JP2005160199A - Apparatus and method for controlling three-phase ac motor - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an apparatus for controlling a three-phase AC motor in which the stability of a control can be held by holding a designed current responding performance in a various q-axis current value region by solving the problems of deteriorating a torque response without obtaining the designed current response performance since an actual proportional gain deviates in the direction of lower than the designed value because the estimated value of a q-axis inductance Lq in a transient region is deviated when an instruction value of the q-axis current is changed stepwise. <P>SOLUTION: The apparatus for controlling the three-phase AC motor estimates the q-q-axis real current estimated value from the q-axis current instruction value when the voltage instruction value is calculated by a PI control from the difference between the current instruction value and the real current value, sets a q-axis proportional gain in the PI control based on the estimated q-axis real current estimated value, and calculates the q-axis voltage instruction value based on the set q-axis proportional gain. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a control device and a control method for a three-phase AC motor.

  As a conventional control device, there is one described in Patent Document 1 below. In this conventional example, in the control of a three-phase AC motor (hereinafter, the motor is referred to as a motor), in order to obtain a stable current response even when the motor parameter varies with simple control, current PI control (PI In the control, the q-axis proportional gain of the control (proportional integral control) is determined by table lookup.

Japanese Patent Laid-Open No. 2003-70280

In the prior art as described above, the proportional gain according to the dq-axis current command value is tabled to determine the proportional gain of the current PI control, so the actual current value of the dq-axis and the dq-axis When the current command value deviates, for example, when the actual current value in the transient region deviates from the command value with respect to the STEP command of the dq-axis current, the q-axis inductance Lq value fluctuates depending on the q-axis actual current Iq value. The proportional gain calculated by multiplying a constant (see FIG. 4 to be described later) deviates from the designed value. Therefore, the following problems arise.
That is, when the q-axis current command value changes stepwise, the estimated value of the q-axis inductance Lq in the transient region is shifted, and the actual proportional gain deviates in a direction lower than the design value. There was a problem that performance was not obtained and torque response was deteriorated.
The present invention has been made to solve the above-described problems, and is a three-phase AC motor that can maintain designed current response performance and control stability in a range of various q-axis current values. It is an object to provide a control device and a control method.

  In order to achieve the above object, in the present invention, when the voltage command value is calculated by PI control from the difference between the current command value and the actual current value, the q-axis actual current estimated value is estimated from the q-axis current command value. Then, the q-axis proportional gain in the PI control is set based on the estimated q-axis actual current estimated value, and the q-axis voltage command value is calculated based on the set q-axis proportional gain.

  With the above configuration, in the present invention, designed current response performance can be maintained and control stability can be maintained in various q-axis current value regions. That is, the actual current value has a slight fluctuation even at the steady state, but the current command value is constant at the steady state. Therefore, by estimating the q-axis actual current estimated value from the q-axis current command value and setting the q-axis proportional gain in PI control based on the estimated q-axis actual current estimated value, the current value that is the basis of the proportional gain setting Is stable, and stable proportional gain setting is possible. Further, by setting the estimated value from the q-axis current command value, it is possible to set a proportional gain based on the actual current value even in a region where the command value and the actual value deviate such as when a STEP command is input. There is.

Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings.
FIG. 1 is a diagram showing an embodiment of a current feedback control block for performing vector control of a three-phase AC motor to which the present invention is applied. In addition, although a present Example demonstrates as a thing which applied the three-phase alternating current motor to the drive motor of an electric vehicle, it is not restricted to this.
In an electric vehicle, an accelerator opening is detected and output using an accelerator opening sensor (not shown). A torque command value calculation unit (not shown) calculates a torque command value T ^* from the accelerator opening detected by the accelerator opening sensor, and outputs the torque command value T ^* as a digital signal.

The current command calculation unit 1 outputs a d-axis current command value Id ^* and a q-axis current command value Iq ^* corresponding to the torque command value T ^* . These current command values are input to the current PI control unit 2.
The current PI control unit 2 performs a proportional-integral calculation based on the deviation between the d-axis current command value Id ^* and the d-axis actual current value Id and outputs a d-axis voltage command value Vd ^* . Similarly, the q-axis current command value The q-axis voltage command value Vq ^* is output based on the deviation between Iq ^* and the q-axis actual current value Iq.

The d-axis voltage command value Vd ^* and the q-axis voltage command value Vq ^* (hereinafter referred to as d-axis q-axis voltage command value when both are collectively) are subjected to non-interference calculation processing as necessary. The two-phase / three-phase converter 3 converts the voltage into the three-phase voltage command values Vu ^* , Vv ^* , Vw ^* .
The above three-phase voltage command values Vu ^* , Vv ^* , Vw ^* are given to the PWM converter 4 and converted into PWM signals.
The inverter 5 converts the power of a DC power source (battery or the like) (not shown) into three-phase AC power in response to the PWM signal, and drives a three-phase motor 6 (hereinafter, the motor is abbreviated as a motor).

  The three-phase currents iu, iv and iw flowing at this time are detected by current sensors 7-1, 7-2 and 7-3, respectively, and the current values Iu, Iv, Convert to Iw. Then, the three-phase two-phase converter 10 converts it into a d-axis actual current value Id and a q-axis actual current value Iq, and feeds back to the current PI control unit 2. In the case of a three-phase alternating current, there is a relationship of iu + iv + iw = 0, so if any two-phase current is detected, the other one phase can be obtained by calculation. Therefore, the current sensors 7-1, 7-2, and 7-3 may have two phases.

  The rotation angle detector 8 detects the rotation angle θ (rotor: rotation angle of the rotor) of the three-phase motor 6. The rotation angle θ is used for coordinate conversion calculation in the two-phase three-phase converter 3 and the three-phase two-phase converter 10. In addition, the rotation number calculation unit 11 calculates the rotation number (rotation speed) of the three-phase motor 6 from the rotation angle θ, and the rotation number is used for calculation in the current command calculation unit 1.

  The above processing is repeated to perform vector control by current feedback of the three-phase motor 6. Note that the current command calculation unit 1, the current PI control unit 2, the two-phase three-phase converter 3, the PWM conversion unit 4, the three-phase two-phase converter 10 and the rotation number calculation unit 11 are electronic devices using, for example, a computer. It can be configured by a circuit.

The present invention relates to the configuration of the current PI control unit 2 in FIG.
FIG. 2 is a block diagram of the current PI control unit as the gist of the present invention.
In FIG. 2, PI calculation is performed for each of the d-axis and q-axis as follows.
First, in the case of the d-axis, as shown in the upper half of FIG. 2, the current deviation err_Id is calculated by subtracting the d-axis actual current value Id from the d-axis current command value Id ^* . Next, the value obtained by multiplying the current deviation err_Id by the proportional gain Kpd and the value obtained by multiplying the current deviation err_Id by the integral gain Kid and adding them are added and output as a d-axis voltage command value Vd ^* .

On the other hand, for the q-axis shown in the lower half of FIG. 2, the q-axis current estimator 21 estimates the q-axis current estimated value Iq ″ from the q-axis current command value Iq ^* (details will be described later).
Next, a q-axis inductance Lq_pi corresponding to the q-axis current estimated value Iq ″ is obtained from an Iq-Lq map 22 (details will be described later) showing the relationship between Iq and Lq obtained in advance through experiments or the like. Thus, the q-axis proportional gain Kpq is calculated by multiplying the q-axis inductance Lq_pi by an arbitrary constant ωc (details will be described later).

Kpq = ωc × Lq_pi
After determining the q-axis proportional gain Kpq in this way, the same calculation as in the case of the d-axis is performed. That is, the q-axis current deviation err_Iq is calculated by subtracting the q-axis actual current value Iq from the q-axis current command value Iq ^* . Next, the multiplier 24 adds the value obtained by multiplying the q-axis current deviation err_Iq by the calculated q-axis proportional gain Kpq and the value obtained by multiplying the q-axis current deviation err_Iq by the integral gain Kid and adding the value. And output as q-axis voltage command value Vq ^* .

Next, the constant ωc, the q-axis current estimator 21 and the Iq-Lq map 22 in FIG. 2 will be described in detail.
FIG. 3 is a diagram showing a current control model when a three-phase motor is placed with a two-phase motor model.
When the above control model is represented by a transfer function, the following (Equation 1) to (Equation 4) are obtained.

Id = Vd / (Ra + Ld × s) (Equation 1)
Vd = {Kpd + Kid / s} × (Id ^* −Id) (Equation 2)
Iq = Vq / (Ra + Lq × s) (Equation 3)
Vq = {Kpq + Kiq / s} × (Iq ^* −Iq) (Expression 4)
However,
Id: d-axis current value Kpd: d-axis proportional gain Kpq: q-axis proportional gain Vd: d-axis voltage value Kid: d-axis integral gain Kiq: q-axis integral gain Iq: q-axis current value Ld: d-axis inductance Lq: q Axis inductance Vq: q-axis voltage value Ra: Winding resistance s: Laplace operator On the d-axis, (Equation 1), (Equation 2),
Id = {Kpd + Kid / s} × (Id ^* −Id) / (Ra + Ld × s) (Equation 5)
Id × (Ra + Ld × s) = {Kpd + Kid / s} × (Id ^* −Id) (Equation 6)
Id × [(Ra + Ld × s) + {Kpd + Kid / s}]
= {Kpd + Kid / s} × Id ^* (Expression 7)
Id / Id ^* = {Kpd + Kid / s}
/ [(Ra + Ld × s) + {Kpd + Kid / s}] (Equation 8)
Id / Id ^* = 1 / [(Ra + Ld × s) / {Kpd + Kid / s} +1] (Equation 9)
Similarly, on the q axis, from (Equation 3) and (Equation 4),
Iq = {Kpq + Kiq / s} × (Iq ^* −Iq) / (Ra + Lq × s) (Equation 10)
Iq × (Ra + Lq × s) = {Kpq + Kiq / s} × (Iq ^* −Iq) (Equation 11)
Iq × [(Ra + Lq × s) + {Kpq + Kiq / s}]
= {KPq + Kiq / s} × Iq ^* (Equation 12)
Iq / Iq ^* = {Kpq + Kiq / s}
/ [(Ra + Lq × s) + {Kpq + Kiq / s}] (Equation 13)
Iq / Iq ^* = 1 / [(Ra + Lq × s) / {Kpq + Kiq / s} +1] (Expression 14)
Here, the PI control gain of each axis is determined using an arbitrary constant ωc. D-axis proportional gain: Kpd = ωc × Ld_pi (A) d-axis integral gain: Kid = ωc × Ra_pi (B) q-axis proportional gain : Kpq = ωc × Lq_pi (C) Equation q-axis integral gain: Kiq = ωc × Ra_pi (D) Equation (D9) is expressed as follows.

Id / Id ^* = 1 / [(Ra + Ld × s)
/ {Ωc × Ld_pi + ωc × Ra_pi / s} +1] (Equation 15)
Id / Id ^* = 1 / [(Ra + Ld × s)
/ {Ωc × (Ld_pi + Ra_pi / s)} + 1] (Equation 16)
Id / Id ^* = 1 / [s × (Ra + Ld × s)
/ {Ωc × (Ld_pi × s + Ra_pi)} + 1] (Equation 17)
Similarly, the equation (14) is as follows.

Iq / Iq ^* = 1 / [(Ra + Lq × s)
/ {Ωc × Lq_pi + ωc × Ra_pi / s} +1] (Equation 18)
Iq / Iq ^* = 1 / [(Ra + Lq × s)
/ {Ωc × (Lq_pi + Ra_pi / s)} + 1] (Equation 19)
Iq / Iq ^* = 1 / [s × [(Ra + Lq × s)]
/ {Ωc × (Lq_pi × s + Ra_pi)} + 1] (Equation 20)
Therefore, Ld_pi and Ra_pi used to determine the d-axis PI control constant, and Lq_pi and Ra_pi used to determine the q-axis PI control constant are expressed as Ld_pi = Ld, respectively.
Lq_pi = Lq
Ra_pi = Ra
Then, the equation (17) becomes as follows.

Id / Id ^* = 1 / [s × (Ra + Ld × s)
/ {Ωc × (Ld × s + Ra)} + 1] (Expression 21)
Id / Id ^* = 1 / [s × (1) / {ωc} +1]
= 1 / [s × (1 / ωc) +1] (Equation 22)
Similarly, the equation (20) is as follows.

Iq / Iq ^* = 1 / [s × (Ra + Lq × s)
/ {Ωc × (Lq × s + Ra)} + 1] (Equation 23)
Iq / Iq ^* = 1 / [s × (1) / {ωc} +1]
= 1 / [s × (1 / ωc) +1] (Equation 24)
As described above, it can be seen that the transfer functions (Equation 22) and (Equation 24) of each of the d-axis and the q-axis become the first-order lag response of the time constant τ = (1 / ωc).

From the above, the values that match the motor parameters in the equations (A), (B), (C), and (D) that determine the PI control gain, that is, Ld_pi = Ld
Lq_Pi = Lq
Ra_Pi = Ra
Is input and the gain is set, the current response of each axis can be controlled to the time constant of the reciprocal of the arbitrary variable ωc.
Therefore, the constant ωc in FIG. 2 represents the reciprocal (ωc = 1 / τ) of the current response time constant design value τ.

Next, the q-axis current estimation unit 21 will be described in detail.
By matching the motor parameters as described above, the current response can be obtained with a first-order lag response corresponding to the time constant design value. Therefore, the q-axis current estimator 21 sets the q-axis current command value Iq ^* to 1 of the time constant τ. A q-axis current estimated value is calculated by performing a calculation of the next filter (for example, the following Expression 25). Any filter calculation other than the filter calculation described here may be used as long as it exhibits a first-order delay.

Y = (e− ^{t / τ} ) × Y _z + (1−e− ^{t / τ} ) × U _z (Equation 25)
Where Y: q-axis current estimated value
Y _z : q-axis current estimated value in the previous calculation
U _z : q-axis current command value in the previous calculation
t: Control cycle
τ: Time constant Next, the Iq-Lq map will be described in detail.
Since the Iq-Lq relationship of the three-phase AC motor normally has the Iq-Lq characteristic shown in FIG. 4, the Lq value is derived based on the characteristic of FIG. 4 in the Iq-Lq map.
Among the motor parameters Ld_pi = Ld Lq_Pi = Lq Ra_Pi = Ra described above, Lq-Iq shows the characteristics shown in FIG. 4, whereas Ld is less current dependent on Id than Lq. There is not much dependency on Ra. Therefore, although a method for calculating only Lq has been described in the present embodiment, each parameter may be corrected based on, for example, the motor temperature in order to improve performance.

Next, FIGS. 5 and 6 are flowcharts showing processing contents when the present invention is applied to the current PI control shown in FIG. This calculation is repeated at a constant cycle.
First, in step 1 of FIG. 5, the d-axis current command value Id ^* , the d-axis actual current value Id, the q-axis current command value Iq ^* , and the q-axis actual current value Iq are read, and the process proceeds to step 2.
In step 2, the d-axis actual current value Id is subtracted from the d-axis current command value Id ^* fetched in step 1 to calculate a d-axis current deviation err_Id, and the process proceeds to step 3.
In step 3, the d-axis proportional error Kerr is multiplied by the d-axis current deviation err_Id calculated in step 2 to calculate a d-axis proportional term, and the process proceeds to step 4.

In step 4, the d-axis current deviation err_Id calculated in step 2 is multiplied by the d-axis integral gain Kid, an integration operation is performed to calculate a d-axis integral term, and the process proceeds to step 5.
In step 5, the d-axis proportional term calculated in step 3 and the d-axis integral term calculated in step 4 are added to calculate a d-axis voltage command value Vd ^*, and the process proceeds to step 6.
In step 6, the q-axis current deviation err_Iq is calculated by subtracting the q-axis actual current value Iq from the q-axis current command value Iq ^* fetched in step 1.

Step 6 in FIG. 5 continues to step 7 in FIG.
In step 7 of FIG. 6, the q-axis current command value Iq ^* fetched in step 1, the previous q-axis current command value Iq ^* _z saved at the previous calculation, and the previous q-axis current estimated value Iq ″ _z and previously stored in memory. The q-axis current estimated value Iq ″ is calculated from the following (Equation 26) using the filter constants α (= e ^{−t / τ} ) and β (= 1−e ^{−t / τ} ), and the process proceeds to Step 8. Transition.
Iq ″ = [α × Iq ″ _z + β × Iq ^* _z ] (Equation 26)
In step 8, the previous value is stored as the previous q-axis current estimated value Iq " _z = Iq" and the previous q-axis current command value Iq ^* _z = Iq ^* for the next calculation, and the process proceeds to step 9.

In step 9, the Lq value corresponding to the q-axis current estimated value Iq ″ calculated in step 8 is calculated from the Iq-Lq map of FIG. 4, and the process proceeds to step 10.
In step 10, the Lq value calculated in step 9 is multiplied by the reciprocal ωc of the filter time constant stored in the memory in advance to calculate the q-axis proportional term gain Kpq, and the process proceeds to step 11.
In step 11, the d-axis current deviation err_Iq calculated in step 6 is multiplied by the q-axis proportional term gain Kpq calculated in step 10 to calculate the q-axis proportional term, and the process proceeds to step 12.

In step 12, the q-axis current deviation err_Iq calculated in step 6 is multiplied by a q-axis integral gain Kiq, an integration operation is performed to calculate a q-axis integral term, and the process proceeds to step 13.
In step 13, the q-axis proportional term calculated in step 11 and the d-axis integral term calculated in step 12 are added to calculate a q-axis voltage command value Vq ^* , and the process ends.

  In steps 9 to 11 in FIG. 6, Lq is mapped from the q-axis current estimated value and Kpq is calculated by multiplying Lq by ωc. However, Kpq obtained by multiplying Lq by ωc in advance is mapped. And Kpq can be mapped from the q-axis current estimated value.

As described above, in the present invention, when the voltage command value is calculated by PI control from the difference between the current command value and the actual current value, the q-axis actual current estimated value is estimated from the q-axis current command value, and estimated. Since the q-axis proportional gain in PI control is set based on the q-axis actual current estimated value and the q-axis voltage command value is calculated based on the set q-axis proportional gain, various q-axis The designed current response performance can be maintained in the current value region, and the control stability can be maintained. That is, the actual current value has a slight fluctuation even at the steady state, but the command value is constant at the steady state. Therefore, by estimating the q-axis actual current estimated value from the q-axis current command value and setting the q-axis proportional gain in PI control based on the estimated q-axis actual current estimated value, the current value that is the basis of the proportional gain setting Is stable, and stable proportional gain setting is possible. Further, by using the estimated value from the q-axis current command value, it is possible to set a proportional gain based on the actual current value even in a region where the command value and the actual value deviate such as when a STEP command is input.
Further, by using a simple estimation method of estimating the q-axis current estimated value with the first-order lag of the q-axis current command value, it is based on the estimated value from the q-axis current command value without increasing the calculation load. Proportional gain can be set.
In addition, there are effects such as being able to obtain a current response with the designed response time constant ωc in various q-axis current value regions.

1 is a diagram showing an embodiment of a current feedback control block that performs vector control of a three-phase AC motor to which the present invention is applied. The block diagram of the electric current PI control part of this invention. The figure which shows the electric current control model at the time of setting a three-phase motor with a two-phase motor model. The characteristic view which shows the relationship of Iq-Lq of a three-phase alternating current motor. FIG. 3 is a part of a flowchart showing processing contents when the present invention is applied to the current PI control shown in FIG. 2. The other part of the flowchart which shows the processing content at the time of applying this invention to the current PI control shown in FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Current command calculating part 2 ... Current PI control part 3 ... Two-phase three-phase converter 4 ... PWM conversion part 5 ... Inverter 6 ... Three-phase motor 7-1, 7-2, 7-3 ... Current sensor 8 ... Rotation Angle detector 9 ... A / D converter 10 ... 3-phase 2-phase converter 11 ... rotational speed calculator

Claims (10)

Current detection means for detecting an actual current that flows through each phase winding of the three-phase motor;
Current conversion means for converting the actual current detected by the current detection means into a d-axis actual current value and a q-axis actual current value;
current command value calculating means for calculating a d-axis current command value and a q-axis current command value;
PI control based on the d-axis current command value and the q-axis current command value calculated by the current command value calculation means, and the d-axis actual current value and the q-axis actual current value converted by the current conversion means Voltage command value calculating means for calculating and calculating a d-axis voltage command value and a q-axis voltage command value;
Motor control means for controlling the motor by controlling the voltage applied to each phase winding of the three-phase motor based on the d-axis voltage command value and the q-axis voltage command value calculated by the voltage command value calculation means; In the control device of the three-phase motor provided,
Q-axis current estimating means for calculating a q-axis current estimated value that is an estimated value of the q-axis actual current value corresponding to the q-axis current command value is provided, and the voltage command value calculating means is proportional to the q-axis in the PI control calculation. Q-axis proportional gain setting means for setting a gain based on the q-axis current estimated value, and calculating a q-axis voltage command value based on at least the q-axis proportional gain set by the q-axis proportional gain setting means 3 Control device for phase AC motor.   2. The control device for a three-phase AC motor according to claim 1, wherein the q-axis current estimated value is a q-axis current estimated value corresponding to a q-axis current command value at the previous PI control calculation.   3. The control device for a three-phase AC motor according to claim 1, wherein the q-axis current estimating unit estimates the q-axis current estimated value as a first-order lag of a q-axis current command value.   The q-axis proportional gain setting means includes q-axis inductance estimating means for estimating q-axis inductance based on a q-axis current estimated value, and the q-axis proportional gain is estimated based on the q-axis inductance estimated by the q-axis inductance estimating means. The control device for a three-phase AC motor according to any one of claims 1 to 3, wherein a gain is set.   The q-axis proportional gain setting unit calculates and sets the q-axis proportional gain by multiplying the q-axis inductance estimated by the q-axis inductance estimation unit by an inverse of a current response time constant. The control device for a three-phase AC motor according to claim 4. A process of detecting an actual current that is a current flowing through each phase winding of the three-phase motor;
A process of converting the detected actual current into a d-axis actual current value and a q-axis actual current value;
processing for calculating a d-axis current command value and a q-axis current command value;
PI control calculation is performed based on the d-axis current command value and the q-axis current command value, and the d-axis actual current value and the q-axis actual current value to calculate the d-axis voltage command value and the q-axis voltage command value. Processing to
A method for controlling a motor by controlling a voltage applied to each phase winding of the three-phase motor based on the d-axis voltage command value and the q-axis voltage command value,
A process of calculating a q-axis current estimated value that is an estimated value of a q-axis actual current value corresponding to the q-axis current command value;
A process of setting a q-axis proportional gain in the PI control calculation based on the q-axis current estimated value;
And a control method for a three-phase AC motor that calculates a q-axis voltage command value based on at least the set q-axis proportional gain.   The method for controlling a three-phase AC motor according to claim 6, wherein the q-axis current estimated value is a q-axis current estimated value corresponding to a q-axis current command value at the time of the previous PI control calculation.   The method for controlling a three-phase AC motor according to claim 6 or 7, wherein the process of estimating the q-axis current estimates a q-axis current estimated value as a first-order lag of a q-axis current command value.   The process of setting the q-axis proportional gain is characterized by estimating a q-axis inductance based on a q-axis current estimated value and setting the q-axis proportional gain based on the estimated q-axis inductance. The method for controlling a three-phase AC motor according to any one of claims 6 to 8.   10. The process of setting the q-axis proportional gain includes multiplying the estimated q-axis inductance by a reciprocal of a current response time constant to calculate and set the q-axis proportional gain. The control method of the three-phase alternating current motor described in 2.

Images