JPH10341598A - Control method for polyphase permanent magnet field synchronous motor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To realize a highly accurate good torque control by calculating a current rotational position angle interpolatively based on an immediately preceding rotational speed and a rotational position angle calcualted at the immediately preceding control period when the calculated rotational position angle of a permanent magnet field synchronous motor is in a divergent region. SOLUTION: A permanent magnet field synchronous motor 1 rotates a rotor according to a rotary field generated by applying an AC three-phase voltage to a three-phase armature winding wound around the armature core of a stator. A voltage sensor 41 detects the DC voltage of a DC power supply 21. A current sensor 42 detects the phase currents Iu, Iw of two out of three feeder lines for supplying power from an inverter 22 to the armature winding of the permanent magnet field synchronous motor 1. A temperature sensor 43 detects the temperature of the armature winding of the stator. Detection output from the temperature sensor 43 is used for correcting the resistance of the armature winding of the permanent magnet field synchronous motor 1.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for controlling a permanent magnet field synchronous motor, and more particularly, to a permanent magnet field synchronous motor that does not use a sensor for detecting the rotational position angle of a rotor of the permanent magnet field synchronous motor. The present invention relates to a method for controlling a synchronous motor.

[0002]

2. Description of the Related Art In recent years, a study by Watanabe et al. On a control method of a permanent magnet field synchronous motor has been known (IEEE Transactions D, Vol. 110, No. 11, published in 1990, p. 119).
3 to P.E. 1200). In this control method, a DC voltage output from a DC power supply is exchanged for an AC voltage by a voltage type inverter, and the AC voltage drives a permanent magnet field synchronous motor.

In this technique, torque control is performed without using a sensor for detecting a rotational position angle of a rotor of a permanent magnet field synchronous motor. That is, in this torque control, the rotational position angle and the rotational speed of the rotor are calculated based on the voltage equation of the permanent magnet field synchronous motor, and the inverter converts the AC voltage into the permanent magnet field according to a torque command input from the outside. This is performed by applying the voltage to a magnetic synchronous motor.

[0004]

In the above control method, the following equation (1) is used to rotate the rotor of the permanent magnet field synchronous motor according to the voltage and current of the permanent magnet field synchronous motor. Calculate the position angle.

[0005]

(Equation 1) In the equation (1), θ represents the rotational position angle of the rotor of the permanent magnet field synchronous motor. ω represents the rotation speed of the rotor. Vx and Vy represent the X and Y components of the voltage in the voltage command vector on the fixed coordinate system, respectively. Ix and Iy represent the X and Y components of the current in the current vector on the fixed coordinate system.

Further, Rd represents the resistance value of the armature winding of the stator of the permanent magnet field synchronous motor on the d-axis in the rotating coordinate system of the rotor. Ld and Lq represent the inductance of the armature winding on the d-axis and the q-axis of the rotating coordinate system.
Here, p represents a differential operator (d / dt). However, when calculating the rotational position angle θ based on the equation (1), if the value of the denominator of the equation (1) becomes substantially zero,
tan θ becomes infinite and the rotational position angle θ diverges. Therefore, there is a problem that an accurate rotational position angle θ cannot be obtained, and the torque controllability of the permanent magnet field synchronous motor deteriorates.

Therefore, in order to cope with such a problem, the present invention takes measures to prevent the divergence of the calculated value of the rotation position angle θ, and calculates the rotation position angle θ accurately over a wide range of rotation speed. Accordingly, it is an object of the present invention to provide a method for controlling a permanent magnet field synchronous motor that obtains good torque controllability.

[0008]

In order to solve the above-mentioned problems, according to the first and second aspects of the present invention, when the calculated value of the rotational position angle of the permanent magnet field synchronous motor is in the divergence region, the latest value is obtained. Instead of the rotation position angle, the current rotation position angle is calculated by interpolation based on the rotation position angle calculated in the immediately preceding control cycle and the immediately preceding rotation speed.

Thus, the accuracy of the calculated value of the rotational position angle can be substantially improved over a wide rotational speed range.
As a result, the calculation accuracy of the rotation speed of the permanent magnet field synchronous motor is improved, and the torque control accuracy of the permanent magnet field synchronous motor is improved, thereby enabling good torque control.
According to the second aspect of the present invention, when it is determined that the magnitude of the voltage command vector is smaller than the predetermined lower limit, a pulse voltage vector is added to the voltage command vector in a plurality of subsequent control cycles. The average value of the pulse voltage vectors to be added in the plurality of control periods is set to substantially zero. Further, when the calculated value of the rotational position angle is in the divergence region, the direction of the pulse voltage vector is switched and added to the voltage command vector so that the calculated value of the rotational position angle deviates from the divergence region to obtain an added voltage command vector. . Then, based on the added voltage command vector, multi-phase AC power is supplied so that a multi-phase current flows through the permanent magnet field synchronous motor, and the torque of the permanent magnet field synchronous motor is controlled.

As a result, accurate and good torque control can be performed even in a region where the rotation speed of the permanent magnet field synchronous motor is low.

[0011]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below with reference to the drawings. FIG. 1 shows an example in which the present invention is applied to a control device of a three-phase permanent magnet field synchronous motor 1. The permanent magnet field synchronous motor 1 is configured such that a salient pole type or reverse salient pole type rotor composed of a permanent magnet is rotatably supported coaxially in a stator. The rotor rotates in accordance with a rotating magnetic field generated by applying a three-phase AC voltage to a three-phase armature winding wound on an armature core of the stator.

The control device includes an AC power generation circuit 2 for driving a permanent magnet field synchronous motor 1 and a control circuit 3. The AC power generation circuit 2 includes a DC power supply 21
And a voltage-type inverter 22.
2 converts a DC voltage supplied from the DC power supply 21 into an AC and applies a phase voltage to each of the three-phase armature windings of the stator of the permanent magnet field synchronous motor 1.

The voltage sensor 41 detects the DC voltage of the DC power supply 21. The current sensor 42 is connected to the inverter 2
2 among the three-phase power supply lines that supply power to the armature winding of the permanent magnet field synchronous motor 1 from the power supply lines 2, Iu, I
w is detected. The temperature sensor 43 detects the temperature of the armature winding of the stator. The detection output of the temperature sensor 43 is used to correct the resistance value of the armature winding of the permanent magnet field synchronous motor 1.

The control circuit 3 includes a microcomputer 31. The microcomputer 31 executes a main control program in accordance with a flowchart shown in FIG. 2, and executes an interrupt control program in accordance with the flowcharts shown in FIGS. Execute During these operations, the microcomputer 31 determines the phase voltage of each phase applied to each armature winding of the permanent magnet field synchronous motor 1 based on the digital conversion output of each of the AD converters 33 to 36 described later. Voltage command vectors (Vu, Vv, Vw) having the command values Vu, Vv, Vw as components are determined, and the permanent magnet field synchronous motor 1 is controlled at a constant control cycle. The main control program and the interrupt control program are stored in the ROM of the microcomputer 31 in advance.

The A / D converter 33 converts the torque command 51 from the outside into a digital signal, and
To enter. The A / D converter 34 converts the detection output of the voltage sensor 41 into a digital signal and inputs it to the microcomputer 31. The A / D converter 35 converts the two detected currents of the current sensor 43 into digital signals, and
To enter. The A / D converter 36 converts the detection output of the temperature sensor 43 into a digital signal,
To enter.

The voltage pulse width modulation circuit 32 (hereinafter referred to as PWM)
The circuit 32) receives a voltage command vector (Vu, Vv, Vw) from the microcomputer 31 and outputs a PWM signal corresponding to the voltage command vector (Vu, Vv, Vw) to the inverter 22 by a triangular wave comparison method.
As a result, the inverter 22 outputs the voltage command vector (V
u, Vv, Vw).

Incidentally, the PWM circuit 32 is provided with a voltage command vector (Vu, Vv, Vw) from the microcomputer 31.
Upon receiving the signal, a synchronization signal is output to the AD converter 35. Thereby, the AD converter 35
Performs the digital conversion process in synchronization with the synchronization signal. Hereinafter, the operation of the present embodiment thus configured will be described.

FIG. 2 is a flowchart showing the operation of the microcomputer 31 after starting. First, step 101
Then, initialization processing in the microcomputer 31 is performed. At this time, the pulse count variable c used in the interrupt control program is cleared, and the switching flag Fd for superimposing the pulse voltage vector is set.
It is cleared that q = 0.

In step 102, the polarity of the magnetic pole of the rotor of the permanent magnet field synchronous motor 1 is determined. This determination is made by instantaneously applying positive and negative voltages to the armature winding of the permanent magnet field synchronous motor 1 and observing changes in the phase current in response thereto. When a voltage is applied to the armature winding, the armature core is magnetized or demagnetized in accordance with the polarity of the field pole of the rotor.

In this case, the magnitude of the change in the detected phase current differs between the case where the magnetization is increased and the case where the magnetization is demagnetized because the magnetization characteristic of the armature core of the stator is nonlinear. Therefore,
The polarity of the magnetic pole of the rotor is determined from the combination of the positive and negative of the voltage applied to the armature winding and the magnitude of the phase current flowing through the armature winding. The details are shown in the contents of the IEEJ Transactions D by Watanabe et al.

In step 103, a digital conversion output of the A / D converter 34 with respect to the detection voltage of the voltage sensor 41 is input to the microcomputer 31. Also, in step 104, A-
The digital conversion output of the D converter 36 is input to the microcomputer 31. In step 105, the initial data of the resistance value of the armature winding of the permanent magnet field synchronous motor 1 stored in the ROM of the microcomputer 31 in advance is corrected based on the armature temperature in step 104. Thereby, the resistance value of the armature winding at the temperature is obtained.

In step 106, the torque command 51 input from the outside is input to the microcomputer 31 via the AD converter 33. Then, step 107
Then, the torque command 51 is converted into a corresponding current, and based on the converted current, the d-axis (in the direction of the magnetic field pole axis) and the q-axis, which are stationary with respect to the rotor of the permanent magnet field synchronous motor 1. The current command vector (I) on coordinates (hereinafter, referred to as rotational coordinates) consisting of (a direction orthogonal to the d-axis).
d ^* , Iq ^* ). Details are described in "High-performance variable control of permanent magnet field synchronous motor with embedded magnet structure" (pp. 278 to 283 of the IEEJ Industrial Application Division, 1992) by a child of Osaka Prefecture University.

In the above steps 103 to 107
This processing is performed by the interrupt control program described below.
Is executed in the extra time, and the current command vector (Id^*,
Iq ^*) And the resistance value of the armature winding
Provided for processing by the system. Above interrupt control program
The interrupt processing of the system is performed by the current sensor 4 of the AD converter 35.
With the completion of the digital conversion for the detection output 2
3 and the flowchart of FIG.

First, in step 201, the current sensor 4
2, three-phase current vectors (Iu, Iv, Iw) on coordinates (hereinafter referred to as fixed coordinates) stationary with respect to the armature core of the stator with respect to the detection output of No. 2. ) Is input to the microcomputer 31. Then, in step 202, the current vectors (Iu, Iv, Iw) are converted into two-phase current vectors (Ix, Iy) on fixed coordinates. The conversion from the three phases to the two phases is performed using the following equations (2) and (3).

[0025]

## EQU2 ## Iy = (2/3) ¹ / 2.Iu + (1/6) ^1/2
・ Iv + (1/6) ^1/2・ Iw

[0026]

## EQU3 ## Ix = (1/2) ¹ / 2.Iv- (1/2) ^1/2
Iw Next, in step 203, the magnitude V of the voltage command vector in the control cycle one cycle before is compared with a constant Vth1 as its lower limit. Also, the pulse count variable c
Is within the range of 1 ≦ c ≦ 5.

Here, when V ≧ Vth1 and (c = 0 or c> 5), the processing of step 205 is performed through step 204 based on the determination of NO in step 203. In step 205, the rotational position angle θ is calculated using the equation (1). Here, as the Ix and Iy, the calculation processing data in step 202 is used, and pI
x and pIy are Ix and Iy of the control cycle one cycle before.
Is divided by the control cycle, the rotational speed ω is the rotational speed ω calculated in step 208 described later in the control cycle one cycle before, and Vx, Vy is the control one cycle before. Step 21 described later in the cycle
Calculate using the numerical value calculated in step 1.

Ld and Lq are the d-axis component and the q-axis component of the rotational coordinates of the armature winding inductance, respectively, and are stored in the ROM of the microcomputer 31 in advance. Rd is the armature winding resistance in the d-axis direction of the rotational coordinate, which is the same as the resistance value of the armature winding calculated in step 105 of FIG. Here, the rotational position angle θ takes a binary value in the range of −π ≦ θ <π, but based on the polarity of the magnetic pole of the permanent magnet field synchronous motor 1 determined at the time of starting, the initial rotational position angle after starting is determined. At the time of calculation, the rotational position angle θ is uniquely determined to any value. Therefore,
In the subsequent control cycle, the rotational position angle θ is determined so that its numerical value is continuous from the first rotational position angle.

In the processing of the next processing routine 206, FIG.
In step 301, the value of the denominator (hereinafter referred to as the denominator Vc) of the equation (1) when the rotational position angle θ is calculated based on the equation (1) in step 205 is determined. That is, if it is smaller than the preset lower limit value Vth2, step 301
Is YES. Then, in step 302,
The rotational position angle θ of the current control cycle is obtained by linearly extrapolating (interpolating) the rotational position angle θ = θold and the rotational speed ω one control cycle before based on the control cycle Tp. Step 2
The value of the rotational position angle θ calculated in 05 is ignored. In step 303, the rotational position angle θ in the current control cycle is updated as θold.

In step 207 in FIG. 4, the current vector (Ix, Iy) on the fixed coordinates is converted into a two-phase current vector (Id, Iq) on the rotating coordinates by the following simultaneous equation (4).

[0031]

Id = (cos θ) · Iy + (sin θ) · Ix Iq = (− sin θ) · Iy + (cos θ) · Ix where θ is the calculation result of step 205 or step 302. Next, in step 208, equation 5
Is calculated based on the following equation.

[0032]

(Equation 5) Here, Vq calculated in step 209 described later in the control cycle one cycle before is used as Vq. φ
a is the magnetic flux of the rotor stored in the ROM of the microcomputer 31 in advance. Rq is the resistance of the armature winding in the q-axis direction of the rotation coordinate, and is the same as the resistance value of the armature winding calculated in step 105 of FIG.

In step 209, step 10 in FIG.
Based on the current command vectors (Id ^* , Iq ^* ) calculated in step 7, vectors (Vd, Vq) of two-phase voltage commands on the rotating coordinates are calculated by so-called PI control. The components Vd, Vq of the voltage command vector (Vd, Vq) are updated as Vd0, Vq0, respectively, in step 210, and the magnitude V of the voltage command vector is calculated by the following equation (6).

[0034]

In [6] ^{V = {(Vd0) 2 +} (Vq0) 2} 1/2 next step 211, the voltage command vector (Vd, Vq) of the two-phase on the rotating coordinates in the simultaneous equation of the following Equation 7 Two-phase voltage command vector (Vx, Vy) on fixed coordinates based on
Is converted to

[0035]

Vy = (cos θ) · Vd− (sin θ) · Vq Vx = (sin θ) · Vd + (cos θ) · Vq Next, the voltage command vector (Vx, Vy) is converted into a three-phase Voltage command vector (Vu, V
v, Vw). Then, in step 213, the voltage command vector (Vu, Vv, Vw) is changed to PWM.
Output to the circuit 32. Thus, the interrupt processing by the interrupt control program ends. This is a normal control process in the interrupt process of the interrupt control program.

Next, a case where the permanent magnet field synchronous motor 1 is driven at a low speed will be described. Step 203 in FIG.
When the magnitude V of the voltage command vector becomes smaller than the lower limit value Vth1, a determination of YES is made. Thereafter, a correction period is set, and a preliminary interrupt process different from the normal interrupt process by the interrupt control program is performed. In this interrupt processing, the steps 201, 202, 211 to 213 are used in common with the normal processing, but the method of obtaining the rotational position angle θ, the voltage command vector (Vd, Vq) and the like is different.

In step 214, the pulse count variable c is determined. During the normal control process, the pulse count variable c is reset at step 204.
Therefore, c = 0 in the first control cycle in which the normal control processing is switched to the preliminary control processing. Therefore, the determination in step 214 is YES, and step 21
5 is performed.

In step 215, the value of the switching flag Fdq determined in steps 307 and 308 described later is determined. Here, when the switching flag Fdq = 0, the determination in step 215 is YES. Accordingly, in step 216, the voltage command vectors (Vd, Vd) in the control cycle immediately before switching from the normal control processing to the preliminary control processing by the interrupt control program.
The pulse voltage vector ΔV is added only to the d-axis direction to the data Vd0 and Vq0 which are the components of q). Then, the addition result is set as a voltage command vector (Vd, Vq).

That is, based on the following equations (8) and (9),
The voltage command vector (Vd, Vq) is calculated).

[0040]

Vd = Vd0 + ΔV

[0041]

Vq = Vq0 In step 215, the switching flag Fdq
= 1, it is determined as NO. In step 217, the pulse voltage vector ΔV is added to the data Vd0 and Vq0.
Is added only in the q-axis direction. Then, the addition result is set as a voltage command vector (Vd, Vq).

That is, the voltage command vectors (Vd, Vq) are calculated based on the following equations (10) and (11). The switching flag Fdq is initialized to Fdq = 0 in step 101, and thereafter, is rewritten only in steps 307 and 308 described later.

[0043]

Vd = Vd0

[0044]

Vq = Vq0 + ΔV Next, in step 222, the rotational position angle in the current control cycle is linearly extrapolated from the rotational position angle θold calculated in the control cycle one cycle before, assuming that the rotational speed ω is constant. Required.

That is, assuming that the control cycle is Tp, (θ + T
p · ω) is the rotational position angle θ in the current control cycle. The calculated rotational position angle θ is updated to θold. In step 232, the value obtained by incrementing the pulse count variable c is divided by a natural number N (indicated by modN), and the remainder of the division result is obtained. And
This remainder is updated as the pulse count variable c. Here, the pulse count variable c is updated to 1. Note that N
Is a natural number of 6 or more, and a sufficiently large number (for example, 20 or more) in which torque ripple generated by superimposing the pulse voltage vector can be ignored.

Then, based on the rotational position angle θ calculated in step 222, the processing of steps 211 to 213 is performed in the same manner as the above-described normal processing. As a result, the voltage command vector (Vd, Vq) on the rotational coordinates set in step 216 or 217 is converted into a three-phase voltage command vector (Vu, Vv, Vw) and output to the PWM circuit 32.

In the next control cycle, the pulse count variable c
= 1, the interrupt control program executes step 2
03 through steps 214 and 218 to step 219
Proceed to. In step 219, the switching flag Fdq is determined as in step 215. When the switching flag Fdq = 0, the determination in step 219 is Y
It becomes ES, and in step 220, the pulse voltage vector (-ΔV) is added to the update data Vd0 and Vq0 only in the d-axis direction. The result of this addition is a voltage vector (Vd,
Vq).

That is, in step 220, the following equation 1
The voltage command vector (Vd,
Vq) is calculated.

[0049]

Vd = Vd0−ΔV

[0050]

Vq = Vq0 In step 219, the switching flag Fdq
When = 1, a determination of NO is made, and the pulse voltage vector (-.DELTA.V) is added to the update data Vd0 and Vq0 only in the q-axis direction.
d, Vq).

That is, in step 221, the voltage command vector (Vd, Vq) is calculated based on the following equations (14) and (15).

[0052]

Vd = Vd0

[0053]

Vq = Vq0−ΔV Therefore, each component V of the voltage command vector (Vd, Vq)
d and Vq largely swing to the positive side in the first control cycle (c = 0) of the preliminary control processing around the update data Vd0 and Vq0, and swing greatly to the negative side in the next control cycle (c = 1).

That is, each component of the voltage command vector (Vd, Vq) is corrected by a pulse time series of 0 in its average in a short period of the first control cycle and the next control cycle. In step 222, the rotational position angle θ is obtained in the same manner as in the first control cycle, and the processing of steps 211 to 213 is performed based on the rotational position angle θ. In step 232, since the pulse count variable c is incremented, in the subsequent control cycle (c = 2),
The interrupt control program proceeds to step 220 or 221 via steps 214, 218 and 219. In the components Vd and Vq of the voltage command vector on the rotating coordinates, when the switching flag Fdq = 0, only the d-axis direction largely swings to the negative side with respect to each of the update data Vd0 and Vq0, and the switching flag Fdq = 1 At this time, only the q-axis direction largely swings to the negative side with respect to each of the update data Vd0 and Vq0.

Further, in the subsequent control cycle (c = 3), the interrupt control program proceeds to step 216 or 217 via steps 214 and 215. Then, in the components Vd and Vq of the voltage command vector, the switching flag Fd
When q = 0, only the update data Vd0, V
q0, the switching flag Fdq
When = 1, only the update data Vd0, Vq in the q-axis direction
It swings greatly to the positive side with respect to 0.

That is, each component of the voltage command vector (Vd, Vq) is corrected by a pulse time series of 0 in its average during the short period composed of these two control periods. When the normal processing is switched to the preliminary control processing, the pulse voltage vector ΔV in the first four consecutive control cycles of the correction period becomes the switching flag F within the four control cycles.
Since the value of dq does not change, in the d-axis or q-axis direction, 1:
-1: -1: 1 ratio. Further, the pulse time series of the second half period is a pulse time series obtained by inverting the pulse voltage vector in the first half period, and the average in each short period is zero.

When the above four control cycles are completed and the pulse counter variable c becomes 4, the interrupt control program goes through steps 203, 214, 218 and 223 to step 22.
Proceed to 4. In step 224, the same processing as step 205 in the normal control processing is performed. That is, the rotational position angle θ is calculated based on the equation (1). At this time, if the magnitude of the voltage command vector is small, the value of both the denominator and the numerator in Equation 1 becomes small, and the accuracy of calculating the rotational position angle θ decreases due to the influence of noise.

Then, steps 216, 217, 22
By superimposing a large pulse voltage vector on the original voltage command vector at 0 and 221, the voltage command vector is made large and the calculation accuracy of the rotational position angle θ is prevented from lowering. However, even in this case where the pulse voltage vector is superimposed, if the pulse voltage vector is superimposed in the same direction, the value of the denominator Vc in the equation (1) may become 0 as shown in FIG. At this time, the calculation result of Expression 1 diverges, and an accurate rotation position angle θ cannot be obtained.

Therefore, the processing routine 22 shown in FIGS.
At 5, countermeasures are taken against this problem. FIG.
In step 304, the value of the denominator Vc at the time when the rotational position angle θ is calculated based on the equation (1) in step 224 is determined. If the rotational position angle θ is smaller than the preset lower limit value Vth3, the processing routine 225 proceeds to step 306, where the switching flag Fdq is determined.

When the switching flag Fdq = 0, the determination in step 306 is YES, and the process proceeds to step 307.
, The switching flag Fdq = 1 is set. Also,
If the determination in step 306 is NO, in step 308, the switching flag Fdq is set to 0. Here, by switching the value of the switching flag Fdq, in steps 215 and 219, the direction in which the pulse voltage vector is superimposed next time is switched between the d-axis direction and the q-axis direction.

That is, when the preliminary control processing of step 214 and subsequent steps is repeated at a low rotation speed of the permanent magnet field synchronous motor 1, when the value of the denominator Vc in the equation (1) becomes small, the pulse voltage is increased. Since the direction in which the vectors are superimposed is switched, the value of the denominator Vc in the equation (1) does not decrease any further. At this time, the value of the denominator Vc in the equation (1) is as shown in FIG. 8, and does not become smaller than the lower limit value Vth3. Therefore, the calculation accuracy of the rotational position angle θ is ensured. Immediately after switching from the normal control processing to the preliminary control processing, the switching flag F at that time is used.
Depending on the value of dq, the value of the denominator of equation (1) may become zero at step 224.

At this time, at step 304, it is determined that the value of the denominator of the equation (1) is smaller than the lower limit value Vth2 (Vth2 <Vth3). Then, at step 305, the rotational position angle θ of the current control cycle is set to 1 It is obtained by linear extrapolation from the rotational position angle θold and the rotational speed ω before the control cycle. At this time,
The result calculated in step 224 is ignored. In step 309, the rotational position angle θ in the current control cycle is θol
Updated to d.

Steps 226, 227, 230, 231
Then, steps 207, 208, and 20 of the normal control process
The same processing as in steps 9 and 210 is performed. That is, step 2
At 26, the current vector (Ix, Ix
y) is the current vector (I
d, Iq). Then, in step 227, the rotation speed ω is calculated based on the equation (5), and in step 230, the voltage command vector (V
d, Vq).

The components of the voltage command vector (Vd, Vq) obtained in step 230 are updated as Vd0 and Vq0 in step 231, and the magnitude V of the voltage command vector is calculated based on the equation (6). Is done. In the next control cycle, the pulse count variable c becomes 5,
Steps 203, 214, 2
The processing proceeds from 223 to step 228 via 18.

In step 228, the rotational position angle θ is
In the control cycle one cycle before, the rotational position angle θ calculated in step 224 (or step 305) and 227
It is obtained by extrapolation in the same manner as in step 222 based on the old and the rotation speed ω. Then, the calculation result is updated as θold. Next, in step 229, the current vector (Ix, Iy) on the fixed coordinates is converted into a two-phase current vector (Id, Iq) on the rotational coordinates in the same manner as in step 207, based on the rotational position angle θ. You. Thereafter, the interrupt control program proceeds to step 230.

In the next control cycle, the pulse count variable c
= 6 (when N> 6), so in step 203,
When the magnitude V of the voltage command vector is larger than the lower limit value Vth1 (when V> Vth1), a determination of NO is made.
Along with this, the processing in the correction period ends, and normal control processing is performed. In addition, the magnitude V of the voltage command vector is lower limit V
When it is smaller than th1 (when V <Vth1), the interrupt control program returns to steps 214, 218 and 223
Then, the process proceeds to step 228.

That is, in the subsequent control cycle, the magnitude V of the updated voltage command vector is smaller than the lower limit (V <Vth
At the time of 1), the processing of steps 228 to 230 is performed as long as the pulse count variable c ≦ N−1. That is,
Assuming that the rotation speed ω is constant, the rotation position angle θ is obtained by linear extrapolation. Then, the current vector (Id, Iq) and the voltage command vector (Vd, Vq) are obtained based on the rotational position angle θ and the rotational speed ω. Thereafter, the processing in the correction period ends.

In practicing the present invention, the following equations (16) to (19) are adopted instead of the above equation (1), and V1 and V2 are calculated based on the equations (16) and (17). Alternatively, the position angle θ may be calculated based on the equation (18) or (19) according to each of V1 and V2. With this, the same operation and effect as those of the above embodiment can be ensured.

[0069]

V1 = Vy− (Rd + pLd) Iy + ωLq
Ix-ωLdIx

[0070]

V2 = −Vx + (Rd + pLd) Ix + ωL
qIy-ωLdIy

[0071]

## EQU18 ## sin θ = V1 / ｛V1 ² + V2 ² ｝ ^1/2

[0072]

Cos θ = V2 / ｛V1 ² + V2 ² ｝ ^1/2 Here, when the denominator of each of the equations (18) and (19) becomes small, the calculation error of the rotational position angle θ increases. For this reason,
In steps 301 and 304 (see FIGS. 5 and 6),
The determination may be made by replacing the denominator in each of the equations 18 or 19 with the denominator Vc.

[Brief description of the drawings]

FIG. 1 is a schematic overall configuration diagram showing an embodiment of the present invention.

FIG. 2 is a flowchart of a main control program executed by the microcomputer of FIG. 1;

FIG. 3 is a first part of a flowchart of an interrupt control program executed by the microcomputer of FIG. 1;

FIG. 4 is a latter part of a flowchart of an interrupt control program executed by the microcomputer of FIG. 1;

FIG. 5 is a detailed flowchart of a processing routine 206 in FIG. 3;

FIG. 6 is a detailed flowchart of a processing routine 225 of FIG. 3;

FIG. 7 is a timing chart showing a temporal change of a denominator of a calculation formula (formula 1) for the rotational position angle θ.

FIG. 8 is a timing chart showing a temporal change of a denominator of an equation for calculating the rotational position angle θ (Equation 1) when a pulse voltage vector is added.

[Explanation of sign numbers]

1: permanent magnet field synchronous motor, 2: AC power generation circuit,
3 ... control circuit, 21 ... DC power supply, 22 ... inverter, 3
DESCRIPTION OF SYMBOLS 1 ... microcomputer, 32 ... PWM circuit, 33-36 ... AD converter, 42 ... current sensor.

Claims (2)

[Claims] 1. A multi-phase AC power is supplied to a multi-phase type permanent magnet field synchronous motor (1) based on a voltage command vector so that a multi-phase current flows. A first step of controlling (213); a second step of determining the magnitude of the voltage command vector (203); and calculating a rotational position angle of the permanent magnet field synchronous motor based on the phase current and the voltage command vector. A third step (205, 224) of calculating the rotation speed of the permanent magnet field synchronous motor based on the phase current, the voltage command vector and the rotation position angle (208, 227); A fifth step (20) of setting the voltage command vector based on the position angle, the rotation speed, and the torque command value;
9 to 212), wherein the first to fifth steps are repeated in a control cycle, wherein the calculated value of the rotational position angle is in a divergence region. In a sixth step (301), the current rotation position angle is calculated by interpolation based on the rotation position angle calculated in the immediately preceding control cycle and the immediately preceding rotation speed, instead of the latest calculated value of the rotation position angle. , 302). A method for controlling a permanent magnet field synchronous motor, comprising: 2. When the magnitude of the voltage command vector is determined to be smaller than a predetermined lower limit value in the second step, a pulse voltage vector is added to the voltage command vector in a plurality of subsequent control cycles. A seventh step (216, 2) in which the average value of the pulse voltage vectors to be added in the plurality of control periods is set to approximately zero as a voltage command vector.
17, 220, 221). In the first step, based on the added voltage command vector, multi-phase AC power is supplied so as to flow a multi-phase current to the permanent magnet field synchronous motor. The torque of the permanent magnet field synchronous motor is controlled, and when the calculated value of the rotational position angle is in the divergence region, the calculated value of the rotational position angle deviates from the divergence region in the seventh step, 2. The control method for a permanent magnet field synchronous motor according to claim 1, wherein the direction of the pulse voltage vector is switched and added to the voltage command vector to obtain the added voltage command vector.