JP5737093B2 - Rotating machine control device - Google Patents

Description

  The present invention relates to a power conversion circuit comprising a switching element that opens and closes between a separate voltage application means for applying each of voltages having different voltage values and a terminal of a rotating machine. It is related with the control apparatus of the rotary machine which controls the controlled variable which has at least one of the electric current which flows through the said rotary machine, the torque of the said rotary machine, and the magnetic flux of the said rotary machine by the on / off operation of the switching element which comprises.

  As this type of control device, for example, as can be seen in Patent Document 1 below, the currents of the three-phase motors when the switching modes defined by the ON / OFF states of the switching elements of the inverter are variously set are predicted. There has been proposed what performs so-called model predictive control in which an inverter is operated in a switching mode capable of minimizing a deviation between a predicted current and a command current. According to this, since the inverter is operated so as to optimize the behavior of the current predicted based on the output voltage of the inverter, the followability to the command current at the time of transition is compared with that by the conventional triangular wave comparison PWM control. And improve. For this reason, model predictive control is considered to be highly useful for applications that require particularly high performance as transient tracking characteristics, such as a motor generator control device as an in-vehicle main unit.

JP 2008-228419 A

  By the way, in the prediction of the current, a current detection value by a current sensor is required as an initial value. However, the output value of the current sensor may include a so-called gain error, which is an error that differs from the actual current amplitude by a predetermined ratio (≠ 1). In this case, the controllability of the model predictive control may be reduced due to the gain error.

  The present invention has been made in the course of solving the above-mentioned problems, and its purpose is to open and close between each of different voltage application means for applying voltages having different voltage values and the terminals of the rotating machine. Regarding a power conversion circuit configured to include a switching element, at least an electric current flowing through the rotating machine, a torque of the rotating machine, and a magnetic flux of the rotating machine by an on / off operation of the switching element configuring the power conversion circuit An object of the present invention is to provide a new control device for a rotating machine that controls a control amount having one.

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

  The invention according to claim 1 is a power conversion circuit configured to include a switching element that opens and closes between a separate voltage application unit that applies each of voltages having different voltage values and a terminal of the rotating machine. A rotating machine that controls a controlled variable having at least one of a current flowing through the rotating machine, a torque of the rotating machine, and a magnetic flux of the rotating machine by turning on / off a switching element constituting the power conversion circuit. Predicting means for temporarily setting a switching mode indicating whether each of the switching elements is in an on state or in an off state in the control device, and performing a prediction regarding the control amount according to each of the temporarily set switching modes And a determinator for determining a switching mode used for actual operation of the power conversion circuit based on a prediction result by the prediction means. And an operation means for operating the power conversion circuit so as to be in the determined switching mode, and an acquisition means for acquiring a detected value of a current flowing through the rotating machine, wherein the prediction process by the prediction means includes: Current prediction processing for performing prediction on the control amount or the current of the rotating machine as a parameter for calculating the control amount, and the norm of the current vector of the rotating machine according to the detected value of the current It further comprises norm feedback means for performing feedback control on the current vector norm predicted by the current prediction processing, corresponding to the switching mode adopted by the operation means.

  When a gain error is superimposed on the detection value, the prediction means has no prediction error between the detection value and the current predicted by the prediction means corresponding to the operation state determined by the determination means. However, divergence can occur. In the above invention, in view of this point, the gain error included in the detected value is grasped using the predicted current as a standard, and the control processing of the control amount is corrected based on the gain error.

  According to a second aspect of the present invention, in the first aspect of the present invention, the feedback control by the norm feedback unit may be configured such that the q-axis component of the current flowing through the rotating machine according to the detected value of the current is the current prediction process. The feedback control is performed on the q-axis component of the current predicted by the method corresponding to the switching mode employed by the operating means.

  There is a positive correlation between the current vector norm and the absolute value of the current vector component in the two-dimensional coordinate system. For this reason, since the q-axis component is also a parameter having a correlation with the norm, it can be used as a feedback control amount.

  The invention described in claim 2 may be as follows.

  In the feedback control by the norm feedback means, the value of the q-axis component for the current flowing through the rotating machine according to the detected value of the current is the value of the q-axis component of the current predicted by the current prediction process. This is a process of performing feedback control to the one corresponding to the switching mode adopted by the operation means.

  The prediction means includes an average voltage calculation means for calculating an average value of the output voltage of the power conversion circuit, and the power conversion according to the temporarily set switching mode for the average value calculated by the average voltage calculation means An instantaneous voltage calculating means for calculating an instantaneous output voltage as a difference between output voltages of the circuit; and a change amount predicting means for predicting a change amount of the control amount based on the instantaneous output voltage, and the norm feedback means provides the The feedback control corresponds to the switching mode adopted by the operation means among the change amounts predicted by the prediction means, with respect to the change amount of the q-axis component for the current flowing through the rotating machine according to the detected value of the current. It is the process which performs feedback control to what to do.

According to a third aspect of the present invention, in the first aspect of the invention, the feedback control by the norm feedback means is predicted by the current of each terminal of the rotating machine according to the detected value of the current and the current prediction process. It is a process of correcting the detection value of the current used in the process of performing the prediction regarding the control amount so that the ratio of the current of each terminal to the target value is feedback-controlled.

  The ratio between the currents is the ratio between the amplitudes if there is no phase shift, and thus the ratio between the norms. For this reason, the feedback control of the ratio to the target value enables the norm feedback control.

  The invention according to claim 4 is the invention according to any one of claims 1 to 3, wherein the norm feedback means includes a current vector flowing through the rotating machine according to the detected value of the current and the current prediction process. Based on the output of the integration element that receives the difference between the norms of each of the current vectors predicted by or the difference between the parameters correlated with the norm, the parameter used in the process of performing the prediction on the control amount is corrected. It is characterized by doing.

  In the above invention, by using the integral element, it is easy to compensate for a steady divergence between the norm of the predicted current vector and the norm corresponding to the detected value.

  According to a fifth aspect of the present invention, in the invention of the fourth aspect, the norm feedback means is configured to correct the parameter used in the process of performing the prediction relating to the control amount, when the parameter of the rotating machine according to the detected value of the current is corrected. It is further characterized by further using an output of a differential element that receives a difference between norms of a current vector and a current vector predicted by the current prediction process or a difference between parameters having a correlation with the norm.

  In the said invention, even if it is a case where a harmonic noise is superimposed on the difference between norms, or the difference between the parameters which have a correlation with this norm, the influence can be suppressed suitably by using a differential element.

  According to a sixth aspect of the present invention, in the invention according to any one of the first to fifth aspects, the prediction means includes a detected value of the current as a current flowing through a part of the terminals of the rotating machine. Current reproduction means for converting the predicted current as the current flowing through the remaining terminals necessary for calculating the current in the dq coordinate system into a current in the dq coordinate system, and the current reproduced by the current reproduction means Is used in the process of performing prediction on the control amount, and the norm feedback means uses the reproduced current as a parameter relating to the norm of the current vector flowing through the rotating machine according to the detected value of the current. It is characterized by.

  In a rotating machine having a plurality of terminals connected to stator windings connected to each other, current information flowing through each of the terminals less than the total number of terminals can be obtained with sufficient current information flowing through the rotating machine. Can be acquired. Here, when only current detection value is insufficient as current information, the shortage can be compensated by using the predicted current. If these are converted into currents in the dq coordinate system, the converted currents include both the detected current value and the predicted current. For this reason, the converted current is a current reflecting the detected value of the current. In view of this point, the above invention includes a current reproducing unit.

  According to a seventh aspect of the invention, in the sixth aspect of the invention, the power conversion circuit includes a switching element that selectively connects a terminal of the rotating machine to each of a positive electrode and a negative electrode of a DC voltage source. The number of terminals to which the output voltage of the DC / AC converter circuit is applied, and the acquisition means is a current flowing through the input terminal of the DC / AC converter circuit. When the voltage vector corresponding to the switching mode adopted by the operation means is an effective voltage vector, the current detection value is output to the three terminals of the rotating machine. The norm feedback means is used as the initial value of the current flowing through any one of the terminals. When the voltage vector corresponding to the winding mode is an effective voltage vector, based on the detected value of the current as a current flowing through one terminal determined according to the adopted switching mode among the three terminals of the rotating machine A parameter relating to a norm of a current vector flowing through the rotating machine according to a detected value of the current is defined.

  In the case of a rotating machine where the number of terminals to which the output voltage of the DC / AC converter circuit is applied is 3, one terminal or two terminals are connected to the positive electrode of the DC voltage source in the operation state represented by the effective voltage vector. And the remaining terminals are connected to the negative electrode of the DC voltage source. For this reason, for the positive electrode or the negative electrode having one connected terminal, the current flowing through the terminal and the detected value of the current match at least in absolute value. In view of this point, in the above invention, the norm-related parameter is determined using the detected current value when the effective voltage vector is adopted.

  The invention according to claim 8 is the invention according to any one of claims 1 to 7, in accordance with a current evaluation result predicted by the prediction means based on a detected value of current acquired by the acquisition means. And a prediction processing correction unit that corrects the control amount prediction processing, wherein the feedback gain of the norm feedback unit is smaller than the feedback gain of the prediction processing correction unit.

  The accuracy of the predicted current can be evaluated by the detected current value at the timing to be predicted. In the said invention, when it estimates that the precision is falling by evaluating the electric current estimated in view of this point, a prediction process can be correct | amended. In addition, at this time, it is possible to achieve compatibility with norm feedback control by making the gain of the norm feedback means smaller than the gain of the prediction processing correction means.

1 is a system configuration diagram according to a first embodiment. FIG. The figure which shows the switching mode concerning the embodiment. The flowchart which shows the process sequence of the model prediction control concerning the embodiment. The flowchart which shows the procedure of the process of the electric current reproduction part concerning the embodiment. The time chart which shows the effect of the embodiment. The time chart which shows the effect of the embodiment. The system block diagram concerning 2nd Embodiment. The system block diagram concerning 3rd Embodiment. The system block diagram concerning 4th Embodiment.

<First Embodiment>
Hereinafter, a first embodiment in which a control device for a rotating machine according to the present invention is applied to a control device for an electric motor will be described with reference to the drawings.

  FIG. 1 shows an overall configuration of an electric motor control system according to the present embodiment. The electric motor 10 is a three-phase synchronous machine.

  The electric motor 10 is connected to the battery 12 via the inverter INV. The inverter INV includes three sets of series connection bodies of switching elements S ¥ p, S ¥ n (¥ = u, v, w), and the connection points of these series connection bodies are U, V, W of the electric motor 10. Connected to each phase. As these switching elements S ¥ # (¥ = u, v, w; # = p, n), an insulated gate bipolar transistor (IGBT) is used in the present embodiment. In addition, a diode D ¥ # is connected in antiparallel to each of these.

  In the present embodiment, the following is provided as detection means for detecting the state of the electric motor 10 and the inverter INV. First, a rotation angle sensor 14 for detecting the rotation angle (electrical angle θ) of the electric motor 10 is provided. Further, a current sensor 16 that detects a current (bus current IDC) flowing through the input terminal (here, the negative side input terminal) of the inverter INV is provided. Furthermore, a voltage sensor 18 for detecting the input voltage (power supply voltage VDC) of the inverter INV is provided.

  The detection values of the various sensors are taken into the control device 20. The control device 20 generates and outputs an operation signal for operating the inverter INV based on the detection values of these various sensors. Here, the signal for operating the switching element S ¥ # of the inverter INV is the operation signal g ¥ #.

  The control device 20 operates the inverter INV so as to control the torque of the electric motor 10 to the required torque T *. Specifically, the inverter INV is operated so that the command current for realizing the required torque T * matches the current flowing through the electric motor 10. That is, in the present embodiment, the torque of the electric motor 10 becomes the final control amount, but in order to control the torque, the current flowing through the electric motor 10 is used as a direct control amount, and this is controlled to the command current. In particular, in the present embodiment, in order to control the current flowing through the motor 10 to the command current, the current of the motor 10 when the switching mode is temporarily set to each of a plurality of modes is predicted, and the actual switching mode of the inverter INV is determined. Perform model predictive control to determine.

  The switching mode indicates whether each of the switching elements S ¥ # constituting the inverter INV is on or off, and includes the eight switching modes 0 to 7 shown in FIG. . For example, the switching mode in which all of the low potential side switching elements Sun, Svn, Swn are in the on state is the switching mode 0, and the switching mode in which all of the high potential side switching elements Sup, Svp, Swp are in the on state. This is switching mode 7. In these switching modes 0 and 7, all phases of the electric motor 10 are short-circuited, and the voltage applied to the electric motor 10 from the inverter INV becomes zero. Therefore, the output voltage vector of the inverter INV is defined as a zero voltage vector. To do. On the other hand, the remaining six switching modes 1 to 6 are defined by an operation pattern in which switching elements that are turned on exist in both the upper arm and the lower arm, and the output voltage vector of the inverter INV is expressed as follows. This is an effective voltage vector.

  FIG. 2B shows voltage vectors V0 to V7 corresponding to the switching modes 0 to 7, respectively. Each of voltage vectors V0 to V7 indicates an output voltage vector of inverter INV in each of switching modes 0 to 7. As shown in the figure, voltage vectors V1, V3, and V5 corresponding to the switching modes 1, 3, and 5 correspond to the positive sides of the U phase, the V phase, and the W phase, respectively.

  Here, the model predictive control will be described in detail.

  Based on the bus current IDC detected by the current sensor 16 shown in FIG. 1, the current reproduction unit 22 calculates the actual currents id and iq of the rotating coordinate system. Further, the rotation angle (electrical angle θ) detected by the rotation angle sensor 14 is input to the speed calculation unit 24, and thereby the rotation speed (electrical angular speed ω) is calculated. On the other hand, the command current setting unit 26 receives the requested torque T * and outputs the command currents id * and iq * in the dq coordinate system. These command currents id *, iq *, actual currents id, iq, electrical angular velocity ω, and electrical angle θ are input to the model prediction control unit 30. The model prediction control unit 30 determines the switching mode of the inverter INV based on these input parameters and outputs it to the operation unit 28. The operation unit 28 generates the operation signal g ¥ # based on the input switching mode and outputs it to the inverter INV.

  Next, details of the processing of the model prediction control unit 30 will be described. The mode setting unit 31 temporarily sets the switching mode of the inverter INV shown in FIG. This process is actually a process of temporarily setting a voltage vector corresponding to the switching mode. The dq conversion unit 32 calculates the voltage vector Vdq = (vd, vq) of the dq coordinate system by performing dq conversion on the voltage vector temporarily set by the mode setting unit 31. In order to perform such conversion, the voltage vector V0 to V7 temporarily set in the mode setting unit 31 is set to, for example, “VDC / 2” when the upper arm is on and “−VDC” when the lower arm is on. / 2 ”. In this case, for example, the voltage vector V0 is (−VDC / 2, −VDC / 2, −VDC / 2), and the voltage vector V1 is (VDC / 2, −VDC / 2, −VDC / 2). .

In the prediction unit 33, the current id when the switching mode of the inverter INV is temporarily set by the mode setting unit 31 based on the voltage vector (vd, vq), the actual current id, iq, and the electrical angular velocity ω. , Iq is predicted. The prediction of the current is performed for each of a plurality of switching modes temporarily set by the mode setting unit 31 based on model expressions expressed by the following expressions (c1) and (c2).
vd = R · id + Ld · (did / dt) −ω · Lq · iq (c1)
vq = R · iq + Lq · (diq / dt) + ω · Ld · id + ω · φ (c2)
Here, the resistance R, the armature flux linkage constant φ, the d-axis inductance Ld, and the q-axis inductance Lq were used.

  On the other hand, the mode determination unit 34 receives the predicted currents ide and iq predicted by the prediction unit 33 and the command currents id * and iq *, and determines the switching mode of the inverter INV. Based on the switching mode thus determined, the operation unit 28 generates and outputs an operation signal g ¥ #.

  FIG. 3 shows a model prediction control processing procedure according to the present embodiment. This process is repeatedly executed in a cycle (control cycle Tc) having a predetermined length.

  In this series of processing, first, in step S10, the electrical angle θ (n) is detected, and the actual currents id (n) and iq (n) are calculated. Also, the voltage vector V (n) determined in the previous control cycle is output. That is, the switching mode of the inverter INV is updated to the switching mode determined in the previous control cycle (the switching mode corresponding to the voltage vector V (n)).

In subsequent step S12, an average voltage vector (vda (n), vqa (n)), which is an average output voltage vector of the inverter INV, is calculated. This is because the following formulas (c3) and (c4) are obtained by substituting the actual currents id (n) and iq (n) into the formulas (c1) and (c2) excluding the term of the differential value of the current. Can be calculated.
vda = R · id−ω · Lqs · iq (c3)
vqa = R · iq + ω · Lds · id + ω · φ (c4)
In subsequent step S14, the current (ide (n + 1), iqe (n + 1)) in one control cycle ahead is predicted. This is a process of predicting what will happen to the current one control cycle ahead based on the voltage vector V (n) output in step S10. This decomposes the voltage vectors (vd, vq) of the above equations (c1) and (c2) into an average voltage vector (vda, vqa) and an instantaneous voltage vector (vd-vda, vq-vqa), and instantaneously This is performed using the following equations (c5) and (c6) in which the voltage vector (vd−vda, vq−vqa) and the differential term of the current in the above equations (c1) and (c2) are equal. it can.
vd−vda = Ld · (did / dt) (c5)
vq−vqa = Lq · (diq / dt) (c6)
Specifically, it can be performed by the following formulas (c7) and (c8) obtained by discretizing the above formulas (c5) and (c6) with the control cycle Tc.
ide (n + 1)
= Tc · {vd (n) −vda (n)} / Ld + id (n) (c7)
iq (n + 1)
= Tc · {vq (n) −vqa (n)} / Lq + iq (n) (c8)
Incidentally, the voltage vector (vd (n), vq (n)) here uses a conversion matrix based on the electrical angle θ (n) detected in step S10 from the voltage vector V (n) output in step S10. Thus, the voltage component on the dq axis is calculated.

  In subsequent steps S16 to S22, a process of predicting a current two control cycles ahead is performed for each of cases where the switching mode (voltage vector V (n + 1)) in the next control cycle is temporarily set in a plurality of ways. That is, first, in step S16, the voltage vector V (n + 1) is temporarily set. In the subsequent step S18, the average voltage vector (vda (n + 1) is used in the same manner as in step S12 using the predicted currents ide (n + 1) and iqe (n + 1) instead of the actual currents id (n) and iq (n). ), Vqa (n + 1)). Further, in step S20, the predicted currents ide (n + 2) and iqe (n + 2) ahead of two control cycles are calculated in the same manner as in step S14. Here, the predicted currents ide (n + 1) and iqe (n + 1) are used instead of the actual currents id (n) and iq (n), and the instantaneous voltage vectors (vd (n + 1) −vda (n + 1), vq (n + 1) are used. ) -Vqa (n + 1)). The voltage vectors (vd (n + 1), vq (n + 1)) here are rotation angles obtained by adding “ωTc” to the electrical angle θ (n) of the voltage vector V (n + 1) temporarily set in step S16. The voltage component on the dq axis is calculated by conversion using a conversion matrix.

  In step S22, it is determined whether or not the calculation of the predicted currents ide (n + 2) and iqe (n + 2) has been completed for all of the switching modes 0 to 7. If a negative determination is made in step S22, the process returns to step S16. On the other hand, when a positive determination is made in step S22, the process proceeds to step S24.

  In step S24, processing for determining the switching mode (voltage vector V (n + 1)) in the next control cycle is performed. Here, the highest switching mode evaluated by the evaluation function J is defined as the final switching mode (voltage vector V (n + 1)). In the present embodiment, the switching mode is evaluated using the evaluation function J, which is evaluated as the difference between the components of the command current vector and the predicted current vector is larger. In detail, as the evaluation function J, a function whose value becomes larger as the evaluation is lower is adopted. Specifically, the evaluation function J is calculated based on the inner product value of the difference between the command current vector (id *, iq *) and the predicted current vector (ide, iqe). This is a method for expressing that the evaluation is lower as the value is larger in view of the fact that the deviation of each component between the command current vector and the predicted current vector can be both positive and negative values.

  Incidentally, the predicted currents ide (n + 2) and iqe (n + 2) for each of the switching modes 0 to 7 are calculated at the time when an affirmative determination is made in step S22. Therefore, eight values of the evaluation function J can be calculated using these eight predicted currents ide (n + 2) and iqe (n + 2).

  In the subsequent step S26, the parameter n specifying the voltage vector, the current, and the electrical angle sampling number is decreased and corrected by "1", thereby updating the parameter n and ending this series of processes once.

  As described above, in the present embodiment, the means for detecting the current of the electric motor 10 is only the current sensor 16 for detecting the bus current IDC. When the switching mode of the inverter INV corresponds to the effective voltage vector, the absolute value of the bus current IDC matches the current of one phase of the electric motor 10. On the other hand, in order to specify the current flowing through the three-phase rotating machine, at least two-phase current information is required. For this reason, in this embodiment, the detection value of the current detection means is smaller than the number of detection values necessary for specifying the current flowing through the electric motor 10. For this reason, in this embodiment, the current flowing through the electric motor 10 is specified in cooperation with the bus current IDC and the predicted currents ide and iqe, and this is used as the input of the prediction unit 33.

  That is, the three-phase conversion unit 21 shown in FIG. 1 converts the predicted currents ide and iqe into three-phase predicted currents iue, ive and iwe. Based on the current switching mode of the inverter INV, the current reproduction unit 22 selectively dq converts a part of the bus current IDC and the predicted currents iue, ive, and iwe as three-phase current values to obtain the actual current. id and iq are calculated and output to the prediction unit 33. The actual currents id and iq output from the current reproduction unit 22 are calculated using not only the bus current IDC but also the predicted currents iue, ive, and iwe. It is not a detected current value, but a mixed value of the detected value and the estimated value.

  FIG. 4 shows the procedure of the selection process performed by the current reproduction unit 22 according to the present embodiment. This process is executed repeatedly in synchronization with the process shown in FIG. 3 and prior to the process of step S10 in FIG.

  In this series of processes, first, in step S30, whether or not the voltage vector V (n) output in the process of step S10 performed subsequent to this series of processes is one of the voltage vectors V1 and V4. Judging. This process is for determining whether or not the absolute value of the bus current IDC matches the absolute value of the U-phase current. That is, in the case of the voltage vector V1, since the switching elements Sup, Svn, Swn are turned on, the bus current IDC is the sum of the current flowing through the V-phase lower arm and the current flowing through the W-phase lower arm. Corresponds to the current flowing through the U-phase upper arm. On the other hand, in the case of voltage vector V4, switching elements Sun, Svp, and Swp are turned on, so that bus current IDC matches the current flowing through the U-phase lower arm.

  If an affirmative determination is made in step S30, it is determined in step S32 whether or not the voltage vector V (n) is the voltage vector V4. If an affirmative determination is made in step S32, the U-phase actual current iu is set to "-IDC" in step S34, whereas if a negative determination is made in step S32, the U-phase actual current iu is set to "IDC" in step S36. " In the present embodiment, these processes correspond to setting the polarity of the phase current to be positive when the current flows from the inverter INV to the electric motor 10. When the processes in steps S34 and S36 are completed, the actual current iu and the predicted currents ive and iwe are selected in step S38.

  On the other hand, if a negative determination is made in step S30, it is determined in step S40 whether or not voltage vector V (n) is voltage vectors V3 and V6. This process is for determining whether or not the absolute value of the bus current IDC matches the absolute value of the V-phase phase current. If an affirmative determination is made in step S40, in the process of steps S42 to S48, the selection process of whether the actual current iv is set to "IDC" or "-IDC" in the manner of the process of steps S32 to S38. Etc.

  Similarly, if a negative determination is made in step S40, it is determined in step S50 whether or not the voltage vector V (n) is the voltage vectors V2 and V5. This process is for determining whether or not the absolute value of the bus current IDC matches the absolute value of the W-phase current. If an affirmative determination is made in step S50, in the processing of steps S52 to S58, the selection process of whether the actual current iw is set to “IDC” or “−IDC” in the manner of the processing of steps S32 to S38. Etc.

  On the other hand, when a negative determination is made in step S50, since the voltage vector V (n) is a zero voltage vector, the predicted currents iue, ive, and iwe are selected in step S60.

  When the processes in steps S38, S48, S58, and S60 are completed, the selected three-phase current values are dq converted in step S62, and this series of processes is temporarily terminated.

  Of the three-phase predicted currents iue, ive, and iwe, those used in the processes of steps S38, S48, S58, and S60 are the predicted currents ide () in the previous cycle of the process shown in FIG. n + 1) and iqe (n + 1) are three-phase converted values. When the predicted currents ide (n + 1) and iqe (n + 1) are input to the prediction unit 33 via the current reproduction unit 22, they are not estimated values of future currents, but are estimated values of current currents. ing. For this reason, the phase of the bus current IDC and the predicted currents iue, ive, and iwe can be synchronized by synchronizing the detection timing of the bus current IDC with the processing of step S10 shown in FIG.

  However, when the voltage vector is changed, when the voltage vector V (n) is output (updated), the bus current IDC does not become the phase current assumed in steps S38, S48, and S58 due to the dead time. There is a fear. Therefore, actually, it is desirable that the bus current IDC detected immediately before the voltage vector is changed be the actual currents iu, iv, iw selected in steps S38, S48, S58. Further, the processing of steps S38, S48, and S58 may be performed prior to detection of the bus current IDC immediately before the voltage vector is changed. In this case, the processing of steps S38, S48, and S58 is processing for determining the handling of the bus current IDC detected in the near future.

  By the way, the detection value of the current sensor 16 may have a so-called gain error that is an error in which the ratio between the amplitude and the actual current amplitude does not become “1”. Therefore, in the present embodiment, the following processing is performed to reduce the influence of this error from the predicted currents ide and iqe.

  That is, the deviation calculating unit 40 shown in FIG. 1 calculates the difference between the values of the same phase for the d-axis predicted current ide and the actual current id. Here, as a method of setting the same phase value, for example, it is assumed that the actual current id is calculated in the previous step S10 of FIG. 3, and the predicted current ide is the previous value for the series of processes of the previous FIG. The predicted current ide (n + 1) in the control cycle may be used. The feedback control unit 42 calculates a compensation amount idcomp as an operation amount for performing feedback control of the predicted current ide to the actual current id. Specifically, the sum of the outputs of the proportional element, the integral element, and the differential element, each having the output value of the deviation calculating unit 40 as an input, is defined as a compensation amount idcomp.

  Similarly, the deviation calculation unit 44 calculates the difference between the values of the same phase for the q-axis predicted current iqe and the actual current iq. The feedback control unit 46 calculates a compensation amount iqcomp as an operation amount for performing feedback control of the predicted current iq to the actual current iq. Specifically, the compensation amount iqcomp is the sum of the outputs of the proportional element, the integral element, and the differential element, each of which receives the output value of the deviation calculation unit 44.

  Then, the correction unit 48 corrects the actual current id output from the current reproduction unit 22 by the compensation amount idcomp, and outputs this to the prediction unit 33. Further, the correction unit 50 corrects the actual current iq output from the current reproduction unit 22 by the compensation amount iqcomp, and outputs this to the prediction unit 33. The output values of the correction units 48 and 50 are the actual currents id and iq calculated in step S10 of FIG.

  According to such a configuration, the actual currents id and iq input to the prediction unit 33 are feedback-controlled to the predicted currents ide and iqe. Here, the predicted currents ide and iqe correspond to the switching mode selected by the mode determination unit 34. When the bus current IDC has a gain error, even if the prediction unit 33 has no prediction error, the vector norm of the predicted currents ide and iq and the actual currents id and iq calculated from the bus current IDC Deviation occurs from the vector norm. For this reason, this deviation can be used as a parameter having a correlation with the gain error. Therefore, by correcting the input of the prediction unit 33 by the feedback operation amount based on the actual currents id, iq and the predicted currents ide, iqe, the prediction unit 33 even when a gain error occurs in the bus current IDC. The prediction accuracy of the predicted currents ide and iq predicted by can be maintained high.

  FIG. 5 shows the effect of this embodiment when the electric motor 10 is a surface magnet synchronous motor (SPMSM), and FIG. 6 shows the present embodiment when the electric motor 10 is an embedded magnet synchronous motor (IPMSM). The effect of As shown in the figure, the error of the true current value iqR and its average value iqRa with respect to the command current iq * is reduced by using the compensation amount iqcomp.

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

  (1) The actual currents id and iq input to the prediction unit 33 are feedback-controlled to the predicted currents ide and iq having the same phase as this. Thereby, the errors of the predicted currents ide and iq due to the gain error of the current sensor 16 can be reduced, and as a result, the controllability of the current of the electric motor 10 can be improved.

  (2) The feedback control units 42 and 46 are configured to include an integral element. As a result, it becomes easy to compensate for a steady divergence between the vector norm of the predicted currents ide and iq and the vector norm of the actual currents id and iq according to the bus current IDC.

  (3) The feedback control units 42 and 46 are configured to include differential elements. Thereby, even when the harmonic noise is superimposed on the difference between the predicted currents ide, iqe and the actual currents id, iq, the influence can be suitably suppressed.

(4) The current reproduction unit 22 is provided. Thus, the predicted currents ide and iqe can be calculated using only the bus current IDC as the detection value.
<Second Embodiment>
Hereinafter, the second embodiment will be described with reference to the drawings with a focus on differences from the first embodiment.

  FIG. 7 shows a system configuration according to the present embodiment. In FIG. 7, the same reference numerals are given for the sake of convenience to those corresponding to the processes and members shown in FIG. 1.

  In the present embodiment, the change amounts of the actual currents id and iq are fed back to the change amounts Δide and Δiqe of the predicted currents ide and iqe (= ide (n + 2) −ide (n + 1), iqe (n + 2) −iqe (n + 1)). Compensation amounts idcomp and iqcomp are calculated as operation amounts for control.

  That is, the change amount calculation unit 53 calculates the difference between the d-axis predicted current ide (n + 1) and the actual current id (n + 2). Incidentally, the predicted d-axis current ide (n + 1) here is calculated in step S14 of FIG. 3, and the actual current id (n + 2) is a value in step S10 after two control cycles. . The deviation calculation unit 40 calculates the difference between the value obtained by subtracting the predicted current ide (n + 1) from the actual current id (n + 2) and the change amount Δide.

Similarly, the change amount calculation unit 52 calculates the difference between the q-axis predicted current iqe (n + 1) and the actual current iq (n + 2). Incidentally, the q-axis predicted current iqe (n + 1) here is calculated in step S14 of FIG. 3, and the actual current iq (n + 2) is a value in step S10 after two control cycles. . The deviation calculation unit 44 calculates the difference between the value obtained by subtracting the predicted current iqe (n + 1) from the actual current iq (n + 2) and the amount of change Δiqe.
<Third Embodiment>
Hereinafter, the third embodiment will be described with reference to the drawings with a focus on differences from the first embodiment.

  In the present embodiment, the bus current IDC is feedback controlled to a predicted current (predicted currents iue, ive, iqe) in a fixed coordinate system.

  FIG. 8 shows a system configuration according to the present embodiment. In FIG. 8, the same reference numerals are assigned for convenience to the processes and members shown in FIG.

  As shown in the figure, the bus current estimation unit 60 receives the switching mode determined by the mode determination unit 34 and the predicted currents iue, ive, and iwe output from the three-phase conversion unit 21, and calculates the predicted bus current IDCe. calculate. Here, when the switching mode corresponds to the effective voltage vector, the predicted bus current IDCe has a value that is equal in absolute value to any one of the predicted currents iue, ive, and iwe. Specifically, when corresponding to each of the voltage vectors V1, V3, and V5, it is assumed that each of the predicted currents iue, ive, and iwe, and when corresponding to each of the voltage vectors V2, V4, and V6, the predicted currents iue, eve, and iwe are determined. Each value is multiplied by “−1”.

  The division unit 62 calculates the ratio between the bus current IDC and the predicted bus current IDCe. The deviation calculation unit 64 calculates the difference between the ratio and 1 and outputs the difference to the feedback control unit 66. The feedback control unit 66 calculates a correction coefficient K of the bus current IDC for feedback control of the ratio to “1”. Specifically, the correction coefficient K is calculated as the sum of the outputs of the proportional element, the integral element, and the derivative element that receive the output of the deviation calculator 64. The correction unit 68 multiplies the bus current IDC by the correction coefficient K and outputs the result to the current reproduction unit 22.

When the switching mode corresponds to the zero voltage vector, the correction coefficient K calculation process is stopped.
<Fourth Embodiment>
Hereinafter, the fourth embodiment will be described with reference to the drawings with a focus on differences from the first embodiment.

  In the case of the first embodiment, the predicted currents ide and iqe are used as norms to compensate for the gain error of the current sensor 16. However, when there is a model error in the prediction unit 33, there is a concern that the gain error of the current sensor 16 cannot be compensated appropriately. Therefore, in the present embodiment, a process for compensating the model error of the prediction unit 33 is added.

  FIG. 9 shows a system configuration according to the present embodiment. Note that, in FIG. 9, the same reference numerals are assigned for convenience to those corresponding to the processes and members shown in FIG.

  The deviation calculating unit 70 calculates a difference between the predicted current ide and the actual current id having the same phase. The feedback control unit 72 calculates an operation amount (model correction amount Mdcomp) for feedback control of the predicted current ide to the actual current id. Specifically, the model correction amount Mdcomp is calculated as the sum of the outputs of the proportional element, the integral element, and the derivative element that receive the output of the deviation calculating unit 70.

  Similarly, the deviation calculation unit 74 calculates the difference between the predicted current iqe and the actual current iq having the same phase. The feedback control unit 76 calculates an operation amount (model correction amount Mqcomp) for feedback control of the predicted current iq to the actual current iq. More specifically, the model correction amount Mqcomp is calculated as the sum of the outputs of the proportional element, the integral element, and the derivative element that receive the output of the deviation calculator 74. The model correction amounts Mdcomp and Mqcomp are input to the prediction unit 33 as correction amounts of the predicted currents ide and iqe. Specifically, for example, the correction amount is used to correct the predicted currents ide (n + 1) and iqe (n + 1) in step S14 of FIG.

Here, the proportional gain, the integral gain, and the differential gain of the feedback control units 72 and 76 that calculate the model correction amounts Mdcomp and Mqcomp are the proportional gain and the integral gain of the feedback control units 42 and 46 that calculate the compensation amounts idcomp and iqcomp, A value larger than the differential gain is set. Thereby, interference between the process in which the predicted currents ide and iqe are corrected by the model correction amounts Mdcomp and Mqcomp using the information obtained from the bus current IDC as a standard, and the process in which the gain error is compensated by using the predicted currents ide and iqe as a standard Can be suitably avoided. Here, the gains of the feedback control units 42 and 46 are set in consideration of the fact that the change of the gain error is very gradual.
<Other embodiments>
Each of the above embodiments may be modified as follows.

"About norm feedback means"
In the second embodiment, when the minimum current / maximum torque control is performed with the electric motor 10 as the SPMSM, the command current id * is zero, and only the q-axis current is fed back (input for calculating the compensation amount). Parameter).

  In the third embodiment, the target value is “1”, but the present invention is not limited to this. For example, as suggested in FIGS. 5 and 6, in view of the fact that the gain error does not become zero by feedback control, it is considered that the gain error can be further reduced by adapting the target value.

  The compensation amounts idcomp, iqcomp and the correction coefficient K are not limited to the sum of the outputs of the proportional element, the integral element and the differential element. For example, the sum of the outputs of the proportional element and the integral element or the output value of the integral element may be used. However, the output value is not limited to using an integral element, and may be an output value of a proportional element.

  Input parameters such as integration elements are not limited to those exemplified in the above embodiments. For example, the norm of the current vector output by the current reproducing unit 22 and the vector norm of the predicted currents ide and iqe of the prediction unit 33 may be used. If the single correction amount calculated from these differences is allocated to the compensation amounts idcomp and iqcomp according to the phase of the command currents id * and iq * set by the command current setting unit 26, for example, in the first embodiment. The effect according to is produced. Moreover, you may perform the gain correction | amendment of the bus current IDC by the way of the said 3rd Embodiment with the single correction amount computed from the ratio of norms.

"Prediction processing correction means"
The model correction amounts Mdcomp and Mqcomp are not limited to the sum of the outputs of the proportional element, the integral element, and the derivative element. For example, the sum of the outputs of the proportional element and the integral element or the output value of the integral element may be used. However, it is not limited to those using integral elements.

"About current reproduction means"
Not only the bus current IDC alone is used to acquire information that is insufficient from the predicted current. For example, only a current sensor that detects a current flowing through one terminal of the three-phase motor 10 may be provided, and insufficient information may be acquired from the predicted current.

"Average voltage calculation method"
For example, the average value of the voltage vector corresponding to the switching mode used for the actual operation of the inverter INV over a predetermined period may be used as the average voltage.

About prediction means
It is not limited to predicting only the control amount generated by the next voltage vector V (n + 1). For example, the control amount by the operation of the inverter INV at the update timing several control cycles ahead may be sequentially predicted.

  Further, the current is not limited to predicting the current on the dq axis, but may be predicting the current in the fixed coordinate system. In this case, there is no need to provide a rotation coordinate component calculation means for converting the predicted currents ide and iq into components of the fixed coordinate system.

About the decision means
For example, in the first embodiment, the absolute value of the difference between the predicted current ide (n + 2) and the command current id * (n + 2) and the difference between the predicted current iqe (n + 2) and the command current iq * (n + 2) The weighted average processing value with the absolute value may be used as a parameter for evaluating the degree of deviation. In short, in order to quantify that the evaluation becomes lower as the degree of divergence is larger, it may be quantified by a parameter having a positive or negative correlation between the degree of divergence and the evaluation.

"About controlled variables"
The control amount used for determining the operation of the inverter INV (the control amount to be evaluated for the degree of deviation from the command value) is not limited to current. For example, torque and magnetic flux may be used. Further, for example, only torque or only magnetic flux may be used. Even in this case, when current is used for torque or magnetic flux prediction, the predicted current is used as reference current information for grasping the error of the detected current value in the manner of each of the above embodiments. Can do.

  In each of the above embodiments, the ultimate control amount of the rotating machine (the control amount that is ultimately required to be a desired amount regardless of whether or not it is a prediction target) is the torque. For example, the rotational speed may be used.

"About rotating machines"
The rotating machine is not limited to a three-phase rotating machine, and may be a four-phase or more rotating machine such as a five-phase rotating machine. For example, in the case of a five-phase rotating machine, current information for four or more phases is required. However, the number of current sensors is not limited to four, and the shortage may be compensated by a predicted current or the like.

  In the above embodiment, it is assumed that the stator windings are star-connected, but the present invention is not limited to this and may be delta-connected. In this case, the terminal and phase of the rotating machine do not match.

  The rotating machine is not limited to an electric motor but may be a generator.

"others"
The DC voltage source is not limited to the battery 12 and may be, for example, an output terminal of a converter that boosts the voltage of the battery 12.

  As a power conversion circuit comprising a switching element that opens and closes between a different voltage applying means for applying each voltage having a voltage value different from each other and a terminal of the rotating machine, the terminal of the rotating machine is connected to a DC voltage. The present invention is not limited to a DC / AC conversion circuit (inverter INV) including a switching element that is selectively connected to each of the positive electrode and the negative electrode of the source. For example, a switching element that selectively opens and closes between a voltage application unit that applies three or more voltages having different values to each phase of a multiphase rotating machine and a terminal of the rotating machine is provided. Also good. An example of a power conversion circuit for applying three or more different voltages to a terminal of a rotating machine is exemplified in Japanese Patent Application Laid-Open No. 2006-174697.

  DESCRIPTION OF SYMBOLS 10 ... Motor generator, 12 ... Battery (one Embodiment of DC voltage source), 14 ... Control apparatus (One Embodiment of the control apparatus of a rotary machine).

Claims (8)

A power conversion circuit configured to include a switching element that opens and closes between a separate voltage application unit that applies each of voltages having different voltage values and a terminal of the rotating machine, and switching that configures the power conversion circuit In a control device for a rotating machine that controls a control amount having at least one of a current flowing through the rotating machine, a torque of the rotating machine, and a magnetic flux of the rotating machine by an on / off operation of an element,
Prediction means for temporarily setting a switching mode indicating whether each of the switching elements is in an on state or an off state, and performing prediction regarding the control amount according to each of the temporarily set switching modes;
A determination unit that determines a switching mode used for an actual operation of the power conversion circuit based on a prediction result by the prediction unit;
Operating means for operating the power conversion circuit to be in the determined switching mode;
Obtaining means for obtaining a detected value of the current flowing through the rotating machine,
The predicting process by the predicting means includes a current predicting process for predicting the control amount or the current of the rotating machine as a parameter for calculating the control amount,
A norm feedback that feedback-controls the norm of the current vector of the rotating machine according to the detected value of the current to the norm of the current vector predicted by the current prediction process, corresponding to the switching mode adopted by the operating means. A control device for a rotating machine, further comprising means.   In the feedback control by the norm feedback means, the q-axis component of the current flowing through the rotating machine according to the detected value of the current is determined by the operating means of the q-axis component of the current predicted by the current prediction process. 2. The control device for a rotating machine according to claim 1, wherein the control is a feedback control to the one corresponding to the adopted switching mode. The feedback control by the norm feedback means feedback controls the ratio of the current of each terminal of the rotating machine according to the detected value of the current and the current of each terminal predicted by the current prediction process to a target value. The control apparatus for a rotating machine according to claim 1, wherein the control device corrects a detection value of the current used in a process of performing prediction regarding the control amount .   The norm feedback means includes a difference between norms of a current vector flowing through the rotating machine according to a detected value of the current and a current vector predicted by the current prediction process or between parameters having a correlation with the norm. The control device for a rotating machine according to any one of claims 1 to 3, wherein a parameter used in a process of performing prediction related to the control amount is corrected based on an output of an integration element having a difference as an input. .   The norm feedback unit includes a current vector of the rotating machine corresponding to a detected value of the current and a current vector predicted by the current prediction process when correcting a parameter used in the process of performing prediction regarding the control amount. 5. The control device for a rotating machine according to claim 4, further comprising using an output of a differential element that receives a difference between norms of each other or a difference between parameters having a correlation with the norm. The prediction means includes the detected value of the current as a current flowing through a part of the terminals of the rotating machine, and the predicted current as a current flowing through the remaining terminals necessary for calculating a current in the dq coordinate system. And a current reproduction means for converting the current to a current in the dq coordinate system, and the current reproduced by the current reproduction means is used for a process of performing prediction regarding the control amount,
The said norm feedback means uses the said reproduced electric current as a parameter regarding the norm of the current vector flowing through the rotating machine according to the detected value of the electric current. The control apparatus of the described rotating machine. The power conversion circuit is a DC AC conversion circuit including a switching element that selectively connects a terminal of the rotating machine to each of a positive electrode and a negative electrode of a DC voltage source,
The number of terminals that are terminals of the rotating machine and to which the output voltage of the DC / AC converter circuit is applied is three;
The acquisition means acquires a detection value of a current flowing through an input terminal of the DC / AC conversion circuit,
In the prediction process, when the voltage vector corresponding to the switching mode adopted by the operation means is an effective voltage vector, the current detection value flows through any one of the three terminals of the rotating machine. It is used as the initial value of current,
When the voltage vector corresponding to the switching mode adopted by the operating means is an effective voltage vector, the norm feedback means is determined according to the adopted switching mode among the three terminals of the rotating machine. The control device for a rotating machine according to claim 6, wherein a parameter relating to a norm of a current vector flowing through the rotating machine according to the detected value of the current is determined based on a detected value of the current as a current flowing through a terminal. . A prediction process correction unit that corrects the control amount prediction process according to a current evaluation result predicted by the prediction unit based on a detected current value acquired by the acquisition unit;
The control device for a rotating machine according to claim 1, wherein a feedback gain of the norm feedback unit is smaller than a feedback gain of the prediction processing correction unit.