JP5758140B2 - Motor control device and motor control method - Google Patents

Description

  The present invention relates to a motor control device and a motor control method.

  As an angle detection device that detects a rotation angle position of a rotor of a motor, an angle detection device that detects a rotation angle by combining a resolver and an angle detection circuit is used. The resolver is a rotation detection device that detects the rotation angle of the rotor based on the phase difference between the rotor coil and the stator coil. The resolver rotates by applying a two-phase sine wave signal (Esin ωt, Ecos ωt (E is the amplitude of the sine wave, ω is an angular velocity)) that is 90 degrees out of phase to two stator coils arranged at right angles to each other. Generates a magnetic field. In this magnetic field, voltages (Esin (ωt + θ), Ecos (ωt + θ)) having a phase difference between the two coils are generated in two rotor coils placed at right angles to each other. The rotation angle θ is detected using the output of the rotor coil.

However, when this resolver output is used as it is as angle detection information, torque fluctuations and speed fluctuations have occurred in the controlled motor due to errors due to the resolver and errors in the circuit that processes the detected signal.
Therefore, in the invention described in Patent Document 1, the resolver generates a pseudo position signal at a specific angle, calculates the angle detection error and the angle detection accuracy using the generated pseudo position signal, and based on the calculated result. An error in the detected angle value is corrected.

Japanese Patent No. 3590624

  However, in the invention described in Patent Document 1, since a pseudo position signal is generated and correction is performed using the generated pseudo position signal, detection errors at rotational positions other than a specific angle may not be reduced. There was a problem. In addition, when the other angles are corrected by interpolation, the fundamental wave component error and the second harmonic component error are limited.

  The present invention has been made in view of the above problems, and provides a motor control device and a motor control method capable of improving detection accuracy even for errors not corresponding to the detection angle of the resolver. It is an object.

In order to achieve the above object, a motor control device according to the present invention includes an angle detection unit that detects and outputs an angle detection signal having a point having a discontinuous differential coefficient, and a discontinuous point of the detected angle detection signal. A discontinuous point detecting unit for detecting, an angle detecting signal in the vicinity of the detected discontinuous point is replaced with an angle detecting signal of a ramp waveform before the discontinuous point and output; and A control signal calculation unit that compares the output signal or the replaced output signal with a reference value and calculates a control signal by linear approximation, and the replacement unit detects a current rotation angle from the angle detection signal. The detected angle obtained by adding 180 degrees to the rotation angle at the sampling time immediately before the current time is compared with the current rotation angle, and added to the rotation angle at the sampling time one time before the current time. When the detected angle is smaller than the current rotation angle, correction is performed by adding 360 degrees to the rotation angle at the sampling time for each of a predetermined number of two or more from the previous one, and the current Approximate interpolation is performed with respect to the rotation angle, and the predetermined number of corrected rotation angles, or a detection angle obtained by subtracting 180 degrees from the rotation angle at the sampling time immediately before the current time, and the current rotation angle, When the detected angle obtained by subtracting 180 degrees from the rotation angle at the sampling time one time before the current is larger than the current rotation angle, a predetermined number of two or more from the current time Each is corrected by subtracting 360 degrees from the rotation angle at the sampling time, and the current rotation angle and the corrected number of rotations are corrected. It is characterized by performing approximate interpolation with respect to the angle.

  ADVANTAGE OF THE INVENTION According to this invention, the angle detection apparatus and angle detection method which can improve a detection precision also with respect to the error which does not respond | correspond to a detection angle are realizable.

It is a control block diagram which shows the motor control apparatus of 1st Embodiment. It is a figure which shows schematic structure of the resolver which concerns on the same embodiment. It is a figure which shows the change of the detection angle with respect to time which concerns on the embodiment. It is a figure which shows schematic structure of the angle correction | amendment part 13 which concerns on the same embodiment. It is a figure explaining the sampling time which the memory | storage part 132 which concerns on the embodiment memorize | stores. It is a figure explaining an example of the information which shows the detection angle of the sampling time which the memory | storage part 132 which concerns on the embodiment memorize | stores. It is a figure explaining the example of an ideal detection signal without an error concerning the embodiment. It is a figure explaining the example of the detection signal with an error concerning the embodiment. It is a figure explaining the example calculated without replacing the discontinuous point of the detection angle which concerns on the same embodiment. It is a figure explaining the method of substitution of the discontinuous point of the detection angle concerning the embodiment. It is a figure explaining another method of substitution of a discontinuous point of detection angle concerning the embodiment concerning the embodiment. It is a figure explaining the sample number n used when performing the least squares method concerning the embodiment. It is a figure which shows an example of the actual value and calculated value of the detection angle which concern on the same embodiment. It is the figure which represented FIG. 13 by the difference actual value and calculated value of a detection angle. It is a figure which shows an example of the measured value and calculated value of the detection angle which concern on 2nd Embodiment. It is the figure which represented FIG. 15 by the difference actual value and calculated value of a detection angle.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In addition, this invention is not limited to the embodiment which concerns, A various change is possible within the range of the technical thought.

[First Embodiment]
The motor control device is a device that controls the motor by receiving power supplied from a battery cell, for example, in an industrial vehicle, an electric vehicle, a hybrid vehicle, a train, a ship, an airplane, a power generation system, or the like.

First, a schematic operation of the motor control device will be described.
In an electric vehicle powered by an electric motor and a hybrid vehicle powered by using an internal combustion engine and an electric motor together (hereinafter referred to as “electric vehicle etc.”), the motor control device has a three-phase structure in order to increase the power use efficiency. Pulse width modulation control (PWM (Pulse Width Modulation) control) is used to modulate the pulse width when controlling the drive current.
In an electric vehicle or the like, a permanent magnet synchronous motor is mainly used, and a three-phase current synchronized with the rotation is passed through the motor. In order to PWM control the three-phase current, an electric pulse having a constant frequency called a carrier signal is used. In this case, the drive current is supplied to the motor as a rectangular wave whose pulse width is modulated in accordance with the timing of the carrier signal, and becomes a three-phase current of a sine wave due to the inductance of the motor.
In such a motor control device, control is performed by PI (Proportional Integral) control using feedback so that the current flowing through the motor becomes the torque of the input torque command. In the PI control, the three phases supplied to the motor are coordinate-converted to the two-axis coordinates of the d-axis and q-axis, and the control is performed with the two phases of the d-axis and q-axis.

FIG. 1 is a control block diagram showing a motor control device of the present embodiment.
As shown in FIG. 1, the motor control apparatus 10 according to the present embodiment includes a resolver 30, an angle detection unit 12, an angle correction unit 13, a speed calculation unit 14, a current command unit 16, a three-phase / two-phase conversion unit 18, A current PI control unit 20, a two-phase / three-phase conversion unit 22, a duty calculation unit 24, a power conversion unit 26, and a current detector 28 are provided. Further, the motor control device 10 controls the motor M.

The motor M is a three-phase motor and is driven by a drive current output from the power conversion unit 26. A resolver 30 is attached to the motor M.
The angle detection unit 12 detects the rotation angle of the motor M at each moment based on the detection signal output from the resolver 30 at each sampling time (hereinafter, the detected rotation angle is referred to as a detection angle), and the detected detection angle is an angle. Output to the correction unit 13. The sampling frequency is, for example, 5 [KHz].
As will be described later, the angle correction unit 13 corrects the detected detection angle and calculates the angle measurement signal θ. The angle correction unit 13 outputs the calculated angle measurement signal θ to the speed calculation unit 14, the 3-phase / 2-phase conversion unit 18, and the 2-phase / 3-phase conversion unit 22.
The speed calculation unit 14 calculates the angular velocity ω of the rotor of the motor M from the angle measurement signal θ output from the angle correction unit 13 and outputs the calculated angular velocity ω to the current command unit 16.

The current detector 28 detects three-phase currents Iu, Iv, and Iw for the motor M, and outputs the detected three-phase currents Iu, Iv, and Iw to the three-phase / two-phase converter 18.
Three-phase currents Iu, Iv, and Iw output from the current detector 28 are input to the three-phase / two-phase converter 18. The three-phase / two-phase converter 18 converts the three-phase currents Iu, Iv, and Iw output from the current detector 28 into a two-phase d-axis component Id and a q-axis component Iq (hereinafter referred to as a detection current). . The three-phase / two-phase conversion unit 18 outputs the converted detection currents Id and Iq to the current PI control unit 20. The d-axis component current (d-axis current) is a component (excitation current) used to generate a magnetic flux in the motor M out of the flowing current when the d-axis is taken in the direction of the magnetic flux. Component). The q-axis component current (q-axis current) is a component corresponding to the torque of the load in the flowing current.
In the present specification, the command value and command signal are represented by a variable with “*” in the upper right.

In the case of an automobile, the current command unit 16 receives a torque command associated with the opening degree of the throttle pedal and the angular velocity ω output from the speed calculation unit 14. The torque command is for commanding torque to be generated by the motor. The current command unit 16 generates a two-phase command current (hereinafter referred to as command current) Id ^* and Iq ^* having a d-axis component and a q-axis component from the value of the torque command and the angular velocity ω. Current command unit 16 outputs generated command currents Id ^* and Iq ^* to current PI control unit 20. The current command unit 16 generates d-axis and q-axis command currents Id ^* and Iq ^* corresponding to the input torque command and the angular velocity ω.

The current PI control unit 20 receives command currents Id ^* and Iq ^* output from the current command unit 16 and detection currents Id and Iq output from the three-phase / two-phase conversion unit 18. Current PI control unit 20 controls currents Iu, Iv, and Iw flowing through motor M so that detection currents Id and Iq, which are control variables, have values corresponding to command currents Id ^* and Iq *.
Current PI control unit 20 calculates deviations ΔId and ΔIq by subtracting detection currents Id and Iq from input command currents Id ^* and Iq ^* , respectively. The current PI control unit 20 uses the calculated deviations ΔId and ΔIq to calculate the d-axis command voltage Vd ^* and the q-axis command voltage Vq ^* according to the following equations (1) to (2). The current PI control unit 20 outputs the calculated command voltages Vd ^* and Vq ^* to the 2-phase / 3-phase conversion unit 22.

Vd ^* = Kp × ΔId + Ki × ∫ (ΔId) dt (1)

Vq ^* = Kp × ΔIq + Ki × ∫ (ΔIq) dt (2)

  In equations (1) and (2), coefficients Kp and Ki are coefficients set in advance.

The 2-phase / 3-phase conversion unit 22 receives the command voltage Vd ^* and the command voltage Vq ^* output from the current PI control unit 20 and the angle measurement signal θ output from the angle correction unit 13. The two-phase / three-phase conversion unit 22 performs coordinate conversion of the input command voltage Vd ^* and the command voltage Vq ^* using the input angle measurement signal θ, and three-phase command voltages Vu ^* , Vv ^* , Vw. ^* Is calculated. The two-phase / 3-phase converter 22 outputs the calculated three-phase command voltages Vu ^* , Vv ^* , Vw ^* to the duty calculator 24.

The three-phase command voltages Vu ^* , Vv ^* , and Vw ^* output from the two-phase / three-phase converter 22 are input to the duty calculator 24. The duty calculator 24 calculates duty signals Du, Dv, and Dw representing drive current signals to be given to the motor from the three-phase command voltages Vu ^* , Vv ^* , and Vw ^* at a timing determined by the carrier frequency fc. The duty calculator 24 outputs the calculated duty signals Du, Dv, Dw to the power converter 26.

  The power converter 26 includes a power control element (power element) such as an IGBT (Insulated Gate Bipolar Transistor) element that performs switching for generating a drive current from the duty signals Du, Dv, and Dw. The power conversion unit 26 generates three-phase drive currents corresponding to the duty signals Du, Dv, and Dw output from the duty calculation unit 24 and supplies the generated three-phase drive currents to the motor M, respectively.

FIG. 2 is a diagram showing a schematic configuration of the resolver according to the present embodiment.
As shown in FIG. 2, the resolver 30 is mounted on the penetrating shaft of the motor M and adjusted according to the rotor magnetic field of the brushless motor. The resolver 30 includes a resolver rotor 32, a primary side coil (rotor) 34, and a primary side coil 34 and two secondary side coils (stator) 36 that are 90 degrees apart from each other. When an AC voltage is applied to the primary side, a voltage is also generated in the secondary coil. The amplitude of the voltage output to the secondary side is sin θ and cos θ, where θ is the rotor angle.
The angle detector 12 calculates the detected angle of the motor M based on the phase of the secondary coil 36. The calculated detection angle is monotonous because the inertia is large in an electric vehicle or the like between the rotation reference angle (0 degrees) and one rotation (360 degrees) of the motor M, and the acceleration is ignored compared to the sampling time. The value increases almost linearly. Accordingly, the calculated value over a plurality of rotations of the motor M becomes a sawtooth waveform as shown in FIG. The detected angle of the motor M can be detected from this calculated value. FIG. 3 is a diagram illustrating a change in the detection angle with respect to time according to the present embodiment. In FIG. 3, the horizontal axis represents time, and the vertical axis represents the detection angle of the motor M. As shown in FIG. 3, the detection angle signal is a sawtooth wave g101 period _{t L.} The detection angle is 0 degree to 360 degrees. The time is a sampling time based on the sampling frequency as will be described later. Then, as shown in FIG. 3, the detection angle increases from 0 degree to 360 degrees from time 0 to time t _L , and at time t _L , the detection angle changes from 360 degrees to 0 degrees. Time t _L is the period of the detection angle by the resolver 30.
For example, when the motor M has 8 poles, the detection period of the resolver is 4 times / rotation. For this reason, when the armature of the motor M is rotating at 1000 [rpm], the cycle is t _L = 15 [msec] (1000 × 4/60 = 67 [Hz]).

FIG. 4 is a diagram illustrating a schematic configuration of the angle correction unit 13 according to the present embodiment.
As shown in FIG. 4, the angle correction unit 13 includes a discontinuous point detection unit 131, a storage unit 132, a replacement unit 133, and a control signal calculation unit 134.

The discontinuous point detection unit 131 stores information indicating the detection angle output by the angle detection unit 12 in the storage unit 132 at each sampling time. The discontinuous point detection unit 131 detects a discontinuous point of the detection angle, as will be described later, based on the current detection angle and the detection angle at the previous sampling time. Note that the discontinuity of the detection angle, in FIG. 3, as in the time t _L, is that the value of the detection angle changes abruptly. That is, as the sawtooth wave (detection angle signal) g101, the period from time 0 to t _L, the detection angle increases at a gradient m, the moment the sampling exceeds the time t _L, the straight line is detected inclination angle m It falls off and suddenly decreases to near 0 degrees. As described above, the discontinuous point detection unit 131 sequentially compares the inclinations of the detection angle at the current sampling time and the detection angle at the previous sampling time in a predetermined sampling time period, and the inclination is not determined. A continuous sampling time is detected as a discontinuous point. That is, the discontinuous point detection unit 131 detects a point where the differential coefficient is discontinuous. Then, the discontinuous point detection unit 131 outputs the detection angle at the current sampling time and information indicating the detected discontinuity point to the replacement unit 133.

  As shown in FIGS. 5 and 6, the storage unit 132 stores the sampling time t and information indicating the detection angle y of the sampling time in association with each other. FIG. 5 is a diagram illustrating the sampling time stored in the storage unit 132 according to the present embodiment. FIG. 6 is a diagram illustrating an example of information indicating the detection angle of the sampling time stored in the storage unit 132 according to the present embodiment.

The replacement unit 133 stores information indicating a predetermined number n (n is a natural number of 1 or more) of detection angles from the previous sampling time based on the information indicating the discontinuous points output from the discontinuous point detection unit 131. Read from the unit 132. The predetermined number n is a number corresponding to the number of poles, acceleration, etc. of the motor M. Here, as an example, the predetermined number n is 16. The previous sampling time is t _n−1 when the current time is t _n in FIGS. 5 and 6, and therefore, the detection angle of the previous sampling time is y _n−1. It is. As described later, the replacement unit 133 detects a predetermined number n of detection angles from the previous sampling time so that the read information indicating the predetermined number of detection angles is continuous with the current sampling time detection angle. Replacement is performed for information indicating. The replacement unit 133 outputs the detection angle at the current sampling time and information indicating the replaced detection angle to the control signal calculation unit 134.

  As will be described later, the control signal calculation unit 134 uses, for example, the least square method using the detection angle of the current sampling time output from the replacement unit 133 and a predetermined number n of detection angles from the previous sampling time. Thus, the angle measurement signal θ at the current sampling time is calculated. The control signal calculation unit 134 outputs the calculated angle measurement signal θ to the speed calculation unit 14, the 3-phase / 2-phase conversion unit 18, and the 2-phase / 3-phase conversion unit 22.

FIG. 7 is a diagram for explaining an example of a signal with an ideal detection angle without error, and FIG. 8 is a diagram for explaining an example of a signal with a detection angle having an error. 7 and 8, the horizontal axis represents time (sampling time), and the vertical axis represents the detection angle. 7 and 8, reference numeral _(t n, _{y n)} is a combination of the detection angle _{y n} at the sampling time _{t n} (hereinafter, referred to as the detection value). 7 and 8, a curve g101 is a curve that passes through the detection values (t _n-5 , y _n-5 ) to (t _n , y _n ) when there is no error.
As shown in FIG. 7, when there is no error or almost no error in the detection angle, the detection values (t _n−5 , y _n−5 ) to (t _n , y _n ) are linear functions y = at + b. Can be expressed.
However, as shown in FIG. 8, the actually detected detection angles have errors, and thus vary, and detection values (t _n−5 , y _n−5 ) to (t _n , Some of y _n ) deviate from the curve g101. Therefore, if this detected angle is used as it is as the angle measurement signal θ, torque fluctuation and speed fluctuation have occurred in the motor M due to detection error.
For this reason, in this embodiment, the angle measurement signal θ is corrected by performing, for example, the least square method on the detected angle, and the detection error is suppressed.

When the detection angle y = at + b is set, the slope a and the y-intercept b of the linear approximation of the detection values (t ₁ , y ₁ ) to (t _n , y _n ) are expressed by the following equations (3) and (4). expressed.

Here, the time (t ₁ , t ₂ ,..., T _n ) and the detected angles (y ₁ , y ₂ ,..., Y _n ) are represented as current time (FIG. 5 and FIG. 6). Setting t _n ) to 0 makes it unnecessary to calculate the slope a.
As shown in FIG. 5, the time t ₁ is the (n-1) before sampling time, time t ₂ is the (n-2) before sampling time. In this way, the angle correction unit 13 assigns each sampling time to each time (t ₁ , t ₂ ,..., T _n−1 ), and assigns the current time to the time t _n . Further, as shown in FIG. 6, the detected angle y ₁ is the detection angle of the (n-1) before sampling, detection angle y ₂ is the detection angle of the (n-2) before sampling. Thus, the angle corrector 13, the detection angle _{(y 1, y 2, ···} , y n-1) Assign the detection angle of each sampling time, the detection angle of the current time to the detected angle _{y n} assign.
By assigning the current time (t _n ) to 0 in this way, since t = 0, the angle correction unit 13 calculates the current detection angle y _out by the following equation (5) by calculating only b that is a y-intercept. It can be calculated as follows.

Further, when the sampling frequency and the sampling number n are constant, the angle correction unit 13 can reduce the amount of calculation by performing the calculation related to the time (t ₁ , t ₂ ,..., T _n ) in advance. Further, when the sampling frequency or the sampling number n is changed, it can be coped with by updating the times t _{1 to} t _n in FIG.

  As shown in FIG. 3, the detection signal based on the detection angle of the resolver 30 is a sawtooth wave. Therefore, when the least square method is applied to this detection signal, as shown in FIG. 9, when the detection angle is switched from 360 degrees to 0 degrees (for example, the period of sampling time t = 17 to 19), the differential coefficient Is discontinuous, and is interpolated as shown by a curve g202. FIG. 9 is a diagram illustrating an example in which the least square method is applied to the angle detection signal according to the present embodiment. In FIG. 9, the horizontal axis represents sampling, and the vertical axis represents the detection angle. A curve g201 is a detected angle with respect to sampling. A curve g202 is a calculated angle obtained by calculating a detection angle with respect to the curve g201 by using the least square method for sampling.

In order to prevent a malfunction like the curve g202 shown in FIG. 9, in this embodiment, a discontinuous point of the differential coefficient is detected, and the detection angle is corrected based on the current detection angle.
A specific example will be described with reference to FIGS. As shown in FIG. 9, one cycle from the detection angle 0 degree to 360 degrees is set to 24 samplings. As shown in FIG. 9, description will be made using the period from the first sampling to the 19th sampling. The sampling number n used for performing the least squares method is 16 samples.
FIG. 10 is a diagram for explaining a detection angle correction method according to the present embodiment. In FIG. 10, the horizontal axis represents time, and the vertical axis represents the detection angle. Curve g301 is detected angle for the sampling time of the sampling time _{t n-15 ~t n-1} periods, curve g302 is the detection angle for the sampling time _{t n} subsequent sampling time. A curve g303 is a detection angle with respect to the sampling time in the period of sampling times tn _{-15 to} tn _-1 after the curve g301 is corrected so as to be continuous with the curve g302.

In FIG. 10, the detection angle at the sampling time t _n is 5 degrees, and the detection angle at the previous sampling time t _n−1 is 350 degrees. In this case, the detection angle _{y n-1 ~y n-15} of the sampling time _{t n-1 ~t n-15} is, in order to continuously and detection angle _{y n} of the sampling time _{t n,} the sampling time _{t n-1} ~ the subtracting detected angle _{y n-1 ~y n-15} from 360 ° t _n-15. The replacement unit 133 determines whether or not a value obtained by subtracting 180 degrees from the detection angle y _{n−1 of the} previous sampling is larger than the current detection angle. When the value obtained by subtracting 180 degrees from the detection angle y _n− 1 of the previous sampling is larger than the current detection angle, the replacement unit 133 detects the detection angle data (y ₁ , y ₂ ,. -Subtract 360 degrees from y _n-1 ).
As a result, the curve g301 is replaced with a curve g303, it is continuous with the sampling time _{t n} subsequent curve g302. As a result, there is no discontinuous point of the differential coefficient, and an appropriate detection angle can be obtained even when the control signal calculation unit 134 applies the least square method to the detected angle.

  Note that the above-described detection and replacement of discontinuous points is not a sinusoidal waveform, but a signal waveform having a discontinuous differential coefficient, for example, a sawtooth waveform in which a predetermined section is represented by a quadratic function instead of a section linear function. The same effect can be obtained with this signal. In other words, when a predetermined number n of detection angles are corrected by performing the least square method, a signal having a detection angle that causes a large error at a specific sampling time by performing the least square method is used. By using the motor control device of the embodiment, it is possible to detect the angle with high accuracy.

FIG. 11 is a diagram for explaining another detection angle correction method according to the present embodiment. In FIG. 11, the horizontal axis represents time, and the vertical axis represents the detection angle. As shown in FIG. 11, the waveform of the reverse sawtooth wave decreases as the sampling time increases during the sampling time t _{n−1 to} t _{n −15} . Curve g401 is detected angle for the sampling time of the sampling time _{t n-15 ~t n-1} periods, curve g402 is the detection angle for the sampling time _{t n} subsequent sampling time. Curve g403 is a detection angle with respect to the sampling time in the period of sampling time tn _{-15 to} tn _-1 after the curve g401 is corrected so as to be continuous with the curve g402.

11, the detection angle is 350 degrees sampling time _{t n,} the detected angle of the sampling time _{t n-1} before one is 5 degrees. In this case, the detection angle _{y n-1 ~y n-15} of the sampling time _{t n-1 ~t n-15} is, in order to continuously and detection angle _{y n} of the sampling time _{t n,} the sampling time _{t n-1} ~ adding 360 ° detection angle _{y n-1 ~y n-15} of t _n-15. The replacement unit 133 determines whether or not a value obtained by adding 180 degrees from the detection angle y _{n−1 of the} previous sampling is larger than the current detection angle. When the value obtained by adding 180 degrees to the detection angle of the previous sampling is smaller than the current detection angle, the replacement unit 133 detects the detection angle data other than the current (y ₁ , y ₂ ,..., Y _{n− 1} ) Add 360 degrees.
As a result, the curve g401 is replaced with a curve g403, it is continuous with the sampling time _{t n} subsequent curve g402. As a result, there is no discontinuous point of the differential coefficient, and an appropriate detection angle can be obtained even when the control signal calculation unit 134 applies the least square method to the detected angle.

  FIG. 12 is a diagram for explaining the number of samples n used when the least square method according to the present embodiment is performed. As shown in FIG. 12, when the control signal calculation unit 134 performs the least squares method, the calculation amount increases as the number of samples n used increases, and the influence of acceleration increases, but the error suppression of low frequencies increases. In addition, the smaller the number n of samples used, the smaller the amount of calculation and the smaller the influence of acceleration, but the lower the frequency error suppression. For this reason, you may make it select the sample number n according to uses, such as a vehicle etc. to which a motor control apparatus is applied.

FIG. 13 is a diagram illustrating an example of an actual measurement value and a calculated value of a detection angle according to the present embodiment. In FIG. 13, the horizontal axis represents sampling, the vertical axis represents detection angle, the curve g501 represents sampling vs. actually detected detection angle, and the curve g502 represents sampling vs. calculated detection angle. The detection angle signal in FIG. 13 is an example of a high-frequency error, for example.
As described above, when the motor control device 10 of the present embodiment is applied, the amplitude of the curve g502, that is, the amplitude due to a high frequency noise-like error is smaller than the curve g501. That is, the detection angle error is reduced. As a result, the torque fluctuation of the motor M is reduced. Further, when the motor control device 10 of the present embodiment is applied, the detection angle can be calculated with high accuracy even when the detection time is switched from 360 degrees to 0 degrees such that the sampling time t is around 40.

FIG. 14 is a diagram representing FIG. 13 by the actual measurement value and the calculated value of the detected angle. In FIG. 14, the horizontal axis represents sampling, the vertical axis represents the difference between the detected angles, the curve g601 represents the difference between the sampling and the actually measured detection angle, and the curve g602 represents the difference between the sampling and the calculated detection angle. FIG. 14 shows an example in which the error in the detected angle is large among the actually measured test data. The difference between the detected angles is the difference between the actually detected detection angle and its previous value (difference in detection angle), and the difference between the calculated angle and its previous value (difference in calculated angle). Further, the actual detection angle shown in FIG. 14 is a case of a high frequency error.
As shown in FIG. 14, the amplitude of the detected angle obtained by actually measuring the curve g601 is about −40 [deg] to +40 [deg]. On the other hand, in the motor control device 10 of the present embodiment, the amplitude of the detected angle is about +5 [deg] to +20 [deg] as indicated by the curve g602. As described above, the motor control device 10 of the present embodiment has an effect of suppressing the variation in the detection angle for each sampling.

As described above, in the present embodiment, the angle detection unit 12 detects the detection angle of the motor M based on the detection signal output from the resolver 30. Then, the discontinuous point detection unit 131 detects discontinuous points of the differential coefficient of the detection angle based on the current detection angle and the detection angle at the previous sampling time. Based on the detected discontinuity of the differential coefficient, the replacement unit 133 samples the previous sampling so that the predetermined number n of detection angles from the first sampling is continuous with the detection angle of the current sampling time. It is determined from the time whether or not the differential coefficient is discontinuous. Then, the replacement unit 133 replaces 360 degrees with respect to a predetermined number n of detection angles from the previous sampling time. Then, the control signal calculation unit 134 calculates the angle measurement signal θ by the least square method using the detected angle at the current sampling time and the predetermined number n of replaced detection angles.
As a result, the motor control device of the present embodiment can calculate the detection angle of the motor with high accuracy at any detection angle regardless of the rotation angle of the resolver. In addition, since it is possible to prevent an error caused by the resolver and an influence of a circuit that processes the detection signal of the resolver, it is not necessary to previously measure a correction value or create a map for each motor control device or resolver. Furthermore, the detection angle of the motor can be calculated with higher accuracy than when the error correction map for the detection angle of the resolver is used. In addition, the motor control device of this embodiment can suppress detection errors that fluctuate to high frequencies.

[Second Embodiment]
In the first embodiment, an example has been described in which an error in which the detection angle fluctuates due to a high frequency error is suppressed. In the present embodiment, a case will be described in which the motor control device of the present embodiment is used for sinusoidal errors (fundamental wave component errors, second harmonic component errors, etc.) corresponding to angles.

FIG. 15 is a diagram illustrating an example of an actual measurement value and a calculated value of a detection angle according to the present embodiment. In FIG. 15, the horizontal axis represents sampling, the vertical axis represents detection angle, the curve g701 represents sampling vs. actually detected detection angle, and the curve g702 represents sampling vs. calculated detection angle. Reference g710 is a sine wave error. The sinusoidal error is a fundamental wave component error or a second harmonic component error in which the detection angle error fluctuates in a sinusoidal manner corresponding to the angle, as indicated by reference numeral g710, during the sampling time 3 to 10. Refers to errors.
Thus, the error in the detection angle of the calculated value of the curve g702 is suppressed with respect to the error in the detection angle of the actual measurement value of the curve g701. Thus, even when there is a sinusoidal error corresponding to the angle, the detection error can be suppressed by using the motor control device 10 of the present embodiment. The configuration of the motor control device 10 and the detection angle processing procedure are the same as those in the first embodiment.

FIG. 16 is a diagram in which FIG. 15 is represented by a difference measurement value and a calculation value of the detection angle. In FIG. 16, the horizontal axis represents sampling, the vertical axis represents the difference between the detected angles, the curve g 801 represents the difference between the sampling and the actually measured detection angle, and the curve g 802 represents the difference between the sampling and the calculated detection angle. The difference between the detected angles is a difference between the actually detected angle and its previous value (detected angle difference), and a difference between the calculated angle and its previous value (calculated angle difference). The actually detected detection angle shown in FIG. 16 is a case of a second harmonic component error corresponding to the angle.
As shown in FIG. 16, the amplitude of the detected angle obtained by actually measuring the curve g801 is about +10 [deg] to +50 [deg]. On the other hand, in the motor control device 10 of the present embodiment, the amplitude of the detected angle is about +25 [deg] to +35 [deg] as indicated by the curve g802. Thus, in the motor control device 10 of the present embodiment, even in the case of a sinusoidal error corresponding to the angle (basic wave component error, second harmonic component error, etc.), the effect of suppressing variation in the detected angle for each sampling. There is.

  As described above, the motor control device according to the present embodiment accurately calculates the detected angle of the motor at any detected angle regardless of the rotation angle of the resolver even when there is a sinusoidal error corresponding to the angle. be able to. Furthermore, the detection angle of the motor can be calculated with higher accuracy than when the error correction map for the detection angle of the resolver is used. In addition, the motor control device of this embodiment can suppress high-frequency detection errors.

  In the present embodiment, the example in which the detection angle is calculated by the least square method of the linear function has been described. For example, when considering the acceleration due to the rotation of the motor M, a least square method of a quadratic function may be used. The sampling n is determined by the designer of the motor control device 10 according to the calculation power of the angle correction unit 13 of the motor control device 10, the acceleration of the motor M, the frequency of the error signal with respect to the detected angle (high frequency noise-like error), and the like. May be selected and set in advance. Alternatively, for example, the control signal calculation unit 134 detects the frequency of the error signal with respect to the detected angle (high-frequency noise-like error), and performs sampling n according to the frequency of the error signal with respect to the detected detection angle (high-frequency noise-like error). May be selected.

  In the present embodiment, the correction of the detection angle signal detected by the resolver attached to the motor M has been described as an example. However, if the detection signal based on the detection angle includes a discontinuous point of the differential coefficient, Angle detection may be performed by means other than the resolver.

  In addition, in order to implement | achieve the function of each part of FIG. 4 of embodiment, it is also possible to perform by the program preserve | saved at ROM etc. connected to CPU of the computer system. Alternatively, it may be realized by hardware using a PLD (programmable logic device), an ASIC (Application Specific Integrated Circuit), or a circuit.

DESCRIPTION OF SYMBOLS 10 ... Motor control apparatus M ... Motor 12 ... Angle detection part 13 ... Angle correction part 14 ... Speed calculation part 16 ... Current command part 18 ... Three phase / two phase conversion 20 ... Current PI control unit 22 ... 2-phase / 3-phase conversion unit 24 ... Duty calculation unit 26 ... Power conversion unit 28 ... Current detector 30 ... Resolver 32 ... Resolver rotor 34 ... Primary coil (rotor)
36 ... Secondary coil (stator)
131: Discontinuous point detection unit 132 ... Storage unit 133 ... Replacement unit 134 ... Control signal calculation unit

Claims (6)

An angle detection unit that detects and outputs an angle detection signal having a point where the differential coefficient is discontinuous;
A discontinuous point detection unit for detecting a discontinuous point of the detected angle detection signal;
A replacement unit that outputs an angle detection signal in the vicinity of the detected discontinuous point by replacing it with an angle detection signal of a ramp waveform before the discontinuous point; and
A control signal calculation unit that compares the output signal of the angle detection unit or the replaced output signal with a reference value and calculates a control signal by linear approximation;
Equipped with a,
The replacement part is:
A current rotation angle is detected from the angle detection signal;
The detected angle obtained by adding 180 degrees to the rotation angle at the sampling time one time before the current is compared with the current rotation angle,
When the detected angle obtained by adding 180 degrees to the rotation angle at the sampling time one time before the current is smaller than the current rotation angle, the sampling time for each of a predetermined number of two or more from the current one Correct by adding 360 degrees to the rotation angle at
Approximate interpolation is performed on the current rotation angle and the corrected predetermined number of rotation angles,
Or
Compare the detected angle obtained by subtracting 180 degrees from the rotation angle at the sampling time immediately before the current and the current rotation angle,
When the detected angle obtained by subtracting 180 degrees from the rotation angle at the sampling time one previous from the current is larger than the current rotation angle, the sampling time for each of a predetermined number of two or more from the previous one Correct by subtracting 360 degrees from the rotation angle at
An approximate interpolation is performed on the current rotation angle and the corrected predetermined number of rotation angles . The replacement part is:
2. The method according to claim 1, wherein, based on an angle detection value based on the angle detection signal, an angle of n periods (n is an integer of 1 or more) is added to the output signal of the angle detection unit for replacement. The motor control apparatus described. The angle detector
The motor control device according to claim 1 or 2, wherein a sampling interval for performing the angle detection is varied according to a rotation speed calculated from the angle detection value. The control signal calculator is
The motor control device according to any one of claims 1 to 3 , wherein the control signal is calculated using a least square method. The angle detection signal having a point where the derivative is discontinuous is
The motor control device according to any one of claims 1 to 4 , wherein the motor control device has a sawtooth waveform. An angle detection step in which the angle detection unit detects and outputs an angle detection signal having a point where the differential coefficient is discontinuous; and
A discontinuous point detecting unit detects a discontinuous point of the detected angle detection signal,
A replacement step of replacing the angle detection signal in the vicinity of the detected discontinuous point with the angle detection signal of the ramp waveform before the discontinuous point and outputting the detected angle detection signal;
A control signal calculation step in which a control signal calculation unit compares the output signal of the angle detection unit or the replaced output signal with a reference value and calculates a control signal by linear approximation;
With
The replacement step includes
Detecting a current rotation angle from the angle detection signal;
Comparing the detected angle obtained by adding 180 degrees to the rotation angle at the sampling time immediately before the current time and the current rotation angle;
When the detected angle obtained by adding 180 degrees to the rotation angle at the sampling time one time before the current is smaller than the current rotation angle, the sampling time for each of a predetermined number of two or more from the current one Correcting by adding 360 degrees to the rotation angle at
Performing approximate interpolation on the current rotation angle and the corrected predetermined number of rotation angles;
Or
Comparing the detected angle obtained by subtracting 180 degrees from the rotation angle at the sampling time immediately before the current time and the current rotation angle;
When the detected angle obtained by subtracting 180 degrees from the rotation angle at the sampling time one previous from the current is larger than the current rotation angle, the sampling time for each of a predetermined number of two or more from the previous one Correcting by subtracting 360 degrees from the rotation angle at
Performing approximate interpolation on the current rotation angle and the corrected predetermined number of rotation angles;
A motor control method comprising:

Images