JP2000217384A - Controller of position-sensorless motor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To realize high efficiency at all times in a system wherein a rotor position is estimated by a model formula of a motor in which a current detected by a current sensor and an applied voltage are used while a position sensor is not used. SOLUTION: A voltage applied by a voltage applying means 4 and a current outputted from a current output means 5 are inputted to an estimated position output means 9, and the position is estimated by calculation using the model formula of a motor. The estimated position output means 9 changes motor constants used in the model formula in accordance with operation conditions. That is, if a current command value or a current value is increased, the inductance in the model formula of the motor is reduced. With this constitution, the change of the inductance caused by the current value can be compensated, so that the controller of a position-sensorless motor which has excellent position estimation precision and always shows high efficiency can be provided.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention enables a motor to be controlled by using a rotor position (rotation angle) calculated and output using a motor model formula and current values and voltage values without using a position sensor. The present invention relates to a control device for a position sensorless motor.

[0002]

2. Description of the Related Art There is a method of estimating a rotor position θ from a motor model using a current detected by a current sensor and an applied voltage without using a position sensor. And the position θ
To generate a 180 ° energization command waveform, and perform 180 ° (sine wave) energization drive. This technique is well known in IEEJ Transactions D, 115, p420 (April 1995) and IEEJ Transactions D, 110, p1193 (November 1990).

[0003] The former will be briefly described.

The d-axis and q-axis voltage equations defined on the rotor magnetic poles of the motor are represented by the following equations.

[0005]

(Equation 1)

In the above equation, id and iq are the d-axis and q-axis components of the current, Ld and Lq are the d-axis and q-axis components of the inductance, and V
d * and Vq * are the d-axis and q-axis components of the voltage, R is the phase winding resistance,
E is the speed electromotive force, and ω is the rotational angular speed. If φ is the armature linkage flux on the d and q axes, then E = ω · φ.

An estimated rotation angle θc (n) and an estimated rotation angular velocity ω (n) are calculated and output in accordance with the following estimation formula derived from the motor model formula of (Equation 1).

[0008]

(Equation 2)

In the above equation, T is a PWM cycle, Kem is an induced voltage constant, and Ke, Kθ, and K are control gains. Furthermore, the subscript (n) indicates the current value, and (n-1) indicates the previous value.
Δθc (n-1) indicates the previously calculated rotational angular velocity for each PWM.

As described above, the position (rotation angle) of the rotor
Without using a position sensor such as an encoder for knowing the rotation angle, it is possible to output the angle and the angular velocity of the rotor by performing calculations using the model formula of the motor.

[0011]

However, in the above-mentioned conventional system, the resistance R, the inductance L, and the induced voltage constant Kem of the motor model are given as constant values. However, in an actual motor, the inductance L changes according to the applied current value. In addition, the voltage value actually used in the calculation is affected by a dead time (a period during which both upper and lower switching elements are turned off to avoid simultaneous conduction of the upper and lower switching elements) and the like, so that the estimated angle is advanced or delayed depending on operating conditions. Therefore, there is a problem that the estimated angle deviates from the actual value, the efficiency is reduced, and the desired torque is not output.

In view of the above problems, the present invention provides an estimated position output means for varying the constants of the motor model equation according to operating conditions such as the current value and the number of revolutions, thereby realizing the actual position of the rotor regardless of the operating state. It is an object of the present invention to provide a highly efficient position sensorless motor control device by accurately estimating the position.

[0013]

In order to solve these problems, a control device for a position sensorless motor according to the present invention comprises:
A rotor, a stator having a plurality of coils wound thereon, voltage applying means for applying a voltage to the coil, current value output means for applying a voltage to the coil by the voltage applying means, and detecting and outputting a current value flowing through the coil. An estimated position output means for calculating and outputting an estimated position based on a motor model formula including a motor constant, a current value output from the current value output means and a voltage value applied by the voltage application means, and outputting a current command value And a current control means for controlling a current flowing through the coil to a current command value using the estimated position of the rotor. The motor constant used in the model formula is changed according to conditions.

[0014]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the present invention will be described below with reference to FIGS.

(Embodiment 1) FIG. 1 is a block diagram of a control device of a position sensorless motor according to a first embodiment.

In FIG. 1, 1 is a motor, 2 is a rotor,
3 is a stator coil, 4 is a voltage applying means, 5 is a current value output means, 6 is a current sensor, 7 is an AD converter, 8 is a three-phase to two-phase converter, 9 is an estimated position output means, 10 is an AD converter, 11 is a speed control unit, 12 is a current command value output unit, 13 is a current value control unit, 14 is a voltage command generation unit, 1
Reference numeral 5 denotes a two-phase / three-phase converter, and 16 denotes a PWM controller.

Hereinafter, the operation will be described in detail.

The motor 1 includes a rotor 2 having a permanent magnet disposed on its surface, and a coil 3 (u) wound around a stator (not shown).
A phase coil 3u, a v-phase coil 3v, and a w-phase coil 3w).

As shown in FIG. 2, the voltage application means 4 includes a DC power supply 20 and switching elements such as transistors (Q1-Q).
6), a diode (D1 to D6) connected in antiparallel with the switching element, and a base drive circuit 21. The switching elements (Q1 to Q6) are turned on and off to apply a voltage from the DC power supply, and coils 3u, 3v, 3
Apply current to w. Here, a base drive circuit 21 for conducting the switching element in response to the ON signal is provided.

Next, the current value output means 5 includes a current sensor 6,
It comprises an AD converter 7 and a three-phase / two-phase converter 8.
First, the current value of two phases of the current flowing through the coil 3 is detected by the current sensor 6 and output as a voltage. The voltage is controlled by an AD converter 7 built in the microcomputer.
Is taken into the microcomputer. Then, the three-phase to two-phase converter 8 calculates the following equation to calculate the two-phase value from the three-phase AC amount (iu, iv).
It is converted into the phase direct current amount (Id, Iq) and output.

[0021]

(Equation 3)

As the rotor angle θc used in the above equation, an estimated angle θc output from the estimated position output means 9 described later is used.

Next, the means for creating the current command values (id *, iq *) will be described. First, a voltage corresponding to the speed command value ω * is taken into the microcomputer through the AD converter 10. Further, the estimated value ω of the angular velocity output from the estimated position output unit 9 is also input to the speed control unit 11. Then, the difference between the speed command value ω * and the estimated angular speed ω is subjected to proportional / integral (PI) control to output a torque command T. Next, the current command value output means 12 converts the torque command T into a current command value (id *, i
q *) and output. Here, the following equation is established between the torque and the current.

[0024]

(Equation 4)

Here, id and iq are the d-axis component and the q-axis component of the current value output from the current value output means 5, respectively. Φa is a value obtained from the induced voltage constant of the permanent magnet, and P is the number of pole pairs of the rotor.

Here, a general surface magnet type motor (SP
In M), since Lq = Ld, the torque command value T is converted into a current command value as follows.

[0027]

(Equation 5)

That is, according to the above equation, the current command value output means 1
2 outputs a current command value.

Next, the current value control means 13 will be described. The current value control means 13 includes a voltage command generation unit 14, a two-phase
It comprises a phase converter 15 and a PWM controller 16. First, the voltage command creating unit 14 outputs the DC values (id *, iq *) output from the current command value output means 12 and the current value output means 5 and (i
d, iq) are converted into voltage command values (Vd *, Vq *) according to the following PI operation and output.

[0030]

(Equation 6)

Here, KP and KI are proportional gains, respectively.
Integral gain. The two-phase to three-phase converter 15 converts the two-phase DC voltage amounts (Vd *, Vq *) to three-phase AC amounts (Vu *, Vv *,
Vw *) according to the following equation and output.

[0032]

(Equation 7)

The rotor angle θc used in the above equation is also the estimated angle θc output from the estimated position output means 9 described later.
Is used.

Next, the PWM controller 16 will be described with reference to FIGS.
This will be described with reference to FIG.

The voltage applying means 4 includes a switching element (Q1
To Q6), an ON or OFF operation is performed to reduce the loss. This method is a well-known PWM control in controlling a switching element of a motor, and will be described.

First, the triangular wave generating circuit 22 generates a triangular wave shown in FIG. Then, the voltage shown in FIG.
*) Is compared with the triangular wave, the switching element Q1 is turned on when the voltage is larger than the triangular wave, and the switching element Q4 is turned on when the voltage is smaller than the triangular wave. (There is a dead time period during which both the switching elements Q1 and Q4 are OFF to avoid this.) That is, an ON or OFF command is given during the PWM period (one cycle of the triangular wave), and the voltage applied to each phase is controlled by the length of the ON period. One cycle of PWM employs a value of about 300 μsec to 50 μsec.

Next, the estimated position output means 9 will be described. The estimated position output means 9 finally calculates and outputs the rotation angle θc and the rotation angular velocity ω (n) using a model formula derived from the following theoretical formula of the motor.

First, a d-axis inductance Ld and a q-axis inductance Lq, which are one of the variables of the motor constant, are obtained by using the following equations.

[0039]

(Equation 8)

In the above equation, Ld # max and Lq # max are the maximum values of the d-axis and q-axis inductances when the applied current is small and the stator is not saturated. Ld # min, Lq
Let #min be the minimum value of inductance, Ld # nom and Ld # nom be the intercepts for calculation, and Kd and Kq be proportional constants.

FIG. 4 shows the relationship between the current i and the inductance L. Here, the change in Ld corresponds to the change in id, and the change in Lq corresponds to the change in iq. As shown in FIG. 4, when the current increases, the inductance of the motor decreases as indicated by the solid line due to the magnetic saturation of the stator. Therefore, the inductances Ld and Lq used in the calculation are corrected and given as indicated by the dotted lines according to (Equation 8).

Next, the intermediate variables Δiγ (n) and Δiδ (n) are obtained from the following equations using Ld and Lq obtained by the above equations.

[0043]

(Equation 9)

As the d-axis component Vd * and the q-axis component Vq * of the voltage, a voltage command value output from the voltage command creating unit 14 is used. Further, T is a PWM cycle, E is an induced voltage proportional to the rotation speed obtained by (Equation 10), and R is a phase winding resistance value [Ω]. Further, since the subscript (n) indicates the current value and (n-1) indicates the previous value, Δθ (n-1) indicates the previously calculated movement amount for each PWM.

Next, Δiγ (n) and Δiδ obtained by (Equation 9)
Substituting (n) into the following equation, the induced voltage E (n) and the moving amount Δ for each PWM
Calculation of θ (n) is performed.

[0046]

(Equation 10)

Here, Kem represents an induced voltage constant per motor rotation speed, and Kθ and Ke represent control gains.

Then, as shown in the following equation, the current movement amount is added to the previous angle, the rotation angle θc and the movement amount Δθ are filtered, and the rotation angular velocity ω (n) is calculated and output.

[0049]

[Equation 11]

As described above, it is possible to output the angle and the angular velocity of the rotor 2 by calculating using the model formula of the motor without using a position sensor such as an encoder for knowing the position (angle) of the rotor 2. It becomes possible. Also,
Since the change due to the current amount of the inductance is taken into account, the position estimation accuracy is good and high efficiency is always realized. For example,
The estimated angle becomes 2 if the value of L used in the calculation is about 30% larger than the actual L, depending on the type of motor and the amount of current.
There may be a delay of about 0 degrees. In this case, the output torque is reduced by about 6%, and the efficiency is reduced.

In the present embodiment, as shown in FIG. 1, the PWM controller 1
Up to 6, analog / digital conversion and arithmetic processing are realized in the microcomputer.

In this embodiment, the inductance Ld, the detected current value output from the current value output means 5 is used.
Lq has been changed. Here, since the current actually applied is controlled by the current control means 13 to follow the current command value, the current command value (id *, iq *) output from the current command value output means 12 is used. It goes without saying that the same effect can be obtained even if the inductance is corrected in (Equation 8).

It is needless to say that this method is effective not only for a motor with a magnet but also for a synchronous reluctance motor (SynRM) composed of iron without a magnet. Here, the development relating to the SynRM motor type is not the main part of the present patent, and therefore will be omitted. For details, see IEEJ Semiconductor Power Conversion Study Group Materials, 1998, No.
See 31. The above document also estimates the position using inductance.

(Embodiment 2) FIG. 5 is a block diagram of a control device for a position sensorless motor according to a second embodiment.

In FIG. 5, 1 is a motor, 2 is a rotor,
3 is a stator coil, 4 is a voltage applying means, 5 is a current value output means, 6 is a current sensor, 7 is an AD converter, 8 is a three-phase to two-phase converter, 10 is an AD converter, 11 is a speed controller, 12 Is a current command value output means, 13 is a current value control means, 14 is a voltage command creation unit, 15 is a two-phase three-phase conversion unit,
6 is a PWM controller, and 19 is an estimated position output means.

Here, the same numbers are assigned to 1 to 8 and 10 to 16 which operate in the same manner as in the first embodiment described above, and the description is omitted. That is, the estimated position output means is changed in the second embodiment from the first embodiment.

Hereinafter, the operation will be described.

As in the first embodiment, the rotor angle θc uses the estimated angle θc output from the estimated position output means 19 described later, and the three-phase / two-phase converter 8 performs exactly the same operation as in (Equation 3). From the three-phase alternating current (iu, iv) to the two-phase direct current (Id, I)
Convert to q) and output.

Further, as in the first embodiment, the rotor angle θc uses the estimated angle θc output from the estimated position output means 19 described later, and the two-phase to three-phase converter 15 uses the two-phase DC voltage (Vd * , Vq *) to three-phase AC quantities (Vu *, Vv *, Vw *) are output in exactly the same manner as in (Equation 3).

Next, the estimated position output means 19 finally calculates and outputs the rotation angle θc and the rotation angular velocity ω (n) using an equation similarly derived from the model equation of the motor. Similarly, in the calculation process, the intermediate variables Δiγ (n) and Δiδ (n) are obtained from (Equation 9), and the induced voltage E (n) and the movement amount Δθ (n) for each PWM are calculated by (Equation 10). . Further, using (Equation 11), the current movement amount is added to the previous angle, the rotation angle θc and the movement amount are filtered, and the rotation angular velocity ω (n) is calculated and output.

Here, Vd * (n-1) and Vq * used in (Equation 9)
Consider the voltage actually applied to (n-1).

First, the operation of the switching elements Q1 and Q4 for one phase of the voltage applying means 4 shown in FIG. 2 for applying a voltage will be described. The switching element Q1 and the switching element Q4 are provided with a simultaneous cut-off period (dead time) for avoiding destruction of the switching element due to simultaneous conduction. FIGS. 6A and 6B show the timing of the conduction and cutoff of the u-phase switching element Q1 and switching element Q4. H (high) indicates conduction (on) and L (low) indicates cut-off (off). Here, in FIGS. 3A and 3B, there are periods A1 and A2 in which both the switching element Q1 and the switching element Q4 are cut off. This simultaneous cutoff period is called dead time. The dead time corresponds to a photocoupler (not shown) for transmitting a signal from a microcomputer included in the base drive circuit 21 of FIG. 2, a circuit for driving a switching element (not shown),
It is provided for several microseconds in order to absorb a time difference such as rising and falling of the switching element and to avoid simultaneous conduction.

Due to the dead time described above, the absolute value of the voltage value actually applied becomes smaller than the absolute value of the command voltage value. Here, the voltage applied to the motor 1 increases as the rotation speed increases, because the induced voltage increases. Assuming that the voltage drop due to the dead time is constant, the effect of the dead time decreases as the voltage increases.

When no voltage is applied due to the influence of the dead time, the current becomes small, and the fourth and fifth terms in (Equation 9) become small, so that ΔIγ becomes negative and E (n) in (Equation 10) becomes large. When E (n) increases, Δ
θ (n) increases, the estimated value advances, and the phase advances.
When the phase advances, the phase β in (Equation 12) shifts from β = 0, and the torque decreases.

[0065]

(Equation 12)

Here, i is a current value obtained by a vector sum of id and iq, and β is a current phase.

FIG. 7 is a graph showing the relationship between the rotational speed and the torque in the case where Kem is constant (Equation 10). As shown in FIG. 7, when a certain Kem1 calculated from the motor is given, the current phase advances too much in the low speed region as shown by the dotted line in FIG. 7, so that the torque decreases. Therefore, when Kem2 larger than Kem1 is given, Δθ is reduced in the low speed region as calculated by the second expression of (Equation 10), phase lead is improved, and a speed-torque characteristic as shown by a solid line in FIG. 7 is obtained. Can be However, if the calculation is performed at a constant Kem2 from the low speed range to the high speed range, the phase will be too late and the torque will decrease when going to the high speed range where the influence of the dead time is reduced.

FIG. 8 shows the relationship between the rotational speed ω and the induced voltage constant Kem. As shown in FIG. 8, Kem2 is used in the low-speed region, and Kem1 is used in the high-speed region.

[0069]

(Equation 13)

In the above equation, ω is the estimated position output means 19
, The previously calculated rotational angular velocity ω (n-1) is used.
ωL and ωH are the rotational speeds of Kem shown in FIG. 7, Kem3 is the intercept value of Kem shown in FIG. 7, Kem2 is the maximum value of Kem, Ke
Let m1 be the minimum value of Kem.

By giving Kem as described above, the torque is maximized at all rotation speeds, and the efficiency is improved.

As described above, it is possible to output the angle and the angular velocity of the rotor 2 by using the motor model formula without using a position sensor such as an encoder for knowing the position (angle) of the rotor 2. It becomes possible. Also,
Since the influence of the dead time due to the number of revolutions is taken into account, the position estimation accuracy is good and high efficiency is always realized.

Since the influence of the dead time is large when the voltage is small as described above, when the voltage output from the voltage command creating unit 14 is small, Kem is increased.
Needless to say, a similar effect can be obtained by reducing the size when the size is large.

When the amount of current is large, the fall of the voltage of the switching element is fast, and when the amount of current is small, the fall of the voltage of the switching element is slow, so that the influence of the dead time is different. Therefore, the same effect can be obtained even if Kem is changed according to the amount of current. Further, it is needless to say that the effect is further increased by changing Kem in consideration of the number of revolutions, the voltage and the current amount.

Here, there is a method called dead time compensation for compensating for the influence of dead time as a constant. However, as described above, since the influence of the dead time changes depending on the amount of current or the like, it is difficult to completely compensate for it, and a phase change occurs. Therefore, by changing Kem, the phase can be adjusted, and the same effect is obtained.

Here, needless to say, the same effect can be obtained even if Vd * and Vq * of (Equation 9) are directly changed in consideration of the influence of the dead time.

Also, the resistance value R of the coil used in (Equation 9) is small at the time of startup because the temperature is low, and increases during operation and increases. When the resistance value is increased, the phase is delayed, and the optimum efficiency phase is not obtained. However, the temperature rise cannot be uniquely determined because it differs depending on the operating conditions and the use environment.However, by setting in advance so that the resistance change is given as a function of the time after the start of the operation, a certain degree of correction becomes possible and the same effect can be obtained. Needless to say,

In this embodiment, the current command value and the DC values (id *, iq *) and (id, iq) output from the current value output means are provided.
Is controlled by PI control. However, it goes without saying that the three-phase current value control means 13 which is conventionally performed by an analog system may be configured (not shown). That is, the AC value (iu, iv) is output from the current value output means 5. It is output from the current value command means 12 (id *, iq *).
Is converted into a two-phase to three-phase signal using the estimated position θc output from the estimated position output means 19 to obtain an AC current command value (iu *, iv *).
Is output. And the error between iu * and iu and iv * and iv
Vu * and Vv * are obtained by control, and Vw * is calculated as Vw * =-Vu * -Vv *
Ask from. It goes without saying that a similar effect can be obtained by a configuration in which these values are given to the PWM controller 16.

[0079]

As described above, according to the present invention, the voltage value applied by the voltage application means 4 and the current value output from the current value output means 5 are input to the estimated position output means 9 and the motor model is input. The position is calculated and estimated from the equation. Here, the estimated position output means 9 changes the motor constant used in the model formula according to the operating condition. That is, when the current command value or the current value increases, the inductance of the model formula of the motor decreases. Alternatively, by providing an estimated position means for increasing the induced voltage constant when the rotational speed is low and decreasing the induced voltage constant when the rotational speed is high, it is possible to compensate for the change due to the current amount of the inductance and the influence of the dead time, thereby improving the position estimation accuracy. A highly efficient position sensorless motor control device can be provided.

[Brief description of the drawings]

FIG. 1 is a block diagram of a control device for a position sensorless motor according to a first embodiment.

FIG. 2 is an explanatory diagram of a voltage application unit and a PWM controller.

FIG. 3 is an explanatory diagram of a voltage command, a triangular wave, and switching.

FIG. 4 is a diagram showing the relationship between current and inductance in the first embodiment.

FIG. 5 is a block diagram of a control device for a position sensorless motor according to a second embodiment.

FIG. 6 (a) is a timing chart of the conduction / interruption of a switching element Q1 according to a second embodiment. FIG. 6 (b) is a switching element Q according to the second embodiment.
2 Conduction / interruption timing diagram

FIG. 7 is a characteristic diagram of rotation speed and torque when Kem is constant in the second embodiment.

FIG. 8 is a diagram illustrating a relationship between a rotation speed and an induced voltage constant according to the second embodiment.

[Explanation of symbols]

 Reference Signs List 1 motor 2 rotor 3, 3u, 3v, 3w stator coil 6 current sensor

 ──────────────────────────────────────────────────続 き Continuing from the front page (72) Inventor Toru Tazawa 1006 Kazuma Kadoma, Kadoma City, Osaka Prefecture Inside Matsushita Electric Industrial Co., Ltd. Terms (reference) 5H560 BB04 BB07 BB12 DA14 DC12 EB01 GG04 UA02 XA02 XA03 XA12 XA13

Claims (4)

[Claims] 1. A rotor, a stator having a plurality of coils wound thereon, voltage applying means for applying a voltage to the coil, and a current value output for detecting and outputting a current value flowing through the coil by the voltage applying means. Means for calculating and outputting an estimated position based on a motor model formula including a motor constant, the current value output from the current value output means, and the voltage value applied by the voltage applying means; A current command value output means for outputting a current command value given by a flow rate, and a current sensor means for controlling a current flowing through the coil to the current command value using the estimated position of the rotor. In the control device, the estimated position output means changes the motor constant used in the model formula according to a motor operating condition. Motor controller. 2. The control device for a position sensorless motor according to claim 1, wherein the estimated position output means reduces the inductance of the motor model when the current command value or the current value increases. 3. The control of the position sensorless motor according to claim 1, wherein the estimated position output means increases the induced voltage constant of the model of the motor when the rotation speed is low and decreases when the rotation speed is high. apparatus. 4. The control device for a position sensorless motor according to claim 1, wherein the estimated position output means decreases the resistance value of the model expression of the motor at the start of the operation and increases it after a certain period of the operation start.