CN1667942A - Rotor position presuming method and apparatus, motor control method, compressor and program - Google Patents

Abstract

The invention provided a rotor position estimating method for easily designing a controller and easily obtaining stability. The rotor position estimating method, for estimating a rotor position [theta] of a permanent magnet motor 20, finds a magnetic flux error between a rotational coordinate axis and its estimated axis, and estimates the rotor position based on the magnetic flux error. The magnetic flux error is preferably the magnetic flux error between a q-axis as the rotational coordinate axis and a [delta]-axis as its estimated axis. A magnetic flux comprises an integration of a difference of a voltage drop Ri[gamma] in a current from a voltage V[gamma] applied to the motor. A magnetic flux [phi][delta] comprises a product of an inductance Lq of a motor winding and a current i[delta]. The magnetic flux error is found by estimating the magnetic fluxes. The magnetic flux error between the rotational coordinate axis and its estimated axis is found on the basis of a magnetic flux equation formed by dividing a voltage equation of the permanent magnet motor by an angular speed [omega] of the rotor. The rotor position is estimated on the basis of the magnetic flux error.

Description

Rotor position presuming method and device, method of motor control, compressor and program

Technical field

Control method, compressor and the program of relevant permanent magnet motor rotor position presuming method of the present invention and device, motor.

Background technology

Have in the permanent magnet motor (DC brushless motor, IPM motor) of salient pole, can be according to the voltage that drives the inverter side acquisition, current signal etc., calculate rotor-position, need not physically position transducer, velocity transducer control motor, promptly adopt so-called no transducer control technology.

As no transducer control method, known have square wave do not have transducer control method, and sinusoidal wave no transducer control method.It is a kind of method that detects induced voltage that square wave does not have transducer control method, though established technically, this method only limits to the rectangular wave drive that drive waveforms is the center with 120 ° of drivings.

Have a kind of demand also quite big at present, that is a target with low noise, high-frequency exactly, wants the drive waveforms of IPM motor is adopted sine wave, drives with position-sensor-free in addition, guarantees reliability, and also reduces cost.Sinusoidal wave no transducer control method is by detecting electric current, calculating induced voltage and come estimated position.

In 180 ° of sinusoidal wave no transducers controls, employing is the d-q axis coordinate system that d axle and begin from the d axle is formed along the q axle on the direction of direction of rotation quadrature by the position of p-m rotor flow direction.The voltage equation of motor model is shown in following formula (1) in the actual rotating coordinate system of this d-q axle.

Formula (1)

$\left(\begin{array}{c}{V}_d\\ V_q\end{array}\right)=\left(\begin{array}{cc}R+p{L}_d& {\mathrm{\&omega;}}_{\mathrm{re}}{L}_q\\ {\mathrm{\&omega;}}_{\mathrm{re}}{L}_d& R+p{L}_q\end{array}\right)\left(\begin{array}{c}{i}_d\\ i_q\end{array}\right)+{\mathrm{\&omega;}}_{\mathrm{re}}{k}_{\mathrm{<figure-callout\; id="0"\; label="E"\; filenames="A20051000706500231.png"\; state="\{\{state\}\}">E</figure-callout>}}\left(\begin{array}{c}0\\ 1\end{array}\right)$

Ke is the induced voltage constant in the formula.

The motor equivalent electric circuit equation of d axle component is considered the generation offset when position-sensor-free, consider induced voltage component Ed as described below, just forms following formula (2).

Formula (2)

V _d＝R·I _d+p(L _d·I _d)-ω _reL _q·I _q+E _d

Utilize above-mentioned formula (2), induced voltage Ed can formula described as follows (3) represent like that.

Formula (3)

E _d＝V _d-R·I _d-p(L _d·I _d)+ωr _eL _q·I _q

Former what submit to is in the data C5-1-5 page or leaf (non-patent literature 1) of topic with " the high-performance motor technology of air-conditioning " in the pass at the production technology center of (strain) Toshiba in technology frontier (Technofrontier) the discussion 2003 sub-venue C-5 motor technology discussion that JMA of civic organization sponsors, and has put down in writing decision angular speed (motor speed) and has made the induced voltage Ed=0 suitable with the estimation error of rotor-position and this value integration is inferred the content of angle (rotor-position).

Technology frontier discussion 2003 sub-venue C-5 motor technology discussion on April 18th, 2003 that a former clever exercise question is sponsored for the JMA of data civic organization of " the high-performance motor technology of air-conditioning " is closed in (non-patent literature 1) (strain) Toshiba

Summary of the invention

But the value of induced voltage Ed for being directly proportional with angular velocity omega (rotating speed).Be induced voltage Ed because and the estimation error of position become the function of rotating speed together, so the problem that exists is that it is complicated that the such design of Controller of axle offset adjuster just becomes, extremely difficult steady running in the operating range of broad.Especially in low regime, owing to reasons such as computational accuracies, the position deduction low precision.

The present invention's purpose is to provide a kind of design of Controller also to obtain control method, compressor and the program of stable rotor position presuming method and device, motor easily easily.

The application's rotor position presuming method is a kind of rotor position presuming method of inferring the permanent magnet motor rotor-position, obtains the magnetic flux error that rotatable coordinate axis and described rotatable coordinate axis are inferred between centers, according to described magnetic flux error, infers described rotor-position.

The application's rotor position presuming method is a kind of rotor position presuming method of inferring the permanent magnet motor rotor-position of salient pole, infer to the difference of the voltage drop that deducts electric current from the motor applied voltage carry out integration and magnetic flux and the magnetic flux of the inductance of motor winding and the long-pending generation of electric current, according to described magnetic flux of inferring, obtain rotatable coordinate axis and described rotatable coordinate axis and infer the magnetic flux error of between centers, according to described magnetic flux error, infer described rotor-position.

The application's rotor position presuming method is a kind of rotor position presuming method of inferring the permanent magnet motor rotor-position of salient pole, remove the magnetic flux equation that described permanent magnet motor voltage equation gets according to angular velocity omega with rotor, obtain rotatable coordinate axis and described rotatable coordinate axis and infer the magnetic flux error of between centers, according to described magnetic flux error, infer described rotor-position.

In the application's the rotor position presuming method, described magnetic flux error be described rotatable coordinate axis be q axle and described rotatable coordinate axis infer the axle be the magnetic flux error of δ between centers.

The application's rotor position presuming method is a kind of rotor position presuming method of inferring the permanent magnet motor rotor-position of salient pole, adopt the approximate expression of difference Δ θ ≈ sin Δ θ of the rotor position angle of the voltage equation of described permanent magnet motor and actual motor and motor model, do not adopt the Δ θ=tan that launches according to described voltage equation simultaneously ^-1The formula of form is obtained the Δ θ of the angular velocity omega that does not depend on rotor, according to the described Δ θ that tries to achieve, infers rotor-position.

Δ θ=tan ^-1The molecule of form, denominator directly do not carry out computing as if not dealing with, and then operand is big, is disadvantageous for the control of carrying out Δ θ 0 with real-time mode.According to above-mentioned the application, can relatively reduce the calculating required time.

In the application's the rotor position presuming method, described magnetic flux error can utilize following formula (4) to try to achieve.

Formula (4)

$\frac{V_{\mathrm{\&gamma;}}-R\&CenterDot;i_{\mathrm{\&gamma;}}}{{\mathrm{\&omega;}}_{\mathrm{re}}}+L_q\&CenterDot;i_{\mathrm{\&delta;}}$

V γ: the γ axle component of armature voltage

R: armature winding resistance

I γ: the γ axle component of armature supply

ω re: the presumed value of rotor velocity or command value (electrical degree)

Lq:q axle inductance

I δ: the δ axle component of armature supply

In the application's the rotor position presuming method, described magnetic flux error can be tried to achieve with following formula (5).

Formula (5)

$\frac{V_{\mathrm{\&gamma;}}-(R+p{L}_d)\&CenterDot;i_{\mathrm{\&gamma;}}}{{\mathrm{\&omega;}}_{\mathrm{re}}}+L_q\&CenterDot;i_{\mathrm{\&delta;}}$

V γ: the γ axle component of armature voltage

R: armature winding resistance

P: differential operator

Ld:d axle inductance

I γ: the γ axle component of armature supply

ω re: the presumed value of rotor velocity or command value (electrical degree)

Lq:q axle inductance

I δ: the δ axle component of armature supply

In the application's the rotor position presuming method, described inductance is for depending on the function of the some amounts in electric current and the rotating speed at least.

The application's method of motor control, be to try to achieve the corresponding described rotatable coordinate axis of the described magnetic flux error obtained with the rotor position presuming method that utilizes above-mentioned the application and the presumed value of inferring the corresponding rotor velocity of between centers site error of described rotatable coordinate axis, presumed value input low pass filter with described rotor velocity, according to the output valve of described low pass filter, carry out and the relevant FEEDBACK CONTROL of described permanent-magnet electric motor speed.

The application's method of motor control is a kind of method of motor control of using described the application's rotor position presuming method, it will make that the error of inferring corresponding value of the detected value of a component and command value with the described rotatable coordinate axis of armature supply is a phase place command value input low pass filter for the output of the current controller of small incidental expenses, according to the output valve and the described rotor-position of inferring of described low pass filter, generate the signal of the phase place of expression voltage instruction.

The application's method of motor control is a kind of method of motor control of using described the application's rotor position presuming method, it infers the corresponding value of detected value of a component to the low pass filter input and the described rotatable coordinate axis of armature supply, according to the deviation of the command value of inferring a component of the rotatable coordinate axis of the output valve of described low pass filter and described armature supply, the command value of formation voltage phase place.

The application's method of motor control is the method for motor control that a kind of control has the permanent magnet motor of salient pole, it obtain rotatable coordinate axis be the q axle to infer axle be the magnetic flux of δ axle, control the magnetic flux that makes described δ axle and converge zero.

In the application's the method for motor control, the magnetic flux of its described δ axle can be to deduct magnetic flux and the inductance L q that the difference of the voltage drop of resistance R and electric current I γ calculates from the presumed value V γ ^ of motor applied voltage and try to achieve with the form of amassing the magnetic flux sum that obtains of electric current I δ by removing with angular velocity omega re^.

In the application's the method for motor control, the magnetic flux of its described δ axle can be to deduct magnetic flux that difference that the voltage drop of resistance R and electric current I γ and inductance L d and I γ change the voltage drop of generation in time calculates and try to achieve with utilizing the form of amassing the magnetic flux sum that obtains of inductance L q and electric current I δ by removing with angular velocity omega re^ from the presumed value V γ ^ of motor applied voltage.According to the application, can improve response characteristic.

In the application's the method for motor control, its described δ axle magnetic flux can be to deduct the voltage drop of resistance R and electric current I γ and positive gain constant K according to removing with angular velocity omega re^ from the presumed value V γ ^ of the motor applied voltage " magnetic flux of calculating after the difference of the long-pending voltage drop that produces that changes in time with electric current I γ and try to achieve according to inductance L q and the form of amassing the magnetic flux sum that obtains of electric current I δ.According to the application,,, system simplifies thereby constituting by not using the such motor constant of Ld." when being zero, just identical in addition at K with foregoing invention.

Method of motor control of the present invention is a kind of method of motor control that does not have the sensor drive permanent magnet motor, it calculates the magnetic flux error, adjust the rotating speed of described permanent magnet motor, the magnetic flux error convergence that makes described computing is zero, by speed estimating value integration to obtaining, the magnetic flux error that makes described computing is zero, thereby calculates the rotor-position of described permanent magnet motor.

The described permanent magnet motor that adopts above-mentioned the application's rotor position presuming method to infer described rotor-position is suitable as the air compressor motor of compressor.Described compressor is applicable to air-conditioning equipment.

Utilize the described permanent magnet motor of described the application's method of motor control control to be suitable as the air compressor motor of compressor, described compressor is applicable to air-conditioning equipment.

Thereby having by calculating the magnetic flux error and controlling, the application's compressor makes that the described magnetic flux error convergence that calculates is the zero air compressor motor that can carry out the position-sensor-free sine wave drive.

Described the application's compression function is applicable to air-conditioning equipment.

The application's program is the program that each step of allowing computer carry out above-mentioned the application's rotor position presuming method is used.

The application's rotor position presuming device is a kind of rotor position presuming device of inferring the permanent magnet motor rotor-position of salient pole, it is the magnetic flux error of inferring between centers of trying to achieve rotatable coordinate axis and described rotatable coordinate axis, according to described magnetic flux error, infer described rotor-position.

According to the application, design of Controller is easy, obtains the good stable characteristic easily.And, the position deduction precision of raising low-speed region.

Description of drawings

Fig. 1 is the computing block diagram during no transition item in the control device of electric motor site error estimator that adopts rotor position presuming method one example of the present invention.

Computing block diagram when in the control device of electric motor site error estimator of Fig. 2 for employing rotor position presuming method one example of the present invention the transition item being arranged.

Fig. 3 adopts the formation block diagram of the control device of electric motor of rotor position presuming method one example of the present invention for expression.

Fig. 4 adopts other formation block diagram of the control device of electric motor of rotor position presuming method one example of the present invention for expression.

Fig. 5 adopts another other of the control device of electric motor of rotor position presuming method one example of the present invention to constitute block diagram for expression.

Fig. 6 is the illustraton of model of the no transducer control of expression usefulness.

Label declaration

10 control device of electric motor

11 adders

12 speed controls

13 adders

14 current controllers

15 voltage generators

16 voltage compensators

17 PWM inverters

21 site error estimators

22 inductance compensation devices

24 speed estimating devices

26 integrators

27 coordinate transform operational parts

28 coordinate transform operational parts

31?LPF

32?LPF

33?LPF

Embodiment

Below, with reference to accompanying drawing detailed description the application's rotor position presuming method one example.

This enforcement forms 180 ° of sinusoidal wave no transducer control technologys of the permanent magnet motor (DC brushless motor, IPM motor) that includes salient pole.As shown in Figure 6, in 180 ° of sinusoidal wave no transducer controls, obtaining by the position on the p-m rotor flow direction is the axle offset Δ θ of d axle and actual rotating coordinate system of d-q axle that leading 90 ° q axle is formed from the d axle along direction of rotation and the imaginary rotor-position γ axle of controlling and γ-δ between centers that leading 90 ° δ axle is formed from the γ axle along direction of rotation, controls this Δ θ and is zero.

The angle θ of rotor (rotor-position) is owing to be amount to the angular velocity omega integration of rotor, so when control position skew (Δ θ) is zero, according to Δ θ, ask to make Δ θ become zero ω (the speed estimating device 24 among aftermentioned Fig. 3), according to this ω control motor.

In this example, as described later, according to the magnetic flux error of inferring out, calculate the offset Δ θ of q axle and δ axle, when the phase place of δ axle is leading phase, reduce the angular speed of inferring on the δ axle, when the phase place on the δ axle is lagging phase, increase the angular speed of inferring on the δ axle, make consistent with corresponding position θ and the q axle of this Δ θ integrated value.

Here, when Δ θ was obtained as the value of be directly proportional with ω (depending on ω), the design of Controller of back level one side of input Δ θ became difficult.In the processes such as gain design of controller, depend on very much the tendency of rotating speed (ω) owing to exist, so the problem that has gain adjustment to become difficult, the convergence state of the presumed value of Δ θ depends on speed.In addition, in this adjustment, also to come by experiment relatively to increase complexity.Thus, when Δ θ obtains as the value that depends on ω, be difficult to guarantee stability, can produce departure again.

On the contrary, in the site error estimator of this example (with reference to the label 21 of Fig. 3), as described later, owing to be not to utilize the induced voltage vector, but utilize magnetic flux vector, so Δ θ can be found the solution as the value that does not depend on ω.Thus, in this example, the design of the controller (speed estimating device 24) of input Δ θ is (available general PI controller usually) just easily, can guarantee enough stability, can prevent the generation of departure again.

Because the magnetic flux of permanent magnet is an induced voltage to the integration of time, so about the inferring of magnetic flux vector, its position is for beginning to lag behind 90 ° position from the induced voltage vector, in addition, it is the magnetic flux (physical quantity) at center that size becomes with the permanent magnet.Thereby, adopt not the magnetic flux vector of the physical quantity that changes because of rotating speed, by inferring the absolute position on electric, just can directly infer the q axle.The result, because of the design of the estimator of calculation speed with rpm-dependent magnetic flux error not as benchmark, so the adjustment of parameter is easy, the result the output of speed control, and the calculating of estimated position on, stability improves, and the driving of motor is stable easily.

The control device of electric motor of the rotor position presuming method that adopts this example is described with reference to Fig. 3.

Among Fig. 3,10 pairs of control device of electric motor have the permanent magnet motor 20 of salient pole not have sensor drive.Control device of electric motor 10 comprises PWM inverter 17, current detector (not shown), coordinate transform operational part 27 and 28, site error estimator 21, speed estimating device 24, integrator 26, speed control 12, current controller 14, voltage generating unit 15, reaches voltage compensator 16.

PWM inverter 17 becomes three-phase alternating voltage with dc voltage conversion.

Current detector (not shown) detects the current i (u, w) of motor 20.

Coordinate transform operational part 27 with its detected current i (u, w) in the enterprising line translation of rotational coordinates.

Site error estimator 21 is inferred the site error Δ θ ^ of rotor.

Speed estimating device 24 is inferred angular velocity omega re^, and making its site error Δ θ ^ that infers is zero.

The output ω re of 26 pairs of these speed estimating devices 24 of integrator " carries out integration, calculates rotor position (γ, δ) ^.

Speed control 12 is used to infer rotor-position and makes speed value ω re ^*With the error of the output ω re^ of speed estimating device 24 be zero.

Current controller 14 is used to make the command value i γ of γ axle (d axle) electric current ^*And the error between the γ shaft current i γ that tries to achieve according to the current information that actual detected goes out is zero.

Voltage instruction value V (m) according to speed control 12 outputs ^*, and the phase place command value V β of current controller 14 output ^*, calculate PWM output valve V (u, v, w) ^*

Voltage generating unit 15 is according to the voltage amplitude instruction V (m) of speed control 12 outputs ^*, and the phase place V (θ) of the voltage instruction of adder 18 output ^*, generate voltage instruction V (u, v, w) to 17 outputs of PWM inverter.

The voltage instruction value V of the input PWM inverter 17 that voltage compensator 16 will be generated by voltage generator 15 (u, v) ^*As input, output (u, is v) carried out voltage presumed value Vmd (u, v) ^ behind phase place, the correction of amplitude for this value V.

(u, v) ^ is in the enterprising line translation of rotational coordinates with the voltage presumed value Vmd of voltage compensator 16 output for coordinate transform operational part 28.

Fig. 3 to Fig. 5 represents that the difference of the control device of electric motor 10 of this example constitutes example.The difference of Fig. 3, Fig. 4, Fig. 5 is different because of the stability raising that makes which output, makes its stable control output insert low pass filter (LPF) for wanting.By inserting LPF, though also some reduction a little of response characteristic sometimes under the situation of preferentially guaranteeing stability, is inserted LPF.Adopt any formation among Fig. 3～Fig. 5 actually, will be different because of the feature of the feature of motor or load.Except the formation about this LPF, remaining system constitutes all identical.Below, with reference to the formation that Fig. 3 illustrates control device of electric motor 10, select need with reference to Fig. 4 or Fig. 5.

Command value to control device of electric motor 10 is angular velocity omega re ^*, and γ shaft current I γ ^*

Adder calculator 11 calculates angular speed command value ω re ^*, and angular speed presumed value ω re^ between deviation.The speed control 12 that this deviation input is made of PI (proportional integral) controller.The output order of speed control 12 outputs is voltage amplitude instruction V (m) ^*This voltage amplitude instruction V (m) ^*Amplitude instruction for the three-phase command voltage of motor 20.

Adder 13 is calculated γ shaft current command value I γ ^*, and according to motor current detect, the deviation of the I γ of computing.The current controller 14 that this deviation input is made of the PI controller.The output order of current controller 14 outputs becomes the command value V β of voltage-phase ^*The command value V β of this voltage-phase ^*Phase bit instruction for the three-phase command voltage of motor 20.As shown in Figure 5, for making motor 20 stable, make this voltage-phase command value V β ^*Output by digital LPF31, reduce oscillating component etc., can make the command voltage phase stabilization.

Adder 18 is tried to achieve the phase bit instruction V β of voltage ^*With the phase place V (θ) of the rotor position of inferring (γ δ) ^ sum as voltage instruction ^*The phase place V of this voltage instruction (θ) ^*Input voltage generating unit 15.In this voltage generating unit 15, for example generate following command voltage waveform.

Formula (6)

$V_u=V{\left(m\right)}^{*}\&CenterDot;\sin (V{\mathrm{\&beta;}}^{*}+\widehat{\mathrm{\&theta;}}\left(\mathrm{\&gamma;\&delta;}\right))$

$V_v=V{\left(m\right)}^{*}\&CenterDot;\sin ({\mathrm{V\&beta;}}^{*}+\widehat{\mathrm{\&theta;}}\left(\mathrm{\&gamma;\&delta;}\right)+2/3\&CenterDot;\mathrm{\&pi;})$

$V_w=V{\left(m\right)}^{*}\&CenterDot;\sin (V{\mathrm{\&beta;}}^{*}+\widehat{\mathrm{\&theta;}}\left(\mathrm{\&gamma;\&delta;}\right)+4/3\&CenterDot;\mathrm{\&pi;})$

To the such voltage instruction V (u, v, w) of PWM inverter 17 outputs.PWM inverter 17 is made of inverter circuit etc., generates the PWM waveform.Because available general PWM inverter commonly used in the past is as this PWM inverter 17, its explanation is omitted.

When driving IPM motor 20, detect the phase current of motor 20 with PWM inverter 17 reality.This testing circuit utilizes operational amplifier to constitute amplifying circuit as the secondary side at CT with CT etc., then can easily obtain phase current is transformed into the value (waveform) of voltage signal.The waveform of this motor phase current is owing to be the analogue value, thus utilize AD converter etc. that it is transformed into digital value, be transformed into can computing value.

Because of the phase current iu of this motor, iw is the electric current from rest frame again, so will it be carried out coordinate transform with coordinate transform operational part 27, is transformed into and infers rotating coordinate system.This transformation matrix is following matrix.

Formula (7)

$\left(\begin{array}{c}{V}_d\\ V_q\end{array}\right)=\left(\begin{array}{cc}\cos \mathrm{\&theta;}& -\sin \mathrm{\&theta;}\\ \sin \mathrm{\&theta;}& \cos \mathrm{\&theta;}\end{array}\right)\left(\begin{array}{c}{V}_{\mathrm{\&gamma;}}\\ V_{\mathrm{\&delta;}}\end{array}\right)$

$\left(\begin{array}{c}{I}_d\\ I_q\end{array}\right)=\left(\begin{array}{cc}\cos \mathrm{\&theta;}& -\sin \mathrm{\&theta;}\\ \sin \mathrm{\&theta;}& \cos \mathrm{\&theta;}\end{array}\right)\left(\begin{array}{c}{I}_{\mathrm{\&gamma;}}\\ I_{\mathrm{\&delta;}}\end{array}\right)$

Because of current i γ, the i δ of this computing and higher harmonic components that produces because of motor 20 or noise stack, institute can reduce higher harmonic components or noise etc. so that this operation result passes through digital LPF32 (Fig. 5) again.According to the characteristic of the characteristic of motor 20 or inverter 17 etc., this LPF32 also can omit.

Calculate current i γ and γ shaft current command value i γ in the adder 13 ^*Between deviation, with this deviation input current controller 14.In addition, current i γ is used for the computing of site error estimator 21 or the computing in the inductance compensation device 22.On the other hand, current i δ is used for the computing of site error estimator 21 or the computing in the inductance compensation device 22.

Inductance compensation in the inductance compensation device 22 is to constitute for saturated or modeling error of compensating inductance etc.The formation of this inductance compensation makes it become the function of one and one above variablees such as comprising electric current (i γ, i δ) or rotating speed at least.This compensation needs only with modes such as approximate expression or forms.

The voltage instruction V of the supply PWM inverter 17 that voltage compensator 16 input is generated by voltage generator 15 (u, v) ^*, output is to this value V (u, v) ^*Carry out voltage presumed value Vmd (u, v) ^ behind phase place, the amplitude compensation.This voltage compensator 16 carries out is compensated for as the compensation that the input and output of considering in the PWM inverter 17 are carried out after non-linear, and carries out accordingly to the output of motor 20 from inverter 17.

Below, describe the computing of site error estimator 21 in detail.

Now the voltage equation with the motor model in the actual rotating coordinate system of d-q axle is shown in following formula (8)

Formula (8)

$\left(\begin{array}{c}{V}_d\\ V_q\end{array}\right)=\left(\begin{array}{cc}R+p{L}_d& -{\mathrm{\&omega;}}_{\mathrm{re}}{L}_q\\ {\mathrm{\&omega;}}_{\mathrm{re}}{L}_d& R+p{L}_q\end{array}\right)\left(\begin{array}{c}{i}_d\\ i_q\end{array}\right)+{\mathrm{\&omega;}}_{\mathrm{re}}{\mathrm{<figure-callout\; id="0"\; label="k\; E"\; filenames="A20051000706500231.png"\; state="\{\{state\}\}">k</figure-callout>}}_E\left(\begin{array}{c}0\\ 1\end{array}\right)$

KE is the induced voltage constant.

γ-δ axle is inferred the voltage equation of the motor model in the rotating coordinate system and is represented with following formula (9).

Formula (9)

$\left(\begin{array}{c}{V}_{\mathrm{\&gamma;}}\\ V_{\mathrm{\&delta;}}\end{array}\right)=\left(\begin{array}{cc}R-{\mathrm{\&omega;}}_{\mathrm{re}}{L}_{\mathrm{\&gamma;\&delta;}}+p{L}_{\mathrm{\&gamma;}}& -{\mathrm{\&omega;}}_{\mathrm{re}}{L}_{\mathrm{\&delta;}}+p{L}_{\mathrm{\&gamma;\&delta;}}\\ {\mathrm{\&omega;}}_{\mathrm{re}}{L}_{\mathrm{\&gamma;}}+p{L}_{\mathrm{\&gamma;\&delta;}}& R+{\mathrm{\&omega;}}_{\mathrm{re}}{L}_{\mathrm{\&gamma;\&delta;}}+p{L}_{\mathrm{\&delta;}}\end{array}\right)\left(\begin{array}{c}{i}_{\mathrm{\&gamma;}}\\ i_{\mathrm{\&delta;}}\end{array}\right)+{\mathrm{\&omega;}}_{\mathrm{re}}{k}_E\left(\begin{array}{c}\sin {\mathrm{\&Delta;\&theta;}}_{\mathrm{re}}\\ \cos {\mathrm{\&Delta;\&theta;}}_{\mathrm{re}}\end{array}\right)$

In the formula, each parameter of inductance L can be represented with following formula (10)

Formula (10)

L _δγ＝L ₁sin2Δθ _re $L_0=\frac{L_d+L_q}{2}$

L _γ＝L ₀+L ₁cos2Δθ _re $L_1=\frac{L_d-L_q}{2}$

L _δ＝L ₀-L ₁cos2Δθ _re

In the formula, make that Δ θ is zero owing to control, thus Δ θ 0, thereby, sin Δ θ 0, cos Δ θ 1.When approximate, obtaining following formula (11) for Δ θ with this.

Formula (11)

$\left(\begin{array}{c}{V}_{\mathrm{\&gamma;}}\\ V_{\mathrm{\&delta;}}\end{array}\right)=\left(\begin{array}{cc}R+p{L}_d& -{\mathrm{\&omega;}}_{\mathrm{re}}{L}_q\\ {\mathrm{\&omega;}}_{\mathrm{re}}\mathrm{Ld}& R+{\mathrm{pL}}_q\end{array}\right)\left(\begin{array}{c}{i}_{\mathrm{\&gamma;}}\\ i_{\mathrm{\&delta;}}\end{array}\right){\mathrm{\&omega;}}_{\mathrm{re}}{k}_E\left(\begin{array}{c}\sin {\mathrm{\&Delta;\&theta;}}_{\mathrm{re}}\\ \cos {\mathrm{\&Delta;\&theta;}}_{\mathrm{re}}\end{array}\right)$

In the formula, induced voltage V is owing to being with the value behind the magnetic flux φ differential, so can obtain following formula (12) as the relational expression for magnetic flux.Promptly, obtain the magnetic flux equation of following formula (12) by remove the voltage equation of above-mentioned formula (11) with angular velocity omega re.

Formula (12)

$\left(\begin{array}{c}{\mathrm{\&phi;}}_{\mathrm{\&delta;}}\\ {\mathrm{\&phi;}}_{\mathrm{\&gamma;}}\end{array}\right)=\frac{1}{{\mathrm{\&omega;}}_{\mathrm{re}}}\left(\begin{array}{c}{V}_{\mathrm{\&gamma;}}\\ V_{\mathrm{\&delta;}}\end{array}\right)=\left(\begin{array}{cc}\frac{R+p{L}_d}{{\mathrm{\&omega;}}_{\mathrm{re}}}& -L_q\\ L_d& \frac{R+p{L}_q}{{\mathrm{\&omega;}}_{\mathrm{re}}}\end{array}\right)\left(\begin{array}{c}{i}_{\mathrm{\&gamma;}}\\ i_{\mathrm{\&delta;}}\end{array}\right)+k_E\left(\begin{array}{c}\sin \mathrm{\&Delta;}{\mathrm{\&theta;}}_{\mathrm{re}}\\ \cos \mathrm{\&Delta;}{\mathrm{\&theta;}}_{\mathrm{re}}\end{array}\right)$

As use the approximate expression of Δ θ re sin Δ θ re, then obtain following formula (13).

Formula (13)

$=\frac{1}{k_E}\{\frac{V_{\mathrm{\&gamma;}}-R\&CenterDot;i_{\mathrm{\&gamma;}}}{{\mathrm{\&omega;}}_{\mathrm{re}}}+L_q\&CenterDot;i_{\mathrm{\&delta;}}\}$

$=K^{\&prime;}\{\frac{V_{\mathrm{\&gamma;}}-R\&CenterDot;i_{\mathrm{\&gamma;}}}{{\mathrm{\&omega;}}_{\mathrm{re}}}+L_q\&CenterDot;i_{\mathrm{\&delta;}}\}(K^{\&prime;}>0)$

Here, establish the transition item of electric current and ignore, p (i) 0.In addition, K ' is equivalent to the inverse of induced voltage constant, can make it become arbitrary constant or the function that satisfies K '＞0.

Because the rotation of DC motor, V becomes the value that is directly proportional with ω, and therefore in above-mentioned formula (13), V γ is the value that is directly proportional with ω.So one of (V γ-Ri the γ)/ω re of above-mentioned formula (13) can not become the value that depends on ω.Thus, can obtain Δ θ re from above-mentioned formula (13), be zero control if make this Δ θ re, just can carry out stable control.

If establish p (i) ≠ 0, consider the transition item again, just become following formula (14).

Formula (14)

$=\frac{1}{k_E}\{\frac{V_{\mathrm{\&gamma;}}-(R+p{L}_d)\&CenterDot;i_{\mathrm{\&gamma;}}}{{\mathrm{\&omega;}}_{\mathrm{re}}}+L_q\&CenterDot;i_{\mathrm{\&delta;}}\}$

$=K^{\&prime;}\{\frac{V_{\mathrm{\&gamma;}}-(R+{\mathrm{pL}}_d)\&CenterDot;i_{\mathrm{\&gamma;}}}{{\mathrm{\&omega;}}_{\mathrm{re}}}+L_q\&CenterDot;i_{\mathrm{\&delta;}}\}(K^{\&prime;}>0)$

$=K^{\&prime;}\{\frac{V_{\mathrm{\&gamma;}}-R\&CenterDot;i_{\mathrm{\&gamma;}}-K^{\&prime;\&prime;}\mathrm{\&Delta;}{i}_{\mathrm{\&gamma;}}}{{\mathrm{\&omega;}}_{\mathrm{re}}}+L_q\&CenterDot;i_{\mathrm{\&delta;}}\}(K^{\&prime;}>0,K^{\&prime;\&prime;}>0)$

Constant arbitrarily or the function of K " energy is as satisfying K "＞0 are controlled.At this moment " for bigger value, compare the occasion that requires stability high with response characteristic, the getting K "=0 of requiring the good occasion of response characteristic, getting K, become and the identical formula of above-mentioned formula (13).

Fig. 1 carries out the formation block diagram of the site error estimator 21 that site error infers for expression utilizes above-mentioned formula (13).Fig. 2 carries out the formation block diagram of the site error estimator 21 that site error infers for expression utilizes above-mentioned formula (14).In Fig. 2, the transition item is corresponding with (Δ i γ/ω re).As shown in Figures 1 and 2, site error estimator 21 comprise the magnetic flux operational part of inferring magnetic flux Φ δ, and according to the site error operational part of the magnetic flux Φ δ estimated position error delta θ that infers out.Among Fig. 1 and Fig. 3, R is the winding resistance of motor.Q axle inductance L q ^*Can try to achieve with inductance compensation device 22.Q axle inductance L q ^*Be i γ, the i δ that obtains by experiment with the motor of belt sensor in advance, the function of ω re.

In Fig. 1, poor (V γ-RI γ) that the magnetic flux operational part of site error estimator 21 is inferred the voltage drop RI γ that deducts electric current I γ from motor applied voltage V γ carries out the magnetic flux ((V γ-RI γ)/ω re) that integration forms, with the inductance L q of motor winding and the long-pending magnetic flux (Lqi δ) of current i δ, according to this magnetic flux of inferring ((V γ-RIr)/ω re+Lqi δ), that asks rotatable coordinate axis (q axle) and rotatable coordinate axis thereof infers the magnetic flux error Φ δ of axle between (δ axle), the site error operational part of site error estimator 21 is inferred rotor-position Δ θ according to this magnetic flux error Φ δ.

In Fig. 2, the magnetic flux operational part of site error estimator 21 is inferred poor (V γ-(R+pLd) I γ) to the voltage drop ((R+pLd) I γ) that deducts electric current from motor applied voltage V γ and is carried out the magnetic flux that integration forms (V γ-(R+pLd)/ω re), long-pending magnetic flux (Lqi δ) with the inductance L q current i δ of motor winding, according to this magnetic flux of inferring ((V γ-(R+pLd) I γ)/ω re+Lqi δ), that obtains rotatable coordinate axis (q axle) and rotatable coordinate axis thereof infers the magnetic flux error Φ δ of axle between (δ axle), the site error operational part of site error estimator 21 is inferred rotor-position Δ θ according to this magnetic flux error Φ δ.

As shown in Figure 3, the site error Δ θ re (Δ θ ^) that is obtained by site error estimator 21 exports to speed estimating device 24.

The speed estimating device 24 that the site error Δ θ ^ input that site error estimator 21 is calculated is made of the PI controller.Speed estimating device 24 calculates and makes this Δ θ ^ is zero angular speed presumed value ω re^.Here, use general usually PI controller as speed estimating device 24.Shown in arithmetic expression in the speed estimating device 24 formula described as follows (15).

Formula (15)

${\widehat{\mathrm{\&omega;}}}_{\mathrm{re}}=K_p\mathrm{\&Delta;}\widehat{\mathrm{\&theta;}}+K_I\&Integral;\mathrm{\&Delta;}\widehat{\mathrm{\&theta;}}\mathrm{dt}$

The magnitude of angular velocity ω re^ that is obtained by speed estimating device 24 exports to adder 11, as mentioned above, is used for the FEEDBACK CONTROL of speed.By this speed feedback control, formation voltage amplitude instruction V (m) ^*Carry out integration with integrator 26 diagonal angle speed estimating value ω re^ in addition, calculate position deduction value θ re (γ δ) ^ of rotor.The position deduction value θ re (γ δ) that this calculates imports the adder 18 and the coordinate converter 27,28 of pressing the phase bit instruction respectively.

Adder 18 generates the voltage-phase instruction V (θ) of input voltage generating unit 15 according to position deduction value θ re (γ δ) ^ ^*Like this, by instruct V (θ) at voltage-phase ^*The offset of the Δ θ ^ that last reflection is inferred by site error estimator 21, thus the offset of Δ θ ^ can on motor 20, be reflected.

Adjust angular speed presumed value ω re^, make the physical location of motor 20 consistent, ask its integrated value promptly to infer position θ re (γ δ) ^ on the coordinate system, carry out FEEDBACK CONTROL, make the position consistency of this d-q axle and γ-δ axle with presumed value.In addition, when this angular speed presumed value ω re^ is used for speed feedback control,, also can allow it pass through LPF33 (Fig. 4) for stable.

As mentioned above, in the control device of electric motor 10 of this example, use the value that goes out by sensor as current value, and use command value or presumed value as magnitude of voltage (not working voltage transducer).Important factor inductance when turning round for the motor that salient pole is arranged (IPM motor) can utilize inductance compensation device 22 to obtain inductance presumed value (command value).About angular velocity omega, use presumed value though consider cambic of expression in this example, also can use command value.

As mentioned above, according to this example,, can obtain irrelevant physical quantity with ω re by carrying out position deduction with magnetic flux.By just adjusting gain easily, can realize to improve the Motor Control of stability with computing simple and that the time is short again with formula (13) or formula (14).If the control device of electric motor of the rotor position presuming method by adopting this example is controlled motor, just can make the motor steady running.As using, then can provide efficient, low noise compressor by the motor of the control device of electric motor control of the rotor position presuming method that adopts this example motor as compressor.In addition, as this compressor being used in (not shown) on the air-conditioning, then can help to reduce the power consumption of air-conditioning.

Compare with above-mentioned non-patent literature 1 described technology, because induced voltage is directly proportional with rotating speed (ω), so the site error Δ θ that infers according to induced voltage is the value that depends on ω, departing from real value may be very big, opposite because the magnetic flux of permanent magnet is not for depending on the intrinsic value of material of ω, so the site error Δ θ that infers according to magnetic flux is for real value or near the value of actual value.Therefore, magnetic flux is suitable for position deduction (site error is inferred).

Also when asking Δ θ re, just can utilize above-mentioned formula (11), launch an accepted way of doing sth (16) and formula (17).

Formula (16)

$\tan \mathrm{\&Delta;}{\mathrm{\&theta;}}_{\mathrm{re}}=\frac{{\mathrm{\&omega;}}_{\mathrm{re}}{K}_E\sin {\mathrm{\&Delta;\&theta;}}_{\mathrm{re}}}{{\mathrm{\&omega;}}_{\mathrm{re}}{K}_E\cos {\mathrm{\&Delta;\&theta;}}_{\mathrm{re}}}=\frac{V_{\mathrm{\&gamma;}}-(R+p{L}_d)\&CenterDot;i_{\mathrm{\&gamma;}}+{\mathrm{\&omega;}}_{\mathrm{re}}{L}_q{i}_{\mathrm{\&delta;}}}{V_{\mathrm{\&delta;}}-{\mathrm{\&omega;}}_{\mathrm{re}}{L}_d\&CenterDot;i_{\mathrm{\&gamma;}}-(R+p{L}_q)\&CenterDot;i_{\mathrm{\&delta;}}}$

Formula (17)

${\mathrm{\&Delta;\&theta;}}_{\mathrm{re}}={\tan }^{-1}\frac{V_{\mathrm{\&gamma;}}-(R+{\mathrm{pL}}_d)\&CenterDot;i_{\mathrm{\&gamma;}}+{\mathrm{\&omega;}}_{\mathrm{re}}{L}_q{i}_{\mathrm{\&delta;}}}{V_{\mathrm{\&delta;}}-{\mathrm{\&omega;}}_{\mathrm{re}}{L}_d\&CenterDot;i_{\mathrm{\&gamma;}}-(R+p{L}_q)\&CenterDot;i_{\mathrm{\&delta;}}}$

Here, in order to carry out the control of Δ θ 0,, also can utilize the following formula (18) of only considering molecule to control even do not consider the denominator of following formula (17) in fact sometimes.

Formula (18)

Δθ _resinΔθ _re＝V _γ-(R+pL _d)·i _γ+ω _reL _qi _δ

In order to carry out the control of Δ θ 0 in real time, to reduce operand sometimes as far as possible, need not above-mentioned formula (17), and as above-mentioned formula (18), ask Δ θ with the approximate formula of sin.

But, the Δ θ=tan of formula (17) ^-1In (), depend on that the relation of ω is offset well (owing to all there is the item that depends on ω to exist, so both cancel each other, just not depending on ω in molecule, denominator), this point is when calculating formula (18), and Δ θ becomes the value that depends on ω.

This is because in the formula of formula (18), calculates induced voltage, therefore becomes the characteristic quantity that is directly proportional with rotating speed, just becomes to have and the identical problem of the occasion that adopts above-mentioned formula (3) (technology of above-mentioned non-patent literature 1).That is, Δ θ in addition, adjusts in the gain of axle offset adjuster to zero convergence variation, carry out different gain adjustment according to the rotating speed of motor 20, becomes complicated on the formation of control.On the contrary,, as mentioned above, carry out position deduction, can infer irrelevant Δ θ with ω by adopting magnetic flux according to this example.

Also have, in the control device of electric motor 10 of this example, preferably ask Δ θ (magnetic flux error) between q axle and the δ axle.The magnetic flux φ that has permanent magnet on the d axle is with φ=0 on the q axle of d axle quadrature.When between q axle and δ axle, asking Δ θ, as long as FEEDBACK CONTROL makes φ (q axle).This point is compared with the occasion of asking Δ θ between d axle and γ axle, again φ=(d axle) carried out FEEDBACK CONTROL, controls easily.This is because if want to ask Δ θ between d axle and γ axle, when magnetic flux φ is carried out FEEDBACK CONTROL, because magnetic flux φ is the intrinsic value of permanent magnet material, so the value of the φ that changes according to the magnetic field intensity because of permanent magnet must change command value to each motor.On the other hand, also can in this example, between d axle and γ axle, ask Δ θ, replace above-mentioned method.

Claims (19)

1. a rotor position presuming method is a kind of rotor position presuming method of inferring the permanent magnet motor rotor-position, it is characterized in that,

Obtain rotatable coordinate axis and described rotatable coordinate axis and infer the magnetic flux error of between centers,, infer described rotor-position according to described magnetic flux error.

2. a rotor position presuming method is a kind of rotor position presuming method of inferring the permanent magnet motor rotor-position of salient pole, it is characterized in that,

Infer to the difference of the voltage drop that deducts electric current from the motor applied voltage carry out integration and magnetic flux and the magnetic flux of the inductance of motor winding and the long-pending generation of electric current, according to described magnetic flux of inferring, obtain rotatable coordinate axis and described rotatable coordinate axis and infer the magnetic flux error of between centers, according to described magnetic flux error, infer described rotor-position.

3. a rotor position presuming method is a kind of rotor position presuming method of inferring the permanent magnet motor rotor-position of salient pole, it is characterized in that,

Remove the magnetic flux equation that described permanent magnet motor voltage equation gets according to angular velocity omega, obtain the magnetic flux error that rotatable coordinate axis and described rotatable coordinate axis are inferred between centers,, infer described rotor-position according to described magnetic flux error with rotor.

4. as each described rotor position presuming method in the claim 1 to 3, it is characterized in that,

Described magnetic flux error be described rotatable coordinate axis be q axle and described rotatable coordinate axis infer the axle be the magnetic flux error of δ between centers.

5. a rotor position presuming method is a kind of rotor position presuming method of inferring the permanent magnet motor rotor-position of salient pole, it is characterized in that,

Adopt the approximate expression of difference Δ θ sin Δ θ of the rotor position angle of the voltage equation of described permanent magnet motor and actual motor and motor model, do not adopt the Δ θ=tan that launches according to described voltage equation simultaneously ^-1The formula of form is obtained the Δ θ of the angular velocity omega that does not depend on rotor, according to the described Δ θ that tries to achieve, infers rotor-position.

6. as each described rotor position presuming method in the claim 1 to 3, it is characterized in that,

Described magnetic flux error can utilize following formula (1) to try to achieve.

Formula (1)

$\frac{V_{\mathrm{\&gamma;}}-R\&CenterDot;i_{\mathrm{\&gamma;}}}{{\mathrm{\&omega;}}_{\mathrm{re}}}+L_q\&CenterDot;i_{\mathrm{\&delta;}}$

V γ: the γ axle component of armature voltage

R: armature winding resistance

I γ: the γ axle component of armature supply

ω re: the presumed value of rotor velocity or command value (electrical degree)

Lq:q axle inductance

I δ: the δ axle component of armature supply

7. as each described rotor position presuming method in the claim 1 to 3, it is characterized in that,

Described magnetic flux error can be tried to achieve with following formula (2).

Formula (2)

$\frac{V_{\mathrm{\&gamma;}}-(R+p{L}_d)\&CenterDot;i_{\mathrm{\&gamma;}}}{{\mathrm{\&omega;}}_{\mathrm{re}}}+L_q\&CenterDot;i_{\mathrm{\&delta;}}$

V γ: the γ axle component of armature voltage

R: armature winding resistance

P: differential operator

Ld:d axle inductance

I γ: the γ axle component of armature supply

ω re: the presumed value of rotor velocity or command value (electrical degree)

Lq:q axle inductance

I δ: the δ axle component of armature supply

8. rotor position presuming method as claimed in claim 2 is characterized in that,

Described inductance is for depending on the function of the some amounts in electric current and the rotating speed at least.

9. a method of motor control is characterized in that,

Try to achieve with utilize as each described rotor position presuming method in the claim 1 to 3 obtain as described in the magnetic flux error corresponding as described in rotatable coordinate axis and as described in the presumed value of inferring the corresponding rotor velocity of between centers site error of rotatable coordinate axis, presumed value input low pass filter with described rotor velocity, according to the output valve of described low pass filter, carry out and the relevant FEEDBACK CONTROL of described permanent-magnet electric motor speed.

10. a method of motor control is a kind of method of motor control of using as each described rotor position presuming method in the claim 1 to 3, it is characterized in that,

To make that the error of inferring corresponding value of the detected value of a component and command value with the described rotatable coordinate axis of armature supply is that the phase place command value is imported low pass filter for the output of the current controller of small incidental expenses, according to the output valve and the described rotor-position of inferring of described low pass filter, generate the signal of the phase place of expression voltage instruction.

11. a method of motor control is a kind of method of motor control of using as each described rotor position presuming method in the claim 1 to 3, it is characterized in that,

Infer the corresponding value of detected value of a component to the low pass filter input and the described rotatable coordinate axis of armature supply, according to the deviation of the command value of inferring a component of the rotatable coordinate axis of the output valve of described low pass filter and described armature supply, the command value of formation voltage phase place.

12. a method of motor control is the method for motor control that a kind of control has the permanent magnet motor of salient pole, it is characterized in that,

Obtain rotatable coordinate axis and be the q axle to infer axle be the magnetic flux of δ axle, control the magnetic flux that makes described δ axle and converge zero.

13. method of motor control as claimed in claim 12 is characterized in that,

The magnetic flux of described δ axle can be to deduct magnetic flux and the inductance L q that the difference of the voltage drop of resistance R and electric current I γ calculates from the presumed value V γ ^ of motor applied voltage and try to achieve with the form of amassing the magnetic flux sum that obtains of electric current I δ by removing with angular velocity omega re^.

14. method of motor control as claimed in claim 12 is characterized in that,

The magnetic flux of described δ axle can be to deduct magnetic flux that difference that the voltage drop of resistance R and electric current I γ and inductance L d and electric current I γ change the voltage drop of generation in time calculates and try to achieve with utilizing the form of amassing the magnetic flux sum that obtains of inductance L q and electric current I δ by removing with angular velocity omega re^ from the presumed value Vr^ of motor applied voltage.

15. method of motor control as claimed in claim 12 is characterized in that,

Described δ axle magnetic flux can be to deduct the voltage drop of resistance R and electric current I γ and positive gain constant K according to removing with angular velocity omega re^ from the presumed value Vr^ of the motor applied voltage " magnetic flux of calculating after the difference of the long-pending voltage drop that produces that changes in time with electric current I γ r and try to achieve according to inductance L q and the form of amassing the magnetic flux sum that obtains of electric current I δ.

16. a method of motor control is a kind of method of motor control that does not have the sensor drive permanent magnet motor, it is characterized in that,

Calculate the magnetic flux error, adjust the rotating speed of described permanent magnet motor, the magnetic flux error convergence that makes described computing is zero, and by the speed estimating value integration to obtaining, the magnetic flux error that makes described computing is zero, thereby calculates the rotor-position of described permanent magnet motor.

17. a compressor is characterized in that,

Have by calculating the magnetic flux error and controlling and make that thereby the described magnetic flux error convergence that calculates is the zero air compressor motor that can carry out the position-sensor-free sine wave drive.

18. a program is characterized in that,

For allowing computer carry out the program of using as each step of each described rotor position presuming method in the claim 1 to 3.

19. a rotor position presuming device is a kind of rotor position presuming device of inferring the permanent magnet motor rotor-position of salient pole, it is characterized in that,

Try to achieve the magnetic flux error of inferring between centers of rotatable coordinate axis and described rotatable coordinate axis,

According to described magnetic flux error, infer described rotor-position.

Images