JP2016195523A - Controller for permanent magnet type synchronous electric motor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a controller capable of stably estimating a magnetic pole position and a speed of a PMSM.SOLUTION: A controller comprises: current detectors 11u, 11w which detect a current of a PMSM 80; an extended induced voltage computing element 31 and an angle difference computing element 32 which compute a position estimation error of a rotor from a current, a terminal voltage and a speed equivalent value of the PMSM 80; a speed estimator 33 which computes a speed by a proportional plus integral operation from a computed value of the position estimation error; an integrator 34 which computes a position estimation value by integrating the speed estimation value; and a gain computing element 35 which computes a proportional gain from the current and speed estimation value of the PMSM 80. The gain computing element 35 has an output limiter 35e which limits an upper-limit value of the proportional gain of the speed estimator 33 with a value proportional to a differential angular frequency, and an output limiter 35f which limits a lower-limit value of an integral time constant with a value inversely proportional to the differential angular frequency.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a technique for calculating the magnetic pole position of a permanent magnet type synchronous motor.

  In order to reduce the cost of a control device for a permanent magnet type synchronous motor, so-called sensorless control that operates without using a magnetic pole position detector for detecting the magnetic pole position of a rotor has been put into practical use. In the sensorless control, torque control and speed control are realized by calculating the magnetic pole position and speed of the rotor from information on the terminal voltage and current of the motor, and performing current control based on these.

In the sensorless control, the following are known as conventional techniques for calculating the magnetic pole position and speed.
For example, in Non-Patent Document 1, the expansion induced voltage generated in the direction orthogonal to the magnetic pole direction of the rotor is calculated, and the magnetic pole position and speed are calculated using the estimation error of the magnetic pole position detected from the calculated value of the expansion induced voltage. Is calculated.

In Patent Document 1, as shown in FIG. 5, the magnetic flux observer 91 in the position / speed estimator 90 is based on the terminal voltage, current, and extended magnetic flux of the motor, and the estimated δ-axis magnetic flux Ψ _δest and γ-axis magnetic flux. The estimated value Ψ _γest is calculated, and the position error δ _est detected by the angle error calculator 93 as the angle of the magnetic flux estimated values Ψ _δest , Ψ _γest is used. A technique for estimating ω ₁ and magnetic pole position θ ₁ is disclosed.
According to Patent Document 1, the position estimation error δ _est calculated from the angles of the δ-axis magnetic flux estimated value Ψ _δest and the γ-axis magnetic flux estimated value Ψ _γest includes not only the position estimation error information but also the differential component (speed) The speed estimation error becomes larger as the motor speed is lower and the load is heavy, and the speed / position estimation system tends to become unstable. This unstable phenomenon also occurs in the case of sensorless control using an extended induced voltage described in Non-Patent Document 1.

In order to avoid the above unstable phenomenon, in Patent Document 1, the q-axis inductance setting value L _q used for the extended magnetic flux calculation of the magnetic flux observer 91 is set so that the speed estimation error included in the calculated value of the position estimation error becomes zero. Is determined by the L _q setter 92 of FIG.

On the other hand, Patent Document 2 discloses a technique for improving the stability of the speed / position estimation system under heavy load by a method different from that of Patent Document 1.
In this prior art, as shown in FIG. 6, the shaft error estimator 96 calculates the shaft error estimated value Δθ _c from the terminal voltage and current of the electric motor 80, and the shaft error command value Δθ _c ^* and the shaft error estimated value Δθ _c are calculated. The speed estimator 97 performs proportional integral control to estimate the speed ω ₁ , and the gain calculator 98 sets the speed estimator 97 based on the control response frequency calculated from the current and the speed estimated value ω _1. The proportional gain _Kp and the integral gain _Ki are set. In FIG. 6, reference numeral 100 denotes a vector control calculation unit.

JP 2011-67066 A (paragraphs [0042] to [0068], FIG. 1 etc.) Japanese Patent Laying-Open No. 2011-91976 (paragraphs [0030] to [0032], FIG. 4, FIG. 5, etc.)

Koji Tanaka and Ichiro Miki, “Position Sensorless Control of Embedded Magnet Synchronous Motor Using Extended Inductive Voltage”, IEEJ Transactions D, Vol. 125, no. 9, pp. 833-838 (2005)

In the prior art described in Patent Document 1, the q-axis inductance set value L _q is calculated in advance so that the value of the complicated evaluation function becomes zero according to the conditions of the γ-axis current and the δ-axis current, and L _q It is necessary to hold the table in the setting device 92 as a table or to calculate the q-axis inductance setting value L _q online by the L _q setting device 92 in accordance with the δ-axis current after setting the γ-axis current to zero. In any case, complicated arithmetic processing is required.
Further, the proportional gain K _p and integral gain K _i of the speed estimator 97 set by the method described in Patent Document 2 are different from the optimum values according to the analysis by the inventors, and are low and under heavy load. There is a possibility that the instability phenomenon of the speed / position estimation system cannot be resolved.

  SUMMARY OF THE INVENTION An object of the present invention is to provide a control device for a permanent magnet type synchronous motor that enables sensorless control that is more stable than conventional ones at low speeds or heavy loads.

In order to solve the above-mentioned problem, the invention according to claim 1 includes means for detecting a current of a permanent magnet synchronous motor,
Means for calculating a rotor position estimation error from the electric current, terminal voltage and speed equivalent value of the motor;
Means for calculating the estimated speed value of the rotor from the calculated value of the position estimation error;
Means for integrating the speed estimate and calculating a position estimate of the rotor;
Means for calculating a differential angular frequency that is a differential component of a position estimation error included in the position estimation error calculation value from the current of the motor and the speed estimation value;
The means for calculating the speed estimated value has a proportional adjustment means,
The means for calculating the differential angular frequency is:
And a means for limiting the upper limit value of the proportional gain in the proportional adjustment means by a value proportional to the differential angular frequency.
Thereby, the stability of the speed / position estimation system when the speed estimation value is calculated by proportional control of the position estimation error calculation value can be improved.

According to a second aspect of the present invention, in the control device for the permanent magnet synchronous motor according to the first aspect, the means for calculating the speed estimated value further includes an integral adjusting means, and the means for calculating the differential angular frequency is And a means for limiting the lower limit value of the integration time constant in the integral adjusting means by a value inversely proportional to the differential angular frequency.
As a result, the stability of the speed / position estimation system when the speed estimation value is calculated by the proportional / integral control of the position estimation error calculation value can be improved.

  According to the present invention, sensorless control that is more stable than conventional ones can be realized at low speeds and heavy loads.

It is a block diagram which shows embodiment of this invention. It is a figure which shows the definition of (gamma) -delta axis orthogonal rotation coordinate system and dq axis orthogonal rotation coordinate system. It is a figure which shows the linear approximation model of the speed and position estimation system in FIG. FIG. 2 is a control block diagram of a gain calculator in FIG. 1. It is a block diagram which shows the principal part of the control apparatus described in patent document 1. It is a block diagram which shows the control apparatus described in patent document 2.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
First, in the control calculation in this embodiment, a virtual γ-δ axis orthogonal rotation coordinate system that rotates at an electrical angular velocity ω ₁ is defined in the control device, and a permanent magnet type is defined according to this γ-δ axis orthogonal rotation coordinate system. Controls the current and voltage of a synchronous motor (hereinafter referred to as PMSM).

FIG. 2 is for explaining the definitions of the γ-δ axis orthogonal rotation coordinate system and the dq axis orthogonal rotation coordinate system.
In FIG. 2, a dq axis orthogonal rotation coordinate system is defined with the N pole direction of the rotor magnetic pole as the d axis and the 90 ° advance direction from the d axis as the q axis, and this dq axis orthogonal rotation coordinate system and γ An angle error (position estimation error) with respect to the −δ axis orthogonal rotation coordinate system is θ _err . Ω _r is the angular velocity of the dq axis (rotational angular velocity of the rotor), and ω ₁ is the angular velocity of the γ-δ axis (speed estimation value).
Since sensorless control cannot directly detect the position of the dq axis, the control device performs a control calculation on the γ-δ axis that is the estimated axis of the dq axis.

Here, the angle error θ _err is _expressed by the following equation ( ₁ ) using the γ-axis angle (position estimation value) θ ₁ with respect to the PMSM u-phase winding and the d-axis angle (with reference to the u-phase winding ( defined as the difference between the magnetic pole position) theta _r.

Next, based on FIG. 1, the structure and function of the control apparatus which concern on this embodiment are demonstrated.
First, PMSM speed control, current control, and voltage control will be described.
In FIG. 1, the difference between the speed command value ω _r ^* and the speed estimated value ω ₁ is calculated by the subtractor 16. The speed regulator 17 performs an adjustment calculation so that the deviation becomes zero, and generates a torque command value τ ^* . The current command calculator 18 calculates the γ-axis current command value i _γ ^* and the δ-axis current command value i _δ ^* so as to generate torque according to the torque command value τ ^* .

On the other hand, phase current detection values i _u and i _{w obtained} by the u-phase current detector 11u and the w-phase current detector 11w are input to the coordinate converter 14, and these phase current detection values i _u and i _w are estimated position values. Coordinates are converted to γ and δ axis current detection values i _γ and i _δ using θ ₁ .
The subtractor 19a obtains a deviation between the γ-axis current command value i _γ ^* and the γ-axis current detection value i _γ, and the γ-axis current regulator 20a performs an adjustment calculation so that the deviation becomes zero, thereby performing the γ-axis A voltage command value v _γ ^* is generated. Further, the subtractor 19b obtains a deviation between the δ-axis current command value i _δ ^* and the δ-axis current detection value i _δ, and the δ-axis current regulator 20b performs an adjustment calculation so that the deviation becomes zero. A δ-axis voltage command value v _δ ^* is generated.
The γ-axis voltage command value v _γ ^* and the δ-axis voltage command value v _δ ^* are converted into phase voltage command values v _u ^* , v _v ^* , v _w ^* by coordinate conversion based on the position estimation value θ ₁ in the coordinate converter 15. It is converted and input to the PWM circuit 13.

The AC voltage of the three-phase AC power supply 50 is converted into a DC voltage by the rectifier circuit 60 and supplied to a power converter 70 such as an inverter.
The PWM circuit 13 generates a gate signal for controlling the output voltage of the power converter 70 to the phase voltage command values v _u ^* , v _v ^* , and v _w ^* . The power converter 70 performs on / off control of a semiconductor switching element such as an internal IGBT based on the gate signal, thereby changing the terminal voltage of each phase of the PMSM 80 to the phase voltage command values v _u ^* , v _v ^* , v _w. Control to ^* .
As a result, the rotational speed of the PMSM 80 is controlled according to the speed command value ω _r ^* .

Next, the estimation operation of the rotor speed and magnetic pole position of the PMSM 80 will be described.
First, the expansion induced voltage calculator 31 calculates the expansion induced voltage generated in the direction orthogonal to the magnetic pole direction of the rotor by Equation 2.

In Equation 2, the γ-axis voltage v _γ and the δ-axis voltage v _δ include the γ-axis voltage command value v _γ ^* and the δ-axis voltage command value output from the γ-axis current regulator 20a and the δ-axis current regulator 20b, respectively. v _δ ^* is used.
In the calculation of Equation 2, the phase voltage or line voltage of the PMSM 80 is measured using a voltage detection circuit (not shown), and the γ-axis voltage and δ-axis calculated from these measured values and the position estimated value θ ₁ are calculated. You may carry out using a voltage. Further, the output of the coordinate converter 14 in FIG. 1 is used for the γ-axis current i _γ and the δ-axis current i _δ in Formula 2, but instead of these, the γ-axis current command value i _γ ^* and the δ-axis current command are used. The value i _δ ^* may be used. Furthermore, although the output of the speed estimator 33 described later is used for the speed estimated value ω ₁ in Equation 2, the speed command value ω _r ^* may be used instead.

Next, a method of calculating the speed and the magnetic pole position using the γ and δ axis expansion induced voltages e _exγest and e _exδest will be described.
Angular error calculator 32, gamma, [delta] axis extended electromotive force _{e exγest,} the position estimation error theta _Errest from _{e Exderutaest} be calculated by Equation 3.

The gain calculator 35 calculates a proportional gain K _Pωest and an integration time constant T _Iωest set in the speed estimator 33 from the γ-axis current i _γ , the δ-axis current i _δ, and the estimated speed value ω ₁ . The operation of the gain calculator 35 will be described in detail later.
The speed estimator 33 performs a proportional / integral adjustment calculation on the position estimation error calculation value θ _errest to obtain the speed estimation value ω ₁ . Specifically, the estimated speed value ω ₁ is calculated by Equation 4.

The integrator 34 integrates the speed estimated value ω ₁ to calculate the position estimated value θ ₁ and outputs it to the coordinate converters 14 and 15.
By these calculations, the speed estimation value ω ₁ and the position estimation value θ ₁ are calculated so that the position estimation error θ _err becomes zero, and these values can be converged to true values.

Next, details of the gain calculator 35 will be described.
When the position estimation error calculation value θ _errest is linearly approximated in the vicinity of the operating point in the steady state, the relationship of Equation 5 is established.

In Equation 5, the differential angular frequency ω _dθerr has the relationship of Equation 6.

As shown in Equation 5, the position estimation error calculation value θ _errest includes not only the position estimation error θ _err but also a component proportional to the speed estimation error ω _err which is a differential component of the position estimation error θ _err . Therefore, the proportional gain K _Pωest and the integration time constant T _Iωest of the speed estimator 33 must be designed in consideration of a component proportional to the speed estimation error ω _err .

From Equations 4 and 5, the linear approximation model of the speed / position estimation system is as shown in FIG. The proportional gain _{K Pomegaest} the integration time constant _{T Aiomegaest,} when the differential angular frequency omega _Dishitaerr is positive, the gain of the open-loop transfer function of the speed and position estimation system | in the Bode diagram of the gain | | _{G 0 G 0} What is necessary is just to design so that the inclination when | is 0 [dB] becomes −20 [dB / dec]. Specifically, it is designed to satisfy the relationship of Equation 7.

From Equation 6 described above, the differential angular frequency ω _dθerr is proportional to the speed estimation value ω ₁ and inversely proportional to the δ-axis current i _δ . Since the δ-axis current i _δ is an increasing function of the torque τ, the differential angular frequency ω _dθerr becomes small at low speed and heavy load. Therefore, from Equation 7, a low speed, during heavy load, in proportion to the absolute value of the differential angular frequency omega _Dishitaerr limits the upper limit of the proportional gain K _Pomegaest, when integrated in inverse proportion to the absolute value of the differential angular frequency omega _Dishitaerr By limiting the lower limit value of the constant T _Iωest (or limiting the upper limit value of the reciprocal of the integral time constant T _Iωest in proportion to the absolute value of the differential angular frequency), the speed / position estimation system can be stabilized. .

FIG. 4 is a control block diagram of the gain calculator 35 in FIG.
In FIG. 4, the differential angular frequency calculator 35 a calculates the differential angular frequency ω _{dθerr according} to Equation 6 from the γ-axis current i _γ , the δ-axis current i _δ, and the estimated speed value ω ₁ . The absolute value calculator 35b calculates the absolute value | ω _dθerr | of the differential angular frequency ω _dθerr .
Incidentally, the calculation of the differential angular frequency omega _Dishitaerr by differential angular frequency calculator 35a, gamma-axis current i _gamma, instead of [delta] -axis current i?, Gamma ^* -axis current value i ^_gamma, [delta] a-axis current value i _[delta] ^* It may be used. Further, the speed command value ω _r ^* may be used instead of the speed estimated value ω ₁ .

The upper limit value K _Pωestmax of the proportional gain K _Pωest in the proportional adjustment means of the speed estimator 33 is calculated by the gain _{35 c} in proportion to the absolute value | ω _dθerr | of the differential angular frequency ω _dθerr . Proportional gain _{K 1} of the gain 35c is set to 1 [times] value of less than a half or less to have a stability margin.
By the gain 35d, the inverse 1 / T _Iωestmin of the lower limit value of the integral time constant T _Iωest in the integral adjusting means of the speed estimator 33 is calculated in proportion to the absolute value | ω _dθerr | of the differential angular frequency ω _dθerr . Proportional gain _{K 2} of the gain 35d is set to _{K 1} value smaller than a quarter or less to have a stability margin.

The initial value K _Pωest0 of the proportional gain and the initial value T _Iωest0 of the integration time constant are designed so as to satisfy the condition of i) of Equation 7. Output limiter 35e sets the initial value _{K Pomegaest0} proportional gain is limited by the upper limit value _{K Pomegaestmax,} calculates a proportional gain _{K Pωest.} Output limiter 35f is a reciprocal ₁ / _T Iωest0 initial value of the integral time constant and limited by the upper limit value ₁ / _T Iωestmin, calculates the reciprocal ₁ / _T Iωest the integration time constant.

According to this embodiment, the proportional gain in the speed estimator 33 is limited in accordance with the absolute value of the differential angular frequency ω _dθerr which becomes smaller at the low speed and heavy load of the _{PMSM 80} , and the speed estimator is inversely proportional to the absolute value. By restricting the lower limit value of the integration time constant T _Iωest at 33, it is possible to suppress the speed estimated value ω _{1 and} thus the position estimated value θ ₁ from changing significantly.

Next, another embodiment of the present invention will be described. In this embodiment, the calculation in the differential angular frequency calculator 35a of the gain calculator 35 is simplified from the above-described Expression 6.
That is, the differential angular frequency ω _dθerr increases at the time of heavy load. Therefore, by calculating the differential angular frequency ω _dθerr using the γ-axis current i _γ (= i _γmax ) and the δ-axis current i _δ (= i _δmax ) at the maximum torque, the calculation of the differential angular frequency ω _dθerr is simplified. Turn into. Specifically, in the differential angular frequency calculator 35a, the differential angular frequency ω _dθerr is calculated by Equation 8.

According to this embodiment, since the differential angular frequency ω _dθerr can be calculated using the _γ- axis current maximum value i _γmax and the δ-axis current maximum value i _δmax fixed corresponding to the maximum torque, the changing γ-axis Compared with Equation 6 for calculating the differential angular frequency ω _dθerr according to the current i _γ and the δ-axis current i _δ , the calculation process can be simplified.

  The present invention can be used not only for sensorless control using an extended induced voltage as in the embodiments, but also as a control device for sensorless control of PMSM using extended magnetic flux as described in Patent Document 1 described above. it can.

11u: u-phase current detector 11w: w-phase current detector 13: PWM circuit 14, 15: coordinate converters 16, 19a, 19b: subtractor 17: speed controller 18: current command calculator 20a: γ-axis current controller 20b: δ-axis current regulator 31: extended induced voltage calculator 32: angle difference calculator 33: speed estimator 34: integrator 35: gain calculator 35a: differential angular frequency calculator 35b: absolute value calculator 35c, 35d: Gain 35e, 35f: Output limiter 50: Three-phase AC power supply 60: Rectifier circuit 70: Inverter 80: Permanent magnet synchronous motor (PMSM)

Claims (2)

Means for detecting the current of the permanent magnet type synchronous motor;
Means for calculating a rotor position estimation error from the electric current, terminal voltage and speed equivalent value of the motor;
Means for calculating the estimated speed value of the rotor from the calculated value of the position estimation error;
Means for integrating the speed estimate and calculating a position estimate of the rotor;
Means for calculating a differential angular frequency that is a differential component of a position estimation error included in the position estimation error calculation value from the current of the motor and the speed estimation value;
The means for calculating the speed estimated value has a proportional adjustment means,
The means for calculating the differential angular frequency is:
A control device for a permanent magnet type synchronous motor, characterized by comprising means for limiting an upper limit value of a proportional gain in the proportional adjustment means by a value proportional to the differential angular frequency. In the control device for the permanent magnet type synchronous motor according to claim 1,
The means for calculating the speed estimated value further includes an integral adjusting means,
The means for calculating the differential angular frequency is:
The controller for a permanent magnet type synchronous motor further comprising means for limiting a lower limit value of an integration time constant in the integral adjusting means by a value inversely proportional to the differential angular frequency.

Images