JP2017011856A - Controller of dynamo-electric machine - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a controller of a dynamo-electric machine ensuring resonance suppression effect, regardless of the value of voltage phase.SOLUTION: In the dq coordinate of a motor generator, the coordinate axis where the change of a current vector for the change of voltage phase is decoupled is set as the λ axis. The λ axis direction component of the current vector is referred to λ axis current Iλr. A controller 30 operates the voltage phase φ so as to feedback control the torque Te of the motor generator to a command torque Trq*. Furthermore, the controller 30 corrects the voltage phase φ based on the λ axis current Iλr, so as to reduce the gain near the electric angular velocity ω, for the frequency characteristics related to the gain of a transfer function where the voltage phase φ is the input and the q axis current Iqr is the output.SELECTED DRAWING: Figure 2

Description

  The present invention relates to a control device applied to a rotating electrical machine electrically connected to a power conversion circuit.

  As this type of control device, as can be seen in Patent Document 1 below, the frequency characteristic relating to the gain of the transfer function that takes as input the voltage phase that is the phase of the voltage vector of the power conversion circuit and outputs the current that flows through the rotating electrical machine. Is known that reduces the gain near the electrical angular velocity of a rotating electrical machine. Specifically, this transfer function has a characteristic of resonating near the electrical angular velocity. The control device removes the low-frequency component of the q-axis current flowing in the synchronous machine as the rotating electrical machine by a high-pass filter, and feedback-controls the torque of the rotating electrical machine to the command torque based on the q-axis current from which the low-frequency component is removed. For correcting the voltage phase. As a result, the gain near the electrical angular velocity is reduced.

JP, 2014-200168, A

  Here, the control device described in Patent Document 1 can reduce the gain near the electrical angular velocity of the transfer function when the voltage phase is within a specific range, but when the voltage phase is outside the specific range. There is a concern that the gain in the vicinity of the electrical angular velocity of the transfer function cannot be reduced and the resonance suppression effect cannot be obtained. As a result, there is a concern that the feedback gain in the feedback control cannot be set large, and the responsiveness in the feedback control cannot be improved.

  The present invention has been made to solve the above problems, and an object of the present invention is to provide a control device for a rotating electrical machine capable of obtaining a resonance suppression effect regardless of the value of the voltage phase.

  Hereinafter, means for solving the above-described problems and the operation and effects thereof will be described.

  The present invention uses the torque of the rotating electrical machine (10) or a parameter having a positive correlation with the torque as a control amount, and operates the voltage vector of the power conversion circuit (20) electrically connected to the rotating electrical machine. In the control device (30) for the rotating electrical machine that controls the control amount, the change in the current vector flowing through the rotating electrical machine with respect to the change in the voltage phase that is the phase of the voltage vector in the rotating coordinate system of the rotating electrical machine is made non-interfering. A phase operation unit (30l) for operating the voltage phase so that the coordinate axis is a non-interference axis, the non-interference axis direction component of the current vector is a non-interference current, and the control amount is feedback controlled to its command value. And a frequency characteristic related to a gain of a transfer function having the voltage phase as an input and the current flowing through the rotating electrical machine or the torque of the rotating electrical machine as an output. A phase correction unit (30k) that corrects the voltage phase operated by the phase operation unit based on the non-interference current so as to reduce a gain in the vicinity of the electrical angular velocity of the rotating electrical machine. Features.

  By using the non-interfering current for correcting the voltage phase operated by the phase operating unit, the real part of the pole of the transfer function can be set to a negative value regardless of the voltage phase value. For this reason, by using a non-interference current for correcting the voltage phase, a resonance suppression effect can be obtained regardless of the value of the voltage phase.

  Therefore, in the above invention, the voltage phase operated by the phase operation unit is corrected based on the non-interference current so as to reduce the gain near the electrical angular velocity of the transfer function. Therefore, a resonance suppression effect can be obtained regardless of the voltage phase value. Thereby, the feedback gain in feedback control can be set large, and the responsiveness in feedback control can be improved.

1 is an overall configuration diagram of a motor control system according to a first embodiment. The block diagram of motor control. The figure which shows the relationship between dq coordinate system, pl coordinate system, and (lambda) o coordinate system. The figure which shows the gain characteristic of a motor transfer function when resonance is large. The figure which shows the motor model of a dq coordinate system. The figure which shows the motor model of pq coordinate system and (lambda) o coordinate system. The figure which shows the locus | trajectory of the pole on a complex plane. The figure for demonstrating the change of the current vector accompanying the change of a voltage vector. The figure for demonstrating (lambda) axis | shaft. The figure which shows the gain characteristic of the motor transfer function at the time of reducing a gain. The time chart which shows the step response at the time of reducing a gain. The figure which shows the calculation method of (lambda) axis current. The figure which shows the effect of 1st Embodiment. The figure which shows the effect of 1st Embodiment. The block diagram of the motor control concerning 2nd Embodiment. The block diagram of the motor control concerning 3rd Embodiment.

(First embodiment)
Hereinafter, a first embodiment in which a control device according to the present invention is applied to a vehicle (for example, an electric vehicle or a hybrid vehicle) including a three-phase rotating electrical machine as an in-vehicle main unit will be described with reference to the drawings.

  As shown in FIG. 1, the motor control system includes a motor generator 10, a three-phase inverter 20 as a “power conversion circuit”, and a control device 30 that controls the motor generator 10. In the present embodiment, the motor generator 10 is an in-vehicle main machine and is connected to drive wheels (not shown). In the present embodiment, a synchronous machine is used as the motor generator 10, and specifically, a permanent magnet type synchronous machine is used. In this embodiment, an IPMSM that is a salient pole machine is used as the motor generator 10.

  The motor generator 10 is connected to a battery 21 as a DC power source via an inverter 20. The output voltage of the battery 21 is, for example, 100 V or more. A smoothing capacitor 22 is provided between the battery 21 and the inverter 20.

  The inverter 20 includes three sets of serially connected bodies of upper arm switches Sup, Svp, Swp and lower arm switches Sun, Svn, Swn. The U phase of the motor generator 10 is connected to a connection point between the U phase upper and lower arm switches Sup and Sun. The V phase of the motor generator 10 is connected to a connection point between the V phase upper and lower arm switches Svp and Svn. The W phase of the motor generator 10 is connected to a connection point between the W phase upper and lower arm switches Swp and Swn. Incidentally, in this embodiment, a voltage-controlled semiconductor switching element is used as each switch Sup, Sun, Svp, Svn, Swp, Swn, and more specifically, an IGBT is used. The free wheel diodes Dup, Dun, Dvp, Dvn, Dwp, Dwn are connected in antiparallel to the switches Sup, Sun, Svp, Svn, Swp, Swn.

  The motor control system further includes a phase current detection unit that detects a current of at least two phases among the phase currents flowing through the motor generator 10. In the present embodiment, the phase current detection unit includes a V-phase current sensor 23V that detects a current flowing in the V-phase of the motor generator 10 and a W-phase current sensor 23W that detects a current flowing in the W-phase. The motor control system includes a voltage sensor 24 and a rotation angle sensor 25. The voltage sensor 24 is a voltage detection unit that detects a power supply voltage of the inverter 20, that is, a DC voltage output from the battery 21. The rotation angle sensor 25 is an angle detection unit that detects the rotation angle (electrical angle θ) of the motor generator 10, and for example, a resolver can be used.

  The control device 30 is composed mainly of a microcomputer, and operates the inverter 20 to feedback control the control amount (torque in this embodiment) of the motor generator 10 to its command value (hereinafter referred to as “command torque Trq *”). . Specifically, the control device 30 controls each switch Sup, Sun, Svp, Svn, Swp, Swn that constitutes the inverter 20 based on the detection values of the various sensors to turn on / off the switches Sup, Sun, Svp, Svn, Swp, Swn. , Gvn, gwp, and gwn are generated, and the generated operation signals gup, gun, gvp, gvn, gwp, and gwn are output to each drive circuit Dr (gate drive circuit) corresponding to each switch. Here, the upper arm side operation signals gup, gvp, gwp and the corresponding lower arm side operation signals gun, gvn, gwn are complementary to each other. That is, the upper arm switch and the corresponding lower arm switch are alternately turned on. The command torque Trq * is, for example, a control device provided outside the control device 30, and is output from a control device higher than the control device 30.

  Next, torque control of the motor generator 10 executed by the control device 30 will be described using FIG.

  The two-phase conversion unit 30a is based on the V-phase current IV detected by the V-phase current sensor 23V, the W-phase current IW detected by the W-phase current sensor 23W, and the electrical angle θ detected by the rotation angle sensor 25. The U-phase current IU, V-phase current IV, and W-phase current IW in the three-phase fixed coordinate system are converted into the d-axis current Idr and the q-axis current Iqr in the two-phase rotational coordinate system (dq coordinate system). The dq coordinate system is an orthogonal two-dimensional rotational coordinate system that rotates at the electrical angular velocity ω of the motor generator 10.

  Torque estimator 30b calculates estimated torque Te of motor generator 10 based on d and q axis currents Idr and Iqr output from two-phase converter 30a. Here, the estimated torque Te may be calculated using a map in which the d-axis current Idr and the q-axis current Iqr are associated with the estimated torque Te, or may be calculated using a model formula.

  The torque deviation calculation unit 30c calculates the torque deviation ΔT by subtracting the estimated torque Te from the command torque Trq *. Note that the estimated torque Te input to the torque deviation calculating unit 30c may be subjected to a low-pass filter process for removing high frequency components.

  Phase setting unit 30d sets voltage phase φ, which is the phase of voltage vector Vnvt of inverter 20, as an operation amount for feedback control of estimated torque Te to command torque Trq *. Specifically, the p-axis voltage Vp is calculated by proportional-integral control with the torque deviation ΔT as an input. In the present embodiment, the voltage phase φ is defined with the positive direction of the d-axis as a reference, and the counterclockwise direction from this reference (the direction rotating from the positive direction of the d-axis to the positive direction of the q-axis) is defined as the positive direction. Has been.

  Here, the transfer function of the motor generator 10 to be controlled by the control device 30 has a resonance characteristic. Specifically, in a frequency characteristic (hereinafter referred to as “gain characteristic”) relating to a gain (gain) of a transfer function having a voltage phase φ as an input and a q-axis current as an output, a gain peak exists near the electrical angular velocity. In this case, there is a concern that the torque controllability of the motor generator 10 may be reduced. For this reason, in this embodiment, in order to reduce this peak, each processing unit such as the p-axis correction amount calculation unit 30j and the p-axis correction unit 30k is provided. Hereinafter, after describing the principle of reducing the peak near the electrical angular velocity, each of the processing units will be described.

  First, the principle of reducing the gain peak will be described.

  The voltage equation of the permanent magnet type synchronous machine is expressed by the following equation (eq1).

In the above equation (eq1), “p” represents a differential operator, “R” represents an armature winding resistance, “Ld” and “Lq” represent d and q axis inductances, and “ω” represents a motor. The electrical angular velocity of the generator 10 is indicated, and “ψ” indicates the effective value of the armature interlinkage magnetic flux of the permanent magnet. “Vd” indicates a d-axis voltage, and “Vq” indicates a q-axis voltage. A voltage amplitude Vn that is an amplitude of the voltage vector Vnvt of the inverter 20 is defined as the square root of the sum of the square value of the d-axis voltage Vd and the square value of the q-axis voltage Vq.

  When the d and q axis voltages Vd and Vp change slightly, the above equation (eq1) is expressed as the following equation (eq2).

In the above equation (eq2), “Δvd” indicates the d-axis voltage change amount, and “Δvq” indicates the q-axis voltage change amount. Further, “Δid” indicates a d-axis current change amount associated with a change in the d-axis voltage Vd, and “Δiq” indicates a q-axis current change amount associated with a change in the q-axis voltage Vq. When the above equation (eq1) is subtracted from the above equation (eq2), the following equation (eq3) is derived.

Solving the above equation (eq3) for d and q-axis current changes Δid and Δiq, the following equation (eq4) is derived.

Here, a pl coordinate system which is a biaxial orthogonal coordinate system is introduced. As shown in FIG. 3, the pl coordinate system is a coordinate system including an l-axis extending from the origin 0 in a direction parallel to the voltage vector Vnvt and a p-axis extending from the origin 0 in a direction orthogonal to the voltage vector Vnvt. The pl coordinate system is a coordinate system obtained by rotating the dq coordinate system about the origin 0 counterclockwise by an angle η (= φ−π / 2). Here, increasing or decreasing the l-axis voltage Vl, which is a component obtained by projecting the voltage vector Vnvt on the l-axis, is equivalent to increasing or decreasing the voltage amplitude Vn. Further, increasing or decreasing the p-axis voltage Vp, which is a component obtained by projecting the voltage vector Vnvt on the p-axis, is equivalent to increasing or decreasing the voltage phase φ.

  Since the pl coordinate system is obtained by rotating the dq coordinate system counterclockwise by an angle η, the following equation (eq5) is established.

In the above equation (eq5), “Δvp” represents the p-axis voltage change amount, and “Δvl” represents the l-axis voltage change amount. Substituting the above equation (eq5) into the above equation (eq4) yields the following equation (eq6).

Here, when the voltage phase φ changes by a minute amount Δφ, the p-axis voltage change amount Δvp is expressed by the following equation (eq7) using the relationship of “sin Δφ≈Δφ”.

From the above equations (eq6) and (eq7), the following equation (eq8) is derived.

As shown in the above equation (eq8), the rational transfer function having the voltage phase φ as an input and the q-axis current Iq that greatly contributes to torque as an output is a complex function of s and is a second-order lag system. . FIG. 4 shows the gain characteristic of this transfer function. The gain characteristic shown in FIG. 4 has a resonance angular velocity near the electrical angular velocity ω, and the gain at the resonance angular velocity is large. For this reason, the feedback gain that is the proportional gain and integral gain of the phase setting unit 30d cannot be set large, and the torque response in the feedback control cannot be increased.

  Here, as shown in FIG. 5, a motor model having d and q-axis voltages Vd and Vq as inputs and d and q-axis currents Id and Iq as outputs is referred to as MODEL1. Further, as shown in FIG. 6, a motor model having p and l-axis voltages Vp and Vl as inputs and λ and o-axis currents Iλ and Io as outputs is referred to as MODEL2. As shown in FIG. 3, the λo coordinate system is a coordinate system obtained by rotating the dq coordinate system about the origin 0 counterclockwise by an angle λ. In order to reduce the gain at the resonance angular velocity, first, consider converting MODEL1 to MODEL2.

  Using the following equation (eq9), the d and q-axis voltage changes Δvd and Δvq are converted into p and l-axis voltage changes Δvp and Δvl in the pl coordinate system.

The following equation (eq10) is used when λ, o-axis current change amounts Δiλ, Δio in the λo coordinate system obtained by rotating the dq coordinate system counterclockwise about the origin 0 by the angle λ are projected on the dq coordinate system. .

Substituting the above equations (eq3) and (eq10) into the above equation (eq9) leads to the following equation (eq11).

However, in the above equation (eq11), “Rs, Rc, Lλo, Loλ” are parameters shown in the following equation (eq12).

Here, as shown in FIG. 6, in order to correct the p-axis voltage Vp input to the motor model MODEL2, a p-axis correction amount calculation unit 30j is introduced. In the present embodiment, the p-axis correction amount calculation unit 30j calculates the phase correction amount Δr as a value obtained by multiplying the λ-axis current Iλ by a proportional gain K (> 0). The phase correction amount Δr using the λ-axis current change amount Δiλ as an input is “K × Δiλ”. When the phase correction amount Δr is added to the left side of the above equation (eq11), the following equation (eq13) is derived.

The above equation (eq13) is transformed into the following equation (eq14).

Solving the above equation (eq14) with respect to λ and o-axis current change amounts Δiλ and Δio, the following equation (eq15) is derived.

Here, in the motor model MODEL2 of FIG. 6, the pole of the transfer function having the p-axis voltage Vp as an input and the λ-axis current Iλ as an output is obtained as s where the denominator on the right side of the above equation (eq15) is zero. s is obtained by the following equation (eq16).

When the above equation (eq16) is arranged for s, the following equation (eq17) is derived.

The solution of the above equation (eq17), that is, the pole of the transfer function is represented by the following equation (eq18) as a complex conjugate value.

Substituting “Lo, Loλ” of the above equation (eq12) into the above equation (eq18), the following equation (eq19) is derived.

In the above formula (eq19), “j” represents an imaginary number. Here, when the following expression (eq20) holds, the above expression (eq19) becomes the following expression (eq21).

When the above equation (eq20) is solved for the angle λ, the following equation (eq22) is derived.

As shown in FIG. 7, the arrangement of the poles of the transfer function derived by substituting the above equation (eq22) for the angle λ of the above equation (eq21) is proportional gain K (that is, the absolute value of the phase correction amount Δr). The trajectory of the pole when is increased from 0 exists in a region where the real part of the pole has a negative value on the complex plane. Further, the arrangement of the poles of the transfer function derived by substituting the above equation (eq22) into the angle λ of the above equation (eq21) indicates that the locus of the pole when the proportional gain K is increased from 0 is the voltage phase φ It will be along the real axis on the complex plane regardless of the value of. Specifically, regardless of the value of the voltage phase φ, as the proportional gain K increases, the real part of the pole moves in the negative direction so as to maintain the value of the imaginary part of the pole. Therefore, by correcting the p-axis voltage Vp with the phase correction amount Δr, the gain near the electrical angular velocity ω in the gain characteristic can be reduced regardless of the value of the voltage phase φ.

  Here, the λ axis that defines the λ-axis current Iλ used for calculating the phase correction amount Δr is a coordinate axis perpendicular to the direction in which the current vector Invt changes when the current voltage phase φ changes by a minute amount Δφ. Corresponds to a non-interfering axis. This can be explained by the relationship between the minute amount Δφ and the λ-axis current Iλ expressed by the following equations (eq23) and (eq24) from the above equations (eq7) and (eq15). In the following equations (eq23) and (eq24), the influence of the resistance R is ignored because it is small, and the term of the differential operator s is excluded to show a steady relationship.

From the above equations (eq12) and (eq20), “Loλ = 0”. Therefore, when the voltage phase changes by a minute amount Δφ, the λ-axis current Iλ does not change and changes in the direction perpendicular to the λ-axis. From this, it can be said that the λ axis is a non-interference axis of the minute amount Δφ.

  FIG. 8 shows the voltage vector Vnvt and the current vector Invt in the dq coordinate system. The current vector is defined as the square root of the sum of the square value of the d-axis current and the square value of the q-axis current. In FIG. 8, the change amount of the current vector Invt when the voltage phase φ is changed by the minute amount Δφ is indicated by “ΔIφ”, and the change amount of the current vector Invt when the voltage amplitude Vn is changed by the minute amount ΔVn. This is indicated by “ΔIvn”. FIG. 9 shows the λ axis extending in the direction orthogonal to the change ΔIφ of the current vector Invt.

  The fact that the gain is reduced by using the phase correction amount Δr will be described using a transfer function. Substituting the above equation (eq15) into the above equation (eq10) leads to the following equation (eq25).

The transfer function of the following equation (eq26) is derived from the above equation (eq25).

Substituting the above equation (eq7) into the above equation (eq26) leads to the following equation (eq27).

The above equation (eq27) represents a transfer function having the voltage phase φ as an input and the q-axis current Iq as an output. This transfer function is obtained by adding a term including a proportional gain K to the denominator of the transfer function of the above equation (eq8). Due to the existence of the term including the proportional gain K, the gain at the resonance angular velocity can be reduced as the proportional gain K is increased as shown in FIG. FIG. 11 shows a torque waveform when the command torque Trq * is changed stepwise in the torque feedback control with the voltage phase φ. By increasing the proportional gain K, the stability of torque control can be improved.

  Next, returning to FIG. 2, each processing unit such as the p-axis correction amount calculating unit 30j for reducing the gain near the electrical angular velocity in the gain characteristic will be described.

  The speed calculation unit 30e calculates the electrical angular speed ω of the motor generator 10 based on the electrical angle θ.

  The command voltage setting unit 30f receives the command torque Trq * and calculates a normalized voltage amplitude “Vl / ω”. In the present embodiment, the normalized voltage amplitude “Vl / ω” is a value obtained by dividing the l-axis voltage Vl by the electrical angular velocity ω. Note that the normalized voltage amplitude may be calculated using a map in which the command torque Trq * and the normalized voltage amplitude are associated with each other.

  The speed multiplication unit 30g calculates the l-axis voltage Vl by multiplying the normalized voltage amplitude “Vl / ω” by the electrical angular speed ω. The l-axis voltage Vl is an operation amount for performing feedforward control of the torque of the motor generator 10 to the command torque Trq *.

  The λ-axis setting unit 30h (corresponding to “non-interference axis setting unit”) calculates the voltage phase φ based on the relationship “φ = η + π / 2”. In the present embodiment, the λ-axis setting unit 30h calculates the voltage phase φ based on the relationship “φ = η + π / 2”. In particular, in the present embodiment, the angle η is subjected to a low-pass filter process in order to remove a noise component of a calculated value of the angle η formed by the d axis and the p axis. For this reason, the voltage phase φ is calculated by the following equation (eq28).

In the above equation (eq28), the l-axis voltage Vl output from the speed multiplication unit 30g and the corrected p-axis voltage Vpc output from the p-axis correction unit 30k described later are used. In the above equation (eq28), the first term on the right side corresponds to the change in the angle η between the d-axis and the p-axis during the period from the previous processing cycle to the current processing cycle. The angle η of the second term on the right side of the above equation (eq28) is “π / 2” from the voltage phase φ calculated in the current processing cycle for each processing cycle in order to calculate the voltage phase φ in the next processing cycle. ”Is updated to the subtracted value. That is, the previously updated angle η is used as the angle η of the second term on the right side of the above equation (eq28). By the arctangent calculation of the above equation (eq28), for example, the first term on the right side of the above equation (eq28) can be calculated between “−π to + π”. Particularly in the present embodiment, when the denominator in the parenthesis of the first term on the right side is 0 and the numerator is a positive value, the value of the first term on the right side is calculated as “π / 2”. On the other hand, when the denominator in the parenthesis of the first term on the right side is 0 and the numerator is a negative value, the value of the first term on the right side is calculated as “−π / 2”. The λ-axis setting unit 30h calculates the angle λ based on the above equation (eq22) using the calculated voltage phase φ as an input.

  Incidentally, the calculation method of the voltage phase φ is not limited to the method described above. For example, the λ-axis setting unit 30h may calculate the voltage phase φ as an operation amount for performing feedback control of the estimated torque Te to the command torque Trq * based on the torque deviation ΔT.

  The λ-axis actual current calculation unit 30i (corresponding to the “non-interference current calculation unit”) includes the d and q-axis currents Idr and Iqr output from the two-phase conversion unit 30a and the angle λ output from the λ-axis setting unit 30h. Are input to the following equation (eq29) to calculate the λ-axis current Iλr (see FIG. 12).

Since the λ axis changes with the change of the driving state of the motor generator 10, the λ axis current Iλr also changes with the change of the driving state of the motor generator 10.

  The p-axis correction amount calculation unit 30j (corresponding to the “phase correction amount calculation unit”) calculates the phase correction amount Δr by multiplying the λ-axis current Iλr by the proportional gain K.

  The p-axis correction unit 30k (corresponding to “phase correction unit”) calculates the corrected p-axis voltage Vpc by subtracting the phase correction amount Δr from the p-axis voltage Vp. That is, the p-axis correction unit 30k adds a correction amount obtained by inverting the sign of the λ-axis current Iλr to the p-axis voltage Vp.

  The operation signal generation unit 30l (corresponding to the “phase operation unit”) is configured to output each operation signal “gup,” based on the corrected p-axis voltage Vpc, the l-axis voltage Vl, and the power supply voltage VINV detected by the voltage sensor 44. Gun, gvp, gvn, gwp, and gwn are generated and output to each drive circuit Dr. In the present embodiment, each operation signal is generated as follows.

  The operation signal generation unit 301 first calculates the voltage phase φ. Here, the voltage phase φ may be calculated by a method similar to the method of calculating the voltage phase φ in the λ-axis setting unit 30h, for example. Then, the operation signal generation unit 301 calculates a sinusoidal three-phase command voltage whose phases are shifted from each other by 120 degrees in electrical angle based on the calculated voltage phase φ and the l-axis voltage Vl equal to the voltage amplitude Vn. To do. Then, the operation signal generation unit 30l performs each operation signal “gup”, “gun”, “gvp”, “gvn” by PWM control based on a magnitude comparison between a signal obtained by standardizing the three-phase command voltage with the power supply voltage VINV and a carrier signal (for example, a triangular wave signal). , Gwp, gwn.

  Subsequently, the effects of the present embodiment will be described with reference to FIGS. 13 and 14.

  FIG. 13 shows the locus of the poles when the proportional gain K is increased from 0 in each case where the voltage phase φ is set to various values. Here, FIG. 13A shows a locus in the present embodiment as shown in FIG. FIGS. 13B and 13C show the trajectories of the poles in the related technology for comparison with the present embodiment. Specifically, FIG. 13B shows a locus of poles when the phase correction amount Δr is calculated as a value obtained by multiplying the d-axis current Idr by the proportional gain K. The complex conjugate pole shown in FIG. 13B is expressed by the following equation (eq30) by setting “λ = 0” in the above equation (eq19).

FIG. 13C shows a locus of poles when the phase correction amount Δr is calculated as a value obtained by multiplying the q-axis current Iqr by the proportional gain K. The complex conjugate pole shown in FIG. 13C is expressed by the following equation (eq31) by setting “λ = π / 2” in the above equation (eq19).

As shown in FIGS. 13B and 13C, when the d and q axis currents Idr and Iqr are used for calculating the phase correction amount Δr, the proportional gain K is increased from 0 according to the value of the voltage phase φ. The trajectory of the poles is different. For example, in FIGS. 13B and 13C, when the voltage phase φ is −30 °, the real part of the pole can be a positive value. For this reason, in the configurations shown in FIGS. 13B and 13C, depending on the value of the voltage phase φ, the gain reduction effect of the resonance angular velocity in the gain characteristics is small, or the torque control is rather unstable.

  On the other hand, according to the present embodiment, as shown in FIG. 13A, the locus of the pole when the proportional gain K is increased from 0 is on the complex plane regardless of the value of the voltage phase φ. It exists in the region where the real part of the pole is negative. Further, according to the present embodiment, the locus of the poles when the proportional gain K is increased from 0 is substantially parallel to the real axis on the complex plane regardless of the value of the voltage phase φ. For this reason, the gain at the resonance angular velocity can be reduced.

  Next, FIG. 14 shows each characteristic corresponding to each of FIGS. 13 (a), (b), and (c). Specifically, the gain characteristic of the transfer function that receives the voltage phase φ and outputs the estimated torque Te, and the gain characteristic of the closed-loop transfer function of the feedback control system that receives the command torque Trq * and outputs the estimated torque Te The step response of the q-axis current Iq when the p-axis voltage Vp is changed stepwise is shown. Here, in each characteristic, the voltage phase φ was set to 30 °, 90 °, and 150 °.

  As shown in FIG. 14A, according to the present embodiment in which the λ-axis current Iλr is used to calculate the phase correction amount Δr, the gain characteristic of the transfer function that receives the voltage phase φ and outputs the estimated torque Te is used. Resonance angular velocity gain can be reduced. For this reason, the gain of the resonance angular velocity in the gain characteristic of the closed-loop transfer function can be reduced, and the step response of the q-axis current is stable without being oscillated. Therefore, the feedback gain in the phase setting unit 30d can be set large, and the torque response in feedback control can be improved.

  On the other hand, as shown in FIG. 14B, in the related technique using the d-axis current Idr for calculating the phase correction amount Δr, the gain characteristic of the transfer function having the voltage phase φ as an input and the estimated torque Te as an output. The resonance angular velocity gain at is increased when the voltage phase φ is 30 °. For this reason, the gain of the resonance angular velocity in the gain characteristic of the closed loop transfer function also increases, and the step response when the voltage phase φ is 30 ° becomes oscillatory. Therefore, the feedback gain in the phase setting unit 30d cannot be set large, and the torque response in feedback control cannot be improved.

  As shown in FIG. 14C, the related art using the q-axis current Iqr to calculate the phase correction amount Δr also increases the resonance angular velocity gain in the gain characteristic of the closed-loop transfer function depending on the value of the voltage phase φ. In addition, the step response becomes oscillatory. In addition, in the figure which shows a step response among FIG.14 (c), since it becomes complicated, description of the value of voltage phase (phi) is abbreviate | omitted.

  According to the present embodiment described above, the gain of the resonance angular velocity in the gain characteristic can be reduced regardless of the value of the voltage phase φ. For this reason, the feedback gain in the phase setting unit 30d can be set large, and as a result, the torque response of the motor generator 10 in the feedback control can be improved.

  Further, according to the present embodiment, the phase correction amount Δr can be calculated by a simple method such as multiplying the λ-axis current Iλr by the proportional gain K.

(Second Embodiment)
Hereinafter, the second embodiment will be described with reference to the drawings with a focus on differences from the first embodiment.

  FIG. 15 shows a block diagram of torque control according to the present embodiment. In FIG. 15, the same processes as those shown in FIG. 2 are denoted by the same reference numerals for the sake of convenience.

  In the present embodiment, control based on the p-axis voltage Vp for feedback control of the estimated torque Te to the command torque Trq * is referred to as phase control. In addition, control based on the l-axis voltage Vl for feedback control of the λ-axis current Iλr to the λ-axis command current Iλ * is referred to as amplitude control.

  In the present embodiment, the control device 30 includes an l-axis correction amount calculation unit 32 that corrects the l-axis voltage Vl. The l-axis correction amount calculation unit 32 is for suppressing interference from phase control to amplitude control. Specifically, as shown in FIGS. 8 and 9, the λ axis is a coordinate axis in the direction in which the change amount of the current vector Invt becomes 0 when the voltage phase φ changes slightly. Therefore, of the current vector change ΔIvn when the voltage amplitude Vn changes by a minute amount ΔVn, the λ-axis component obtained by mapping the change ΔIvn to the λ-axis is not affected by the change in the voltage phase φ. It is. Therefore, the interference from the phase control to the amplitude control can be suppressed by using the λ-axis current Iλ to correct the l-axis voltage Vl.

  Based on the above, specific processing of the l-axis correction amount calculation unit 32 will be described with reference to FIG.

  The command current setting unit 32a sets d and q-axis command currents Id * and Iq * for realizing the command torque Trq * based on the command torque Trq *. In the present embodiment, currents for realizing minimum current / maximum torque control are set as d and q-axis command currents Id * and Iq *.

  The λ-axis command current calculation unit 32b calculates the following equation (eq32) based on the command currents Id * and Iq * output from the command current setting unit 32a and the angle λ output from the λ-axis setting unit 30h. First, the λ-axis command current Iλ * is calculated.

In the present embodiment, the λ-axis setting unit 30h calculates the voltage phase φ based on the following equation (eq33).

In the above equation (eq33), the corrected p-axis voltage Vpc and the corrected l-axis voltage Vlc output from the l-axis correction unit 30n described later are used.

  The λ-axis current deviation calculation unit 32c calculates the λ-axis current deviation ΔIλ by subtracting the λ-axis current Iλr from the λ-axis command current Iλ *. Note that the λ-axis current Iλr input to the λ-axis current deviation calculation unit 32c may be subjected to low-pass filter processing for removing high frequency components.

  Based on the λ-axis current deviation ΔIλ, the amplitude correction amount calculation unit 32d calculates an amplitude correction amount Δl as an operation amount for performing feedback control of the λ-axis current Iλr to the λ-axis command current Iλ *. Specifically, the amplitude correction amount Δl is calculated by proportional-integral control using the λ-axis current deviation ΔIλ as an input.

  The l-axis correction unit 30n calculates the corrected l-axis voltage Vlc by adding the amplitude correction amount Δl calculated by the amplitude correction amount calculation unit 32d to the l-axis voltage Vl output from the speed multiplication unit 30g. .

  The operation signal generation unit 301 (corresponding to “phase operation unit” and “amplitude operation unit”) generates each operation signal gup based on the corrected p-axis voltage Vpc, the corrected l-axis voltage Vlc, and the power supply voltage VINV. , Gun, gvp, gvn, gwp, gwn are generated and output to each drive circuit Dr.

  According to the present embodiment described above, since interference from phase control to amplitude control can be suppressed, the feedback gain in the phase setting unit 30d and the amplitude correction amount calculation unit 32d can be set large. For this reason, the torque response of the motor generator 10 can be further improved.

(Third embodiment)
Hereinafter, the third embodiment will be described with reference to the drawings with a focus on differences from the first embodiment. In the present embodiment, the phase correction amount Δr is calculated based on d, q-axis currents Idr, Iqr, and voltage phase φ instead of the λ-axis current Iλr.

  FIG. 16 shows a block diagram of torque control according to the present embodiment. In FIG. 16, the same processes as those shown in FIG. 2 are given the same reference numerals for the sake of convenience.

  The phase calculation unit 30p calculates the voltage phase φ based on the l-axis voltage Vl and the corrected p-axis voltage Vpc by the same process as the λ-axis setting unit 30h of the second embodiment.

  The p-axis correction amount calculation unit 30q receives the voltage phase φ and the d and q-axis currents Idr and Iqr as inputs, and calculates the phase correction amount Δr based on the following equation (eq34).

The p-axis correction amount calculation unit 30q variably sets each of the d-axis coefficient Kd and the q-axis coefficient Kq in the above equation (eq34) based on the voltage phase φ. The phase correction amount Δr proportional to the λ-axis current Iλr is calculated by combining a value obtained by multiplying the d-axis current Idr by the d-axis coefficient Kd and a value obtained by multiplying the q-axis current Iqr by the q-axis coefficient Kq. Because. Specifically, for example, the locus of the pole when the voltage phase φ is 150 ° in FIG. 13B and the locus of the pole when the voltage phase φ is 150 ° in FIG. By combining the ratios, it is possible to obtain a polar trajectory along the real axis shown in FIG. Hereinafter, a method for setting the d and q axis coefficients Kd and Kq according to the present embodiment will be described.

  When the voltage phase φ is 90 °, the trajectory of the pole when the voltage phase φ is 90 ° in FIG. 13B and the trajectory of the pole shown in FIG. Therefore, when the voltage phase φ is 90 °, the d-axis coefficient Kd is set to the first specified value K1, which is a positive value, and the q-axis coefficient Kq is set to 0.

  When the voltage phase φ is 180 °, the locus of the poles when the voltage phase φ is 180 ° in FIG. 13C is substantially the same as the locus of the poles shown in FIG. Therefore, when the voltage phase φ is 180 °, the d-axis coefficient Kd is set to 0, and the q-axis coefficient Kq is set to the second specified value K2 that is a positive value.

  When the voltage phase φ is 0 °, the locus when the sign of the proportional gain K is made negative with respect to the locus of the pole when the voltage phase φ is 0 ° in FIG. 13C, and FIG. The traces of the poles shown substantially coincide. Therefore, when the voltage phase φ is 0 °, the d-axis coefficient Kd is set to 0, and the q-axis coefficient Kq is set to a third specified value K3 that is a negative value.

  According to the values of the d and q axis coefficients Kd and Kq described above when the voltage phase φ is 0 °, 90 °, and 180 °, when the voltage phase φ is 0 ° to 90 °, the larger the voltage phase φ, The d-axis coefficient Kd is set to increase from 0 toward the first specified value K1, and the q-axis coefficient Kq is set to increase from the third specified value K3 to 0 as the voltage phase φ increases. When the voltage phase φ is 90 ° to 180 °, the d-axis coefficient Kd is set to decrease from the first specified value K1 toward 0 as the voltage phase φ increases, and the q-axis coefficient increases as the voltage phase φ increases. Kq is set larger from 0 toward the second specified value K2.

  The phase correction amount Δr calculated by the p-axis correction amount calculation unit 30q is input to the p-axis correction unit 30k.

  Also according to the present embodiment described above, the gain of the resonance angular velocity in the gain characteristic can be reduced regardless of the value of the voltage phase φ, and the torque response in the feedback control can be improved.

(Other embodiments)
Each of the above embodiments may be modified as follows.

  In the first embodiment, the λ axis is defined as a coordinate axis perpendicular to the direction in which the current vector Invt changes when the current voltage phase φ changes by a minute amount Δφ, but is not limited thereto. For example, the λ axis may be defined as a coordinate axis in a direction in which the change in the current vector is 0 when the current voltage phase φ is slightly changed.

  In each of the above embodiments, the feedback control in the phase setting unit 30d may be performed, for example, only by integral control or by proportional integral derivative control. In the second embodiment, the feedback control in the amplitude correction amount calculation unit 32d may be performed only by integral control, or may be performed by proportional integral derivative control.

  In the first embodiment, the voltage amplitude may not be variable according to the command torque Trq * and the electrical angular velocity ω, but may be a fixed value. That is, there is no need for feedforward control of voltage amplitude. Here, as the fixed value, for example, an output upper limit value of the inverter 20 may be used for the purpose of using the inverter 20 utilization range to the maximum, or a value less than the output upper limit value for the purpose of reducing loss and noise. May be used. As a value less than the above output upper limit value, for example, an output voltage at which the output voltage of the inverter 20 becomes 1 pulse or more (“High-efficiency motor control by harmonic modulation type pulse-saving drive: Kimihisa Furukawa et al., 7 others, Heisei 22 Annual Electrical Society Industrial Application Division Conference, 1-134, pp. I-627 to I-632 ”) or the maximum output value in the linear region of the inverter 20 may be used.

  A configuration using the amplitude correction amount Δl described in the second embodiment may be applied to the third embodiment.

  When the difference between the d-axis inductance Ld and the q-axis inductance Lq is small, “Ld / Lq” is a value close to 1. Therefore, in the above equation (eq22), the λ axis can be obtained based only on the voltage phase φ. When motor generator 10 is SPMSM, “Ld / Lq” is 1, so that the λ axis can be obtained based only on voltage phase φ.

  As the voltage phase φ for obtaining the λ axis in the above equation (eq22), a value obtained by adding “π / 2” to the angle η may be used instead of the value given by the above equation (eq28). A value in which the voltage phase φ corresponding to the angular velocity ω and the command torque Trq * is recorded in advance may be used.

  The q-axis current has a positive correlation with the motor generator 10 torque. Therefore, the transfer function of the above equation (eq8) of the first embodiment can be replaced with a transfer function having the voltage phase φ as an input and torque as an output.

  The method for generating the operation signal is not limited to the one based on the PWM control, and for example, the method described below may be used. For example, the control device 30 includes storage means (for example, a memory such as a ROM) in which a line voltage pattern for realizing the voltage amplitude Vn is stored in advance. In such a configuration, the line voltage pattern stored in the storage means is converted into a pulse pattern for the gate of each switching element, and the operation signal is set by setting the output timing of the converted pulse pattern based on the voltage phase φ. Generate.

  Further, the method for generating the operation signal is not limited to a method of calculating a sinusoidal three-phase command voltage whose phases are shifted from each other by 120 degrees in electrical angle. For example, a method of superposing a third-order high frequency on a sine wave may be used.

  In each of the above embodiments, the torque of the motor generator 10 is estimated and this estimated value is used for motor control. However, the present invention is not limited to this. For example, a torque detection unit that detects the torque of the motor generator 10 may be provided in the control system, and the detection value of the torque detection unit may be used for motor control.

  The rotary electric machine is not limited to IPMSM, but may be SPMSM or a wound field type synchronous machine. Here, for example, when the SPMSM is employed in the first embodiment, the torque of the rotating electrical machine is determined by the q-axis current. For this reason, the control amount may be a q-axis current instead of the torque. Further, the rotating electrical machine is not limited to being used as an in-vehicle main machine, but may be used as an in-vehicle auxiliary machine such as an electric power steering device or an electric motor constituting an air-conditioning electric compressor. In addition, the rotating electrical machine is not limited to a vehicle-mounted type.

  DESCRIPTION OF SYMBOLS 10 ... Motor generator, 20 ... Inverter, 30 ... Control apparatus.

Claims (8)

By using the torque of the rotating electrical machine (10) or a parameter having a positive correlation with the torque as a controlled variable, and operating the voltage vector of the power conversion circuit (20) electrically connected to the rotating electrical machine, the controlled variable In the control device (30) of the rotating electrical machine that controls
In the rotating coordinate system of the rotating electrical machine, the coordinate axis on which the change of the current vector flowing through the rotating electrical machine with respect to the change of the voltage phase that is the phase of the voltage vector is made non-interfering, and the non-interfering axis of the current vector The direction component is a non-interference current,
A phase operation unit (30l) for manipulating the voltage phase to feedback-control the control amount to its command value;
The non-interference so as to reduce the gain in the vicinity of the electrical angular velocity of the rotating electrical machine with respect to the frequency characteristic related to the gain of the transfer function that takes the voltage phase as an input and outputs the current flowing through the rotating electrical machine or the torque of the rotating electrical machine as an output. A control apparatus for a rotating electrical machine, comprising: a phase correction unit (30k) that corrects the voltage phase operated by the phase operation unit based on a current.   2. The control device for a rotating electrical machine according to claim 1, wherein the non-interference axis is a coordinate axis perpendicular to a direction in which the current vector changes when the current voltage phase is slightly changed.   2. The control device for a rotating electrical machine according to claim 1, wherein the non-interference axis is a coordinate axis in a direction in which a change amount of the current vector becomes 0 when the current voltage phase slightly changes. A non-interference current calculation unit (30i) for calculating the non-interference current;
A phase correction amount calculation unit (30j) that calculates a phase correction amount proportional to the non-interference current calculated by the non-interference current calculation unit,
The said phase correction part correct | amends the said voltage phase so that the gain of the said electrical angular velocity vicinity may be reduced based on the said phase correction amount calculated by the said phase correction amount calculation part. The control apparatus of the rotary electric machine as described in the item. A non-interference axis setting unit (30h) for setting the non-interference axis based on the voltage phase or the voltage phase and the inductance of the rotating electrical machine;
The non-interference current calculation unit calculates the non-interference current based on the non-interference axis set by the non-interference axis setting unit and a current flowing through the rotating electrical machine in the rotating coordinate system. The control apparatus of the rotary electric machine described. A phase correction amount calculation unit (30q) that calculates a phase correction amount proportional to the non-interference current based on the voltage phase, a d-axis current flowing in the rotating electrical machine, and a q-axis current flowing in the rotating electrical machine;
The said phase correction part correct | amends the said voltage phase so that the gain of the said electrical angular velocity vicinity may be reduced based on the said phase correction amount calculated by the said phase correction amount calculation part. The control apparatus of the rotary electric machine as described in the item.   In the phase correction unit, the locus of the pole of the transfer function when the absolute value of the phase correction amount is increased from 0 exists in a region where the real part of the pole has a negative value on a complex plane. In addition, the arrangement of the poles can be realized such that the locus of the poles when the absolute value of the phase correction amount is increased from 0 follows the real axis on the complex plane regardless of the value of the voltage phase. The control apparatus for a rotating electrical machine according to any one of claims 4 to 6, wherein the voltage phase is corrected based on the phase correction amount. A non-interference current calculation unit (30i) for calculating the non-interference current;
An amplitude operation unit (30l) that operates a voltage amplitude that is an amplitude of the voltage vector in order to feedback-control the non-interference current calculated by the non-interference current calculation unit to a command current according to the command value. The control device for a rotating electrical machine according to any one of Items 1 to 7.