JP2001161099A - Control scheme for synchronous motor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To solve problems that arise associated with convergence computation when d- and q-axis currents are determined based on only torque commands and voltage conditions in controlling IPM and problems of voltage saturation and magnetic saturation. SOLUTION: Q-axis current computation estimates from a magnetic flux table 11 a magnetic flux component corresponding to the current of a rotating coordinate system that rotates in synchronism with a field system, and an output torque computation part 13 estimates from a q-axis current and the result of the magnetic flux estimation a torque that is produced in this state of current. A torque current computation part 14 determines a q-axis current component that corrects the component of error between the estimated torque and a torque command so that the estimated torque is matched with the torque command, adds up the q-axis current component and the present q-axis current, and outputs the result as a new torque current command. An inductance table 12 calculates self-inductance on the d-axis and the q-axis when the current of the rotating coordinate system varies, and uses it for q-axis current computation. In addition, a voltage saturation prevention controlling means is included which, if the output voltage is increased and exceeds a voltage limit value, produces a demagnetization current.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a synchronous motor (IP) using a permanent magnet embedded in a field core as a field source.
M: Interiormounted Permanen
The present invention relates to a control method of TmagnetMotor), particularly to a torque control method.

[0002]

2. Description of the Related Art (1) Features of IPM It is known that IPM is efficient and small in size for induction motors. Factors of high efficiency are as follows: (a) Induction motors required an exciting current component to generate magnetic flux, but this is not present in synchronous motors. Therefore, copper loss due to the primary resistance is reduced. (B) The induction motor has a copper loss due to the resistance of the secondary circuit due to the principle of passing a current through the secondary circuit, but does not exist in the synchronous motor.

[0003] The factor of miniaturization is that the thickness of the secondary conductor of the induction machine is sufficient only by the thickness of the permanent magnet, so that the radius on the field side is reduced and the outer dimensions of the motor can be reduced.

(2) Type of IPM Synchronous motors are classified based on the shape of the field core, and are classified into SPM (Surfacemounted Pe).
rmanentmagnetMotor), IPM (I
interiormountedPemmentmag
netMotor).

[0005] The SPM is formed by attaching a permanent magnet in the shape of a cylinder to a cylindrical iron core on the surface thereof. The armature inductance viewed from the terminal has a cylindrical synchronous machine characteristic in which the inductance components of the d axis (field axis) and the q axis (axis orthogonal to the field axis) are equal.

On the other hand, in the case of the IPM in which the permanent magnet is embedded in the core, the armature inductance is
It has salient-pole synchronous machine characteristics in which the inductance components of the d-axis (field axis) and the q-axis (axis orthogonal to the field axis) are different.

The present invention relates to a control system having saliency such as IPM, particularly to a torque control system. The following document 1 describes the characteristics of IPM.

Reference 1 "Maximum torque control of PM type brushless DC motor", Keita Hatanaka, Douki, Shigeo Morimoto, Yoji Takeda, Tagao Hirasata (Osaka Prefectural University), Semiconductor Power Conversion Research Group, SPC-91-6 (3) Relational expression under linear condition of IPM Since IPM has salient pole characteristics, reluctance torque due to saliency is generated in addition to torque due to magnetic flux and current of a normal permanent magnet. There is a control method for improving the output torque of the motor by making effective use.

First, a description will be given of a relational expression in a linear case where the parameters of the motor are constant values. The voltage-current equation when the inductance is a fixed constant is as follows.

[0010]

(Equation 1)

Here, a two-dimensional orthogonal coordinate system (dq axes) that rotates in synchronization with the field poles is defined as shown in FIG. Also,
Regarding the motor, the equivalent circuit of FIG. 7 is handled.

In the above equation, the inductance has a leakage component and a rotor interlinkage component, and is expressed by the following equation.

[0013]

(Equation 2)

The torque equation can be obtained by multiplying the magnetic flux term having the electric angular velocity ω by the current transposed from the left side, and is expressed by the following equation.

[0015]

(Equation 3)

Although not directly related to the present invention,
Assuming that the leakage inductances on the d and q axes are equal,
Since the leakage inductance included in the winding inductance is equally divided between the two inductances in the second term of the equation (3), either the primary inductance or the secondary inductance may be used in the torque equation. Understand.

[0017]

(Equation 4)

The first term of the above equation (4) is a torque component generated by the permanent magnet and a current component orthogonal thereto, and
The term becomes a term of reluctance torque due to saliency.

(4) Relational expression between current and torque (maximum torque condition) According to the above expression (3), the torque is changed by the current components of both the d and q axes due to the influence of the reluctance torque. . When this is expressed as a contour line of the torque on the d and q current coordinates, it becomes a hyperbolic shape as shown in FIG.

When the equations (1) and (3) are transformed into equations using torque as a parameter, and the torque equation of the equation (3) is transformed so that the left side becomes iq, the following equation (5) is obtained. Becomes

[0021]

(Equation 5)

Here, λd−id · (Ld−Lq) is a denominator, and when a contour line of constant torque is drawn, a hyperbola as shown in FIG. 8 is obtained.

If an auxiliary circle centered on the origin is drawn on the current coordinates shown in FIG. 8 so as to be in contact with the contour line of the torque, the minimum current condition at which the contact can generate torque is obtained. Operation at this point provides highly efficient operating conditions with minimal copper loss.

The physical meaning of this point is that, when the current vector having a constant amplitude is rotated only in phase on the auxiliary circle, the torque is represented by the phase of the current vector on the horizontal axis and the torque on the vertical axis as shown in FIG. It is easy to understand when you draw. In fact, the minimum current operating point in FIG. 8 is a condition for generating the maximum torque under the condition that the current amplitude is constant. For this reason, the method of controlling the current vector on the locus of the maximum torque in FIG. 8 is often referred to as “maximum torque control”, and this method will be used in the following description.

(5) Relationship between Current and Voltage Since the output voltage is limited in the constant output range, it is necessary to supply a demagnetizing current. In order to consider the voltage-current equation of the above equation (1) in a steady state, the term of p is approximated to zero, and the magnitude of the voltage vector is obtained as the following equation (6).

[0026]

(Equation 6)

Further, when deformed, the following equation (7) is obtained, and the contour line of the voltage on the current coordinates becomes an elliptic equation.

[0028]

(Equation 7)

FIG. 10 illustrates this ellipse.
When there is a voltage limit, the limit range becomes elliptical, and the voltage exceeds the limit value for a current vector that goes out of the range. Therefore, it is necessary to satisfy two conditions, that is, the torque expression of Expression (3) and the voltage of Expression (7).

When these conditions are shown on the same current coordinates,
As shown in FIG. 11, the intersection of the voltage and torque contours is the operating point in the constant output range.

In order to find this operating point, equations (3) and (7) must be solved simultaneously, which is algebraically complicated. The calculation time is limited, and there are actually non-linear elements such as inductance. Therefore, the calculation must be performed by convergence calculation.

The above is a description of the characteristics of the IPM when it is assumed that the constant is constant and linear.

(6) Conventional Basic Control Method Reference 2 below discusses developing a practical torque command into q- and d-axis current components.

Reference 2 "High-efficiency variable speed drive of IPM motor", Naotake Shibata, Toshiaki Idemitsu, Ken Kameyama (Yaskawa Electric), Transactions of the Institute of Electrical Engineers of Japan, MID-98-12.

[0035]

In the above-mentioned document 2, a torque command and a voltage condition (voltage limit due to voltage saturation) are given, and a q and d-axis current command that satisfies them is obtained. Since the relationship between voltage, current, and torque is not a linear relationship, trigonometric functions and square roots can be used,
An approximate solution is obtained by performing convergence calculation. The first problem is how to construct this convergence operation.

As a second problem, when voltage saturation occurs, it is necessary to supply a demagnetizing current to control the terminal voltage. The terminal voltage may increase due to the voltage component due to the armature reaction magnetic flux, and the voltage may exceed the outputtable voltage of the converter and become a voltage saturated state. In such a case, current control becomes impossible and becomes unstable.

Although there is a method of setting the induced electromotive force itself low so as not to cause voltage saturation, the current rating is increased by the reduced voltage, and the capacity required for the driver is increased.

Therefore, the rated voltage of the motor is set to a high value, the rated current is set to a high value, the DC power supply voltage and the output voltage are monitored, and when the voltage reaches the saturation, the demagnetization is performed. Control for preventing saturation of the terminal voltage by flowing a current is required.

The third problem is due to magnetic nonlinearity. In the above-mentioned literature, it is assumed as a precondition that the inductance of the motor is constant. But,
Inside the iron core of the actual IPM, the magnetic flux of the permanent magnet and the armature reaction magnetic flux are combined, so that a magnetic saturation of the iron core occurs due to a change in the magnetic flux, and the inductance component changes due to a load current or the like.

An example of the non-linearity of the magnetic flux is shown as a contour line of each magnetic flux component tail of the d and q axes on the dq coordinate of the current.
It is an example of FIG. 12, FIG.

Originally, if the inductance is constant, the contour lines of the magnetic flux are parallel to each axis and at equal intervals. However,
Actually, due to the magnetic saturation inside the iron core, d, q
Due to the interference of the currents of the axial components, the magnetic flux changes nonlinearly, resulting in a curved characteristic. The reason for the non-equal spacing is that the value of the inductance is changed due to magnetic saturation or the like.
This means that there is an inter-axis interference magnetic flux, such as the generation of an axial magnetic flux. This can be expressed as follows.

Conventionally, as inductance, Ld, L
Only q existed, and the relation between the current and the magnetic flux was considered to be the following equation. Here, one diagonal component was regarded as zero.

[0043]

(Equation 8)

However, there is actually interference between the d-q axes, and the inductance in the 2 rows and 2 columns is as follows.
Furthermore, each inductance component changes non-linearly with a current.

[0045]

(Equation 9)

With such non-linearity, the output torque cannot be controlled in accordance with the torque command, the voltage cannot be controlled, voltage saturation occurs, and the current control system itself becomes uncontrollable and becomes unstable. I do.

In order to perform control even when there is such a nonlinearity of the magnetic flux, the following reference 3 proposes a control system for feeding back the voltage and the torque component.

Reference 3 "Torque feedback control method of salient-pole permanent magnet synchronous motor, Keiichi Kondo, Koichi Matsuoka (RTRI), Yosuke Nakazawa, Hideyuki Shimizu (Toshiba), IEEJ, Poem D, 119th Edition, p. 119, p. .1115, 1999] However, this paper also has the following problems.

The PI calculation is used in the portion for calculating the current command by feeding back the torque. The convergence is not optimally controlled because the PI calculation remains fixed and does not correspond to a change in inductance. Therefore, the response of the actual output torque to a change in the torque command must be set lower.

In Reference 3, since the voltage feedback affects only the d-axis and the torque feedback affects only the q-axis, the torque fluctuation caused by the influence of the d-axis feedback component is corrected by direct calculation. However, the correction is indirectly performed by the torque feedback calculation in the next and subsequent times. Therefore, it is assumed that the response is further delayed.

In Reference 3, high torque response performance is not required because it is developed for railway use. In such a case, such a system can be applied, but an industrial variable speed system is used. In order to use it, it is required to improve the responsiveness of the torque control.

An object of the present invention is to provide a control method for a synchronous motor that solves the above problems.

[0053]

SUMMARY OF THE INVENTION According to the present invention, a synchronous motor using a permanent magnet as a field source is controlled in torque according to a torque command, and a d-axis current command is calculated for the torque command according to a maximum torque condition. And a means for calculating a q-axis current command from the torque command and the d-axis current command, wherein the means for calculating the q-axis current command rotates in synchronization with a field. A magnetic flux estimating unit for estimating a magnetic flux component corresponding to a current of the rotating coordinate system by a function or a table approximation; an estimated torque calculating unit for estimating a torque generated in the current state from the current and the magnetic flux estimation result; A current change calculator for obtaining a q-axis current component for correcting an error component between the estimated torque and the torque command so that the obtained torque matches the torque command, and adding the q-axis current component to a current q-axis current. Characterized by comprising a current adder for outputting a new torque current command Te.

A series of calculations of the magnetic flux estimation, the torque calculation, the current change amount calculation, and the current addition are repeated a plurality of times so that the torque estimation converges on the torque command, and the result of the convergence calculation is output as a current command. It is characterized by the following.

Also provided is a self-inductance calculating section for calculating self-inductances of the d-axis and the q-axis when the current of the rotating coordinate system changes, and using the self-inductance in the q-axis current calculation by the current change amount calculating section. It is characterized by having.

When the output voltage applied to the synchronous motor increases beyond the voltage limit value, a voltage saturation prevention control means for generating a demagnetizing current is provided, and a d-axis current command output according to a pattern such as a maximum torque condition is provided. And a new d-axis current command.

Further, the present invention is characterized in that the limit value of the q-axis current is variably reduced in accordance with the d-axis current so that the current command becomes equal to or less than the absolute value limit value of the d-axis current.

When the demagnetizing current reaches the demagnetizing current limit value of the synchronous motor and cannot be increased any more, or when the demagnetizing current is still higher than the voltage limit value, the q-axis current command is further reduced. It is characterized by having.

[0059]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS (First Embodiment) FIG. 1 shows a control block in the case where a speed control system is configured using IPM. Each part is composed of the following elements.

The speed control amplifier 1 calculates a torque command by PI (proportional / integral) calculation or the like so that the speed detection follows the speed command. Torque-current command converter 2
Converts a torque command and a demagnetizing current command from a voltage saturation prevention control unit described later into d- and q-axis current commands. The current control amplifier 3 calculates a voltage to generate an actual current according to the current command. The power amplifier 4 power-amplifies the output voltage of the current control amplifier 3 by PWM modulation or the like. IP
M5 is a motor for actually driving the load. The position sensor 6 detects the rotational position (phase) of the IPM 5. The phase detector 7 converts a signal output from the position sensor 6 into an actual phase angle θ. The speed detector 8 calculates the speed by differentiating the position detection data (θ). The absolute value converter 9
The terminal voltage of the IPM 5 is calculated from the output voltage of the power amplifier 4 or the voltage command output from the current control amplifier 3. The voltage saturation prevention controller 10 controls the demagnetization current so that the terminal voltage does not reach the output voltage limit of the power amplifier.

The present embodiment proposes a new control method for the torque-current command converter 2 and the voltage saturation prevention controller 10 in FIG. For other parts, a control method similar to the conventional one can be applied. In addition to the speed control application, the present invention can be applied to other applications such as a torque control system and a position control system.

Hereinafter, each part of the torque-current command converter 2 and the voltage saturation prevention controller 10 will be described in detail.

(1A) Defining a Nonlinear Function as Magnetic Flux for Current to Consider Nonlinearity First, a method of treating the magnetic flux nonlinearity corresponding to the magnetic flux table 11 in FIG. 2 will be described.

Here, the above equation (9) is treated as a function approximation or a magnetic flux table. The current (i, i) of the rotating coordinate system (d, q coordinates) rotating in synchronization with the field
d, iq), and a “non-linear flux function” that outputs magnetic flux components (出力 d (id, iq), Λq (id, iq)) on the dq coordinates at that time. Actually, there are various methods of expressing nonlinearity as described below.

A method of approximating nonlinearity by function approximation using higher-order equations, trigonometric functions, and logarithmic functions.

A method of preparing two types of two-dimensional tables of d-axis magnetic flux and q-axis magnetic flux and interpolating them to approximate nonlinearity.

In the present embodiment, since the same function can be realized by applying any nonlinear expression method, these are abstractly referred to as “non-linear flux functions”.

In order to distinguish the magnetic flux of the approximated “non-linear magnetic flux function” from the actual magnetic flux, the table data function is handled by the following functional expression using a capital letter Λ.

[0069]

(Equation 10)

The inductance table 12 shown in FIG.
Is a part for calculating a self-inductance component obtained by differentiating the function of Expression (10) with a current component.

In the inductance matrix, the mutual inductance component, which is an interference term between axes, has a smaller value than the other self-inductance components. Therefore, in the example of FIG. 2, only the self-inductance is used. Even if this function is boldly approximated to an average fixed value as explained in the next section, the function intended by the present invention can be realized.
An example using “inductance calculation” will be described.

(1B) Calculation Formula of Torque Considering Non-Linearity of Magnetic Flux Using the non-linear magnetic flux function with the current of the preceding paragraph as an input,
The torque can be calculated. The output torque calculator 13 in FIG. 2 is a block that performs this torque calculation. First, a torque equation is derived. When the nonlinear magnetic flux function of the equation (10) is used, the voltage-current equation of the equation (1) becomes the equation (11).

[0073]

[Equation 11]

Here, the first term is the voltage drop due to the resistance of the armature winding and the leakage inductance, and the second term is the voltage component due to the permanent magnet and the armature reaction magnetic field. Since the two terms take into account the nonlinearity, the inter-axis interference term is also included in the nonlinear magnetic East function.

In order to calculate the torque based on this equation, the mechanical energy may be obtained by using the second term, and the mechanical energy may be divided by the angular velocity ω. Then, the following equation is obtained.

[0076]

(Equation 12)

As described above, even in a system having a non-linear inductance, the torque can be calculated by directly using the magnetic flux. The output torque calculator 13 executes this torque calculation.

(1C) Calculation method of d-axis current command (maximum torque condition) When the current is known, the torque can be obtained by the above equation (11), but in actual control, , Given is the torque command, from which id,
The problem is how to calculate iq. Therefore, a current command calculation method will be described next.

In this section, the d-axis current (demagnetizing current) component will be described first. As the d-axis current, a function for outputting a d-axis current corresponding to a current locus in the constant torque range in FIG. This function can be calculated off-line in advance if the nonlinearity of the magnetic flux is known.

Considering the accuracy of the maximum torque function, even if a result deviated from the maximum torque condition is output, the torque characteristic near the maximum torque actually becomes gentle as shown in FIG. , It is possible to obtain a torque substantially equal to the ideal point.

Therefore, the accuracy of the d-axis current function is not so required. Efficiency substantially equal to the maximum torque condition can be realized even if a slight error is included. Therefore, the inductance or the magnetic flux may be calculated by approximating it linearly.

When voltage saturation occurs, the current control itself becomes uncontrollable. Therefore, in order to set the order of the voltage suppression control higher than the torque, a command is calculated before the torque current calculation, and the torque current is calculated from the result. It is configured to calculate.

A function calculation of the d-axis current is performed by the demagnetizing current table 17 and the absolute value calculator 18 shown in FIG. In addition,
In the case of a motor in which the permanent magnet itself is demagnetized when an excessive demagnetizing current flows, the motor may be protected by limiting the maximum value of the electromagnetic current with the demagnetizing current limiter 19.

(1D) Conversion method from torque command to current command (application of convergence calculation) Since the q-axis current component needs to directly control the torque with respect to the d-axis current, torque control accuracy is required. . In addition, it is necessary to increase the demagnetizing current beyond the maximum torque condition in order to prevent voltage saturation. Therefore, the method of calculating the torque current command must be able to cope with the change in the d-axis current.

Further, since there is also a non-linearity in which the magnetic flux characteristics change depending on the current, a calculation method using a convergence calculation is applied here.

In the above-mentioned reference 3, the PI calculation is simply used, but in the present embodiment, the current required for outputting the torque by predicting the change in the magnetic flux using the inductance is calculated, and the calculated current is further calculated. A method of repeatedly converging was adopted.

In the present embodiment, since the convergence calculation is performed by using the derivative of the magnetic flux called inductance, it can be said that this is a kind of Newton-Raphson method.

Here, the inductance component used is
It is used only for estimating the magnetic flux when the current changes, and the final torque accuracy is based on the magnetic flux table 11 described above. That is, when the accuracy of the inductance is low, the convergence is inferior, but the torque accuracy at the time of final convergence is not affected. In addition, the interference inductance component between the q axes is ignored for the sake of simplicity of calculation.

First, the previous current commands id, iq are known, and the magnetic flux components {d (id, iq),
It is assumed that q (id, iq) has also been calculated, and that the torque T generated by the previous current command has already been calculated by the following equation (12). In response to this, a new torque command T ′ and a d-axis current command id ′ (demagnetization current command)
Is given, the q-axis current iq ′ is determined.

Assuming that the current has changed from id, iq to id ′, iq ′, the current change Δid, Δiq
The magnetic flux Λd '(id, iq), Λq' (id, i
How q) changes is estimated by applying the following equation that is linearly approximated using inductance.

[0091]

(Equation 13)

[0092]

[Equation 14]

Originally, the interference inductance between the axes also exists. However, when this is taken into consideration, the torque current estimation formula to be obtained later becomes a quadratic formula, and the value of the interference inductance is small and the influence is small. Here, it is approximated to zero.

An equation for obtaining a new q-axis current iq ′ from the magnetic flux equation for the previous currents id and iq and the torque equation of equation (12) is obtained as follows.

[0095]

(Equation 15)

When the above equation is replaced with the difference torque from the previous torque component, it can be expressed by a change term of the current as follows.

[0097]

(Equation 16)

Further, when the equation for obtaining Δiq is modified,

[0099]

[Equation 17]

As described above, since the change amount of the q-axis current can be obtained by the equation (17), the next time iq 'can be obtained by adding the q-axis current command of the previous value.

As described above, since the torque current component is obtained by predicting the change in the magnetic flux by the inductance, the magnetic flux calculation using the nonlinear magnetic flux function and the torque calculation using the same are performed again using the obtained current command. Check that it matches the command value.

If the torque calculation value is different from the command value, the calculation is repeated in the same procedure using the current current again. Then, the current and magnetic flux conditions finally coincide with the torque command. This is the convergence calculation method applied in the present embodiment, and corresponds to the torque current calculation unit 14 in FIG.

Although the convergence method for obtaining the change current from the inductance is fast, the convergence method may be excessively large than the true value due to an error of the inductance or the like, and an excessive amount may be generated. Then, since hunting or convergence cannot be achieved at the time of convergence, the current output is actually applied with a relaxation gain by the low-pass filter 15 of FIG. 2 to achieve stabilization.

When the present invention is applied to a general-purpose inverter, if the number of repetitions of the convergence operation is increased, the operation time is restricted. In some cases, the convergence operation may be performed only once. At this time, it is also effective to add a first-order lag or cushion to the torque command itself by the low-pass filter 16 of FIG. 2 to suppress a sudden change. As a result, since the current change (flux change) does not occur significantly, the current command also changes slowly, and the error decreases even if the number of convergence calculations is small.

If a delay component is inserted into the torque command, the response performance itself is naturally limited. Therefore, when high-speed response is required as in a servomotor, it is necessary to adopt a method of increasing the number of convergence calculations. There is. Inserting the first-order lag into the torque command is suitable for an application that does not require much responsiveness, such as an energy-saving application.

(Second Embodiment) In the first embodiment,
An example in which the inductance value in each current state is used in consideration of the nonlinearity of the inductance has been described. In this case, optimal convergence is obtained. However, when the variation in inductance is small, the inductance may be an appropriate fixed value such as an average value. The inductance is used only for the current prediction calculation for convergence, and is not used for the torque calculation itself. That is, although the inductance affects the convergence, it does not affect the torque accuracy.

Therefore, in this embodiment, the inductance is set to a constant value. FIG. 3 shows a block diagram thereof. 2 is different from FIG. 2 in that the inductance table 12 for calculating the inductance is omitted.

(Third Embodiment) The present embodiment is a case where the voltage saturation prevention control is added.

In the constant output range, it is necessary to keep the voltage constant even when the rotational speed increases, and in the case of IPM, it is necessary to supply a demagnetizing current.

Therefore, in this embodiment, when the output voltage rises above the set voltage, the negative d-axis current (demagnetizing current) is increased to prevent the terminal voltage from rising. Add control functions. Figure 4 shows the block diagram.
Shown in

FIG. 4 shows an example in which a voltage saturation prevention control unit 20 is added to FIG. The demagnetizing current output from the control unit 20 is obtained as an addition to the id current command. The control unit 20 calculates the absolute value component of the terminal voltage from the output voltage V1, and this component is calculated as the voltage saturation set value Vlim
When it exceeds it, the demagnetizing current is increased by PI calculation.

(Fourth Embodiment) In the third embodiment,
Although demagnetizing current feedback is added to prevent voltage saturation, the output current increases by the amount of the demagnetizing current. Since the ordinary power amplifier 4 has a current limitation, even if the demagnetization current increases, it is necessary to perform a limiter process on the current command so as to be within the current limitation.

Actually, if the demagnetizing current itself is reduced, voltage saturation occurs and the current control itself becomes unstable, so that it cannot be applied.

Therefore, in the present embodiment, the current on the q-axis side corresponding to the torque current is reduced. FIG. 5 shows a control block diagram. This figure is different from the block diagram of the third embodiment in that a torque current limiter value calculation unit 21 is provided as an addition of a current limiting function.

The arithmetic unit 21a calculates the absolute value of the d-axis current Id_ref, and thereafter calculates the limit value of the q-axis current by the following equation.

[0116]

(Equation 18)

Thus, if the q-axis current is limited to I'q_lim or less, the absolute value of the current will not exceed I1_lim.

Next, in some cases, the limit of the demagnetizing current of the electromagnetic current which is the d-axis current is set so that the permanent magnet is not demagnetized. In this case, the voltage cannot be reduced using the demagnetizing current. In this case, the terminal voltage is reduced by decreasing the q-axis current and reducing the voltage component due to the q-axis current and the inductance. Suppress.

Therefore, in 21b, when the d-axis current increases beyond the demagnetization current limit value of the voltage saturation prevention demagnetization current control unit 20, the q-axis current limit value output from 21a is further reduced. I have.

In 21c, since the d-axis current limit value obtained by combining the outputs of 21a and 21b is an absolute value, the d-axis current limit value is limited to a negative value.

At 21d, the positive and negative sides of the q-axis current command are actually limited according to the limit value (absolute value) of the q-axis current output from 21c.

[0122]

According to the IPM, the saliency of the iron core causes the inductance to change due to the armature reaction of the current, and the IPM often has nonlinearity. Therefore, in the present invention, the relationship between the current and the magnetic flux is provided as a non-linear function in the controller, and the torque accuracy can be improved by calculating the torque using the magnetic flux function.

Further, a current command operation with good convergence can be realized from the relational expression between torque and current using inductance, and as a result, the response to the torque command can be improved.

Further, in order to prevent voltage saturation, a demagnetizing current feedback control function is added, and operation is possible even in a constant output range.

In addition, since the current command itself is appropriately limited in response to the current limitation of the converter, the output torque is reduced, but the current control can always maintain the control state. No unstable phenomenon due to saturation occurs.

[Brief description of the drawings]

FIG. 1 is a control block diagram of an IPM showing an embodiment of the present invention.

FIG. 2 is a block diagram of a torque-current command converter according to the embodiment.

FIG. 3 is a block diagram of a torque-current command converter according to the embodiment.

FIG. 4 is a block diagram to which current saturation prevention control is added according to the embodiment;

FIG. 5 is a block diagram to which a torque current limiter value calculation according to the embodiment is added.

FIG. 6 is an explanatory diagram of coordinate axes.

FIG. 7 is an equivalent circuit diagram of the IPM.

FIG. 8 is a torque-current vector diagram.

FIG. 9 is a diagram showing a current phase difference angle-torque characteristic.

FIG. 10 is a voltage contour diagram on current coordinates.

FIG. 11 is a contour diagram of torque and voltage on current coordinates.

FIG. 12 is a contour diagram of a D-axis magnetic flux.

FIG. 13 is a contour diagram of a Q-axis magnetic flux.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Speed amplifier 2 ... Torque-current command conversion part 3 ... Current control amplifier 4 ... Power amplifier 5 ... IPM 6 ... Position sensor 7 ... Position detector 8 ... Speed detector 9 ... Absolute value change part 10, 20 ... Voltage saturation Prevention control unit 11 Magnetic flux table 12 Inductance table 13 Output torque calculation unit 14 Torque current calculation unit 15, 16 Low-pass filter 17 Demagnetization current table 18 Absolute value circuit 19 Demagnetization current limiter 21 Torque current Limiter value calculator

Claims (6)

[Claims] 1. A means for controlling a synchronous motor using a permanent magnet as a field source according to a torque command according to a torque command, and calculating a d-axis current command according to a maximum torque condition or the like with respect to the torque command. A synchronous motor control method including means for calculating a q-axis current command from a d-axis current command, wherein the means for calculating the q-axis current command is based on a current of a rotating coordinate system that rotates in synchronization with a field. A magnetic flux estimating unit for estimating the generated magnetic flux component by a function or a table approximation, an estimated torque calculating unit for estimating a torque generated in the current state from the current and the magnetic flux estimation result, and matching the estimated torque with a torque command. A current change amount calculating section for obtaining a q-axis current component for correcting an error component between the estimated torque and the torque command; and adding the q-axis current component to a current q-axis current to obtain a new torque current Control system of the synchronous motor, characterized in that a current adding unit for outputting as. 2. A series of calculations of the magnetic flux estimation, the torque calculation, the current change amount calculation, and the current addition output a convergence calculation result a plurality of times repeatedly as a current command so that the torque estimation converges on the torque command. The control method for a synchronous motor according to claim 1, wherein: 3. A self-inductance calculator for calculating d-axis and q-axis self-inductances at the time of a current change in the rotating coordinate system, and using the self-inductance for q-axis current calculation by the current change amount calculator. 3. The control method for a synchronous motor according to claim 1, wherein 4. A d-axis current command output according to a pattern such as a maximum torque condition is provided with a voltage saturation prevention control means for generating a demagnetizing current when an output voltage applied to the synchronous motor increases beyond a voltage limit value. 4. A new d-axis current command is obtained by adding
The control method of the synchronous motor according to any one of the above. 5. The method according to claim 4, wherein the limit value of the q-axis current is variably reduced according to the d-axis current so that the current command becomes equal to or less than the absolute value limit value of the d-axis current. The control method of the synchronous motor described. 6. When the demagnetizing current reaches a demagnetizing current limit value of the synchronous motor and cannot further increase the demagnetizing current,
5. The system according to claim 4, further comprising: means for further reducing the q-axis current command when the voltage is still higher than the voltage limit value.
Or the control method of the synchronous motor according to 5.