JP5375693B2 - Rotating machine control device - Google Patents

Description

  The present invention controls the amount of control of the rotating machine by operating the output voltage of a power conversion circuit including a switching element that electrically connects each of a positive electrode and a negative electrode of a DC power supply to a terminal of the rotating machine. Predicting means for predicting the control amount of the rotating machine when setting the operation state of the power conversion circuit, and determining the actual operation state of the power conversion circuit based on the predicted control amount, the determined The present invention relates to a control device for a rotating machine including operating means for operating the power conversion circuit so as to be in an operating state.

  As this type of control device, for example, as shown in Patent Document 1 below, there is one that performs model predictive control. In this apparatus, first, a trajectory of torque at sampling timings k to k + N when a switching state sequence at each sampling timing k, k + 1,... K + N−1 (N ≧ 1) is temporarily set is predicted. Next, a subsequent locus of torque is predicted by extrapolation based on the locus of torque at sampling timings k + N−1 to k + N. Next, based on the number of samplings n required for the torque predicted by extrapolation to be out of the predetermined range, the number of switching of the switching state determined by the temporary setting of the switching sequence is divided. Then, the switching state at the sampling timing k in the switching sequence that minimizes the division value is determined as the actual switching state at the sampling timing k. As a result, the number of switching of the switching state can be limited.

JP 2006-174697 A

  The division value becomes smaller as the period until the extrapolated value of torque deviates from the predetermined range is longer. For this reason, even in a switching sequence in which the total number of switching states in the sampling timings k to k + N−1 is large, the divided value is small when the number of times of sampling n is large. For this reason, in the said apparatus, the number of switching of a switching state cannot necessarily be fully restrict | limited. In particular, since the prediction accuracy of the behavior of the torque by extrapolation is not necessarily high, the above-described device has a variable weight for evaluating the switching number of the switching state based on the prediction with low accuracy about the torque. There is a problem in reliability as processing for suppressing the number of switching.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to provide a control device for a rotating machine that can suitably suppress the number of switching states by model predictive control.

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

  According to the first aspect of the present invention, the control amount of the rotating machine is controlled by operating the output voltage of a power conversion circuit including a switching element that electrically connects each of the positive electrode and the negative electrode of the DC power source to the terminal of the rotating machine. Therefore, based on the predicted control amount, predicting means for predicting the control amount of the rotating machine when the operation state of the power conversion circuit is set, determine the actual operation state of the power conversion circuit, An average voltage direction calculating means for calculating the direction of an average output voltage vector of the power conversion circuit in a control device for a rotating machine comprising operating means for operating the power conversion circuit so as to be in the determined operation state The operating means includes an effective voltage vector representing the operating state, and an angle formed with a direction calculated by the average voltage direction calculating means is less than a specified angle of 20 degrees or less. On condition that there is to be, characterized in that it comprises a zero-voltage vector priority means that the highest priority relating to the determination of the actual operating state the operating state represented by the zero voltage vector.

  The average output voltage vector is considered to be an appropriate output voltage vector for controlling the control amount. On the other hand, when there is an effective voltage vector having a small angle with the average output voltage vector, the contribution of the effective voltage vector is small in realizing the average effective voltage vector. In the above invention, in view of this point, switching of the switching state is suppressed by giving priority to the operation state represented by the zero voltage vector under such a situation.

  The invention according to claim 2 is the invention according to claim 1, wherein the difference between the predicted value of the control amount and the command value of the control amount when the current operation state is set as the operation state of the power conversion circuit is An allowable range determining means for determining whether or not the allowable range is deviated, and when the operating means is determined to be within the allowable range by the allowable range determining means, the current operating state of the power conversion circuit is determined. State maintaining priority means for determining that the priority is high is further provided, and the zero voltage vector priority means has a logical product of that it is determined that the zero voltage vector priority means is out of the allowable range and that there is a predetermined angle or less. In this case, the priority of the operation state expressed by the zero voltage vector is the highest.

  In the above invention, since the priority of the current operation state is high on condition that the difference is determined to be within the allowable range, switching of the operation state is suppressed while maintaining controllability of the control amount. can do.

  According to a third aspect of the present invention, in the second aspect of the present invention, the zero voltage vector priority means sets the zero voltage vector to the zero voltage vector when the priority of the operation state expressed by the zero voltage vector is the highest. When the operation state represented by the prediction means is set as the prediction target and the difference between the predicted value and the command value is within the allowable range, the operation state represented by the zero voltage vector is actually The operation state is determined as follows.

  According to a fourth aspect of the present invention, in the second or third aspect of the present invention, when the operation means determines that the operation state is determined to be out of the allowable range, a voltage vector representing an operation state to be predicted is a zero voltage vector. The effective voltage vector priority means for making the priority of at least one of the operation states expressed by a pair of effective voltage vectors having a small angle with the average output voltage vector the highest. Features.

  When the difference deviates from the allowable range due to the operation state expressed by the zero voltage vector, it is desirable to use the operation state expressed by the effective voltage vector in order to make the difference the allowable range. Setting this difference within the allowable range means selecting an appropriate operation state for controlling the controlled variable, but an appropriate output voltage vector for controlling the controlled variable is an average output voltage vector. It is believed that there is. The minimum effective voltage vector necessary for realizing the average output voltage vector is the pair of effective voltage vectors. In the above invention, by paying attention to this point, it is possible to give priority to an effective voltage vector that is particularly appropriate for making the difference within the allowable range.

  The invention according to claim 5 is connected to the switching element whose switching state is switched when determining the actual operation state of the power conversion circuit in the invention according to any one of claims 1 to 4. The apparatus further comprises prohibiting means for prohibiting the number of terminals of the rotating machine from becoming larger than one.

  In the above invention, the number of switching states can be suppressed.

  The invention according to claim 6 is the invention according to claim 5, wherein the prediction means excludes an operation state prohibited by the prohibition means from a prediction target by the prediction means.

  In the said invention, a calculation load can be reduced suitably.

  The invention according to claim 7 is the invention according to any one of claims 1 to 6, wherein the prediction means excludes the operation state having the low priority from the prediction target by the prediction means, The operation means determines an actual operation state from among the prediction targets.

  In the said invention, a calculation load can be reduced by excluding the operation state with low priority from a prediction object.

  The invention according to claim 8 is the invention according to any one of claims 1 to 7, wherein the prediction means predicts the control amount when setting an operation state of the power conversion circuit. And second predicting means for predicting an initial value used for prediction by the first predicting means with an operation state already determined to be used by the operating means as an input.

  In predicting the control amount in the setting period when the operation state of the power conversion circuit is set with high accuracy, the initial value of the prediction process at the time of approximation as much as possible to the adoption time of the above setting (operation state update timing) ( It is desired to obtain a control amount or a physical quantity that can be calculated. However, since it is necessary to predict the control amount before the time when the operation is actually performed, it is impossible or difficult to obtain the initial value by detection at this time or the time approximated to this time as much as possible. . In this regard, in the above-described invention, by providing the second prediction unit, the initial value used for the prediction of the first prediction unit can be set to an appropriate value as much as possible. For this reason, the control amount can be predicted with high accuracy. For this reason, especially in the case where the above-described allowable range determining means is provided, the control performed based on the determination by the allowable range determining means can be made appropriate.

  Incidentally, when the second prediction means is not provided, the control amount prediction is approximated to the prediction accuracy when the second prediction means is provided for a plurality of control destinations, but control of one control cycle ahead from the next update timing. It has been found by the inventors that the accuracy of quantity prediction is inferior to that provided with the second prediction means.

1 is a system configuration diagram according to a first embodiment. FIG. The figure which shows the voltage vector expressing the operation state of an inverter. The time chart which shows the problem of model predictive control. The figure which compares the switching pattern of triangular wave comparison PWM control and model prediction control. The figure which shows the normative switching transition concerning the said embodiment. The flowchart which shows the process sequence of the model prediction control concerning the embodiment. The flowchart which shows the procedure of the prediction process of the electric current in the said model prediction control. The figure which shows the process which specifies the area | region of the average voltage vector concerning the embodiment. The flowchart which shows the procedure of the change examination process of the voltage vector in the said model prediction control. The time chart which shows the effect of the embodiment. The figure which shows the effect of the same embodiment. The time chart which shows the effect of the embodiment. The flowchart which shows the process sequence of the model prediction control concerning 2nd Embodiment. The figure which shows the permission conditions of the change examination process of the voltage vector concerning the embodiment. The flowchart which shows the process sequence of the model prediction control concerning 3rd Embodiment.

(First embodiment)
Hereinafter, a first embodiment in which a rotating machine control device according to the present invention is applied to a hybrid vehicle control device will be described with reference to the drawings.

  FIG. 1 shows the overall configuration of a motor generator control system according to this embodiment. The motor generator 10 is a three-phase permanent magnet synchronous motor. The motor generator 10 is a rotating machine (saliency pole machine) having saliency. Specifically, the motor generator 10 is an embedded magnet synchronous motor (IPMSM).

  The motor generator 10 is connected to the high voltage battery 12 via the inverter IV. The inverter IV includes a series connection body of the switching elements Sup and Sun, a series connection body of the switching elements Svp and Svn, and a series connection body of the switching elements Swp and Swn. The motor generator 10 is connected to the U, V, and W phases, respectively. In the present embodiment, insulated gate bipolar transistors (IGBTs) are used as the switching elements Sup, Sun, Svp, Svn, Swp, and Swn. In addition, diodes Dup, Dun, Dvp, Dvn, Dwp, and Dwn are connected in antiparallel to these.

  In this embodiment, the following is provided as detection means for detecting the state of the motor generator 10 and the inverter IV. First, a rotation angle sensor 14 for detecting the rotation angle (electrical angle θ) of the motor generator 10 is provided. Further, a current sensor 16 that detects currents iu, iv, and iw flowing through the phases of the motor generator 10 is provided. Further, a voltage sensor 18 for detecting an input voltage (power supply voltage VDC) of the inverter IV is provided.

  The detection values of the various sensors are taken into the control device 20 constituting the low voltage system via an interface (not shown). The control device 20 generates and outputs an operation signal for operating the inverter IV based on the detection values of these various sensors. Here, the signals for operating the switching elements Sup, Sun, Svp, Svn, Swp, Swn of the inverter IV are the operation signals gup, gun, gvp, gvn, gwp, gwn.

  The control device 20 operates the inverter IV to control the torque of the motor generator 10 to the required torque Tr. Specifically, the inverter IV is operated so that a command current for realizing the required torque Tr is obtained. That is, in this embodiment, the torque of the motor generator 10 becomes the final control amount, but in order to control the torque, the current flowing through the motor generator 10 is directly controlled as a command current. . In particular, in this embodiment, in order to control the current flowing through the motor generator 10 to the command current, the current flowing through the motor generator 10 when the operation state of the inverter IV is temporarily set is predicted, and the predicted current and the command current are calculated. Model predictive control is performed to determine the actual operation state of the inverter IV based on the difference.

  Specifically, the phase currents iu, iv, iw detected by the current sensor 16 are converted into actual currents id, iq in the rotating coordinate system by the dq converter 22. Further, the electrical angle θ detected by the angle sensor 14 becomes an input to the speed calculation unit 23, and thereby the rotational speed (electrical angular speed ω) is calculated. On the other hand, the command current setting unit 24 receives the required torque Tr and outputs command currents idr and iqr in the dq coordinate system. The command currents idr and iqr, the actual currents id and iq, and the electrical angle θ are input to the model prediction control unit 30. Based on these input parameters, the model prediction control unit 30 determines a voltage vector Vi that defines the operation state of the inverter IV and inputs the voltage vector Vi to the operation unit 26. The operation unit 26 generates the operation signal based on the input voltage vector Vi and outputs it to the inverter IV.

  Here, the voltage vectors expressing the operation state of the inverter IV are eight voltage vectors shown in FIG. For example, a voltage vector representing an operation state (indicated as “lower” in the drawing) in which the low-potential side switching elements Sun, Svn, Swn are turned on is the voltage vector V0, and the high-potential side switching elements Sup, A voltage vector representing an operation state (indicated as “upper” in the drawing) in which Svp and Swp are turned on is a voltage vector V7. These voltage vectors V0 and V7 are for short-circuiting all phases of the motor generator 10 and are called zero voltage vectors because the voltage applied to the motor generator 10 from the inverter IV becomes zero. On the other hand, the remaining six voltage vectors V1 to V6 are defined by an operation state in which switching elements that are turned on exist in both the upper arm and the lower arm, and are called effective voltage vectors. Yes. FIG. 2B is a diagram in which the effective voltage vectors V1 to V6 are converted into a fixed two-dimensional coordinate system with the zero voltage vectors V0 and V7 as the origin. As illustrated, each of the voltage vectors V1, V3, and V5 corresponds to the positive side of the U phase, the V phase, and the W phase, respectively.

  Next, details of the processing of the model prediction control unit 30 will be described. In the operation state setting unit 31 shown in FIG. 1, the operation state of the inverter IV is set. Here, voltage vectors V0 to V7 shown in FIG. 2 are set as the operation state of inverter IV. The dq conversion unit 32 calculates the voltage vector Vdq = (vd, vq) in the dq coordinate system by performing dq conversion on the voltage vector set by the operation state setting unit 31. In order to perform such conversion, the voltage vectors V0 to V7 in the operation state setting unit 31 are, for example, “upper” as “VDC / 2” and “lower” as “−VDC / 2” in FIG. It can be expressed as above. 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 operation state of the inverter IV is set to the state set by the operation state setting unit 31 based on the voltage vector (vd, vq), the actual currents id, iq, and the electrical angular velocity ω. , Iq is predicted. Here, the voltage equation expressed by the following (c1) and (c2) is discretized from the following state equations (formulas (c3) and (c4)) obtained by solving the current differential term. Predict.
vd = (R + pLd) id−ωLqiq (c1)
vq = ωLdid (R + pLq) iq + ωφ (c2)
pid
= − (R / Ld) id + ω (Lq / Ld) iq + vd / Ld (c3)
piq
= −ω (Ld / Lq) id− (Rd / Lq) iq + vq / Lq−ωφ / Lq (c4)
Incidentally, in the above formulas (c1) and (c2), the resistance R, the differential operator p, the d-axis inductance Ld, the q-axis inductance Lq, and the armature linkage flux constant φ are used.

  The prediction of the current is performed for each of a plurality of operation states set by the operation state setting unit 31.

  On the other hand, the operation state determination unit 34 inputs the currents ide and iq predicted by the prediction unit 33 and the command currents idr and iqr, and determines the operation state of the inverter IV. In one of the determination processes, the evaluation function J is used. That is, each operation state set by the operation state setting unit 31 is evaluated by the evaluation function J, and the operation state having the highest evaluation is selected. As this evaluation function J, in the present embodiment, a function whose value increases as the evaluation becomes lower is adopted. Specifically, the evaluation function J is calculated based on the inner product value of the difference between the command current vector Idqr = (idr, iqr) and the predicted current vector Idqe = (ide, iqe). This is a technique for expressing that the evaluation is lower as the value is larger in view of the fact that the deviation of each component between the command current vector Idqr and the predicted current vector Idqe can be both positive and negative values. As a result, it is possible to construct an evaluation function J in which the evaluation becomes lower as the difference between the components of the command current vector Idqr and the predicted current vector Idqe is larger.

  If the evaluation function J is used, an operation state in which the difference between the predicted current vector Idqe and the command current vector Idqr can be selected in each control cycle Tc can be selected, while the switching state is switched simultaneously. The number of phases of the generator 10 may increase. In this case, since the surge is increased, the withstand voltage required for the switching element is increased.

  Here, the inventors initially considered control for limiting the number of phases that can be switched simultaneously to 1 or less. Specifically, in order to maintain controllability to the command current, the above limitation is performed when the error between the command current and the predicted current is within the allowable range, and when the error is outside the allowable range, Of these, a process of selecting the one having the highest evaluation by the evaluation function J was considered. A switching example of the switching state at this time is shown in FIG. As shown in the figure, it has been found that in this case, the switching frequency between operation states expressed by different effective voltage vectors is high, and as a result, the number of switching of the switching state per unit time is increased. An example of switching of the switching state at this time and an example of switching of the switching state by the conventional triangular wave comparison PWM processing are shown in comparison with FIG.

  As shown in FIG. 4A, in the triangular wave comparison PWM process, the actual current vector Idq is subtracted from the command current vector Idqr by a pattern in which two effective voltage vectors and one zero voltage vector are periodically switched. The end point of the error vector edq (the coordinate component when the start point of the error vector edq is the origin) is displaced. In the figure, an average voltage vector Va is also shown.

  Here, the average voltage vector Va is a fundamental wave component having an electrical angular frequency in the output voltage of the inverter IV. In other words, the inverter IV switches the switching state at a time interval shorter than one electrical angular cycle, so that the output voltage simulates a sinusoidal voltage having an electrical angular frequency component. The sinusoidal voltage simulated by the inverter IV is the average voltage vector Va. Incidentally, the norm of the average voltage vector Va is a physical quantity proportional to the modulation rate and the voltage utilization rate. Here, the modulation rate is a Fourier coefficient of the fundamental wave component for the output voltage of the inverter IV. In calculating the Fourier coefficient, the amplitude center of the fundamental wave and the median value of the fluctuation range of the output voltage of the inverter IV are matched.

  The end point of the error vector edq is displaced in the vector direction obtained by subtracting the voltage vector expressing the operation state of the inverter IV from the illustrated average voltage vector Va. Therefore, when the vector representing the operation state is a zero voltage vector, the end point of the error vector edq is displaced in the direction of the average voltage vector Va. In particular, in the triangular wave comparison PWM process, the end point of the error vector edq is a pair of effective voltages having a small angle with the average voltage vector Va when the start point of the effective voltage vector and the average voltage vector Va is the start point of the error vector edq. When the operation state represented by the zero voltage vector is selected when it is within the region surrounded by the vector (V3, V4 in the figure) and the region reversed by “180 °”, it is represented by the zero voltage vector. The period for which the operation state is kept is long. Moreover, the end point of the error vector edq is displaced by the operation state expressed by the zero voltage vector and the norm of the error vector edq is increased to some extent by the displacement of the region surrounded by the pair of effective voltage vectors. Therefore, a period in which the actual current is larger than the command current and a period in which the actual current is smaller than the command current are alternately visited. For this reason, the average electric current over one period of this pattern can be made to follow a command current suitably.

  On the other hand, in the example of model predictive control shown in FIG. 4B, the switching state is frequently switched. This is due to the fact that the operation state that can obtain the highest evaluation by the evaluation function J is selected each time. As described above, when model predictive control is used, the number of switching states tends to increase by selecting an optimal solution on a microscopic time scale. However, this does not necessarily occur in model predictive control. For example, the operation state in the next control cycle Tc is predicted based on the predicted currents up to a plurality of control cycles Tc ahead, and the tendency to select the optimal solution on the microscopic time scale by extending the prediction period. It can be suppressed. However, in this case, the calculation load of the control device 20 increases. Incidentally, the error vector edq in FIG. 4B is obtained by subtracting the actual current vector Idq from the command current vector Idqr.

  Therefore, the present inventor considered that the average voltage vector Va is referred to when determining the operation state in the next control cycle Tc. Here, it is considered that average voltage vector Va is appropriate for setting actual currents flowing through motor generator 10 as command currents idr and iqr. For this reason, if the average voltage vector Va is referred to, it is considered that selection of the optimum operation state in a time scale longer than the control cycle Tc can be realized by model predictive control without extending the prediction period.

  FIG. 5A shows the transition of the operation state that is prioritized by the model predictive control in the present embodiment. As shown in the figure, at a point P1 where the current error deviates from the allowable range (the norm of the error vector edq is greater than the threshold value eth), one of a pair of effective voltage vectors having a small angle with the average voltage vector Va. (V3 in the figure) is selected. Thereafter, the other (V4 in the figure) of the pair of effective voltage vectors is selected at the point P2. A zero voltage vector (V7 in the figure) is selected at a point P3 where the current error deviates from the allowable range again (the norm of the error vector edq is greater than the threshold value eth). Thereby, like the triangular wave comparison PWM process, the period of the operation state expressed by the zero voltage vector can be lengthened, and the number of switching of the switching state can be reduced.

  As the point P2, an appropriate point is selected in order to control the point P3 to be a region inverted by “180 °” from the region surrounded by the pair of effective voltage vectors having a small angle with the average voltage vector Va. Should. In view of this point, in the present embodiment, as shown in FIG. 5B, the time point when the magnitude relationship between the norm | Idqr | of the command current vector Idqr and the norm | Idqe | Let P2.

  FIG. 6 shows a processing procedure of model predictive control according to the present embodiment. This process is repeatedly executed by the control device 20 at the control cycle Tc.

  In this series of processing, first, in step S10, the voltage representing the current (current) operation state is used as the voltage vector V (n + 1) representing the operation state at the next update timing among the update timings visited every control cycle Tc. A vector V (n) is temporarily set. In the subsequent step S12, a process of predicting a predicted current vector Idqe (n + 2) ahead of one control cycle Tc when an operation state expressed by the voltage vector V (n + 1) is adopted at the next update timing is performed. .

  FIG. 7 shows details of this processing.

  In this series of processes, first, in step S12a, the electrical angle θ (n) and the actual currents id (n) and iq (n) are detected, and the voltage vector V (n determined in the previous control cycle Tc is detected. ) Is output. In the subsequent step S12b, 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 S12a. Here, the currents ide (n + 1) and iqe (n + 1) are calculated by using the model expressed by the above equations (c3) and (c4) discretized by the forward difference method with the control cycle Tc. . At this time, the actual currents id (n) and iq (n) detected in step S12a are used as initial values of the current, and the voltage vector V (n) is used as the voltage vector on the dq axis. The one obtained by dq conversion using θ (n) detected in step 1 is used.

  In the subsequent step S12c, a process of predicting a current two control cycles ahead is performed when the voltage vector V (n + 1) at the next update timing is set. That is, the predicted currents ide (n + 2) and iqe (n + 2) are calculated in the same manner as in step S12b. However, here, the predicted currents ide (n + 1) and iqe (n + 1) calculated in step S12b are used as the initial value of the current, and the voltage vector V (n + 1) is used as the voltage vector on the dq axis. The electrical angle θ (n) detected in step S12a is subjected to dq conversion by an angle obtained by adding ωTc. When the process of step S12c is completed, the process returns to the process of FIG.

  In step S14 of FIG. 6, an error vector edq is calculated by subtracting the predicted current vector Idqe (n + 2) from the command current vector Idqr. In the subsequent step S16, an average voltage vector Va is calculated. Here, the average voltage vector Va is calculated by inputting the command current vector Idqr to the equations (c1) and (c2) from which the differential operator p is removed. That is, in consideration of the fact that the average current flowing through the motor generator 10 is the command currents idr and iqr except for the current ripple caused by the switching of the switching state, the command currents idr and iqr flow through the motor generator 10 steadily. An average voltage vector Va is calculated as a voltage applied thereto.

  In subsequent step S18, it is determined whether or not the current error is within an allowable range (whether or not norm | edq | of error vector edq is equal to or smaller than threshold value eth). Here, the threshold value eth is desirably variably set according to the state quantity of the motor generator 10 (current amplitude, electrical angular velocity ω, etc.). If it is determined that it is within the allowable range, it is determined in step S20 whether or not the magnitude relationship between the norm | Idqr | of the command current vector Idqr and the norm | Idqe | of the predicted current vector Idqe is reversed. If the state is reversed, the state transition permission flag F is set to “1”. However, the condition for setting the state transition permission flag F to “1” is that the current voltage vector V (n) is one of a pair of effective voltage vectors having a small angle with the average voltage vector Va. Add further conditions. That is, the state transition permission flag F is set to “1” when the logical product of the condition that the current voltage vector V (n) is one of the above and the condition that it is inverted is true. The

  Here, the determination as to whether or not one of the above can be made based on the process of specifying the existence region of the average voltage vector Va. That is, as shown in FIG. 8, when the starting point of the average voltage vector Va and the effective voltage vectors V1 to V6 is shared, any of the regions S1 to S6 surrounded by the pair of adjacent effective voltage vectors V1 to V6 is used. This can be done based on whether the average voltage vector Va exists. Here, the average voltage vector Va and the voltage vector V1 are obtained by converting (Vαa, Vβa) a component of the rotating two-dimensional coordinate system of the average voltage vector Va into a component of a fixed two-dimensional coordinate system (αβ coordinate system). By calculating the formed angle θva, it is specified which of the regions S1 to S6 exists. As shown in the figure, these regions S1 to S6 all have an angle region of “π / 3”.

  If it is specified in any of the regions S1 to S6 in this way, a pair of effective voltage vectors having a small angle with the average voltage vector Va is automatically determined.

  If a negative determination is made in step S18 of FIG. 6 or an affirmative determination is made in step S20, the process proceeds to step S22, and a process for examining the change of the voltage vector V (n + 1) at the next update timing is performed. Do. On the other hand, when the process of step S22 is completed, or when a negative determination is made in step S20, this series of processes is temporarily terminated.

  FIG. 9 shows details of the process in step S22.

  In this series of processing, first, in step S30, it is determined whether or not the state transition permission flag F is “1”. If it is determined that the state transition permission flag F is “1”, in step S32, the current voltage vector V (n) that is not the current voltage vector V (n) is selected from the pair of effective voltage vectors having a small angle with the average voltage vector Va. The operation state represented by (solid line in the figure) is assumed to be the highest priority and is considered.

  On the other hand, if a negative determination is made in step S30, it is determined in step S34 whether or not the current voltage vector V (n) is an effective voltage vector. This process is for giving priority to a specific effective voltage vector at the point P1 in FIG. That is, when a negative determination is made in step S34, in step S36, the number of switching phases of the switching state from the current voltage vector V (n) of the pair of effective voltage vectors having a small angle with the average voltage vector Va is determined. Considering that the priority of the operation state expressed by the person who becomes “1” or less is the highest priority, this is considered. For example, when the pair of effective voltage vectors is the effective voltage vectors V3 and V4 and the current voltage vector is the zero voltage vector V0, the number of switching phases to the effective voltage vector V3 is “1”, while the effective voltage vector Since the number of switching phases to V4 is “2”, the effective voltage vector V3 is considered.

  On the other hand, when an affirmative determination is made in step S34, in step S38, there is an effective voltage vector Vi whose angle with the average voltage vector Va is A (≦ 20 °) or less, and the current voltage vector V It is determined whether or not the logical sum of the voltage vector immediately before switching to (n) being an effective voltage vector is true. Here, the second condition is for giving priority to the zero voltage vector at the point P3 in FIG. The first condition is that the zero voltage vector in view of the fact that the effective voltage vector Vi hardly contributes to the generation of the average voltage vector Va when there is an effective voltage vector Vi having a small angle with the average voltage vector Va. Is to prioritize. When the logical sum is true, in step S40, the priority of the operation state expressed by the zero voltage vector in which the number of switching phases of the switching state from the current voltage vector V (n) is “1” or less is set. As the highest, this is considered. For example, when the current voltage vector V (n) is V4, the operation state expressed by the zero voltage vector V7 is considered, and when the current voltage vector V (n) is V3, the zero voltage vector The operation state expressed by V0 is considered.

  When the processes in steps S32, S36, and S40 are completed, the process proceeds to step S42. In step S42, a predicted current Idqe vector (n + 2) is calculated for the case where the operation state expressed by the voltage vector to be examined is set, and the norm | edq | of the error vector edq is a threshold value. It is determined whether it is equal to or less than eth. And when it is judged that it is below threshold value eth, the voltage vector made into examination object in Step S46 is adopted.

  On the other hand, when a negative determination is made in step S42 or step S38, among all those in which the number of switching phases in the switching state from the current voltage vector V (n) is “1” or less in step S44. The one having the highest evaluation by the evaluation function J is adopted. For example, when the current voltage vector V (n) is the effective voltage vector V3, the one having the highest evaluation by the evaluation function J among the effective voltage vectors V2, V3, V4 and the zero voltage vector V0 is adopted.

  When the processes in steps S46 and S44 are completed, this series of processes is temporarily terminated.

  FIG. 10 shows the effect of this embodiment. Specifically, FIG. 10A shows the current flowing through the U phase of the motor generator 10 according to the present embodiment and the transition of the voltage vector, and FIG. 10B shows a comparative example with reference to FIG. The case where the above control is performed is shown. As shown in the figure, in this embodiment, the number of switching states can be suitably reduced.

  FIG. 11 shows a comparison result of the present embodiment and the triangular wave comparison PWM processing at a plurality of measurement points. Here, FIG. 11A shows the measurement points, FIG. 11B shows the number of times of switching at each measurement point, and FIG. 11C shows the effective value of the harmonic current at each measurement point. . Here, FIG.11 (b) and FIG.11 (c) have shown the average value in the three measurement points in four torque values which are mutually different in Fig.11 (a). Further, the effective value of the harmonic current in FIG. 11C is obtained by integrating the square root of the square of the difference between the actual current flowing through the U phase of the motor generator 10 and the command value over one cycle of the electrical angle. It is.

  As shown in the figure, the number of switching of the switching state is smaller than the triangular wave comparison PWM process at any measurement point. Also, the effective value of the harmonic current can be reduced to be equal to or greater than that of the triangular wave comparison PWM processing.

  FIG. 12 shows an example of switching of the switching state according to this embodiment. As shown in the figure, in this embodiment, the voltage vector may be changed in the same manner as in the triangular wave comparison PWM process, or the number of voltage vector transitions may be smaller than in the triangular wave comparison PWM process.

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

  (1) Based on the direction of the average voltage vector Va, the priority for determining the actual operation state of each operation state of the inverter IV is set. Thereby, the optimal operation state in a time scale larger than the time scale of the control cycle Tc can be easily grasped.

  (2) The current operation state has the highest priority on the condition that the current error is determined to be within the allowable range (the norm | edq | of the error vector edq is determined to be equal to or less than the threshold eth). It was. Thereby, switching of an operation state can be suppressed, maintaining the controllability of an electric current.

  (3) A voltage vector that represents an operation state that is a prediction target when the current error is determined to be out of the allowable range (the norm | edq | of the error vector edq is determined to be larger than the threshold eth) is valid. If the logical product of the condition that it is a voltage vector and the condition that the voltage vector that represents the operation state before switching to the current operation state is an effective voltage vector is true, it is expressed as a zero voltage vector The priority of the operation state to be performed is the highest. As a result, it is possible to perform control that keeps the error within an allowable range for a long time using the zero voltage vector.

  (4) A voltage vector that represents an operation state that is a prediction target when the current error is determined to be out of the allowable range (the norm | edq | of the error vector edq is determined to be larger than the threshold eth) is valid. When the logical product of the condition that a voltage vector is present and the condition that an effective voltage vector that has an angle with the average voltage vector Va that is equal to or smaller than the specified angle A is true, a zero voltage vector The priority of the operation state expressed by is assumed to be the highest. As a result, it is possible to perform control that keeps the error within an allowable range for a long time using the zero voltage vector.

  (5) When the current error is determined to be out of the allowable range (the norm | edq | of the error vector edq is determined to be larger than the threshold value eth), the voltage vector expressing the operation state that is the prediction target is zero. In the case of a voltage vector, the highest priority is given to the operation state in which the number of switching phases in the switching state is “1” or less among the operation states represented by a pair of effective voltage vectors having a small angle with the average voltage vector Va. did. Accordingly, it is possible to give priority to an effective voltage vector that is particularly appropriate for setting the current error within an allowable range.

  (6) When the magnitude relationship between the norm | Idqr | of the command current vector Idqr and the norm | Idqe | of the predicted current vector Idqe is inverted, a pair of effective voltage vectors adjacent to the average voltage vector Va and the current The highest priority is given to those that do not correspond to the operation state. As a result, the end point of the error vector edq can be controlled to the region side that is “180 °” reversed from the region where the average voltage vector Va is present.

  (7) The number of phases of the motor generator 10 whose switching state is switched when the difference between the predicted value and the command value regarding the operation state determined to be out of the allowable range and the highest priority is not within the allowable range Among the operation states in which is 1 or less, the one with the highest evaluation by the evaluation function J is set as the next operation state. Thereby, when an error cannot be within an allowable range depending on an operation state having the highest priority, an appropriate operation state can be selected while suppressing the number of switching states.

  (8) When the state transition permission flag F is “1”, the difference between the predicted value and the command value related to the operation state to be examined does not fall within the allowable range. Among the operation states in which the number is 1 or less, the operation state having the highest evaluation by the evaluation function J is set as the next operation state. Thereby, when an unexpected behavior occurs, it is possible to select an appropriate operation state while suppressing the number of switching states.

  (9) The number of phases of the motor generator 10 whose switching state is switched is prohibited from being larger than 1. Thereby, the magnitude | size of the surge resulting from switching of a switching state can be reduced.

  (10) The operation state in which the number of phases of the motor generator 10 to which the switching state is switched is greater than 1 is excluded from the prediction target. Thereby, calculation load can be reduced suitably.

  (11) The average voltage vector Va was calculated with the command currents idr and iqr as inputs. Thereby, the direction of the average output voltage vector can be calculated appropriately.

  (12) The initial value of the current is predicted with the operation state already determined to be used as the operation state of the inverter IV as an input. Thereby, the prediction of the current by the model predictive control can be performed with high accuracy.

  (13) Current detection values (id (n), iq (n)) are used as inputs, and current initial values (ide (n + 1), iqe (n + 1)) at the next operation state update timing (update timing n + 1). Predicted. As a result, the current can be predicted with higher accuracy.

  (14) The operation state at the update timing n + 1 is determined based on the predicted current value when one control cycle Tc has elapsed from the update timing n + 1 of the next operation state. Thereby, the operation state of inverter IV can be determined suitably.

  (15) The current (id (n), iq (n)) detected at the current update timing is the current used for the prediction process of the initial value of the current prediction according to the operation state setting at the next update timing. . As a result, the current prediction process associated with the setting of the operation state at the next update timing can be made substantially the same as the current initial value prediction process used for the prediction. For this reason, there are merits that the design of the prediction processing means becomes easy and the processing means (arithmetic program) is partially shared.

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

  FIG. 13 shows a processing procedure of model predictive control according to the present embodiment. This process is repeatedly executed by the control device 20 at a predetermined cycle, for example. In FIG. 13, processes corresponding to the processes shown in FIG. 6 are given the same step numbers for convenience.

  In this series of processing, the condition for setting the state transition permission flag F to “1” is different. In other words, instead of the condition that the magnitude relationship between the norm | Idqr | of the command current vector Idqr and the norm | Idqe | Employs a condition that the line reaches a straight line (two-dot chain line in FIG. 14) orthogonal to the average voltage vector Va and passing through the origin.

  Also according to the present embodiment described above, it is possible to obtain effects according to the above-described effects of the first embodiment.

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

  FIG. 15 shows a processing procedure of model predictive control according to the present embodiment. This process is repeatedly executed by the control device 20 at a predetermined cycle, for example. In FIG. 15, processes corresponding to the processes shown in FIG. 6 are given the same step numbers for convenience.

  In this series of processing, processing for calculating the average voltage vector Va is different. That is, in step S16a, the voltage vectors V (n), V (n−1),... V (representing the operation state actually adopted at the past N (> 2) update timings up to the previous update timing. The average voltage vector Va is calculated by the simple moving average process of (n−N). Here, the number of times N is variably set according to the electrical angular velocity ω. This is a setting corresponding to the fact that the control cycle Tc is constant. That is, since the average voltage vector Va is a voltage vector corresponding to the fundamental frequency, when this is expressed by a simple moving average of a plurality of voltage vectors arranged in time series, a period (time) during which the moving average process is performed. Is preferably variably set according to the fundamental frequency.

  Also according to the present embodiment described above, it is possible to obtain effects according to the above-described effects of the first embodiment.

(Other embodiments)
Each of the above embodiments may be modified as follows.
<Zero voltage vector priority means>
The conditions in step S38 in FIG. 9 are as follows: the condition that the vector representing the actual operation state when the norm of the error vector edq is larger than the threshold value eth is an effective voltage vector, and the effective voltage vector. Only a condition that an angle formed with the average voltage vector Va is equal to or less than the specified angle A may be used.

  Further, the zero vector priority means is not limited to that constructed in the manner of each of the above embodiments. For example, a means for constructing the evaluation function J so that the evaluation of the operation state expressed by the zero voltage vector is likely to be high may be used. This is, for example, an evaluation function relating to an operation state expressed by a zero voltage vector when the above condition is satisfied, with the final evaluation function J obtained by multiplying the evaluation function J exemplified in the above embodiment by a weighting coefficient. This can be done by making the weighting coefficient multiplied by J the smallest.

  The condition that gives priority to the zero voltage vector is not limited to the condition that “the norm of the error vector edq is larger than the threshold value eth”. For example, the condition may be that the end point of the error vector edq (the coordinate component of the vector) reaches a straight line that passes through the start point of the error vector edq and extends in the direction of the average voltage vector Va. This condition is particularly effective in the overmodulation region where the modulation rate of the inverter IV exceeds 1. This is because the end point of the error vector edq tends to be displaced toward the vicinity of the origin after P2 shown in FIG. 5, and the direction of the average voltage vector Va when the norm of the error vector edq is larger than the threshold value eth. This is because the distance from the straight line extending to

Further, for example, the condition may be that the end point of the error vector edq is entered in a region surrounded by a pair of effective voltage vectors whose directions are opposite to those of the pair of effective voltage vectors close to the average voltage vector Va.
<About state maintenance priority means>
The state maintenance priority means is not limited to maintaining the current operation state when the norm of the error vector edq is equal to or less than the threshold eth and the state transition permission flag F is off. For example, it may be a means for constructing the evaluation function J when a predetermined condition is satisfied so that the evaluation of the current operation state is likely to be high. This is performed, for example, by making the evaluation function J exemplified in the above embodiment multiplied by the weighting coefficient the final evaluation function J, and making the weighting coefficient multiplied by the evaluation function J related to the current operation state the smallest. be able to.

However, a configuration without the state maintenance priority unit itself may be used.
<About effective voltage vector priority means>
The effective voltage vector priority means is not limited to those exemplified in the above embodiments. For example, on the condition that the norm of the error vector edq is larger than the threshold eth and the voltage vector representing the operation state at that time is a zero voltage vector, the effective voltage vector having a small angle with the average voltage vector Va A means for constructing the evaluation function J so that the evaluation regarding the operation state expressed by can be easily increased. For example, the evaluation function J related to the operation state expressed by the effective voltage vector when the above condition is satisfied is obtained by multiplying the evaluation function J exemplified in the above embodiment by a weighting coefficient as the final evaluation function J. This can be done by making the weighting coefficient multiplied by the smallest.
<About prohibited means>
The prohibition means for prohibiting the number of terminals of the rotating machine connected to the switching element whose switching state is switched from being larger than 1 is not limited to that configured by the logic illustrated in FIG. For example, switching to an operation state in which all of the operation states in which the number of terminals of the rotating machine connected to the switching element whose switching state is switched is 1 or less is set as a control amount prediction target in each control cycle and is not a prediction target It may be a logic that prohibits. Even in this case, it is possible to perform the same control as the processing of FIG. 8 by providing the zero voltage vector priority means, the effective voltage vector priority means, and the like.

Moreover, it is not necessary to provide the prohibition means for prohibiting the number of terminals of the rotating machine connected to the switching element whose switching state is switched from being larger than one. At this time, for example, it may be prohibited that the number of terminals of the rotating machine connected to the switching element whose switching state is switched is larger than two. However, there is no need to provide such a prohibition at all.
<About average voltage direction calculation means>
1. Means for Inputting Parameters Related to Current Mean voltage direction calculation means is not limited to a means for inputting a command current to a voltage equation from which a differential operation term of current is removed. For example, the detected value of the current flowing through the motor generator 10 may be input. However, at this time, it is desirable to filter the detected current value.

  The average voltage vector on the dq axis calculated from the voltage equation on the dq axis in which the term including the differential operation value of the current is deleted is not limited to αβ conversion. For example, a voltage equation for calculating an average voltage vector on the dq axis may be αβ converted. In this case, the input is a current on the αβ axis, and the output is an average voltage vector having components in the αβ coordinate system. Incidentally, it is desirable that the input current at this time is obtained by subjecting the command currents idr and iqr to αβ conversion. However, the αβ component of the actual current may be used.

  Similarly, it may be configured to include means for taking a three-phase current as an input and using an output as a component of an average voltage vector in a three-dimensional fixed coordinate system. Incidentally, it is desirable that the input current at this time is obtained by three-phase conversion of the command currents idr and iqr. However, it may be a three-phase component of an actual current.

  Further, the average voltage direction calculation means is not limited to the one in which the current differential term of the voltage equation is deleted. For example, the following expression (c5) including a differential operation term of the command current Iαβr = (iαr, iβr) on the αβ axis may be used.

  Incidentally, the above equation (c5) is obtained by removing the current vector Iαβ from the following equation (c6) showing the relationship between the current vector Iαβ and the current voltage vector Vi, and setting the voltage vector Vi as a zero voltage vector. Is.

Further, as an average voltage direction calculation means that receives current-related parameters as input, the change direction during the period in which the operation state of the inverter IV is the zero voltage vector for the error vector edq obtained by subtracting the detected current vector Idq from the command current vector Idqr It may be calculated.
2. Means for Inputting Voltage-Related Parameters The average voltage direction calculating means for inputting the voltage-related parameters is not limited to that calculated by the simple moving average processing of the voltage vector representing the actual operation state of the inverter IV. For example, for a thing more than a predetermined number of times, it may be calculated by an exponential moving average with a small weighting factor. However, the averaging process itself is not essential for specifying the direction of the average voltage vector.
3. Definition of Average Voltage Vector Va The average voltage vector Va is not limited to a fundamental wave component having an electrical angular frequency in the output voltage of the inverter IV. For example, regardless of whether or not it is a fundamental wave component, in general, it is shorter than one electrical angle cycle for the output voltage of the inverter IV required for realizing a command value of a control amount such as current, and the control cycle Tc. It is good also as an average value in a longer time scale. Specifically, for example, in an overmodulation region, as an average value of a voltage for realizing a superimposition of a sixth harmonic current on a command value of the current assuming the fundamental wave (the command current idr, iqr) Also good. Note that the sixth harmonics are superimposed on the command value in the overmodulation region, for example, “PMSM high response torque control system based on model predictive control in consideration of inverter overmodulation region: Hozumi, Ishida, Michiki , Okuma, 2009 IEEJ Industrial Application Division Conference.
<About controlled variable>
The control amount used for determining the operation of the inverter IV based on the command value and the predicted value is not limited to the current. For example, torque and magnetic flux may be used. In this case, the magnetic flux command value may be set so that, for example, maximum torque control can be realized. Even when such processing is performed, it is effective to examine the change of the operation state based on the average voltage vector Va when the torque or the magnetic flux is out of the allowable range.

  Also, for example, current and torque may be used. Here, if both the d-axis current and the q-axis current are determined, the torque is uniquely determined. However, it is also possible to set both of these as prediction targets. Thereby, for example, the predicted value of the current is used only for the process of setting the value of the state transition permission flag F, and for the other processes, instead of the magnitude comparison between the norm of the error vector edq and the threshold value eth, It is possible to compare the difference between the predicted value of torque and the command value and the threshold value for torque deviation.

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 rotation speed may be used.
<About prediction means>
In each of the above embodiments, the current detection timing is synchronized with the update timing of the operation state of the inverter IV, but the present invention is not limited to this. For example, the current may be detected at a central timing between each pair of update timings adjacent in time series. Even in this case, it is effective to predict the current at the next update timing based on the detected current as the initial value of the current prediction associated with the setting of the operation state at the next update timing.

  In each of the above embodiments, the detection timing of the electrical angle θ is synchronized with the update timing of the operation state of the inverter IV, but is not limited thereto. For example, the electrical angle θ may be detected at a central timing between each pair of update timings adjacent in time series.

  In each of the above embodiments, the control amount one control cycle ahead is predicted from the update timing of the operation state of the inverter IV, but the present invention is not limited to this. For example, the control amount at an intermediate time point in the period from when the operation state is updated until one control cycle elapses may be predicted.

  -The method of discretizing a model in a continuous system is not limited to using a difference method such as a forward difference method. For example, a linear multi-stage method with N (≧ 2) stages, a Runge-Kutta formula, or the like may be used.

  The model used for predicting the current is not limited to a model that ignores iron loss, and may be a model that takes this into consideration.

  -The model used for predicting current etc. is not limited to the model based on the fundamental wave. For example, a model including higher-order components for inductance and induced voltage may be used. In addition, as a means for predicting current and the like, not only a model formula but also a map may be used. At this time, the input parameters of the map may be voltage (vd, vq) and electrical angular velocity ω, and may further include temperature and the like. Here, the map is storage means in which output parameter values corresponding to discrete values for input parameters are stored.

In each of the above embodiments, the control amount due to the operation of the inverter IV at the next update timing (the timing one control cycle ahead) of the operation state of the inverter IV is predicted, but the present invention is not limited to this. For example, the operation state at the update timing one control cycle ahead may be determined by sequentially predicting the control amount by the operation of the inverter IV at the update timing several control cycles ahead. This process can be used in step S44 of FIG. Incidentally, in this case, the evaluation function J may be, for example, the sum of the evaluation function J that receives the error vector edq using the operation state at each update timing. At this time, the evaluation of the evaluation function J when the norms of the error vectors edq are the same may be made the same for all the update timings. For example, the contribution rate to the final evaluation may be reduced in the far future.
<Others>
The predetermined condition for turning on the state transition permission flag F is not limited to those exemplified in the above embodiments. For example, the end point of the error vector edq is displaced from the region surrounded by a pair of effective voltage vectors having a small angle with the average voltage vector Va among the effective voltage vectors starting from the point where the error vector edq becomes zero. This may be a condition. Further, for example, after the end point of the error vector edq is displaced from the region surrounded by the pair of effective voltage vectors, the error vector edq may be further displaced from the region.

  Even if the process using the state transition permission flag F is not performed, it is possible to obtain an effect according to the effect (4) of the first embodiment.

  The rotating machine is not limited to an embedded magnet synchronous machine, and may be an arbitrary synchronous machine such as a surface magnet synchronous machine or a field winding type synchronous machine. Furthermore, it is not limited to a synchronous machine, but may be an induction rotating machine such as an induction motor.

  -As a rotary machine, not only what is mounted in a hybrid vehicle but the thing mounted in an electric vehicle may be sufficient. Further, the rotating machine is not limited to the one used as the main machine of the vehicle.

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

  DESCRIPTION OF SYMBOLS 10 ... Motor generator, 20 ... Control apparatus, IV ... Inverter, Swp, Swn ... Switching element.

Claims (8)

In order to control the control amount of the rotating machine by operating the output voltage of the power converting circuit provided with a switching element that electrically connects each of the positive electrode and the negative electrode of the DC power source to the terminal of the rotating machine, Predicting means for predicting a control amount of the rotating machine when an operation state is set, and an actual operation state of the power conversion circuit is determined based on the predicted control amount, and the determined operation state is obtained. In a control device for a rotating machine comprising operating means for operating the power conversion circuit as described above,
Average voltage direction calculating means for calculating an average output voltage vector direction of the power conversion circuit,
On the condition that the operation means includes an effective voltage vector representing the operation state, an angle formed with the direction calculated by the average voltage direction calculation means is less than a specified angle of 20 degrees or less. An apparatus for controlling a rotating machine, comprising: zero voltage vector priority means for setting an operation state expressed by a zero voltage vector to a highest priority for determining the actual operation state. Tolerance range determination means for determining whether or not a difference between the predicted value of the control amount and the command value of the control amount when the current operation state is set as the operation state of the power conversion circuit is out of the allowable range. Prepared,
The operation means further includes a state maintenance priority means for determining that the priority of the current operation state of the power conversion circuit is high when it is determined to be within the allowable range by the allowable range determination means,
The zero voltage vector priority means is an operation state represented by the zero voltage vector when a logical product of that it is determined that the zero voltage vector is out of the allowable range and that there is an angle that is equal to or less than the specified angle is true. The control device for a rotating machine according to claim 1, wherein the priority of the rotating machine is the highest.   The zero voltage vector priority unit sets the operation state represented by the zero voltage vector as a prediction target of the prediction unit when the priority of the operation state represented by the zero voltage vector is the highest. The operation state represented by the zero voltage vector is determined as an actual operation state when the difference between the predicted value and the command value is within the allowable range. Control device for rotating machine.   When the voltage vector expressing the operation state that is the prediction target when it is determined to be out of the allowable range is a zero voltage vector, the operation means is a pair of small angles formed with the average output voltage vector. 4. The control apparatus for a rotating machine according to claim 2, further comprising effective voltage vector priority means for setting the priority of at least one of the operation states expressed by the effective voltage vector of the highest to the highest.   When determining the actual operation state of the power conversion circuit, the power conversion circuit further comprises prohibiting means for prohibiting the number of terminals of the rotating machine connected to the switching element whose switching state is switched from being larger than one. The control device for a rotating machine according to any one of claims 1 to 4.   6. The control apparatus for a rotating machine according to claim 5, wherein the predicting unit excludes an operation state prohibited by the prohibiting unit from a prediction target by the predicting unit. The predicting means excludes the low priority operation state from the prediction target by the predicting means,
The control device for a rotating machine according to claim 1, wherein the operation means determines an actual operation state from among the prediction targets. The prediction means includes
First prediction means for predicting the control amount when setting the operation state of the power conversion circuit;
8. A second prediction unit that predicts an initial value used for prediction by the first prediction unit using an operation state that has already been determined to be used by the operation unit as input. 8. The control apparatus of the rotary machine of Claim 1.