voltages v _dh ^* and v _qh ^* The direction is referred to as “first elliptical long axis direction”.

The second band-pass filter 18 is included in the γδ-axis current detection values i _γ and i _δ output from the coordinate conversion unit 10, and the γδ axis of the harmonic component for the same frequency component as the high-frequency voltages v _dh ^* and v _qh ^*. The high-frequency currents i _γh2 and i _δh2 are extracted and output. The harmonic component in this embodiment is a frequency component that is twice the same frequency component as the high-frequency voltages v _dh ^* and v _qh ^* . That is, the magnetic pole position / rotation speed estimation unit 11 sets the harmonic component as the second harmonic component with respect to the frequencies of the high-frequency voltages v _dh ^* and v _qh ^* .

The second digital Fourier transform processing unit 19 uses the γδ-axis high-frequency currents i _γh2 and i _δh2 extracted by the second bandpass filter 18 as real-frequency components and imaginary-axis components having the same frequency components as the high-frequency voltages v _dh ^* and v _qh ^*. The components i _{γh2_Re} , i _{γh2_Im} , i _{δh2_Re} , and i _{δh2_Im} are decomposed and output.

Based on the imaginary axis components i _{γh2_Re} , i _{γh2_Im} , i _{δh2_Re} , and i _{δh2_Im} output from the second digital Fourier transform processing unit 19, the second ellipse analysis unit 20 converts the γδ axis high-frequency currents i _γh2 and i _δh2 to the γδ axis. The locus drawn above, that is, the γδ axis current detection values i _γ and i _δ are approximated to an ellipse (the ellipse of the high-frequency current vector locus). Then, the first ellipse analysis unit 14 calculates and outputs the second ellipse major axis direction (deviation) θ _ζ2 of the high-frequency current vector locus with respect to the γ-axis by the following equation (5).

In the following description, the major axis direction of the elliptical shape in the high-frequency current vector locus obtained by using the γδ-axis high-frequency currents i _γh1 and i _δh1 of the harmonic components for the same frequency component with respect to the high-frequency voltages v _dh ^* and v _qh ^* Is referred to as “second elliptical major axis direction”.

The elliptical long axis direction θ _ζ ′ with respect to the γ axis is added to the magnetic pole position correction amount Δθ _ζ output from the magnetic pole position correction amount calculation unit 17. Thereby, the elliptical long axis direction θ _ζ ′ with respect to the γ axis is corrected to the elliptical long axis direction θ _ζ of the high-frequency current vector locus with respect to the γ axis. The elliptical long axis direction θ _ζ of the high-frequency current vector locus with respect to the γ axis is supplied to the PI amplifier 15.

The PI amplifier 15 obtains the rotational speed estimated value ω ^ by PI amplification of the elliptical long axis direction θ _ζ of the high-frequency current vector locus with respect to the γ-axis, and outputs it to the command value generation unit 1.

  The integrator 16 integrates the estimated rotational speed value ω ^ output from the PI amplifier 15 to obtain the estimated magnetic pole position θ ^ of the rotor and outputs it to the coordinate conversion unit 10.

  Here, the principle of obtaining the rotor magnetic pole position estimated value θ ^ and the rotational speed estimated value ω ^ will be described. FIG. 4 shows concentrated winding when the magnet (S, N) of the rotor 9a in the electric motor 9 is one pole pair and the number of teeth (stator 9b) on the fixed side is six (1) to (6). An IPM type electric motor 9 is shown. 5 and 6 show the distributions of inductances Lr and Ls of the electric motor 9.

  The inductance Lr caused by the rotor 9a of the electric motor 9 is distributed as shown in FIG. 5 with respect to the parts (a), (b), (c), and (d) of the rotor 9a shown in FIG. On the other hand, the inductance Ls caused by the stator 9b is distributed as shown in FIG. 6 with respect to the parts (1) to (6) of the stator 9b shown in FIG.

When the stator 9b has a distributed winding structure, the inductance Ls caused by the stator 9b has a substantially flat distribution as shown by the dotted line in FIG. As a result, the combined inductance of the inductance Lr caused by the rotor 9a and the inductance Ls caused by the stator 9b is minimized in the parts (a) and (c) of the rotor 9a, that is, in the d-axis direction. (B), (d), that is, the maximum in the q-axis direction. Therefore, the locus of the high-frequency current vector represented by the γδ-axis current detection values i _γ and i _δ becomes an ellipse that swells in the d-axis direction as shown in FIG. Therefore, the first ellipse analysis unit 14 is based on the imaginary axis components i _{γh1_Re} , i _{γh1_Im} , i _{δh1_Re} , i _{δh1_Im} based on the γδ axis current detection values i _γ and i _δ obtained from the first digital Fourier transform processing unit 13. Approximating the high-frequency current vector locus with an ellipse and calculating the elliptical long axis direction of the high-frequency current vector locus to calculate the elliptical long axis direction (deviation) θ _ζ ′ with respect to the γ axis, and finally the rotor The magnetic pole position estimated value θ ^ of 9a can be estimated.

  However, when the stator 9b has a concentrated winding structure, the inductance Ls caused by the stator 9b is the maximum at a position where the stator 9b (tooth) is present, and the minimum between the stators 9b, and the difference due to the position of the stator 9b is different. growing. For this reason, the combined inductance of the inductance Lr caused by the rotor 9a and the inductance Ls caused by the stator 9b is not necessarily minimized in the parts (a) and (c) of the rotor 9a, that is, in the d-axis direction. Is the smallest at a point deviated within a certain range. Therefore, as shown in FIG. 8, the high-frequency current vector locus has an elliptical shape in which the inductance swells in the minimum direction and the size expands and contracts around the d-axis. Thus, the major axis direction (γ) when the high-frequency current vector locus is approximated by an ellipse is a direction deviated from the d-axis direction such as an ellipse indicated by a dotted line in FIG.

  Therefore, in order to accurately estimate the magnetic pole position in the motor 9 in which the stator 9b has a concentrated winding structure, the true magnetic pole position (d-axis) that varies according to the magnetic pole position of the rotor 9a and the high-frequency current vector locus It is necessary to correct an error (deviation) from the first ellipse major axis direction (γ).

  For this reason, the motor control device shown as the first embodiment has a high-frequency current distortion component in which the elliptical shape of the high-frequency current vector locus deviates from the d-axis, the true magnetic pole position, and the elliptical shape as shown in FIG. Focusing on the fact that both the errors from the direction of the ellipse major axis are due to the stator structure and have a certain relationship, the ellipse-shaped first ellipse major axis in the true magnetic pole position and the high-frequency current vector locus Correct the direction error. This correction will be described below.

The first ellipse major axis direction (deviation) θ _ζ ′ and the second ellipse major axis direction (deviation) θ _ζ2 are supplied to the magnetic pole position correction amount calculation unit 17 of FIG. 3 described above. By using the first ellipse long axis direction (deviation) theta _zeta 'and the second ellipse major axis direction (deviation) theta _{?? 2,} the magnetic pole position correction amount calculating unit 17 calculates the magnetic pole position correction amount [Delta] [theta] _zeta.

As shown in FIG. 9, the magnetic pole position correction amount calculation unit 17 is supplied with a value obtained by subtracting the first ellipse major axis direction (deviation) θ _ζ ′ from the second ellipse major axis direction (deviation) θ _ζ2 . A position correction amount map reference unit 21 is provided. The magnetic pole position correction amount map reference unit 21 includes a first ellipse major axis direction (deviation) θ _ζ ′ that is a major axis direction when a vector locus drawn by a fundamental wave component in a high frequency current is approximated to an ellipse and a high frequency current. The difference between the second ellipse major axis direction (deviation) θ _ζ2 , which is the major axis direction when the vector locus drawn by the harmonic component is approximated by an ellipse, and the excitation direction and the first ellipse major axis direction (deviation) θ _ζ ′ And the error (magnetic pole position correction amount Δθ _ζ ) are stored as map data. Then, the magnetic pole position correction amount map reference unit 21 refers to the map data in response to the input of the difference between the second ellipse major axis direction (deviation) θ _ζ2 and the first ellipse major axis direction (deviation) θ _ζ ′. The magnetic pole position correction amount Δθ _ζ is calculated and output.

The magnetic pole position correction amount map reference unit 21 preferably stores map data for each operating point of the electric motor 9. In this case, the magnetic pole position correction amount map reference section 21 selects the map data based on the operating point of the motor 9, the magnetic pole position correction amount [Delta] [theta] _zeta to correct an error of the first ellipse long axis direction (deviation) theta _zeta ' Is calculated. The magnetic pole position correction amount map reference unit 21 may calculate the magnetic pole position correction amount Δθ _ζ by performing linear interpolation according to the operating point of the electric motor 9.

For example, as shown in FIG. 10, the map data includes a difference (θ _ζ '−θ _ζ2 [°] between the first elliptical long axis direction (deviation) θ _ζ ′ and the second elliptical long axis direction (deviation) θ _ζ2 . ) And the magnetic pole position correction amount Δθ _ζ [°]. The relationship described in such map data will be described with reference to FIGS.

As shown in FIGS. 11 (a) to 15 (a), the rotor 9a indicated by a dotted line in the figure is positioned at the position (1) in FIG. 11 (a), the position (2) in FIG. Suppose that it rotates towards position (3) in (a). In this case, the second ellipse major axis direction (deviation) θ _ζ2 due to the vector locus of the secondary component of the high-frequency current changes as shown in FIGS. 11B to 15B. That is, the second ellipse major axis direction (deviation) θ _ζ2 is + 90 ° in FIG. 11B, 0 ° in FIG. 13B, and −90 ° in FIG. 15B according to the position of the rotor 9a. It will change. Thus, the second ellipse major axis direction (deviation) θ _ζ2 has a characteristic of rotating in the direction opposite to the rotation direction of the rotor 9a. Further, the error θ _ζe between the first ellipse major axis direction (deviation) θ _ζ ′ and the true magnetic pole position varies in one cycle according to the position of the rotor 9a. _Therefore , the difference between the first elliptical long axis direction (deviation) θ _ζ ′ and the second elliptical long axis direction (deviation) θ _ζ2 and the error θ between the first elliptical long axis direction (deviation) θ _ζ ′. _{The relationship} with the magnetic pole position correction amount Δθ _ζ that cancels _ζe can be obtained as shown in FIG.

"Operation of motor control device"
Next, processing for estimating the position of the electric motor 9 including the magnetic pole position and the rotation speed of the rotor 9a by the electric motor control device described above will be described with reference to FIG.

The motor control device supplies AC power from the inverter 7 to the motor 9 based on the torque command value T ^* and the DC voltage V _dc to drive the motor 9 every predetermined period (for example, several msec). The process after step S1 is started.

In step S <b> 1, the magnetic pole position / rotation speed estimation unit 11 acquires the γδ axis current detection values i _γ and i _δ from the coordinate conversion unit 10. At this time, the coordinate conversion unit 10 obtains the U-phase and V-phase current values i _u and iv from the current sensors 8 _u and 8 _v and calculates the γδ-axis current detection values i _γ and i _δ. It supplies to the rotation speed estimation part 11.

In the next step S2, the magnetic pole position / rotation speed estimation unit 11 uses the first bandpass filter 12 to detect the high frequency voltages v _dh ^* and v _qh from the γδ axis current detection values i _γ and i _δ acquired in step S1. _Γδ- axis high-frequency currents i _γh1 and i _δh1 having the same high frequency component (fundamental wave component) as ^* are extracted. Similarly, the second band-pass filter 18 uses the second-order components (harmonics) for the high-frequency components that are the same as the high-frequency voltages v _dh ^* and v _qh ^* from the γδ-axis current detection values i _γ and i _δ acquired in step S1. The component γδ-axis high-frequency currents i _γh2 and i _δh2 are extracted.

In the next step S3, the magnetic pole position / rotation speed estimation unit 11 performs the digital Fourier transform processing (DFT) by the first digital Fourier transform processing unit 13 and the second digital Fourier transform processing unit 19 to extract the γδ extracted in step S2. The axial high-frequency currents i _γh1 and i _δh1 and the γδ-axis high-frequency currents i _γh2 and i _δh2 are separated into a real axis component and an imaginary axis component.

In the next step S4, the magnetic pole position / rotation speed estimation unit 11 uses the first ellipse analysis unit 14 to calculate the high frequency current (fundamental wave) vector from the change in the values (magnitudes) of the γδ axis high frequency currents i _γh and i _δh. The locus is obtained by ellipse approximation. Then, the first ellipse analyzer 14 calculates the first ellipse major axis direction (deviation) θ _ζ ′ of the high-frequency current vector locus with respect to the γ-axis.

Similarly, in step S5, the second ellipse analysis unit 20 obtains a high-frequency current (second-order harmonic) vector locus by elliptical approximation from changes in the values (magnitudes) of the γδ-axis high-frequency currents i _γh and i _δh. . Then, the first ellipse analysis unit 14 calculates the second ellipse major axis direction (deviation) θ _ζ2 of the high-frequency current vector locus with respect to the γ-axis.

In the next step S6, the magnetic pole position correction amount calculator 17 calculates the first ellipse major axis direction (deviation) θ _ζ ′ calculated in step S4 and the second ellipse major axis direction calculated in step S5 ( Deviation) Reference is made to the map data stored in the magnetic pole position correction amount map reference unit 21 based on the difference from _θζ2 . Thereby, the magnetic pole position correction amount map reference unit 21 calculates and outputs the magnetic pole position correction amount Δθζ corresponding to the difference.

Here, the magnetic pole position correction amount calculation unit 17 uses not only the case of using a map, but also the difference between the first elliptical long axis direction (deviation) θ _ζ ′ and the second elliptical long axis direction (deviation) θ _ζ2. Processing according to a predetermined arithmetic expression for obtaining the magnetic pole position correction amount Δθ _ζ may be performed.

In the next step S7, the magnetic pole position / rotation speed estimation unit 11 corrects the magnetic pole position obtained in step S6 from the elliptical long axis direction θ _ζ ′ of the high-frequency current vector locus with respect to the γ axis obtained in step S4. Subtract the amount Δθ _ζ to calculate the elliptical long axis direction θ _ζ of the high-frequency current vector locus for the new γ-axis.

In the next step S10, the PI amplifier 15 calculates the rotational speed estimated value ω ^ by PI amplification of the elliptical long axis direction θ _ζ of the high-frequency current vector locus with respect to the γ axis calculated in step S9.

  In the next step S11, the PI amplifier 15 calculates the magnetic pole position estimated value θ ^ by integrating the rotational speed estimated value ω ^ calculated in step S10.

Thus, the motor control device can obtain the estimated rotational speed value ω ^ and the estimated magnetic pole position θ ^ using the γδ-axis current detection values i _γ and i _δ. The motor 9 can be controlled using the estimated magnetic pole position value θ ^.

“Effect of the first embodiment”
As explained above, according to the motor control device shown as the first embodiment of the present invention,
When the electric motor 9 is driven, the high frequency voltages v _dh ^* and v _qh ^* are superimposed on the γδ axis voltage command values v _γ ^* and v _δ ^* generated by the current control unit 2. When the magnetic pole position estimated value θ ^ and the rotational speed estimated value ω ^ as the state of the motor 9 are _obtained , the same frequency as the high frequency voltages v _dh ^* and v _qh ^* is _obtained from the γδ axis current detection values i _γ and i _δ. of the ?? axes high frequency current i _Ganmaetchi1, the i _{Delta] hl,} the ?? axes high frequency current i _Ganmaetchi2 harmonic components, extracts the i _{[Delta] h2.} Next, the motor control device approximates the _extracted vector locus of the γδ-axis high-frequency currents i _γh1 , i _δh1 , i _γh2 , i _δh2 with an ellipse. Then, the magnetic pole position correction amount Δθ _ζ is obtained using the first ellipse major axis direction (deviation) θ _ζ ′ and the second ellipse major axis direction (deviation) θ _{ζ2 of} the ellipse, and the first ellipse major axis direction ( Deviation) θ _ζ 'is corrected. Next, based on the corrected first ellipse major axis direction (deviation) θ _ζ ′, the magnetic pole position estimated value θ ^ can be obtained.

  According to such an electric motor control device, even if the position sensor for detecting the position of the rotor 9a is not provided and the state of the rotor 9a is estimated, the current of the harmonic component is thereby generated by the rotor. The state of the rotor can be accurately estimated by utilizing the fact that there is a correlation with the rotation of the rotor.

Further, according to this motor control device, since the harmonic component is the second harmonic component with respect to the frequency of the high frequency voltage v _dh *, v _qh *, a high-precision and high-resolution current detector can be added, The estimation accuracy of the magnetic pole position and speed of the rotor 9a can be improved without shortening the detection interval of the high-frequency current.

Further, according to this motor control device, the _relationship between the difference between the first elliptical long axis direction (deviation) θ _ζ ′ and the second elliptical long axis direction (deviation) θ _ζ2 and the magnetic pole position correction amount Δθ _ζ is mapped. Since it is stored as data and the magnetic pole position correction amount Δθ _ζ is obtained based on the map data, it is possible to improve the estimation accuracy of the magnetic pole position / speed of the rotor 9a.

Furthermore, according to this motor control device, the map data for obtaining the magnetic pole position correction amount Δθ _ζ is stored for each operating point of the motor 9, and the map data is selected based on the operating point of the motor 9. Even if the operating point of the electric motor 9 changes, the first ellipse major axis direction (deviation) θ _ζ ′ can be accurately corrected.

[Second Embodiment]
Next, an electric motor control device shown as the second embodiment will be described. In addition, about the part similar to the above-mentioned 1st Embodiment, the detailed description is abbreviate | omitted by attaching | subjecting the same code.

The motor control device shown as the second embodiment constitutes a magnetic pole position correction amount calculation unit 17 as shown in FIG. The magnetic pole position correction amount calculation unit 17 calculates a magnetic pole position correction amount Δθ _ζ using a difference between the first elliptical long axis direction (deviation) θ _ζ ′ and the second elliptical long axis direction (deviation) θ _ζ2. A correction amount calculation unit 22 is provided. The correction amount calculator 22 calculates a sine wave having a phase component based on the difference between the first elliptical long axis direction (deviation) θ _ζ ′ and the second elliptical long axis direction (deviation) θ _ζ2. A magnetic pole position correction amount Δθ _ζ is calculated based on the sine wave.

The magnetic pole position correction amount Δθ _ζ with respect to the difference between the first elliptical long axis direction (deviation) θ _ζ ′ and the second elliptical long axis direction (deviation) θ _ζ2 shown in FIG. Approximate by synthesis. This sine wave is
ΣK _i sin {2i (θ _ζ2 −θ _ζ ') + ψ _i }
It is represented by Then, the correction amount calculation unit 22 calculates a trigonometric function according to the above equation, combines the trigonometric functions, and calculates a magnetic pole position correction amount Δθ _ζ according to the characteristics shown in FIG.

Here, the correction amount calculation unit 22 may calculate a sine wave using the amplitude K _i and the phase offset ψ _i of the sine wave stored in advance for each operating point of the electric motor 9. Then, the correction amount calculation unit 22 calculates the magnetic pole position correction amount Δθ _ζ using the calculated sine wave. At this time, as K _i and ψ _i , optimal numerical values may be mapped according to the operating point of the electric motor 9, and may be calculated by switching or linear interpolation according to the operating point of the electric motor 9.

Further, the correction amount calculation unit 22 generates a plurality of sine waves having a phase component based on the difference between the first elliptical long axis direction (deviation) θ _ζ ′ and the first elliptical long axis direction (deviation) θ _ζ ′. It is possible to calculate and combine the orders (1 to i) and obtain the magnetic pole position correction amount Δθ _ζ based on the combined sine wave.

Thereby, according to the electric motor control apparatus shown as 2nd Embodiment, in addition to the effect shown in 1st Embodiment, the calculation for calculating _| requiring magnetic pole position correction amount (DELTA) (theta) _ζ by combining a sine wave can be performed. Thus, even if the first ellipse major axis direction (deviation) θ _ζ ′ is large, the first ellipse major axis direction (deviation) θ _ζ ′ is corrected with high accuracy and the magnetic pole position estimated value θ ^ is obtained. Can do. Further, it is possible to improve the estimation accuracy of the magnetic pole position and speed of the rotor while suppressing the amount of memory used.

Furthermore, according to this motor control device, the optimum values of the amplitude and phase offset of the sine wave for obtaining the magnetic pole position correction amount Δθ _ζ are stored for each operating point of the motor 9, thereby operating the motor 9. Even if the point changes, the first ellipse major axis direction (deviation) θ _ζ ′ can be accurately corrected.

[Third Embodiment]
Next, an electric motor control device shown as a third embodiment will be described. Note that parts similar to those in the above-described embodiment are denoted by the same reference numerals, and detailed description thereof is omitted.

  In the motor control device shown as the third embodiment, the magnetic pole position correction amount calculation unit 17 replaces the first bandpass filter 12 and the second bandpass filter 18 with each other as shown in FIG. It has.

The window function processing unit 23 inputs the γδ axis current detection values i _γ and i _δ , performs processing using a predetermined window function, sets a signal in a predetermined frequency section to “0”, and sets other frequency sections as The γδ-axis high-frequency currents i _γhw and i _δhw having a finite value are calculated and output. The first digital Fourier transform processing unit 13 and the second digital Fourier transform processing unit 19 are the same as the high-frequency voltages v _dh ^* and v _qh ^* by the γδ-axis high-frequency currents i _γhw and i _δhw supplied from the window function processing unit 23. The fundamental wave component and the harmonic component of the harmonic component are extracted.

  Thereby, in addition to the effect shown in 1st Embodiment, the motor control apparatus shown as 3rd Embodiment reduces arithmetic processing load compared with the case where the 1st band pass filter 12 and the 2nd band pass filter 18 are used. can do.

  The above-described embodiment is an example of the present invention. For this reason, the present invention is not limited to the above-described embodiment, and various modifications can be made depending on the design and the like as long as the technical idea according to the present invention is not deviated from this embodiment. Of course, it is possible to change.

DESCRIPTION OF SYMBOLS 1 Command value generation part 2 Current control part 3 High frequency voltage generation part 4 Coordinate conversion part 5 PWM conversion part 6 DC power supply 7 Inverter 8u, 8v Current sensor 9 Electric motor 9a Rotor 9b Stator 10 Coordinate conversion part 11 Magnetic pole position and rotation speed Estimator 12 First bandpass filter 13 First digital Fourier transform processor 14 First ellipse analyzer 15 PI amplifier 16 Integrator 17 Magnetic pole position correction amount calculator 18 Second bandpass filter 19 Second digital Fourier transform processor 20 Second ellipse analysis unit 21 Magnetic pole position correction amount map reference unit 22 Correction amount calculation unit 23 Window function processing unit

Claims (8)

Voltage command value calculating means for calculating the voltage command value of the excitation direction component and the voltage command value of the torque direction component of the motor as a voltage command value for controlling the motor to a desired state;
High-frequency voltage generating means for superimposing a high-frequency voltage of a frequency different from the drive frequency of the motor on the voltage command value calculated by the voltage command value calculating means;
Drive means for driving the electric motor by generating a drive signal for driving the switching element of the inverter based on the voltage command value on which the high-frequency voltage is superimposed by the high-frequency voltage generating means, and converting the DC voltage to the AC voltage by the inverter. When,
Current detecting means for detecting a current supplied from the driving means to the electric motor;
A rotor state estimating means for estimating a state of a rotor in the electric motor based on the current detected by the current detecting means;
Control means for controlling the drive means to drive the electric motor based on the rotor position estimated by the rotor state estimation means,
The rotor state estimation means includes
The excitation direction component current and torque direction component current of the fundamental wave component of the same frequency as the high frequency voltage included in the current detected by the current detection means are extracted, and the current detected by the current detection means includes the current Extract the excitation direction component current and torque direction component current of the harmonic component for the same frequency as the high frequency voltage,
A high frequency current vector locus is approximated by an ellipse using the extracted excitation direction component current and torque direction component current of the fundamental wave component, and a first ellipse major axis direction that is a deviation of the ellipse major axis direction from the excitation direction is obtained. The second ellipse length is a deviation of the ellipse major axis direction from the excitation direction by approximating the high frequency current vector locus by an ellipse using the extracted excitation direction component current and torque direction component current of the extracted harmonic component. Find the axial direction,
Based on the difference between the first ellipse major axis direction and the second ellipse major axis direction, a correction amount in the first ellipse major axis direction is calculated,
The motor control device according to claim 1, wherein the first elliptical long axis direction is corrected based on the correction amount, and the rotor position is estimated based on the corrected first elliptical long axis direction.   The motor control apparatus according to claim 1, wherein the rotor state estimation unit uses the harmonic component as a second harmonic component with respect to a frequency of the high-frequency voltage. Storage means for storing, as map data, a relationship between a difference between the first elliptical long axis direction and the second elliptical long axis direction and a correction amount in the first elliptical long axis direction;
The rotor state estimating means calculates a difference between the first ellipse major axis direction and the second ellipse major axis direction, and based on the map data stored in the storage means, the first ellipse major axis direction and A correction amount in the first elliptical long axis direction corresponding to a difference from the second elliptical long axis direction is calculated, and the first elliptical long axis direction is corrected based on the calculated correction amount. The motor control device according to claim 1 or 2. The storage means stores the map data for each operating point of the electric motor,
The motor control apparatus according to claim 3, wherein the rotor state estimation unit selects map data based on an operating point of the motor.   The rotor state estimation means calculates a sine wave having a phase component based on a difference between the first elliptical long axis direction and the second elliptical long axis direction, and based on the calculated sine wave, The motor control device according to claim 1, wherein a correction amount in one elliptical long axis direction is calculated. The rotor state estimation means calculates and synthesizes a sine wave having a phase component based on a difference between the first ellipse major axis direction and the second ellipse major axis direction for a plurality of orders. 6. The motor control device according to claim 5 , wherein a correction amount in the first elliptical long axis direction is calculated based on a sine wave.   The rotor state estimation means calculates a sine wave using the amplitude and phase offset of the sine wave stored in advance for each operating point of the electric motor, and uses the sine wave to calculate the first elliptical length. The motor control device according to claim 5, wherein a correction amount in the axial direction is calculated. As the voltage command value for controlling the electric motor to a desired state, the voltage command value of the excitation direction component and the voltage command value of the torque direction component of the electric motor are calculated, and the drive frequency of the electric motor with respect to the voltage command value is When superposing high frequency voltages of different frequencies and driving the electric motor by driving the switching element of the inverter based on the voltage command value on which the high frequency voltage is superimposed,
The excitation direction component current and the torque direction component current of the fundamental wave component having the same frequency as the high frequency voltage included in the current supplied to the motor are extracted and the same as the high frequency voltage included in the current supplied to the motor. Extract the excitation direction component current and torque direction component current of the harmonic component with respect to the frequency,
A high frequency current vector locus is approximated by an ellipse using the extracted excitation direction component current and torque direction component current of the fundamental wave component, and a first ellipse major axis direction that is a deviation of the ellipse major axis direction from the excitation direction is obtained. The second ellipse length is a deviation of the ellipse major axis direction from the excitation direction by approximating the high frequency current vector locus by an ellipse using the extracted excitation direction component current and torque direction component current of the extracted harmonic component. Find the axial direction,
Based on the difference between the first ellipse major axis direction and the second ellipse major axis direction, a correction amount in the first ellipse major axis direction is calculated,
An electric motor state estimation method, wherein the first elliptical long axis direction is corrected based on the correction amount, and the rotor position is estimated based on the corrected first elliptical long axis direction.

Images