JP2013219988A - Control device of rotating machine - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a control device of a rotating machine which can prevent an opportunity in which a detection value of a current sensor can be corrected from being restricted.SOLUTION: When it is determined that electric angle speed ω of a motor generator is higher than a second angle speed ωB, an error correction coefficient Kt¥ is calculated on the basis of an integration valueΔiθ¥over one electric angle cycle of a current absolute value. On the other hand, when it is determined that the electric angle speed ω is higher than zero and less than a first angle speed ωA, the error correction coefficient Kt¥ is calculated on the basis of an integration valueΔVθ¥ over one electric angle cycle of a voltage absolute value. Then, a U-phase current calculation value is corrected by multiplying the U-phase current calculation value and V-phase, W-phase current detection values by the calculated error correction coefficient Kt¥.

Description

  The present invention relates to a control device for a rotating machine that is applied to a control system for a rotating machine that includes a multiphase rotating machine that is electrically connected to an inverter and a current detection unit that detects a phase current flowing through the rotating machine.

  Conventionally, in a control system for a rotating machine comprising a multi-phase rotating machine electrically connected to an inverter and a current detector for detecting a phase current flowing through the rotating machine, the rotating machine is controlled based on a detection value of the current detector. The technology to do is known. Current detectors include current-to-voltage conversion devices such as current transformers and high-precision resistors, operational amplifiers that amplify voltages, and analog circuits such as AD converters. In an analog circuit, a phenomenon is also known in which an error (hereinafter referred to as a gain error) in which a deviation between an actual current value and a detected value of a current detector becomes large due to temperature characteristics and transient characteristics. If the rotating machine is controlled based on the detected current value including the gain error, there is a possibility that the torque fluctuation of the rotating machine due to the gain error occurs.

  In order to cope with such a problem, a technique for correcting a current detection value including a gain error has been proposed as seen in Patent Document 1 below. This technology will be described. First, among the U, V, and W phases of the three-phase rotating machine, for example, U and V at the timing when the V-phase current becomes 0 and the timing when the W-phase current becomes 0, respectively. V-phase and W-phase currents are detected, and a correction coefficient for the detection value is calculated based on the current detection values at these timings. Then, the current detection value is corrected by the calculated correction coefficient. By controlling the rotating machine based on the corrected current detection value in this way, it is possible to suppress torque fluctuation.

Japanese Patent No. 4708444

  Here, according to the technique described in Patent Document 1, although it is possible to calculate a correction coefficient capable of suppressing torque fluctuation, an operation state suitable for calculating the correction coefficient based on the operation state of the rotating machine. Is required. That is, the opportunity for correcting the current detection value is limited.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to provide a control device for a rotating machine that can avoid the restriction of the opportunity to correct the current detection value.

  In order to solve the above-mentioned problem, the invention according to claim 1 includes a multi-phase rotating machine (10) electrically connected to the inverter (INV), and current detecting means (18u) for detecting a phase current flowing through the rotating machine. , 18v, 18w), and a correction means (26u, 26v, 26w, 54u, 54v, 54w, 56u, 56v, 56w) for correcting the detection value of the current detection means, Command voltage calculation means for calculating a command voltage (V ¥ *; ¥ = u, v, w) to be applied to each phase of the rotating machine based on the detection value of the current detection means corrected by the correction means; And an operation means (36) for operating the inverter based on the detection value of the current detection means corrected by the correction means, and calculated by the detection value of the current detection means and the command voltage calculation means. At least one of the received command voltages is defined as a correction parameter, and the correction means determines the correction parameter amplitude equivalent amount (Δiθ ¥, ΔVθ ¥) or the correction parameter amplitude (amp). The detection value of the current detection means is corrected based on a comparison of values between phases of the rotating machine.

  If a gain error is included in the detection value of the current detection means, the amplitude of the detection value deviates from the actual phase current amplitude. For this reason, the amplitude of the command voltage calculated based on the detected value including the gain error also deviates from the amplitude of the command voltage when the gain error is not included. And the deviation of the amplitude of the detected value or command voltage tends to increase as the gain error increases. Paying attention to this point, the amplitude or the amplitude equivalent amount of the correction parameter that is at least one of the detection value of the current detection means and the command voltage is a parameter for quantifying the gain error.

  Here, in the above invention, the detection value of the current detection means can be corrected by the correction means. Furthermore, the comparison between the amplitude equivalent amounts or the amplitudes performed by the correcting means can be performed, for example, during normal control of the rotating machine. For this reason, in the said invention, it can avoid that the opportunity which can correct | amend the said detected value is restrict | limited.

1 is a system configuration diagram according to a first embodiment. FIG. The figure which shows a gain error. The flowchart which shows the procedure of the process of the electric current process part concerning the embodiment. The flowchart which shows the procedure of the process of the voltage process part concerning the embodiment. 6 is a flowchart showing a processing procedure of a correction coefficient calculation unit according to the embodiment. The figure which shows the outline | summary of the gain error correction concerning the embodiment. The figure which shows the outline | summary of the gain error correction concerning the embodiment. The figure which shows the effect of the error correction process concerning the embodiment. The figure which shows the outline | summary of the gain error correction concerning 2nd Embodiment. The figure which shows the outline | summary of the gain error correction concerning the embodiment. The flowchart which shows the procedure of the process of the electric current process part concerning 3rd Embodiment. The flowchart which shows the procedure of the process of the voltage process part concerning the embodiment. The flowchart which shows the procedure of the error correction process concerning 4th Embodiment. The figure which shows the outline | summary of the integration method concerning 5th Embodiment. The system block diagram concerning 7th Embodiment. The figure which shows the voltage vector expressing the operation state of an inverter. The flowchart which shows the procedure of the process of the voltage process part concerning 7th Embodiment. The figure which shows the map used for calculation of the applied voltage concerning the embodiment. The system block diagram concerning 8th Embodiment. The figure which shows the angle error concerning the embodiment. 6 is a flowchart showing a procedure of error correction processing according to the embodiment. The figure which shows the correction coefficient calculation part concerning 9th Embodiment. 6 is a flowchart showing a procedure of error correction processing according to the embodiment.

(First embodiment)
DESCRIPTION OF EMBODIMENTS Hereinafter, a first embodiment in which a control device according to the present invention is applied to a vehicle equipped with a rotating machine as an in-vehicle main machine will be described with reference to the drawings.

  FIG. 1 shows the overall configuration of a motor generator control system according to this embodiment.

  The motor generator 10 as the in-vehicle main machine is a three-phase permanent magnet synchronous motor, more specifically, an embedded magnet synchronous motor (IPMSM).

  The motor generator 10 is connected to the high voltage battery 12 and the capacitor 14 via the inverter INV. The high voltage battery 12 is a direct current power source whose terminal voltage is, for example, 100 V or more. The inverter INV includes three sets of series connection bodies of switching elements S ¥ p, S ¥ n (¥ = u, v, w). The connection points of these series connection bodies are the U, V of the motor generator 10. , W phase. In the present embodiment, insulated gate bipolar transistors (IGBT) are used as the switching elements S ¥ # (# = p, n). Further, a diode D ¥ # is connected in antiparallel to each of the switching elements S ¥ #. Furthermore, the U, V, and W phase coils included in the motor generator 10 are Y-connected by connecting one end of each of the coils at a neutral point.

  In the present embodiment, the following is provided as detection means for detecting the state of the motor generator 10 and the inverter INV. First, a rotation angle sensor 16 (for example, a resolver) that detects the rotation angle (electrical angle θ) of the motor generator 10 is provided. Further, current sensors 18v and 18w for detecting the V-phase current and the W-phase current flowing in the V-phase and W-phase of the motor generator 10 are provided. Here, as the current sensors 18v and 18w, for example, a sensor provided with a current transformer or a resistor can be adopted. Further, a voltage sensor 20 that detects an input voltage (power supply voltage VDC) of the inverter INV is provided.

  The detection values of the various sensors are taken into the control device 22 constituting the low voltage system via an interface (not shown). Based on the detection values of these various sensors, the control device 22 generates and outputs an operation signal g ¥ # for operating the switching element S ¥ # included in the inverter INV.

  The control device 22 operates the inverter INV so as to control the output torque of the motor generator 10 to the required torque Tr. Specifically, the inverter INV is operated so that the command current for realizing the required torque Tr matches the current flowing in the motor generator 10. That is, in the present embodiment, the output torque of the motor generator 10 becomes the final control amount. In order to control the output torque, the current flowing through the motor generator 10 is used as a direct control amount, which is used as the command current. To control. In particular, in this embodiment, current vector control is performed to control the current flowing through the motor generator 10 to a command current. Hereinafter, the process regarding current vector control among the processes which the control apparatus 22 performs is demonstrated.

  Based on Kirchhoff's law, the U-phase current calculation unit 24 calculates the U-phase from the V-phase current and the W-phase current detected by the current sensors 18v and 18w (hereinafter, the V-phase current detection value iv and the W-phase current detection value iw). Calculate the current. Hereinafter, the U-phase current iu thus calculated is referred to as a U-phase current calculated value iu.

  The error correction units 26u, 26v, and 26w convert the U-phase error correction coefficient Ktu, the V-phase error correction coefficient Ktv, and the W-phase error correction coefficient Ktw into the U-phase current calculation value iu and the V-phase current detection values iv and W, respectively. The U-phase current calculation value iu, the V-phase current detection value iv, and the W-phase current detection value iw are corrected by multiplying the phase current detection value iw. The correction of the detected current value will be described in detail later.

  The two-phase conversion unit 28 uses the corrected U-phase current calculation value iu, V-phase current detection value iv, and W-phase current detection value iw based on the detected value of the electrical angle θ by the rotation angle sensor 16 as the current in the rotating coordinate system. To a d-axis current idr and a q-axis current iqr.

  The command current calculation unit 30 calculates a d-axis command current id * and a q-axis command current iq *, which are current command values of the rotating coordinate system, based on the required torque Tr.

  The command voltage calculation unit 32 uses the d and q command voltages Vd *, d as the operation amount for feedback control of the d axis current idr and the q axis current iqr to the d axis command current id * and the q axis command current iq *. Vq * is calculated. Specifically, the command voltage Vd * on the d axis is calculated by proportional integral control based on the deviation Δid between the d axis current idr and the d axis command current id *, and the q axis current iqr and q axis command current iq * are calculated. The command voltage Vq * on the q axis is calculated by proportional-integral control based on the deviation Δiq.

  The three-phase conversion unit 34 converts the command voltages Vd * and Vq * on the d and q axes into the three-phase command voltages V ¥ * (¥ = u, v, and v) in the fixed coordinate system of the motor generator 10 based on the electrical angle θ. w). These command voltages V ¥ * are operation amounts for performing feedback control of the d and q axis currents idr and iqr to the command currents id * and iq *.

  The operation unit 36 generates the operation signal g ¥ # for setting the three-phase output voltage of the inverter INV to a voltage simulating the command voltage V ¥ *. In the present embodiment, the operation signal g is controlled by sinusoidal PWM control based on a comparison of the magnitude of the command voltage V ¥ * with the input voltage VDC of the inverter INV detected by the voltage sensor 20 and a triangular wave carrier. Generate \ #. The operation unit 36 outputs the generated operation signal g ¥ # to the inverter INV via an interface (not shown). As a result, sinusoidal voltages that are 120 degrees out of phase with each other at the electrical angle of the motor generator 10 are applied to the U, V, and W phases of the motor generator 10, and the phases are 120 degrees out of phase with each other at the electrical angle. A sinusoidal phase current flows.

  Next, error correction processing according to the present embodiment will be described.

  This process is a process for suppressing torque fluctuations of motor generator 10 caused by a gain error included in V phase current detection value iv and W phase current detection value iw. Here, as shown in FIG. 2, the gain error means the actual value and the V-phase current detection value iv as the actual value of the phase current (V-phase current, W-phase current) flowing through the motor generator 10 increases. , An error that increases the deviation from the detected W-phase current value iw. That is, the gain error in a situation where a sine wave phase current flows through the motor generator 10 is an AC component error centered on 0A. When the gain error is included in the current detection value, the command voltage V ¥ * includes an AC component error centered on 0V. Hereinafter, the mechanism in which the gain error occurs, the mechanism in which the torque fluctuation of the motor generator 10 due to the gain error occurs, and the error correction process will be described in detail in this order.

  First, a mechanism that causes a gain error will be described.

  The detection of the phase current flowing through the motor generator 10 is usually performed by converting current to voltage by the current sensors 18v and 18w, amplifying the voltage by an operational amplifier (not shown), and discretizing the amplified voltage by an A / D converter (not shown). And quantization. A gain error occurs due to the conversion to the voltage or the amplification of the voltage.

  Next, a mechanism that causes torque fluctuation of the motor generator 10 due to the gain error will be described.

  When the V-phase current detection value iv and the W-phase current detection value iw including the gain error are converted into the d-axis current idr and the q-axis current iqr in the two-phase conversion unit 28, the two-phase conversion unit 28 converts them into the d-axis current idr and q-axis current iqr. A fluctuation component (hereinafter referred to as current secondary fluctuation) having an angular velocity twice as high as the electric angular velocity of the motor generator 10 is included. When the switching element S ¥ # is turned on / off by the command voltage V ¥ * based on the d-axis current idr and the q-axis current iqr including the secondary current fluctuation, the output torque of the motor generator 10 includes the electric angular velocity of the motor generator 10. A fluctuation component having a double angular velocity (hereinafter referred to as torque secondary fluctuation) is included. That is, torque fluctuation due to gain error occurs.

  In particular, in this embodiment, since the U-phase current calculation value iu is calculated based on the V-phase current detection value iv and the W-phase current detection value iw, at least one of these detection values includes a gain error. The U-phase current calculation value iu may deviate from the actual U-phase current. In this case, the torque fluctuation of the motor generator 10 due to the gain error may become significant.

  In order to deal with such a problem, the error correction process is performed. In the present embodiment, the control device 22 includes a current processing unit 38, a voltage processing unit 40, and a correction coefficient calculation unit 42 in order to perform error correction processing. Hereinafter, this process will be described with reference to FIGS. The current processing unit 38 and the voltage processing unit 40 are provided with a memory as storage means.

  FIG. 3 is a diagram illustrating a procedure of processing executed by the current processing unit 38 in the error correction processing. This process is repeatedly executed at a predetermined control cycle Tc.

  In this series of processing, first, in step S10, the U-phase current calculation value iu and the V and W-phase current detection values iv and iw are acquired, and the U, V, and W-phase absolute current values | i ¥ | (¥ = u, v, w) is calculated. As a specific method for calculating the absolute value, for example, the absolute value is calculated by changing the highest-order code bit from “1” to “0” for the signed variable storing the acquired data. It is sufficient to adopt a technique to do this.

  In the subsequent step S12, the U-phase current calculation value iu and the integrated values Δi ¥ of the V and W phase current detection values iv and iw are calculated. This is performed by adding the multiplication value of the control cycle Tc and the current absolute value | i ¥ | to the previous integration value Δi ¥.

  In the subsequent step S14, it is determined whether or not the integration interval of the integration value Δi ¥ is one electrical angle cycle. If it is determined that one electrical angle cycle has been reached, the process proceeds to step S16, where the integrated value Δi ¥ is held as an integrated value Δiθ ¥ over one electrical angle cycle.

  If a negative determination is made in step S14, or if the process of step S16 is completed, this series of processes is temporarily terminated.

  Next, a procedure of processing executed by the voltage processing unit 40 in the error correction processing will be described with reference to FIG. This process is repeatedly executed at a predetermined control cycle Tc.

  In this series of processing, first, in step S18, U, V, and W phase command voltages V ¥ (¥ = u, v, w) are acquired, and U, V, and W phase voltages that are their respective absolute values. The absolute value | V ¥ | is calculated.

  In the subsequent step S20, the respective integral values ΔV ¥ of the U, V, W phase voltage absolute values | V ¥ | are calculated. This is performed by adding a multiplication value of the control period Tc and the current voltage absolute value | V ¥ | to the previous integration value ΔV ¥, similarly to the processing in step S12 of FIG.

  In the subsequent step S22, it is determined whether or not the integration interval of the integration value ΔV ¥ is one electrical angle cycle. If it is determined that one electrical angle cycle has been reached, the process proceeds to step S24, and the integration value ΔV ¥ is held as an integration value ΔVθ ¥ over one electrical angle cycle.

  If a negative determination is made in step S22, or if the process in step S24 is completed, this series of processes is temporarily terminated.

  Subsequently, a procedure of processing executed by the correction coefficient calculation unit 42 in the error correction processing will be described with reference to FIG. This process is repeatedly executed at a predetermined control cycle Tc.

  In this series of processes, first, in step S26, the electric angular velocity ω of the motor generator 10 is equal to or higher than the first angular velocity ωA higher than 0 and equal to or lower than the second angular velocity ωB higher than the first angular velocity ωA. Determine whether or not. Here, the electrical angular velocity ω is calculated as a differential operation value of the electrical angle θ detected by the rotation angle sensor 16 in the speed calculation unit 44 provided in the control device 22 as shown in FIG. Used.

  In the present embodiment, the first angular velocity ωA and the second angular velocity ωB are fixed values determined based on the responsiveness of the control device 22. Specifically, the first angular velocity ωA can sufficiently secure the responsiveness of the feedback control for canceling the secondary current fluctuation included in the d and q axis currents idr and iqr in the command voltage calculation unit 32. It is set to an assumed value, and the second angular velocity ωB has sufficient feedback control response for canceling the secondary current fluctuation included in the d and q axis currents idr and iqr in the command voltage calculation unit 32. The value is assumed to be unable to be secured.

  If a negative determination is made in step S26, the process proceeds to step S28 to determine whether or not the electrical angular velocity ω is higher than the second angular velocity ωB. This process is a process for determining whether or not to perform an error correction process based on the integral value Δiθ ¥ of the current absolute value | i ¥ |.

  If an affirmative determination is made in step S28, error correction processing based on the integral value Δiθ ¥ of the current absolute value | i ¥ | is performed in subsequent steps S30 to S38. Hereinafter, the reason why the integral value Δi ¥ is used for the error correction process will be described with reference to FIG.

  FIG. 6A-1 shows one electrical angle period of each of the U-phase current calculation value iu, the V-phase current detection value iv, the W-phase current detection value iw, the actual V-phase current irv, and the actual W-phase current irw. 6 (a-2) shows the absolute values of the U, V, and W phase currents | i ¥ |, the absolute value of the actual V phase current irv, and the absolute value of the actual W phase current irw. The transition of the integrated value with each electrical angle θ as a variable is shown. In the example shown in FIG. 6A-1, a gain error in which the absolute value of the V-phase current detection value iv is larger than the actual absolute value of the V-phase current irv is included in the V-phase current detection value iv. Thus, a gain error in which the absolute value of the W-phase current detection value iw is smaller than the actual absolute value of the W-phase current irw is included in the W-phase current detection value iw. For this reason, the actual U-phase current iru and the calculated U-phase current value iu coincide.

  As shown in FIG. 6 (a-1), when a gain error is included in the V-phase current detection value iv and the W-phase current detection value iw, the electrical angle θ is changed as shown in FIG. 6 (a-2). The integrated values Δiθv and Δiθw over one electrical angle period of the current absolute values | iv | and | iw | of the V and W phases are the absolute values of the U phase currents | iu | (the absolute values of the actual U phase currents iru) ) Deviates from the integral value Δiθu over one electrical angle period. Specifically, the integrated value Δiθv of the V-phase current absolute value | iv | is larger than the integrated value Δiθu of the U-phase current absolute value | iu |, and the W-phase current absolute value | iw The integrated value Δiθw of | is small. From these results, the integral value Δiθ ¥ over one electrical angle period of the current absolute values | i ¥ | of the U, V, and W phases is a parameter for quantifying the gain error. For this reason, the integral value Δiθ ¥ is used for error correction processing.

  FIG. 6 (b-1) shows the transition of the command voltages Vu *, Vv *, and Vw * over one electrical angle period, and FIG. 6 (b-2) shows the U, V, and W phase transitions. The transition of the integrated value with each electrical angle θ of the voltage absolute value | V * | as a variable is shown.

  As shown in FIG. 6B-2, the difference between the integral values ΔVθ ¥ over one electrical angle period of the U, V, and W phase voltage absolute values | V ¥ | is very small, and these integral values ΔVθ ¥. The gain error cannot be quantified. This is due to the following reason.

  The situation in which an affirmative determination is made in step S28 of FIG. 5 is that the responsiveness for feedback control of the d-axis current idr and the d-axis current idr to the command currents id * and iq * is not sufficient in the command voltage calculation unit 32. It is. Under such circumstances, it is possible to calculate the d-axis command voltage Vd * and the q-axis command voltage Vq * so as to cancel the secondary current fluctuation included in the d-axis current idr and the q-axis current iqr due to the gain error. Can not. That is, the difference between the three-phase command voltage V ¥ * based on these command voltages Vd * and Vq * and the ideal command voltage for setting the output torque of the motor generator 10 as the required torque Tr tends to be small. . Since the integral values over one electrical angle period of the absolute value of the ideal command voltage are the same, the difference between the integral values ΔVθ ¥ of the voltage absolute value | V ¥ | For this reason, in the situation in which an affirmative determination is made in step S28, the responsiveness of the feedback control of the command voltage calculation unit 32 at the current electrical angular velocity of the motor generator 10 becomes low, and the integral value ΔVθ ¥ of the voltage absolute value | V ¥ | The gain error cannot be quantified.

  On the other hand, in the situation in which an affirmative determination is made in step S28, the actual U, V, W phase currents iru, irv, irw and the U, V, W phase currents when the ideal command voltage is applied. Although the deviation is small, the V and W phase current detection values iu and iv are deviated from the actual V and W phase currents irv and irw when the ideal command voltage is applied due to a gain error. Therefore, the gain error can be quantified by the integral value Δiθ ¥ of the current absolute value | i ¥ |. Therefore, as described above, the integral value Δiθ ¥ is selected as a parameter used for the error correction process.

  Returning to FIG. 5, in step S30, the first switch SW1 is turned off and the second switch SW2 is turned on. In step S32, an integral value Δiθ ¥ over one electrical angle period of the current absolute value | i ¥ | is acquired from the current processing unit 38.

  In the subsequent step S34, an intermediate value is selected as the current reference value from the integral value iθ ¥ of the U, V and W phase current absolute values | i ¥ |. In step S36, the U, V, and W phase second correction coefficients KA ¥ are obtained by dividing the current reference value by the integral value Δiθ ¥ of the U, V, and W phase current absolute values | i ¥ |. calculate. That is, the second correction coefficient KA ¥ is calculated based on a comparison between phases of the integral value Δiθ ¥ of the current absolute value | i ¥ |.

  In the subsequent step S38, the error correction coefficient Kt ¥ for the U, V, and W phases is updated with the second correction coefficient KA ¥ for the U, V, and W phases, and the updated error correction coefficient Kt ¥ is updated to the error correction unit 26u, It outputs to 26v and 26w. As a result, the U-phase current calculation value iu, the V-phase current detection value iv, and the W-phase current detection value iw are corrected in the error correction units 26u, 26v, and 26w. That is, the U-phase current calculation value iu, the V-phase current detection value iv, and the W-phase current detection value iw are corrected so that the maximum value and the minimum value of the integral value Δiθ ¥ of each phase approach the intermediate value. It becomes.

  On the other hand, if a negative determination is made in step S28, the process proceeds to step S40 to determine whether or not the electrical angular velocity ω is higher than 0 and lower than the first angular velocity ωA. This process is a process for determining whether or not to perform an error correction process based on the integral value ΔVθ ¥ of the voltage absolute value | V ¥ |.

  If an affirmative determination is made in step S40, error correction processing based on the integral value ΔVθ ¥ of the voltage absolute value | V ¥ | is performed in subsequent steps S42 to S50. Hereinafter, the reason why the integrated value ΔVθ ¥ is used for the error correction process will be described with reference to FIG.

  FIG. 7 (b-1) shows the transition of the three-phase command voltage V ¥ * over one electrical angle period, and FIG. 7 (b-2) shows the U, V, and W phase absolute voltage values | V. The transition of the integral value with each electrical angle θ as a variable is shown. In the example shown in FIGS. 7B-1 and 7B-2, the absolute value of the V-phase current detection value iv is larger than the absolute value of the actual V-phase current irv, as in FIG. Is included in the V-phase current detection value iv, and a gain error in which the absolute value of the W-phase current detection value iw is smaller than the actual absolute value of the W-phase current irw is a W-phase current detection value iw. Shows the situation that is included.

  The situation in which an affirmative determination is made in step S40 of FIG. 5 is a situation in which the command voltage calculation unit 32 has sufficient responsiveness for feedback control of the d-axis current idr and the d-axis current idr to the command currents id * and iq *. It is. Under such circumstances, it is possible to calculate the d-axis command voltage Vd * and the q-axis command voltage Vq * so as to cancel the secondary current fluctuation included in the d-axis current idr and the q-axis current iqr due to the gain error. These command voltages Vd * and Vq * include a fluctuation component having an angular velocity that is twice the electrical angular velocity of the motor generator 10. Therefore, as shown in FIG. 7 (b-1), the three-phase command voltage V ¥ * based on these command voltages Vd * and Vq * and the ideal for setting the output torque of the motor generator 10 to the required torque Tr. There is a tendency for the deviation from a typical command voltage to increase.

  Therefore, as shown in FIG. 7B-2, the integrated values ΔVθv and ΔVθw over one electrical angle period of the voltage absolute values | Vv | and | Vw | of the V and W phases with the electrical angle θ as a variable are It will deviate from the integral value ΔVθu of the U-phase voltage absolute value | Vu |. Specifically, the integrated value ΔVθv of the V-phase voltage absolute value | Vv | is larger than the integrated value ΔVθu of the U-phase voltage absolute value | Vu |, and the W-phase voltage absolute value | Vw | The integral value ΔVθw becomes small. From these results, the integral value ΔVθ ¥ over one electrical angle period of the U, V, and W phase voltage absolute values | V ¥ | becomes a parameter for quantifying the gain error. Therefore, the integral value ΔVθ ¥ is used for error correction processing.

  FIGS. 7 (a-1) and 7 (a-2) show diagrams corresponding to FIGS. 6 (a-1) and 6 (a-2).

  As shown in FIG. 7A-2, the difference between the integral values Δiθ ¥ over one electrical angle period of the current absolute values | i ¥ | of the U, V, and W phases is very small, and these integral values Δiθ ¥ The gain error cannot be quantified. This is due to the following reason.

  The situation in which an affirmative determination is made in step S40 of FIG. 5 is a situation in which the command voltage calculation unit 32 has sufficient responsiveness for feedback control of the d-axis current idr and the d-axis current idr to the command currents id * and iq *. It is. Under such circumstances, as shown in FIG. 7 (a-1), the deviation between the amplitudes of the U-phase current calculation value iu and the V and W-phase current detection values iv and iw becomes small. For this reason, as shown in FIG. 7A-2, the difference between the integral values Δiθ ¥ of the absolute current value | i ¥ | becomes small, and the gain error cannot be quantified by these integral values Δiθ ¥. Therefore, in the situation where an affirmative determination is made in step S40, as described above, the integral value ΔVθ ¥ is selected as the parameter used for the error correction process.

  Returning to FIG. 5, in step S42, the first switch SW1 is turned on and the second switch SW2 is turned off. In step S44, an integral value ΔVθ ¥ over one electrical angle period of the voltage absolute value | V ¥ | is acquired from the voltage processing unit 40.

  In the subsequent step S46, an intermediate value is selected as the voltage reference value from the integral value ΔVθ ¥ of the U, V, and W phase voltage absolute values | V ¥ |. In subsequent step S48, the first correction coefficient KV ¥ for the U, V, and W phases is calculated by dividing the voltage reference value by the integral value ΔVθ ¥. In step S50, the U, V, and W phase error correction coefficients Kt ¥ are updated with the first U, V, and W phase correction coefficients KV ¥, and the updated error correction coefficient Kt ¥ is updated to the error correction unit 26u. , 26v, 26w. That is, the U-phase current calculation value iu, the V-phase current detection value iv, and the W-phase current detection value iw are corrected so that the maximum value and the minimum value of the integral value ΔVθ ¥ of each phase approach the intermediate value. It becomes.

  On the other hand, if an affirmative determination is made in step S26, an error correction process based on both the integral value Δiθ ¥ of the current absolute value | i ¥ | and the integral value ΔVθ ¥ of the voltage absolute value | V ¥ | is performed. Here, the reason why the two integral values Δiθ ¥ and ΔVθ ¥ are used is that the contribution as a parameter for quantifying the gain error is approximately the same for each of the integral values Δi ¥ and ΔV ¥.

  Specifically, in step S52, both the first switch SW1 and the second switch SW2 are turned on.

  In subsequent steps S32 to S36, the second correction coefficient KA ¥ is calculated based on the integral value Δiθ ¥ of the current absolute value | i ¥ |. In steps S44 to S48, the first correction coefficient KV ¥ is calculated based on the integral value ΔVθ ¥ of the voltage absolute value | V ¥ |.

  In the subsequent step S54, the U, V, W phase error correction coefficient Kt \ is updated by the multiplication value of the U, V, W phase first correction coefficient KV \ and the second correction coefficient KA \, and updated. The error correction coefficient Kt ¥ is output to the error correction units 26u, 26v, and 26w.

  When a negative determination is made in step S40 or when the processes in steps S38, S50, and S54 are completed, the series of processes is temporarily ended.

  Incidentally, the situation in which a negative determination is made in step S40 is, for example, a situation in which the rotational speed of the motor generator 10 is “0”.

  FIG. 8 shows the effect of the error correction process.

  As illustrated, according to the error correction process, the gain error included in the V-phase current detection value iv and the W-phase current detection value iw is eliminated, and the U-phase current calculation value iu caused by the gain error is actually Is eliminated from the U-phase current iru. As a result, fluctuations in the output torque of the motor generator 10 are eliminated.

  According to the embodiment described in detail above, the following effects can be obtained.

  (1) When it is determined that the electrical angular velocity ω of the motor generator 10 is higher than the second angular velocity ωB, the error correction coefficient Kt ¥ is based on the integral value Δiθ ¥ over one electrical angular period of the current absolute value | i ¥ | Was calculated. On the other hand, when it is determined that the electrical angular velocity ω is higher than 0 and less than the first angular velocity ωA, the error correction coefficient Kt is based on the integral value ΔVθ ¥ over one electrical angular period of the voltage absolute value | V ¥ |. ¥ was calculated. Therefore, a parameter that can appropriately quantify the gain error according to the operating state of the motor generator 10 can be used for calculating the error correction coefficient Kt ¥. Then, the U-phase current calculation value iu and the like can be appropriately corrected by multiplying the error correction coefficient Kt ¥ by the U-phase current calculation values iu and V and the W-phase current detection values iv and iw.

  Further, the integration values Δiθ ¥ and ΔVθ ¥ are used for calculating the error correction coefficient Kt ¥, for example, when noise is mixed in the V and W phase current detection values iv and iw and the command voltage V ¥ *. However, the influence of noise on the error correction coefficient Kt ¥ can be reduced as much as possible. On the other hand, with the correction method described in Patent Document 1, it is not possible to mitigate the influence of noise contamination on the current detection value on the correction of the current detection value.

  (2) When it is determined that the electrical angular velocity ω is greater than or equal to the first angular velocity ωA and less than or equal to the second angular velocity ωB, the integrated value Δiθ ¥ of the current absolute value | i ¥ | and the absolute voltage value | V ¥ An error correction coefficient Kt ¥ was calculated based on both of the integral values ΔVθ ¥ of |. As a result, the accuracy of calculating the error correction coefficient Kt ¥ can be increased in a situation where the contribution as a parameter for quantifying the gain error is approximately the same in the integral values Δiθ ¥ and ΔVθ ¥.

  (3) The integral values Δiθ ¥ and ΔVθ ¥ used for the error correction process are set as integral values over one electrical angle period. According to the integral value thus calculated, the gain error can be accurately quantified.

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

  In this embodiment, instead of the integrated value Δiθ ¥ (¥ = u, v, w) of the absolute current value | i ¥ | and the integrated value ΔVθ ¥ of the absolute voltage value | V ¥ |, the calculated U-phase current value iu, The integral value over one electrical angle cycle of the current square value i \ ^ 2, which is the square value of the V-phase current detection value iv and the W-phase current detection value iw, and the U, V, W-phase command voltage V ** The integral value over one electrical angle period of the voltage square value V \ ^ 2 that is the square value of is used for the calculation of the error correction coefficient Kt \. Even in this case, as shown in FIG. 9 and FIG. 10, the gain error can be quantified by these integral values. 9 and 10 correspond to FIGS. 6 and 7 above.

  Incidentally, the error correction processing according to the present embodiment can be performed by processing according to the processing shown in FIGS. Specifically, the process of step S10 of FIG. 3 is replaced with a process of calculating the U-phase current calculation value iu and the square value i \ ^ 2 of the V and W phase current detection values iv and iw. The process of step S18 is replaced with a process of calculating the square value V ¥ ^ 2 of the command voltage V ¥ *.

  As described above, the error correction processing according to the present embodiment can achieve the same effects as those of the first embodiment.

  When the U-phase current calculation value iu and the V and W-phase current detection values iv and iw are larger than “1”, the current square value i ¥ ^ 2 is larger than the current absolute value | i ¥ |. For this reason, the integral value of the current square value i \ ^ 2 is larger than the integral value of the current absolute value | i \ |. Thereby, in addition to the above-described effect, it is possible to achieve an effect of increasing the gain error quantification accuracy when the value is larger than “1”. The same applies to the integral value of the voltage square value V \ ^ 2.

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

  In this embodiment, instead of the integrated value Δiθ ¥ (¥ = u, v, w) of the absolute current value | i ¥ | and the integrated value ΔVθ ¥ of the absolute voltage value | V ¥ |, the calculated U-phase current value iu, The exponential moving average value iema ¥ of the amplitude of the V phase current detection value iv and the W phase current detection value iw and the exponential moving average value Vema ¥ of the amplitude of the command voltage V ¥ * are used for the calculation of the error correction coefficient Kt ¥. Here, the amplitude is an absolute value of a positive or negative peak value that appears once in a half cycle of the electrical angle θ.

  FIG. 11 shows a procedure of processing executed by the current processing unit 38 in the error correction processing.

  In this series of processing, in step S55, the U phase current calculation value iu and the amplitude amp of the V and W phase current detection values iv and iw are calculated. Here, the amplitude amp is a value obtained by adding the absolute values of the positive peak value and the negative peak value in one electrical angle period for each of the U-phase current calculation value iu and the V and W-phase current detection values iv and iw. It is sufficient to calculate as “½”.

  In subsequent step S56, the amplitude amp corresponding to each phase calculated in step S55 is stored in its own memory.

  In the subsequent step S57, the exponential moving average value iema ¥ is calculated based on a plurality of past amplitudes amp stored in the memory. Here, as the plurality of past amplitudes amp used to calculate the exponential moving average value iema ¥, for example, an amplitude in an N electrical angle period (N is an integer of 2 or more) may be used.

  In addition, when the process of step S57 is completed, this series of processes is once complete | finished.

  FIG. 12 shows a procedure of processing executed by the voltage processing unit 40 in the error correction processing.

  In this series of processing, the amplitude amp of the command voltage V ¥ * is calculated in step S58. Here, the amplitude amp may be calculated by a method similar to the process in step S55 of FIG.

  In subsequent step S59, the amplitude amp corresponding to each phase calculated in step S58 is stored in its own memory.

  In the subsequent step S60, the exponential moving average value Vema ¥ is calculated based on a plurality of past amplitudes amp stored in the memory, similarly to the processing in step S57 of FIG.

  In addition, when the process of step S60 is completed, this series of processes is once complete | finished.

  Incidentally, the process in the correction coefficient calculation unit 42 can be performed by a process according to the process shown in FIG. Specifically, the process of step S32 in FIG. 5 is replaced with a process of acquiring the exponential moving average value ema ¥ corresponding to each phase from the current processing unit 38, and the process of step S34 is replaced with the above process corresponding to each phase. It replaces with the process which selects an intermediate value as an electric current reference value among exponential moving average values iema \. The same applies to steps S44 and S46.

  According to the above-described method for calculating the error correction coefficient Kt ¥, the amplitude of the U-phase current calculated value iu and the like stored most recently greatly affects the calculation of the error correction coefficient Kt ¥. For this reason, for example, even when the operating state of the motor generator 10 changes suddenly, the influence of the sudden change on the error correction coefficient Kt ¥ can be suppressed as much as possible.

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

  In the present embodiment, during the transient operation of the motor generator 10, the integration of the absolute current value | i ¥ | and the absolute voltage value | V ¥ | is stopped, and the calculation of the error correction coefficient Kt * is stopped. Then, during the transient operation, using the previously calculated error correction coefficient Kt *, the error correction units 26u, 26v, and 26w use the U-phase current calculation value iu, the V-phase current detection value iv, and the W-phase current detection value iw. The process which correct | amends is performed. This process is performed in view of the fact that the accuracy of quantifying the gain error decreases depending on the integral value Δiθ ¥ based on the detected V-phase current value iv during the transient operation of the motor generator 10.

  FIG. 13 shows the procedure of the above processing. This process is repeatedly executed at a predetermined control cycle Tc.

  In this series of processing, first, in step S61, the condition that the absolute value of the change amount ΔTr of the required torque Tr is larger than the first specified amount ΔTth and the absolute value of the change amount Δω of the electrical angular velocity ω are the second specified value. It is determined whether or not the logical sum of the condition that it is larger than the amount Δωth is true. This process is a process for determining whether or not the motor generator 10 is in a transient operation.

  If an affirmative determination is made in step S61, the process proceeds to step S62, and the calculation of the integrated values Δi ¥ and ΔV ¥ of the current absolute value | i ¥ | and the voltage absolute value | V ¥ | is stopped. Thus, during the transient operation of the motor generator 10, the integrated values Δiθ ¥ and ΔVθ ¥ over one electrical angular period output from the current processing unit 38 and the voltage processing unit 40 to the correction coefficient calculation unit 42 shift to the transient operation. Are held at the integral values Δiθ ¥ and ΔVθ ¥ immediately before.

  If a negative determination is made in step S61, or if the process of step S62 is completed, this series of processes is temporarily terminated.

  As described above, in the present embodiment, by performing the processing shown in FIG. 13, the calculation of the error correction coefficient Kt ¥ can be stopped during the transient operation of the motor generator 10 in which the gain error quantification accuracy decreases. . And at the time of an excessive driving | operation, correction | amendment of U phase electric current calculation value iu etc. can be continued using the error correction coefficient Kt ¥ immediately before shifting to an excessive driving | operation. That is, it is possible to suppress the torque fluctuation caused by the gain error in the transient state as much as possible.

(Fifth embodiment)
Hereinafter, a fifth embodiment will be described with reference to the drawings, focusing on differences from the first embodiment.

  In the present embodiment, the integration interval Tint of the absolute current value | i ¥ | and the absolute voltage value | V ¥ | is N electrical angular periods (N is an integer of 2 or more). This setting is for increasing the accuracy of gain error quantification based on the integral values Δiθ ¥ and ΔVθ ¥.

  Specifically, the current absolute value | i ¥ | will be described as an example. In order to increase the gain error quantification accuracy, the U-phase current calculation value iu and the V-phase current detection value iv used to calculate the integral value Δiθ ¥. It is desirable to use values at various electrical angles θ as the W-phase current detection value iw. However, as shown in FIG. 14A, when the sampling number of the V-phase current calculation value iv and the like used for calculating the integral value is insufficient, the electrical angle θ corresponding to the V-phase current calculation value iv and the like becomes a quantification of the gain error. It will not be sufficient to increase the accuracy of the process. As a result, errors are likely to occur between the integrated values iθ ¥ of the U, V, and W phases.

  On the other hand, as shown in FIG. 14B, when the integration interval Tint of the integral value is set to N electrical angle periods (N = 4 is illustrated in the figure), the V-phase current calculated value iv used for calculating the integral value. The number of samplings can be increased. As a result, as long as an integral multiple of the sampling period does not become the electrical angle period, the electrical angle θ such as the V-phase current to be sampled becomes various values, and the error between the integral values iθ ¥ can be reduced.

  Thus, according to the present embodiment, by setting the integral interval Tint of the integral value to the N electrical angle period, phase currents and the like at various electrical angles can be used for calculating the error correction coefficient Kt ¥. As a result, the gain error quantification accuracy based on the integral values Δiθ ¥ and ΔVθ ¥ can be further increased.

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

  In this embodiment, instead of the integrated value Δiθ ¥ (¥ = u, v, w) of the absolute current value | i ¥ | and the integrated value ΔVθ ¥ of the absolute voltage value | V ¥ |, the calculated U-phase current value iu, The amplitude amp of the V-phase current detection value iv and the W-phase current detection value iw and the amplitude amp of the command voltage V ¥ * are used to calculate the error correction coefficient Kt ¥.

  Incidentally, the error correction processing according to the present embodiment can be performed by processing according to the processing shown in FIGS.

  Thus, in the present embodiment, the amplitude amp is used to calculate the error correction coefficient Kt ¥. Even in this case, the U-phase current calculation value iu and the like can be corrected.

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

  In the present embodiment, the control mode of the motor generator 10 is changed from that based on current vector control. Specifically, in order to control the current flowing through the motor generator 10 to a command current, the current of the motor generator 10 is predicted for a case where the operation state of the inverter INV is temporarily set to each of a plurality of types, and the temporary current is estimated based on the predicted current. Evaluate the set operation state. Then, model predictive control is performed in which the highly evaluated operation state is adopted as the actual operation state of the inverter INV.

  FIG. 15 shows an overall configuration of a motor generator control system according to the present embodiment. In FIG. 15, the same members as those in FIG. 1 are given the same reference numerals for the sake of convenience.

  As illustrated, in this embodiment, in addition to the current sensors 18v and 18w, a current sensor 18u that detects a U-phase current is provided. In the present embodiment, the detection value of the current sensor 18u is referred to as a U-phase current detection value iu.

  Next, processing related to model prediction control among processing performed by the control device 22 will be described.

  The model prediction control unit 45 outputs the d and q axis command currents id * and iq * output from the command current calculation unit 30, the d and q axis currents idr and iqr output from the two-phase conversion unit 28, the electrical angle θ, Then, the electric angular velocity ω output from the speed calculation unit 44 is input, and a voltage vector Vi that defines the operation state of the inverter INV is determined and output to the operation unit 36. The operation unit 36 generates the operation signal g ¥ # based on the input voltage vector Vi and outputs it to the inverter INV.

  Here, the voltage vectors expressing the operation state of the inverter INV are eight voltage vectors shown in FIG. For example, the voltage vector representing the operation state (indicated as “lower” in the figure) in which all of the low-potential side switching elements Sun, Svn, Swn are in the on state is the voltage vector V0, and the high-potential side switching element A voltage vector representing an operation state (indicated as “upper” in the drawing) in which all of Sup, Svp, and Swp are in an on state is a voltage vector V7. These voltage vectors V0 and V7 are for short-circuiting all phases of the motor generator 10, and are referred to as zero voltage vectors because the voltage applied to the motor generator 10 from the inverter INV becomes zero. On the other hand, the remaining six voltage vectors V1 to V6 are defined by an operation pattern in which switching elements that are turned on exist in both the upper arm and the lower arm, and are referred to as effective voltage vectors. . As shown in FIG. 16B, each of the voltage vectors V1, V3, and V5 corresponds to the positive side of the U phase, the V phase, and the W phase, respectively.

  Returning to the description of FIG. 15, the temporary setting unit 46 temporarily sets the operation state of the inverter INV. Here, voltage vectors V0 to V7 shown in FIG. 16 are temporarily set as the operation state of inverter INV.

  The dq conversion unit 48 calculates the voltage vector Vdq = (vd, vq) of the rotating coordinate system by performing dq conversion on the voltage vector set by the temporary setting unit 46. In order to perform such conversion, the voltage vectors V0 to V7 in the temporary setting unit 46 are set to, for example, “upper” as “VDC / 2” and “lower” as “−VDC / 2” in FIG. Can be expressed as In this case, for example, the voltage vector V0 is (−VDC / 2, −VDC / 2, −VDC / 2), and the voltage vector V1 is (VDC / 2, −VDC / 2, −VDC / 2). .

In the prediction unit 50, when the operation state of the inverter INV is temporarily set by the temporary setting unit 46 based on the voltage vector (vd, vq), the d and q axis currents idr, iqr, and the electrical angular velocity ω. The currents idr and iqr at are predicted. Here, basically, the voltage equations expressed by the following (eq1) and (eq2) are discretized from the following state equations (eq3) and (eq4) obtained by solving the current differential term. Predict current.
vd = (R + pLd) idr−ωLqiq (eq1)
vq = ωLdidr + (R + pLq) iqr + ωφ (eq2)
pidr
= − (R / Ld) idr + ω (Lq / Ld) iqr + vd / Ld (eq3)
piqr
= −ω (Ld / Lq) idr− (Rd / Lq) iqr + vq / Lq−ωφ / Lq
... (eq4)
Incidentally, in the above equation, the resistance R, the differential operator p, the d-axis inductance Ld, the q-axis inductance Lq, and the armature flux linkage constant φ are used.

  The prediction of the current in the prediction unit 50 is performed for each of a plurality of operation states temporarily set by the temporary setting unit 46.

  On the other hand, the operation state setting unit 52 receives the predicted currents ide and iq predicted by the prediction unit 50 and the command currents id * and iq * as input, and determines the operation state of the inverter INV. Here, a voltage vector that minimizes the evaluation function J among the eight voltage vectors Vi (i = 0 to 7) is determined as the operation state. As the evaluation function J, a function whose value increases as the evaluation becomes lower is adopted. Specifically, the evaluation function J is calculated based on the inner product value of the difference between the command current vector Idqr = (id *, iq *) and the predicted current vector Idqe = (ide, iqe). This is one method for expressing that the evaluation is lower as the value is larger in view of the fact that the deviation of each component between the command current vector Idqr and the predicted current vector Idqe can be both positive and negative values. As a result, it is possible to construct an evaluation function J in which the evaluation becomes lower as the difference between the components of the command current vector Idqr and the predicted current vector Idqe is larger.

  Based on the operation state thus determined, the operation unit 36 generates an operation signal g ¥ # and outputs it to the inverter INV.

  Next, error correction processing according to the present embodiment will be described.

  In the present embodiment, the voltage vector Vi set each time by the operation state setting unit 52 and the power supply voltage VDC detected by the voltage sensor 20 are input, and the command voltage V ¥ * for each phase is calculated and calculated. An integral value ΔVθ ¥ of the voltage absolute value | V ¥ | used for calculation of the first correction coefficient KV ¥ is calculated as an integral value over one electrical angle period of the command voltage V ¥ *.

  FIG. 17 shows a procedure of processing executed by the voltage processing unit 40 in the error correction processing. This process is repeatedly executed at a predetermined control cycle Tc. In FIG. 17, the same steps as those shown in FIG. 4 are given the same step numbers for the sake of convenience.

  In this series of processing, first, in step S64, the current voltage vector V (n) is acquired. In the subsequent step S66, the map operation is performed on the command voltage V ¥ * (n) (¥ = u, v, w) of each phase from the voltage vector V (n). FIG. 18 shows this map. As shown in the figure, the map defines the relationship between each voltage vector Vi (i = 0 to 7) and the command voltages Vu *, Vv *, and Vw *. Here, for example, regarding the voltage vector V1, it becomes a map value as follows.

  First, the line voltage Vuv between the U phase and the V phase, the line voltage Vvw between the V phase and the W phase, and the line voltage Vwu between the W phase and the U phase are expressed by the following equations.

Vuv = vu-vv = VDC-0 = VDC
Vvw = vv-vw = 0-0 = 0
Vwu = vw−vu = 0−VDC = −VDC
Here, by using the relationship of “vu + vv + vw = 0”, the value of the map shown in FIG. 18 is obtained. In the present embodiment, the voltage processing unit 40 constitutes a command voltage calculation unit.

  Returning to FIG. 17, in step S68, the integrated value ΔV ¥ of the command voltage V ¥ * of each phase is calculated. This is performed by adding the product of the control cycle Tc and the current command voltage V ¥ * (n) to the previous integrated value ΔV ¥, as in step S20 of FIG. Thereafter, the process proceeds to step S22.

  If a negative determination is made in step S22, or if the process in step S24 is completed, this series of processes is temporarily terminated.

  Thus, in the present embodiment, the U, V, and W phase current detection values i ¥ can be appropriately corrected by the error correction process even in the configuration that performs the model predictive control.

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

  In the present embodiment, in addition to the gain error, an error correction process is performed to suppress the torque fluctuation of the motor generator 10 caused by the angle error included in the electrical angle θ detected by the rotation angle sensor 16.

  FIG. 19 shows an overall configuration of a motor generator control system according to the present embodiment. In FIG. 19, the same members as those in FIG. 1 are given the same reference numerals for the sake of convenience.

  As shown in the drawing, the correction coefficient calculation unit 42 calculates the first correction coefficient KV ¥ (¥ = u, v, w), the first correction coefficient calculation unit 42a, and the second correction coefficient KA ¥. And a second correction coefficient calculation unit 42b for calculating.

  The first error correction units 54u, 54v, 54w multiply the U-phase current calculation value iu, the V-phase current detection value iv, and the W-phase current detection value iw by the first correction coefficient KV ¥ for the U, V, and W phases. By doing so, these values i ¥ are corrected.

  The second error correction units 56u, 56v, and 56w multiply the output values of the first error correction units 54u, 54v, and 54w by the second correction coefficient KA ¥ for the U, V, and W phases, thereby outputting the output. Correct the value.

  Next, the angle error, the influence of the angle error on torque fluctuation and the error correction processing for suppressing the torque fluctuation will be described.

  First, the angle error will be described.

  In the present embodiment, the angle error is a sinusoidal error having a period of an electrical angle half cycle or M electrical angle period (M: positive integer) of the motor generator 10. In the present embodiment, as shown in FIG. 20, it is assumed that an angle error Δθ having the same cycle (M = 1) as one electrical angle cycle of the motor generator 10 is included in the electrical angle θ.

  Next, the influence of the angle error on torque fluctuation will be described.

  In a situation where the electrical angle θ detected by the rotation angle sensor 16 includes the angle error Δθ, the U-phase current calculation values iu and V, W-phase current detection values iv, iw including the gain error in the two-phase converter 28. Is converted to dq. As a result, the present inventors have examined that the d and q axis currents idr and iqr output from the two-phase converter 28 include a fluctuation component having the same angular velocity as the electric angular velocity, the second-order current fluctuation, and the like. ing. Then, in a situation where the responsiveness for feedback control of the d-axis current idr and the d-axis current idr to the command currents id * and iq * is sufficient, the calculation is based on the d and q-axis currents idr and iqr including the fluctuation component. The command voltages Vd * and Vq * thus converted are converted into a three-phase command voltage V ¥ * by the three-phase converter 34. As a result, a DC component error (DC error) and a fluctuation component error (fluctuation error) having an amplitude corresponding to the amplitude of the angle error Δθ and an angular velocity twice as large as the electrical angular velocity (fluctuation error). It has been investigated by the present inventors that it is included in **.

  When the motor generator 10 is controlled based on the command voltage V ¥ * including the fluctuation error, the phase current includes a fluctuation component due to the fluctuation error, thereby causing torque fluctuation. Further, when the command voltage V ¥ * includes the variation error and the phase current includes a variation component due to the variation error, the first voltage based on the integrated value Vθ ¥ of the voltage absolute value | V ¥ | There is also a possibility that the calculation accuracy of the second correction coefficient KA ¥ based on the correction coefficient KV ¥ and the integrated value Δiθ ¥ of the current absolute value | i ¥ |

  In order to solve such a problem, first, error correction processing based on the integral value ΔVθ ¥ of the voltage absolute value | V ¥ | is preferentially performed in order to remove the influence of the angle error Δθ. Thereafter, an error correction process is performed based on the integral value Δiθ ¥ of the current absolute value | i ¥ |. As a result, the inconvenience caused by the fluctuation error is solved.

  FIG. 21 shows a procedure of error correction processing executed by the correction coefficient calculation unit 42. In the present embodiment, the initial values of the first correction coefficient KV ¥ and the second correction coefficient KA ¥ are set to “1”.

  In this series of processes, first, in step S70, an integral value ΔVθ ¥ over one electrical angle cycle of the voltage absolute value | V ¥ | (¥ = u, v, w) is calculated. Here, the integral value ΔVθ ¥ may be calculated by the process shown in FIG. 4 of the first embodiment.

  In the subsequent step S72, the first correction coefficient KV ¥ is calculated based on the integral value ΔVθ ¥, and is output to the first error correction units 54u, 54v, 54w. Here, the first correction coefficient KV ¥ may be calculated by the processing shown in steps S46 and S48 in FIG.

  In the subsequent step S74, the deviation of the amplitudes of the three-phase command voltage V ¥ * when the correction with the first correction coefficient KV ¥ is performed (the deviation of the amplitude between Vu * and Vv *, and the difference between Vv * and Vw *). It is determined whether or not the amplitude deviation between * and the amplitude deviation between Vw * and Vu * is less than a predetermined value γ (> 0). Here, in the present embodiment, specifically, the amplitude of the command voltage V ¥ * is, for example, an average value of the command voltage V ¥ * in a period that is an integral multiple of one electrical angle cycle and a command in one electrical angle cycle. The absolute value of the difference from the maximum or minimum voltage. The predetermined value γ may be set to a value slightly larger than 0 or set to a value close to 0.

  If an affirmative determination is made in step S74, the process proceeds to step S76, and an integral value Δiθ ¥ over one electrical angle period of the current absolute value | i ¥ | is calculated. Here, the integral value Δiθ ¥ may be calculated by the processing shown in FIG. 3 of the first embodiment.

  In the subsequent step S78, the second correction coefficient KA ¥ is calculated based on the integral value Δiθ ¥, and is output to the second error correction units 56u, 56v, 56w. Here, the second correction coefficient KA ¥ may be calculated by the processing shown in steps S34 and S36 in FIG.

  If a negative determination is made in step S74, or if the process in step S78 is completed, this series of processes is temporarily terminated.

  According to the error correction processing described above, first, the correction based on the integral value ΔVθ ¥ of the voltage absolute value | V ¥ | is performed, so that the U-phase current calculated values iu and V due to the variation error caused by the angle error Δθ. , Fluctuations in the W-phase current detection values iv and iw can be suppressed. Then, after the above-described fluctuation due to the angle error Δθ is suppressed, the correction based on the integral value Δiθ ¥ of the current absolute value | i ¥ | is performed, so that the U-phase current calculation values iu and V, W due to the gain error are performed. Variations in the detected phase current values iv and iw can be suppressed. Thereby, torque fluctuation of motor generator 10 caused by angle error Δθ and gain error can be suitably suppressed.

(Ninth embodiment)
Hereinafter, the ninth embodiment will be described with reference to the drawings with a focus on differences from the eighth embodiment.

  In the present embodiment, the correction based on the integral value ΔVθ ¥ (¥ = u, v, w) of the voltage absolute value | V ¥ | is prioritized over the correction based on the integral value Δiθ ¥ of the current absolute value | i ¥ |. To change.

  FIG. 22 shows a configuration of the correction coefficient calculation unit 42 according to the present embodiment.

  As shown in the figure, in the present embodiment, the first correction coefficient calculation unit 42a sets a target value of the first correction coefficient KV ¥ (hereinafter, referred to as a first target value KVtgt), and the first correction coefficient KV ¥ is feedback-controlled to the first target value KVtgt. Further, in the second correction coefficient calculation unit 42b, a target value of the second correction coefficient KA ¥ (hereinafter referred to as a second target value KAtgt) is set, and the second correction coefficient KA ¥ is set as the second target value KAtgt. Feedback control. Here, the transfer function in the feedback control of the first and second correction coefficient calculators 42a and 42b is a first-order lag element, and the time constant TV of the transfer function in the first correction coefficient calculator 42a is the second correction factor. It is set smaller than the time constant TA of the transfer function in the coefficient calculation unit 42b.

  FIG. 23 shows a procedure of error correction processing executed by the correction coefficient calculation unit 42. In FIG. 23, the same processes as those shown in FIG. 21 are given the same reference numerals for the sake of convenience. The initial values of the first correction coefficient KV ¥ and the second correction coefficient KA ¥ are set to “1”.

  In this series of processes, the first target value KVtgt is set in step S80 after the processes of steps S70 and S72 are completed. In the present embodiment, an intermediate value is set as the first target value KVtgt in the first correction coefficient KV ¥ calculated in the process of step S72.

  Thereafter, the process proceeds to step S82 via the processes of steps S76 and S78, and the second target value KAtgt is set. In the present embodiment, an intermediate value of the second correction coefficient KA ¥ calculated in the process of step S78 is set as the second target value KAtgt.

  In addition, when the process of step S82 is completed, this series of processes is once complete | finished.

  Since the time constant TV in the first correction coefficient calculation unit 42a is set smaller than the time constant TA in the second correction coefficient calculation unit 42b, according to the error correction process described above, the first correction coefficient The time required for KV ¥ to converge to the first target value KVtgt is shorter than the time required for the second correction coefficient KA ¥ to converge to the second target value KAtgt. That is, the correction based on the integral value ΔVθ ¥ (¥ = u, v, w) of the voltage absolute value | V ¥ | takes precedence over the correction based on the integral value Δiθ ¥ of the current absolute value | i ¥ |. . Such error correction processing can also suitably suppress the torque fluctuation of the motor generator 10 caused by the angle error Δθ and the gain error.

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

  The unifying means is not limited to positively unifying the signs of the current detection value and the command voltage, and may be negatively unified.

  The addition value calculation means is not limited to the one that calculates the exponential moving average value, but may be, for example, a simple moving average value of a plurality of past values (current detection value, command voltage).

  In the first embodiment, the integration interval may be a half cycle of the electrical angle θ. Further, the integration interval may be a period that is a period that is equal to or greater than a half cycle of the electrical angle θ and that is not synchronized with a period that is an integral multiple of the half cycle of the electrical angle θ.

  The selection means is not limited to a selection based on the electrical angular velocity ω. For example, the operation signal g ¥ # is generated by sine wave PWM control when the electrical angular velocity ω is less than a predetermined speed, and the operation signal g ¥ # is generated by overmodulation control when the electrical angular velocity ω is equal to or higher than the predetermined speed. In this configuration, the selection may be made based on the modulation factor (the amplitude of the command voltage V ¥ * / the input voltage VDC). Here, overmodulation control means that the command voltage V ¥ * is multiplied by a coefficient larger than “1” when the modulation factor becomes a specified value larger than “1” (eg, “1.15”). In this control, the command voltage V ¥ * is increased and corrected, and the predetermined speed is set lower as the required torque Tr of the motor generator 10 increases.

  -As a judgment means, it is not restricted to what was illustrated to said each embodiment. For example, the condition of the electrical angular velocity ω may be removed in step S61 in FIG. 13 of the fourth embodiment. For example, instead of the change amount ΔTr of the required torque Tr, a condition that the change amount of the command currents id * and iq * is larger than a specified value may be set.

  In FIG. 5 of the first embodiment, the first angular velocity ωA and the second angular velocity ωB are set to the same value, and the electrical angular velocity ω is equal to or higher than the second angular velocity ωB in the process of step S28. The process from step S26 to step S54 may be removed instead of the process for determining whether or not. Further, one or more of error correction processing based on the integral value ΔVθ ¥ of the voltage absolute value | V ¥ |, error correction processing based on the integral value Δiθ ¥ of the current absolute value | i ¥ |, and error correction processing based on both of them. A control logic that performs two may be employed.

  In the third embodiment, instead of the exponential moving average value of the amplitude amp, the voltage absolute value | V ¥ | and the current absolute value | i ¥ |, or the exponential moving average of the current square value and the voltage square value The value may be used for calculating the error correction coefficient Kt ¥. In this case, for example, the integration interval for calculating the voltage absolute value | V ¥ | etc. may be M electrical angle period (M: positive integer).

  In each of the above embodiments, the command current for controlling the output torque of the motor generator 10 to the required torque Tr is calculated. However, the present invention is not limited to this. For example, a command current for controlling the rotation speed of the motor generator 10 to the required speed may be calculated. In this case, in the process of step S61 of FIG. 13 in the fourth embodiment, a condition that the absolute value of the change amount of the required speed is larger than a predetermined value may be set instead of the condition of the required torque Tr. .

  The determining means is not limited to that exemplified in the eighth embodiment. For example, the weighted average processing value of the absolute value of the difference between the predicted current ide and the command current id * and the absolute value of the difference between the predicted current iqe and the command current iq * is expressed as the degree of divergence between the predicted current and the command current. It may be a parameter to be evaluated.

  In the eighth and ninth embodiments, only the error correction processing based on the integral value ΔVθ ¥ of the voltage absolute value | V ¥ | may be performed. That is, the second correction unit may not be provided. Even in this case, fluctuations in the current detection value and torque caused by the angle error Δθ can be suppressed.

  21. Whether the absolute value of the difference between the minimum value and the maximum value of the first correction coefficient KV ¥ of the U, V, and W phases is less than the predetermined value in the process of step S74 of FIG. 21 in the eighth embodiment. It may be replaced with a process for determining whether or not. This is because the absolute value of the difference between the minimum value and the maximum value of the first correction coefficient KV ¥ is less than a predetermined value, and the deviation between the amplitudes of the three-phase command voltage V ¥ * is less than the predetermined value γ. It is based on being able to relate to a situation.

  The angle error Δθ is not limited to a sinusoidal error having the same cycle as one electrical angle cycle of the motor generator 10, but may be an electrical angle half cycle or N electrical angle cycle (N: an integer of 2 or more) of the motor generator 10. It may be a sinusoidal error having a period. Even when such an angle error Δθ is included in the electrical angle θ, according to the error correction processing shown in the eighth and ninth embodiments, the fluctuation caused by the angle error Δθ due to the integrated value ΔVθ ¥. The error can be quantified.

  -About correction | amendment amount etc., as a correction method based on the comparison of the value of the phases of a rotary machine, not only what was illustrated by said each embodiment but the method demonstrated below, for example may be employ | adopted.

  In step S34 of FIG. 5 of the first embodiment, the maximum value is selected as the current reference value among the integral values Δiθ ¥ of the U, V, and W phase current absolute values | i ¥ |. In step S36, the second correction coefficient KA ¥ may be calculated by dividing the integral value Δiθ ¥ by the current reference value. In this case, the U phase current calculation value iu and the V and W phase current detection values iv and iw are corrected so that the amplitude of the fundamental wave becomes small. The same applies to the correction method based on the integral value ΔVθ ¥ of the voltage absolute value | V ¥ |.

  In the first embodiment, the second correction coefficient KA ¥ is calculated based on the ratio between the integral values Δiθ ¥ of the current absolute value | i ¥ |. However, the present invention is not limited to this. For example, the second correction coefficient KA ¥ may be calculated based on the difference between the integral values Δiθ ¥. Specifically, for example, the difference between the integral values of the remaining two phases with respect to the intermediate value among the integral values Δiθ ¥ is calculated. Then, as the difference with respect to the intermediate value is a positive value and the absolute value of the difference is larger, the second correction coefficient KA ¥ of the phase is set to a smaller value (where 0 <KA ¥ <1). The second correction coefficient KA ¥ for the phase may be set to a larger value as the difference with respect to the intermediate value is a negative value and the absolute value of the difference is larger (where KA ¥> 1). The same applies to the calculation method of the first correction coefficient KV ¥.

  In the first embodiment, the method of correcting the V-phase current detection value iv and the like by multiplying the error correction coefficient Kt ¥ by the V-phase current detection value iv and the like is adopted, but the present invention is not limited to this. For example, a method may be employed in which a correction value associated with the electrical angle θ is calculated based on the integral value Δiθ ¥, and the detected value is corrected by adding the calculated correction value to the current detection value or the like each time. .

  As a control method for calculating the command voltage based on the detected current value, for example, instantaneous current value control may be employed instead of the current vector control.

  -In the said 1st Embodiment, you may provide the current sensor which detects a U-phase electric current similarly to the said 7th Embodiment.

  The rotation angle sensor 16 is not limited to a resolver but may be a rotary encoder, for example. Even in this case, if the angle error Δθ is included in the electrical angle θ detected by the rotation angle sensor 16, the application of the error correction processing according to the eighth and ninth embodiments is effective.

  -The rotating machine is not limited to Y-connected, but may be Δ-connected.

  -The rotating machine is not limited to an embedded magnet synchronous motor. For example, a surface magnet synchronous motor or a field winding synchronous motor may be used. The rotating machine is not limited to a synchronous rotating machine but may be an induction rotating machine.

  -The rotating machine is not limited to a three-phase rotating machine, and may be, for example, a rotating machine having four or more phases. Further, the rotating machine is not limited to the rotating machine as the in-vehicle main machine, and may be a rotating machine that drives a compressor for air conditioning. Furthermore, the rotating machine is not limited to that mounted on the vehicle.

  DESCRIPTION OF SYMBOLS 10 ... Motor generator, 18v ... Current sensor, 18w ... Current sensor, 26u, 26v, 26w ... Error correction part, 36 ... Operation part, INV ... Inverter.

Claims (15)

Application to a control system for a rotating machine comprising a multiphase rotating machine (10) electrically connected to an inverter (INV) and current detecting means (18u, 18v, 18w) for detecting a phase current flowing through the rotating machine And
Correction means (26u, 26v, 26w, 54u, 54v, 54w, 56u, 56v, 56w) for correcting the detection value of the current detection means;
Command voltage calculation means for calculating a command voltage (V ¥ *; ¥ = u, v, w) to be applied to each phase of the rotating machine based on the detection value of the current detection means corrected by the correction means;
Operation means (36) for operating the inverter based on the detection value of the current detection means corrected by the correction means;
Defining at least one of the detection value of the current detection means and the command voltage calculated by the command voltage calculation means as a correction parameter;
The correction unit is configured to detect the current detection unit based on a comparison of values of the phases of the rotating machine with respect to the amplitude equivalent amount (Δiθ ¥, ΔVθ ¥) of the correction parameter or the amplitude (amp) of the correction parameter. A control device for a rotating machine, wherein a detection value is corrected. Unifying means for unifying the signs of the correction parameters;
An addition value calculating means for calculating an integral value or a moving average value of a plurality of past values using the electrical angle (θ) of the rotating machine as a variable for the correction parameter with the same sign Further comprising
2. The control device for a rotating machine according to claim 1, wherein the correction unit corrects a detection value of the current detection unit based on a comparison of the integral value or the moving average value of the phases of the rotating machine. .   3. The control device for a rotating machine according to claim 2, wherein the addition value calculation means calculates the integral value or the moving average value over a period obtained by multiplying a half cycle of the electrical angle by an integer.   3. The rotation according to claim 2, wherein the addition value calculation means calculates the integral value or the moving average value over a period obtained by multiplying one cycle of the electrical angle by N times (N is an integer of 2 or more). Machine control device. The correction means includes
Correction value calculation means for calculating a correction value of the detection value of the current detection means based on the comparison of the integral value or the moving average value of the phases;
Correcting the detection value of the current detection means based on the correction value calculated by the correction value calculation means;
Determining means for determining whether or not the rotating machine is in a transient operation;
Holding means for holding the previous correction value calculated by the correction value calculation means when it is determined by the determination means that it is during the transient operation;
The correction means corrects the detection value of the current detection means based on the correction value held by the holding means in a period determined by the determination means as the transient operation. The control apparatus of the rotary machine of any one of -4.   The unifying unit calculates an absolute value (| i \ |, | V \ |; \ = u, v, w) of the correction parameter or a square value of the correction parameter to calculate the correction parameter. The control device for a rotating machine according to any one of claims 2 to 5, wherein the reference numerals are unified. The correction means includes
Correction value calculation means for calculating a correction coefficient (Kt ¥; ¥ = u, v, w) of the detection value of the current detection means based on the comparison of the values of the phases;
The rotation according to claim 1, wherein the detection value is corrected by multiplying the detection value of the current detection unit by the correction coefficient calculated by the correction value calculation unit. Machine control device.   The correction value calculation means calculates the correction coefficient based on a ratio between a reference phase value and a residual phase value of the phase of the rotating machine for the amplitude equivalent amount or the amplitude. The control device for a rotating machine according to claim 7.   The correction means selects the detection value of the current detection means as the correction parameter based on the determination that the electrical angular velocity (ω) of the rotating machine is in a predetermined high angular velocity region, and the electrical angular velocity is predetermined. The control of the rotating machine according to any one of claims 1 to 8, further comprising selection means for selecting the command voltage as the correction parameter based on being determined to be in the low angular velocity region. apparatus. The control system further includes angle detection means (16) for detecting an electrical angle (θ) of the rotating machine,
The command voltage calculation means includes
Current conversion means (28) for converting the detection value of the current detection means corrected by the correction means into a current in a rotating coordinate system based on the detection value of the angle detection means;
Means (32) for calculating a command voltage of the rotating coordinate system based on the current of the rotating coordinate system;
Means (34) for converting the rotational coordinate system command voltage into a fixed coordinate system command voltage based on the detection value of the angle detection means;
The command voltage as the correction parameter is a command voltage of the fixed coordinate system,
10. The rotating machine according to claim 1, wherein the operating unit operates the inverter so that an applied voltage of each phase of the rotating machine is a command voltage of the fixed coordinate system. Control device. Prediction means (50) for predicting the current flowing through the rotating machine when the operation state of the inverter is temporarily set based on the detection value of the current detection means corrected by the correction means;
Based on the current predicted by the prediction means, further comprising a determination means (52) for evaluating an operation state corresponding to the predicted current and determining an operation state with a high evaluation as the operation state of the inverter;
The operation means operates the inverter so as to be in the operation state determined by the determination means,
10. The control device for a rotating machine according to claim 1, wherein the command voltage calculation unit calculates the command voltage based on the operation state determined by the determination unit. The control system further includes angle detection means (16) for detecting an electrical angle (θ) of the rotating machine,
The command voltage calculation means includes
Current conversion means (28) for converting the detection value of the current detection means corrected by the correction means into a current in a rotating coordinate system based on the detection value of the angle detection means;
Means (32) for calculating a command voltage of the rotating coordinate system based on the current of the rotating coordinate system;
Means (34) for converting the rotational coordinate system command voltage into a fixed coordinate system command voltage based on the detection value of the angle detection means;
The operation means operates the inverter to set the applied voltage of each phase as a command voltage of the fixed coordinate system,
The correction unit is a first correction unit that corrects the detected value of the current detection unit based on a comparison of values of the phases with respect to the amplitude equivalent of the command voltage in the fixed coordinate system or the amplitude of the command voltage. 42. The control device for a rotating machine according to claim 1, further comprising: 42a). The correction means is a second correction means for correcting the detection value of the current detection means based on the comparison of the values of the phases with respect to the amplitude equivalent of the detection value of the current detection means or the amplitude of the detection value. 42b)
13. The rotating machine control device according to claim 12, further comprising priority means for prioritizing correction by the first correction means over correction by the second correction means.   The priority unit starts correction by the second correction unit based on the fact that the deviation between the amplitudes of the command voltages of each phase becomes less than a predetermined value by performing correction by the first correction unit. The rotating machine control device according to claim 13. The first correction means includes
The correction value (KV ¥; ¥ = u, v, w) of the detection value of the current detection means based on the comparison of the values of the phases with respect to the amplitude equivalent of the command voltage in the fixed coordinate system or the amplitude of the command voltage First correction value calculating means for calculating
First feedback control means for feedback-controlling the correction value calculated by the first correction value calculation means to its target value (KVtgt),
Based on the correction value feedback-controlled by the first feedback control means, the detection value of the current detection means is corrected,
The second correction means includes
Second correction for calculating a correction value (KA ¥) of the detection value of the current detection unit based on the comparison of the values of the phases with respect to the amplitude equivalent amount of the detection value of the current detection unit or the amplitude of the detection value A value calculating means;
Second feedback control means for feedback-controlling the correction value calculated by the second correction value calculation means to its target value (KAtgt),
Based on the correction value feedback-controlled by the second feedback control means, the detection value of the current detection means is corrected,
14. The rotating machine according to claim 13, wherein the priority unit sets a time constant of feedback control by the first feedback control unit to be smaller than a time constant of feedback control by the second feedback control unit. Control device.