JP2017038444A - Control device for AC rotary electric machine - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an AC rotary electric machine that improves responsiveness to a torque command value.SOLUTION: A control device for an AC rotary electric machine comprises: a selecting section that selects phase current values at an arithmetic angle in a predetermined range equal to or longer than an electric angle of 1/3 period and shorter than an electric angle of 1/3 period, for the phase current values of three phases; a synthesizing section that uses one of three phases as a reference phase, superimposes the selected phase current value in the reference phase and a selected phase current value in another phase with a shift at an electric angle of 120°, and uses each of the superimposed phase current values as a phase current value for the reference phase; a coefficient calculating section that calculates a Fourier coefficient that integrates calculated values based on each phase current value of the reference phase for an electric angle of one period for each arithmetic angle; a primary current calculating section that calculates a primary current value from the calculated Fourier coefficient; a dq current calculating section that calculates dq current by converting the calculated primary current value; and an operating section that operates an inverter such that the amount to be controlled based on the calculated dq current follows a target value.SELECTED DRAWING: Figure 4

Description

  The present invention relates to a control device for an AC rotating electrical machine that controls energization of the AC rotating electrical machine based on a detected value of a phase current detected by a current sensor.

  Conventionally, in control of an AC rotating electrical machine, a phase current value is detected, and a control amount based on the detected phase current value is fed back to a target value. In such feedback control, when a high-order component is superimposed on the phase current, the high-order component is also superimposed on the operation command for the switching element of each phase of the inverter as a result of the feedback control. As a result, noise components such as switching noise also contain higher order components, which causes a problem in terms of quietness. Further, if the phase current is offset, torque fluctuation and output fluctuation occur, which is not desirable.

  Therefore, the control device for an AC motor described in Patent Document 1 expands the phase current detection value as a function of the electrical angle and expands the Fourier series, detects the primary current that is the primary component of the phase current, and detects the detected primary current. Based on the feedback control. Specifically, the control device integrates a calculated value based on the phase current detection value at the integrated angle over one electrical angle period in an integrated angle set by dividing the electrical angle k period into N, thereby obtaining a Fourier. The coefficient is calculated.

JP 2014-132815 A

  The control device detects a primary current based on a phase current detection value at an integrated angle over one electrical angle cycle. Therefore, when the torque command value is changed, the primary current drags the phase current before the torque command value is changed from when the torque command value is changed until one cycle of the potential angle elapses. Therefore, a decrease in responsiveness when changing the torque command value may be a problem. In particular, in the low rotational speed region, the time for the primary current to drag the phase current before the torque command value is changed is longer than that in the high rotational speed region. .

  In view of the above circumstances, it is a main object of the present invention to provide a control device for an AC rotating electrical machine that can improve the response to a torque command value.

  The invention according to claim 1 is a three-phase AC rotating electrical machine, an inverter that drives the AC rotating electrical machine, and a current sensor that detects phase current values of two or more phases of the AC rotating electrical machine. A control device for an AC rotating electrical machine that is applied to a rotating electrical machine system that includes a calculation angle setting unit that divides one cycle of three-phase electrical angle into N pieces (N is a natural number) and sets a calculation angle at the same timing. A selection unit for selecting the phase current values at the calculation angles over a predetermined range of electrical angle 1/3 cycle or more and less than electrical angle 1/2 cycle for the three-phase phase current values; One of the three phases is set as a reference phase, and the selected phase current value in a phase other than the reference phase of the three phases is set to an electrical angle of 120 with respect to the selected phase current value in the reference phase. Shift and overlay And combining a calculated value based on each phase current value of the reference phase over one cycle of the electrical angle for each of the calculation angles. A coefficient calculation unit that calculates a Fourier coefficient, a primary current calculation unit that calculates a primary current value based on the calculated primary component of the Fourier coefficient, and a dq conversion of the calculated primary current value. A dq current calculation unit that calculates a d-axis current value and a q-axis current value, and a control amount based on the calculated d-axis current value and the q-axis current value so as to follow a target value of the control amount. And an operation unit for operating a plurality of switching elements constituting the inverter.

  According to the first aspect of the invention, the calculation angle is set by dividing one period of the three-phase electrical angle into N pieces. For the three-phase current values, the phase current values at the calculation angles over a predetermined range of an electrical angle of 1/3 cycle or more and less than an electrical angle 1/2 cycle are selected. Then, the selected phase current value in the other phase other than the reference phase is overlapped with the selected phase current value in the reference phase shifted by an electrical angle of 120 °, and the superimposed phase current values are the reference phase values. Phase current value. Further, for each calculation angle, calculated values based on the phase current values of the reference phase are integrated over one electrical angle cycle, and a Fourier coefficient is calculated. Based on the calculated primary component of the Fourier coefficient, a primary current value of each phase is calculated, and the calculated primary current value is converted into a dq current value. Then, the switching element of the inverter is operated so that the control amount based on the dq current value follows the target value.

  The phase current values of the respective phases have a relationship that the electrical angle is shifted by 120 °. Therefore, the phase current value of the other phase shifted by an electrical angle of 120 ° with respect to the phase current value of the reference phase can be regarded as the phase current value of the reference phase. Therefore, a phase current value over one electrical angle cycle of the reference phase can be obtained from a three-phase phase current value over a predetermined range of electrical angle 1/3 period or more and less than electrical angle 1/2 period. That is, the Fourier coefficient can be calculated from the phase current value of less than 1/2 electrical angle cycle. Therefore, when the torque command value is changed, the period in which the primary current value drags the phase current value before the torque command value change is compared to the case where the Fourier coefficient is calculated from the detected value of the phase current value over one electrical angle cycle. Can also be shortened. Therefore, the responsiveness to the torque command value can be improved.

  The invention according to claim 2 is a three-phase AC rotating electrical machine, an inverter that drives the AC rotating electrical machine, a current sensor that detects phase current values of two or more phases of the AC rotating electrical machine, and , A control device for an AC rotating electrical machine that is applied to a rotating electrical machine system that divides one cycle of a three-phase electrical angle into N pieces (N is a natural number) and sets a computation angle at the same timing A setting unit, a selection unit that selects each of the phase current values in the calculation angle over a predetermined range of an electrical angle of 1/6 cycle or more and less than an electrical angle of 1/4 cycle for the three-phase phase current values; A translation inversion unit that performs translation inversion for shifting the current value by an electrical angle of 180 ° and inversion with respect to a central axis passing through the center in the amplitude direction of the phase current, and using one of the three phases as a reference phase, the reference Selected before in phase The phase current value, the phase current value obtained by translationally inverting the selected phase current value in the reference phase, and the selected phase current value in a phase other than the reference phase of the three phases are electrically A combining unit that uses a phase current value shifted by 120 ° and a phase current value obtained by translationally reversing the phase current value shifted by 120 ° in the electrical angle, for each calculation angle; A coefficient calculation unit that calculates a Fourier coefficient by integrating calculated values based on each phase current value of the reference phase over one period of the electrical angle, and a primary current based on the calculated primary component of the Fourier coefficient A primary current calculation unit that calculates a value, a dq current calculation unit that calculates a d-axis current value and a q-axis current value by performing dq conversion on the calculated primary current value, and the calculated d-axis current value And a control amount based on the q-axis current value is So as to follow the target value, and an operation unit for operating a plurality of switching elements constituting the inverter.

  According to the second aspect of the present invention, the calculation angle is set by dividing one period of the three-phase electrical angle into N pieces. For the three-phase current values, the phase current values at the calculation angles over a predetermined range of an electrical angle of 1/6 cycle or more and less than an electrical angle 1/4 cycle are selected. Then, the selected phase current value of the reference phase, the phase current value obtained by translationally inverting the selected phase current value of the reference phase, and the phase current value shifted by 120 ° in the selected phase current value of the other phase. The current value and the phase current value obtained by translationally reversing the phase current value shifted by 120 ° in the electrical angle of the other phase are set as the reference phase phase current value. Further, for each calculation angle, calculated values based on the phase current values of the reference phase are integrated over one electrical angle cycle, and a Fourier coefficient is calculated. Based on the calculated primary component of the Fourier coefficient, a primary current value of each phase is calculated, and the calculated primary current value is converted into a dq current value. Then, the switching element of the inverter is operated so that the control amount based on the dq current value follows the target value.

  The phase currents of the respective phases have a relationship that is deviated by 120 ° in electrical angle. Therefore, the phase current value of the other phase shifted by an electrical angle of 120 ° with respect to the phase current value of the reference phase can be regarded as the phase current value of the reference phase. In addition, if the magnitude of the voltage applied to the AC rotating electric machine is the same, a value obtained by translationally inverting a certain phase current value can be regarded as a phase current value at an operation angle shifted by 180 ° from the phase current value. Therefore, a phase current value over one electrical angle cycle of the reference phase can be obtained from a three-phase phase current value over a predetermined range of 1/6 electrical angle or more and less than 1/4 electrical angle. That is, the Fourier coefficient can be calculated from the phase current value less than the electrical angle quarter cycle. Therefore, when the torque command value is changed, the period in which the primary current value drags the phase current value before the torque command value change is compared to the case where the Fourier coefficient is calculated from the detected value of the phase current value over one electrical angle cycle. Can also be shortened.

The figure which shows schematic structure of the motor system which concerns on each embodiment. The figure explaining the control mode of an AC motor. The figure which shows the correspondence of the operation state of an AC motor, and control mode. The block diagram which shows the structure of the control apparatus of the AC motor corresponding to the PWM control mode which concerns on 1st Embodiment. The figure which shows an example of the calculation angle | corner of a Fourier coefficient. The figure which shows the calculation object of the (a) Fourier coefficient based on the past, and (b) the calculated primary current value. The figure which shows the calculation object of the (a) Fourier coefficient which concerns on 1st Embodiment, (b) the phase current value over the synthesized 1 period, and (c) the calculated primary current value. (A) A selected phase current value in the V phase, (b) a selected phase current value in the W phase, (c) a selected phase current value in the U phase, and (d) a combined phase current value. . The flowchart which shows the process sequence which calculates dq electric current which concerns on 1st Embodiment. (A) The time chart which shows the characteristic of the rotation position with respect to time, (b) The characteristic of the phase current of the V phase with respect to time. The block diagram which shows the structure of the control apparatus of the AC motor corresponding to the rectangular wave control mode which concerns on 1st Embodiment. The block diagram which shows the structure of the control apparatus of the alternating current motor corresponding to the PWM control mode which concerns on 2nd Embodiment. The block diagram which shows the structure of the control apparatus of the AC motor corresponding to the rectangular wave control mode which concerns on 2nd Embodiment. The figure which shows (a) voltage and (b) phase current. (A) Selected phase current value in V phase, (b) Selected phase current value in W phase, (c) Selected phase current value in U phase, (d) Selected phase current in each phase The phase current value obtained by synthesizing the values, (e) the synthesized phase current value, and the phase current value obtained by translation inversion. The flowchart which shows the process sequence which calculates dq electric current which concerns on 2nd Embodiment. (A) selected phase current value in V phase, (b) selected phase current value in W phase, (c) selected phase current value in U phase, (d) synthesized when there is an offset error The figure which shows a phase current value. The figure which shows the detected value of the phase current value containing an offset error. (B) The figure which shows the estimated value estimated from the detected value of the phase current value with an offset error. The block diagram which shows the function which correct | amends a phase current value with an offset error. The flowchart which shows the process sequence which calculates dq electric current which concerns on 3rd Embodiment.

  Hereinafter, embodiments embodying a control device for an AC rotating electrical machine will be described with reference to the drawings. The control device for an AC rotating electrical machine according to each embodiment is assumed to be applied to a motor system that drives a hybrid vehicle. In the following embodiments, parts that are the same or equivalent to each other are denoted by the same reference numerals in the drawings, and the description of the same reference numerals is used.

(First embodiment)
First, a motor system according to the present embodiment will be described with reference to FIG. The motor system (rotating electrical machine system) according to this embodiment includes a motor 30, an inverter 20, a DC power supply 10, current sensors 31 and 32, a rotational position sensor 33, and a control device 40. Torque command value Trq * is input to control device 40 from vehicle control device 100.

  MG30 (AC rotating electrical machine) includes a stator and a rotor around which a three-phase coil of U phase, V phase, and W phase is wound, and has a three-phase function as a motor and a function as a generator. It is a motor generator. The MG 30 is a travel drive source for the hybrid vehicle. MG30 may be a three-phase motor having only a function as an electric motor.

  The inverter 20 is a bridge circuit in which three series bodies in which two switching elements are connected vertically are connected in parallel. As the switching element, for example, a gate drive type IGBT, a MOS transistor, or the like can be employed. The operation signals gua, gub, gva, gvb, gwa, gwb transmitted from the control device 40 to be described later are input to the gate terminals of the respective switching elements, whereby the on / off states of the respective switching elements are operated. As a result, AC voltages Vu, Vv, Vw are applied to the coils of each phase of MG30, and driving of MG30 is controlled.

  The DC power supply 10 is, for example, a lithium ion or nickel hydride secondary battery, and exchanges power with the MG 30 via the inverter 20. The DC voltage of the DC power supply 10 becomes the system voltage VH input to the inverter 20. When a boost converter is connected between the DC power supply 10 and the inverter 20, the DC voltage boosted by the boost converter becomes the system voltage VH input to the inverter 20.

  The current sensor 31 is provided in the V-phase coil, detects the V-phase current Iv at a predetermined cycle, and transmits the detected value of the phase current Iv to the control device 40. The current sensor 32 is provided in the W-phase coil, detects the W-phase current Iw at a predetermined cycle, and transmits the detected value of the phase current Iw to the control device 40. Here, since the sum of the three-phase phase currents is always zero according to Kirchhoff's law, if the V-phase current Iv and the W-phase current Iw are known, the remaining U-phase current Iu is obtained by calculation. It is done. In the present embodiment, the current sensors 31 and 32 are provided in the two phases of the V phase and the W phase, but the current sensors 31 and 32 may be provided in any two of the three phases.

  The rotational position sensor 33 is provided near the rotor of the MG 30, detects the rotational position θ of the rotor, and transmits the detected rotational position θ to the control device 40. The rotational position θ is a value represented by an electrical angle, and the rotational speed Nr of the MG 30 is calculated from the rotational position θ. As the rotational position sensor 33, a resolver, an encoder, or the like can be employed.

  The vehicle control device 100 is a control device that controls the entire hybrid vehicle above the control device 40, and is mainly configured by a microcomputer including a CPU, a ROM, a RAM, an I / O, and the like. The vehicle control device 100 detects the driving state of the vehicle based on detection values of various sensors such as an accelerator sensor, a brake sensor, and a shift position sensor. Then, vehicle control device 100 calculates a torque command value Trq * corresponding to the detected driving state of the vehicle, and transmits the calculated torque command value Trq * to control device 40.

  The control device 40 (control device for an AC rotating electrical machine) is mainly configured by a microcomputer including a CPU, a ROM, a RAM, an I / O, and the like. The control device 40 realizes various functions to be described later by software processing for executing a program stored in the ROM by the CPU and hardware processing by a dedicated electric circuit, and controls driving of the MG 30.

  As shown in FIG. 2, the control device 40 selects and implements one of the three control modes according to the modulation factor m as the control mode of the MG 30. The modulation factor m is defined by the following equation (1). Vr is the magnitude of the voltage vector applied to the MG 30 on the rotational coordinate plane.

  When the modulation factor m is less than 0 to less than 1.27, the control device 40 performs the sine wave PWM control mode and the overmodulation PWM control mode. The sine wave PWM control mode is a mode in which the switching element of each phase of the inverter 20 is controlled based on a comparison between the voltage command value of each phase of the sine wave and the carrier wave, and the amplitude of the sine wave voltage command is that of the carrier wave. It is performed in the range below the amplitude. On the other hand, in the overmodulation PWM control mode, PWM control similar to the sine wave PWM control mode is performed in a range where the amplitude of the sine wave component of the voltage command is larger than the amplitude of the carrier wave. In the sine wave PWM control mode and the overmodulation PWM control mode, “current feedback control” is performed in which the amplitude and phase of the voltage applied to the MG 20 are controlled by feedback of the current flowing through the MG 30.

  Further, when the modulation factor m = 1.27, the control device 40 executes the rectangular wave control mode. The rectangular wave control mode is a mode in which one pulse of a rectangular wave is applied to the MG 30 with a ratio of the upper arm on period and the lower arm on period being 1: 1 within a certain period. In the rectangular wave control mode, since the amplitude of the voltage applied to the MG 30 is fixed, “torque feedback control” that controls the phase of the rectangular wave voltage pulse based on the deviation between the torque command value Trq * and the estimated torque value Trqe. To implement.

  FIG. 3 shows the correspondence between the operating state of the MG 30 and each control mode. In the low rotational speed region I, the sine wave PWM control mode is performed and torque fluctuation is suppressed. In the middle rotational speed range II where the rotational speed Nr of the MG 30 increases and the system voltage VH exceeds the maximum value, the control mode is switched from the sine wave PWM control mode to the overmodulation PWM control mode, and the output in the middle rotational speed range II Raise. Further, when the rotation speed Nr of the MG 30 increases, the overmodulation PWM control mode is switched to the rectangular wave control mode, and the output in the high rotation speed region III is increased. Hereinafter, the PWM control mode including the sine wave PWM control and the overmodulation PWM control mode and the rectangular wave control mode will be described.

<PWM control>
First, the PWM control mode will be described. In the PWM control mode, since current feedback control is performed, the feedback control amount is a current value. FIG. 4 shows a configuration of the control device 40 corresponding to the PWM control mode. The control device 40 includes a current command generation unit 41, an FF term calculation unit 42, deviation calculation units 43 and 44, PI control units 45 and 46, addition units 47 and 48, a two-phase three-phase conversion unit 49, an rφ conversion unit 50, The functions of the voltage DUTY conversion unit 51 and the PWM unit 52 are provided. Further, the control device 40 includes functions of a selection unit 53, a synthesis unit 54, a Fourier coefficient calculation unit 55, a primary current calculation unit 56, a dq current calculation unit 57, and a calculation angle setting unit 71.

  The current command generator 41 generates current command values Id * and Iq * of the rotating coordinate system based on the torque command value Trq * and the current command map. The current command map is a map showing the correspondence between the torque command value Trq * and the current command values Id * and Iq *, and is stored in advance in a storage device. In the PWM control mode, the current command values Id * and Iq * correspond to the target value of the control amount.

  The FF term calculation unit 42 calculates feed-forward terms Vd_ff and Vq_ff of the d-axis voltage command value and the q-axis voltage command value based on the generated current command values Id * and Iq *, respectively (Japanese Patent Laid-Open No. 2014-2014). No. 132815). The deviation calculation units 43 and 44 represent the current deviations ΔId and ΔIq between the generated current command values Id * and Iq * and the actual currents Id and Iq of the rotating coordinate system calculated by the dq current calculation unit 57 described later. Calculate each. The PI control units 45 and 46 calculate the d-axis voltage command value and the feedback terms Vd_fb and Vq_fb of the q-axis voltage command value by proportional integral control so that the calculated current deviations ΔId and ΔIq converge to 0, respectively. . Adders 47 and 48 add feedforward terms Vd_ff and Vq_ff and feedback terms Vd_fb and Vq_fb, respectively, to calculate voltage command values Vd * and Vq * of the rotating coordinate system, respectively. The actual currents Id and Iq (the actual currents Id and Iq are collectively referred to as dq current) correspond to the control amount.

  The two-phase / three-phase converter 49 uses the rotational position θ received from the rotational position sensor 33 to calculate the calculated voltage command values Vd * and Vq * of the rotational coordinate system and the voltage command values Vu * and Vu * of the fixed coordinate system. Convert to Vv *, Vw *. The rφ converter 50 converts the voltage command values Vd * and Vq * in the rotating coordinate system into a command voltage vector having a magnitude Vr on the rotating coordinate plane and a phase φq with reference to the q axis. The phase φq is defined as positive in the counterclockwise direction from the q axis.

  The voltage DUTY conversion unit 51 converts the voltage command values Vu *, Vv *, Vw * of each phase into command duties Du, Dv, Dw based on the rotational position θ, the calculated phase φq, and the modulation factor m. . The modulation factor m is calculated by multiplying the calculated vector Vr, the system voltage VH, and a coefficient. The PWM unit 52 generates operation signals gua, gub, gva, gvb, gwa, gwb based on the command duties Du, Dv, Dw, and transmits the generated operation signals to the switching elements of the inverter 20. Each operation signal is a signal for controlling on / off of each switching element of the inverter 20. In the PWM control mode, the functions from the FF term calculation unit 42 to the PWM unit 52 described above correspond to the operation unit.

  By turning on / off each switching element with an operation signal, three-phase AC voltages Vu, Vv, Vw are applied to the coils of each phase of MG30. Along with the application of the three-phase AC voltages Vu, Vv, Vw, torque corresponding to the torque command value Trq * is output from the MG 30, and phase currents Iu, Iv, Iw flow through the coils of each phase of the MG 30.

  The control device 40 detects the phase currents Iu, Iv, Iw of each phase, calculates the actual currents Id, Iq, and feeds back the actual currents Id, Iq to the current command values Id *, Iq *. At this time, high-order components may be superimposed on the phase currents Iu, Iv, and Iw, or the phase currents Iu, Iv, and Iw may be offset. When the actual currents Id and Iq are calculated from the phase currents Iu, Iv, and Iw on which the higher-order components are superimposed, the higher-order components are also superimposed on the command duties Du, Dv, and Dw, and noise components such as switching noise of the inverter 20 associated therewith Also include higher order components. In particular, in the overmodulation PWM control mode, since the high-order component is included in the PWM pulse, there is a problem that the high-order component is easily superimposed on the phase current value and noise is increased. Further, when the phase current value is offset, torque fluctuation and power fluctuation occur.

  Therefore, the control device 40 expands the phase current value as a function of the rotational position θ and develops a Fourier series. Formulas (2) and (3) show general formulas for Fourier series expansion. Expression (2) is an expression for calculating the n-th order Fourier coefficient an of the cos function and the n-order Fourier coefficient bn of the sin function. F (θ) shown in Expression (3) is a periodic function having a period 2π with θ as a variable, and is represented by the sum of trigonometric functions. Note that the letter “n” in the formulas (2) and (3) is used independently only by these formulas and is different from “n” used in other places.

  In equation (3), a0 / 2, which is a zeroth-order component, is a direct current component and corresponds to the offset amount at the center of the current amplitude with respect to the current 0 (A). The second and higher order components correspond to higher order components. Therefore, the control device 40 extracts only the primary component of the phase current, calculates the actual currents Id and Iq from the primary current value that is the primary component of the extracted phase current, and performs current feedback control.

  In general, the Fourier coefficient is calculated based on N phase current values at calculation angles set by dividing one period of electrical angle into N (N is a natural number). For example, as shown in FIG. 5, when a calculation angle is set by dividing one period of electrical angle into 12 parts, in a certain calculation angle, 12 of the calculation angle and 11 calculation angles immediately before the calculation angle. The Fourier coefficient is calculated based on the phase current value. Each calculation angle becomes the calculation start timing of the Fourier coefficient. In the present embodiment, it is assumed that the detection timings of the phase currents Iv and Iw by the current sensors 31 and 32 are synchronized with the calculation angle that is the calculation start timing.

  FIG. 6A shows a range of phase current values to be calculated in a conventional Fourier coefficient calculation process. At time points t0, t1, t2, and t3, the interval between adjacent time points is an electrical angle half cycle. Conventionally, based on the equation (2), the calculated value based on the phase current value at the calculation angle over one electrical angle cycle is integrated over one electrical angle cycle to calculate the Fourier coefficient. FIG. 6B shows the primary current value calculated from the primary component of the calculated Fourier coefficient.

  Conventionally, when the torque command value Trq * is changed, the Fourier coefficient, and thus the primary current value, before the torque command value Trq * is changed from when the torque command value Trq * is changed until one electrical angle cycle elapses. It is influenced by the phase current value. For example, as shown in FIG. 6A, when the torque command value Trq * is changed at time t1, and the dq current at time t2 is calculated, the phase current value from time t0 to time t1 before the command value change, The primary current value is calculated based on the phase current value from the time point t1 to the time point t2 after the command value is changed. Therefore, the calculated primary current value has a relatively large error with respect to the actual primary current value at time t2. As a result, the actual torque response to the torque command value Trq * is reduced.

  In FIG. 6A, when the command value is changed at time t1, the calculated primary current value is dragged from the phase current value before the command value change until the dq current is calculated at time t3. Become. In particular, in the low rotational speed range, the time required for the rotor to rotate by one electrical angle period is longer than in the high rotational speed range, and therefore, when the torque command value Trq * is changed, the primary current value is commanded. The time for dragging the phase current value before the value change becomes longer. Therefore, particularly in the low rotation speed region, a decrease in responsiveness when the torque command value Trq * is changed becomes a problem, and an improvement in responsiveness when the torque command value is changed is desired.

  Here, the phase currents of the phases are in a relationship of being shifted from each other by an electrical angle of 120 °. For example, it is assumed that the U-phase phase current Iu is shifted by 120 ° in the electrical angle direction and the W-phase phase current Iw is shifted by 120 ° in the retarded direction with respect to the V-phase current Iv. In this case, the U-phase phase current Iu shifted by 120 ° in the retarded direction is equal to the V-phase phase current Iv, and the W-phase phase current Iw shifted by 120 ° in the advanced angle direction is the V-phase phase current Iv. It becomes equal to the phase current Iv.

  Therefore, in each phase, control device 40 selects phase current values Ivs, Iws, and Ius at calculation angles over one-third cycle of electrical angle. Then, the control device 40 uses the V phase of the three phases as a reference phase, and converts the phase current values Ius and Iws selected in the other two phases to an electrical angle of 120 ° with respect to the phase current value Ivs selected in the reference phase. I decided to overlap and superimpose. The calculation angle over the electrical angle 1/3 cycle is the calculation angle of the same timing in three phases. FIG. 7A shows a range of phase current values of each phase selected as a calculation target in the Fourier coefficient calculation processing according to the present embodiment, and FIG. 7B shows the selected three-phase phase current values. The phase current value obtained by synthesizing is shown. Each phase current value superimposed on the reference phase can be regarded as a phase current value over one electrical angle cycle of the reference phase. FIG. 7C shows the primary current value calculated based on the phase current value over one electrical angle cycle of the combined reference phase.

  In this way, by synthesizing the phase current values over one-third cycle of the electrical angle of each phase and calculating the phase current value over one cycle of the electrical angle of the reference phase, when the torque command value Trq * is changed, 1 The period during which the next current value drags the phase current value before the change of the torque command value Trq * is shortened to 1/3 of the conventional value. For example, as shown in FIG. 7A, when the torque command value Trq * is changed at the time point t1 and the dq current at the time point t2 is calculated, the phase of each phase in the period from the time point t2 to 1/3 cycle before. Based on the current value, the primary current value is calculated. Therefore, when calculating the dq current at time t2, the primary current value does not drag the phase current value before the command value change. Therefore, in the present embodiment, compared to the prior art, a decrease in the actual torque response to the torque command value Trq * when the torque command value Trq * is changed is suppressed. Hereinafter, a method for calculating the primary current value in the present embodiment will be described. In this embodiment, the V phase is described as a reference phase, but the reference phase may be any phase.

  The calculation angle setting unit 71 divides one cycle of the electrical angle of the three-phase current into N pieces (N is a natural number), and sets the calculation angle at the same timing.

  The selector 53 selects a phase current value at a calculation angle over a period of 1/3 electrical angle for each of the three-phase phase current values. FIGS. 8A to 8C show diagrams in which the phase current values at the calculation angles ranging from 0 ° to 120 ° are selected. A selected phase current value in the V phase is indicated by a black circle, and an unselected phase current value is indicated by a white circle. A selected phase current value in the W phase is indicated by a black square, and an unselected phase current value is indicated by a white square. The selected phase current value in the U phase is indicated by a black triangle, and the unselected phase current value is indicated by a white triangle.

  The combining unit 54 uses the V phase as a reference phase and superimposes the phase current value Ius selected in the U phase on the phase current value Ivs selected in the V phase while shifting the electrical angle by 120 ° in the retard direction. The phase current value Iws selected in the phase is superposed with an electrical angle of 120 ° shifted in the advance direction. Then, the phase current value for three phases synthesized with the V phase as a reference is set as the phase current value in one cycle of the electrical angle of the V phase. FIG. 8D shows the selected phase current values of the W phase and the U phase shown in FIGS. 8B and 8C with respect to the selected phase current value of the V phase shown in FIG. It is the figure which synthesize | combined each shifted | deviated the electrical angle 120 degrees in the advance angle direction and the retard angle direction.

  The Fourier coefficient calculation unit 55 (coefficient calculation unit) calculates the Fourier coefficient by integrating the calculated values based on the phase current values of the reference phase synthesized by the synthesis unit 54 over one electrical angle period for each calculation angle. .

  The primary current calculation unit 56 calculates the primary current values Iu_1, Iv_1, and Iw_1 for each phase based on the Fourier coefficient calculated by the Fourier coefficient calculation unit 55. Specifically, a V-phase primary current value Iv_1 is calculated from the calculated primary coefficients a1 and b1 of Fourier coefficients. Then, the V-phase primary current value Iv_1 is shifted by 120 ° in the advance direction to calculate the U-phase primary current value Iu_1, and the V-phase primary current value Iv_1 is shifted by 120 ° in the retardation direction to shift the W-phase. The primary current value Iw_1 is calculated. Details of the calculation of the Fourier coefficient and the calculation of the primary current value of each phase will be described later.

  The dq current calculation unit 57 converts the calculated three-phase primary current values Iu_1, Iv_1, and Iw_1 into an actual current Id (d-axis current value) and an actual current Iq (q-axis current value) of the rotary seat system. To do.

  Next, a method for calculating the dq current from the detected value of the phase current value will be described with reference to the flowchart of FIG. 9 and the time chart of FIG.

  10A shows the characteristic of the rotational position θ with respect to time, and FIG. 10B shows the characteristic of the V-phase phase current Iv (θ) with respect to time. In FIG. 10A, the rotational position θ increases linearly in proportion to the time in the range of 0 to 360 °, and changes with time in a sawtooth shape that reaches 360 ° and returns to 0 ° at the same time. The vertical axis indicates the (n−1) th, nth, and (n + 1) th calculation angles θ [n−1], θ [n], and θ [n + 1]. The sample interval, which is the interval between consecutive calculation angles, is constant at Δθ = 360 / N, and is an interval obtained by equally dividing one period of electrical angle by N. Therefore, in one electrical angle cycle, calculation angles from calculation angles θ [1] to θ [N] (= θ [0]) are set. In one continuous electrical angle cycle, the calculation angle θ [0] of the subsequent cycle is the same rotational position θ as the calculation angle θ [N] of the previous cycle. N calculation angles are set as calculation angles for one electrical angle cycle.

  Hereinafter, the processing procedure for calculating the dq current from the detected value of the phase current value according to the present embodiment will be described with reference to the flowchart of FIG. This processing procedure is repeatedly executed by the control device 40 for each calculation angle.

  First, the value of n is incremented by 1 from the value of n at the previous processing (S10). Then, it is determined whether n is larger than N (S11). When n is N or less (S11: NO), the process proceeds to S13 as it is. When n is larger than N (S11: YES), n = 1 is set (S12), and the process proceeds to S13.

  Subsequently, the V-phase phase current value Iv (θ [n]) detected by the current sensor 31 and the W-phase phase current value Iw (θ [n]) detected by the current sensor 32 are acquired (S13). ). The phase current value Iv (θ [n]) and the phase current value Iw (θ [n]) are the V-phase phase current value and the W-phase phase current value at the calculation angle θ [n], respectively.

  Subsequently, using the phase current value Iv (θ [n]) and the phase current value Iw (θ [n]) acquired in S13, the phase current value of the U phase at the calculation angle θ [n] according to Kirchoff's law. Iu (θ [n]) is calculated (S14).

  Subsequently, a phase current value over a period of 1/3 of the electrical angle of each phase is selected (S15). Specifically, the phase current value of each phase at the calculation angles θ [n], θ [n−1], θ [n−2]... Over the most recent electrical angle 1/3 period before the calculation angle θ [n]. Select.

  Subsequently, with the V phase as the reference phase, the phase current value selected in the W phase and the phase current value selected in the U phase in the advance direction and the retard direction, respectively, with respect to the phase current value selected in the V phase. The electrical angle is shifted by 120 ° and synthesized (S16).

  Subsequently, first-order Fourier coefficients a1 and b1 are calculated using the equation shown in FIG. 9 for the phase current value of the V phase over one electrical angle cycle synthesized in S16. The expression shown in FIG. 9 is an expression obtained by discretizing Expression (2) and Expression (3). The phase current Iv (θ [n]) (n = 1, 2,... N) is the selected phase current value in the V phase, the phase current value in the U phase synthesized in the V phase, and the W in the V phase. Combined with the phase current value of the phase. The calculation angle θ [n] (n = 1, 2,... N) is obtained by sequentially selecting the selected calculation angle θ [n] over a period of 1/3 electrical angle and the selected calculation angle θ [n] ± 120 °. It will be a side by side. Iv (θ [n]) · cos θ [n] · (θ [n] −θ [n−1]) is integrated from n = 1 to N to calculate the Fourier coefficient a1. Further, Iv (θ [n]) · sin θ [n] · (θ [n] −θ [n−1]) is integrated from n = 1 to N to calculate the Fourier coefficient b1. Then, the V-phase primary current value Iv_1 (θ [n]) is calculated from the calculated Fourier coefficients a1 and b1 (S17).

  In the present embodiment, Iv (θ [n]) (n = 1, 2,... N) corresponds to each phase current value of the reference phase. Further, Iv (θ [n]) · cos θ [n] · (θ [n] −θ [n−1]) and Iv (θ [n]) · sin θ [n] · (θ [n] −θ [n-1]) corresponds to a calculated value based on each phase current value of the reference phase.

  Subsequently, the V-phase primary current Iv_1 (θ [n]) calculated in S17 is shifted by an electrical angle of 120 ° in the advance direction to calculate the U-phase primary current Iv_1 (θ [n]). Further, the W-phase primary current Iw_1 (θ [n]) is calculated by shifting the V-phase primary current Iv_1 (θ [n]) by 120 ° in the retarding direction (S18).

  Subsequently, the three-phase primary current values Iu_1 (θ [n]), Iv_1 (θ [n]), and Iw_1 (θ [n]) calculated in S18 are represented by the matrix shown in FIG. The actual currents Id and Iq are converted (S19). This process is complete | finished above. The U-phase primary current value Iu_1 or the W-phase primary current value Iw_1 is calculated from the V-phase primary current value Iv_1, and the calculated 2-phase primary current value and Kirchhoff's law are shown in FIG. The actual currents Id and Iq may be calculated from the matrix shown.

<Rectangular wave control>
Next, the rectangular wave control mode will be described. In the rectangular wave control mode, torque feedback control is performed, so the feedback control amount is torque. FIG. 11 shows the configuration of the control device 40 corresponding to the rectangular wave control mode. The control device 40 includes functions of a deviation calculation unit 61, a PI control unit 62, a rectangular pulse generation unit 63, and a signal generation unit 64. Furthermore, the control device 40 includes functions of a selection unit 53, a synthesis unit 54, a Fourier coefficient calculation unit 55, a primary current calculation unit 56, a dq current calculation unit 57, a torque estimation unit 65, and a calculation angle setting unit 71. Since the functions of the selection unit 53, the synthesis unit 54, the Fourier coefficient calculation unit 55, the primary current calculation unit 56, the dq current calculation unit 57, and the calculation angle setting unit 71 are the same as those in the PWM control mode, description thereof is omitted.

  The deviation calculating unit 61 calculates a torque deviation ΔTrq between the torque command value Trq * and a torque estimated value Trqe estimated by a torque estimating unit 65 described later. The torque estimated value Trqe is an estimated value of actual torque. The PI controller 62 calculates the voltage phase command value φq * by proportional integral control so that the torque deviation ΔTrq converges to 0 so that the estimated torque value Trqe follows the torque command value Trq *. In the rectangular wave control mode, the estimated torque value Trqe corresponds to the control amount, and the torque command value Trq * corresponds to the target value of the control amount.

  The rectangular pulse generator 63 generates a rectangular wave based on the voltage phase command value φq *, which is the command value of the phase φq, and the rotational position θ, and voltage command values Vu *, Vv *, Vw * for each phase. Is generated. The signal generator 64 generates operation signals gua, gub, gva, gvb, gwa, gwb based on the voltage command values Vu *, Vv *, Vw * and the modulation factor m of each phase, and generates the generated operation signals. It transmits to each switching element of the inverter 20. In the rectangular wave control mode, the functions from the deviation calculation unit 61 to the signal generation unit 64 described above correspond to the operation unit.

  The torque estimation unit 65 calculates the estimated torque value Trqe based on the actual currents Id and Iq calculated by the dq current calculation unit 57 and the torque map or mathematical expression. The torque map is a map showing the correspondence between the actual currents Id and Iq and the actual torque, and is stored in advance in the storage device.

  In addition, although the structure corresponding to PWM control mode and the structure corresponding to rectangular wave control mode were each shown as a structure of the control apparatus 40, the control apparatus 40 respond | corresponds to both PWM control mode and rectangular wave control mode. It may be configured to. That is, the control device 40 has a function corresponding to PWM control from the current command generation unit 41 to the PWM unit 52 and a function corresponding to rectangular wave control from the deviation calculation unit 61 to the signal generation unit 64, respectively. A function corresponding to PWM control and a function corresponding to rectangular wave control may be switched according to the modulation factor m.

  According to 1st Embodiment described above, there exist the following effects.

  -By superimposing the phase current values of the three phases, if there is a detected value of the phase current value over one-third cycle of the electrical angle of each phase, the phase current value over one cycle of the electrical angle of the reference phase can be obtained. Therefore, when the torque command value is changed, the period in which the primary current value drags the phase current value before the torque command value change is calculated from the case where the Fourier coefficient is calculated from the detected value of the phase current value of one electrical angle cycle. In comparison, it can be reduced to 1/3. Therefore, the responsiveness to the torque command value can be improved.

  -The primary current of the reference phase can be calculated from the calculated Fourier coefficient, and the primary current of the other two phases can be obtained by shifting the primary current of the reference phase by 120 electrical degrees in the advance direction and the retard direction, respectively. Can be calculated.

(Second Embodiment)
Next, a difference between the control device 40 according to the second embodiment and the first embodiment will be described with reference to FIGS. FIG. 12 shows the configuration of the control device 40 corresponding to the PWM control mode according to the second embodiment, and FIG. 13 shows the configuration of the control device 40 corresponding to the rectangular wave control mode according to the second embodiment. The control device 40 according to the second embodiment is different in the selection range of the phase current value of each phase by the selection unit 53 and further has the function of the translation inversion unit 58.

  When the torque command value Trq * is constant, as shown in FIG. 14A, the magnitude of the voltage command value of the predetermined phase at the electrical angle of 0 ° to 180 ° and the electrical angle of 180 ° to 360 in one electrical angle cycle. The magnitude of the voltage command value of the predetermined phase at ° is equal. In this case, as shown in FIG. 14B, the phase current value at the electrical angle of 0 ° to 180 ° shifts the phase current value at the electrical angle of 180 ° to 360 ° by 180 °, and the amplitude direction of the phase current. The value is inverted with respect to the central axis that passes through the center of. Here, shifting the phase current value by an electrical angle of 180 ° and reversing it with respect to the central axis is referred to as translational reversal. By performing the translational reversal, when the torque command value Trq * is constant, the phase current value at the calculation angle shifted by an electrical angle of 180 ° is estimated with high accuracy. Further, when the torque command value Trq * is changed, the period during which the primary current value drags the phase current value before the change of the torque command value Trq * is shortened.

  The control device 40 according to the present embodiment calculates both the synthesis of the phase current values selected in each phase and the translational inversion of the phase current values to calculate the phase current value of the reference phase over one electrical angle cycle. Thereby, compared with 1st Embodiment, the phase current value selected in each phase can be made into the phase current value in a half electrical angle period. As a result, the response of the actual torque to the torque command value Trq * can be further improved than in the first embodiment. This will be described in detail below.

  The selection unit 53 selects the phase current value at the calculation angle over the electrical angle 1/6 cycle for the phase current value of each phase. FIGS. 15A to 15C show diagrams in which phase current values at calculation angles ranging from 0 ° to 60 ° are selected. A selected phase current value in the V phase is indicated by a black circle, and an unselected phase current value is indicated by a white circle. A selected phase current value in the W phase is indicated by a black square, and an unselected phase current value is indicated by a white square. The selected phase current value in the U phase is indicated by a black triangle, and the unselected phase current value is indicated by a white triangle.

  The combining unit 54 uses the V phase as a reference phase and superimposes the selected phase current value Ius in the U phase on the selected phase current value Ivs in the V phase while shifting the electrical angle by 120 ° in the retard direction. The selected phase current value Iws in the phase is overlapped by shifting the electrical angle by 120 ° in the advance direction. FIG. 15D shows the selected phase current values in the W phase and U phase shown in FIGS. 15B and 15C with respect to the selected phase current values in the V phase shown in FIG. It is the figure which synthesize | combined each shifted | deviated the electrical angle 120 degrees in the advance direction and the retard direction. A combination of the phase current values selected in the three phases is a phase current value for an electrical angle half cycle.

  The translation inversion unit 58 inverts the phase current value obtained by synthesizing the phase current values selected in the three phases. The combining unit 54 adds the combined phase current value and the phase current value obtained by translationally inverting the combined phase current value to obtain the reference phase phase current value over one electrical angle cycle. That is, the combining unit 54 advances the selected phase current value in the V phase, the phase current value obtained by translationally inverting the selected phase current value in the V phase, and the selected phase current value in the W phase and the U phase in the advance direction. And a phase current value shifted by 120 ° in the electrical angle direction and a phase current value obtained by translationally reversing the phase current value shifted by the electrical angle of 120 ° in the advance direction and the retard direction by one period of electrical angle. The phase current value of the V phase is FIG. 15E shows the phase current value of the V phase over one electrical angle cycle calculated in this way. White circles indicate phase current values obtained by translationally reversing selected phase current values in the V phase. The white squares indicate the phase current values obtained by shifting the selected phase current value in the W phase by 120 ° in the advance angle direction and translationally reversing it. The white triangle indicates the phase current value obtained by shifting the selected phase current value in the U phase by 120 ° in the retarding direction and translationally reversing it.

  For each phase, after calculating the phase current value obtained by translationally reversing the phase current value selected in each phase, the phase current value selected in each phase and the phase current value selected in each phase are translated and inverted. The phase current value of the reference phase in one electrical angle cycle may be calculated by combining the obtained phase current values.

  Next, a processing procedure for calculating the dq current from the detected value of the phase current value according to the present embodiment will be described with reference to the flowchart of FIG. This processing procedure is repeatedly executed by the control device 40 for each calculation angle.

  First, in S30 to S34, the same processing as S10 to S14 is performed. Subsequently, a phase current value over a period of 1/6 electrical angle of each phase is selected (S35). Specifically, the phase current value of each phase at the calculation angles θ [n], θ [n−1], θ [n−2]... Over the latest electrical angle 1/6 period before the calculation angle θ [n]. Select.

  Subsequently, with the V phase as the reference phase, the phase current value selected in the W phase and the phase current value selected in the U phase in the advance direction and the retard direction, respectively, with respect to the phase current value selected in the V phase. The electrical angle is synthesized by shifting 120 ° (S36).

  Subsequently, the phase current value of the V phase over one electrical angle cycle is calculated from the synthesized phase current value of the V phase and the phase current value obtained by translationally inverting the synthesized V phase phase current value (S37).

  Subsequently, first-order Fourier coefficients a1 and b1 are calculated for the V-phase phase current value calculated in S37 for one electrical angle cycle, as in the process of S17. The phase current Iv (θ [n]) (n = 1, 2,... N) is a selected phase current value in the V phase, a phase current value obtained by translationally inverting the selected phase current value in the V phase, and W The phase current value obtained by shifting the selected phase current value in the phase and the U phase by 120 ° in the advance angle and the retard direction, respectively, and the phase current value obtained by shifting the electrical angle in the advance direction and the retard direction by 120 °, respectively. The phase current value obtained by translation inversion is combined. The calculation angle θ [n] (n = 1, 2,... N) is the selected calculation angle θ [n] over a period of 1/6 electrical angle, the selected calculation angle θ [n] + 180 °, and the selected calculation. The angle θ [n] ± 120 ° and the selected calculation angle θ [n] ± 120 ° + 180 ° are sequentially arranged.

  Subsequently, in S39 and S40, the same processing as in S18 and S19 is performed. This process is complete | finished above.

  According to 2nd Embodiment described above, there exist the following effects.

  -By superimposing the phase current values of the three phases and reversing the translation, if there is a detected value of the phase current value over 1/6 electrical angle of each phase, the phase current value over 1 electrical angle cycle of the reference phase is obtained. be able to. Therefore, when the torque command value is changed, the period in which the primary current value drags the phase current value before the torque command value change is calculated from the case where the Fourier coefficient is calculated from the detected value of the phase current value of one electrical angle cycle. Compared to 1/6. Therefore, the responsiveness to the torque command value can be improved.

(Third embodiment)
Next, a difference between the control device 40 according to the third embodiment and the first embodiment and the second embodiment will be described with reference to FIGS. As illustrated in FIG. 19, the control device 40 according to the third embodiment includes functions of offset error calculation units 72 and 73 and correction units 74 and 75. FIG. 19 shows only functions of the control device 40 between the MG 30 and the selection unit 53. The configuration of the control device 40 other than the portion shown in FIG. 19 is the same as that of any of FIGS.

  As described above, the phase current value of the MG 30 is a signal in which the primary component is the main component and the offset error a0 / 2, which is the zeroth-order component, and the higher-order component are superimposed. As shown in Expression (3), the phase current value can be obtained by performing Fourier series expansion of the phase current value and extracting the primary component, thereby obtaining a phase current value that does not include an offset error or a higher-order component. However, when calculating the phase current value of one phase by superimposing the three-phase phase current values, for example, if the offset error is included only in the W phase, the offset error included in the combined phase current value is removed. Can not do it. Further, even when a phase current value over a predetermined range is selected and the phase current value over one electrical angle period is estimated by translationally inverting the selected phase current value, an offset error included in the phase current value cannot be removed. .

  FIG. 17A shows the phase current value in the V phase with black circles. In FIG. 17B, the phase current value including the offset error in the W phase is indicated by a black square, and the phase current value not including the offset error is indicated by a solid line. FIG. 17C shows the phase current value in the U phase with a black triangle. FIG. 17D shows the phase current values obtained by synthesizing the phase current values over the electrical angles of 0 ° to 120 ° selected in each phase with reference to the V phase. The synthesized phase current value includes an offset error in the phase current value at an electrical angle of 120 ° to 240 ° which was originally the phase current value of the W phase.

  In FIG. 18A, the phase current value including the offset error is indicated by a black square, and the phase current value not including the offset error is indicated by a solid line. In FIG. 18B, a phase current value including an offset error over an electrical angle half cycle is indicated by a black square, and a phase current value obtained by translationally inverting the phase current value is indicated by a white square. As shown in FIGS. 18A and 18B, when the selected phase current value includes an offset error that is offset in the positive direction, the phase current value obtained by translationally inverting the selected phase current value is Therefore, an offset error offset in the negative direction is included.

  As shown in FIG. 17B and FIG. 18A, if an equal offset error is not included over one electrical angle period, the offset error cannot be calculated correctly even if Fourier series expansion is performed. For this reason, when the three phases do not include an offset error evenly, even if the combined phase current value is Fourier series expanded, the offset error cannot be calculated correctly and the offset error cannot be removed. As a result, the extracted primary component also includes an error.

  In addition, even if the offset error is included evenly in the three phases, if the phase current value including the offset error is translated and inverted, the synthesized phase current value includes the offset error unevenly. The offset error cannot be calculated correctly and the offset error cannot be removed. As a result, the extracted primary component also includes an error.

  Therefore, the offset error calculation units 72 and 73 are not the phase current values over one electrical angle cycle estimated from the V phase and W phase current values over the electrical angle half cycle, but the V phase and W phase over one electrical angle cycle. V-phase and W-phase offset errors are calculated based on the detected phase current value. However, if the offset error is calculated based on the detected value of the phase current value over one electrical angle cycle, as described above, the phase current value before and after the torque command value Tr * is changed when the torque command value Tr * is changed. The offset error is calculated using both of the detected values. For this reason, when the torque command value Tr * is changed, an offset error calculation error increases, which may impair high response. Here, in general, the offset error does not change greatly before and after the change of the torque command value Tr *.

  Therefore, the offset error calculation units 72 and 73 calculate the offset error when the change amount of the torque command value Tr * in the predetermined period is equal to or smaller than the threshold value, that is, when the torque command value Tr * is not changed. Update to error. Further, when the change amount exceeds the threshold value, that is, when the torque command value Tr * is changed, the offset error calculation units 72 and 73 maintain the offset error at the value of the offset error at the previous calculation. Thus, an offset error having a large calculation error calculated based on both the detected value of the phase current value before and after the torque command value Tr * is not used. Note that a change amount of the estimated torque value Tre may be used instead of the torque command value Tr *.

  The correction units 74 and 75 correct the detected values of the V-phase and W-phase phase current values using the V-phase and W-phase offset errors calculated by the offset error calculation units 72 and 73. Specifically, the correction units 74 and 75 correct the detected values of the V-phase and W-phase phase current values by adding an offset error, respectively. The estimating unit 53 translates and reverses the detected value of the phase current value over the electrical angle half cycle after correction to calculate the phase current value over one electrical angle cycle.

  Next, a method for calculating the dq current from the detected phase current value according to the third embodiment will be described with reference to the flowchart of FIG. This processing procedure is repeatedly executed by the control device 40 for each calculation angle. This processing procedure is a processing procedure when the offset error calculation units 72 and 73 and the correction units 74 and 75 are applied to the second embodiment. Similarly, the offset error calculation unit 72, 73 and correction units 74 and 75 may be applied.

  First, in S50 to S53, the same processing as S30 to S33 is performed. Next, using the equation shown in FIG. 20, from the detected value Ivd of the V-phase current and the detected value Iwd of the W-phase current over one electrical angle cycle, the V-phase offset error aV0 / 2 and the W-phase The offset error aW0 / 2 is calculated (S54). Note that Ivd (θ [n]) · (θ [n] −θ [n−1]) and Iwd (θ [n]) · (θ [n] −θ [n−1]) are V phase. And a calculated value based on the detected value of the phase current value of the W phase.

  Subsequently, it is determined whether or not the amount of change in the torque command value Trq * within a predetermined period is within a threshold (S55). If the change amount is within the threshold (S55: YES), the offset error is updated (S56). On the other hand, when the amount of change exceeds the threshold (S55: NO), the offset error is maintained at the value at the previous processing (S57).

  Subsequently, in S59 to S65, the same processing as S34 to S40 is performed. This process is complete | finished above.

  According to 3rd Embodiment described above, while there exists an effect of 1st Embodiment or 2nd Embodiment, there exist the following effects.

  Even when the detected value of the phase current value over the selected predetermined range includes an offset error, the offset error can be calculated with high accuracy and the offset error can be removed. As a result, a highly accurate primary current can be calculated.

  When the amount of change in the torque command value * in the predetermined period exceeds the threshold value, the offset error is not updated, so there is no possibility of impairing the high response of the primary current calculation.

(Other embodiments)
-In 1st Embodiment, when combining the phase current value of several phases, a part of phase current value may overlap. Specifically, when synthesizing three-phase phase current values, it is only necessary to select a phase current value in a calculation angle over a predetermined range of electrical angle 1/3 period or more and electrical angle 1/2 period or less in each phase. . For the portion where the phase current values of a plurality of phases overlap, any one phase current value may be selected as the reference phase phase current value.

  In the second embodiment, when the phase current values of a plurality of phases are combined and translated and inverted, a part of the phase current values may overlap. Specifically, when the phase current values of the three phases are synthesized and translated and inverted, the phase current values at the calculation angles over a predetermined range of the electrical angle 1/6 period or more and the electrical angle 1/4 period or less are obtained in each phase. Just choose. For a portion where a plurality of phase current values or translationally reversed phase current values overlap, any one of the phase current values may be selected as the reference phase phase current value.

  -Each of the three-phase phase current values may be detected by a current sensor. In this case, in the flowcharts of FIG. 9 and FIG. 16, the detected values of the three-phase phase currents are acquired in the processes of S13 and S33, respectively, and the processes of S14 and S34 are omitted.

  The detection timing of the phase current value by the current sensors 31 and 32 and the calculation angle may not be synchronized. In this case, the detected phase current value may be linearly complemented to calculate the phase current value at the timing of the calculation angle and use it for the calculation of the Fourier coefficient.

  DESCRIPTION OF SYMBOLS 20 ... Inverter, 30 ... MG, 31, 32 ... Current sensor, 40 ... Control apparatus.

Claims (6)

A three-phase AC rotating electric machine (30), an inverter (20) for driving the AC rotating electric machine, and current sensors (31, 32) for detecting phase current values of two or more phases of the AC rotating electric machine, respectively A control device (40) for an AC rotating electrical machine applied to a rotating electrical machine system comprising:
A calculation angle setting unit that divides one cycle of three electrical angles into N pieces (N is a natural number) and sets a calculation angle at the same timing;
For the three-phase phase current value, a selection unit that selects each of the phase current values in the calculation angle over a predetermined range of electrical angle 1/3 cycle or more and electrical angle 1/2 cycle,
One phase of the three phases is set as a reference phase, and the selected phase current value in a phase other than the reference phase in the three phases is selected with respect to the selected phase current value in the reference phase. An electrical angle 120 ° shifted and superimposed, and a combined unit that sets the superimposed phase current values as the phase current values of the reference phase,
A coefficient calculation unit that calculates a Fourier coefficient by integrating a calculated value based on each phase current value of the reference phase over one cycle of the electrical angle for each calculation angle;
A primary current calculation unit for calculating a primary current value based on the calculated primary component of the Fourier coefficient;
A dq current calculation unit for calculating a d-axis current value and a q-axis current value by performing dq conversion on the calculated primary current value;
An operation unit that operates a plurality of switching elements constituting the inverter so that a control amount based on the calculated d-axis current value and the q-axis current value follows a target value of the control amount;
A control device for an AC rotating electrical machine. A three-phase AC rotating electric machine (30), an inverter (20) for driving the AC rotating electric machine, and current sensors (31, 32) for detecting phase current values of two or more phases of the AC rotating electric machine, respectively A control device (40) for an AC rotating electrical machine applied to a rotating electrical machine system comprising:
A calculation angle setting unit that divides one cycle of three electrical angles into N pieces (N is a natural number) and sets a calculation angle at the same timing;
For the three-phase phase current values, a selection unit that selects each of the phase current values in the calculation angle over a predetermined range that is equal to or greater than 1/6 electrical angle and less than 1/4 electrical angle;
A translation reversing unit that performs translation reversal that shifts the phase current value with respect to a central axis passing through the center in the amplitude direction of the phase current while shifting the electrical angle by 180 °;
One phase of the three phases is set as a reference phase, the selected phase current value in the reference phase, the phase current value obtained by translating the selected phase current value in the reference phase, and the 3 The phase current value obtained by shifting the selected phase current value in the phase other than the reference phase among the phases by an electrical angle of 120 °, and the phase current value obtained by translating and reversing the phase current value shifted by the electrical angle of 120 ° And a combining unit that sets the phase current value of the reference phase as
A coefficient calculation unit that calculates a Fourier coefficient by integrating a calculated value based on each phase current value of the reference phase over one cycle of the electrical angle for each calculation angle;
A primary current calculation unit for calculating a primary current value based on the calculated primary component of the Fourier coefficient;
A dq current calculation unit for calculating a d-axis current value and a q-axis current value by performing dq conversion on the calculated primary current value;
An operation unit that operates a plurality of switching elements constituting the inverter so that a control amount based on the calculated d-axis current value and the q-axis current value follows a target value of the control amount;
A control device for an AC rotating electrical machine. The current sensor to be detected has two phases, and the remaining one phase is calculated from the two phases,
The selection unit respectively selects the phase current values at the calculation angles over a predetermined range of electrical angle 1/3 period or more and electrical angle 1/2 period for the three-phase phase current values,
The synthesizing unit converts the selected phase current values of two phases other than the reference phase into electrical angles in the advance direction and the retard direction, respectively, with respect to the selected phase current value of the reference phase. The control device for an AC rotating electrical machine according to claim 1, wherein the superposition is performed while shifting by 120 °.   The primary current calculation unit calculates the primary current of the reference phase from the calculated Fourier coefficient, and the calculated primary current of the reference phase has an electrical angle of 120 degrees in the advance direction and the retard direction, respectively. The control device for an AC rotating electrical machine according to any one of claims 1 to 3, wherein the primary current of the other two phases is calculated by shifting the primary current. The calculated value based on the detected value of the phase current value at the calculation angle over one cycle of the electrical angle is integrated to calculate the Fourier coefficient of the zero-order component, and the phase current is calculated based on the calculated Fourier coefficient of the zero-order component An offset error calculation unit for calculating an offset error of the detected value of the value;
The control device for an AC rotating machine according to claim 1, further comprising: a correction unit that corrects the phase current value detected by the current sensor using the calculated offset error.   The offset error calculation unit updates the offset error to the calculated offset error when the amount of change in torque of the AC rotating electric machine during a predetermined period is equal to or less than a threshold, and when the amount of change exceeds the threshold The control device for an AC rotating electrical machine according to claim 5, wherein the offset error is maintained at a value at the time of the previous calculation.

Images