阅读说明：本技术 电动机控制装置 (Motor control device ) 是由 家造坊勋 折井将彦 森辰也 泽田诚晋 久保建太 于 2019-05-30 设计创作，主要内容包括：本发明提供一种电动机控制装置,其能够减小转矩脉动直到高频率而不增大不必要的噪声和振动。本发明的电动机控制装置根据施加电压及电流的检测值来推定交链磁通,根据电流的检测值及交链磁通的推定值来推定输出转矩,从输出转矩的推定值提取出脉动分量,根据脉动分量的提取值来计算电压指令校正值,将电压指令校正值叠加至电压指令的基本值,计算出叠加后的电压指令,根据叠加后的电压指令,对绕组施加电压。(The invention provides a motor control device capable of reducing torque ripple up to a high frequency without increasing unnecessary noise and vibration. A motor control device estimates a linkage magnetic flux from detected values of an applied voltage and a current, estimates an output torque from the detected value of the current and the estimated value of the linkage magnetic flux, extracts a ripple component from the estimated value of the output torque, calculates a voltage command correction value from the extracted value of the ripple component, superimposes the voltage command correction value on a basic value of a voltage command, calculates a superimposed voltage command, and applies a voltage to a winding based on the superimposed voltage command.)

1. A motor control device is characterized by comprising:

a basic voltage command calculation unit that calculates a basic value of the voltage command;

a current detection unit that detects a current flowing through a winding of the motor;

a magnetic flux estimation unit that estimates a linkage magnetic flux that links to the winding, based on a voltage applied to the winding and a detected value of the current;

a torque estimation unit that estimates an output torque of the motor based on a detected value of the current and an estimated value of the linkage magnetic flux;

a pulsation extraction unit that extracts a pulsation component from the estimated value of the output torque;

a ripple control unit that calculates a voltage command correction value based on the extracted value of the ripple component;

a superimposing unit that superimposes the voltage command correction value on a basic value of the voltage command and calculates a superimposed voltage command; and

and a voltage applying unit that applies a voltage to the winding in accordance with the superimposed voltage command.

2. The motor control device according to claim 1,

the pulsation extraction unit extracts a pulsation frequency component of a preset rotation order from the estimated value of the output torque.

3. The motor control device according to claim 1 or 2,

the pulsation extraction unit performs fourier transform on the estimated value of the output torque at a pulsation frequency of a preset rotation order based on the rotation angle of the motor, calculates fourier coefficients of a cosine wave and a sine wave, and calculates the fourier coefficients of the cosine wave and the sine wave as the extracted value of the pulsation component.

4. The motor control device of claim 3,

the ripple control unit calculates a cosine wave multiplier by multiplying a cosine wave of the ripple frequency of the rotation order by a fourier coefficient of the cosine wave, calculates a sine wave multiplier by multiplying a sine wave of the ripple frequency of the rotation order by a fourier coefficient of the sine wave, and calculates a value obtained by multiplying a total value of the cosine wave multiplier and the sine wave multiplier by a control gain as the voltage command correction value.

5. The motor control device of claim 3,

the ripple control unit calculates a control value of a cosine wave by performing a control calculation based on a fourier coefficient of the cosine wave, calculates a cosine wave multiplier by multiplying the cosine wave of the ripple frequency of the rotation order by the control value of the cosine wave, calculates a control value of a sine wave by performing a control calculation based on a fourier coefficient of the sine wave, calculates a sine wave multiplier by multiplying the sine wave of the ripple frequency of the rotation order by the control value of the sine wave, and calculates a value obtained by summing the cosine wave multiplier and the sine wave multiplier as the voltage command correction value.

6. The motor control device according to claim 4 or 5,

the ripple control unit calculates a phase delay caused by the calculation delay based on the rotation angle of the motor, and calculates a cosine wave of the ripple frequency of the rotation order and a sine wave of the ripple frequency of the rotation order after the phase delay is compensated.

7. The motor control device according to claim 1 or 2,

the ripple extracting unit calculates, as the extracted value of the ripple component, a value obtained by performing a band-pass filtering process of passing a component of a ripple frequency of a preset rotation order and attenuating components other than the ripple frequency of the preset rotation order on the estimated value of the output torque,

the ripple control unit multiplies the extracted value of the ripple component by a control gain to calculate the voltage command correction value.

8. The motor control device according to any one of claims 1 to 7,

the magnetic flux estimating unit calculates an estimated value of the interlinkage magnetic flux based on an integrated value of the applied voltage and an integrated value of the detected value of the current.

9. The motor control device according to any one of claims 1 to 8,

the torque estimation unit calculates an estimated value of the output torque based on a multiplication value of the estimated value of the interlinkage magnetic flux and the detected value of the current.

10. The motor control device according to any one of claims 1 to 8,

the torque estimation unit calculates an estimated value of the output torque based on a multiplication value of a differential value obtained by differentiating the estimated value of the interlinkage magnetic flux by a rotation angle and a detection value of the current.

11. The motor control device according to any one of claims 1 to 10,

the basic voltage command calculation unit calculates a d-axis voltage command basic value and a q-axis voltage command basic value as basic values of the voltage command,

the superimposing unit superimposes a voltage command correction value on the q-axis voltage command basic value, calculates a superimposed q-axis voltage command, and directly sets the d-axis voltage command basic value as a superimposed d-axis voltage command.

Technical Field

The present application relates to a motor control device.

Background

The motor outputs torque based on the flux interlinking with the motor winding and the current flowing through the winding. However, since the interlinkage magnetic flux of most motors pulsates in synchronization with the rotation angle of the motor, the output torque pulsates in synchronization with the rotation angle. Various devices have been proposed to suppress such output torque ripple.

As an example of the conventional device, as in patent document 1, after a voltage drop in a winding resistance is subtracted from an applied voltage, a linkage flux is integrated with time to estimate a linkage flux, an inner product of the estimated linkage flux and a winding current is calculated, a ripple of a torque is estimated, a gain is multiplied to calculate a current correction value, and a current command is corrected.

As another example of the conventional device, as disclosed in patent document 2, an induced voltage is estimated by inverting a voltage equation of a motor, a torque is estimated from the induced voltage, the estimated torque is fourier-transformed by a frequency of a torque ripple, a torque ripple component is extracted, a torque correction value is calculated, and the torque correction value is fed back to a torque command.

Documents of the prior art

Patent document

Patent document 1: japanese patent laid-open No. 2009-268267

Patent document 2: international publication No. 2017/167667

Disclosure of Invention

Technical problem to be solved by the invention

In the device of patent document 1, although the output torque can be estimated up to a high frequency, there is a problem that a response from the current command to the current actually flowing through the motor is delayed, the applied voltage of the motor cannot be corrected up to a sufficiently high frequency, and the high-frequency torque ripple cannot be reduced because the correction value is fed back to the current command. Further, since the estimated torque ripple includes unnecessary noise and vibration, and is increased and fed back as the current correction value, there is also a problem that the noise and vibration included in the current and torque of the motor are increased.

In the device as disclosed in patent document 2, since the induced voltage including the current differential is estimated by inverting a voltage equation indicating the response from the induced voltage to the current of the motor, high frequency components are superimposed and reduced by using a low-pass filter. Therefore, there are problems as follows: the induced voltage and the output torque cannot be estimated up to a sufficiently high frequency, and the torque ripple at the high frequency cannot be reduced.

Further, since the estimated torque ripple component is fed back to the torque command, as in patent document 1, the applied voltage of the motor cannot be corrected to a sufficiently high frequency due to a response delay from the torque command to the current actually flowing through the motor, and the torque ripple at a high frequency cannot be reduced. Further, although the voltage command is corrected to reduce the ripple of the induced voltage and reduce the ripple of the current, torque ripple cannot be sufficiently reduced against the problem of the ripple of the interlinkage magnetic flux because torque is generated by the interlinkage magnetic flux and the current. In addition, in the case of the method of estimating the induced voltage for estimating the linkage flux and the output torque, as described above, an estimator such as an observer may be used in addition to the low-pass filter, but in this case as well, the upper limit of the torque frequency that can be estimated is given by the gain existing in the observer and the response frequency specified in the filter, and a high frequency cannot be estimated.

The present application has been made to solve the above-mentioned problems, and an object thereof is to reduce torque ripple up to a high frequency without increasing unnecessary noise and vibration.

Technical scheme for solving technical problem

The motor control device according to the present application includes:

a basic voltage command calculation unit that calculates a basic value of the voltage command;

a current detection unit that detects a current flowing through a winding of the motor;

a magnetic flux estimation unit that estimates a linkage magnetic flux that links to the winding, based on a voltage applied to the winding and a detected value of the current;

a torque estimation unit that estimates an output torque of the motor based on a detected value of the current and an estimated value of the linkage magnetic flux;

a pulsation extraction unit that extracts a pulsation component from the estimated value of the output torque;

a ripple control unit that calculates a voltage command correction value based on the extracted value of the ripple component;

a superimposing unit that superimposes the voltage command correction value on a basic value of the voltage command and calculates a superimposed voltage command; and

and a voltage applying unit that applies a voltage to the winding in accordance with the superimposed voltage command.

Effects of the invention

According to the motor control device of the present application, since the interlinkage magnetic flux is estimated based on the detected value of the current and the applied voltage, and the output torque is estimated based on the detected value of the current and the estimated value of the interlinkage magnetic flux, it is possible to estimate the interlinkage magnetic flux and the output torque up to a high frequency while suppressing unnecessary amplification of noise and vibration. Further, since the pulsation component is extracted from the estimated value of the output torque and the voltage command correction value is calculated, the influence of unnecessary noise and vibration can be reduced. Further, since the voltage command can be directly corrected by the voltage command value correction value, the torque ripple can be reduced to a high frequency without being affected by the response delay of the current feedback control system.

Drawings

Fig. 1 is a schematic configuration diagram of a motor and a motor control device according to embodiment 1.

Fig. 2 is a schematic block diagram of the motor control device according to embodiment 1.

Fig. 3 is a hardware configuration diagram of the motor control device according to embodiment 1.

Fig. 4 is a block diagram of a ripple extracting unit and a ripple control unit according to embodiment 1.

Fig. 5 is a time chart for explaining the effect of reducing the torque ripple component according to embodiment 1.

Fig. 6 is a timing chart for explaining noise immunity according to embodiment 1.

Fig. 7 is a time chart for explaining the effect of reducing the torque ripple component according to embodiment 3.

Fig. 8 is a time chart for explaining the effect of reducing the torque ripple component according to embodiment 4.

Detailed Description

1. Embodiment mode 1

A motor control device 30 (hereinafter, simply referred to as a control device 30) according to embodiment 1 will be described with reference to the drawings. Fig. 1 is a schematic configuration diagram of a motor 1, an inverter 20, and a control device 30 according to embodiment 1.

1-1. Electric motor

The motor 1 is a permanent magnet synchronous motor having a stator provided with three-phase windings Cu, Cv, and Cw of U-phase, V-phase, and W-phase, and a rotor provided with permanent magnets. The stator is provided with three phase windings Cu, Cv and Cw. The three-phase windings Cu, Cv and Cw are provided as star connections. In addition, the three-phase winding may be a delta connection.

The motor 1 has a rotation angle sensor 2 that outputs an electric signal corresponding to the rotation angle of the rotor. The rotation angle sensor 2 is a hall element, an encoder, a resolver, or the like. The output signal of the rotation angle sensor 2 is input to the control device 30.

1-2. Inverter with a voltage regulator

The inverter 20 has a plurality of switching elements and performs dc/ac conversion between the dc power supply 25 and the three-phase winding. The inverter 20 is provided with 3 sets of series circuits (arms) obtained by connecting in series a positive side switching element 23H (upper arm) connected to the positive side of the dc power supply 25 and a negative side switching element 23L (lower arm) connected to the negative side of the dc power supply 25, opposite to the three-phase windings. The inverter 20 has 3 positive-side switching elements 23H and 3 negative-side switching elements 23L, and 6 switching elements in total. Also, a connection point connecting the positive-side switching element 23H and the negative-side switching element 23L in series is connected to the winding of the corresponding phase.

As the switching element, an IGBT (Insulated Gate Bipolar Transistor) having a diode 22 connected in antiparallel, a MOSFET (Metal Oxide Semiconductor Field Effect Transistor) having a diode function connected in antiparallel, or the like is used. The gate terminal of each switching element is connected to control device 30. Each switching element is turned on or off by a control signal output from control device 30.

The smoothing capacitor 24 is connected between the positive side and the negative side of the inverter 20. The voltage sensor 3 outputs an electric signal corresponding to the dc voltage of the dc power supply 25. The output signal of voltage sensor 3 is input to control device 30.

The current sensor 4 outputs an electric signal corresponding to the current flowing through each phase winding. The current sensor 4 is provided on a wire connecting the series circuit of the switching elements and each phase of the winding. The output signal of current sensor 4 is input to control device 30.

As dc power supply 25, a chargeable and dischargeable power storage device (for example, a lithium ion battery, a nickel hydride battery, and an electric double layer capacitor) is used. In addition, the direct-current power supply 25 may be provided with a DC-DC converter, which is a direct-current power converter for stepping up or stepping down a direct-current voltage.

1-3. Control device

Control device 30 controls inverter 20 to control motor 1. As shown in fig. 2, the control device 30 includes a current detection unit 31, a rotation detection unit 32, a voltage detection unit 33, a basic voltage command calculation unit 34, a magnetic flux estimation unit 35, a torque estimation unit 36, a ripple extraction unit 37, a ripple control unit 38, a superimposing unit 39, a voltage application unit 40, and the like, which will be described later. Each function of the control device 30 is realized by a processing circuit provided in the control device 30. Specifically, as shown in fig. 3, the control device 30 includes, as Processing circuits, an arithmetic Processing device 90 (computer) such as a CPU (Central Processing Unit), a storage device 91 that exchanges data with the arithmetic Processing device 90, an input circuit 92 that inputs an external signal to the arithmetic Processing device 90, an output circuit 93 that outputs a signal from the arithmetic Processing device 90 to the outside, and the like.

The arithmetic processing device 90 may include an ASIC (Application Specific Integrated Circuit), an IC (Integrated Circuit), a DSP (Digital Signal Processor), an FPGA (Field Programmable Gate Array), various logic circuits, various Signal processing circuits, and the like. A plurality of arithmetic processing units 90 of the same type or different types may be provided to execute the respective processes in a shared manner. The storage device 91 includes a RAM (Random Access Memory) configured to be able to Read and write data from and to the arithmetic processing device 90, a ROM (Read Only Memory) configured to be able to Read data from the arithmetic processing device 90, and the like. The input circuit 92 is connected to various sensors and switches such as the rotation angle sensor 2, the voltage sensor 3, and the current sensor 4, and includes an a/D converter and the like that inputs output signals of these sensors and switches to the arithmetic processing device 90. The output circuit 93 is connected to an electrical load such as a gate drive circuit that drives the switching elements to be turned on and off, and includes a drive circuit and the like that outputs a control signal from the arithmetic processing device 90 to these electrical loads.

The functions of the control units 31 to 40 and the like in fig. 2 included in the control device 30 are realized by the arithmetic processing device 90 executing software (program) stored in the storage device 91 such as a ROM and cooperating with other hardware of the control device 30 such as the storage device 91, the input circuit 92, and the output circuit 93. The setting data such as the control gain Kv, the rotation order n, and the dead time Δ Td used by the respective control units 31 to 40 are stored in the storage device 91 such as a ROM as a part of software (program). The respective functions of the control device 30 will be described in detail below.

1-3-1. Each detection part

The rotation detecting unit 32 detects a rotation angle θ (magnetic pole position) of the rotor. In the present embodiment, the rotation detecting unit 32 detects the rotation angle θ (magnetic pole position) of the rotor at the electrical angle based on the output signal of the rotation angle sensor 2.

The voltage detection unit 33 detects the power supply voltage VDC of the dc power supply 25. In the present embodiment, the voltage detection unit 33 detects the power supply voltage VDC based on the output signal of the voltage sensor 3.

The current detection unit 31 detects winding currents Iu, Iv, and Iw flowing through the three-phase windings. In the present embodiment, the current detection unit 31 detects the currents Iu, Iv, and Iw flowing from the inverter 20 to the phase windings Cu, Cv, and Cw based on the output signal of the current sensor 4. Here, a vector including the U-phase current detection value Iu, the V-phase current detection value Iv, and the W-phase current detection value Iw is set as the three-phase current detection value Iuvw.

The current detection unit 31 performs three-phase to two-phase conversion on the three-phase current detection value Iuvw, and calculates an α -axis current detection value I α and a β -axis current detection value I β. Here, a vector formed by the α -axis current detection value I α and the β -axis current detection value I β is defined as the α β -axis current detection value I α β. The current detection unit 31 converts the rotational coordinate of the α β -axis current detection value I α β based on the magnetic pole position θ, and calculates a d-axis current detection value Id and a q-axis current detection value Iq. Here, a vector formed by the d-axis current detection value Id and the q-axis current detection value Iq is referred to as a dq-axis current detection value Idq.

The α axis is set in the winding direction of the U phase, and the β axis is set in the direction advanced by 90(π/2) in electrical angle from the α axis. The d-axis is determined in the direction (magnetic pole position) of the N-pole of the permanent magnet provided to the rotor, and the q-axis is determined in the direction advanced 90 (pi/2) in electrical angle from the d-axis.

1-3-2. Basic voltage command calculation unit

The basic voltage command calculation unit 34 calculates a basic value of the voltage command. In the present embodiment, the basic voltage command calculation unit 34 calculates the d-axis voltage command basic value Vd0 and the q-axis voltage command basic value Vq0 as the basic values of the voltage command. Here, a vector constituted by the d-axis voltage command basic value Vd0 and the q-axis voltage command basic value Vq0 is defined as a dq-axis voltage command basic value Vdq 0.

The basic voltage command calculation unit 34 includes a current command calculation unit 34a and a current control unit 34 b. The current command calculation unit 34a calculates a d-axis current command Id _ ref and a q-axis current command Iq _ ref. Here, a vector composed of the d-axis current reference Id _ ref and the q-axis current reference Iq _ ref is referred to as a dq-axis current reference Idq _ ref. In order to calculate the dq-axis current command Idq _ ref, a known current vector control method such as maximum torque current control, flux weakening control, and Id 0 control is used.

The current control unit 34b calculates a current deviation between the dq-axis current command Idq _ ref and the dq-axis current detection value Idq, and performs control calculation such as PID control based on the current deviation to calculate a dq-axis voltage command basic value Vdq 0.

1-3-3. Magnetic flux estimating unit

The magnetic flux estimating unit 35 estimates a linkage magnetic flux linked with the winding based on the applied voltage applied to the winding and the detected current value. In the present embodiment, the magnetic flux estimation unit 35 calculates an estimated value Φ α β of the α β axis linkage magnetic flux representing the linkage magnetic flux in the α β axis coordinate system.

The calculation principle is explained. The voltage equation expressed by the α β axis coordinate system is shown below. Here, the α β axis applied voltage V α β is a vector composed of the α axis applied voltage V α and the β axis applied voltage V β. The α β -axis linkage flux Φ α β is a vector composed of the α -axis linkage flux Φ α and the β -axis linkage flux Φ β. R is the resistance value of the winding.

[ mathematical formula 1]

As shown in the formula (1), the α β axis applied voltage V α β is equal to the total value of the voltage drop R · I α β obtained by the product of the resistance value R of the winding and the α β axis current I α β and the electromotive force obtained by the time derivative of the α β axis interlinkage magnetic flux Φ α β.

Here, the α β -axis interlinkage magnetic flux Φ α β is formed by a sum of a magnetic flux component L · I α β of an inductance obtained by multiplying the inductance L of the armature and the α β -axis current I α β and an α β -axis magnet interlinkage magnetic flux Φ α β which is an α β -axis interlinkage magnetic flux obtained by the magnet of the rotor, as shown in the following mathematical expression.

[ mathematical formula 2]

Φαβ＝L·Iαβ+Φmαβ …(2)

When the equation (2) is substituted into the interlinkage magnetic flux Φ α β of the equation (1), it is understood that the equation (1) also includes a differential term of the current I α β. The electromotive force obtained by time-dividing the magnet interlinkage magnetic flux Φ α β is a so-called induced voltage. That is, the voltage equation can be said to represent the response from the applied voltage and the induced voltage to the current flowing through the winding.

On the other hand, both sides of the formula (1) are integrated and arranged to obtain the following formula. That is, the α β -axis interlinkage magnetic flux Φ α β can be calculated using the integrated value of the α β -axis applied voltage V α β and the integrated value of the α β -axis current I α β.

[ mathematical formula 3]

Φαβ＝∫(Vαβ-R·Iαβ)dt＝∫Vαβdt-R∫Iαβdt …(3)

In the present embodiment, the magnetic flux estimation unit 35 calculates an estimated value Φ α β of the α β -axis interlinkage magnetic flux from the integrated value of the α β -axis applied voltage V α β and the integrated value of the α β -axis current detection value I α β. Specifically, the magnetic flux estimation unit 35 performs calculation using equation (3). The magnetic flux estimating unit 35 performs three-phase to two-phase conversion on the three-phase voltage commands Vu, Vv, Vw calculated by the voltage applying unit 40, which will be described later, and calculates an α -axis applied voltage V α and a β -axis applied voltage V β.

1-3-4. Torque estimating unit

The torque estimating unit 36 estimates the output torque Te of the motor based on the detected current value and the estimated value of the interlinkage magnetic flux.

In the present embodiment, the torque estimating unit 36 calculates the estimated value Te of the output torque from the product of the estimated value of the interlinkage magnetic flux and the detected current value. As shown in the following equation, the torque estimating unit 36 calculates an estimated value Te of the output torque by using an outer product (vector product) of an estimated value Φ α β of the α β -axis interlinkage magnetic flux and the α β -axis current detection value I α β.

[ mathematical formula 4]

Te＝Φα·Iβ-Φβ·Iα …(4)

1-3-5. Pulsation extraction section

The pulsation extraction unit 37 extracts a pulsation component from the estimated value Te of the output torque. The ripple extracting unit 37 extracts a ripple frequency component of a preset rotation order n from the estimated value Te of the output torque. Here, the pulsation frequency of the rotation order n is a frequency n times the rotation frequency of the rotor in electrical angle.

In the present embodiment, the ripple extraction unit 37 performs fourier transform on the estimated value Te of the output torque at a ripple frequency of a preset rotation order n according to the rotation angle θ of the motor, calculates fourier coefficients an and bn of a cosine wave and a sine wave, and calculates the fourier coefficients an and bn of the cosine wave and the sine wave as extraction values of the ripple component.

Here, fourier transform is explained. A mathematical expression obtained by performing fourier series expansion on the estimated value Te of the output torque is shown as follows.

[ math figure 5]

When the known fourier transform that integrates the cosine wave of the pulsation frequency of the rotation order n multiplied by the equation (5) in the period of the rotation order 1 is performed, the fourier coefficient an of the cosine wave of the rotation order n can be calculated.

[ mathematical formula 6]

Similarly, by multiplying equation (5) by a sine wave of a pulsation frequency of rotation order n and performing known fourier transform integrating in a period of rotation order 1, it is possible to calculate a fourier coefficient bn of the sine wave of rotation order n.

[ math figure 7]

Therefore, as shown in fig. 4, the ripple extracting unit 37 multiplies the estimated value Te of the output torque by a cosine wave of 2 and the rotation order n, performs a Low Pass Filter (LPF) on the multiplication value, performs fourier transform, and calculates a fourier coefficient an of the cosine wave. The ripple extracting unit 37 multiplies the estimated value Te of the output torque by a sine wave of 2 and the number n of revolutions, performs a Low Pass Filter (LPF) on the product, performs fourier transform, and calculates a fourier coefficient bn of the sine wave. Low-pass filtering processing is used instead of the integration processing of the period of rotation order 1 of equations (6) and (7). Therefore, the cutoff frequency of the low-pass filtering process is set to a frequency higher than the frequency of the rotation order 1 (the rotation angular frequency (rotation angular velocity) ω of the rotor). The cutoff frequency of the low-pass filtering process may also be changed in accordance with the rotational angular velocity ω of the rotor. Alternatively, a moving average process of rotating the period of order 1 may be used instead of the low-pass filtering process.

The pulsation extracting unit 37 of fig. 4 will be described in detail. The estimated value Te of the output torque is input to the gain 101 set to 2, and the estimated value Te of the output torque is set to 2 times. The rotation angle θ is input to the gain 113 set to the rotation order n, and the rotation angle θ is set to be n times. The number n of rotation steps is set to a natural number of 1 or more in advance. In addition, since the rotation angle θ in fig. 4 is an angle based on the U-phase winding, a rotation angle in which the phase of the pulsating component is added to the rotation angle θ may be used instead.

The output value of the gain 113 (n times the rotation angle n θ) is input to the sine function 102, and a sine wave of the rotation order n is calculated. The output value of the gain 113 (n times the rotation angle n θ) is input to the cosine function 103, and a cosine wave of the rotation order n is calculated. The output value (sine wave) of the sine function 102 and the output value (2 Te times the estimated value of the output torque) of the gain 101 are input to the multiplier 104, and multiplication processing is performed on the 2 input values. The output value (cosine wave) of the cosine function 103 and the output value (2 Te times the estimated value of the output torque) of the gain 101 are input to the multiplier 105, and multiplication processing is performed on 2 input values. The output value of the multiplier 104 is input to a Low Pass Filter (LPF)106, and the input value is subjected to low pass filtering processing to calculate a fourier coefficient bn of a sine wave. The output value of the multiplier 105 is input to a low-pass filter (LPF)107, and the input value is subjected to low-pass filtering processing to calculate a fourier coefficient an of a cosine wave.

1-3-6. Pulsation control section

The ripple control unit 38 calculates a voltage command correction value based on the extracted value of the ripple component. In the present embodiment, as shown in fig. 4, the cosine wave multiplier Ya is calculated by multiplying the cosine wave of the pulsation frequency of the rotation order n by the fourier coefficient an of the cosine wave, and the sine wave multiplier Yb is calculated by multiplying the sine wave of the pulsation frequency of the rotation order n by the fourier coefficient bn of the sine wave. Then, the ripple control unit 38 calculates a value obtained by multiplying the total value Yab of the cosine wave multiplier Ya and the sine wave multiplier Yb by the control gain Kv as the voltage command correction value Vc.

The total value Yab of the sine wave and the cosine wave is a value obtained by restoring only the pulsation component of the frequency of the rotation order n, and removing the noise components of the other frequencies.

There is a dead time Δ Td due to calculation delay from the detection of the current and the rotation angle to the calculation of the estimated value of the interlinkage magnetic flux, the estimated value of the output torque, the fourier coefficient, and the voltage command correction value, and the reflection of the estimated value, the fourier coefficient, and the voltage command correction value on the applied voltage. Therefore, the ripple control unit 38 calculates a phase delay Δ θ d due to the calculation delay from the rotation angle θ of the motor, and calculates a cosine wave and a sine wave of the ripple frequency of the rotation order n compensated for the phase delay Δ θ d.

The magnitude of the control gain Kv is set so that the gain crossover frequency of the open-loop characteristic obtained by multiplying the transfer characteristic from the voltage command after linear approximation to the output torque by the control gain Kv becomes larger than the maximum value of the pulsation frequency to be reduced. Since the pulsation frequency is proportional to the rotational speed of the motor, the pulsation frequency at the rated rotational speed of the motor becomes the maximum value of the pulsation frequency. In the present embodiment, it is realized that there is almost no response delay in the transfer characteristic from the actual output torque to the estimated value of the output torque, and therefore, the transfer characteristic from the voltage command to the output torque is almost equal to the transfer characteristic from the voltage command to the estimated value of the output torque. This provides a very significant effect of suppressing the torque ripple to the vicinity of the maximum value of the ripple frequency.

The pulsation controller 38 of fig. 4 will be described in detail. The rotation angle θ is input to the differentiator 114, and the rotation angular velocity ω is calculated. The output value (rotational angular velocity ω) of the differentiator 114 is input to a gain 116 set to the rotation order n, and the rotational angular velocity ω is set to n times. The output value of the gain 116 (n-fold value n ω of the rotational angular velocity) is input to a gain 117 set to a dead time Δ Td due to the computation delay, and multiplication processing of the dead time Δ Td is performed on the n-fold value n ω of the rotational angular velocity. The output value of the gain 117 is Δ Td × n × ω, and is a value obtained by converting the dead time Δ Td into the phase Δ θ d of the oscillation frequency of the rotation order n. The dead time Δ Td is set to, for example, a natural number multiple of the carrier period.

The output value of gain 113 (n times the number of revolutions n θ) and the output value of gain 117 (phase delay Δ θ d) are input to adder 118, and the phase delay Δ θ d is added to the rotation angle n θ of the rotation order n to calculate the rotation angle of the rotation order n after the phase delay compensation. By adding this phase delay Δ θ d, the phase is advanced by the amount of the phase delay Δ θ d, thereby compensating for the phase delay.

The output value of the adder 118 (the rotation angle of the rotation order n after the phase delay compensation) is input to the sine function 108, and a sine wave of the rotation order n after the phase delay compensation is calculated. The output value of the adder 118 is input to the cosine function 109, and a cosine wave of the rotation order n after phase delay compensation is calculated. The output value (sine wave) of the sine function 108 and the output value (fourier coefficient bn of the sine wave) of the low-pass filter 106 are input to a multiplier 110, and multiplication processing is performed on 2 input values to calculate a sine wave multiplication value Yb. The output value (cosine wave) of the cosine function 109 and the output value (fourier coefficient an of cosine wave) of the low-pass filter 107 are input to the multiplier 111, and multiplication processing is performed on 2 input values to calculate a cosine wave multiplication value Ya. Then, the output value (sine wave multiplication value Yb) of the multiplier 110 and the output value (cosine wave multiplication value Ya) of the multiplier 111 are input to the adder 112, and the 2 input values are added to calculate the total value Yab of the sine wave and the cosine wave. The output value (total value Yab) of the adder 112 is input to a gain 115 set as a control gain Kv, and the total value Yab is multiplied by the control gain Kv to calculate a voltage command correction value Vc.

1-3-7. Superposition section

The superimposing unit 39 superimposes the voltage command correction value on the voltage command basic value to calculate a superimposed voltage command. In the present embodiment, as shown in fig. 2, the superimposing unit 39 superimposes the voltage command correction value Vc on the q-axis voltage command basic value Vq0, calculates the superimposed q-axis voltage command Vq, and directly sets the d-axis voltage command basic value Vd0 as the superimposed d-axis voltage command Vd. The voltage command correction value Vc is subtracted from the q-axis voltage command basic value Vq0 by a subtractor 391 to calculate a superimposed q-axis voltage command Vq. Superimposed dq-axis voltage command Vdq is a vector composed of superimposed d-axis voltage command Vd and superimposed q-axis voltage command Vq.

1-3-8. Voltage applying part

The voltage applying unit 40 applies a voltage to the winding in accordance with the superimposed voltage command. In the present embodiment, the voltage application unit 40 includes a coordinate conversion unit 40a and a PWM control unit 40 b.

The coordinate conversion unit 40a performs fixed coordinate conversion and two-phase and three-phase conversion on the superimposed dq-axis voltage command Vdq based on the magnetic pole position θ, and calculates three-phase voltage commands Vu, Vv, Vw.

The PWM control unit 40b compares each of the three-phase voltage commands Vu, Vv, Vw with a carrier wave (triangular wave) that oscillates at the amplitude of the power supply voltage VDC/2 centered on 0 on the carrier frequency, and turns on the rectangular pulse wave when the voltage command exceeds the carrier wave and turns off the rectangular pulse wave when the voltage command falls below the carrier wave. The PWM control unit 40b outputs control signals corresponding to the rectangular pulse wave phases of the three phases to the inverter 20, and turns on and off the switching elements of the inverter 20.

1-3-9. Effect

In the magnetic flux estimating unit 35, as shown in equation (3), the interlinkage magnetic flux is directly estimated from the integrated value of the applied voltage and the integrated value of the current, and the estimated response frequency is not limited by a low-pass filter or an observer, so that the interlinkage magnetic flux of a sufficiently high frequency can be estimated with high accuracy. Further, since the torque estimation unit 36 directly estimates the output torque from the multiplication value of the flux and the current as shown in expression (4), the output torque can be estimated with high accuracy up to a sufficiently high frequency without impairing the information of the flux estimated to be high frequency and the current including high frequency components. Here, the sufficiently high frequency means a frequency higher than the ripple component of the interlinkage magnetic flux and the output torque at the rated rotation speed of the motor.

The ripple extracting unit 37 performs fourier transform on the estimated value of the output torque at a predetermined ripple frequency of the rotation order n to calculate a fourier coefficient of the ripple frequency of the rotation order n, and the ripple control unit 38 recovers only a torque ripple component having the ripple frequency of the rotation order n based on the fourier coefficient and amplifies the torque ripple component to calculate the voltage command correction value Vc.

Further, since the voltage command correction value Vc is fed back to the voltage command, the applied voltage of the motor can be corrected to a high frequency without being affected by a response delay of the current feedback control system as in patent document 1 fed back to the current command. Further, in the pulsation control section 38, the control gain Kv is set so that the gain cross frequency is higher than the pulsation frequency of the rotation order n at the rated rotation speed, whereby the torque pulsation component can be reduced up to a sufficiently high frequency.

In the configuration as in patent document 2, since the induced voltage including the current differential is estimated by inverting the voltage equation representing the response from the induced voltage to the current of the motor, high-frequency components are superimposed, and the high-frequency components of the induced voltage are reduced by using a low-pass filter or an observer. Therefore, there are problems as follows: the induced voltage and the output torque cannot be estimated up to a sufficiently high frequency, and the torque ripple component at a high frequency cannot be reduced.

< effect of reducing Torque ripple component >

The effect of reducing the torque ripple component according to the present embodiment will be described with reference to fig. 5. Fig. 5 shows waveforms of torque ripple components when the rotation speed is increased as time elapses so that the ripple frequency of the rotation order n rises from 50Hz to 1500Hz in the motor of the rotation order n at the rated rotation speed having the ripple frequency of 1500 Hz. The upper diagram of fig. 5 shows a comparative example in which voltage command correction value Vc is not superimposed, and the lower diagram shows an example of the present embodiment in which voltage command correction value Vc is superimposed. In comparison with the above comparative example, it is found that: in the present embodiment below, the torque ripple component is reduced. In this example, since the control gain Kv is set so that the gain crossover frequency becomes 1500Hz, although the effect of reducing the torque ripple component decreases as approaching 1500Hz, the torque ripple component can be reduced up to 1500 Hz.

< noise resistance >

Next, the effect of suppressing amplification of noise and vibration in the present embodiment will be described with reference to fig. 6. Fig. 6 shows the q-axis current measured without torque ripple in order to facilitate the observation of the influence of noise, and is a case where random noise is included in the detected current. The upper diagram of fig. 6 shows a comparative example in which the noise contained in the estimated value of the output torque is estimated and amplified without reducing the noise contained in the current and negatively fed back as the voltage command correction value, and the lower diagram shows an example of the present embodiment in which the influence of the noise contained in the current is reduced, the output torque is estimated, only the ripple frequency component contained in the estimated value of the output torque is extracted, and the voltage command correction value is calculated. In the upper and lower diagrams, the scale of the vertical axis is the same. In the upper comparative example, the noise of the current is greatly amplified as compared with the case where the voltage command is not corrected. Since the noise of the current is amplified as the magnitude of the control gain is larger, a tradeoff between reducing the torque ripple component to a sufficiently high frequency and suppressing the amplification of the current noise is made. In the present embodiment below, the noise of the current is not amplified, and the magnitude of the control gain Kv is increased, the torque ripple component can be reduced up to a sufficiently high frequency.

According to the present embodiment, since the ac linkage flux is directly estimated using the current and the voltage without the need to pass through the induced voltage, the ac linkage flux and the output torque can be estimated up to a sufficiently high frequency, and since the voltage command correction value for reducing the torque ripple component of the rotation order n is calculated by extracting the ripple frequency component of the rotation order n and the voltage command is corrected without amplifying unnecessary noise and vibration, the applied voltage of the motor can be corrected up to a sufficiently high frequency without being limited by the response delay of the current feedback control system. Therefore, an unprecedented significant effect that torque ripple can be reduced up to a sufficiently high frequency is obtained without amplifying unnecessary noise and vibration.

Further, since the q-axis voltage command is corrected by the voltage command correction value Vc without correcting the d-axis voltage command, only one voltage command needs to be corrected, and torque ripple can be reduced by a simple operation.

1-3-10. Examples of transfer

In the above embodiment, the magnetic flux estimation unit 35 calculates the estimated value Φ α β of the α β -axis interlinkage magnetic flux represented by the α β -axis coordinate system. However, the magnetic flux estimation unit 35 may calculate the estimated value Φ dq of the dq-axis linkage magnetic flux represented by the dq-axis coordinate system. In this case, the magnetic flux estimating unit 35 calculates an estimated value Φ dq of the dq-axis linkage magnetic flux from an integrated value of the dq-axis applied voltage Vdq and an integrated value of the dq-axis current detection value Idq, as shown by the following equation. The magnetic flux estimating unit 35 uses the superimposed dq-axis voltage command Vdq calculated by the superimposing unit 39 and uses the dq-axis current detection value Idq calculated by the current detecting unit 31 as the dq-axis applied voltage Vdq. The dq-axis linkage flux Φ dq is a vector composed of the d-axis linkage flux Φ d and the q-axis linkage flux Φ q.

[ mathematical formula 8]

Φdq＝∫(∫dq-R·Idq)dt＝∫Vdq dt-R∫Idq dt …(8)

In response to this, the torque estimating unit 36 may calculate the estimated value Te of the output torque by using an outer product (vector product) of the estimated value Φ dq of the dq-axis linkage magnetic flux and the dq-axis current detection value Idq, as shown in the following equation.

[ mathematical formula 9]

Te＝Φd·Iq-Φq·Id …(9)

2. Embodiment mode 2

Next, motor control device 30 according to embodiment 2 will be described. The same components as those in embodiment 1 are omitted from description. The basic configuration of the control device 30 according to the present embodiment is the same as that of embodiment 1, but the configuration of the torque estimating unit 36 is different from that of embodiment 1.

In the present embodiment, the torque estimation unit 36 uses an estimation formula that can include more pulsation components than in embodiment 1. That is, an estimation formula constructed based on the principle of obtaining the generated output torque by partially differentiating the magnetic energy by the rotation angle is used.

Therefore, the torque estimating unit 36 calculates the estimated value Te of the output torque from a multiplication value of a differential value obtained by differentiating the estimated value of the interlinkage magnetic flux by the rotation angle and the current detection value. As shown in items 1 and 2 on the right side of the following equation, the torque estimating unit 36 calculates the estimated value Te of the output torque from an 1/2-fold value of the inner product of the α β -axis current detection value I α β and a differential value obtained by differentiating the estimated value Φ α β of the α β -axis interlinkage magnetic flux by the rotation angle. In the right-hand items 3 and 4 of the following equations, 1/2 times the inner product of the α β axis current detection value I α β and a differential value obtained by differentiating the α β axis magnet interlinkage magnetic fluxes Φ m α, Φ m β by the rotation angle are added. However, when the ratio of the pulsation of the magnet interlinkage magnetic flux with respect to the total interlinkage magnetic flux is small, the items 3 and 4 may be omitted.

[ mathematical formula 10]

Here, regarding the angular differential values of the α β axis magnet interlinkage magnetic fluxes Φ m α, Φ m β of the items 3 and 4, the torque estimating unit 36 calculates the angular differential values of the α β axis magnet interlinkage magnetic fluxes Φ m α, Φ m β corresponding to the current rotation angle with reference to table data or a sine function in which the relationship between the angular differential values of the α β axis magnet interlinkage magnetic fluxes Φ m α, Φ m β and the rotation angle is set in advance.

Alternatively, as shown in items 1 and 2 on the right side of the following equation, the torque estimating unit 36 may calculate the estimated value Te of the output torque from an inner product of the α β -axis current detection value I α β and a differential value obtained by differentiating the estimated value Φ α β of the α β -axis interlinkage magnetic flux by the rotation angle. In terms 3, 4, and 5 on the right side of the following equation, 1/2 times the second order of the α β axis current detection value I α β and the differential values obtained by differentiating the α β axis inductances L α, L β, and L α β by the rotation angle are subtracted. However, when the ratio of the pulsation of the magnetic flux generated by the inductance to the total interlinkage magnetic flux is small, the items 3, 4, and 5 may be omitted.

[ mathematical formula 11]

Here, regarding the angle differential values of the α β axis inductances L α, L β, and L α β in the items 3, 4, and 5, the torque estimation unit 36 calculates the angle differential values of the α β axis inductances L α, L β, and L α β corresponding to the current rotation angle by referring to table data or a sine function in which the relationship between the angle differential values of the α β axis inductances L α, L β, and L α β and the rotation angle is set in advance.

3. Embodiment 3

Next, motor control device 30 according to embodiment 3 will be described. The same components as those in embodiment 1 are omitted from description. The basic configuration of the control device 30 according to the present embodiment is the same as that of embodiment 1, but the configurations of the pulsation extracting unit 37 and the pulsation control unit 38 are different from those of embodiment 1.

In the present embodiment, as shown in fig. 7, the ripple extracting unit 37 calculates, as the extracted value Ter of the ripple component, a value obtained by performing, on the estimated value Te of the output torque, a band-pass filter process of passing a component of the ripple frequency of the preset rotation order n and attenuating components other than the ripple frequency of the rotation order n. The ripple control unit 38 multiplies the extracted value Ter of the ripple component by the control gain Kv to calculate a voltage command correction value Vc.

Fig. 7 is explained in detail. The rotation angle θ is input to the differentiator 114, and the rotation angular velocity ω is calculated. The output value (rotational angular velocity ω) of the differentiator 114 is input to a gain 116 set to the rotation order n, and the rotational angular velocity ω is set to n times. The n-fold value n ω of the rotational angular velocity is input to the Band Pass Filter (BPF)121, and is set as the pass frequency of the band pass filter 121. The estimated value Te of the output torque is input to a band-pass filter (BPF)121, band-pass filter processing is performed on the estimated value Te of the output torque, and a torque ripple component Ter is output. Then, the torque ripple component Ter is input to a gain 115 set as a control gain Kv, and the torque ripple component Ter is multiplied by the control gain Kv to calculate a voltage command correction value Vc.

In the case of this embodiment as well, as in embodiment 1, the ripple component of the rotation order n can be extracted reliably, and noise can be reduced. The noise reduction amount can be adjusted by the characteristic design of the band-pass filter. Therefore, the torque ripple can be suppressed up to a sufficiently high frequency without amplifying unnecessary noise and vibration.

4. Embodiment 4

Next, the motor control device 30 according to embodiment 4 will be explained. The same components as those in embodiment 1 are omitted from description. The basic configuration of the control device 30 according to the present embodiment is the same as that of embodiment 1, but the configuration of the pulsation control unit 38 is different from that of embodiment 1.

In the present embodiment, the ripple control unit 38 performs a control operation based on the fourier coefficient an of the cosine wave to calculate the control value Ua of the cosine wave, multiplies the control value Ua of the cosine wave by the cosine wave of the ripple frequency of the rotation order n to calculate the cosine wave multiplication value Ya, performs a control operation based on the fourier coefficient bn of the sine wave to calculate the control value Ub of the sine wave, multiplies the control value Ub of the sine wave by the sine wave of the ripple frequency of the rotation order n to calculate the sine wave multiplication value Yb, and calculates a voltage command correction value Vc as a value obtained by summing up the cosine wave multiplication value Ya and the sine wave multiplication value Yb.

In the present embodiment, as in embodiment 1, the ripple control unit 38 calculates a phase delay Δ θ d due to the calculation delay from the rotation angle θ of the motor, and calculates a cosine wave of the ripple frequency of the rotation order n and a sine wave of the ripple frequency of the rotation order n compensated for the phase delay Δ θ d.

Fig. 8 is explained in detail. The configuration of the pulsation extracting unit 37 is the same as that of embodiment 1 of fig. 4, and therefore, the description thereof is omitted. As in embodiment 1, the rotation angle θ is input to the differentiator 114, and the rotation angular velocity ω is calculated. The output value (rotational angular velocity ω) of the differentiator 114 is input to a gain 116 set to the rotation order n, and the rotational angular velocity ω is set to n times. The output value of the gain 116 (n-fold value n ω of the rotational angular velocity) is input to a gain 117 set to a dead time Δ Td due to the computation delay, and multiplication processing of the dead time Δ Td is performed on the n-fold value n ω of the rotational angular velocity.

The output value of gain 113 (n times the number of revolutions n θ) and the output value of gain 117 (phase delay Δ θ d) are input to adder 118, and the phase delay Δ θ d is added to the rotation angle n θ of the rotation order n to calculate the rotation angle of the rotation order n after the phase delay compensation. By adding this phase delay Δ θ d, the phase is advanced by the amount of the phase delay Δ θ d, thereby compensating for the phase delay.

The output value of the adder 118 (the rotation angle of the rotation order n after the phase delay compensation) is input to the sine function 108, and a sine wave of the rotation order n after the phase delay compensation is calculated. The output value of the adder 118 is input to the cosine function 109, and a cosine wave of the rotation order n after phase delay compensation is calculated.

In the present embodiment, the control operation in the pulsation control unit 38 is a proportional operation. The fourier coefficient an of the cosine wave is input to a gain 120 set to a proportional gain Ka of the cosine wave, and the fourier coefficient an of the cosine wave is multiplied by the proportional gain Ka. The fourier coefficient bn of the sine wave is input to a gain 119 set as a proportional gain Kb of the sine wave, and the fourier coefficient bn of the sine wave is multiplied by the proportional gain Kb. The control operation in the pulsation control unit 38 may be any control operation other than a proportional operation, for example, a proportional-integral operation, and in the case of the proportional-integral operation, the integrators are added in parallel to the gains 119 and 120, respectively.

The output value (sine wave) of the sine function 108 and the output value of the gain 119 are input to the multiplier 110, and multiplication processing is performed on the 2 input values to calculate a sine wave multiplication value Yb. The output value (cosine wave) of the cosine function 109 and the output value of the gain 120 are input to the multiplier 111, and multiplication processing is performed on 2 input values to calculate a cosine wave multiplication value Ya. The output value (sine wave multiplication value Yb) of the multiplier 110 and the output value (cosine wave multiplication value Ya) of the multiplier 111 are input to the adder 112, and the added value of the 2 input values is output as the voltage command correction value Vc.

As shown in the example of fig. 8, if the control operation of the pulsation control unit 38 is a proportional operation, a process equivalent to that of embodiment 1 is obtained. If the control operation of the pulsation control unit 38 is a proportional-integral operation, the amplification factor can be increased until the torque pulsation component is eliminated by the integral operation, and therefore the magnitude of the stable torque pulsation can be further reduced.

[ other embodiments ]

In each of the above embodiments, the superimposing unit 39 superimposes only the voltage command correction value Vc on the q-axis voltage command basic value Vq0, but may superimpose the correction value on the d-axis voltage command basic value Vd0, normally, since the contribution degree of the q-axis voltage command to the torque is high, and thus, it is relatively simple to superimpose only the q-axis voltage command, but for a motor or the like having a large reluctance torque, the voltage command correction value Vc may be distributed to the d-axis correction value in accordance with the ratio of the reluctance torque to the total torque, and superimposed on the d-axis voltage command basic value Vd 0. In this case, the torque ripple can be reduced, and the same effects as those of the respective embodiments can be obtained.

In the above embodiments, the control system for calculating the voltage command in the dq axis coordinate system is used, but the control system for calculating the voltage command in the α β axis coordinate system or the UVW phase coordinate system may be used. In this case, the voltage command correction value is captured as a value of the dq-axis coordinate system, and is converted into a voltage command correction value of the α β -axis coordinate system or the UVW-phase coordinate system by coordinate conversion, and may be superimposed on the voltage command of each coordinate system. In this case, the same effects as those of the respective embodiments can be obtained.

The motor control device of each of the above embodiments can be applied to a control device of an electric power steering for assisting steering of an automobile, and in this case, since torque pulsation of the motor is reduced and unnecessary noise and vibration are not amplified, noise and pulsation of a steering wheel felt by an automobile driver can be reduced.

Although various exemplary embodiments and examples have been described in the present application, the various features, modes, and functions described in 1 or more embodiments are not limited to the application to specific embodiments, and may be applied to the embodiments alone or in various combinations. Therefore, numerous modifications not illustrated are also considered to be included in the technical scope disclosed in the present specification. For example, the present invention also includes a case where at least 1 component is modified, added, or omitted, and a notch where at least 1 component is extracted and components of other embodiments are combined.

Description of the reference symbols

20 inverter, 25 dc power supply, 30 motor control device, 31 current detection unit, 32 rotation detection unit, 33 voltage detection unit, 34 basic voltage command calculation unit, 35 magnetic flux estimation unit, 36 torque estimation unit, 37 ripple extraction unit, 38 ripple control unit, 39 superposition unit, 40 voltage application unit, I α β α β axis current detection value, Idq dq axis current detection value, Kv control gain, Te output torque estimation value, Vc voltage command correction value, Vdq0 dq axis voltage command basic value, Ya cosine wave multiplication value, Yb sine wave multiplication value, Yab total value, Ua remainder wave control value, Ub sine wave control value, an cosine wave fourier coefficient, bn sine wave fourier coefficient, n rotation order, Φ α β α β axis interlinkage estimation value, Φ dq dq axis interlinkage magnetic flux estimation value, θ rotation angle.