JP4984472B2 - Control device for electric power steering device - Google Patents

Description

  The present invention relates to a control device for an electric power steering apparatus that applies a steering assist force by a motor to a steering system of an automobile or a vehicle, and more particularly, an electric motor with a dead time compensation function that improves the dead time compensation of an inverter for driving a motor. The present invention relates to a control device for a power steering device.

  An electric power steering device that energizes an automobile or vehicle steering device with the rotational force of a motor assists the steering shaft or rack shaft with a transmission mechanism such as a gear or a belt via a reduction gear. Energize power. In order to supply a current to the motor so that the motor generates a desired torque, an inverter or the like is used in the motor drive circuit.

  Here, FIG. 4 shows a basic configuration of an electric power steering device disclosed in Japanese Patent Application Laid-Open No. Hei 8-142288 (Patent Document 1) and the like, and FIG. 5 shows details of the motor drive circuit 20. In FIG. 4, the steering torque T detected by the torque sensor 10 is input to the phase compensator 2 to calculate a torque command value TA. Next, the torque command value TA is input to the steering assist command value calculator 3, and the steering assist command value Iref is calculated by the steering assist command value calculator 3 in consideration of the vehicle speed V detected by the vehicle speed sensor 11. The current i of the motor 1 is detected by the motor current detection circuit 8 and fed back to the subtractor 4A, and an error from the steering assist command value Iref is calculated. The error is proportional-integral controlled by the proportional calculator 6 and the integral calculator 7. The steering assist command value Iref is input to the differential compensator 5 for improving the transient response, and the outputs of the differential compensator 5, the proportional calculator 6 and the integral calculator 7 are added by the adder 4B to control the current. A value E is calculated. The motor drive circuit 20 supplies current to the motor 1 based on the current control value E.

  As shown in FIG. 5, the motor drive circuit 20 includes an inverter unit composed of FET1 to FET4 which are switching elements, and a gate control unit which controls each gate of the FET1 to FET4. In the inverter section, an upper and lower arm composed of FET1 and FET3 and an upper and lower arm composed of FET2 and FET4 are arranged, and an H bridge circuit is formed with a configuration in which the motor 1 is connected between connection points of the upper and lower arms. In the gate control unit, the current control value E is input to the conversion unit 21, timing signals D1 to D4 for the FET1 to FET4 are generated, and input to the gate drive circuits 22A, 23A, 22B, and 23B to drive the FET1 to FET4. A possible gate signal is generated. Note that a voltage (Vdc) is supplied from the battery 12 to each of the FET1 to FET4. However, the timing signals D3 and D4 generated by the converter 21 are not directly input to the gate drive circuits 22B and 23B, but are input to the dead time circuits 24 and 25, respectively, for the following reason.

  The upper and lower arms FET1 and FET3 of the inverter are alternately turned on and off, and similarly, FET2 and FET4 are alternately turned on and off. However, the FET is not an ideal switch and does not instantly turn on and off as instructed by the gate signals D1 to D4, and requires a turn-on time and a turn-off time. For this reason, when an instruction to turn on FET1 and an instruction to turn off FET3 are made at the same time, FET1 and FET3 are turned on at the same time and the upper and lower arms are short-circuited. Therefore, when an off signal is given to the gate drive circuit 22A so that the FET1 and FET3 are not turned on at the same time, an on signal is not given immediately to the gate drive circuit 22B, and the dead time circuit 24 performs the interval between the dead times Td. In this case, the on / off signal is given to the gate drive circuit 22B to prevent the FET1 and FET3 from being vertically short-circuited. The same applies to FET2 and FET4.

  However, the existence of this dead time Td causes a problem of insufficient torque and torque ripple for the control of the electric power steering apparatus, and this problem will be described.

First, FIG. 6 shows the relationship between dead time Td, turn-on time Ton, and turn-off time Toff. The gate signal D is an on / off signal for the basic FET1 and FET3. However, actually, the gate signal D1 is given to the FET1, and the gate signal D2 is given to the FET2, so that the dead time Td is secured. The phase voltage composed of FET1 and FET2 is Van. Even if an on signal by the gate signal D1 is given, the FET 1 is not turned on immediately, but it is turned on with a turn-on time Ton, and even when an off signal is given, it is not turned off immediately and a turn-off time Toff is required. ing. Vdc is the power supply voltage of the inverter. Therefore, the total delay time Ttot is expressed as the following formula 1.
(Equation 1)
Ttot = Td + Ton-Toff

Here, the turn-on time Ton and the turn-off time Toff vary depending on the type and capacity of the FET and battery used, and the dead time Td is generally larger than the turn-on time Ton and the turn-off time Toff.

The influence of the dead time Td on the voltage is as follows. As shown in FIG. 6, the actual gate signals D1 and D3 differ from the ideal gate signal D due to the dead time Td. Therefore, a distortion voltage ΔV is generated. The distortion voltage ΔV is expressed by Equation 2 when the motor current direction is positive (when the current direction flows from the power source to the motor), and the motor current direction is negative. In the case of (when the direction of current flows from the motor to the power source), Equation 3 is obtained.
(Equation 2)
−ΔV = − (Ttot / Ts) · (Vdc / 2)
However, Ts is the reciprocal Ts = 1 of the PWM frequency fs when the inverter is PWM-controlled.
/ Fs.

(Equation 3)
ΔV = (Ttot / Ts) · (Vdc / 2)

When the above formulas 2 and 3 are expressed by one formula, the following formula 4 is obtained.
(Equation 4)
ΔV = −sign (Is) · (Ttot / Ts) · (Vdc / 2)
Here, sign (Is) represents the polarity of the motor current.

What is derived from Equation 4 is that the distortion voltage ΔV has a higher frequency fs, and the larger the power supply voltage Vdc, the greater the influence of the dead time Td.

  Here, the influence of the dead time Td on the distortion voltage ΔV has been described, but the current or torque also has an undesirable influence due to the dead time Td. As for the distortion current, a phenomenon (zero clamping phenomenon) in which the current sticks to near zero when the current changes from positive to negative or from negative to positive is caused by the dead time Td. This is because the load (motor) is an inductance, and the voltage decrease due to the dead time Td tends to keep the current at zero.

Further, the influence of the dead time Td on the torque appears in an insufficient torque output and an increase in torque ripple. That is, the distortion current generates low-order harmonics, which leads to an increase in torque ripple. Further, the torque output shortage occurs because the actual current affected by the dead time Td is smaller than the ideal current.
Japanese Patent Laid-Open No. 08-142848 WO 2005/023626 A1 Ben Brahim, “The analysis and compensation of dead-time effects in the three phase PWM inverters”, Proceedings of the IEEE-IECON98, Volume 2, pages 792-792

  In order to prevent the undesired influence of the dead time Td as described above, various dead time compensations have been considered. The basic idea is to compensate for the distortion voltage ΔV shown in Equation 4. Therefore, in order to compensate Equation 4, correction is performed with the following correction voltage (compensation value) Δu of Equation 5.

(Equation 5)
Δu = sign (Is) · (Ttot / Ts) · (Vdc / 2)

Here, the problem is that the polarity sign (Is) of the current Is cannot be detected correctly. When measuring the polarity of the current Is, it is difficult to correctly measure the polarity of the current Is due to the noise of PWM control and the above-described zero clamping phenomenon of the current.

Many conventional dead time compensations are described in, for example, Ben Brahim, “Ben Brahim, The analysis and compensation of dead-time effects”, Journal of IEEE-IECON 98, Volume 2, pages 792 to 797. in the three phase PWM inverters, Proceedings of the IEEE-IECON98, Volume
2, pages 792-792) (Non-Patent Document 1). However, the method disclosed in Non-Patent Document 1 is a complicated method, requires additional hardware, and does not take measures that take into account changes in load current such as motor current. Therefore, dead time compensation to prevent short circuit between the upper and lower arms of the inverter leads to motor voltage, current distortion, torque output shortage and torque ripple increase. Conventionally, improvement measures for dead time compensation are complicated and hardware Incomplete dead time compensation that does not take into account the effect of motor load current.

  As a solution to such a problem, there is one disclosed in WO 2005/023626 A1 (Patent Document 2). The outline will be described below.

  FIG. 7 shows an example of a basic configuration of a control device for an electric power steering device with a dead time compensation function. A current command value calculation circuit 31 calculates a current command value Iref based on a steering torque signal Tref generated on the steering shaft. Is done. On the other hand, the current Imes of the motor 30 is detected by the current detector 36 and fed back to the subtracting unit 32 to calculate an error between the current command value Iref and the motor current Imes, and is input to the current control circuit 33 to be input to the voltage command value u. Is calculated. The inverter 35 is PWM controlled based on the voltage command value u, and a dead time compensation circuit 50 is added to the basic control configuration. The dead time compensation circuit 50 receives the current command value Iref, calculates a compensation value Δu, and adds it to the voltage command value u, which is the output of the current control circuit 33, by the adder 34. The addition result is input to the inverter 35 as a voltage command value with dead time compensation.

  Next, an example of the dead time compensation circuit 50 will be described with reference to FIG. 8. The current command value Iref is input to the reference model circuit 51 and the model current Imod is output. The model current Imod is input to the polarity determination circuit 53 and the absolute value circuit 55 in the compensation value amount calculation circuit A through the hysteresis circuit 52. The compensation value Δu output from the dead time compensation circuit 50 is composed of polarity and quantity (hereinafter referred to as “compensation value quantity Δu2”).

  The model current Imod from the reference model circuit 51 is input to the hysteresis circuit 52 and output as a model current Imodh having hysteresis characteristics. The model current Imodh is used to determine the polarity determination circuit 53 for determining the polarity and the compensation value amount Δu2. Input to the absolute value circuit 55. The reason for providing the hysteresis circuit 52 is to prevent the polarity from becoming unstable (chattering of compensation value) when the load current passes through the zero point, and to enable stable control.

The model current Imodh having hysteresis characteristics is input to the polarity determination circuit 53, and the determination polarity sign (Imodh) is output in the form of “+1” or “−1”. The reference model circuit 51 receives the current command value Iref and calculates the model current Imod. The transfer function MR (s) of the reference model circuit 51 is as shown in the following equation (6).
(Equation 6)
MR (s) = 1 / (1 + Tc · s)
Here, Tc = 1 / (2π · fc), and fc is a cutoff frequency of the current control loop.

The first-order lag function of Equation 6 is a model function of a current control loop in which the function 1 / (R + s · L) indicating the motor 30 in FIG. 7 is derived based on the current control circuit 33, the inverter 35, and the current detector 36. .

Here, the relationship between the current command value Iref and the actual motor current Imes is as shown in FIG. 9, and the actual motor current Imes contains a lot of noise, which makes it difficult to determine the polarity near zero current. I'm making things. Therefore, without using the actual current Imes, the motor current Imod is generated via the first-order lag circuit based on the current command value Iref without noise, and the hysteresis is given by the hysteresis circuit 52. The model current Imodh from the hysteresis circuit 52 is input to the polarity determination circuit 53, and sign (Imodh) which is the polarity of the model current Imodh is calculated. The polarity sign (Imod) takes a value of either (+1) or (−1) as shown in Equation 7. It is very difficult to correctly determine the polarity by measuring the actual motor current and inverter current due to noise or the like, but there is no such concern if it is determined using the model current.
(Equation 7)
sign (Imodh) = "+ 1" or "-1"

Further, the compensation value amount Δu2 is obtained by the compensation value amount calculation circuit A as follows. The model current Imodh is input to the absolute value circuit 55, the absolute value | Imodh | is input to the change value calculation circuit 56, and the change value Δu1 is calculated. The reason for obtaining the absolute value | Imodh | of the model current Imodh is that the polarity needs to be unified since the compensation value amount is calculated. The absolute value | Imod | is input to the change value calculation circuit 56, and the change value calculation circuit 56 calculates the change value Δu1. The input / output relationship of the change value calculation circuit 56 is expressed by the following formula 8.
(Equation 8)
Δu1 = Reg · | (Imodh−Ic) |
Here, Reg indicates an equivalent resistance, and Imodh> Ic.

In Equation 8, Δu1 = 0 when Ic>Imod> 0. That is, when the model current Imodh is small, the change value Δu1 is not compensated. That is, the change value Δu1 expressed by Equation 8 is proportional to the model current Imodh, and the change value Δu1 in consideration of the change in the motor current due to the change in the motor load affects the compensation value Δu.

On the other hand, a fixed value Δu0 is set in the fixed value setting circuit 54. The fixed value Δu0 is a value of the following formula 9.
(Equation 9)
Δu0 = (Ttot / Ts) · (Vdc / 2)

Here, the total delay time Ttot is Ttot = Td + Ton-Toff as shown in the equation (1). The dead time Td, turn-on time Ton, and turn-off time Toff are values determined by the type of switching element used in the inverter. For example, in the case of an FET, the turn-on time Ton and turn-off time Toff increase as the rated voltage and the rated current increase. Increases and the dead time Td also increases.

  Next, the change value Δu1 and the fixed value Δu0 are added by the adding unit 58, and a compensation value amount Δu2 (= Δu0 + Δu1) is calculated.

Finally, the polarity signal sign (Imod) from the polarity determination circuit 53 and the compensation value amount Δu2 = (Δu0 + Δu1) from the compensation value amount calculation circuit A are multiplied by a multiplication circuit 57 as a polarity applying circuit, and the following equation 10 Accordingly, a compensation value Δu having polarity is calculated.
(Equation 10)
Δu = sign (Imodh) · Δu2
= Sign (Imod) · (Δu0 + Δu1)

The compensation value Δu calculated in this way is added by the adder 34 to the voltage command value u which is the output of the current control circuit 3 shown in FIG. The meaning that the compensation value Δu is added to the voltage command value u is to improve voltage distortion, current distortion, and torque ripple due to the dead time Td for preventing the upper and lower arms from being short-circuited in the basic control indicated by the voltage command value u. The control is performed in consideration of the compensation value Δu.

  The compensation value Δu that has passed through the hysteresis circuit 52 is illustrated in FIG. Since the model current has less noise, the hysteresis width can be made smaller than when the actual motor current is used, so that more accurate dead time compensation is possible.

  However, when the model current Imod or Imodh is in the vicinity of 0 [A], there is a portion in which the fixed value −Δu0 rapidly changes from the fixed value −Δu0. In a brushless motor driven by a sine wave, the current command value for each phase is a sine wave. When the assist is small, the amplitude of the sin wave current command value is small, the timing for inputting the compensation value for the dead time may be shifted, and a steering torque fluctuation may occur. In order to prevent this effect, the dead time compensation value is reduced when the current command value is small. However, if the dead time compensation value is reduced, the amount of assist near the steering center is insufficient, resulting in a feeling of friction. Furthermore, since the current command value becomes small even when the steering wheel returns, the dead time compensation value becomes small and the distortion of the motor current increases, thereby deteriorating the steering sound.

  The present invention has been made under the circumstances described above, and the object of the present invention is to distort motor voltage by using dead time compensation that is simple in configuration and takes into account the influence of motor load current and angle information. Another object of the present invention is to provide a control device for an electric power steering apparatus that has little distortion of current, torque ripple, and that does not deviate dead time compensation timing even when the steering speed and current command value are small.

The present invention includes a current command value calculation unit that calculates a d-axis current command value Idref and a q-axis current command value Iqref based on a torque command value, a dead time circuit, and a dead time compensation function. An inverter that drives a motor that applies a steering assist force to the motor, and a current control unit that converts the d-axis current command value Idref and the q-axis current command value Iqref into voltage command values for each phase of the motor and inputs them to the inverter; And a control device for an electric power steering apparatus with a dead time compensation function for controlling the motor by vector control, and the object of the present invention is to calculate a compensation value based on the torque command value. A current advance calculation unit for calculating a current advance of the d-axis current command value Idref and the q-axis current command value Iqref; An advance angle calculation unit for calculating an advance amount based on the angular velocity of the data, and a dead time compensation timing for calculating a dead time compensation timing based on the addition result of the motor angle, the advance amount and the current advance And calculating a dead time by multiplying the compensation value from the dead time compensation timing calculating unit by the compensation value and adding the result to the voltage command value of each phase of the motor .

Further, the above object of the present invention is that the current advance calculation unit calculates the current advance according to θ = atan (Idref / Iqref), or the advance angle calculation unit calculates k as a coefficient. As described above, the advance amount = k · angular velocity is calculated, or the compensation timing is “1” or “−1”, which is more effectively achieved.

  According to the present invention, since the dead time is compensated based on the angle information, the timing of the dead time compensation is not easily shifted even when the assist is small, and the steering noise when the steering wheel returns can be reduced. The feeling of sticking due to the fact that the current does not rise due to insufficient compensation can be eliminated.

  In the present invention, the compensation timing is obtained from the current command value, the angle, and the angular velocity to compensate for the dead time, so that the compensation can be performed without deviation at the timing at which the polarity of the current direction is switched.

  First, vector control of a motor, which is a premise of the present invention, will be described with reference to FIG.

The rotor position detector 111 attached to the motor 110 detects the rotor position (angle) θe, and the rotor position θe and the angular velocity (electrical angular velocity) ωe are input to the vector control phase command value calculation unit 100 and are input to the three-phase excitation coil. On the other hand, the motor 110 is rotationally driven by a two-phase excitation method in which the excitation coils are sequentially switched one by one while energizing two phases simultaneously.

The motor drive control device based on vector control is based on the vector control phase command value calculation unit 100, the current command values Iaref, Ibref, Icref and the motor phase currents Ia, Ib, Ic from the vector control phase command value calculation unit 100. Subtractor 112-1, 112-2, 112-3 for obtaining phase current deviation, PI controller 113 for proportional-integral control of each phase current deviation obtained by subtractor 112-1 to 112-3, and PI control PWM control unit 114 that inputs PWM command values Va, Vb, and Vc and performs PWM control, and inverter 115 that rotates and drives each phase command current to motor 110 based on the PWM control of PWM control unit 114 It consists of

Then, the dq-axis current command value calculation unit 103 in the vector control phase command value calculation unit 100 determines the current command values Idref and Iqref of the vector control dq-axis component, and then the dq-axis component of the dq-axis component. The current command values Idref and Iqref are converted into respective phase current command values Iaref, Ibref, and Icref, and the feedback control unit is configured to close all by phase control instead of dq axis control.

The current control unit A is composed of subtraction units 112-1, 112-2, and 112-3 for obtaining each phase current deviation, and a PI control unit 113 for inputting each phase current deviation. Current detectors 116-1, 116-2, 116-3 are arranged between the motor 110 and the motor phase currents Ia, Ib, detected by the current detectors 116-1, 116-2, 116-3. A feedback circuit B is formed in which Ic is fed back to the subtraction units 112-1, 112-2, and 112-3, respectively.

The vector control phase command value calculation unit 100 includes a dq axis current command value calculation unit 103 that calculates current command values Idref and Iqref of dq axis components of vector control, and a two-phase as each phase current command calculation unit. / 3 phase conversion unit 104. Further, the vector control phase command value calculation unit 100 includes a rotor position (rotation angle) θe detected by a rotor position detector 111 such as a resolver, an angular velocity ωe calculated by the differentiation circuit 117, and a torque sensor ( A torque command value Tref determined based on the torque detected in (not shown) is input, and a phase command value by vector control is calculated.

Next, drive control of the motor 110 will be described.

First, the vector phase command value calculation unit 100 inputs the rotor position θe and the angular velocity ωe, and the dq axis current command value calculation unit 103 determines the q axis current command value Iqref based on the torque command value Tref and the torque coefficient Kt. It calculates according to following formula 11.
(Equation 11)
Iqref = Tref / Kt
The d-axis current command value Idref is a function of the q-axis current command value Iqref, and is expressed as the following formula 12.
(Equation 12)
Idref = f (Iqref)

The function f (Iqref) is determined by the motor output requirement specification.

As shown in Equation 11, the q-axis current command value Iqref is derived from a motor output equation that the motor output corresponds to the electric power. The dq axis current command values Idref and Iqref are converted into respective phase current command values Iaref, Ibref, and Icref by the two-phase / three-phase conversion unit 104. The determinant C2 is a constant determined by the rotor position θe of the motor 110, as shown in Equations 13 and 14.

この後のフィードバック制御系はｄ−ｑ軸制御ではなく、各相制御になっている。つまり、電流検出器１１６−１，１１６−２，１１６−３で検出されたモータ相電流Ｉａ，Ｉｂ，Ｉｃと、各相電流指令値Ｉａｒｅｆ，Ｉｂｒｅｆ，Ｉｃｒｅｆを減算部１１２−１，１１２−２，１１２−３で減算して各偏差を算出する。次に、各相電流偏差をＰＩ制御部１１３で制御してインバータ１１５の指令値、即ちＰＷＭ制御部１１４のデューティを表わす電圧値Ｖａ，Ｖｂ，Ｖｃが算出され、その値に基づいてＰＷＭ制御回路１１４がインバータ１１５をＰＷＭ制御してモータ１１０を駆動する。

The subsequent feedback control system is not a dq axis control but a phase control. In other words, the motor phase currents Ia, Ib, Ic detected by the current detectors 116-1, 116-2, 116-3 and the phase current command values Iaref, Ibref, Icref are subtracted by the subtractors 112-1, 112-2. , 112-3 to calculate each deviation. Next, each phase current deviation is controlled by the PI control unit 113 to calculate the command value of the inverter 115, that is, the voltage values Va, Vb, Vc representing the duty of the PWM control unit 114, and the PWM control circuit based on the values. 114 drives the motor 110 by PWM control of the inverter 115.

  Here, the compensation of the dead time Td is effective by compensating for the timing at which the sign of the motor current switches, but when the motor is rotating, the angle signal is read and a compensation signal (compensation value Δut) is output. The delay until is not negligible, and the compensation timing tends to be delayed. Therefore, it is effective to compensate for the dead time based on the angle signal advanced according to the rotational speed. Also, during field-weakening control, that is, when the d-axis current command value (Idref) is not 0, the phase of each motor phase current command value (Iaref, Ibref, Icref) is advanced compared to when Idref = 0, and the amount of advance Depends on the magnitudes of the d-axis current command value Idref and the q-axis current command value Iqref. As a result, the phase of the motor current also advances, so the angle at which the polarity of the current switches is also advanced, and it is necessary to compensate for the dead time in accordance with this advanced angle.

FIG. 2 shows a configuration example of the present invention, in which a dead time compensation unit 50 is provided, and the compensation value Δu from the dead time compensation unit 50 is input to the multiplication unit 201. The d-axis current command value Idref and the q-axis current command value Iqref are input to the current advance calculation unit 202, and the d-axis current command value Idrefl and the q-axis current command value Iqrefl subjected to the current advance process are input to the addition unit 203. The For example, since the angle advance amount based on the d-axis current command value Idref is Tanθ = Idref / Iqref, it is calculated as θ = atan (Idref / Iqref). Further, the rotor position (angle) θe is input to the adding unit 203, and the angular velocity ωe is calculated by the advance angle calculating unit 204, and the advanced advance amount ωel is input to the adder unit 203. The advance amount is an advance amount ωel = k · angular velocity = k · ωe with k as a coefficient, and the advance amount ωel from the advance calculation unit 204 is an angle. The addition result RS in the addition unit 203 is RS = θ + θe + ωel = atan (Idref / Iqref) + θe + k · ωe, and this addition result RS is input to the dead time compensation timing calculation unit 200 to calculate the dead time compensation timing.
The compensation time RSd for dead time compensation calculated by the dead time compensation timing calculation unit 200 is input to the multiplication unit 201, multiplied by the compensation value Δu from the dead time compensation unit 5, and the compensation value Δut subjected to timing compensation is PI controlled. The adder 205 adds the voltage command values Va, Vb, and Vc from the unit 113 to compensate for the dead time.

FIG. 3 shows an example of timing of dead time compensation, and the horizontal axis indicates the angle (electrical angle) θe, and when the phase currents A, B, and C are switched, that is, atan (Idref / Iqref) at the angle θe. ) The compensation timing RSd is switched at an angle obtained by adding + k · ωe. In this way, the dead time compensation timing calculation unit 200 outputs the compensation timing RSd of “1” or “−1”, and the dead time compensation unit 50 outputs the direction of the compensation value and the compensation amount. Is done. For this reason, the compensation value Δut is output with the determined direction and timing. That is, the compensation value Δut is output at the timing when the current direction switches between positive and negative, and the compensation direction becomes the current direction.

It is a block diagram which shows the structural example of the vector control system used as the premise of this invention. It is a block block diagram which shows the Example of this invention. It is a time chart which shows the operation example of this invention. It is a block diagram which shows the structural example of control of an electric power steering apparatus. It is a block connection diagram showing an example of the gate circuit configuration in consideration of the dead time of the inverter. It is a time chart which shows the relationship of the dead time, turn-on time, and turn-off time in switching of an inverter. It is a block diagram which shows the basic structural example of the electric power steering apparatus provided with the dead time compensation function. It is a block diagram which shows the structural example of the conventional dead time compensation circuit. It is a characteristic view which shows the example of a waveform of electric current command value Iref and actual motor current Imes. It is a figure which shows the example of a compensation value characteristic of the conventional dead time compensation.

Explanation of symbols

1, 30, 110 Motor 2 Phase compensator 3 Steering assist command value calculator 5 Differential compensator 6 Proportional calculator 7 Integral calculator 8 Motor current detection circuit 10 Torque sensor 11 Vehicle speed sensor 12 Battery 20 Motor drive circuit 21 Converter 22A , 22B, 23A, 23B Gate drive circuit 24, 25 Dead time circuit 31 Current command value calculation circuit 33 Current control circuit 35 Inverter 50 Dead time compensation unit 51 Reference model circuit 52 Hysteresis circuit 53 Polarity determination circuit 54 Fixed value setting circuit 55 Absolute Value circuit 56 Change value calculation circuit 57 Polarization application circuit 100 Vector control phase command value calculation unit 103 dq-axis current command value calculation unit 104 2-phase / 3-phase conversion unit 200 Dead time compensation timing calculation unit 201 Multiplication unit 202 Current advance Calculation unit 203 Addition unit 204 Advance calculation unit

Claims (4)

A current command value calculation unit for calculating the d-axis current command value Idref and the q-axis current command value Iqref based on the torque command value, and a dead time compensation function with a built-in dead time circuit, and a steering assist force in the steering mechanism An inverter that drives a motor that provides a motor, and a current control unit that converts the d-axis current command value Idref and the q-axis current command value Iqref into a voltage command value for each phase of the motor and inputs the voltage command value to the inverter, In the control device of the electric power steering device with a dead time compensation function for driving and controlling the motor by vector control,
Based on a dead time compensation unit that calculates a compensation value based on the torque command value, a current advance calculation unit that calculates a current advance of the d-axis current command value Idref and the q-axis current command value Iqref, and an angular velocity of the motor An advance angle calculation unit that calculates an advance angle amount, and a dead time compensation timing calculation unit that calculates a dead time compensation timing based on an addition result of the motor angle, the advance angle amount, and the current advance. ,
A control apparatus for an electric power steering apparatus, wherein dead time compensation is performed by multiplying the compensation timing from the dead time compensation timing calculation section and the compensation value and adding the result to a voltage command value of each phase of the motor. 2. The control device for an electric power steering apparatus according to claim 1, wherein the current advance calculation unit calculates a current advance in accordance with θ = atan (Idref / Iqref). 3. 2. The control device for an electric power steering apparatus according to claim 1, wherein the advance angle calculation unit calculates an advance angle amount = k · angular velocity by using k as a coefficient. 4. The control device for an electric power steering apparatus according to claim 2, wherein the compensation timing is “1” or “−1”.

Images