JP2015195715A - Estimation method of q-axis inductance, and estimation device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To more accurately estimate q-axis inductance.SOLUTION: An estimation method of q-axis inductance includes: (a) controlling a motor (4) so as to generate torque variation; (b) calculating an estimate (θ^) for a magnetic pole position of the motor (4) on the basis of current values q(i, iand i) of the motor (4) and an estimate (L^) of the q-axis inductance; and (c) updating the estimate (L^) of the q-axis inductance on the basis of the variation in the estimate (θ^) of the magnetic pole position caused by the torque variation.

Description

  The present invention relates to a technique for controlling a motor.

  In order to control the motor, it is necessary to know the magnetic pole position of the motor. The position of the magnetic pole can be known by using the position sensor, but it is expensive, and if the position sensor fails, the motor may not be controlled. For this reason, it is often performed to estimate the magnetic pole position without using a position sensor. In addition, since the motor device constants vary, an error occurs in the magnetic pole position estimated from the motor voltage equation. For example, Patent Document 1 describes a technique for correcting the q-axis inductance by regarding the estimated value of the magnetic pole position when the motor current becomes zero as the correct magnetic pole position, and estimating the magnetic pole position more accurately. Yes.

JP 2005-130691 A

Chen Zhen, 3 others, “Sensorless position estimation method and stability of salient pole type brushless DC motor”, 1998 IEEJ National Conference on Industrial Applications, No. 59, 1998, p. 179-182 Kiyoshi Sakamoto and two others, “Position Sensorless Control of IPM Motors by Direct Estimation of Axis Error”, IEEJ Semiconductor Power Conversion / Industrial Electric Power Application Joint Study Group, November 2000, SPC-00-67, p. . 73-76

  However, according to the technique of Patent Document 1, there is a problem that it is difficult to accurately estimate the q-axis inductance when the motor current does not become zero.

  An object of the present invention is to estimate the q-axis inductance more accurately even if the motor current does not become zero.

In the first q-axis inductance estimation method according to the present invention, (a) the motor (4) is controlled so that torque fluctuation occurs, and (b) the current value (i _u , i _v , i _w ) and an estimated value (L _q ^) of the q-axis inductance, an estimated value (θ _e1 ^) of the magnetic pole position of the motor (4) is obtained, and (c) the magnetic pole generated by the torque fluctuation Based on the fluctuation of the estimated position value (θ _e1 ^), the estimated q-axis inductance value (L _q ^) is updated.

According to this, the estimated value (L _q ^) of the q-axis inductance is updated based on the fluctuation of the estimated value (θ _e1 ^) of the magnetic pole position caused by the torque fluctuation. For this reason, even if the current of the motor (4) does not become zero, the estimated value (L _q ^) of the q-axis inductance can be obtained more accurately.

  According to the second q-axis inductance estimation method of the present invention, in the first q-axis inductance estimation method, the inverter circuit (80) that drives the motor (4) rectifies the AC supplied from the AC power supply. The pulsating direct current obtained as described above is input, and in the process (a), control is performed so that the torque fluctuation occurs in synchronization with the pulsation of the input voltage of the inverter circuit (80).

  According to this, since torque fluctuations are generated in synchronization with the pulsation of the input voltage of the inverter circuit (80), there is no need to generate a phase for causing torque fluctuations.

  In the third q-axis inductance estimation method according to the present invention, in the second q-axis inductance estimation method, the frequency of the torque fluctuation is twice the frequency of the AC supplied from the AC power supply.

  According to this, it can be utilized that the torque fluctuation of the motor (4) has a frequency twice as high as the AC frequency supplied by the AC power supply.

A fourth q-axis inductance estimation method according to the present invention is the first q-axis inductance estimation method, wherein in the processing of (c), the torque fluctuation of the estimated value (θ _e1 ^) of the magnetic pole position is performed. And the estimated value (L _q ^) of the q-axis inductance is updated so that the same frequency component as the torque fluctuation approaches zero.

According to this, the estimated value (L _q ^) of the q-axis inductance is updated so that the component having the same frequency as the torque fluctuation approaches zero. For this reason, the estimated value (L _q ^) of the q-axis inductance can be obtained more accurately by updating.

  A fifth q-axis inductance estimation method according to the present invention is the first to fourth q-axis inductance estimation method, wherein the processing of (b) and (c) satisfies a predetermined condition. Repeat until

According to this, since the estimated value (L _q ^) of the q-axis inductance is updated until a predetermined condition is satisfied, a more accurate estimated value (L _q ^) of the q-axis inductance can be obtained.

A sixth q-axis inductance estimation method according to the present invention is the first or fourth q-axis inductance estimation method. In the process (c), the estimated value (θ _e1 ^) of the magnetic pole position is smoothed. Then, the difference between the result (θ _ef1 ^) and the estimated value (θ _e1 ^) of the magnetic pole position is obtained as the fluctuation of the estimated value (θ _e1 ^) of the magnetic pole position caused by the torque fluctuation.

According to this, the fluctuation of the estimated value (θ _e1 ^) of the magnetic pole position caused by the torque fluctuation can be obtained.

The seventh q-axis inductance estimation method according to the present invention is the first q-axis inductance estimation method. In the processing of (c), the estimated value (θ _{e1) of the} magnetic pole position generated by the torque fluctuation is used. The fluctuation component of ^) is used after removing the influence of the fluctuation of the rotational speed of the motor (4) caused by the torque fluctuation.

According to this, since the influence of the fluctuation of the rotational speed of the motor (4) caused by causing the torque fluctuation is eliminated, the estimated value (L _q ^) of the q-axis inductance can be obtained more accurately.

The q-axis inductance estimation apparatus according to the present invention includes a fluctuation generator (40, 214) that controls a motor (4) so that torque fluctuation occurs, and current values (i _u , i _v , i _w ) of the motor (4). ), And a magnetic pole position estimation unit (34) for _obtaining an estimated value (θ _e1 ^) of the magnetic pole position of the motor (4) based on the estimated value (L _q ^) of the q-axis inductance, and the torque variation And an inductance estimator (50; 250) for updating the estimated value (L _q ^) of the q-axis inductance based on the fluctuation of the estimated value (θ _e1 ^) of the magnetic pole position.

According to this, the inductance estimation unit (50; 250) updates the estimated value (L _q ^) of the q-axis inductance based on the variation of the estimated value (θ _e1 ^) of the magnetic pole position caused by the torque variation. . For this reason, even if the current of the motor (4) does not become zero, the estimated value (L _q ^) of the q-axis inductance can be obtained more accurately.

According to the present invention, the q-axis inductance (L _q ^) can be estimated more accurately even if the motor current does not become zero. Therefore, estimation of the magnetic pole position (θ _e1 ^) of the motor (4) and current control of the motor (4) can be performed more accurately.

FIG. 1 is a block diagram illustrating a configuration example of a power conversion device according to an embodiment of the present invention. FIG. 2 is a block diagram illustrating a configuration example of the voltage command generation unit of FIG. FIG. 3 is an explanatory diagram showing the relationship between the magnetic pole position and the coordinate axis of the motor shown in FIG. FIG. 4 is a block diagram illustrating a configuration example of the fluctuation generation unit and the inductance estimation unit of FIG. FIG. 5A is a graph showing an example of the estimated value of the magnetic pole position obtained by the magnetic pole position estimating unit in FIG. FIG. 5B is a graph illustrating an example of the output of the phase generation unit in FIG. FIG. 5C is a graph showing an example of the fluctuation component of the estimated value of the magnetic pole position caused by the torque fluctuation. FIG. 6 is a graph showing an example of an approximate expression of q-axis inductance. FIG. 7 is a graph showing another example of an approximate expression of q-axis inductance. FIG. 8 is a flowchart illustrating an example of processing in the voltage command generation unit of FIG. FIG. 9 is a diagram illustrating an example of a limit range of the d-axis current value. FIG. 10 is a block diagram showing a configuration of a modification of the voltage command generation unit of FIG. FIG. 11 is a block diagram showing a configuration of a modification of the inductance estimation unit of FIG. FIG. 12A is a graph showing an example of the estimated value of the magnetic pole position obtained by the magnetic pole position estimating unit in FIG. FIG. 12B is a graph showing an example of the phase used in the inductance estimation unit of FIG. FIG. 12C is a graph showing an example of the fluctuation component of the estimated value of the magnetic pole position caused by the torque fluctuation. FIG. 13 is a block diagram illustrating a configuration of a modification of the power conversion device of FIG. 1. FIG. 14 is a circuit diagram showing an example of the switching circuit of the matrix circuit of FIG. FIG. 15 is a circuit diagram showing another example of the switching circuit of the matrix circuit of FIG.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. Components shown with the same reference numbers in the drawings are identical or similar components.

  FIG. 1 is a block diagram illustrating a configuration example of a power conversion device according to an embodiment of the present invention. 1 includes a PWM (pulse width modulation) modulator (6), a voltage command generator (10), a reactor (72), a converter circuit (76), and a capacitor (78). And an inverter circuit (80). The power converter (100) converts the AC power supplied from the AC power supply (2) into three-phase AC power and drives the motor (4). As the motor (4), for example, a so-called IPM (interior permanent magnet) motor is employed. A motor (4) drives the compressor of an air conditioning apparatus, for example.

  The converter circuit (76) is connected to the AC power source (2) via the reactor (72). In this example, the converter circuit (76) is a diode bridge rectifier having four diodes. These diodes convert the alternating current from the alternating current power source (2) into a direct current by full-wave rectification and output it. The capacitor (78) is connected between the output nodes (N1, N2) of the converter circuit (76).

  The inverter circuit (80) converts the direct current output from the converter circuit (76) into alternating current and outputs the alternating current to the motor (4). The inverter circuit (80) has a plurality of legs (82U, 82V, 82W) connected in parallel. The U-phase leg (82U) includes an upper arm switching circuit (84) and a lower arm switching circuit (86) connected in series to the upper arm switching circuit (84). The leg (82U) outputs a drive current (IU) to the motor (4) from the intermediate node (NU) to which the upper arm switching circuit (84) and the lower arm switching circuit (86) are connected.

  The V-phase leg (82V) and the W-phase leg (82W) are configured in the same manner as the U-phase leg (82U). The leg (82V) outputs drive current (IV) from the intermediate node (NV) to the motor (4). The leg (82W) outputs drive current (IW) from the intermediate node (NW) to the motor (4).

The voltage command generation unit (10) includes the speed command value (ω _ref ), the voltage value (v _u , v _v , v _w ) obtained by the PWM modulation unit (6), and the current of each phase of the motor (4). Based on the values (i _u , i _v , i _w ), voltage command values (v _{u_ref} , v _{v_ref} , v _{w_ref} ) are generated and output to the PWM modulation unit (6). The current value (i _u , i _v , i _w ) is measured using, for example, a shunt resistor connected in series to each leg (82U, 82V, 82W). Based on the voltage command values (v u — _ref , v v — _ref , v w — _ref ), the PWM modulation unit (6) _generates voltage values (v _u , v _v , v _w for each phase of the drive signal (DS) and the motor (4). ) And output. The drive signal (DS) includes a plurality of signals each controlling each switching element of the inverter circuit (80). These signals are amplified as necessary and given to the control terminals of the corresponding switching elements.

  FIG. 2 is a block diagram illustrating a configuration example of the voltage command generation unit of FIG. 2 includes a speed control unit (12), a current command generation unit (14), an adder (15), a current control unit (16), and a coordinate conversion unit (22, 24, 26, 28), a magnetic pole position estimation unit (34), a speed calculation unit (36), a fluctuation generation unit (40), and an inductance estimation unit (50). The magnetic pole position estimation unit (34), the fluctuation generation unit (40), and the inductance estimation unit (50) operate as a q-axis inductance estimation device.

  FIG. 3 is an explanatory diagram showing the relationship between the magnetic pole position and the coordinate axis of the motor shown in FIG. FIG. 3 shows the α-axis, β-axis, d-axis, q-axis, d ^ -axis, q ^ -axis, and origin (C) that is the rotation axis of the rotor of the motor (4). The α axis and the β axis are stationary and are orthogonal to each other at the origin (C).

The d-axis is an axis from the origin (C) to the north pole of the rotor of the motor (4), and the d-axis is orthogonal to the q-axis at the origin (C). The d-axis indicates the magnetic pole position (θ _e ) of the motor (4). The magnetic pole position (θ _e ) is an angle formed by the α axis and the d axis, and is measured counterclockwise from the α axis. The d-axis rotates counterclockwise at the rotational speed (ω) of the motor (4).

The d ^ axis is the axis from the origin (C) to the estimated N pole of the rotor of the motor (4), and the d ^ axis is orthogonal to the q ^ axis at the origin (C). The d ^ axis indicates the estimated value (θ _e1 ^) of the magnetic pole position of the motor (4). The estimated value (θ _e1 ^) of the magnetic pole position is an angle formed by the α axis and the d ^ axis, and is measured counterclockwise from the α axis. The d ^ axis rotates counterclockwise at the estimated rotational speed (ω ^) of the motor (4).

  In the present specification, the letters α, β, d, q, d ^ or q ^ representing the axis may be attached to symbols representing current, voltage, and the like. The code | symbol which each character attached | subjected shows the component of the direction of the axis which the character represents. The angle is indicated by an electrical angle.

This will be described with reference to FIG. The speed control unit (12) _obtains and outputs a torque command value (T _{m_ref} ) based on the input speed command value (ω _ref ) and the estimated rotational speed (ω ^) of the motor (4). The current command generator (14) _{obtains and outputs} a d-axis current command value (i _{d_ref} ) and a q-axis current command value (i _{q_ref1} ) based on the torque command value (T _{m_ref} ). The fluctuation generation unit (40) generates and outputs a phase (θ _s ) and a current value (i _{q — s} ).

Adder (15), q-axis current command value _(i q_ref1), adds the generated current value fluctuation generator (40) to (i _{Q_s),} q-axis current command value addition result (i _{Q_ref2)} Output as. The current control unit (16) includes a current command value (i _{d_ref,} i _{q_ref2} ), a d-axis current value (i _d ), a q-axis current value (i _q ), and a motor ( Based on the q-axis inductance (L _q ^) in 4), the voltage command value (v _{d_ref,} v _{q_ref} ) is obtained and output.

The coordinate converter (22) uses the estimated value (θ _e1 ^) of the magnetic pole position obtained by the magnetic pole position estimator (34) to _perform a two-phase-three _{operation on the} voltage command value (v _{d_ref} , v _{q_ref} ). Phase conversion is performed, and the obtained three-phase voltage command values (v _{u_ref} , v _{v_ref} , v _{w_ref} ) are output to the PWM modulation section (6). The coordinate conversion unit (24) performs three-phase to two-phase conversion on the detected current value (i _u , i _v , i _w ) of the motor (4) and obtains the obtained current value (i _α , i _β ) is output.

The coordinate conversion unit (26) is obtained by performing three-phase to two-phase conversion on the voltage values (v _u , v _v , v _w ) of the motor (4) output from the PWM modulation unit (6). The voltage value (v _α , v _β ) is output. The coordinate conversion unit (28) performs coordinate conversion from the α-β coordinate to the dq coordinate for the current value (i _α , i _β ) using the estimated value (θ _e1 ^) of the magnetic pole position, The obtained d-axis current value (i _d ) and q-axis current value (i _q ) are output.

Based on the q-axis current value (i _q ), the estimated value of the magnetic pole position (θ _e1 ^), and the phase (θ _s ) generated by the fluctuation generation unit (40), the inductance estimation unit (50) Estimate and output the q-axis inductance (L _q ^) in 4). The magnetic pole position estimator (34) generates a motor based on the current value (i _α , i _β ), voltage value (v _α , v _β ), and q-axis inductance estimated value (L _q ^) of the motor (4). Estimate and output the magnetic pole position (θ _e1 ^) in (4). The magnetic pole position (θ _e1 ^) can be estimated by a known method. Details thereof will be described later. The speed calculation unit (36) calculates and outputs the estimated rotational speed (ω ^) of the motor (4) based on the estimated value (θ _e1 ^) of the magnetic pole position.

  FIG. 4 is a block diagram illustrating a configuration example of the fluctuation generation unit and the inductance estimation unit of FIG. The variation generation unit (40) in FIG. 4 includes an amplitude generation unit (42), a phase generation unit (44), and a calculation unit (46). 4 includes an arithmetic unit (51, 53, 58, 64, 67), an integrating unit (52, 54, 66), a filter (62), and a subtractor (63, 69, 93), an initial value storage unit (68), and a variation calculation unit (92).

  FIG. 5A is a graph showing an example of the estimated value of the magnetic pole position obtained by the magnetic pole position estimating unit (34) of FIG. FIG. 5B is a graph illustrating an example of the output of the phase generation unit in FIG. FIG. 5C is a graph showing an example of the fluctuation component of the estimated value of the magnetic pole position caused by the torque fluctuation.

This will be described with reference to FIGS. 4 and 5A to 5C. The amplitude generation unit (42) generates and outputs a predetermined amplitude value (i _s ). The phase generator (44) generates and outputs a phase (θ _s ) that changes at a constant rate, for example, as shown in FIG. Arithmetic unit (46), the phase values determined from (theta _s) (sin [theta _s) of the amplitude value by multiplying the (i _s), the resulting current value (i _{Q_s)} of FIG. 2 adder (15) Output to. As a result, the q-axis current command value (i _{q_ref2} ) varies, and the torque of the motor (4) varies. That is, the fluctuation generation unit (40) controls the motor (4) so that the torque of the motor (4) fluctuates at a predetermined frequency. The calculation unit (46) outputs the phase (θ _s ) to the calculation units (51, 53, 64).

Arithmetic unit (51), the phase (theta _s) value obtained from the (sin [theta _s) by multiplying the q-axis current value (i _q), and outputs the obtained value. The integration unit (52) integrates the output of the calculation unit (51) and outputs the obtained value (i _{q — sin} ). Arithmetic unit (53), the phase values determined from (theta _s) (cos [theta] _s) by multiplying the q-axis current value (i _q), and outputs the obtained value. The integration unit (54) integrates the output of the calculation unit (53) and outputs the obtained value (i _{q_cos} ). The computing unit (58) calculates the phase difference (δ) between the torque fluctuation of the motor (4) and the component of the predetermined frequency included in the q-axis current.
δ = tan ^-1 (i _{q_sin} / i _{q_cos} )
To obtain and output.

ａ及びｂは実数係数である。 The filter (62) is, for example, a low-pass filter (LPF), and smoothes the estimated value (θ _e1 ^) of the magnetic pole position obtained by the magnetic pole position estimating unit (34), and the obtained result is passed through the filter. Output as magnetic pole position (θ _ef1 ^). The subtracter (63) subtracts the magnetic pole position (θ _ef1 ^) after passing through the filter from the estimated value (θ _e1 ^) of the magnetic pole position, and outputs the obtained difference (Δθ _e1 ^). The difference (Δθ _e1 ^) represents a fluctuation component of the estimated value (θ _e1 ^) of the magnetic pole position caused by the torque fluctuation by the fluctuation generator (40). The processing of the filter (62) is expressed by, for example, z conversion of the equation (1).

a and b are real coefficients.

The inductance estimator (50) _converts the fluctuation component of the estimated value (θ _e1 ^) of the magnetic pole position caused by the torque fluctuation by the fluctuation generator (40) into the motor ( Use after removing the effect of fluctuation of rotation speed in 4). More specifically, the fluctuation calculation unit (92) multiplies the torque command value (T _{m_ref} ) output from the speed control unit (12) by sinθ _s in the same manner as the fluctuation generation unit (40), thereby changing the fluctuation. Determine the torque command value to be used. The fluctuation calculation unit (92) obtains the difference between the current fluctuating torque command value and the fluctuating torque command value sampled immediately before as the torque fluctuation (ΔT).

The fluctuation calculation unit (92) uses the input moment of inertia (J) of the rotor of the motor (4) to calculate the equation of motion,
dω / dt = ΔT / J
From this, the rate of change (dω / dt) in the rotational speed of the motor (4) is obtained. From the obtained value, it can be seen that the rotational speed of the motor (4) changes due to the torque fluctuation by the fluctuation generator (40). The fluctuation calculation unit (92) obtains and outputs the change (Δθ _v ) of the magnetic pole position as the influence of the fluctuation of the rotation speed of the motor (4) by second-order integrating the rate of change (dω / dt) of the rotation speed. . The subtracter (93) subtracts the change (Δθ _v ) of the magnetic pole position from the output (Δθ _e1 ^) of the subtracter (63) and outputs the result to the computing unit (64). As described above, the fluctuation calculation unit (92) and the subtracter (93) can remove the influence of the fluctuation in the rotational speed of the motor (4) caused by the torque fluctuation.

The arithmetic unit (64) multiplies the output of the subtracter (93) by a value (sin (θ _s + δ)) obtained from the phase (θ _s ) and the phase difference (δ), and is generated by torque fluctuation. The component having the same frequency as the torque variation applied to the motor (4) included in the variation component (Δθ _e1 ^) of the estimated value (θ _e1 ^) of the magnetic pole position is obtained and output. The integration unit (66) integrates the output of the calculation unit (64) and outputs the obtained value.

FIG. 6 is a graph illustrating an example of an approximate expression of the q-axis inductance (L _q ). As shown in FIG. 6, the q-axis inductance (L _q ) is the current value (i) (q-axis current value (i _q ) or phase current value (i _u , i _v , i) of the motor (4). _w )) to change. Therefore, the inductance estimation unit (50)
L _q ^ = f1
_{= G * (L q0 + L} q1 * i + L q2 * i 2) (g, L q0, L q1, L q2 is a real number coefficient) ... (approximation formula A)
Is used to change the coefficient g to an appropriate value. The coefficients L _q0 , L _q1 and L _q2 are obtained in advance for the motor (4). The coefficient (L _q0 ) is, for example, the q-axis inductance when the motor (4) is not energized. Hereinafter, an example in which the q-axis current value (i _q ) is used as the current value (i) of the motor (4) will be described.

The initial value storage unit (68) in FIG. 4 stores 1 as the initial value of the coefficient (g) and outputs it. The subtracter (69) subtracts the output of the integration unit (66) from the initial value, and outputs the obtained result as a coefficient (g). The computing unit (67) uses the approximate expression A described above to obtain and output an estimated value (L _q ^) of the q-axis inductance based on the coefficient (g) and the q-axis current value (i _q ).

FIG. 7 is a graph showing another example of the approximate expression of the q-axis inductance (L _q ). The calculation unit (67) is an approximate expression:
L _q ^ = f2
= G * ( _Lq0 + _Lq1 * i) (g, _Lq0 , _Lq1 are real coefficients) ... (approximate equation B)
May be used. The coefficients L _q0 and L _q1 are obtained in advance for the motor (4). In this case, the lower limit value of the expression f2 may be the value (LL).

  The inductance estimating unit (50) may not include the fluctuation calculating unit (92) and the subtracter (93). In this case, the calculation unit (64) uses the output of the subtracter (63).

  FIG. 8 is a flowchart showing an example of processing in the voltage command generator (10) of FIG. The voltage command generator (10) in FIG. 2 performs the following operation.

In block S12, the inductance estimator (50) outputs an initial value of q-axis inductance (for example, L _q0 + L _q1 * i _q + L _q2 * i _q ² ) as an estimated value (L _q ^) of the q-axis inductance.

In block S14, the fluctuation generating unit (40) outputs the current value (i _{q_s} ) to the adder (15) in FIG. 2 to vary the q-axis current command value (i _{q_ref2} ), and the motor (4) To start torque fluctuation. That is, the fluctuation generation unit (40) controls the motor (4) so that the torque of the motor (4) fluctuates at a predetermined frequency. Here, the fluctuation generator (40) fluctuates the torque to such an extent that no large speed fluctuation occurs in the motor (4). For example, the variation rate of the q-axis current value (i _q ) is made sufficiently larger than the variation rate of the estimated rotational speed (ω ^) of the motor (4). The fluctuation generation unit (40) continues such control that causes torque fluctuation in the motor (4) until the process of block S24 described later is performed.

In block S16, the magnetic pole position estimating unit (34) determines the current value (i _α , i _β ), the voltage value (v _α , v _β ), and the estimated value (L _q ^) of the q-axis inductance of the motor (4). The estimated value (θ _e1 ^) of the magnetic pole position of the motor (4) is obtained based on As described above, the current values (i _α , i _β ) are obtained from the phase current values (i _u , i _v , i _w ) of the motor (4).

In block S18, the inductance estimating unit (50) is caused by the torque fluctuation due to variations generator described above (40), fluctuations in the estimated value of the magnetic pole position (theta _e1 ^), i.e., the estimated value of the magnetic pole position (theta _e1 The component of the predetermined frequency (frequency of torque fluctuation) included in ^) is obtained. More specifically, the filter (62) smoothes the estimated value (θ _e1 ^) of the magnetic pole position obtained by the magnetic pole position estimating unit (34), and the result is _{converted to} the magnetic pole position (θ _ef1 ^) after passing through the filter. ) Subtractor (63), the estimated value of the magnetic pole position (theta _e1 ^) pole position after passing through the filter from (theta _ef1 ^) and subtracting the resulting difference of ([Delta] [theta] _e1 ^), caused by the torque fluctuation, Output as a fluctuation component of the estimated value (θ _e1 ^) of the magnetic pole position.

As described above, the fluctuation calculation unit (92) obtains and outputs the change in the magnetic pole position (Δθ _v ) as an influence of the fluctuation of the rotation speed of the motor (4) caused by the torque fluctuation by the fluctuation generation unit (40). . The subtracter (93) subtracts the change (Δθ _v ) of the magnetic pole position from the output (Δθ _e1 ^) of the subtracter (63) and outputs the result to the computing unit (64). Note that the processing of the fluctuation calculation unit (92) and the subtracter (93) may be omitted.

The calculation unit (51, 53) and the integration unit (52, 54) have the same frequency as the torque fluctuation included in the processing of the block S14 described above, which is included in the q-axis current value (torque current value) (i _q ). _Find the components (i _{q_sin} , i _{q_cos} ). The calculation unit (58) obtains the phase difference (δ) of these frequency components with respect to the phase (θ _s ) of torque fluctuation. The calculation unit (64) multiplies the difference (Δθ _e1 ^) by the value (sin (θ _s + δ)), and uses the difference (Δθ _e1 ^) for the component having the same frequency as the torque fluctuation. _{Is obtained} as a frequency component of torque fluctuation included in the estimated value (θ _e1 ^).

実数係数（ｋ）は積分部（66）における積分のゲインを表し、係数（ｋ）は例えば正の値である。このため、差（Δθ_e1^）に含まれる、トルク変動と同じ周波数の成分が正の場合はｑ軸インダクタンス（L_q^）が減少し、この成分が負の場合はｑ軸インダクタンス（L_q^）が増加する。 In block S20, the integration unit (66), the calculation unit (67), and the subtracter (69) cause the q-axis inductance estimated value (L _q ^) so that the same frequency component as the torque fluctuation approaches zero. Update. Specifically, the coefficient g of the above approximate expression A or approximate expression B is updated. The processing of the integration unit (66) is expressed by, for example, z conversion of Expression (2).

The real coefficient (k) represents the gain of integration in the integration unit (66), and the coefficient (k) is, for example, a positive value. Thus, included in the difference ([Delta] [theta] _e1 ^), if the component of the same frequency as the torque fluctuation of a positive decrease q-axis inductance (L _q ^) is, if this component is negative q-axis inductance (L _q ^) Will increase.

In block S22, the inductance estimation unit (50) determines whether or not the update of the estimated value (L _q ^) of the q-axis inductance should be terminated. The inductance estimation unit (50) includes a determination unit, and the determination unit has, for example, an absolute value of a difference between the q-axis inductance (L _q ^) before and after the update in the process of block S20 as a predetermined value. If the condition of smaller is satisfied, it is determined that the update should be terminated. If this condition is not satisfied, the process returns to block S16. In this way, the magnetic pole position estimation unit (34) and the inductance estimation unit (50) repeat the processes of blocks S16 to S20 while updating the q-axis inductance (L _q ^) until a predetermined condition is satisfied, The q-axis inductance (L _q ^) converges.

In block S24, the fluctuation generation unit (40) causes the motor (4) to end the torque fluctuation. Specifically, the fluctuation generating unit (40) stops outputting the current value (i _{q_s} ). q-axis current command value (i _{Q_ref1)} the current value generated by the fluctuation generator (40) by adding the (i _{Q_s),} and hold and torque fluctuations in the motor (4), the efficiency of the motor (4) Because it gets worse.

In block S26, the magnetic pole position estimator (34) obtains an estimated value (θ _e1 ^) of the magnetic pole position using the updated estimated value (L _q ^) of the q-axis inductance. Through the above processing, the q-axis inductance (L _q ^) can be estimated more accurately, and the magnetic pole position (θ _e1 ^) of the motor (4) can be estimated more accurately. The current control unit (16) controls the motor (4) using the estimated value (L _q ^) of the q-axis inductance. For this reason, the current control of the motor (4) can be performed more accurately. In the above processing, the current of the motor (4) need not be zero.

  In addition, when the rotational speed of the motor (4) changes by a predetermined value or more, or when the load of the motor (4) changes by a predetermined value or more, the processing of the blocks S14 to S26 may be resumed.

A method of estimating the magnetic pole position (θ _e1 ^) in the magnetic pole position estimation unit (34) will be described. Equation (3) represents a voltage equation on the α-β coordinate. Here, R _a is an armature resistance, L _d is a d-axis inductance, θ _e is a magnetic pole position, p is a time differential operator, and φ _a is a field main magnetic flux.

If λ _α and λ _β are defined by equation (4), the magnetic pole position (θ _e ) can be expressed by equation (5), and equation (3) can be transformed into equation (6). By substituting this into equation (5), the magnetic pole position (θ _e ) can be obtained. Such a technique is described in Non-Patent Document 1, for example.

As can be seen from the equation (6), λ _α and λ _β are observable quantities (i _α , i _β , v _α , v _β ), and armature resistance (R _a ) and q-axis inductance, which are device constants. Determined using (L _q ). For this reason, the magnetic pole position (θ _e ) obtained by Equation (5) also depends on the armature resistance (R _a ) and the q-axis inductance (L _q ). However, in a somewhat high speed operation, the voltages (v _α , v _β ) are large and R _a i _α and R _a i _β can be ignored. Therefore, the magnetic pole position (θ _e ) depends only on the q-axis inductance (L _q ) among the device constants. Therefore, if the estimated value (L _q ^) of the q-axis inductance is known, by adopting this as the q-axis inductance (L _q ) of the equation (6), the magnetic pole position (θ _e ) of the equation (5) Can be adopted as the magnetic pole position (θ _e1 ^).

As a method of estimating the magnetic pole position (θ _e1 ^) in the magnetic pole position estimation unit (34), the following method may be used. Equation (7) represents the voltage equation on the d ^ -q ^ coordinates. Here, θ _err is an _error of the estimated axes (d ^ axis and q ^ axis) with respect to the d axis and the q axis, ω is an angular frequency of the d axis and the q axis, and ω ^ is an angular frequency of the estimated axis.

Assuming that the angular frequency and the current are constant, an approximation that ignores the term including the differential operator (p) is applied to Equation (7). Under this approximation, Equation (7) is solved for the axial error (θ _err ) to obtain Equation (8). The angular frequency (ω ^) of the estimated axis is adjusted using, for example, PLL control so that the axis error (θ _err ) becomes zero. Then, the estimated value (θ _e1 ^) of the magnetic pole position is obtained by integrating with respect to time using the adjusted angular frequency (ω ^) of the estimated axis. Such a technique is described in Non-Patent Document 2, for example.

Similarly to the case of _obtaining the magnetic pole position (θ _e1 ^) using the equations (3) to (6), the term including the armature resistance (R _a ) can be ignored, and the estimated value of the q-axis inductance (L _q If ^) is known, this is adopted as the q-axis inductance (L _q ) in Equation (8), whereby the magnetic pole position (θ _e ) obtained by the method is adopted as the magnetic pole position (θ _e1 ^). be able to.

FIG. 9 is a diagram illustrating an example of a limit range of the d-axis current value. The power conversion device (100) of FIG. 1 uses the fact that the error in the magnetic pole position changes due to the change in the q-axis current when there is an error in the q-axis inductance (L _q ), thereby making the q-axis inductance (L _q ). Therefore, so as not to cancel the error change in the magnetic pole position caused by a change in the q-axis current (i _q) by a change in the d-axis current (i _d), to limit the d-axis current (i _d) as shown in FIG. 9 May be.

As an example, q-axis inductance error (ΔL _q ) is positive, and q-axis current value change (Δi _q ) when q-axis current value (i _q ) changes from i _q1 to i _q2 is positive Will be described. If the magnetic pole position error when the q-axis current value (i _q ) is i _q1 is θ ₁ , the q-axis current value (i _q ) increases and the magnetic pole position error changes to the negative side. Here, for example, the magnetic pole position estimator (34) uses the d-axis current value (i _d ) at which the magnetic pole position error is θ ₁ when the q-axis current value (i _q ) is i _q2 as the d-axis current. Obtained as the limit value (I _d _lim).

Since the d-axis current value (i _d) is negative, in order to change the error of the magnetic pole position caused by a change in the q-axis current (i _q) is not canceled out, the d-axis current value (i _d) is It is necessary to set a value larger than the d-axis current limit value (I _d _lim) (that is, a value closer to zero than the d-axis current limit value (I _d _lim)). Therefore, the current control unit (16) performs control so that the d-axis current value (i _d ) is larger than the d-axis current limit value (I _d _lim). The range of the d-axis current value (i _d ) for other combinations of the q-axis inductance error (ΔL _q ) and the q-axis current value change (Δi _q ) shown in FIG. Can do.

Since the d-axis current of the motor (4) is limited as shown in FIG. 9, the fluctuation of the estimated value (θ _e1 ^) of the magnetic pole position caused by the torque fluctuation by the fluctuation generator (40) is caused by the change of the d-axis current. It will not be completely countered. For this reason, the estimated value (L _q ^) of the q-axis inductance can be reliably obtained based on this variation.

Note that, when estimating the q-axis inductance, the position fluctuation is obtained by integrating for a predetermined time, so there is no problem even if the d-axis current value (i _d ) exceeds the limit of the d-axis current for a minute time.

  Next, a modified example will be described. FIG. 10 is a block diagram showing a configuration of a modification of the voltage command generation unit of FIG. The voltage command generation unit (210) of FIG. 10 is used in place of the voltage command generation unit (10) of FIG. 2 in the power conversion device (100) of FIG. The voltage command generation unit (210) of FIG. 10 includes a current command generation unit (214) and an inductance estimation unit (250) instead of the current command generation unit (14) and the inductance estimation unit (50), and an adder ( Although the point which does not have 15) and a fluctuation | variation production | generation part (40) differs from the voltage command production | generation part (10) of FIG. 2, other points are comprised similarly. The current command generation unit (214), the magnetic pole position estimation unit (34), and the inductance estimation unit (250) operate as a q-axis inductance estimation device, and the current command generation unit (214) also operates as a fluctuation generation unit.

  FIG. 11 is a block diagram showing a configuration of a modification of the inductance estimation unit of FIG. The inductance estimation unit (250) in FIG. 11 includes a calculation unit (251, 253, 264) instead of the calculation unit (51, 53, 64), and further includes a calculation unit (256). It is different from 50), but the other points are configured similarly.

  FIG. 12A is a graph showing an example of the estimated value of the magnetic pole position obtained by the magnetic pole position estimating unit in FIG. FIG. 12B is a graph showing an example of the phase used in the inductance estimation unit of FIG. FIG. 12C is a graph showing an example of the fluctuation component of the estimated value of the magnetic pole position caused by the torque fluctuation.

  Here, it is assumed that the capacitor (78) of FIG. 1 has a relatively small capacity. In such a case, the direct current obtained by rectifying the alternating current supplied by the alternating current power supply (2) pulsates, and the voltage between the node (N1) and the node (N2) that is the input voltage of the inverter circuit (80) Pulsates. In the case of FIG. 1, this voltage pulsates at a frequency twice that of the AC frequency supplied by the AC power source (2) of FIG.

As described above, the inverter circuit (80) drives the motor (4) using the direct current (pulsating flow) including the alternating current component of the alternating current power source (2), and the current command generation unit (214) A d-axis current command value (i _{d_ref} ) and a q-axis current command value (i _{q_ref} ) are obtained and output based on T _{m_ref} ) and the power supply phase (θ _in ). At this time, the d-axis current command value (i _{d_ref} ) and the q-axis current command value (i _{q_ref} ) fluctuate at twice the frequency of the AC supplied from the AC power supply (2). For this reason, the motor (4) is controlled so that the torque fluctuates at a frequency twice as high as the AC frequency supplied by the AC power supply (2). That is, the current command generator (214) controls the motor (4) so that torque fluctuation occurs in synchronization with the pulsation of the input voltage of the inverter circuit (80). The frequency of torque fluctuation is twice the frequency of the alternating current supplied by the alternating current power source (2).

This will be described with reference to FIGS. 11 and 12A to 12C. Instead of the phase (θ _s ), the power source phase (θ _in ) is input to the inductance estimation unit (250). The power supply phase (θ _in ) is the phase of the AC power supply (2) in FIG. The power conversion device (100) may have a zero-cross detection circuit, and the power supply phase (θ _in ) can be obtained by this zero-cross detection circuit, for example.

The calculation unit (256) doubles the power supply phase (θ _in ) and outputs the obtained phase (2θ _in ) to the calculation unit (251,253,264). Calculation unit (251) is a phase (2 [Theta] _in) value obtained from (Sin2shita _in) by multiplying the q-axis current value (i _q), and outputs the obtained value. Calculation unit (253) is a phase (2 [Theta] _in) value obtained from (cos _in) by multiplying the q-axis current value (i _q), and outputs the obtained value.

The calculation unit (264) calculates the difference (Δθ _e1 ^) output from the subtracter (63) from the value (−cos (2θ _in + δ)) obtained from the phase (2θ _in ) and the phase difference (δ). To obtain a component having the same frequency as the torque fluctuation of the motor (4) included in the difference (Δθ _e1 ^) and output it. The other components of the inductance estimation unit (250) are the same as those of the inductance estimation unit (50) of FIG.

According to this modification, it can be used that the torque fluctuation of the motor (4) is twice the frequency of the alternating current supplied by the alternating current power source (2), and the voltage command generator (10 4), it is not necessary to generate the phase (θ _s ) for generating the torque fluctuation, so that the fluctuation generating section (40) as shown in FIG. 4 is not necessary. For this reason, the structure of an apparatus can be simplified.

  FIG. 13 is a block diagram illustrating a configuration of a modification of the power conversion device of FIG. 1. In the power converter (300) of FIG. 13, the voltage command generator (10) of FIG. 2 or the voltage command generator (210) of FIG. 10 may be used. The power conversion device (300) includes a PWM modulation unit (306), a voltage command generation unit (10, 210), a reactor circuit (372), a capacitor circuit (378), and a matrix circuit (380). The power converter (300) directly converts AC power supplied from the three-phase AC power source (302) into three-phase AC power, and drives the motor (4).

The matrix circuit (380) has nine switching circuits (87), and the drive current (IU, IV, IW) of the motor (4) from the alternating current supplied from the alternating current power supply (302) according to the drive signal (DS). Is generated and supplied to the motor (4). Based on the voltage command values (v _{u_ref} , v _{v_ref} , v _{w_ref} ) generated by the voltage command generation unit (10 or 210), the PWM modulation unit (306) is _configured with each of the drive signal (DS) and the motor (4). The phase voltage values (v _u , v _v , v _w ) are obtained and output. The drive signal (DS) includes a plurality of signals that respectively control the switching circuits (87) of the matrix circuit (380). In this manner, the voltage command generation unit (10, 210) can also be used in the matrix circuit (380) of FIG.

  FIG. 14 is a circuit diagram showing an example of the switching circuit (87) of the matrix circuit (380) of FIG. The circuit of FIG. 14 has two transistors (88) and two diodes (89). FIG. 15 is a circuit diagram showing another example of the switching circuit (87) of the matrix circuit (380) of FIG. The circuit of FIG. 15 has one transistor (88) and four diodes (89).

  Each functional block in this specification may typically be realized by hardware. Alternatively, some or all of each functional block can be implemented in software. For example, such a functional block can be realized by a processor and a program executed on the processor. In other words, each functional block described in the present specification may be realized by hardware, may be realized by software, or may be realized by any combination of hardware and software.

  The above embodiments are essentially preferable examples, and are not intended to limit the scope of the present invention, its application, or its use.

  As described above, the present invention is useful for a q-axis inductance estimation method and estimation apparatus, a motor control apparatus for estimating a magnetic pole position, and the like.

4 Motor
10,210 Voltage command generator
34 Magnetic pole position estimation unit
40 Fluctuation generator
50,250 Inductance estimation unit
80 Inverter circuit
214 Current command generator (variation generator)
L _q ^ q-axis inductance estimated value θ _e magnetic pole position θ _e1 ^ magnetic pole position estimated value
i _u , i _v , i _w Motor phase current value

Claims (8)

(A) Control the motor (4) so that torque fluctuation occurs,
(B) The estimated value (θ of the magnetic pole position of the motor (4) based on the current value (i _u , i _v , i _w ) of the motor (4) and the estimated value of the q-axis inductance (L _q ^) _e1 ^)
(C) updating the estimated value (L _q ^) of the q-axis inductance based on the fluctuation of the estimated value (θ _e1 ^) of the magnetic pole position caused by the torque fluctuation; Estimation method. In claim 1,
The inverter circuit (80) that drives the motor (4) receives a pulsating direct current obtained by rectifying the alternating current supplied by the alternating current power supply,
In the process (a), the q-axis inductance estimation method is characterized in that the torque fluctuation is controlled in synchronization with the pulsation of the input voltage of the inverter circuit (80). In claim 2,
The q-axis inductance estimation method, wherein the frequency of torque fluctuation is twice the frequency of the AC supplied from the AC power supply. In claim 1,
In the process (c) above,
_Find the component of the same frequency as the torque fluctuation in the estimated value (θ _e1 ^) of the magnetic pole position,
A method for estimating a q-axis inductance, comprising updating the estimated value (L _q ^) of the q-axis inductance so that a component having the same frequency as the torque fluctuation approaches zero. In any one of Claims 1-4,
A method of estimating a q-axis inductance, wherein the processes (b) and (c) are repeated until a predetermined condition is satisfied. In claim 1 or 4,
In the processing of (c), the estimated value (θ _e1 ^) of the magnetic pole position is smoothed, and the difference between the result (θ _ef1 ^) and the estimated value of the magnetic pole position (θ _e1 ^) is _calculated as the torque fluctuation. The q-axis inductance estimation method is characterized in that it is obtained as a fluctuation in the estimated value (θ _e1 ^) of the magnetic pole position caused by the above. In claim 1,
In the process of (c), the fluctuation component of the estimated value (θ _e1 ^) of the magnetic pole position caused by the torque fluctuation is considered to be the influence of the fluctuation of the rotational speed of the motor (4) caused by the torque fluctuation. A method for estimating q-axis inductance, which is used after being removed. A fluctuation generator (40, 214) that controls the motor (4) so that torque fluctuations occur;
Based on the current value (i _u , i _v , i _w ) of the motor (4) and the estimated value (L _q ^) of the q-axis inductance, the estimated value (θ _e1 ^) of the magnetic pole position of the motor (4) A magnetic pole position estimation unit (34) for obtaining
An inductance estimation unit (50; 250) that updates the estimated value (L _q ^) of the q-axis inductance based on the variation of the estimated value (θ _e1 ^) of the magnetic pole position caused by the torque variation. An apparatus for estimating q-axis inductance.

Images