JP2017038443A - Control device for AC rotary electric machine - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a control device for an AC rotary electric machine that improves responsiveness to a torque command value.SOLUTION: A control device for an AC rotary electric machine comprises: an arithmetic angle setting section that sets an arithmetic angle by dividing an electric angle k period (k is a natural number) by N (N is a natural number); an estimating section that estimates a phase current value for a remaining electric angle (k-m/2) period on the basis of a detection value of a phase current value for an electric angle m/2 period (m is a natural number of m≤k) of the electric angle k period; a coefficient calculating section that calculates a Fourier coefficient by integrating calculated values based on each phase current value of a specific phase for the electric angle k period for each arithmetic angle; a primary current calculating section that calculates a primary current value on the basis of a primary component of the calculated Fourier coefficient; a dq current calculating section that calculates a dq current value 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 value 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 the electrical angle k period at the 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 the primary current based on the phase current detection value at the integrated angle over the electrical angle k period. Therefore, when the torque command value is changed, the detected primary current drags the phase current before the torque command value is changed until the potential angle k period elapses after the torque command value is changed. . 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 described in claim 1 is a rotating electrical machine comprising a three-phase AC rotating electrical machine, an inverter that drives the AC rotating electrical machine, and a current sensor that detects a phase current value of at least one phase of the AC rotating electrical machine. A control device for an AC rotating electric machine applied to a system, wherein an electrical angle k period (k is a natural number) is divided into N (N is a natural number) to set an arithmetic angle, and the electrical angle Based on the detected value of the phase current value in the calculation angle over the electrical angle m / 2 period (m is a natural number of m ≦ k) in the k period, the remaining electrical angle (k in the electrical angle k period) -M / 2) An estimation unit that estimates the phase current value at the calculation angle over a period, and combines the detected value of the phase current value with the estimated value to obtain the phase current value over the electrical angle k period; For each calculation angle, each phase current value of a predetermined phase A coefficient calculation unit that calculates a Fourier coefficient by integrating the calculated values over the electrical angle k period, and a primary current calculation unit that calculates a primary current value based on the calculated primary component of the Fourier coefficient; A dq current calculation unit for calculating the d-axis current value and the q-axis current value by dq conversion of the calculated primary current value, and a control based on the calculated d-axis current value and the q-axis current value An operation unit that operates a plurality of switching elements constituting the inverter so that the amount follows a target value of the control amount.

  According to the first aspect of the present invention, the calculation angle is set by dividing the electrical angle k cycle into N pieces. Further, the phase current value at the calculation angle over the remaining (k−m / 2) cycles is estimated based on the detected value of the phase current value at the calculation angle over the electrical angle m / 2 cycle of the electrical angle k cycle. . Then, the detected value of the phase current value and the estimated value are combined to obtain a phase current value over the electrical angle k period. Then, for each calculation angle, the calculated values based on the phase current values are integrated over the electrical angle k period, and the Fourier coefficient is calculated. Thereby, the Fourier coefficient can be calculated from the detected value of the phase current value over the k / 2 period. That is, the Fourier coefficient can be calculated from the detected value of the phase current value over a period of ½ or less, compared to the case where the Fourier coefficient is calculated from the detected value of the phase current value over the electrical angle k period.

  Further, a primary current value is calculated based on the calculated primary component of the Fourier coefficient, 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 controlled variable based on the converted dq current value follows the target value. Therefore, when the torque command value is changed, the primary current value is equal to the phase current value before the torque command value change compared to the case where the Fourier coefficient is calculated from the detected value of the phase current value over the electrical angle k period. The drag period can be reduced to half or less. Therefore, the responsiveness to the torque command value can be improved.

  The invention according to claim 5 is a three-phase AC rotating electrical machine, an inverter that drives the AC rotating electrical machine, a current sensor that detects a phase current value of at least one phase of the AC rotating electrical machine, and the AC rotating electrical machine. And a rotational position sensor for detecting the rotational position of the rotor of the AC rotating electrical machine control device applied to a rotating electrical machine system, wherein N electrical angle k periods (k is a natural number) (N is a natural number) The phase angle at a predetermined calculation angle when it is determined that the phase of the phase current and the rotational position of the rotational position sensor are synchronized with each other. An estimation unit that calculates the maximum amplitude of the phase current based on the detected value of the value and the detected value of the rotational position, and estimates the phase current value at the calculation angle over the electrical angle k cycle from the calculated maximum amplitude And said performance For each corner, a coefficient calculation unit that calculates a Fourier coefficient by accumulating a calculated value based on the estimated value of the phase current value of a predetermined phase over the period of the electrical angle k, and a primary component of the calculated Fourier coefficient A primary current calculation unit that calculates a primary current value; a dq current calculation unit that calculates a d-axis current value and a q-axis current value by dq conversion of 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 d-axis current value and the q-axis current value follows a target value of the control amount.

  According to the fifth aspect of the present invention, the calculation angle is set by dividing the electrical angle k cycle into N pieces. When the phase of the phase current and the rotational position of the rotational position sensor are synchronized, the maximum amplitude of the phase current can be calculated from at least one set of the detected value of the phase current value and the detected value of the rotational position. Therefore, when it is determined that the detected value of the phase current value and the detected value of the rotational position are synchronized, the phase current is detected based on the detected value of the phase current value and the detected value of the rotational position at a predetermined calculation angle. And the phase current value over the electrical angle k period is estimated from the calculated maximum amplitude of the phase current. Then, for each calculation angle, the calculated value based on the estimated value of the phase current value is integrated over the electrical angle k period, and the Fourier coefficient is calculated. Further, a primary current value is calculated based on the calculated primary component of the Fourier coefficient, the calculated primary current value is converted into a dq current value, and a control amount based on the converted dq current value is obtained. The switching element of the inverter is operated so as to follow the target value. Therefore, since the primary current can be calculated based on at least one set of phase current value detection values and rotation position detection values, the responsiveness to the torque command value can be improved.

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 alternating current motor corresponding to PWM control mode. 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 (a) Fourier coefficient which concerns on this embodiment, (b) the estimated phase current value over 1 period, and (c) the calculated primary current value. The figure which shows (a) voltage and (b) phase current. (A) The detected value of the phase current value of a U phase, (b) The figure which shows the detected value and estimated value of the phase current value of a U phase. The figure which shows the detected value of the phase current value of (a) V phase, and the detected value and estimated value of the (b) phase current value of V phase. (A) The detected value of the phase current value of W phase, (b) The figure which shows the detected value and estimated value of the phase current value of W phase. The flowchart which shows the process sequence which calculates dq electric current which concerns on 1st Embodiment. The time chart which shows the characteristic of the rotation position with respect to time (a) which concerns on 1st Embodiment, and the characteristic of the phase current of the V phase with respect to (b) time. The flowchart which shows the process sequence which calculates the dq electric current which concerns on the modification of 1st Embodiment. The block diagram which shows the structure of the control apparatus of the AC motor corresponding to a rectangular wave control mode. (A) 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 2nd Embodiment. The flowchart which shows the process sequence which calculates the dq electric current which concerns on the modification of 2nd Embodiment. The flowchart which shows the process sequence which calculates dq electric current which concerns on 3rd Embodiment. The time chart which shows the characteristic of the rotation position with respect to time which concerns on 3rd Embodiment, and the characteristic of the phase current of the V phase with respect to (b) time.

  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. Furthermore, the control device 40 includes functions of an estimation unit 53, a Fourier coefficient calculation unit 54, a primary current calculation unit 55, a dq current calculation unit 56, 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 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 56 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 an electrical angle k period (k is a natural number) into N (N is a natural number). For example, as shown in FIG. 5, when k = 1 and the calculation angle is set by dividing one period of the electrical angle into 12, the calculation angle and 11 calculation parameters immediately before the calculation angle are calculated. A Fourier coefficient is calculated based on the twelve phase current values at the calculation angle. 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 the conventional Fourier coefficient calculation process when k = 1. 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 is changed between the torque command value Trq * and the period of one electrical angle after the torque command value Trq * is changed. It is affected by the previous 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.

  Therefore, the estimation unit 53 determines the remaining electrical angle (k−m / m) based on the detected value of the phase current value in the calculation angle over the electrical angle m / 2 period of the electrical angle k period out of the electrical angle k period. 2) The phase current value at the calculation angle over the period was estimated. m is a natural number satisfying m ≦ k. Then, the estimation unit 53 combines the detected value of the phase current value over the period of m / 2 and the estimated value of the phase current value over the remaining electrical angle (km−2) period to obtain a phase over the electrical angle k period. The current value. The Fourier coefficient calculation unit 54 (coefficient calculation unit) integrates the calculated values based on the phase current value calculated over the electrical angle k cycle of the predetermined phase calculated by the estimation unit 53 over the electrical angle k cycle, and the Fourier coefficient of the predetermined phase. The primary component of is calculated. In this embodiment, k = 1, and the estimation unit 53 estimates the phase current value over the remaining electrical angle half cycle based on the detected value of the phase current value over the electrical angle half cycle.

  A specific method for estimating the phase current value by the estimation unit 53 will be described below. When the torque command value Trq * is constant, as shown in FIG. 7A, 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. 7B, 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.

  Therefore, as shown in FIGS. 8A and 8B, the estimation unit 53 shifts the detected value of the phase current value over the electrical angle half cycle by 180 ° and reverses it with respect to the central axis, and the rest. Calculate an estimate of the phase current value over the electrical angle half cycle. Here, shifting the detected value of 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. Then, the estimation unit 53 combines the detected value of the phase current value over the electrical angle half cycle and the detected value of the phase current value over the remaining electrical angle half cycle to obtain the phase current value over one electrical angle cycle. As described above, when the value obtained by translationally reversing the phase current value of the electrical angle half cycle is the phase current value of the remaining electrical angle half cycle, an ideal state is obtained in the Fourier series expansion. FIG. 8C shows the primary current value calculated based on the calculated phase current value over one electrical angle cycle. In FIGS. 9A and 9B, the detected value of the U-phase phase current value is indicated by a black circle, and the estimated value obtained by translationally inverting the detected value is indicated by a white circle. 10A and 10B show the detected value of the V-phase phase current value by a black square, and the estimated value obtained by translationally inverting the detected value by a white square. In FIGS. 11A and 11B, the detected value of the W-phase phase current value is indicated by a black triangle, and the estimated value obtained by translationally inverting the detected value is indicated by a white triangle.

  When the torque command value Trq * is constant by calculating the phase current values Iue, Ive, Iwe over one electrical angle cycle of each phase from the detected value of the phase current value for the electrical angle half cycle, The phase current value in the electrical angle half cycle is estimated with high accuracy. Further, when the torque command value Trq * is changed, the period in which the primary current value drags the phase current value before the change of the torque command value Trq * is shortened to ½ of the conventional value. For example, as shown in FIG. 8A, when the torque command value Trq * is changed at time t1, and the dq current at time t2 is calculated, the detected value of the phase current value from time t1 to time t2, and time t1 From this, the primary current value is calculated based on the estimated value estimated from the detected value of the phase current value at time t2. 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. Conventionally, the period in which the primary current value drags the phase current value before the command value change is from the time point t1 to the time point 3, whereas in this embodiment, the dragging period is shortened from the time point t1 to the time point t2. Is done. Therefore, compared with the prior art, a decrease in response of the actual torque to the torque command value Trq * when the torque command value Trq * is changed is suppressed.

  The primary current calculation unit 55 calculates the primary current values Iu_1, Iv_1, and Iw_1 of each phase based on the Fourier coefficients au1, bu1, av1, bv1, aw1, bw1 of the phases calculated by the Fourier coefficient calculation unit 54. Calculate each. 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 56 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.

  The calculation angle setting unit 71 sets the calculation angle by dividing one cycle of the electrical angle into N pieces.

  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. 12 and the time chart of FIG.

  FIG. 13A shows the characteristic of the rotational position θ with respect to time, and FIG. 13B shows the characteristic of the V-phase phase current Iv (θ) with respect to time. In FIG. 13 (a), 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, a processing procedure for calculating the dq current from the detected value of the phase current 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). The calculated U-phase phase current value Iu (θ [n]) is set as a detected value of the U-phase phase current value at the calculation angle θ [n].

  Subsequently, phase current values Iue, Ive, Iwe over one electrical angle cycle of each phase are estimated from the phase current values over the electrical angle half cycle of each phase (S15). Specifically, for each phase, the phase current detected at calculation angles θ [n], θ [n−1], θ [n−2]... Over the most recent electrical angle half cycle before the calculation angle θ [n]. A phase current value Iue, Ive, Iwe over one electrical angle cycle is estimated by combining the detected value of the value and an estimated value obtained by translationally inverting the detected value of the phase current value. When N is an even number, N / 2 estimated values are calculated from the latest N / 2 detected values. When N is an odd number, (N + 1) / 2 estimated values are calculated from the latest (N + 1) / 2 detected values. When N is an odd number, a portion where the detected value and the estimated value overlap is generated, and either one is selected.

  Subsequently, the first-order Fourier coefficients av1 and bv1 are respectively calculated using the formula shown in FIG. 12 for the phase current value Ive for one cycle of the V-phase electrical angle calculated in S15. The expression shown in FIG. 12 is an expression obtained by discretizing Expressions (2) and (3). Ive (θ [n]) · cos θ [n] · (θ [n] −θ [n−1]) is integrated from n = 1 to N to calculate the Fourier coefficient av1. Further, Ive (θ [n]) · sin θ [n] · (θ [n] −θ [n−1]) is integrated from n = 1 to N to calculate the Fourier coefficient bv1.

  Then, a V-phase primary current value Iv_1 (θ [n]) is calculated from the calculated Fourier coefficients av1 and bv1 (S16). Similarly, first-order Fourier coefficients au1, bu1, aw1, bw1 are calculated for the phase current values Iue and Iwe of the U-phase and W-phase electrical angles calculated in S15, and the U-phase and W-phase primarys are calculated. Current values Iu_1 (θ [n]) and Iw_1 (θ [n]) are calculated. FIG. 12 representatively shows equations for calculating the V-phase Fourier coefficients av1 and bv1, but calculates the U-phase Fourier coefficients au1 and bu1 and the W-phase Fourier coefficients aw1 and bw1. The formula is similar.

  In this embodiment, Ive (θ [n]) (n = 1, 2,... N) corresponds to each phase current value of the V phase. Further, Ive (θ [n]) · cos θ [n] · (θ [n] −θ [n−1]) and Ive (θ [n]) · sin θ [n] · (θ [n] −θ [n−1]) corresponds to a calculated value based on each phase current value of the V phase. The same applies to the U phase and the W phase.

  Subsequently, the three-phase primary current values Iu1 (θ [n]), Iv_1 (θ [n]), and Iw_1 (θ [n]) calculated in S16 are represented by the matrix shown in FIG. The actual currents Id and Iq are converted (S17). This process is complete | finished above.

  Next, a modified example of the processing procedure for calculating the dq current from the detected value of the phase current 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 S20 to S22, processing similar to S10 to S12 is performed. Subsequently, only the phase current value Iv (θ [n]) of the V phase detected by the current sensor 31 is acquired (S23). Subsequently, similarly to S15, the phase current value Ive over one electrical angle period is calculated by translationally inverting the detected value of the phase current value Iv (θ [n]) over the most recent electrical angle half cycle (S24).

  Subsequently, since the phase current values of the respective phases are shifted from each other by an electrical angle of 120 °, the phase current value Ive calculated in S24 over one cycle of the electrical angle is shifted by an electrical angle of 120 ° in the advance direction to obtain an electrical angle. The U-phase phase current value Iue over one cycle is calculated. Further, the phase current value Ive is shifted by an electrical angle of 120 ° in the retard direction, and the W-phase phase current value Iwe over one electrical angle cycle is calculated (S25). Subsequently, in S26 and S27, processing similar to S16 and S17 is performed. Above, this processing procedure is complete | finished.

  Furthermore, the processing procedure for calculating the dq current from the detected value of the phase current value may be modified as follows. In S13, only the phase current value of any one phase, for example, the phase current value Iv (θ [n]) of the V phase is acquired, and the process of S14 is omitted, and one cycle of the electrical angle of the V phase is obtained in S15. Only the phase current value Ive over the range may be estimated. Further, the process of S25 may be omitted. Then, only the V-phase primary current value Iv_1 is calculated in S16 and S26, and the V-phase primary current value Iv_1 is shifted by an electrical angle of 120 ° in the advance direction and the retard direction, respectively. The value Iu_1 and the W-phase primary current value Iw_1 may be calculated. Alternatively, 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 two-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.

  Moreover, you may perform the process of S24 and the process of S25 in reverse order. That is, the phase current value Iv (θ [n]) of the V phase is shifted by an electrical angle of 120 °, and the phase current values Iu (θ [n] + 120 °), Iw (θ [n] −120) of the U phase and the W phase. And then the phase current value over one electrical angle cycle may be calculated from the phase current value over the electrical angle half cycle of each phase.

<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. 15 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 an estimation unit 53, a Fourier coefficient calculation unit 54, a primary current calculation unit 55, a dq current calculation unit 56, a torque estimation unit 65, and a calculation angle setting unit 71. The functions of the estimation unit 53, the Fourier coefficient calculation unit 54, the primary current calculation unit 55, the dq current calculation unit 56, and the calculation angle setting unit 71 are the same as in the PWM control mode, and thus the 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 56 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.

  A Fourier coefficient is calculated based on the detected value of the phase current value over the electrical angle half cycle. Therefore, when the torque command value Trq * is changed, the primary current value is the phase current before the torque command value change compared to the case where the Fourier coefficient is calculated from the detected value of the phase current value over one electrical angle cycle. The period of dragging the value can be halved. Therefore, the responsiveness to the torque command value Trq * can be improved.

  -The phase current value over the remaining half of the electrical angle is estimated by shifting the detected value of the phase current value over the half cycle of the electrical angle within 180 degrees of the electrical angle and reversing it relative to the central axis. be able to. Further, the response to the torque command value Trq * can be improved by setting the detected value of the phase current used for the calculation of the Fourier coefficient to the minimum necessary electrical angle half cycle.

  -From the detection value of the phase current value of one phase over a half cycle of electrical angle, the phase current value over one cycle of electrical angle of that phase can be estimated, and the phase current value over one cycle of electrical angle of the other two phases can be estimated Can do.

  -The phase current value of each phase has a relationship that the electrical angle is shifted by 120 °. Therefore, the phase current value over one period of electrical angle of a predetermined phase consisting of the detected value of the phase current value over a half electrical angle and the estimated value of the phase current value over a half electrical angle period, respectively, in the advance direction and the retard direction. By shifting the electrical angle by 120 °, the phase current value over one cycle of the electrical angle of the other two phases can be calculated.

(Second Embodiment)
Next, a difference of the control device 40 according to the second embodiment from the first embodiment will be described with reference to FIGS. 16 to 19. As shown in FIG. 17, the control device 40 according to the second embodiment includes functions of offset error calculation units 72 and 73 and correction units 74 and 75. FIG. 17 shows only the function of the control device 40 between the MG 30 and the estimation unit 53. The configuration of the control device 40 other than the portion shown in FIG. 17 is the same as that of FIG. 4 or FIG.

  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 a phase current value over an electrical angle half cycle is selected and the phase current value over one electrical angle cycle is estimated by translationally inverting the selected phase current value, an offset error included in the phase current value can be removed. Can not.

  In FIG. 16A, 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. FIG. 16B shows a phase current value including an offset error over an electrical angle half cycle by a black square, and a phase current value obtained by translationally inverting the phase current value by a white square. As shown in FIGS. 16A and 16B, 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. 16A, 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. Therefore, when the selected phase current value includes an offset error, even if the estimated phase current value over one period of the electrical angle is expanded in Fourier series, the offset error is correctly calculated and the offset error is removed. I can't. 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 value of the phase current value according to the second 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 S33, processing similar to S10 to S13 is performed. Then, using the equation shown in FIG. 18, 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 (S34). 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 value (S35). If the amount of change is within the threshold (S35: YES), the offset error is updated (S36). On the other hand, when the amount of change exceeds the threshold (S35: NO), the offset error is maintained at the value at the previous processing (S37).

  Subsequently, offset errors Ivof and Iwof are respectively added to the detected values Ivd and Iwd of the phase current values of the V phase and the W phase for correction to calculate the phase current values Iv and Iw of the V phase and the W phase (S38). ).

  Subsequently, in S39 to S42, the same processing as S14 to S17 is performed. This process is complete | finished above.

  Next, a method for calculating the dq current from the detected value of the phase current value according to the modification of the second 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 S50 to S53, the same processing as S20 to S23 is performed. Next, using the equation shown in FIG. 19, the V-phase offset error aV0 / 2 is calculated from the detected value Ivd of the V-phase current over one electrical angle cycle (S54).

  Subsequently, in S55 to S57, processing similar to S35 to S37 is performed.

  Subsequently, an offset error Ivof is added to the detected value Ivd of the V-phase current to correct it, and a V-phase current value Iv is calculated (S58).

  Subsequently, in S59 to S62, the same processing as S24 to S27 is performed. This process is complete | finished above.

  According to 2nd Embodiment described above, while there exist the effect of 1st Embodiment, there exist the following effects.

  Even when an offset error is included in the detected value of the phase current value over the selected electrical angle half cycle, the offset error can be accurately calculated and 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.

(Third embodiment)
Next, a difference between the control device 40 according to the third embodiment and the first embodiment will be described with reference to FIGS. 20 and 21. The control device 40 according to the second embodiment is different in the estimation method of the phase current value by the estimation unit 53.

  As shown in FIG. 21, when the rotational position θ detected by the rotational position sensor 33 is 0 to 360 ° and the phase current phase of any phase is 0 to 360 °, the rotational position θ From the detected value and the detected value of the phase current value, the maximum amplitude A of the phase current can be estimated. If the maximum amplitude A is estimated, the phase current value at each calculation angle can be estimated. Therefore, the control device 40 determines whether or not the phase of any phase current and the rotational position θ detected by the rotational position sensor 33 are synchronized.

  The estimation unit 53 detects the phase current value at a predetermined calculation angle when it is determined that the phase current phase of any phase and the rotational position θ detected by the rotational position sensor 33 are synchronized. Based on the value and the rotational position θ detected by the rotational position sensor 33, the maximum amplitude A of the phase current is calculated. In the present embodiment, it is assumed that the phase of the V-phase current and the rotational position θ are synchronized.

  Hereinafter, a processing procedure for calculating the dq current from the detected value of the phase current value 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 S32, the same processing as S10 to S12 is performed. Subsequently, the V-phase phase current value Iv (θ [n]) detected by the current sensor 31 and the rotational position θ detected by the rotational position sensor 33 are acquired (S33).

  Subsequently, the maximum amplitude A [n] is calculated from the phase current value Iv (θ [n]) and the rotational position θ acquired in S33 and the equation of Iv (θ [n]) = A [n] · sin θ [n]. Is calculated (S34). Subsequently, the amplitude A_av is calculated by averaging the m amplitudes A [n], A [n−1],... (M is a natural number) calculated at the latest calculation angle (S35). In addition, m = 1 may be sufficient. That is, the maximum amplitude A [n] calculated in S34 may be used as the amplitude A_av as it is.

  Subsequently, using the amplitude A_av calculated in S35, phase current values Iue, Ive, and Iwe of each phase at the calculation angle over one electrical angle cycle are calculated (S36).

  Subsequently, in S37 and S38, the same processing as in S16 and S17 is performed. This process is complete | finished above. Note that a modified example of the processing procedure of the first embodiment may be appropriately applied to this processing procedure.

  According to the second embodiment to be described above, since the primary current value can be calculated based on at least one set of detected phase current values and the detected value of the rotational position θ, the responsiveness to the torque command value Trq * is improved. Can be improved.

(Other embodiments)
The Fourier coefficient may be calculated by integrating calculated values based on the phase current value over the electrical angle k cycle over the electrical angle k cycle. In this case, the detected value of the phase current value over one of the electrical angles 1/2, 1, 3/2,..., M / 2 period (m ≦ k) is repeatedly translated and inverted, and the remaining (k−m / 2) What is necessary is just to estimate the phase current value over a period. For example, in the case of k = 3, from the detected value of the phase current over one of the electrical angle half cycle, electrical angle 1 cycle, electrical angle 3/2 cycle, the remaining electrical angle 5/2 cycle, electrical angle 2 cycle, What is necessary is just to estimate the phase current value over the electrical angle 3/2 period.

  -Each of the three-phase phase currents may be detected by a current sensor. In this case, in the flowchart of FIG. 12, the detected values of the three-phase phase current values are respectively acquired in the process of S13, and the process of S14 is 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 value of the phase current value is linearly complemented, the detected value of the phase current value at the calculation angle timing is calculated, and used for the calculation of the Fourier coefficient.

  DESCRIPTION OF SYMBOLS 20 ... Inverter, 30 ... MG, 31, 32 ... Current sensor, 33 ... Rotation position sensor, 40 ... Control apparatus.

Claims (7)

A three-phase AC rotating electric machine (30), an inverter (20) that drives the AC rotating electric machine, and a current sensor (31, 32) that detects a phase current value of at least one phase of the AC rotating electric machine. An AC rotating electrical machine control device (40) applied to a rotating electrical machine system,
A calculation angle setting unit that sets a calculation angle by dividing an electrical angle k cycle (k is a natural number) into N (N is a natural number);
Based on the detected value of the phase current value at the calculation angle over the electrical angle m / 2 period (m is a natural number of m ≦ k) of the electrical angle k period, the remaining electricity of the electrical angle k period Estimating the phase current value at the calculation angle over an angular (k−m / 2) period, and combining the detected value and the estimated value of the phase current value to make the phase current value over the electrical angle k period And
A coefficient calculation unit that calculates a Fourier coefficient by integrating calculated values based on the phase current values of a predetermined phase over the electrical angle k period 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 electrical angle k period is one electrical angle period;
The estimation unit shifts the detected value of the phase current value at the calculation angle over the electrical angle half cycle with respect to the central axis passing through the center in the amplitude direction of the phase current while shifting the electrical angle by 180 °, so that the electrical angle 1 The control apparatus for an AC rotating electrical machine according to claim 1, wherein the phase current value at the calculation angle over the remaining electrical angle half cycle of the cycle is estimated. The current sensor detects a phase current value flowing in only one phase,
The estimation unit is configured to determine the 1 in the calculation angle over the electrical angle (km−2) period based on the detected value of the phase current value of the one phase in the calculation angle over the electrical angle m / 2 period. The control apparatus for an AC rotating electrical machine according to claim 1 or 2, wherein the phase current value of a phase and the phase current value of a phase different from the one phase at the calculation angle over the electrical angle k period are estimated.   The control of the AC rotating electric machine according to claim 3, wherein the estimation unit estimates the phase current value of the different phase by shifting the detected value and the estimated value of the phase current value of the one phase by an electrical angle of 120 °. apparatus. The calculated value based on the detected value of the phase current value in the calculation angle over the electrical angle k period is integrated to calculate a zero-order component Fourier coefficient, and the phase current is calculated based on the calculated zero-order component Fourier coefficient. 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 detected value of the phase current value 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. A three-phase AC rotating electric machine (30), an inverter (20) for driving the AC rotating electric machine, a current sensor (31, 32) for detecting a phase current value of at least one phase of the AC rotating electric machine, and the AC An AC rotating electrical machine control device (40) applied to a rotating electrical machine system comprising a rotational position sensor (33) for detecting a rotational position of a rotor of the rotating electrical machine,
A calculation angle setting unit that sets a calculation angle by dividing an electrical angle k cycle (k is a natural number) into N (N is a natural number);
When it is determined that the phase of the phase current and the rotational position of the rotational position sensor are synchronized, based on the detected value of the phase current value and the detected value of the rotational position at a predetermined calculation angle, An estimator that calculates a maximum amplitude of the phase current and estimates the phase current value at the calculation angle over the electrical angle k period from the calculated maximum amplitude;
A coefficient calculation unit that calculates a Fourier coefficient by integrating a calculated value based on an estimated value of the phase current value of a predetermined phase over the electrical angle k period 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.

Images