JP2009159751A - Thrust ripple compensator for linear motor and motor controller - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a thrust ripple compensator for linear motors wherein uneven motor feed is reduced even though the thrust ripple of a linear motor is deficient in periodicity or load fluctuates. <P>SOLUTION: The thrust ripple compensator for linear motors includes: a thrust ripple compensation unit 20 that generates a compensated thrust command based on a thrust command and the position of a linear motor; a current command converter 12 that is input with the compensated thrust command and outputs a current command; and a current amplifier 13 that causes the current of the linear motor to follow the current command. The thrust ripple compensation unit 20 is comprised of: a phase calculator 22 that is input with the position of the linear motor and outputs a Fourier basic phase; a thrust ripple calculator 21 that computes a thrust ripple compensation value based on the Fourier basic phase and the thrust command; and an adder 23 that adds the thrust ripple compensation value to the thrust command and outputs a compensated thrust command. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a thrust ripple compensator for a linear motor that reduces a feed ripple caused by the ripple by reducing a thrust ripple caused by the linear motor.

In the servo motor, thrust is obtained by changing the current flowing in the coil in accordance with the relative position between the mover and the stator.
For example, in a servo motor such as a synchronous linear motor, a permanent magnet is disposed on either side of the mover or the stator, and a coil is disposed on the opposite side, so that the relative position between the mover and the stator is Accordingly, in order to perform dq conversion and obtain a constant thrust, a constant q-axis current command is given, and current feedback control is performed to pass a current through the coil.
However, cogging and torque constant fluctuations exist due to factors such as non-uniformity of the core, disturbance of magnetization, and end effects. Therefore, even if a constant q-axis current command is given, the thrust generated by the servo motor is Depends on the relative position of the stator and the stator. Therefore, in order to improve the servo control performance, such compensation of thrust ripple is indispensable.

In a conventional torque ripple correction method for a servo motor, a torque command control unit compensates for a torque change equivalent to torque ripple from a torque command generated by a speed loop compensator (see, for example, Patent Document 1).
Also, a first correction amount, which is a sine wave signal made in accordance with the electrical angle of the servo motor, and a second correction, which is a sine wave signal made in accordance with a specific movement distance of the machine driven by the servo motor. Some have added quantity to the torque command (see, for example, Patent Document 2).
FIG. 4 is a block diagram showing the configuration of the torque ripple correction method for the servo motor of the first prior art. In FIG. 4, 31 is a speed loop compensator, 32 is a torque constant, 33 is a motor and mechanical system, and 36 is a torque command control unit.
The torque command control unit 36 calculates the torque command Tc calculated by the speed loop compensator 31, the rotor phase θm that is a spatial positional relationship between the rotor magnet of the servo motor and the armature winding, and the current value that flows through the armature winding. Based on the torque ripple Tl (θm, I) that depends on I, a corrected torque command Tc ′ is calculated.
Now, assuming that the torque constant 32 is Kt, the torque Tm is
Tm = Kt · Tc + Tl (θm, I)
It becomes.
On the other hand, when the torque ripple correction method of the first prior art is applied, the torque command Tc is corrected by the torque command control unit 36 to become a corrected torque command Tc ′.
Tm = Kt · Tc ′ + Tl (θm, I)
It becomes.
The correction amount of the torque command included in the correction torque command Tc ′ is determined by a correction term C that compensates for the torque ripple Tl (θm, I).
Tc ′ = Tc · (1 + C)
It can be expressed as.
In this case, the torque Tm is
Tm = Kt · Tc · (1 + C) + Tl (θm, I)
= Kt.Tc + Kt.Tc.C + Tl (.theta.m, I)
And
C = −Tl (θm, I) / (Kt · Tc)
Thus, the torque ripple Tl (θm, I) can be corrected.

FIG. 5 is a block diagram showing a configuration of a torque ripple compensator for a servo motor of the second prior art. In FIG. 5, 41 is a motor, 42 is an encoder, 43 is a current control means, 44 is a PWM power converter, 45 is a current detection means, 46 is a motor electrical angle calculation means, 47 is a first frequency f, amplitude D · Phase φ input means, 48 is first sine wave calculation means, 49 is motor position calculation means, 50 is second frequency f / amplitude D / phase φ input means, 51 is second sine wave calculation means, and 52 is First addition means 53 is a second addition means.
The motor electrical angle calculator 46 calculates the motor electrical angle θe from the motor position pos detected by the encoder 42. In the first frequency f / amplitude D / phase φ input means 47, the frequency (integer multiple of the motor electrical angular frequency), amplitude, and phase can be freely selected, and these are input to the first sine wave calculation means 48. To do. The first sine wave calculation means 48 includes a sin table, and based on the motor electrical angle θe and the input frequency f, amplitude D, and phase φ,
D ・ sin (f ・ θe−φ)
However, f is a positive integer.
Make a sine wave.
Since the first sine wave can be made m (m is a natural number), the first torque ripple correction amount Tcnp1 is

It becomes.
The motor position calculation means 49 calculates the motor position Le within the specific movement range L from the motor position pos. In the frequency f / amplitude D / phase φ input means 50, the frequency (multiple of an integer multiple of the frequency with a specific distance of the motor as one cycle), amplitude, and phase can be freely selected. To the sine wave calculation means 51. The values of the frequency f, the amplitude D, and the phase φ are, for example, read torque data when moving within a range of the specific distance or more at a constant speed, and performs a Fourier transform calculation for each value of the frequency f. It is obtained by obtaining the value of the phase φ. The second sine wave calculating means 48 includes a sin table, and based on the motor position Le and the input frequency f, amplitude D, and phase φ,
D · sin {2π · f · (Le / L) −φ}
However, f is a positive integer.
Make a sine wave.
Since the second sine wave can be made k (k is a natural number), the second torque ripple correction amount Tcmp2 is

It becomes.
In this way, the first torque ripple correction amount Tcnp1 and the second torque ripple correction amount Tcmp2 are obtained, and a value obtained by adding them is used as the torque ripple correction amount.
JP-A-7-284286 (page 8, FIG. 1) Japanese Patent Laying-Open No. 2005-80482 (page 6, FIG. 1)

The torque ripple correction method of the servo motor of the first prior art uses a value obtained by multiplying the torque command generated by the speed loop compensator by a coefficient for compensating for the torque change equivalent due to the torque ripple. There is a problem that the torque ripple that exists when 0 is not compensated.
In addition, since the torque ripple compensation device for the servo motor of the second prior art calculates the torque command compensation value based only on the electrical angle and the rotation angle of the servo motor, the torque ripple that changes depending on the current when the load changes can be obtained. There was a problem that it could not be compensated.
In addition, in the above two prior arts, since a rotating motor is assumed and periodic ripples depending on the rotation angle or electrical angle are compensated, the linearity such as a linear motor with little ripple periodicity is detected. There was also a problem that could not be adapted.
The present invention has been made in view of such a problem. Even if the thrust ripple is present when the thrust command is 0, the thrust ripple varies depending on the current when the load changes. However, an object of the present invention is to provide a thrust ripple compensation device and a motor control device that can compensate for thrust ripples with high accuracy and reduce motor feed unevenness even if the periodicity of the ripples is small.

In order to solve the above problems, the present invention is configured as follows.
The invention described in claim 1 includes a thrust ripple compensator that generates a compensation thrust command (F ^* ) based on the thrust command (F ₀ ^* ) and the position (X) of the linear motor, and the compensation thrust command (F ^* ) ^. ) And a current command converter for outputting a current command, and a linear motor thrust ripple compensator comprising a current amplifier that causes the current of the linear motor to follow the current command,
The thrust ripple compensator receives a position (X) of the linear motor and outputs a phase (θ) of a Fourier fundamental wave (hereinafter referred to as “Fourier fundamental phase (θ)”), and the Fourier A thrust ripple calculator that outputs a thrust ripple compensation value (F _h ) based on the basic phase (θ) and the thrust command (F ₀ ^* ), and the thrust ripple compensation value (F _{0 in the} thrust command (F ₀ ^* )). _h ), and an adder that outputs the compensation thrust command (F ^* ).

According to a second aspect of the present invention, in the thrust ripple compensator for a linear motor according to the first aspect, the phase calculator calculates the Fourier fundamental phase (θ) by the following equation. Is.
θ = mod {(X−X _s ), (X _e −X _s )} · 2π
Where X _s is the start position of the movement range to be compensated, X _e is the end position of the movement range to be compensated, and mod {P, Q} is a function for obtaining a remainder obtained by dividing P by Q.

According to a third aspect of the present invention, there is provided the thrust ripple compensation device for a linear motor according to the first aspect, wherein the thrust ripple calculator uses the Fourier basic phase (θ) and the thrust command (F ₀ ^* ). Based on this, the thrust ripple compensation value (F _h ) is calculated by the following equation.

However, N is a fixed natural number. Also, _An is

B _n is

M is a constant natural number, and a _nm and b _nm (n = 1, 2,..., N; m = 0, 1,..., M) are constants.

  According to a fourth aspect of the present invention, in the motor control device, the linear motor thrust ripple compensating device according to any one of the first to third aspects is provided.

According to the invention described in any one of claims 1 to 3, the thrust command is compensated by calculating a thrust ripple compensation value with a small error from the actual thrust ripple based on the thrust command and the position of the linear motor. Even if the thrust ripple that exists when the torque is 0, the thrust ripple that changes depending on the current when the load changes, even if the periodicity of the thrust ripple of the linear motor is small, high accuracy It is possible to provide a thrust ripple compensation device that compensates for thrust ripple and reduces motor feed unevenness.
According to the fourth aspect of the present invention, the thrust ripple compensator is provided that performs compensation by calculating a thrust ripple compensation value with a small error from the actual thrust ripple based on the thrust command and the position of the linear motor. Therefore, even if the thrust ripple exists when the thrust command is 0 or the thrust ripple changes depending on the current when the load changes, the periodicity of the thrust ripple of the linear motor is small. In addition, it is possible to provide a motor control device that can compensate for thrust ripples with high accuracy and reduce motor feed unevenness.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
Although various functions and means are built in the actual motor control device, only the functions and means related to the present invention will be described and described in the figure. Hereinafter, the same reference numerals are assigned to the same names as much as possible, and the duplicate description is omitted.

FIG. 1 is a block diagram showing a configuration of a thrust ripple compensator for a linear motor according to the present invention. In FIG. 1, 1 is a motor control device, 4 is a linear motor, 5 is a load, 6 is a position detector, 12 is a current command converter, 13 is a current amplifier, 20 is a thrust ripple compensator, and 21 is a thrust ripple. A calculator 22 is a phase calculator.
The motor control device 1 includes a current command converter 12, a current amplifier 13, and a thrust ripple compensation unit 20, and the thrust ripple compensation unit 20 inputs a thrust command F ₀ ^* and a linear motor position X, Based on the thrust command F ₀ ^* and the position X of the linear motor, a compensation thrust command F ^* is calculated and output to the current command converter 12.
The current command converter 12 converts the compensation thrust command F ^* into a current command I ^* and outputs it to the current amplifier 13.
Current amplifier 13 flows the motor current _{I m} to the linear motor 4 based on the current command ^{I *.}
The linear motor 4 generates thrust and drives a load.
The position detector 5 is attached to the linear motor 4, detects the position X of the linear motor 4, and outputs it to the thrust ripple compensation unit 20.

The thrust ripple compensation unit 20 includes a thrust ripple calculator 21, a phase calculator 22, and an adder 23, and the phase calculator 22 inputs the position X of the linear motor 4 and puts it in the position X of the linear motor 4. Based on this, the Fourier basic phase θ is calculated and output to the thrust ripple calculator 21.
The thrust ripple calculator 21 receives the thrust command F ₀ ^* and the Fourier fundamental phase θ, calculates a thrust ripple compensation value F _h based on the thrust command F ₀ ^* and the Fourier fundamental phase θ, and outputs the thrust ripple compensation value F _h to the adder 23. To do.
The adder 23 adds the thrust ripple compensation value F _h on the thrust instruction F ₀ ^* to calculate a compensation thrust command F ^*, and outputs the current command converter 12.

Hereinafter, the compensation principle of thrust ripple will be described.
First, the relationship between the thrust ripple _Fr , the thrust command F ^*, and the position X of the linear motor is analyzed.
The thrust F of the linear motor includes an electromagnetic force F _e generated by a current flowing through the coil of the linear motor, and an attractive force (generally referred to as cogging) F _c generated between the permanent magnet and the magnetic material. Ideally linear motors, electromagnetic force F _e is proportional constant constant motor current I _m (i.e., force constant is a constant), it is desirable that the cogging F _c is 0, the actual linear motor Then, the force constant varies depending on the motor current _Im and the position X of the linear motor, and the cogging F _c varies depending on the position X of the linear motor.

The present invention differs from the prior art in that a phase calculator 22 that inputs a linear motor position X and outputs a Fourier fundamental phase θ, and a thrust ripple compensation value F based on the Fourier fundamental phase θ and a thrust command F ₀ ^*. includes a thrust ripple calculator 21 which outputs the _h, and thrust command F ₀ ^* thrust ripple compensation section 20 constituted by an adder 23 which adds the thrust ripple compensation value F _h outputs a compensation thrust command F ^* in, Compensation is performed by considering both the thrust ripple that exists when the thrust command is 0 and the thrust ripple that changes depending on the current when the load changes.

FIG. 2 is an approximate equivalent block diagram of FIG. In FIG. 2, 14 is a thrust constant, 15 is a cogging function, 16 is an adder, 17 is a subtractor, and 18 is a model of a machine moving part including a mover of a linear motor.
Generally, since the characteristics of the current amplifier 13 is good, since the motor current I _m can be regarded as equal to the current command I ^*, modeling of the current amplifier 13 are omitted.
The thrust constant 14 is represented by a function K (I _m , X). The cogging function 15 is represented by a function c (X). The model 18 of the machine movable part including the linear motor movable element is represented by a transfer function from the acceleration force F _m to the position X of the linear motor. The sum of the electromagnetic force F _e and the cogging F _c is the thrust F generated by the motor.
Also, minus the load force F _L from the thrust F is the acceleration force F _m for accelerating the machine actuators.

According to FIG. 2, the thrust F can be expressed as shown in Equation (1).
F = F _e + F _c = I _m · K (I _m , X) + c (X) (1)
Moreover, Formula (2) is materialized.
I _m = I ^* = F ^* / K ₀ (2)
Equation (2) is substituted into Equation (1), and thrust F can be expressed as a function of compensation thrust command F ^* and position X of linear motor 4 as shown in Equation (3).
F = (F ^* / K ₀ ) · K (F ^* / K ₀ , X) + c (X) = f (F ^* , X) (3)
From equation (3), the thrust F is not follow the compensation thrust command F ^*, its ripple varies with position X of the compensation thrust command F ^* and the linear motor 4.
Therefore, in order to generate a constant thrust F in the linear motor, it is necessary to include a ripple compensation value in the compensation thrust command F ^* .

Next, a method for calculating the Fourier basic phase θ and the thrust ripple compensation value F _h will be described.
In the case where the thrust F is a periodic function of the position X of the linear motor 4, the start position of the period is X _s and the end position is X _e .
In the case where the thrust F is not a periodic function of the position X of the linear motor 4, the start position of the movement range is X _s and its end position is X _e to compensate.
Therefore, the Fourier basic phase θ is given as shown in Equation (4).
θ = mod {(X−X _s ), (X _e −X _s )} · 2π (4)
Here, mod {P, Q} represents a function for obtaining a remainder obtained by dividing P by Q.
Then, it denotes the thrust ripple compensation value F _h in the Fourier series of Equation (5).

However, N is a fixed natural number. The Fourier coefficients A _n (n = 1, 2,..., N) and B _n (n = 1, 2,..., N) are the original thrust commands F _{0 as} shown in the equation (6). ^It shall be expressed by the polynomial of ^* .

Here, M is a constant natural number, and a _nm and b _nm (n = 1, 2,..., N; m = 0, 1,..., M) are constants.
Therefore, the compensation thrust command F ^* is as shown in Expression (6 ′).

The compensation parameters a _nm and b _nm are constants, and are determined when the system for driving the linear motor is determined. Therefore, it is not necessary to determine each time the linear motor is operated.
Next, a method for determining the compensation parameters a _nm and b _nm will be described.
The position / speed control system is configured by feeding back the position X of the linear motor 4.
FIG. 3 shows an example in which the position / velocity control system used to determine the compensation parameter is configured. In FIG. 3, 11 is a position / speed controller, and 30 is a compensation parameter determination unit.
The position / speed controller 11 receives the position command X ^* and the position X of the linear motor 4, calculates a thrust command F ₀ ^* based on the position command X ^* and the position X of the linear motor, and the compensation parameter determination unit 30. Output to.
The compensation parameter determination unit 30 determines the compensation parameters a _nm and b _nm based on the thrust command F ₀ ^* and the position X of the linear motor 4.

First, the position command X ^* is set so that the section from the start position X _s to the end position X _e enters the constant speed section so that the motor speed is constant between the start position X _s and the end position X _e. The control gain of the position / velocity controller 11 is sufficiently increased, or the position / velocity controller is repeatedly included with a compensator.
Next, while moving the linear motor, the linear motor records the start position X _s and the end position X thrust command F and the corresponding position X of the linear motor when moving between _{e 0} ^*.
Further, _An and _Bn are calculated by Fourier analysis based on the data. Moreover, to calculate the _{A n} and _{B n} by changing the number of times the load force _{F L.}
Finally, a _nm and b _nm are calculated using the least square method so that the approximation error between _An and _Bn is minimized.
In this way, the compensation parameters a _nm and b _nm are determined.

In this way, when operating a linear motor, the value close to the thrust command used at the time of ripple measurement is used as the new thrust command, so that the linear motor moves almost the same as at the time of ripple measurement. Can do.
Further, since the motor speed at the time of the ripple measurement is controlled to be constant, since the acceleration force F _m is zero, the thrust F becomes equal to the load force F _L, i.e., thrust ripple had become 0.
Accordingly, the thrust ripple generated by the compensated linear motor as shown in FIG. 1 is also reduced, so that the motor feed unevenness can be reduced.

  The linear motor thrust ripple compensation device according to the present invention is based on the thrust ripple existing when the thrust command is 0 and the current when the load changes, as compared with the torque ripple compensation device of the conventional servo motor. Compensation takes into account both the changing thrust ripple and reduces the motor feed unevenness even if the torque ripple is large or the load changes when the torque command is 0 There is an effect of making it.

The block diagram which shows the structure of the thrust ripple compensation apparatus of the linear motor of this invention Approximate equivalent block diagram of FIG. An example of a position / velocity control system used to determine compensation parameters Block diagram showing the configuration of the torque ripple correction method of the servo motor of the first prior art Block diagram showing the configuration of a torque ripple compensator for a servo motor of the second prior art

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Motor controller 4 Linear motor 5 Load 6 Position detector 12 Current command converter 13 Current amplifier 14 Thrust constant 15 Cogging function 16 Adder 17 Subtractor 18 Machine moving part model 20 Thrust ripple compensation part 21 Thrust ripple calculator 22 Phase calculator 30 Compensation parameter determination unit 31 Speed loop compensator 32 Torque constant 33 and mechanical system 36 Torque command control unit 41 Motor 42 Encoder 43 Current control unit 44 PWM power converter 45 Current detection unit 46 Motor electrical angle calculation unit 47 1 frequency f / amplitude D / phase φ input means 48 first sine wave calculation means 49 motor position calculation means 50 second frequency f / amplitude D / phase φ input means 51 second sine wave calculation means 52 first Adder 53 Second adder

Claims (4)

A thrust ripple compensator that generates a compensation thrust command (F ^* ) based on the thrust command (F ₀ ^* ) and the position (X) of the linear motor, and inputs the compensation thrust command (F ^* ) and outputs a current command In a thrust ripple compensator for a linear motor comprising a current command converter and a current amplifier that causes the current of the linear motor to follow the current command,
The thrust ripple compensator receives a position (X) of the linear motor and outputs a phase (θ) of a Fourier fundamental wave (hereinafter referred to as “Fourier fundamental phase (θ)”), and the Fourier A thrust ripple calculator that outputs a thrust ripple compensation value (F _h ) based on the basic phase (θ) and the thrust command (F ₀ ^* ), and the thrust ripple compensation value (F _{0 in the} thrust command (F ₀ ^* )). _h ) and an adder that outputs the compensation thrust command (F ^* ) by adding the same to the thrust ripple compensation device for a linear motor. The thrust ripple compensator for a linear motor according to claim 1, wherein the phase calculator calculates the Fourier fundamental phase (θ) by the following equation.
θ = mod {(X−X _s ), (X _e −X _s )} · 2π
Where X _s is the start position of the movement range to be compensated, X _e is the end position of the movement range to be compensated, and mod {P, Q} is a function for obtaining a remainder obtained by dividing P by Q. The thrust ripple calculator calculates the thrust ripple compensation value (F _h ) by the following equation based on the Fourier basic phase (θ) and the thrust command (F ₀ ^* ). The thrust ripple compensator for linear motors described in 1.

However, N is a fixed natural number. Also, _An is

B _n is

M is a constant natural number, and a _nm and b _nm (n = 1, 2,..., N; m = 0, 1,..., M) are constants.   A motor control device comprising the linear motor thrust ripple compensation device according to claim 1.