JP2002171799A - Vector control method and vector control device for ac motor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a vector control method and a vector control device, which are a drive control method and a drive control device for an AC motor with reverse salient pole characteristics and can eliminate a position sensor, such as an encoder and the like hitherto attached to the rotor of the motor. SOLUTION: A vector position estimating device 13 detects or estimates a high frequency in-phase current vector and an image-phase current vector, by using a stator current vector, etc., containing high-frequency components. Then, by using both the in-phase and image-phase current vectors, a cosine value and a sine value of the intermediate angle of an angle between both the current vectors or their estimated values are determined and then outputted. The outputted signals are utilized to generate a rotation signal for a vector rotation device, in order to achieve the vector control without using a position sensor.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a vector control method and an apparatus for an AC motor having a reverse salient pole characteristic, such as a permanent magnet type synchronous motor and a rotor slot open type induction motor. In particular, a position estimator is used in place of the position detector attached to the rotor to secure cosine and sine information of the rotor position necessary for the vector rotator for vector control, and further, to drive the estimator. The present invention relates to a vector control method and an apparatus for generating a necessary signal by using at least a high-frequency current generated according to a high-frequency voltage applied to a motor.

[0002]

2. Description of the Related Art Stabilization of stator current is indispensable for an AC motor to exhibit high control performance, and a vector control method has conventionally been known as a control method for this purpose. The vector control method is a method of dividing and controlling a stator current contributing to torque generation as a d-axis component and a q-axis component of a current vector on a rotating dq coordinate system including a d-axis and a q-axis orthogonal to each other. It has a control step.

The rotating dq coordinate system at this time is a coordinate system synchronized with the position of the rotor magnetic flux, that is, the phase (in the case of the permanent magnet type synchronous motor, the reverse salient pole position of the rotor itself) with zero spatial phase difference. Is generally adopted. That is, it is common to employ a rotating dq coordinate system in which the phase of the rotor magnetic flux is selected on the d-axis and the axis orthogonal to this is selected on the q-axis. In order to maintain the rotating dq coordinate system in a synchronized state having no spatial phase difference with the phase of the rotor magnetic flux, it is generally necessary to know the position of the rotor. In order to know this accurately, a position detector represented by an encoder is traditionally mounted on a rotor.

FIG. 18 shows a vector control method using a rotor position detector for a permanent magnet type synchronous motor, which is a typical AC motor having a reverse salient pole characteristic, and is mounted on the permanent magnet type synchronous motor. FIG. 1 is a block diagram schematically showing one example. 1a is an AC motor (synchronous motor), 2 is a rotor position detector, 3 is a power converter, 4 is a current detector, 5a and 5b are 3 phase 2 phase converters, 2 phase 3 phase conversion respectively. 6a and 6b are vector rotators, 7 is a cosine sine signal generator, 8 is a current controller, 9 is a command converter, 10 is a speed controller, and 11 is a speed detector. ing. In FIG. 18, various devices from 4 to 9 constitute a vector control device. In this figure, to ensure simplicity,
A 2x1 vector signal closely related to the present invention is represented by one thick signal line. The following block diagram representation follows this.

In particular, the rotor position detector 2 detects the rotor position as a position angle with respect to the center of the U-phase winding, and 7 outputs the cosine / sine signal to the vector rotators 6a and 6b. , Means for generating a rotation signal for the rotation dq coordinate system. In an AC motor having reverse salient pole characteristics, generally, an angle indicating the direction of the rotor reverse salient pole is selected as a reference rotor position. For this reason, as is well known to those skilled in the art, generally, the rotor position and the reverse salient pole position of the rotor may be considered synonymously. In the description of the present invention, the rotor position and the reverse salient pole position of the rotor are used synonymously. Therefore, in the example of FIG. 18 in which the permanent magnet synchronous motor is used, the rotor position means the phase of the rotor magnetic flux.

The rotor position and the rotor speed have an integral and derivative relationship with each other. As is well known to those skilled in the art, the speed information is obtained from a rotor position detector such as an encoder, like the position information. ing. Reference numeral 11 denotes a speed detector which realizes such speed detecting means. 4, 5a, 5b, 6a, 6b,
A current control step of dividing the stator current into a d-axis component and a q-axis component on a rotating dq coordinate system and controlling each of them to follow d-axis and q-axis current command values; Is constituted.

The three-phase current detected by the current detector 4 is 3
After being converted into 2-phase currents on the fixed αβ coordinate system phase 2 phase converter 5a, is converted by the vector rotator 6a 2-phase currents i _d of the rotating dq coordinate _system, the i _q, feeding the current controller 8 Can be
The current controller 8 converts the conversion currents _id , _iq It is converted into a two-phase voltage command value in the αβ coordinate system and sent to the two-phase three-phase converter 5b. In step 5b, the two-phase signal is converted into a three-phase voltage command value, which is output to the power converter 3 as a command value. The power converter 3 generates electric power according to the command value, and
and apply this to ing.

FIG. 18 shows an example in which a speed control system is configured. Therefore, a torque command value is obtained as an output of the speed controller 10 which receives a speed command value and a speed detection value. As is well known to those skilled in the art, when the control purpose is generated torque and a speed control system is not configured, the speed controller 1
0, the speed detector 11 is unnecessary. In this case, the torque command value is directly applied from the outside.

FIG. 19 shows a case where a vector control method using a rotor position detector is applied to an open-slot type induction motor, which is a type of AC motor having reverse salient pole characteristics, and is mounted on the motor. FIG. 1 is a block diagram schematically showing a typical example. Vector control of this example and FIG.
The major difference between the vector control and the permanent magnet type synchronous motor described using is that the AC motor is changed to the rotor open slot type induction motor 1b and the magnetic flux phase estimator 12 is introduced. This is the same as in the case of FIG. In an induction motor, slip exists between the rotor position and the phase of the rotor magnetic flux, and the rotor position and the rotor magnetic flux phase are not the same. The phase of the rotor magnetic flux has a relationship of integration and differentiation with the angular frequency of the rotor magnetic flux, and the angular frequency of the rotor magnetic flux is determined as the sum of the rotor speed and the slip angle frequency. Further, the slip angle frequency is a two-phase current i in the rotating dq coordinate system.
It is estimated from _d and _iq . The magnetic flux phase estimator 12 outputs the estimated value of the rotor magnetic flux phase obtained using this relationship to the cosine sine signal generator 7.

[0010]

In order to realize a conventional vector control method for an AC motor, a rotor position detector for detecting a rotor position is indispensable as described in the representative example. It is. However, mounting a rotor of a rotor position detector such as an encoder has inevitably caused the following problems.

The first problem is a reduction in the reliability of the motor system. Rotor position detectors such as encoders have significantly lower mechanical robustness when compared to motor bodies. That is, the mounting of the rotor position detector significantly reduces the mechanical reliability of the motor system. Deterioration of the robustness of the motor system due to the mounting of the rotor position detector is not only the mechanical aspect, but also the electrical aspect seen in the incorporation of power supply noise into the rotor position detector signal,
Further, the same occurs on the thermal side seen in the rise in temperature of the rotor position detector due to the heat generated by the rotor.
As described above, by attaching the rotor position detector such as the encoder to the motor rotor, the reliability of the motor system has been significantly reduced.

A second problem is an increase in motor space. Although depending on the volume of the motor alone, mounting the rotor position detector on the rotor increases the volume of the motor in the axial direction by several percent to tens of percent.

A third problem is to secure a wiring for a power supply line for operating the rotor position detector and a signal line for receiving a detection signal and a space for the wiring. As a matter of course, wiring is necessary for operating the rotor position detector and obtaining the rotor position information therefrom. In addition, the signal line is generally required to be as robust as the power line for driving the motor body in order to avoid the above-mentioned reduction in mechanical, electrical, and thermal reliability as much as possible. As a result, wiring of a signal line having substantially the same size as an original power line per motor is required, and furthermore, a space for this is required.

The fourth problem is an increase in various costs. In small motors, the cost of the rotor position detector may already be higher than the motor itself at the time of manufacture.
The cost of wiring associated with the rotor position detector is not negligible for small motors. Further, an increase in maintenance cost to cope with the decrease in reliability inevitably occurs. These various costs increase with the number of motors used. In particular, the maintenance cost has a characteristic that increases exponentially according to the number.

The above problem is directly or indirectly caused by the rotor position detector, and is inevitably solved if a so-called sensorless vector control method that does not require a rotor position detector is established. . In fact, several distinctive methods for sensorless vector control of an AC motor for this purpose have already been reported. The latest survey results of sensorless vector control method for AC motors and technology development in Japan and abroad related to the device are described in the literature (edited by the Technical Committee on New Technology Research for AC Motor Drive System of IEEJ, Technical Report No. 7 of IEEJ).
No. 60, a recent technical trend in AC motor driving, published in February 2000). As described therein, the sensorless vector control method of the AC motor includes a method of estimating the rotor position using a fundamental wave component of a voltage and a current for generating a torque, and a method of using a high frequency that does not contribute to the torque generation. The method is roughly divided into a method of separately injecting power and using a high-frequency component included in the voltage and current.
In particular, with respect to the method of applying a high-frequency voltage targeted by the present invention, although potential performance is expected, the history of research and development is short and sufficient performance has not always been obtained. In particular, since the characteristics of an AC motor fluctuate according to the operating state, in other words, the motor parameters expressing the characteristics of the motor fluctuate. It has been demanded.

SUMMARY OF THE INVENTION The present invention has been made under the above background, and has as its object to provide a vector control method and apparatus which do not require a rotor position detector such as an encoder for an AC motor having reverse salient pole characteristics. Another object of the present invention is to provide a vector control method and a vector control method capable of accurately or efficiently estimating the cosine and sine values of the rotor position in a state robust to variations in the motor parameters.

[0017]

In order to achieve the above-mentioned object, according to the first aspect of the present invention, a stator current contributing to torque generation is generated by a d-axis and a q-axis orthogonal to each other specified by a vector rotator. A current control step of dividing and controlling the d-axis component and the q-axis component of the current vector on the configured rotating dq coordinate system, and applying a high-frequency voltage that can be treated as a rotating high-frequency voltage vector as at least a part of the stator voltage Using a high-frequency current vector generated in response to the application of the high-frequency voltage vector, and rotating in the same direction as the high-frequency magnetic flux vector generated in response to the application of the high-frequency voltage vector. And the mirror current vector rotating in the opposite direction to the high-frequency magnetic flux vector is detected or estimated. A cosine / sine value of an intermediate angle between the vector and the mirror phase current vector or an estimated value thereof is used as a cosine / sine estimated value of a rotor position of the AC motor for generating a rotation signal of the vector rotator. It is characterized by doing.

According to a second aspect of the present invention, there is provided the vector control method for an AC motor according to the first aspect, wherein the in-phase current vector or the estimated value of the in-phase current vector and the mirror-phase current vector or the estimated value of the in-phase current vector are calculated. Is used to determine the cosine / sine value of the double angle of the intermediate angle or its estimated value, and the cosine / sine value of the intermediate angle or the estimated value thereof is determined from the determined value of the double angle cosine / sine. It is characterized by the following.

The third aspect of the present invention provides the first and second aspects.
The vector control method for an AC motor according to claim 1, wherein the cosine / sine value of the intermediate angle or the estimation of the cosine / sine value of the intermediate angle is determined according to an expected magnitude of the cosine / sine value of the intermediate angle. The method of determining a value is changed.

According to a fourth aspect of the present invention, there is provided a vector control method for an AC motor according to the first aspect, wherein the in-phase current vector or an estimated value of the in-phase current vector and the mirror-phase current vector, Generate two vectors with the in-phase estimated value and determine the cosine / sine value of the intermediate angle or the estimated value thereof in proportion to the composite vector obtained by adding the two identical norm vectors. It is characterized by doing so.

According to a fifth aspect of the present invention, there is provided the vector control method for an AC motor according to the first aspect, wherein the in-phase current vector or the estimated value of the in-phase current vector and the mirror-phase current vector, Two vectors with this and the in-phase estimated value are generated, and a combined vector obtained by vector subtraction of the two same norm vectors is π / 2
(Rad) The method is characterized in that the cosine / sine value of the intermediate angle or its estimated value is determined in proportion to the subtracted composite vector after the rotation.

The invention according to claim 6 is based on claim 1, claim 2,
A vector control method for an AC motor according to claim 3, 4, or 5, wherein the high-frequency voltage vector is a sine high-frequency voltage vector.

According to a seventh aspect of the present invention, a stator current contributing to torque generation is represented by a d-vector of a current vector on a rotating dq coordinate system composed of a d-axis and a q-axis which are orthogonal to each other and designated by a vector rotator. A vector control device for an AC motor having current control means for dividing and controlling an axis component and a q-axis component, and high-frequency voltage applying means for applying a high-frequency voltage that can be treated as a rotating high-frequency voltage vector as at least a part of the stator voltage Using a stator current vector generated in response to the application of the high-frequency voltage vector, a common-mode current vector that rotates in the same direction as the high-frequency magnetic flux vector also generated in response to the application of the high-frequency voltage vector,
Means for detecting or estimating a mirror phase current vector rotating in a direction opposite to the high-frequency magnetic flux vector, and a cosine / sine estimated value of a rotor position of the AC motor which can be used for generating a rotation signal of the vector rotator. Means for generating a cosine / sine value of an intermediate angle between the in-phase current vector and the mirror-phase current vector or an estimated value thereof.

Next, the operation of the present invention will be described. The direction of the center of the stator U-phase winding of the AC motor is the base axis, that is, α.
Consider a fixed αβ coordinate system in which the axis is an axis and the sub-axis orthogonal to the axis is the β axis. The direction from the base axis to the sub-axis is defined as a positive direction. It is assumed that the reverse salient pole position of the rotor (that is, the rotor position to be a reference) instantaneously makes an electrical angle θ with respect to the α-axis. FIG. 1 shows, as an example, a state of a rotor reverse salient pole position (N pole position) of a permanent magnet type synchronous motor. On the fixed αβ coordinate system, the electric characteristics of the AC motor having the reverse salient pole characteristics targeted by the present invention are as follows (1):
It can be expressed by equation (2).

Ν ₁ = R ₁ i ₁ + sφ ₁ (1)

## EQU2 ## φ ₁ = φ _i + φ _m (2)

[0025] (1), the _{R 1} is the winding resistance of the stator in (2), s is meant differential operator d / dt.
Further, ν ₁ , i ₁ , φ ₁ are a stator voltage vector, a stator current vector, and a stator linkage flux (stator flux) vector, respectively. φ _i and φ _m indicate components constituting the stator magnetic flux vector φ ₁ , φ _i indicates a magnetic flux vector directly generated by the stator current, and φ _m indicates a rotor linked from the rotor side. Each means a magnetic flux vector. These vectors are all 2x1 vectors.

When performing the vector control, as described with reference to FIGS. 18 and 19, the signal on the fixed αβ coordinate system is converted into the signal on the rotation dq coordinate system using the vector rotator 6a, and It is necessary to convert the signal on the dq coordinate system into a signal on the fixed αβ coordinate system using the vector rotator 6b. The characteristics of the vector rotator 6b at this time are expressed by the following equation (3).

[Equation 3] The characteristics of the vector rotator 6a are expressed as this transposed matrix. Θ _φ in the equation (3) is the phase of the rotor magnetic flux evaluated on the fixed αβ coordinate system. That is, as clearly shown by the equation (3), the vector rotator needs the cosine and sine of the rotor magnetic flux phase evaluated on the fixed αβ coordinate system.

In the conventional vector control method for a synchronous motor, the rotor magnetic flux phase θ _φ and the rotor position θ are the same as described with reference to FIG. In order to obtain a sine value, a position detector is mounted on the rotor, and the cosine and sine value of the rotor position θ have been directly obtained.
In a conventional vector control method for an induction motor, FIG.
As described above, in order to obtain the cosine and sine value of the rotor magnetic flux phase, a position detector is mounted on the rotor, and the rotor speed has been obtained through an approximate differentiation process. Instead, the present invention seeks to obtain a cosine / sine estimate of the rotor position and an estimate of the rotor speed without using a rotor position detector.

In the present invention, a high-frequency voltage that can be treated as a rotating high-frequency voltage vector is applied to the AC motor as at least a part of the stator voltage. In the present invention for applying a high-frequency voltage, the voltage, current,
Each vector of the magnetic flux can be expressed separately as a fundamental component (low frequency component) that contributes to torque generation and has the same frequency as the electrical angular velocity of the rotor and a high frequency component caused by the applied high frequency voltage, as follows. it can.

(Equation 4)

Equation (4) The first term on the right side of each equation means a low-frequency component having the same frequency as the electrical angular velocity of the rotor, and this is clearly indicated by a footnote s. On the other hand, each expression
The term means a high frequency component, which is clearly indicated by a footnote h. All signals in equation (4) are 2 × 1 vectors. As is clearly shown in the final expression of the expression (4), the rotor magnetic flux vector φ _m directly related to the rotor is, of course, a signal on the low frequency side. These two components have greatly different frequencies, and can be separated using a filter having appropriate characteristics.

The high-frequency voltage vector ν _1h , the high-frequency current vector i _1h , and the high-frequency magnetic flux vector φ _1h , which are the high-frequency components of the stator voltage, current, and magnetic flux vectors, are expressed by the following equations (1) and (2). As is easily understood by extracting the high-frequency component using the relationship of the expression (4), the relationship of the following expression (5) holds.

Ν _1h = R ₁ i _1h + sφ _1h (5) At this time, the following equation (6) is given between the high-frequency magnetic flux vector φ _1h and the high-frequency current vector i _1h from the inverse salient pole characteristic. Relationship is established.

(Equation 6) Q (θ) in the equation (6) is a 2 × 2 mirror matrix defined by the following equation (7):

(Equation 7) The _L a, _{L b} denotes the inductance. The salient pole characteristics of the AC motor targeted by the present invention are represented by a mirror matrix Q (θ). Particularly reverse saliency is expressed in negative L _{b <0} inductance L _b present integrally with the mirror matrix.

The present invention uses the high-frequency current vector i _1h in the equations (5) and (6) to _obtain a high-frequency magnetic flux vector φ
_Common- mode current vector i rotating in the same direction as _1h
1ha and a mirror-phase current vector i _1hb rotating in the opposite direction to the high-frequency magnetic flux vector, and calculating the cosine / sine value of the intermediate angle between the in-phase current vector and the mirror-phase current vector or its estimated value by rotation. This is an estimated value of the child position.

In the present invention, the common-mode current vector i _1ha ,
The mirror phase current vector i _1hb is defined as in equations (8) and (9), respectively.

(Equation 8)

(Equation 9) The in-phase current vector i _1ha and the mirror-phase current vector i _1hb defined by the expressions (8) and (9) satisfy the following expression (10) with respect to the high-frequency current vector.

I _1h = i _1ha + i _1hb (10) Further, these three current vectors satisfy the relationship of the expression (6).

Next, it will be described that the intermediate value of the angles formed by the in-phase and mirror-phase current vectors can be used as the estimated value of the rotor position (that is, the position of the reverse salient pole of the rotor) θ. For the sake of simplicity, first, a unit vector constituted by the cosine and sine value of θ is defined as in equation (11).

(Equation 11) RF magnetic flux vector phi _1h, when the phase and theta _a, can be expressed as follows using the same.

(Equation 12) At this time, the in-phase current vector i _1ha is in phase with the high-frequency magnetic flux vector φ _1h from the equation (8), and can be evaluated as in the following equation (13).

(Equation 13)

On the other hand, the mirror phase current vectors are (9), (1)
Considering equation (2), it can be re-evaluated as equation (14).

[Equation 14] In (14), the reverse salient pole characteristic, _-L b> 0
Is established.

Equations (13) and (14) clearly explain that the in-phase and mirror-phase current vectors are in opposite phases with respect to the opposite salient pole position of the rotor. In other words, it indicates that an intermediate value between the phases of the two current vectors can be treated as an estimated value of the rotor position (that is, the reverse salient pole position) θ. The cosine and sine of the estimated value of the rotor position θ are, of course,
This is the estimated value of the sine. The present invention intends to use the estimated values of the cosine and sine obtained in this way for generating a rotation signal of a vector rotator necessary for forming a rotation dq coordinate system. As described above, the relationships among the unit vector, the high-frequency magnetic flux vector, the high-frequency current vector, the in-phase current vector, and the mirror-phase current vector indicating the rotor position described with reference to the equations (8) to (14) are provided for a clearer understanding. , In the form of a vector diagram.

The stator current vector can be detected without difficulty by the current detector 4 and the three-phase to two-phase converter 5a conventionally used in the vector control method or the vector control device. In addition, the in-phase current vector and the mirror-phase current vector can be detected from the stator current vector or estimated from the related signal by utilizing the frequency characteristics as specifically described in an embodiment described later. can do. As is clear from the above description, the estimation of the rotor position requires only the phase information of both the in-phase and mirror-phase current vectors, and does not need this amplitude information. Independence on amplitude information in rotor position estimation is
When considered in combination with the equations (8) and (9), it means that the rotor position can be estimated without depending on the inductance, that is, the motor parameter, in other words, in a state robust to the fluctuation of the motor parameter. are doing.

As is apparent from the above description, according to the present invention of claim 1 or claim 7, a vector rotator for vector control without using a rotor position detector mounted on a stator. An effect is obtained that an estimated value of the rotor position required for generating the rotation signal, and thus an estimated value of the cosine and sine of the rotor position and an estimated value of the rotor speed can be obtained. In addition, an effect is obtained that it can be obtained in a state that is robust to variations in the motor parameters. This operation is performed when the angular frequency of the applied high-frequency voltage vector is sufficiently high with respect to the electrical angular velocity of the rotor, that is, when the related signals are in a state where frequency separation is possible as specified in the equation (4). However, the present invention is not limited to when the rotor is stopped or when the rotor is stopped at an extremely low speed, but can be obtained in a wide operating range.

Next, the operation of the second aspect of the present invention will be described. Using the equations (8)-(14), it has been described that the in-phase vector and the mirror-phase vector are in opposite phases with respect to the reverse salient pole position θ of the rotor. This relationship is expressed by the rotor's reverse salient pole position θ, the in-phase and mirror phase current vector phases θ _a ,
using theta _b, it can be expressed as (15).

2θ = θ _a + θ _b (15)

The rotation signal required for the vector rotator for vector control is not the phase itself but the cosine and sine value of the phase, as described using the equation (3). That is, not only the expression (15) but also the relationship of the following expression (16) is important in application.

(Equation 16)

The right side of equation (16) can be calculated directly from the in-phase and mirror-phase current vectors. For example, the following equation (17) may be simply used.

[Equation 17] Equation (17) indicates that the double-angle cosine and sine value of the rotor reverse salient pole position is obtained by the normalized in-phase and mirror-phase current vectors.
In other words, it clearly shows that the determination is made only by the phase information of both vectors, and the consistency with the equation (16) is clear. Note that J in the expression (17) is a 2 × 2 alternation matrix defined by the following expression (18).

(Equation 18)

On the other hand, as is well known, regarding the double angle,
The trigonometric function relation of the following equation (19) generally holds.

[Equation 19] From this, it is clear that if the cosine / sine value of the double angle of the reverse salient pole position is known, the cosine / sine value of the reverse salient pole position can be determined using the relationship of equation (19).

According to a second aspect of the present invention, there is provided the vector control method according to the first aspect, wherein the cosine of a double angle of the intermediate angle is obtained from the in-phase current vector or the estimated value and the mirror current vector or the estimated value. The sine value or its estimated value is determined, and the cosine / sine value of the intermediate angle or its estimated value is determined from the determined cosine / sine value of the double angle. (17) - (19) As is apparent from the above description with reference to, according to the present invention, phase current vector, each phase theta _a mirror-phase current _vector, theta _b without calculating,
From these vectors, the effect is obtained that the estimated values of the cosine and sine required for generating the rotation signal of the vector rotator can be directly calculated. The calculation of the position from the vector requires an inverse operation of a trigonometric function, which is a type of nonlinear function. As is well known, this type of inverse operation causes a large error depending on the phase or requires a large amount of operation. Sometimes. However, according to the second aspect of the present invention, since this kind of inverse operation is not required, the estimated value of the cosine / sine of the rotor position can be determined with relatively high accuracy and with a relatively light calculation amount. The effect described is obtained. In other words, according to the second aspect of the present invention, the operation described in the first aspect can be obtained with relatively high accuracy and with a relatively light calculation amount.

Next, the operation of the present invention will be described. When calculating the cosine / sine value or the estimated value of the rotor reverse salient pole position from the cosine / sine value of the double angle of the rotor inverted salient pole position or the estimated value thereof, the first line of the equation (19) is used. In the case of using the relationship, a method of solving the square root is inevitably required. Further, when using the relationship in the second row of the equation (19), the division is required, although the solution of the square root is not required. In general, the amount of operation required for division is smaller than the amount of operation required for solving the square root, and therefore it is desirable to use the second row in an effort. However, division has a characteristic that causes a large error when the absolute value of the denominator is extremely small, so that it is practically necessary to avoid this as much as possible. As understood from the above description, for example, when the absolute value of the cosine value increases, the determination method shown in the following equation (20) is desirable.

(Equation 20) On the other hand, when the absolute value of the sine value increases, the following (2)
The determination method shown in the expression 1) is desirable.

(Equation 21)

According to a third aspect of the present invention, there is provided the vector control method according to the first or second aspect, wherein the cosine of the intermediate angle
According to the expected magnitude of the estimated value of the sine, the method of determining the cosine / sine value of the intermediate angle or the estimated value thereof from the determined value of the cosine / sine of the double angle is changed. As a result, as is clear from the above description using the equations (20) and (21), the vector rotator is maintained while maintaining the highest calculation accuracy and further reducing the calculation amount rationally. The effect of being able to determine the cosine and sine values required for the generation of the rotation signal or an estimated value thereof can be obtained. Consequently, according to the third aspect of the present invention, the effects described in the first and second aspects can be obtained with the highest calculation accuracy and with a reasonably reduced calculation amount.

Next, the operation of the present invention according to claim 4 will be described. The present invention according to claim 4 is the vector control method according to claim 1, wherein the in-phase current vector or a vector having the same direction as the estimated value and the mirror-phase current vector or the estimated Generate two vectors with the value and a vector having the same direction, and determine the estimated values of the cosine and sine of the intermediate angle in proportion to the composite vector obtained by adding the two same norm vectors. I am trying to do it.

FIG. 3 shows the state of the vector synthesis. In the drawing, the unit vector u a reverse salient pole position of the rotor and the cosine-sine value element (theta), the two vectors were identical the norm i _{1ha /} ‖i _1ha ‖ and i
_1hb / { _i 1hb _} . Also, a composite vector obtained by adding these vectors is represented by _ａa . It can be easily understood from the figure that the resultant vector has the same direction as the reverse salient pole of the rotor. The additive synthesis vector zeta _a has the same direction as the unit vector indicating the reverse salient pole position, strictly speaking, it can be described using equations, as follows.

(Equation 22) The expression (22) supports that the sum vector ζa is _a scalar multiple of the unit vector u (θ) indicating the reverse salient pole position, that is, is proportional to u (θ). Things.

The cosine / sine value of the reverse salient pole position of the rotor is
It is clear from equation (22) that the estimation can be made immediately according to the following relationship.

(Equation 23) That is, as specified in claim 4 of the present invention, the cosine and sine value of the intermediate angle or the estimated value thereof may be determined in proportion to the additive composite vector. In the present invention, (2)
As is apparent from the expression 3), the "proportional" relationship between the two vectors is captured, including not only the plus sign that the sgn function can take, but also the minus sign.

According to the fourth aspect of the present invention, (22)
It is easier to understand from the above explanation using equation (23). Except for the special case, the effect of being able to determine the cosine and sine of the intermediate angle required for generating the rotation signal of the vector rotator or the estimated value thereof with extremely simple calculation is obtained. As a result, according to the fourth aspect of the present invention, the operation described in the first aspect can be achieved by a very simple operation.
The method of avoiding the problem when the absolute value of the inner product of the high-frequency magnetic flux vector and the unit vector is small, and the method of determining the sign depending on the sgn function will be described later in connection with the embodiment related to the present invention. It will be explained in detail and in detail.

Next, the operation of the present invention will be described. According to a fifth aspect of the present invention, there is provided the vector control method according to the first aspect, wherein the in-phase current vector or the vector having the same direction as the estimated value, and the mirror-phase current vector or the estimated current value are used. Generates two vectors, a value and a vector having the same direction, and rotates a composite vector obtained by subtracting the two same norm vectors by π / 2 (rad), and is proportional to the subtracted composite vector after the rotation. Thus, the estimated values of the cosine and the sine of the intermediate angle are determined.

FIG. 4 shows the state of the vector synthesis. In the figure, the direction of the reverse salient pole of the rotor is represented by a unit vector u.
In (θ), two current vectors having the same norm are represented by i
_1ha / _{‖i 1ha} ‖ it is expressed in and _{i 1hb} / ‖i _1hb ‖ capital. Also, it expresses the resultant vector of these vectors subtraction zeta _s. The direction of the subtractive synthesis vector zeta _s is the direction perpendicular to the rotor opposite salient pole position,
It can be easily understood from FIG. This can be strictly described as follows using mathematical expressions.

(Equation 24) Equation (24) indicates that the direction of the subtraction composite vector points in the direction perpendicular to the unit vector u (θ) indicating the reverse salient pole position, and that the size is a scalar multiple thereof, and that the validity of the explanation with reference to FIG. This is a proof. The alternation matrix J defined by the equation (18) and used in the equation (24) has a function of rotating a vector acting on the matrix by π / 2 (rad).

The cosine and sine of the reverse salient pole position of the rotor are
It is clear from equation (24) that the estimation can be immediately made according to the following relationship.

(Equation 25) That is, as manifested in claim 5 of the present invention, the subtraction resultant vector zeta _s using a skew-symmetric matrix J π / 2 (rad)
The rotation may be performed, and the cosine / sine value of the intermediate angle or the estimated value thereof may be determined in proportion to the subtracted composite vector after the rotation. In the present invention, not only + sign that the sgn function can take but also −
It captures the "proportional" relationship between the two vectors, including the sign.

According to the fifth aspect of the present invention, (24)
It is easier to understand than the above explanation using equation (25). Except for special situations where the values are insignificant, very simple calculations
The effect is obtained that the cosine and sine values of the intermediate angle required for generating the rotation signal of the vector rotator or the estimated value thereof can be determined. As a result, according to the fifth aspect of the present invention, the operation described in the first aspect can be achieved by an extremely simple operation. A method for avoiding the problem when the absolute value of the alternating inner product of the high-frequency magnetic flux vector and the unit vector is small, and sgn
The method of determining the sign depending on the function will be described later specifically and in detail with reference to an embodiment related to the present invention.

Next, the operation of claim 6 of the present invention will be described. In consideration of the high-frequency property of the applied high-frequency voltage vector, the second term on the right side is dominant in the expression (5) showing the relationship between the high-frequency voltage vector, the high-frequency current vector, and the high-frequency magnetic flux vector. That is, the relationship of the expression (5) can be substantially expressed by the following expression (26) or (27) if the high frequency property of the signal is considered.

Ν _1h = sφ _1h (26)

[Equation 27]

As the expression (27) means, the high-frequency magnetic flux vector φ _1h indicates the integral value of the high-frequency voltage vector ν _1h. Therefore, if the applied high-frequency voltage vector has a sine wave shape, the high-frequency magnetic flux vector φ _1h Also has a sine wave shape. _Consequently , the in-phase current vector i _1ha and the mirror-phase current vector i _1hb defined by the equations (8) and (9) also have a sine wave shape. As a result, the detection or estimation of the cosine and sine values of the phase using both the in-phase and mirror-phase current vectors is stabilized.

As is clear from the above description, claim 6
According to the present invention, since the applied high-frequency voltage vector is a sine high-frequency voltage vector, the detection or estimation of the cosine and sine values of the phase using both the in-phase and mirror-phase current vectors is stabilized. Is obtained. As a result, according to the present invention of claim 6, claim 1, claim 2,
The operation described in claim 3, claim 4, and claim 5,
It can be achieved stably.

As described above, according to the present invention,
It has been explained that the cosine and sine values of the rotor position can be estimated in a state that is robust to fluctuations of the motor parameters. In addition, it was revealed that it has excellent characteristics in terms of estimation accuracy, calculation amount, and stability. This superior characteristic is also accepted in rotor speed estimation. This is because the rotor speed is in a relationship between the rotor position and the differential integration, and the cosine / sine value u of the rotor position is
It is clear from (θ) that it can be calculated according to the relationship of the following equation (28).

[Equation 28] Where ω _2m is the mechanical angular velocity of the rotor, and N _p is the number of pole pairs.

[0057]

Embodiments of the present invention will be described below in detail with reference to the drawings. Vector controller and AC motor (synchronous motor) to which vector control method of the present invention is applied
FIG. 5 shows the basic structure of one embodiment of the present invention. The basic difference between this structure and the structure based on the conventional control method described with reference to FIG. 18 is that the rotor position detector 2 and the cosine sine signal generator 7
And a velocity estimator 14 instead of the velocity detector 11 are newly introduced. Further, a high-frequency voltage commander 15 for applying a high-frequency voltage vector and a high-frequency current removing filter 16 for eliminating the high-frequency current vector i _1h resulting from mixing into the current controller are additionally provided. It is in. The other devices are basically the same as the device based on the conventional control method shown in FIG. 18, and the operation principle is the same as the conventional device. For example, the processing and generation procedure of the current vector i _1s and the voltage command vector ν _1s synchronized with the rotor speed in the embodiment of FIG. 5 are the same as the current vector i ₁ , voltage command is the same as the procedure of processing and generating vector [nu _1.

The core of the present invention resides in the vector position estimator 13. The speed estimator 14 is an estimator for estimating the speed of the rotor from the cosine / sine estimated value of the position, which is the output of the vector position estimator 13, and may utilize a conventional method well known to those skilled in the art. For example, the speed estimation method according to the equation (28) may be used in a form in which the differential processing is replaced with an approximate differential processing. In the present embodiment, an example of speed control is shown for comparison with the device according to the conventional method of FIG. 18, but it is also applicable to torque control as apparent to those skilled in the art. In the following, first, the high-frequency voltage commander 15 and the high-frequency current removal filter 16 introduced in connection with the application of the high-frequency voltage vector will be described, and then the vector position estimator 13 which is the core of the present invention will be described in detail. I do.

FIG. 6 shows the internal structure of one embodiment of the high-frequency voltage commander 15 for applying a high-frequency voltage vector. First, a high-frequency angular frequency omega _h input to the phase generator 15a, to determine the phase theta _h of the high frequency voltage vector. Although output serving high frequency phase theta _h of the phase generator 15a is basically the integral value of the high-frequency angular frequency omega _h, phase onset The frequency phase is input to a cosine sine signal generator 15b.
b generates its cosine and sine value u (θ _h ). Increasing the gain K _h to the cosine-sine values achieve the required voltage level The high-frequency voltage vector ν _1h is applied to the electric motor 1a.
At this time, the high-frequency voltage command unit 15, as best seen in FIG. 6, the high-frequency angular frequency omega _h and cosine-sine value u
(Θ _h ) can be output simultaneously.

In response to the application of the high-frequency voltage vector, a high-frequency current vector i _1h is generated, and this is included in the stator current vector i ₁ . In vector control, it is necessary to perform current control only on the stator current vector that contributes to torque generation. That is, it is necessary to control the current only for the fundamental wave component (low frequency component) i _1s included in the stator current vector i ₁ shown in Expression (4). The high-frequency current elimination filter 16 has been introduced for this purpose, and its role is to play the high-frequency current vector i _1h
In the current controller is eliminated, and a stator current vector i _1s contributing to torque generation is extracted.

Therefore, the high-frequency current removing filter 16 basically needs to have a filter characteristic capable of eliminating the high-frequency current vector i _1h . For example, this can be achieved with a band-stop filter or a low-pass filter. Band-stop filter F to the high frequency angular frequency ω _h and the center frequency of the band-stop (s)
Can be realized, for example, according to the following equation (29).

(Equation 29) Further, the low-pass filter F (s) having ω _h as a cutoff angular frequency can be realized, for example, according to the following equation (30).

[Equation 30]

[0062] In the example embodiment shown in FIG. 5, as figure clearly shows, the vector location estimator, the stator current vector i ₁ from the three-phase to two-phase converter 5a, a high-frequency voltage command 15 More high-frequency voltage vector information (high-frequency angular frequency ω _h , cosine / sine value u (θ _h )) is obtained as an input. The output of the vector position estimator is a cosine / sine value of an intermediate angle between the in-phase current vector and the mirror-phase current vector or an estimated value thereof, as will be described in detail later. In the example, this is used as the rotation signal itself for the two vector rotators as clearly shown in the figure.

FIG. 7 shows the internal structure of the vector position estimator 13. The vector position estimator 13 is mainly composed of two devices: an in-phase mirror phase current vector generator 13a and a cosine sine generator 13b. The in-phase mirror-phase current vector generator 13a detects or estimates an in-phase current vector and a mirror-phase current vector from a high-frequency current vector and outputs the same. That is, the in-phase mirror-phase current vector generator 13a includes an in-phase current vector rotating in the same direction as the high-frequency magnetic flux vector generated in response to the application of the high-frequency voltage vector, and a mirror-phase current vector rotating in the opposite direction to the high-frequency magnetic flux vector. Is realized. The cosine sine generator 13b obtains a detected value or an estimated value of the in-phase current vector and the mirror-phase current vector,
The cosine / sine value of the intermediate angle between these angles is output. That is, as the estimated value of the rotor position of the AC motor that can be used to generate the rotation signal of the vector rotator, the cosine / sine value of the intermediate angle between the mirror-phase current vector and the in-phase current vector or the estimated value thereof is generated. Means to do that.

FIG. 8A shows an example of a typical embodiment of the in-phase mirror-phase current vector generator 13a. In this example, the stator current vector i ₁ including the high-frequency current vector i _1h and the high-frequency angular frequency ω _h from the high-frequency voltage commander 15 are used as input signals, and the cosine and sine value u (θ of the high-frequency voltage vector are used. _h ) is not used.
The in-phase mirror phase current vector generator 13a includes two D-factor filters F (D) 1 which are a kind of variable characteristic multivariable filter.
3a-1 and 13a-2 are configured as main elements.
13a which receives the sign-inverted high-frequency angular frequency - [omega] _h
13a-2 receives the in-phase current vector i _1ha and the high-frequency angular frequency ωh without sign inversion, and 13a-2 detects and outputs the mirror-phase current vector i _1hb .

Here, the D-factor filter used in the present invention will be described. The D factor used in this filter is defined by the following equation (31) using the unit matrix I and the alternation matrix J.

[Equation 31] For example, a third-order D-factor filter F (D) is expressed by the following equation (32)-(34) using a D-factor.

F (D) = A ⁻¹ (D) B (D) (32)

A (D) = D ³ + a ₂ D ² + a ₁ D + a ₀ I (33)

B (D) = b ₃ D ³ + b ₂ D ² + b ₁ D + b ₀ I (34)

The above-mentioned D-factor filter applies a frequency characteristic F (s + j) to a scalar signal which is each component of a 2 × 1 vector.
ω _h ). This equivalent relationship is shown in FIG. Therefore, F (s) is designed to have a low-pass characteristic, and this coefficient is set to a D-factor filter F
If utilized for (D), F (D) will exhibit bandpass characteristics with -ω _h as the center frequency. Moreover, it has the property of polarity separation. The in-phase current vector and the mirror-phase current vector can be separated and detected by the band pass characteristic of the polarity separation. Realization of D-factor filter 1
As an example of the embodiment, a specific realization using an inverse factor D- ¹ and an inverse factor of a D factor for a third-order filter is shown in FIG.
(A) and (b) respectively.

The in-phase mirror phase current vector generator 13a
Instead of the factor filter, the configuration can be made by utilizing the vector rotator accompanying filter. Regarding the characteristics of the vector rotator entrainment filter, refer to literatures (Analysis of general characteristics of vector rotator entrainment filter for three-phase signal processing, Shin-China-Shinji, Journal of the Institute of Electrical Engineers of Japan, Vol. 120, No. 7,
pp. 947-948, Heisei 12), detailed description of the characteristics will be omitted. Introducing only the conclusion, the vector rotator entrainment filter exhibits a filter characteristic equivalent to a D-factor filter.

FIG. 10 shows an in-phase mirror-phase current vector generator 1.
As an example of another embodiment of 3a, an example using a vector rotator entrainment filter has been shown. The format of the vector rotator used in the present embodiment is as defined in Expression (3). However, as the cosine and sine signals for the vector rotator, the high-frequency voltage The cosine sine value u (θ _h ) is obtained directly and used. In this case, the high frequency angular frequency ω _h coming from the high-frequency voltage command unit 15 does not need to use. The filter F (sI) in FIG. 10 is obtained by replacing D with sI in the equations (32) to (34), that is, two ordinary low-pass filters are arranged in parallel so as to act on each scalar component of the vector signal. It was done.

In the embodiment shown in FIG. 10, the vector rotator accompanying filter enclosed by a broken line exhibits a filter characteristic equivalent to a D-factor filter. From the comparison between FIG. 8A and FIG. 10, the mutual relationship between the two embodiments is obvious.

When the high-frequency voltage vector ν _1h is a sinusoidal high-frequency voltage vector as shown in the embodiment of FIG. 6, the high-frequency magnetic flux vector φ _1h and the high-frequency magnetic flux vector φ _1h are obtained from the relations of equations (8) and (27). about π / 2 (r toward the rotary direction with respect to the high frequency voltage vector [nu _1h of the common mode current vector _{i 1ha}
ad). Estimation of the rotor position according to the present invention requires only phase information of both in-phase and mirror-phase current vectors and does not require amplitude information, as has been made clear in the description of the operation of the present invention. Rather, a current vector with a norm of 1 is more convenient. If this point is taken into consideration, the common-mode current vector whose norm has been normalized to 1 can be estimated according to the following equation (35).

(Equation 35)

In equation (35), i _1ha / ３５i _1ha
、 Is the cosine sine generator 1 inside the vector position estimator 13
3b in place of _i1ha . FIG. 11 shows an example of the embodiment of the in-phase mirror-phase current vector generator 13a using the in-phase current vector estimator 13a-4 using the relationship of the equation (35). In the present embodiment, the stator current vector, the high-frequency angular frequency of the high-frequency voltage vector, and the sine and cosine values of the phase of the same vector are input, and the estimated value of the in-phase current vector and the detected value of the mirror-phase current vector are output. I have. The same phase as the high-frequency angular frequency is
What is necessary is just to use the thing from the high frequency voltage commander 15. In this example, the D-factor filter is used for detecting the mirror phase current vector. However, a vector rotator accompanying filter may be used.

FIG. 12 shows an embodiment of the cosine sine generator 13b. 13b-1 in the figure is a double-angle cosine sine generator, and 13b-2 is an intermediate-angle cosine sine generator. Reference numeral 13b-3 denotes a determiner for generating a selection signal used for selecting a determination method in the intermediate angle cosine sine generator.

The double-angle cosine sine generator 13b-1 receives the in-phase current vector, the mirror-phase current vector or its estimated value as an input, and determines the cosine / sine value of the double angle of the intermediate angle of both vectors or its estimated value. And output. The determination process at this time is performed in accordance with the equation (17) used in the description of the operation of the present invention. (17)
The contents of the determination processing including the normalization processing of the in-phase current vector and the mirror-phase current vector, which are expressed by the expressions, are more than obvious to those skilled in the art.

The intermediate angle cosine sine generator 13b-2 receives as input the double angle cosine sine value output by the double angle cosine sine generator 13b-1 or its estimated value, and uses this to obtain the intermediate angle cosine.・ The sine value or its estimated value is determined and output.

According to the present invention, the cosine / sine value of the intermediate angle or its estimated value is determined from the cosine / sine value of the double angle or its estimated value according to the expected magnitude of the cosine / sine value of the intermediate angle. I try to change the way. For example,
The intermediate angle cosine sine generator 13b-2 has the following (36)-
Four types of determination methods as shown in the equation (39) are prepared, and one of the four types of determination methods is selected according to the expected magnitude of the cosine / sine value of the intermediate angle. It has become.

[Equation 36]

(37)

[Equation 38]

[Equation 39] In the equations (36)-(39), considering the fact that the magnitude of the cosine / sine value of the intermediate angle is directly dependent on the value of the intermediate angle itself, the selection condition of the determination method is expected. The value of the intermediate angle is shown on the rightmost wing of each equation.

The determiner 13b-3 determines the expected magnitude of the cosine / sine value of the intermediate angle and plays a role in selecting the above-described determination method. It is practical that the processing of each device in the present invention is performed digitally. As a specific example of the embodiment, FIG. 12 shows a case in which digital processing is taken into consideration and the expected value at the present time is determined using the determined value of the cosine and sine of the intermediate angle one control cycle before. An example is illustrated. For the sake of simplicity, let the current time be k,
One control cycle before is the time (k-1), and the determined value of the cosine / sine of the intermediate angle at the time (k-1) is expressed as u (θ, k-1). In FIG. 12, z- ¹ is a delay element for one control cycle, and has a function of delaying an input signal by one control cycle and outputting the delayed signal. The operation after this element is as follows. First, the cosine / sine determination value of the intermediate angle at the time point (k-1) is processed according to the following equation (40), and the determination indices d ₁ (k) and d ₂ (k) at the time point k are calculated. Generate.

(Equation 40) Next, using this determination index, it is determined which of equations (36)-(39) should be adopted as the determination method at the time point k. The judgment based on the indexes d ₁ (k) and d ₂ (k) is shown in FIG.
In accordance with the method described in The method of FIG.
The judgment is made only by the sign of the judgment index, and the third row is the output (selection result) in response to the input of the first and second rows (the sign of the judgment index). As described above, the present invention described with reference to the embodiment has high utility in which rational adoption can be easily performed. In order to correctly determine the cosine / sine value of the intermediate angle from the cosine / sine value of the double angle, as will be understood from the repeatability of the determiner 13b-3 described above, the position of the inverted salient pole It is assumed that the initial estimated value of (N pole) is correctly recognized without being mistaken for the S pole. The initial position estimation is performed before the drive control of the electric motor, and various methods of the initial position estimation that can be directly used for this are described in the literature (IEEJ Technical Report No. 76).
No. 0) has already been described in detail, and may be used.

FIG. 14 shows a second embodiment of the cosine sine generator 13b. In the present embodiment, when information on both in-phase and mirror-phase current vectors is input, first, processing for equalizing norms of both vectors is performed. Specifically, normalization processing is performed so that the norm becomes 1. Next, according to the first expressions of the expressions (22) and (24) shown in the description of the operation of the present invention, an addition composite vector ζa and a subtraction composite vector ζs are simultaneously generated. Subtractive synthesis further multiplied by J in vector zeta _s, the zeta _a and Jzeta _s decision normalizer 13b-4
Enter

The processing of multiplying the subtracted combined vector by J is performed by rotating the subtracted combined vector by π / 2 (rad),
This is for determining a cosine / sine value of an intermediate angle between the in-phase current vector and the mirror-phase current vector or an estimated value thereof in proportion to the subtracted combined vector after rotation. In the case of the additive composite vector, the cosine / sine value of the intermediate angle or the estimated value thereof is determined in proportion to the additive composite vector itself, so that the process of multiplying by J is unnecessary.

In the decision normalizer 13b-4, the following (4)
1) According to the equation, a composite vector ζ is selected.

(Equation 41) As shown in Expressions (22) and (24), the added combined current vector and the subtracted combined current vector may be extremely small in the relationship between the high-frequency magnetic flux vector and the rotor reverse salient pole position. In this case, the rotor position cannot be properly estimated, as has been clarified in the description of the operation of the present invention. The selection process specified in the equation (41) takes this point into consideration, and utilizes the mutually complementary characteristic of both combined vectors that the addition combined vector and the subtraction combined vector are not reduced at the same time. In this case, a combined current vector is selected. With this selection process, the rotor position can be accurately estimated. In the decision normalizer 13b-4, the combined vector ζ obtained by the selection processing is further normalized so that the norm becomes 1, and the normalized combined vector _{ｎ n} defined by the following equation (42).
Output.

(Equation 42)

This embodiment utilizes the relationship between the second expressions of the expressions (23) and (25) used in the description of the operation of the present invention. In the processing described above, (2
The generation of the unit vectors in the equations (3) and (2) is completed. Next, the code to be multiplied by the unit vector must be determined. As shown in Expressions (23) and (25), the sign is determined by the relative positional relationship between the high-frequency magnetic flux vector and the rotor salient pole position. That is, it is not determined by the absolute position of the rotor salient pole. In the present invention, the high-frequency magnetic flux vector rotates relatively fast with respect to the rotor salient pole position. Therefore, in the present invention, the change in the sign may be regarded as a change in the high-frequency magnetic flux vector, and further, a change in the common-mode current vector and the mirror-phase current vector directly related to the change. 13b-5 in the embodiment of FIG. 14 is a code determiner configured from this viewpoint.

As described in the embodiment with reference to FIG. 12, it is practical that the processing of each device in the present invention is performed digitally. The processing of the sign determinator in FIG. 14 considers digital processing as a practical embodiment,
An example is shown in which the current sign is determined using the estimated cosine / sine value of the rotor position one control cycle before. For the sake of simplicity, the current time is set to k, the control cycle before is set to (k-1), and the estimated cosine / sine value of the rotor position at (k-1) is u (θ, k −1), and the normalized composite vector at the time point k is expressed as ζ _n (k). In FIG. 14, z- ¹ is a delay element for one control cycle, and has a function of delaying an input signal by one control cycle and outputting the delayed signal. In the sign determination unit 13b-5, first, u
The sign of (θ, k-1) is assumed to be correct and u (θ, k-
The inner product of 1) and the current ζ _n (k) is taken, and then the sign is determined. The sign of the inner product can be easily obtained by processing the sgn function. The rotation of the high-frequency magnetic flux vector is
If the inner product is negative, this is due to the rotation of the high-frequency magnetic flux vector, since it is much faster than the rotation of the rotor salient poles. Therefore, when the sign of the inner product is inverted, the sign of the composite vector _{ｎ n} (k) at the current time is an error, and the sign of ζ _n (k) needs to be inverted. ζ
_{In order to} invert the sign of _n (k), ζn (k) should be multiplied by -1. Since -1 for sign inversion is obtained as an output of the sgn function processing, it may be used. FIG.
The sign determination unit in FIG. 2 expresses this processing step in a block diagram.

The above-described sign determination of the repetitive form is based on the premise that the initial sign determination is correctly performed. Although the initial code determination is performed before the drive control of the motor, various methods of initial position estimation that can be directly used for this are already described in detail in the literature (IEEJ Technical Report No. 760) and the like. You can make use of this.

FIG. 15 shows another embodiment of the present invention which is directed to a synchronous motor, but is an alternative to FIG. In the figure, in order to avoid the congestion in the figure, the detected current and the voltage command value on the rotating dq coordinate system are collectively expressed by vector signal lines. The difference between the embodiment of FIG. 15 and FIG. 5 is that the vector position estimator 1 related to the rotor position estimation is different.
3. The point is that the high-frequency voltage commander 15 and the high-frequency current removal filter 16 are moving from the fixed αβ coordinate system to the rotating dq coordinate system. The configuration method of the three devices of the vector position estimator 13, the high-frequency voltage commander 15, and the high-frequency current removal filter 16 is basically the same as the case shown in the embodiment of FIG. The cosine sine generator 13b, which is an internal element of the vector position estimator 13, outputs a cosine sine estimated value of the inverted salient pole position on the adopted coordinate system. Therefore, FIG. 5 outputs a cosine sine estimated value of the reverse salient pole position on the fixed αβ coordinate system,
In FIG. 15, a cosine sine estimated value of the reverse salient pole position on the rotating dq coordinate system is output. On the other hand, the vector rotators 6a and 6b
Requires the cosine sine value of the inverted salient pole position evaluated on the fixed αβ coordinate system, so that the embodiment of FIG. 15 additionally requires a coordinate conversion process. This additional processing can be easily implemented as described below.

Assuming that all the processes of the vector position estimator 13 are performed digitally, the output of the cosine sine generator 13b at the time point k is converted to u (θ) as shown in the embodiment of FIGS. And On the other hand, the final output of the vector position estimator 13 at the time point k is u (θ (k)). u
(Θ (k)) is represented by the following (4) as a coordinate conversion value of u (θ).
It is obtained by the processing of equation 3).

U (θ (k)) = R (θ (k−1)) u (θ) (43) FIG. 16 shows a configuration example of the vector position estimator 13 including the processing of the equation (43). Was. In the figure, the coordinate conversion process specified by the equation (43) is performed by the vector rotator 13c.

Next, an embodiment in which the vector control device based on the vector control method of the present invention is applied to an induction motor having reverse salient poles will be described. FIG. 17 shows the basic structure. The basic difference between this structure and the structure according to the conventional control method shown in FIG. 19 is that a vector position estimator 13 replaces the rotor position detector 2 and a speed estimator 14 replaces the speed detector 11. It has been newly introduced. Further, a high-frequency voltage commander 15 for applying a high-frequency voltage vector and a high-frequency current removing filter 16 for eliminating the high-frequency current vector i _1h resulting from mixing into the current controller are additionally provided. It is in. The other devices are basically the same as the device based on the conventional control method shown in FIG. 19, and the operation principle is the same as the conventional device.
For example, the current vector i in the embodiment of FIG.
The processing and generation procedure of _1s and the voltage command vector ν _1s , and the generation procedure of the rotation signal for the vector rotator are the same as those of the conventional apparatus described with reference to FIG. The devices newly introduced in connection with the present invention are the vector position estimator 13, the speed estimator 14, the high-frequency voltage commander 15, and the high-frequency current removal filter 16, as described above. These newly introduced devices in the embodiment of FIG. 17 are the same as those in the embodiment of FIG. 5, as will be easily understood by those skilled in the art by comparing FIG. 17 with FIG. Therefore, the description of the newly introduced device is the same as that of FIG.

As described above, in the present invention, the embodiment of the synchronous motor described with reference to FIG.
7 corresponds to the embodiment of the induction motor described with reference to FIG. Further, in the present invention, the embodiment of the synchronous motor described with reference to FIG. 5 is easily developed into the embodiment of the synchronous motor described with reference to FIG. As will be easily understood by those skilled in the art, an embodiment of the induction motor corresponding to the embodiment of the synchronous motor described with reference to FIG. 15 can be easily obtained by utilizing this correspondence. More specifically, the devices newly introduced in FIG. 17 may be moved from the fixed αβ coordinate system to the rotating dq coordinate system. The manner of movement at this time is the same as that of the embodiment in FIG.

The vector position estimator according to the present invention has been described in detail and in detail using a plurality of embodiments with reference to various drawings. As has been stated repeatedly in the description, the vector position estimator of the present invention is preferably constructed digitally given the recent significant advances in digital technology. The digital configuration includes a hardware configuration and a software configuration, but any one of the present invention can be configured as will be apparent to those skilled in the art.

[0088]

As is clear from the above description, the present invention has the following effects. In particular, the present invention of claims 1 and 7 uses a high-frequency current vector generated in response to the application of a high-frequency voltage vector, and uses a high-frequency current vector rotating in the same direction as the high-frequency magnetic flux vector, and a direction opposite to the high-frequency magnetic flux vector. To detect or estimate the rotating mirror-phase current vector, and use the cosine and sine of the intermediate angle between the in-phase current vector and the mirror-phase current vector or the estimated value to generate the rotation signal of the vector rotator. I have to. As a result, according to the present invention of claim 1 or claim 7, without using the rotor position detector mounted on the stator,
An effect is obtained that an estimated value of the rotor position necessary for generating a rotation signal of the vector rotator for vector control, and consequently, an estimated value of the cosine and sine of the rotor position and an estimated value of the rotor speed can be obtained. . In addition, an effect is obtained that the motor parameters can be obtained in a state that is robust to the fluctuations. This effect can be obtained in a wide operation range without being limited to when the rotor is stopped or at an extremely low speed equivalent to when the rotor is stopped, if the angular frequency of the applied high-frequency voltage vector is sufficiently higher than the electrical angular velocity of the rotor. As a result of this operation, the vector rotator, which is indispensable for the vector control of the AC motor, can be normally operated in a wide operating range while being robust against parameter fluctuations. An effect is obtained that vector control of the AC motor can be performed without using a detector. Further, in the vector control of the AC motor, the reliability of the motor system is reduced, the volume in the axial direction is increased, wiring problems, and various costs are conventionally caused by mounting the rotor position detector on the rotor. The effect of being able to overcome various problems such as an increase in the number is obtained.

In particular, a second aspect of the present invention is the vector control method according to the first aspect, wherein the cosine of the double angle of the intermediate angle is calculated from the in-phase current vector or the estimated value thereof and the mirror current vector or the estimated value. Since the sine value or its estimated value is determined, and the cosine / sine value of the intermediate angle or its estimated value is determined from the determined cosine / sine value of the double angle, the in-phase current vector and the mirror-phase current vector are determined. An effect is obtained in that the estimated values of the cosine and sine required for generating the rotation signal of the vector rotator can be directly calculated from these vectors themselves without calculating the position. In addition, since an inverse operation for calculating the rotor position is not required, an effect that the estimated value of the cosine / sine of the rotor position can be determined with relatively high accuracy and with a relatively small amount of calculation can be obtained. Further, according to the second aspect of the present invention, the operation described in the first aspect can be rationally secured. As a result of such an operation, the effect of claim 1 can be achieved with relatively high accuracy and a relatively small amount of calculation.

According to a third aspect of the present invention, there is provided the vector control method according to the first and second aspects, wherein the cosine / sine of the double angle is adjusted according to the expected magnitude of the estimated value of the cosine / sine of the intermediate angle. The method for determining the cosine / sine value of the intermediate angle or the estimated value of the intermediate angle from the determined value is changed, and the rotation signal of the vector rotator is maintained while maintaining the highest calculation accuracy while reducing the calculation amount. Has the effect of being able to determine the estimated values of the cosine and sine required to generate. As a result of this operation, according to the present invention of claim 3, claims 1 and 2
Is obtained while maintaining the highest calculation accuracy while reducing the amount of calculation.

According to a fourth aspect of the present invention, there is provided the vector control method according to the first aspect, wherein the norm is equalized, and the in-phase current vector or a vector having the same direction as the estimated value, the mirror-phase current vector or the Generate two vectors of the estimated value and the vector having the same direction, and calculate the cosine and sine of the intermediate angle or the estimated value in proportion to the composite vector obtained by adding the two identical norm vectors. I'm trying to decide. Thus, except for the special case where the absolute value of the inner product of the high-frequency magnetic flux vector and the unit vector of the salient pole position is small, the calculation is extremely simple,
The effect is obtained that the estimated values of the cosine and the sine required for generating the rotation signal of the vector rotator can be determined. As a result, the effect of claim 1 can be achieved by an extremely simple operation.

According to a fifth aspect of the present invention, there is provided the vector control method according to the first aspect, wherein the in-phase current vector having the same norm, the vector having the same direction as the estimated value, the mirror-phase current vector or the Generate two vectors, the estimated value and the vector having the same direction, rotate the combined vector obtained by subtracting the two identical norm vectors by π / 2 (rad), and proportional to the subtracted combined vector after the rotation. Then, the cosine and sine values of the intermediate angle or their estimated values are determined. Thus, except for a special case where the absolute value of the alternating dot product between the high-frequency magnetic flux vector and the unit vector is small, the estimated value of the cosine and sine required for generating the rotation signal of the vector rotator can be calculated with extremely simple calculation. Can be determined. As a result, the effect of claim 1 can be achieved by an extremely simple operation.

In the case of the present invention according to claims 4 and 5, the use of the high-frequency magnetic flux vector is limited in a region where the same component as the rotor reverse salient pole or the vertical component is small as a compensation for the reduction in the amount of calculation. Is done. In order not to lose the effects of the present invention of claims 4 and 5, a practical method for avoiding this area has been specifically shown and described in connection with the embodiments of claims 4 and 5. .

According to the present invention, since the applied high-frequency voltage vector is a sine high-frequency voltage vector,
The effect is obtained that the detection or estimation of the cosine and sine values of the phase using both the in-phase and mirror-phase current vectors becomes stable. As a result, according to the present invention of claim 6, claim 1, claim 2, claim 3, claim 4, and claim 5
The effect of being able to stably achieve the effect of the above is obtained.

[0095]

[Brief description of the drawings]

FIG. 1 is a vector diagram showing a rotor reverse salient pole position (electrical angle) on a fixed αβ coordinate system.

FIG. 2 shows high-frequency magnetic flux, high-frequency current, in-phase current, mirror-phase current, and rotor reverse salient pole position 1 on a fixed αβ coordinate system.
Vector diagram showing a relationship example

FIG. 3 is a vector diagram showing one example of a relationship between an additive combined vector and a direction of a rotor reverse salient pole on a fixed αβ coordinate system.

FIG. 4 is a vector diagram showing an example of a relation between a subtraction combined vector and a direction of a rotor reverse salient pole on a fixed αβ coordinate system.

FIG. 5 is a block diagram showing a basic configuration of a vector control device according to the embodiment;

FIG. 6 is a block diagram showing a basic configuration of a high-frequency voltage commander according to the embodiment;

FIG. 7 is a block diagram illustrating a schematic configuration of a vector position estimator according to the embodiment;

8A is a block diagram illustrating a schematic configuration of an in-phase mirror-phase current vector generator according to an embodiment, and FIG. 8B is a diagram illustrating equivalent characteristics of a D-factor filter.

FIGS. 9A and 9B are a block diagram showing a method for realizing an inverse factor D- ¹ and a block diagram showing an embodiment of a D-factor filter using D- ¹ . FIGS.

FIG. 10 is a block diagram showing a schematic configuration of an in-phase mirror-phase current vector generator according to the embodiment;

FIG. 11 is a block diagram showing a schematic configuration of an in-phase mirror-phase current vector generator according to the embodiment;

FIG. 12 is a block diagram illustrating a schematic configuration of a cosine sine generator according to the embodiment;

FIG. 13 shows an example of a relation between a decision index used for selecting a decision method in the intermediate angle cosine sine generator and a selection result

FIG. 14 is a block diagram showing a schematic configuration of a cosine sine generator according to the embodiment;

FIG. 15 is a block diagram showing a basic configuration of a vector control device according to the embodiment;

FIG. 16 is a block diagram illustrating a schematic configuration of a vector position estimator according to the embodiment;

FIG. 17 is a block diagram showing a basic configuration of a vector control device according to the embodiment;

FIG. 18 is a block diagram showing a schematic configuration of a conventional vector control device.

FIG. 19 is a block diagram showing a schematic configuration of a conventional vector control device.

[Explanation of symbols]

 1a AC motor (synchronous motor) 1b AC motor (induction motor) 2 Rotor position detector 3 Power converter 4 Current detector 5a Three-phase two-phase converter 5b Two-phase three-phase converter 6a Vector rotator 6b Vector rotator 7 Cosine sine signal generator 8 Current controller 9 Command converter 10 Speed controller 11 Speed detector 12 Magnetic flux phase estimator 13 Vector position estimator 13a In-phase mirror phase current vector generator 13a-1 D-factor filter 13a-2 D Factor filter 13a-3 Vector rotator companion filter 13a-4 In-phase current vector estimator 13b Cosine sine generator 13b-1 Double angle cosine sine generator 13b-2 Intermediate angle cosine sine generator 13b-3 Judge 13b-4 Judgment Normalizer 13b-5 Sign determiner 13c Vector rotator 14 Speed estimator 15 High-frequency voltage commander 15 Phase generator 15b cosine and sine signal generator 16 frequency current removing filter

 ──────────────────────────────────────────────────続 き Continuing on the front page F term (reference)

Claims (7)

[Claims] 1. A stator current contributing to torque generation is represented by a d-axis and a q
On a rotating dq coordinate system composed of axes, the current vector d
A vector control method for an AC motor, comprising: a current control step of dividing and controlling an axis component and a q-axis component, and a high-frequency voltage applying step of applying a high-frequency voltage that can be treated as a rotating high-frequency voltage vector as at least a part of a stator voltage. And using a high-frequency current vector generated in response to the application of the high-frequency voltage vector, an in-phase current vector rotating in the same direction as the high-frequency magnetic flux vector generated in response to the application of the high-frequency voltage vector, and the high-frequency magnetic flux vector. Detects or estimates the mirror-phase current vector rotating in the opposite direction, and calculates the cosine / sine value of the intermediate angle between the in-phase current vector and the mirror-phase current vector or its estimated value, and calculates the rotation of the AC motor. AC motor characterized in that it is used as a cosine / sine estimated value of a child position for generating a rotation signal of the vector rotator. Machine vector control method. 2. A cosine / sine value of twice the intermediate angle or an estimated value thereof is obtained by using the in-phase current vector or the estimated value of the in-phase current and the mirror current vector or the estimated value of the in-phase current vector. 2. The vector control method for an AC motor according to claim 1, wherein the cosine / sine value of the intermediate angle or an estimated value thereof is determined from the determined value of the double angle cosine / sine. 3. A method for determining the cosine / sine value of the intermediate angle or the estimated value thereof from the double-valued cosine / sine value according to the expected magnitude of the cosine / sine value of the intermediate angle. 3. The method according to claim 1, wherein
A vector control method for the AC motor as described in the above. And generating two vectors, i.e., the in-phase current vector or the in-phase estimated value thereof and the mirror-phase current vector or the in-phase estimated value having the same norm. 2. A vector control method for an AC motor according to claim 1, wherein the cosine / sine value of the intermediate angle or an estimated value thereof is determined in proportion to a composite vector obtained by vector addition of the same norm vector. . 5. A method for generating two vectors, i.e., an in-phase current vector or an in-phase estimated value thereof and a mirror-phase current vector or an in-phase estimated value having the same norm. A composite vector obtained by vector subtraction of the same norm vector is rotated by π / 2 (rad), and a cosine / sine value of the intermediate angle or an estimated value thereof is determined in proportion to the subtracted composite vector after rotation. The vector control method for an AC motor according to claim 1, wherein: 6. A vector control method for an AC motor according to claim 1, wherein said high-frequency voltage vector is a sine high-frequency voltage vector. 7. A stator current contributing to torque generation is defined by d-axis and q-axis orthogonal to each other indicated by a vector rotator.
On a rotating dq coordinate system composed of axes, the current vector d
A vector control device for an AC motor having current control means for dividing and controlling an axis component and a q-axis component, and high-frequency voltage applying means for applying a high-frequency voltage that can be treated as a rotating high-frequency voltage vector as at least a part of the stator voltage A stator current vector generated in response to the application of the high-frequency voltage vector, an in-phase current vector rotating in the same direction as the high-frequency magnetic flux vector also generated in response to the application of the high-frequency voltage vector, and the high-frequency magnetic flux vector Means for detecting or estimating a mirror phase current vector rotating in the opposite direction to the direction of rotation, and the cosine and sine estimated values of the rotor position of the AC motor that can be used for generating a rotation signal of the vector rotator. Means for generating a cosine / sine value of an intermediate angle between the current vector and the mirror phase current vector or an estimated value thereof. A vector control device for an AC motor, comprising: