JP2012095485A - Rotary electric machine control system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a rotary electric machine control system which can generate a PWM signal at a sine wave control mode, an overmodulation control mode and a rectangular wave control mode by a common method.SOLUTION: A PWM signal generation device 56 of the rotary electric machine control system 10 includes: a voltage command vector generation portion 66; a region determination portion 68 determining to which region in a plurality of phase regions of a spatial vector system a prediction phase as timing for generating the PWM signal by using a voltage command vector belongs; an angle ratio calculation portion 70 calculating a contribution ratio of the voltage command vector to a switching vector specifying a prediction phase region to which the prediction phase belongs as an angle ratio to a control angle period Δθ; and a switching phase calculation portion 72 which sequentially calculates a switching phase between the switching vectors in the prediction phase region at the calculated angle ratio to generate the PMW signal in accordance with a switching changeover order.

Description

  The present invention relates to a rotating electrical machine control system, and more particularly, to a rotating electrical machine control system that performs pulse width modulation using a space vector method.

  In the three-phase rotating electrical machine, the operation control is performed by controlling the on / off timing of the switching elements constituting the respective phase arms of the three-phase inverter connected thereto. As the on / off control signal of the switching element constituting each phase arm, a PWM (Pulse Wide Modulation) signal that is pulse-width modulated as compared with a triangular wave is used with the control voltage waveform of each phase of three phases as a fundamental wave. be able to. As a method for obtaining on / off timing in a control period corresponding to the period of a triangular wave without using a triangular wave, a PWM signal generation method using a space vector method is known.

  For example, Patent Document 1 discloses that for a PWM control voltage source inverter, an u-phase, v-phase, and w-phase sine wave signal is compared with a triangular wave having a carrier period T to replace each conventional transistor on and off, and instantaneously. It describes pulse width modulation using the space vector method. Here, the voltage vector representing the voltage pattern output from the inverter unit is changed from V0 to V7 in the on / off state of the transistors Q1, Q2 of the transistor modules Tu, Tv, Tw corresponding to the u phase, v phase, and w phase. Using the pulse width calculation formula for each carrier period T, the pulse width of the voltage vector representing the voltage pattern in the interval of π / 3 = 60 degrees of the phase angle Φ is obtained, and each transistor module Tu, Tv, Tw is obtained. It is introduced to turn on and off. Of these V0 to V7, V0 and V7 are called zero vectors, and V1 to V6 are called basic voltage vectors.

In Patent Document 2, as a control method of a three-phase inverter that drives a three-phase motor, a UVW coordinate system that is a fixed coordinate system every 120 degrees, an αβ coordinate system that is an orthogonal fixed coordinate system, and an orthogonal rotating coordinate system are used. After describing coordinate conversion with a certain dq coordinate system, the voltage vector locus of the motor voltage Va on the αβ coordinate system becomes a regular hexagon centered on the origin on the αβ coordinate system when overmodulation control is executed. It shows that the maximum value of the magnitude of the voltage vector that can be output as the motor voltage Va after using overmodulation is (2/3) ^1/2 × Vdc. Here, six axes rotated by n × 60 degrees from the α axis in the αβ coordinate system are defined as the a ₀ axis to the a ₅ axis. The a ₀ to a ₅ axes are the axial directions of the basic voltage vector of the space vector method. Here, it is described that when the end point of the voltage command vector exceeds the regular hexagonal locus, voltage command vector correction for correcting the a-axis component of the voltage command vector is performed.

  Patent Document 3 discloses that a three-phase PWM signal generator uses a space vector method and has two basic voltage vectors having a phase difference of 60 degrees sandwiching a voltage command vector, and the two basic voltage vectors. When a PWM signal is produced using a third basic voltage vector that has a phase difference of 60 degrees and a basic voltage vector close to the voltage command vector, the central basic voltage It is stated that the value given as the generation time ratio to the basic voltage vector far from the voltage command vector among the two types of basic voltage vectors excluding the vector is determined based on the minimum generation time ratio at which current can be detected. Yes.

  In Patent Document 4, as a drive control device for an AC motor, a d-axis voltage command and a q-axis voltage command are converted into phase voltages so that the fundamental amplitude of a pulse width modulation voltage applied to the AC motor is in accordance with the voltage command amplitude. It is described that linear correction is performed in accordance with the synchronization number K, which is the number of triangular wave carriers per one command cycle.

Japanese Patent No. 3821145 JP 2008-113494 A JP 2008-113494 A JP 2008-31420 A

  As the operation control mode of the rotating electrical machine, there are a sine wave control mode, an overmodulation control mode, and a rectangular wave control mode, depending on the voltage value of the DC voltage supplied to the three-phase inverter circuit and the magnitude of the voltage amplitude of the fundamental wave. Thus, the operation control mode is appropriately selected. In the sine wave control mode used when the voltage amplitude of the fundamental wave is small, when the carrier frequency of the pulse width modulation can be sufficiently large, the PWM signal uses a triangular wave comparison PWM, and the pulse over one period of the fundamental wave This becomes a width-modulated pseudo sine wave. Therefore, the counter or timer that manages the pulse width does not require a very large number of bits.

Incidentally, synchronous PWM is known as a method for reducing the number of times of switching without generating a beat in the phase current of the rotating electrical machine in order to suppress the heat generation of the switching elements constituting the inverter. In synchronous PWM, when the rotating electrical machine is in an extremely low rotation range, the carrier frequency of the inverter is lowered in synchronization with the rotation speed, and a counter or timer that manages the pulse width may require a large number of bits. In addition, in the rectangular wave control mode in which the fundamental wave voltage is (2/3) ^1/2 V _dc , beats are generated with a small number of switching times in one electrical cycle. A large number of bits is required for the counter or timer for managing the.

  The synchronization PWM synchronization number used in the sine wave control mode and overmodulation control mode can be set so that the number of bits of the counter or timer for managing the pulse width does not become too large. However, the counter value or timer value in the rectangular wave control mode cannot be reduced only by devising the setting of the number of synchronizations of the synchronous PWM. Therefore, for example, a method of switching the control mode when the angles coincide is adopted.

  As described above, since the use of the PWM signal is different in the sine wave control mode, the overmodulation control mode, and the rectangular wave control mode, a control circuit suitable for each mode is used, and these controls are performed when the control mode is switched. Switching between circuits takes place. Although it is convenient that these three control modes can be used consistently with a common control method, as described above, since the PWM signal generation method is different in the prior art, there is no common control method as it is. It has become difficult.

  An object of the present invention is to provide a rotating electrical machine that enables PWM signal generation in a sine wave control mode, an overmodulation control mode, and a rectangular wave control mode in a common method without requiring a counter or timer having a particularly large number of bits. To provide a control system.

  The rotating electrical machine control system according to the present invention has two switching patterns for turning on and off the switching elements constituting the three-phase inverter circuit connected to the three-phase rotating electrical machine and each phase arm of the three-phase inverter circuit. Corresponding to six types of zero vectors and six types of basic voltage vectors, and sequentially arranging these six types of basic voltage vectors on the phase plane with a phase difference of 60 degrees from each other, the six phase regions are divided into phase regions And a control device that generates a PWM signal for each phase arm using a space vector method from I to phase region VI. The control device includes two basic voltage vectors that define one phase region and two Zero vector as switching vector, switching vector switching order for generating PWM signal as switching switching order A storage unit that predetermines and stores each of the phase region I to the phase region VI and a control angle cycle Δθ as a cycle on the phase plane for generating the PWM signal in a three-phase inverter circuit in the case of full cycle control. The control angle cycle setting is set to 2π / K, which is an angle obtained by dividing 2π of one phase of the voltage command vector by an integer number of synchronization K, and in the case of half cycle control, π / K which is 1/2 of that. And an interrupt phase synchronizer that synchronizes the interrupt phase with any one of six basic voltage vectors on the phase plane, with the phase at the starting point of the control angle period as the interrupt phase, The phase current of the three-phase rotating electrical machine and the electrical angle of the rotor are acquired at the timing, the d-axis current value and the q-axis current value are calculated based on the acquired phase current, the d-axis command current value, the d-axis current value, D based on the deviation between A voltage command value is obtained, a q-axis voltage command value is obtained based on a deviation between the q-axis command current value and the q-axis current value, and based on the d-axis voltage command value, the q-axis voltage command value, and the electrical angle of the rotor. A voltage command vector generation unit that generates a voltage command vector and a timing for generating a PWM signal using the generated voltage command vector, predicting that a voltage vector command exists in a predicted phase that is advanced by nΔθ from the interrupt phase, A region determination unit for determining which region of the phase region I to VI the predicted phase belongs to, and a phase region to which the predicted phase belongs as a predicted phase region, and a contribution ratio of the voltage command vector to the switching vector defining the predicted phase region An angle ratio calculation unit that calculates an angle ratio with respect to the control angle period Δθ, and a switching switching order in the predicted phase region are read out, and the switching order is followed A switching phase calculation unit that sequentially calculates a switching phase between the respective switching vectors in the predicted phase region at the angle ratio calculated by the angle ratio calculating means, and generates a PMW signal in the phase region. To do.

Further, in the rotating electrical machine control system according to the present invention, the angle ratio calculation unit is on a phase plane defined by the phase region I to the phase region VI when the DC voltage applied to the three-phase inverter circuit is V _dc . Since the radius of the inscribed circle of the hexagon is V _dc / (2) ^{1/2 and} the radius of the envelope circle connecting the vertices of the hexagon is (2/3) ^1/2 V _dc , the phase region I If the hexagonal origin on the phase plane defined by the phase region VI is the origin of the voltage command vector, and the magnitude of the voltage command vector is inside the hexagon, the voltage command vector Corresponds to the overmodulation control mode, and the basic voltage vector component of the voltage command vector extends outward beyond the envelope circle. When the square wave control mode When the control mode is an overmodulation control mode, the angle ratio of two zero vectors among the switching vectors in the predicted phase region is set to zero, and Among the calculated values of the angle ratio of the voltage command vector, the calculated angle ratio of one of the basic voltage vectors is directly used as the angle ratio of one of the basic voltage vectors, and the angle ratio of the other basic voltage vector is from 1 to one. Overmodulation angle ratio calculation means that subtracts the angle ratio of the basic voltage vector, and when the control mode is the rectangular wave control mode, the angle ratio of two zero vectors of the switching vectors in the predicted phase region are both zero. The angle ratio of the voltage command vector to the two basic voltage vectors is ½ A square wave angular ratio calculating means preferably comprises a.

  In the rotating electrical machine control system according to the present invention, the region determination unit sets the expected phase as n = 1.5 in the case of half-cycle control, and n = 1.25 and n = 1 in the case of full-cycle control. It is preferable to set .75 and two expected phases.

  Further, in the rotating electrical machine control system according to the present invention, the angle ratio calculation unit may determine which of the two basic voltage vectors when the predicted phase is equal in phase from the two basic voltage vectors constituting the predicted phase region. It is preferable that the calculated angle ratio of one of the basic voltage vectors is directly used as the angle ratio of the one basic voltage vector, and the angle ratio of the other basic voltage vector is equal to the angle ratio of the one basic voltage vector.

  Further, in the rotating electrical machine control system according to the present invention, the angle ratio calculation unit is configured such that the predicted phase is within a predetermined phase range from one of the two basic voltage vectors constituting the predicted phase region. In addition, the calculated angle ratio of one basic voltage vector is directly used as the angle ratio of the one basic voltage vector, the angle ratio of the other basic voltage vector is set to zero, and is within a predetermined phase range from the other basic voltage vector. In some cases, it is preferable that the calculated angular ratio of the other basic voltage vector is directly used as the angular ratio of the other basic voltage vector, and the angular ratio of one basic voltage vector is zero.

  The rotating electrical machine control system according to the present invention preferably includes a synchronization number switching unit that changes the synchronization number when the interrupt phase matches the phase of the basic voltage vector determined in advance according to the synchronization number K.

  Further, in the rotating electrical machine control system according to the present invention, the storage unit sets the d-axis voltage command value and q so that the fundamental wave amplitude corresponding to the generated PWM signal matches the voltage command amplitude corresponding to the voltage command vector. The linear correction value for correcting the shaft voltage command value is stored for each synchronization number K, and the voltage command vector generation unit reads the linear correction value corresponding to the synchronization number K, and d-axis voltage command value and q-axis voltage command value It is preferable to generate a voltage command vector by correcting.

With the above configuration, the rotating electrical machine control system includes a control device that generates a PWM signal for each phase arm using the space vector method. A PWM signal based on the space vector method is generated as follows. That is, the phase at the start point of the control angle period is set as an interrupt phase, and the interrupt phase is synchronized with any one of six basic voltage vectors on the phase plane so that the phase current of the three-phase rotating electrical machine and the electrical Get the horns and Using these, a voltage command vector is generated, and as a timing for generating a PWM signal using the generated voltage command vector, it is predicted that the voltage vector command exists in a predicted phase advanced by nΔθ from the interrupt phase, and the predicted phase To which phase region I to VI belongs. Then, the phase region to which the predicted phase belongs is defined as the predicted phase region, the contribution ratio of the voltage command vector to the switching vector defining the predicted phase region is calculated as the angle ratio with respect to the control angle period Δθ, and the predicted phase region is calculated with the calculated angle ratio. The PMW signal is generated by sequentially calculating the switching phase between the switching vectors at.
In this way, the PWM signal is generated using the angle ratio with respect to the control angle period Δθ as the contribution ratio of the voltage command vector to the switching vector that defines the predicted phase region, so that a counter or timer having a particularly large number of bits is not required. In addition, the angle ratio data may be simply calculated. Since the same applies to the sine wave control mode, the overmodulation control mode, and the rectangular wave control mode, PWM signal generation in each control mode can be performed by a common method.

  In addition, since the interrupt phase matches the phase of the basic voltage vector, it is possible to ensure the synchronism between the electrical cycle of the rotor and the control cycle. For example, even if one pulse drive is performed in one cycle as in the rectangular wave control, Fluctuation of the voltage actually applied to the rotating electrical machine is suppressed, and occurrence of a beat, which is a phenomenon that the motor current vibrates, can be suppressed.

  In the rotating electrical machine control system, the control mode is determined by comparing the magnitude of the voltage command vector with an envelope circle connecting the hexagonal inscribed circle on the phase plane and each vertex of the hexagon. When the control modes are the overmodulation control mode and the rectangular wave control mode, the angle ratios of the two zero vectors are both zero. In addition, in the overmodulation control mode, priority is given to one of the calculated values of the angle ratio of the voltage command vector to the two basic voltage vectors, and the other is a value obtained by subtracting one angle ratio from 1. To do. In the case of the rectangular wave control method, the angle ratio of the voltage command vector to the two basic voltage vectors is both ½. In this manner, PWM signal generation in each control mode of the sine wave control mode, the overmodulation control mode, and the rectangular wave control mode can be performed by a common method.

  In the rotating electrical machine control system, the expected phase is set as n = 1.5 in the case of half-cycle control, and two expected phases are set as n = 1.25 and n = 1.75 in the case of full-cycle control. To do. In this method, in the conventional PWM signal generation in a triangular wave, from the timing of data acquisition such as phase current to the calculation of the fundamental wave signal indicating the voltage command value, in the case of half-cycle control, it is performed in the period from the valley of the triangular wave to the peak. In the case of full-cycle control, it corresponds to performing the period from the trough of the triangular wave to the trough and then comparing the triangular wave signal and the fundamental wave signal to obtain the switching timing. Therefore, the procedure for obtaining the angle ratio by the space vector method can correspond to the signal processing procedure at the time of conventional PWM signal generation.

  In the rotating electrical machine control system, when the predicted phase is equal in phase from the two basic voltage vectors constituting the predicted phase region, the calculation angle ratio of one of the two basic voltage vectors Is the angle ratio of one basic voltage vector, and the angle ratio of the other basic voltage vector is equal to the angle ratio of the one basic voltage vector. As described above, since the expected phase is calculated based on the detected value of the electrical angle of the rotor, the sensor that detects the electrical angle of the rotor detects it with an appropriate number of bits for 60 degrees that is the phase region angle. If this is not possible, an error may occur. Since this is a problem with the detection accuracy of the sensor, the setting of the expected phase is equiphase from the two basic voltage vectors, and the calculation is proceeded assuming that the detection value of the sensor is equiphase.

  In the rotating electrical machine control system, when the predicted phase is within a predetermined phase range from one of the two basic voltage vectors constituting the predicted phase region, the calculation angle of one of the basic voltage vectors When the ratio is the angle ratio of one basic voltage vector as is, the angle ratio of the other basic voltage vector is zero, and the other basic voltage vector is within a predetermined phase range, the other basic voltage vector is calculated. The angle ratio is directly used as the angle ratio of the other basic voltage vector, and the angle ratio of one basic voltage vector is set to zero. As described above, since the predicted phase is calculated based on the detected value of the electrical angle of the rotor, the sensor that detects the electrical angle of the rotor detects it with an appropriate number of bits for 60 degrees that is the angle of the phase region. If this is not possible, an error may occur. Here, for the phase range near the basic voltage vector, the calculation is advanced with the angle ratio set to zero.

  In the rotating electrical machine control system, the synchronization number is changed when the interrupt phase matches the phase of the basic voltage vector determined in advance according to the synchronization number K. Depending on the number of synchronizations, even if the first interrupt phase is set to be synchronized with the basic voltage vector, the subsequent interrupt phase may not necessarily match the phase of the basic voltage vector. Even in such a case, the change in the synchronization number synchronizes with the phase of the predetermined basic voltage vector, so that the synchronization of control can be ensured even if the synchronization number is changed.

  In the rotating electrical machine control system, the d-axis voltage command value and the q-axis voltage command value are corrected so that the fundamental wave amplitude corresponding to the generated PWM signal matches the voltage command amplitude corresponding to the voltage command vector. The linear correction value that is the correction value is stored for each synchronization number K and used. For example, in the overmodulation control mode, the increase in the fundamental wave amplitude is less than the increase in the voltage command amplitude. By performing linear correction, the fundamental wave amplitude can be set to the voltage command amplitude.

It is a figure which shows the structure of the rotary electric machine control system of embodiment which concerns on this invention. It is a figure which shows the relationship between six basic voltage vectors in a space vector system, six phase areas, and d-axis q-axis. It is a figure which shows the relationship between the circle _{| round |} yen prescribed | regulated by six phase area | regions of a space vector system, and DC voltage _Vdc applied to a three-phase inverter circuit. It is a figure explaining the interruption phase at the time of half cycle control with the number of synchronizations K = 6. In the rotary electric machine control system of embodiment which concerns on this invention, it is a figure explaining the interruption phase at the time of half cycle control by the synchronous number K = 6 in relation to the hexagon of a space vector system. In the rotary electric machine control system of embodiment which concerns on this invention, it is a figure explaining the prediction phase at the time of half cycle control by the synchronous number K = 6. 6 is a flowchart for explaining a procedure for generating a PWM signal when the number of synchronizations is K = 6 and half-cycle control is performed in the rotating electrical machine control system according to the embodiment of the present invention. In the rotating electrical machine control system according to the embodiment of the present invention, it is a diagram for generally explaining the relationship between two basic voltage vectors and an α axis and a β axis. It is a figure explaining a mode when the alpha axis is made to correspond with one of two basic voltage vectors in FIG. In the rotary electric machine control system of embodiment which concerns on this invention, it is a figure explaining the angle ratio of the voltage command vector with respect to two basic voltage vectors. In the rotary electric machine control system of embodiment which concerns on this invention, it is a figure explaining the angle ratio of the voltage command vector with respect to two basic voltage vectors at the time of half number control with the synchronous number K = 6. FIG. 11 is a diagram showing a switching order of switching vectors when the number of synchronizations is K = 6 and half cycle control is performed in the space vector method. It is a figure which shows the content of FIG. 12 with a flowchart system. It is a figure explaining the phase area | region I at the time of half cycle control by the number K of synchronizations = 6 about the content of FIG. In the rotating electrical machine control system according to the embodiment of the present invention, the angle ratio of the voltage command vector to the two basic voltage vectors is set for each region obtained by dividing the phase region I when the synchronization number K = 6 and half-cycle control is performed. It is a figure explaining each. FIG. 16 is a diagram illustrating a boundary state between a sine wave control mode and an overmodulation control mode in which the magnitude of the voltage command vector is increased and the zero vector is eliminated from the state of FIG. 15. In the rotary electric machine control system of embodiment which concerns on this invention, it is a figure which shows the result of the PWM signal generation | occurrence | production of each phase arm at the time of the sine wave control mode under half cycle control with the synchronous number K = 6. FIG. 17 is a diagram for explaining a state where the magnitude of the voltage command vector is increased from the state of FIG. 16 and the overmodulation control mode is entered. In the rotary electric machine control system of embodiment which concerns on this invention, it is a figure which shows the result of the PWM signal generation | occurrence | production of each phase arm about the time of the overmodulation control mode under half cycle control with the synchronous number K = 6. FIG. 19 is a diagram for explaining a state in which the magnitude of a voltage command vector is increased from the state of FIG. 18 and a rectangular wave control mode is entered. In the rotary electric machine control system of embodiment which concerns on this invention, it is a figure which shows the result of the PWM signal generation | occurrence | production of each phase arm about the time of the rectangular wave control mode under half cycle control with the synchronous number K = 6. It is a figure explaining the interruption phase at the time of half cycle control with the number of synchronization K = 9. In the rotary electric machine control system of embodiment which concerns on this invention, it is a figure explaining the interruption phase at the time of half cycle control by the synchronous number K = 9 in relation to the hexagon of a space vector system. In the rotary electric machine control system of embodiment which concerns on this invention, it is a figure which shows the result of the PWM signal generation | occurrence | production of each phase arm about the time of the sine wave control mode under half cycle control with the synchronous number K = 9. In the rotary electric machine control system of embodiment which concerns on this invention, it is a figure which shows the result of the PWM signal generation | occurrence | production of each phase arm about the time of the overmodulation control mode under half cycle control with the synchronous number K = 9. In the rotary electric machine control system of embodiment which concerns on this invention, it is a figure which shows the result of the PWM signal generation | occurrence | production of each phase arm at the time of the rectangular wave control mode under half cycle control with the synchronous number K = 9. In the rotary electric machine control system of embodiment which concerns on this invention, it is a figure which shows the example of the synchronous number K = 3, 6, 9, 12 as an example of a linear correction map. 27 is a diagram illustrating an example of the synchronization number K = 15, 18, 21, and 27 as an example of the linear correction map. 28 is a diagram illustrating an example of the synchronization number K = 33, 39, 45, 51 as an example of the linear correction map. 28 is a diagram illustrating an example of the synchronization number K = 63, 75, 87 as an example of the linear correction map. It is a figure explaining the interruption phase at the time of full cycle control with the number of synchronous K = 6. In the rotating electrical machine control system according to the embodiment of the present invention, it is a diagram for explaining an interrupt phase in the case of full cycle control with a synchronization number K = 6 in relation to a hexagon of a space vector method. In the rotary electric machine control system of embodiment which concerns on this invention, it is a figure explaining the prediction phase at the time of full cycle control by the synchronous number K = 6. 7 is a flowchart for explaining a procedure for generating a PWM signal when the number of synchronizations is K = 6 and full-cycle control is performed in the rotating electrical machine control system according to the embodiment of the present invention. In the rotary electric machine control system of embodiment which concerns on this invention, it is a figure explaining the angle ratio of the voltage command vector with respect to two basic voltage vectors at the time of full number control with the synchronous number K = 6. 7 is a map showing the switching order of switching vectors when the number of synchronizations is K = 6 and full-cycle control is performed in the space vector method. It is a figure which shows the map of FIG. 36 with a flowchart system. FIG. 37 is a diagram illustrating the contents of the map of FIG. 36 for the phase region I when the number of synchronizations K = 6 and full cycle control is performed. It is a figure explaining the interruption phase at the time of the total number control with the number of synchronization K = 9. In the rotating electrical machine control system according to the embodiment of the present invention, the interrupt phase at the time of full cycle control with the synchronization number K = 9 will be described in relation to the hexagon of the space vector method. In the rotary electric machine control system of embodiment which concerns on this invention, it is a figure explaining the example which divides one phase area into two of the area | region A of equiphase vicinity, and the other area | region B from two basic voltage vectors. is there. In the rotating electrical machine control system according to the embodiment of the present invention, one phase region is converted from two basic voltage vectors to a region B near the same phase, two regions A near the basic voltage vector, and the other regions C. It is a figure explaining the example divided into. In the rotary electric machine control system of embodiment which concerns on this invention, it is a figure explaining the phase from which the synchronous number switching is performed from the synchronous number K = 6 by half cycle control. FIG. 44 is a diagram illustrating a phase in which synchronization number switching is performed from the synchronization number K = 9 in half-cycle control, as in FIG. 43. FIG. 44 is a diagram illustrating a phase in which synchronization number switching is performed from the synchronization number K = 6 in full-cycle control, as in FIG. 43. FIG. 46 is a diagram illustrating the phase in which the synchronization number is switched from the synchronization number K = 9 in the full cycle control, as in FIG. 45. The rotary electric machine control system of embodiment which concerns on this invention WHEREIN: The figure explaining the interruption phase from which a prediction phase will be 90 degree | times among two types of interruption phases in the case of the number of synchronization K = 9 by half cycle control. It is. It is a figure explaining the phase from which synchronous number switching is performed. Similarly to FIG. 47, when the number of synchronizations is K = 15 in the half-cycle control, it is a diagram illustrating an interrupt phase at which the predicted phase is 90 degrees among the two types of interrupt phases.

Embodiments according to the present invention will be described below in detail with reference to the drawings. In the following, a motor / generator mounted on a vehicle will be described as a rotating electrical machine, but a rotating electrical machine used for purposes other than mounting on a vehicle may be used. In the following description, the inverter circuit connected to the rotating electrical machine will be described as having an arm for each phase and supplied with V _dc, which is the terminal voltage of the capacitor. This is a basic element. In addition to the capacitor, a power storage device, a voltage converter, or the like may be connected as an element that supplies DC power to the inverter circuit.

  Specific numerical values of the synchronization number K described below are examples, and of course, other synchronization numbers may be used.

  Below, the same code | symbol is attached | subjected to the same element in all the drawings, and the overlapping description is abbreviate | omitted. In the description in the text, the symbols described before are used as necessary.

  FIG. 1 is a diagram illustrating the configuration of the rotating electrical machine control system 10. The rotating electrical machine control system 10 is a system that controls the operation of the rotating electrical machine 12 and includes an inverter circuit 20 connected to the rotating electrical machine 12 and a control block 50. In FIG. 1, the control block 50 is a portion excluding the inverter circuit 20, the rotating electrical machine 12, the resolver 14 provided in the rotating electrical machine 12, and the current detection units 16 and 18.

  The rotating electrical machine 12 is a motor / generator (M / G) mounted on a vehicle and functions as a motor when electric power is supplied from the inverter circuit 20 side, and when driven by an engine (not shown), or This is a three-phase synchronous rotating electric machine that functions as a generator during braking of a vehicle.

  That is, the rotating electrical machine 12 functions as a motor by AC power supplied from the inverter circuit 20 side when used as a starter for starting an engine (not shown) for vehicle running or when powering. During braking, the regenerative energy is recovered by functioning as a generator, converted into DC power via the inverter circuit 20, and supplied to the power storage device. Control of the operation of the rotating electrical machine 12 is performed by operation control of the inverter circuit 20 by the control device.

Resolver 14 provided in the rotary electric machine 12 is a device for detecting the angular position of the rotor of the rotary electric machine 12, the detected data as electrical angle theta _e of the rotor is transmitted to the control device.

The current detectors 16 and 18 provided in the rotating electrical machine 12 are current sensors that detect each phase current of the rotating electrical machine 12. Specifically, it has a function of detecting a current flowing through a power line connecting each phase arm of the inverter circuit 20 and each phase coil of the rotating electrical machine 12. When the one-side terminals of the phase coils of the rotating electrical machine 12 are connected to each other at a neutral point, the sum of the phase currents including the sign is 0, so that two phase currents are detected. It is enough. In FIG. 1, two current detection units 16 and 18 are shown to detect two of a U-phase current value I _U and a V-phase current value I _V. The remaining W-phase current value I _W is obtained by I _W = − (I _U + I _V ). The detected phase currents are transmitted to the control device.

  The inverter circuit 20 is a circuit connected to the rotating electrical machine 12 and includes a plurality of switching elements that operate under the control of the control block 50, and has a function of performing power conversion between AC power and DC power. Have. That is, the inverter circuit 20 has an AC / DC conversion function that converts the AC three-phase regenerative power from the rotating electrical machine 12 into DC power and supplies it as a charging current to the power storage device when the rotating electrical machine 12 functions as a generator. . Further, when the rotating electrical machine 12 functions as a motor, the rotating electrical machine 12 has an orthogonal conversion function that converts DC power from the power storage device side into AC three-phase driving power and supplies the AC power to the rotating electrical machine 12 as AC driving power.

  When the rotating electrical machine 12 is a three-phase system, a series connection element is used in which two switching elements are connected in series to each phase, and a diode is connected in reverse to each switching element. When the serial connection elements for each phase are referred to as U-phase arm 22, V-phase arm 24, and W-phase arm 26, respectively, inverter circuit 20 includes U-phase arm 22, V-phase arm 24, and W-phase arm 26. It has the structure mutually connected in parallel. Each phase power line is drawn from the connection point of the two switching elements in the U-phase arm 22, the V-phase arm 24, and the W-phase arm 26, and is connected to the other terminal of each phase coil of the rotating electrical machine 12.

The power storage device provided in the inverter circuit 20 has a function of supplying direct-current power between the positive-side bus bar and the negative-side bus line in which the U-phase arm 22, the V-phase arm 24, and the W-phase arm 26 of the inverter circuit 20 are connected in parallel. It has a direct current power supply. In FIG. 1, the voltage of this DC power is shown as V _dc .

  When the rotary electric machine 12 is driven by the inverter circuit 20, as a control method thereof, a control mode called a sine wave control mode, a control mode called an overmodulation control mode, and a control mode called a rectangular wave control mode are properly used. Has been done. That is, in order to achieve both high output and miniaturization of the rotating electrical machine 12, a rectangular wave control mode using one-pulse switching is required, a sine wave control mode having excellent characteristics in a low speed region, and a medium speed The operation of the rotating electrical machine 12 is optimally controlled while performing mode switching between the overmodulation control modes used in the region.

  Here, the sine wave control mode and the overmodulation control mode are current feedback control, and in the conventional technology, the PWM command is output to the rotating electrical machine 12 by comparing the voltage command value and the carrier wave (carrier). is there. On the other hand, the rectangular wave control mode is a control for outputting a one-pulse switching waveform to the rotating electrical machine 12 according to the electrical angle, and the voltage amplitude is fixed to the maximum value, and the torque is feedback controlled by controlling the phase. .

  The sine wave control mode is sometimes called a PWM control mode because a typical PMW signal becomes a pseudo sine wave signal. When the PWM signal is literally a pulse width modulation signal, both the overmodulation control mode and the sine wave control mode use the pulse width modulation signal. In the following, in order to distinguish the PWM signal from the PWM control mode, the PWM control mode is used. The term “sine wave control mode” will be used.

  The control block 50 includes a plurality of arithmetic devices, and particularly includes a storage unit 48 and a PWM control device 52. The rotating electric machine is controlled by the PWM signal generated by the PWM control device 52 through the control of the inverter circuit 20. 12 functions to control the operation. The control block 50 uses the sine wave control mode, overmodulation control mode, and rectangular wave control mode as the operation control modes of the rotating electrical machine 12 consistently for these three control modes without changing the circuit. be able to.

  The control block 50 is a circuit that specifically controls the operation of the rotating electrical machine 12 via the inverter circuit 20, and the contents thereof are shown in the control block diagram in FIG. 1. The rotating electrical machine control device can be configured by hardware, but can also be configured by software. Of course, a part may be configured by hardware.

The torque command value T ^* and the rotational angular velocity command value ω ^{* in} the control block diagram of FIG. 1 are the operation target command value 30 for the rotating electrical machine 12. These command values are calculated by estimating the user's requested torque and requested vehicle speed from the accelerator, brake, etc. of the vehicle not shown.

The current command generator 32 compares the actual rotational angular velocity ω of the rotating electrical machine 12 with the rotational angular velocity command value ω ^*, and uses the previously created table or the like to determine the torque command value T ^* as the d-axis current command value I _d. ^It has a function to calculate as a set of ^* and q-axis current command value I _q ^* .

Here, an angle-speed conversion section 80 has a function of converting the electrical angle theta _e of the rotor detected by the resolver 14 to the rotational angular velocity omega. The converted rotational angular velocity ω is transmitted to the current command generation unit 32 described above.

Subtraction unit 36, a subtractor having a function of calculating a d-axis current deviation [Delta] I _d by subtracting the d-axis current value I _d from the d-axis current command value I _d ^*, q axis current instruction value I _q ^* from the q It is a subtractor that subtracts the shaft current value I _q to calculate the q-axis current deviation ΔI _q .

Here, the dq coordinate conversion unit 82 has a function of calculating the d-axis current value I _d and the q-axis current value I _q based on the phase current values I _U , I _V , and I _W of the rotating electrical machine 12. That is, it is an arithmetic unit having a function of converting the current from the UVW coordinate system to the dq coordinate system. As described above, since the U-phase current value I _U and the V-phase current value I _V are detected by the current detectors 16 and 18, the remaining W-phase current value I _W is calculated from the two phase current values, from the electrical angle theta _e of the rotor detected by the three phase current values and the resolver 14, the coordinate transformation calculation, d-axis current value I _d and the q-axis current value I _q is calculated. The calculated d-axis current value I _d and q-axis current value I _q are transmitted to the subtractor 36.

The current control unit 38, the d-axis current deviation [Delta] I _d and the q-axis current deviation [Delta] I _q, seek control deviation corresponding thereto by performing a proportional integral control under a predetermined gain, the d-axis in accordance with the control deviation The voltage command value V _d ^* and the q-axis voltage command value V _q ^* are calculated.

The linear correction unit 42 matches the d-axis voltage command value V _d ^* and the q-axis voltage command value V _q ^* so that the fundamental wave amplitude corresponding to the generated PWM signal matches the voltage command amplitude corresponding to the voltage command vector ^. Is an arithmetic unit having a function of correcting The correction value is stored in the storage unit 48 as a linear correction value for each synchronization number K. The linear correction unit 42 reads the linear correction value corresponding to the synchronization number K from the storage unit 48 and corrects the d-axis voltage command value V _d ^* and the q-axis voltage command value V _q ^* .

The αβ coordinate conversion unit 44 is an arithmetic device having a function of converting the d-axis voltage command value V _d ^* and the q-axis voltage command value V _q ^* after linear correction into voltage command values Vα and Vβ in the αβ coordinate system. Here, the αβ coordinate system is an orthogonal fixed coordinate system used in the space vector method. The dq coordinate system is an orthogonal rotation coordinate system, and the UVW coordinate system is a fixed coordinate system having three reference axes having a phase of 120 degrees with respect to each other.

  The PWM control device 52 includes a synchronization control device 54 and a PWM signal generation device 56 as its internal configuration. Since these generate a PWM signal using the space vector method, the space vector method will be briefly described here.

  As described above, when the switching elements constituting the U-phase arm 22, V-phase arm 24, and W-phase arm 26 of the inverter circuit 20 are turned on / off, the upper element on / lower element off is 1, the upper element off / lower element on. If 0 is 0, there are 8 combinations of 1 and 0 of each of the three phase arms. That is, the state of the inverter circuit 20 can be indicated by eight types of switching patterns, and these eight types of switching patterns are associated with two types of zero vectors and six types of basic voltage vectors. Then, these six types of basic voltage vectors are sequentially arranged on the phase plane with a phase difference of 60 degrees from each other, thereby dividing the six phase regions. A method of generating a PWM signal for each phase arm based on the relationship between the six phase regions I to the phase region VI and the voltage command vector V of the inverter circuit 20 is a PWM signal generation method using a space vector method.

  The synchronization control device 54 includes a synchronization number setting unit 60 for setting the synchronization number K, and six basic voltages on the phase plane, with the phase at the start point of the control angle period Δθ determined by the synchronization number as the interrupt phase. An interrupt phase synchronization unit 62 that synchronizes with any one phase of the vector and a synchronization number switching unit 64 that switches the synchronization number K according to the control mode of the rotating electrical machine 12 are configured.

  Here, the control angle period Δθ is a period on the phase plane for generating the PWM signal. In the case of full cycle control, 2π / K, which is an angle obtained by dividing 2π of one phase of a voltage command vector in a three-phase inverter circuit by an integer number of synchronization numbers K, is a control angle cycle Δθ, and in the case of half cycle control In this case, π / K which is a half of the control angle period is Δθ.

  The PWM signal generating device 56 generates a voltage command vector V from the voltage command values Vα and Vβ in the αβ coordinate system, and a prediction phase for predicting the timing for generating a PWM signal using the voltage command vector. Is a region determination unit 68 that determines which region from phase region I to VI in the space vector method belongs to, and a phase region to which the predicted phase belongs is a predicted phase region, and a voltage command vector for a switching vector that defines the predicted phase region The angle ratio calculation unit 70 that calculates the contribution ratio as an angle ratio with respect to the control angle period Δθ, and the switching switching order in the predicted phase area are read, and in the predicted phase area at the angle ratio calculated by the angle ratio calculating means according to the switching order. Calculate the switching phase between each switching vector sequentially, Configured to include a switching 換位 phase calculating unit 72 which generates a PMW signal in the phase domain.

  These functions of the control block 50 can be realized by executing software. Specifically, it can be realized by executing a rotating electrical machine control program. Some of these functions may be realized by hardware.

  The storage unit 48 has a function of storing a rotating electrical machine control program. Further, it has a function of storing linear correction value data associated with the number of synchronizations described in the linear correction unit 42. In addition, the switching phase calculation unit 72 has a function of storing the switching switching order associated with the phase region. Here, the switching order is a switching order of switching vectors for generating a PWM signal using two basic voltage vectors and two zero vectors defining one phase region as switching vectors.

As described above, in the configuration of FIG. 1, each phase current value as an actual current value in the rotating electrical machine 12 is converted into the d-axis current value I _d and the q-axis current value I _q and fed back to the subtraction unit 36. The d-axis current command value I _d ^* is compared with the d-axis current value I _d . Then, the voltage command value is calculated so that the deviation between the command current value and the actual current value is zero, a voltage command vector is generated from the voltage command value, and the control angle period Δθ corresponding to the voltage command vector The angle ratio to is obtained. On / off of the U-phase arm 22, the V-phase arm 24, and the W-phase arm 26 of the inverter circuit 20 is determined based on the angle ratio, and thereby, each phase current value to the rotating electrical machine 12 is determined. Each phase current value is fed back to the subtractor 36 again as described above. By repeating this, the rotating electrical machine 12 can perform an operation following the torque command value T ^* and the rotational angular velocity command value ω ^* .

  The operation of the above configuration, in particular, each function of the PWM control unit device 52 will be described in detail below with reference to the drawings. First, a brief description of the space vector method will be given with reference to FIGS. 2 and 3, and then the content of PWM signal generation in the case of half-cycle control will be described with reference to FIGS. The linear correction will be described with reference to FIGS. Thereafter, the contents of PWM signal generation in the case of full-cycle control will be described with reference to FIGS. The division of the phase region for improving accuracy, the change of the synchronization number, etc. will be described later.

  FIG. 2 is a diagram illustrating a relationship among six basic voltage vectors, six phase regions, and d-axis and q-axis in the space vector method. The six basic voltage vectors in the space vector method are (0, 0, 0) and (0) of eight combinations of the U phase arm 22, the V phase arm 24, and the W phase arm 26 of the inverter circuit 20. This is a vector corresponding to six except two of 1,1,1).

  (1, 0, 0) is when the U-phase arm 22 is in the upper element on / lower element off state, and (0, 1, 0) is when the V-phase arm 24 is in the upper element on / lower element off state ( 0, 0, 1) is a state in which the W-phase arm is in the upper element on / lower element off state. (1, 1, 0), (0, 1, 1), (0, 0, 1) are intermediate states. The basic voltage vectors representing these states are read by converting the 3-bit state into a decimal value, and (1, 0, 0) is V1, (0, 1, 0) is V2, (1, 1, 0) ) As V3, (0,0,1) as V4, (1,0,1) as V5, and (0,1,1) as V6.

  Since these six basic voltage vectors have one phase, the phase between the basic voltage vectors is 60 degrees. That is, when V1 is taken as a reference, V3, V2, V6, V4, and V5 are sequentially arranged on the phase plane with a phase difference of 60 degrees counterclockwise. By the arrangement of the six basic voltage vectors, six phase regions I, II, III, IV, V, and VI are partitioned in a phase plane.

That is, a phase region I partitioned by two basic voltage vectors V1 and V3, a phase region II partitioned by two basic voltage vectors V3 and V2, a phase region III partitioned by two basic voltage vectors V2 and V6, A phase region IV defined by two basic voltage vectors V6 and V4, a phase region V defined by two basic voltage vectors V4 and V5, and a phase region VI defined by two basic voltage vectors V5 and V1.
In this way, the hexagon 90 in the space vector method is formed. Note that two of V0 representing (0, 0, 0) and V7 representing (1, 1, 1) are called zero vectors and are arranged at the origin of a hexagon of the space vector method.

  The UVW coordinate system is a fixed coordinate system composed of three basic axes having a phase difference of 120 degrees between the U axis corresponding to V1, the V axis corresponding to V2, and the W axis corresponding to V4. . The αβ coordinate system is a fixed coordinate system including an α axis corresponding to V1 and a β axis orthogonal to the α axis. The dq coordinate system is a rotational coordinate system composed of two basic axes, d-axis and q-axis that are orthogonal to each other.

In FIG. 2, the electrical angle θ _e is the rotation angle of the d axis with respect to the U axis of the UVW coordinate system, and φ _V is the phase of the voltage command vector based on the d axis. Therefore, in the αβ coordinate system, the phase of the voltage command vector V is represented by (φ _V + θ _e ). Here, since the resolver 14 detects the rotation angle of the rotor with respect to the stator, it corresponds to the rotation angle of the d-axis with respect to the U-axis of the UVW coordinate system, and is a value corresponding to the electrical angle θ _e in FIG. is there. Since φ _V is the phase of the voltage command vector V with respect to the d axis, it is calculated from the d axis voltage command value V _d ^* and the q axis voltage command value V _q ^* .

FIG. 3 is a diagram showing a relationship between a circle defined by six phase regions of the space vector method and a DC voltage V _dc applied to the three-phase inverter circuit. As shown in FIG. 3, the radius of the inscribed circle 92 for each side 94 of the hexagon 90 on the phase plane defined by the phase region I to the phase region VI is V _dc / (2) ^1/2 . The radius of the envelope circle 96 connecting the vertices of the square 90 is (2/3) ^1/2 V _dc .

  Here, when the origin of the hexagon 90 on the phase plane is the origin of the voltage command vector V, and the magnitude of the voltage command vector V is within the hexagon 90, the voltage command vector V corresponds to the sine wave control mode. Corresponds to the overmodulation control mode when the basic voltage vector component of the voltage command vector is inside the envelope circle, and the basic voltage vector component of the voltage command vector exceeds the envelope circle and goes outside When it extends, it corresponds to the rectangular wave control mode.

  Next, regarding the half-cycle control, the specific content of PWM signal generation when the number of synchronizations K = 6, 9 will be described. Here, the synchronization number K = 6 will be described in detail, and the synchronization number K = 9 will be described focusing on the differences.

  FIG. 4 is a diagram for explaining the interrupt phase when the synchronization number K = 6 and half-cycle control is performed. The interrupt phase is a phase on the phase plane of the space vector system that indicates the timing for obtaining the phase current of the rotating electrical machine and the electrical angle of the rotor for generating the PWM signal. The interrupt phase corresponds to the starting point of each control angle period Δθ that is a period of PWM signal generation.

  In conventional PWM signal generation using a triangular wave, the control method for performing the period from the trough of the triangular wave to the peak period from the timing of data acquisition of phase current etc. to the calculation of the fundamental wave signal indicating the voltage command value is a half cycle control, The control method performed in the valley to valley period is called full cycle control. In accordance with this distinction, the control angle period Δθ in the space vector method is an angle obtained by dividing 2π of one phase of the voltage command vector in the three-phase inverter circuit by an integer synchronization number K in the case of full-cycle control. / K, and in the case of half-cycle control, it is π / K, which is 1/2 of that. As described above, the setting of the synchronization number K and the control angle cycle Δθ is executed by the function of the synchronization number setting unit 60 of the synchronization control device 54.

  Therefore, when the synchronization number K = 6 and half-cycle control is performed, the interrupt phase is set at intervals of π / 6 = 180 degrees / 6 = 30 degrees. In FIG. 4, interrupt phases at intervals of 30 degrees are indicated by broken lines and triangle marks. Thus, by setting the control angle period Δθ to a value obtained by dividing π or 2π by the number of synchronizations, the synchronization between the electrical period of each phase corresponding to the rotation period of the rotor and the control period can be secured. As in the case of control, even if one pulse drive is performed in one cycle, fluctuations in the voltage actually applied to the rotating electrical machine 12 are suppressed, and occurrence of a beat, which is a phenomenon that the motor current vibrates, can be suppressed.

FIG. 5 is a diagram for explaining the interrupt phase in the case of half-cycle control with the synchronization number K = 6 in relation to the hexagon 90 of the space vector method. Here, although the phase of the starting point of the control angle period Δθ is an interrupt phase, the interrupt phase is synchronized with any one of six basic voltage vectors on the phase plane. This is the same content as ensuring the synchronism of the electrical cycle and control cycle of each phase described above. Specifically, as described in FIG. 2, the phase of the value obtained by adding the control angle period Δθ to (φ _V + θ _e ), which is the phase of the voltage command vector V in the αβ coordinate system, is represented by six basic voltage vectors. It is made to correspond to any one phase.

  FIG. 5 shows an example in which the interrupt phase matches the phase of V1. The square marks taken on the envelope circle 96 correspond to the 12 interrupt phases every 30 degrees described in FIG. As described above, the function of causing the interrupt phase to coincide with any one phase of the basic voltage vector is executed by the interrupt phase synchronization unit 62 of the synchronization control device 54.

  FIG. 6 is a diagram for explaining a predicted phase when the synchronization number K = 6 and half-cycle control is performed. Here, as described above, the interrupt phase is assumed to coincide with the phase of V1. The predicted phase is a phase of timing for generating a PWM signal using the generated voltage command vector V.

  As described above, in the PWM signal generation using the triangular wave of the prior art, from the timing of acquiring the data such as the phase current to the calculation of the fundamental wave signal indicating the voltage command value, in the case of half-cycle control, from the triangular wave trough to the peak In the case of full cycle control, the period is from the trough to trough of the triangular wave, and then the triangular wave signal and the fundamental wave signal are compared to determine the switching timing. If this conventional procedure is made to correspond to the case of the space vector method, the predicted phase comes after the control angle period Δθ. Therefore, the predicted phase is assumed to be advanced by nΔθ from the interrupt phase, and it is predicted that the voltage vector command exists in the predicted phase. Specifically, in the case of half-cycle control, n = 1.5, and the predicted phase is a position advanced by 1.5 Δθ from the interrupt phase. In FIG. 6, the predicted phase is shown as the predicted angle after 1.5 cycles.

FIG. 7 is a flowchart for explaining a procedure for generating a PWM signal when the synchronization number K = 6 and half-cycle control is performed. Here, when an interrupt occurs, the switching switching angle and the switching vector in the current control angle cycle are written in the interrupt phase (S10). Together, wherein the switching switching angle in the next control angle period, the acquisition of the electrical angle theta _e of acquisition and the rotor of each phase current of the rotary electric machine 12 for writing switching vector is performed.

Acquisition of each phase current is a current detection unit 16, obtains the electrical angle theta _e of the rotor is performed by the resolver 14. Next, the voltage command value after linear correction is obtained by the functions of the angle / speed conversion unit 80, the dq coordinate conversion unit 82, the current command generation unit 32, the subtraction unit 36, the current control unit 38, and the linear correction unit 42 of the control block 50. V _d and V _q are calculated. Then, the voltage command values Vα and Vβ of the αβ coordinate system are created by the function of the αβ coordinate conversion unit 44, and the voltage command vector V is generated by the function of the voltage command vector generation unit 66 of the PWM signal generation device 56. However, at this time, as a phase region in which the voltage command vector V exists, a predicted phase that is a predicted angle after 1.5 cycles from the interrupt phase is used. That is, the voltage command values Vα and Vβ in the αβ coordinate system in the predicted phase that is the predicted angle after 1.5 cycles are calculated from the interrupt phase, and a voltage command vector is created based on this (S12).

The αβ coordinate conversion unit 44 uses the following equations (1) and (2) to convert the d-axis voltage command value V _d and the q-axis voltage command value V _q after linear correction into the α-axis voltage command value in the predicted phase. Vα and β-axis voltage command value Vβ are converted. This is performed in order to calculate a voltage average value at the next interruption. Since the predicted phase is 1.5Δθ from the interruption phase of the αβ coordinate system, the electrical angle θ _e is used in the dq coordinate system to be (θ _e + 1.5Δθ). Expression (1) is for the case of normal rotation of the rotating electrical machine 12, and Expression (2) is for the case of reverse rotation of the rotating electrical machine 12. Here, the time when the voltage command vector V rotates counterclockwise on the phase plane is the forward rotation, and the time when the voltage command vector V rotates clockwise is the reverse rotation.

The voltage command vector V is calculated as (Vα ² + Vβ ² ) ^1/2 , and the d-axis reference phase φ _V is calculated as tan ⁻¹ (Vβ / Vα).

Then, the function of the region determination unit 68 of the PWM signal generation device 56 determines which region of the phase regions I to VI the predicted phase belongs to (S14). Specifically, the control angle period Δθ = π / synchronization number = 180 degrees / 6 = 30 degrees is used to determine in which phase region the phase advanced by nΔθ = 1.5Δθ = 45 degrees from the phase of the interrupt phase. To do. In the above case, since the interruption period is the phase of V1, it can be determined that the phase advanced 45 degrees therefrom is in the phase region I. In an actual dq coordinate system, the phase of (θ _e + φ _V + 1.5Δθ) is determined in which phase region using the electrical angle θ _e and the phase φ _V of the voltage command vector based on the d axis. .

Next, the phase region to which the predicted phase that is the predicted angle after 1.5 cycles belongs is set as the predicted phase region, and the contribution ratio of the voltage command vector to the switching vector that defines the predicted phase region is calculated as the angle ratio with respect to the control angle cycle Δθ (S16). This process is executed by the function of the angle ratio calculation unit 70 of the PWM signal generation device 56. Here, in FIG. 7, the contribution ratio of the voltage command vector to the switching vector is shown as an angle ratio with respect to the control angle period Δθ, and as SW angles, Δθ _max , Δθ _min , Δθ _zero .

Δθ _max is the angle ratio of the voltage command vector to the maximum angle vector whose phase is advanced among the two basic voltage vectors defining the predicted phase region, and Δθ _min is the voltage for the minimum angle vector whose phase is delayed The angle ratio of the command vector, and Δθ _zero is the angle ratio of the voltage command vector with respect to the two zero vectors.
Here, the angle ratio is used to determine the length of the switching period for how to switch the switching state of each phase arm during the control angle period Δθ. The length of the switching period is determined not by time but by the angle ratio in the control angle period Δθ.

For example, in phase region I, the two basic voltage vectors are V1 and V3, and V1 is (1, 0, 0), so that the U-phase arm is in the 0 state, the V-phase arm is in the 0 state, The arm is in the 1 state. Similarly, V3 is (1, 1, 0) when the U-phase arm is 1, the V-phase arm is 1, and the W-phase arm is 0. Since the maximum angle vector whose phase is advanced is V3, Δθ _max indicates a ratio of how long this (1, 1, 0) state continues. Similarly, Δθ _min indicates the ratio of how long the (1, 0, 0) state of V1 continues, and (0, 0, 0) and V7 ( Δθ _zero indicates the ratio of how long the state of 1,1,1) lasts.

  For this purpose, it is necessary to obtain a relational expression between the voltage command vector V, the two basic voltage vectors, and the two zero vectors. This will be described with reference to FIGS.

FIG. 8 is a diagram showing the relationship between the two basic voltage vectors V1 and V3 and the αβ coordinate system for the phase region I. Here, V1 is shown as Vθ _min and V3 is shown as Vθ _max . The distinction between min and max is set to max when the phase is advanced and min as the delayed phase. Therefore, the phase region I is as described above, but in the phase region II, V3 is Vθ _min and V2 is Vθ _max . Similarly, in the other phase regions, the maximum angle vector whose phase is advanced is Vθ _max and the minimum angle vector whose phase is delayed is Vθ _min . The contents described below for the phase region I can be similarly applied to other phase regions by distinguishing the two basic voltage vectors of each phase region from Vθ _min and Vθ _max .

FIG. 8 shows an example of a state where Vθ _min that is V1 is advanced by a phase γ with respect to the α axis. Here, assuming that the α-axis component of Vθ _min is Vαθ _min and the β-axis component is Vβθ _min , Vαθ _min = Vθ _min · cos γ, and Vβθ _min = Vθ _min · sin γ. Similarly, when the α axis component of Vθ _max is Vαθ _max and the β axis component is Vβθ _max , Vαθ _max = Vθ _max · cos (60 ° + γ), Vβθ _max = Vθ _max · sin (60 ° + γ) It is.

Thus the α-axis component of V [theta] _min at Vαθ _min, represents a β-axis component in Vβθ _min, the α-axis component of V [theta] _max in Vαθ _max, represents a β-axis component in Vβθ _max, similarly α axis zero vector The component is represented by V ₀ α, and the β-axis component is represented by V ₀ β. Then, assuming that the angle ratio, which is the contribution ratio of Vθ _max , is Δθ _max , the angle ratio of Vθ _min is Δθ _min, and the angle ratio of the zero vector is Δθ _zero , Vα obtained by Expressions (1) and (2) , Vβ can be written as the following general formulas (3) and (4).

Since Δθ _max , Δθ _min, and θ _zero indicate the contribution ratio of Δθ in angle, the relationship of Expression (5) is established.

Here, since V ₀ α and V ₀ β are zero, the following equation (6) is obtained from equation (4).

Substituting this into equation (3) yields equation (7).

By rewriting this, Δθ _min is obtained as shown in Equation (8).

Similarly, Δθ _max is obtained as shown in equations (9), (10), and (11).

In this way, general formulas for Δθ _min and Δθ _max are obtained. If these two are obtained, Δθ _zero can be obtained from equation (5).

Here, when the phase region I is considered, in FIG. 8, γ = 0 °. The state of Vαθ _min , Vβθ _min , Vαθ _max , and Vβθ _max at that time is shown in FIG. As shown here, Vαθ _min = Vθ _min · cos 0 ° = Vθ _min , Vβθ _min = 0, Vαθ _max = Vθ _max · cos 60 ° = (1/2) Vθ _max , Vβθ _max = Vθ _max · sin 60 ° = (3 ^1/2 / 2) Vθ _max . In the case of the phase region II, Vαθ _min , Vβθ _min , Vαθ _max , and Vβθ _max can be obtained similarly with γ = 60 ° in FIG. For other phase regions, Vαθ _min , Vβθ _min , Vαθ _max , and Vβθ _max can be obtained similarly by setting γ to a corresponding angle.

In FIG. 9, when the magnitudes of Vθ _min and Vθ _max are assumed to be 1, and the denominators of equations (8) and (11) are calculated, equation (12) is obtained.

As described in FIG. 3, the magnitudes of Vθ _min and Vθ _max are (2/3) ^1/2 V _dc . Here, _assuming that (V _dc / 2) = 1, where (V _dc / 2) is a standardized voltage unit, the magnitudes of Vθ _min and Vθ _max are 2 × (2/3) ^1/2. It becomes. When this is applied to equation (12), the denominators of equations (8) and (11) are (3 ^1/2 ) / 2 multiplied by 2 × (2/3) ^1/2 , (2 ^{1 / 2} ).

Equation (8), V [theta] _min and V [theta] _max in (11), it indicates the contribution of the [Delta] [theta] as a unit, indicating V [theta] _min and V [theta] _max at an angle ratio control angle period [Delta] [theta], equation (13) It becomes. In this equation, (V _dc / 2) is taken as a standardized voltage unit, and Vα, Vβ, Vαθ _min , Vβθ _min , Vαθ _max , and Vβθ _max are also (V _dc / 2) = 1V. Is the value indicated by the standardized voltage.

Specifically, FIG. 10 shows how Δθ _min and Δθ _max are obtained for the voltage command vector V in the phase region I. As shown in FIG. 10, the contribution ratio of the voltage command vector V to Vθ _min is {Vα− (1/3 ^1/2 ) Vβ}, and the contribution ratio of the voltage command vector V to Vθ _max is (2 / 3 ^1/2 ) Vβ}. In order to match this value with the standardized voltage, it is sufficient to divide by 2 × (2/3) ^1/2 that is the magnitude of Vθ _min and Vθ _max .

When the result is expressed as an angle ratio with respect to the control angle period Δθ, the angle ratio of the voltage command vector V to Vθ _min is (Δθ / 2 ^1/2 ) [{(3 ^1/2 / 2) Vα− (1 / 2) Vβ], Vβθ _max = Vθ _max · sin 60 ° = sin 60 ° = (3 ^1/2 / 2), Vαθ _max = Vθ _max · cos 60 ° = cos 60 ° = (1/2) , Which coincides with Δθ _min in equation (13). Similarly, when the contribution ratio of the voltage command vector V to Vθ _{max is} expressed as an angle ratio with respect to the control angle period Δθ, Vβ is obtained, which is Vαθ _min = Vθ _min · cos 0 ° = cos 0 ° = 1, Vθ _min , Vβθ. _By re-setting _min = 0, the value agrees with Δθ _max in the equation (13).

As shown in FIG. 10, by normalizing the voltage and normalizing the magnitude of Δθ to the magnitudes of Vθ _min and Vθ _max , the Vθ _min component of the voltage command vector V becomes Δθ _{min as} it is, and the voltage command vector V It can be seen that the Vθ _max component of A becomes Δθ _max as it is.

In the above description, the phase region I has been described. However, in general formulas of Δθ _min and Δθ _max including other phase regions, the phase between the α axis and Vθ _min is defined as θ _min and the phase between the α axis and Vθ _{max. Is} given by equation (14) where θ _max is θ _max .

FIG. 11 is a diagram showing the states of Δθ _min and Δθ _max of the voltage command vector V when the number of synchronizations is K = 6 and the interrupt phase is made to match the basic voltage vector of Vmin in the half cycle control. Note that Vθ _min and Vθ _max are the minimum angle vector that is delayed in phase and the maximum angle vector that is advanced in phase among the two basic voltage vectors that divide each phase region. Hereinafter, it will be simply expressed as Vmin and Vmax.

In this case, since the predicted phase is 45 degrees from the phase of Vmin, it is predicted that the voltage command vector V exists at that phase. Here, if the magnitude of Δθ is the magnitude of Vmin and Vmax, the Vmin component of the voltage command vector V becomes the angle ratio Δθ _min , and the Vmax component becomes the angle ratio Δθ _max . The remainder obtained by subtracting (Δθ _min + Δθ _max ) from the magnitude of Vmin or Vmax is Δθ _zero .

  When the angle ratio of the voltage command vector V with respect to each switching vector is calculated in this way, the process returns to FIG. 7 again to select the switching vector in accordance with a predetermined switching switching order, and the selected switching vector The period is determined by the angle ratio, the switching phase between the switching vectors is calculated, and the PWM signal is generated (S18). This process is executed by the function of the switching phase calculation unit 72 of the PWM signal generation device 56. In FIG. 7, the switching switching angle calculation and switching vector selection in the next cycle are shown. The switching angle is the switching phase.

  Since the switching order of the switching vectors is predetermined in each phase region, the switching vectors are sequentially selected according to this switching order. FIG. 12 is a diagram showing the switching order of switching vectors in the half cycle control in association with each phase region. This switching switching order is divided into half-cycle control and full-cycle control, and is stored in advance in the storage unit 48 in association with each phase region.

  In FIG. 12, the region is a phase region. Also, the synchronization number timing is used to distinguish each phase region because it is divided into two in half-cycle control. The synchronization number timing 1 is a region on the Vmin side, and the synchronization number timing 2 is a region on the Vmax side. The minimum angle and the maximum angle are the phase of Vmin and the phase of Vmax based on the α axis. The selection vector indicates two basic voltage vectors among the four switching vectors in each phase region, and the minimum angle is Vmin and the maximum angle is Vmax. The output vector order is a switching order of switching vectors, and indicates that switching vectors are sequentially selected and switched in this order in the direction of advancing phase.

  For example, in the phase region I, the switching vectors are two basic voltage vectors V1 and V3, and two zero vectors V0 and V7. These switching vectors are selected and sequentially switched in the order of V7-V3-V1-V0-V0-V1-V3-V7 from the phase V1 to the phase V3. When the synchronization number K = 6, the angle between the phase V1 and the phase V3 is exactly the control angle period Δθ. Therefore, in this one control angle period, the U-phase arm 22 and the V-phase arm of the inverter circuit 20 24. Switching of the W-phase arm 26 is switched in this order. The switching interval is set by the angle ratio calculated in S16.

  As described above, the switching switching order is stored in advance in the storage unit 48, but the storage format may be other than the lookup table format as shown in FIG. FIG. 13 shows the switching order in a flowchart format, but it may be stored in the storage unit 48 in this format. FIG. 14 shows the switching order in the map format, but it may be stored in the storage unit 48 in this format.

In the flowchart of FIG. 13, first, the phase region is specified (S22). The specific procedure for the phase region has already been described in S14. The result of S22 is divided into two types. For example, in the case of the phase number I synchronization number timing 1, the process proceeds to S24, and in the case of the phase area I synchronization number timing 2, the process proceeds to S26. In the case of S24, V7 is first selected and the angle ratio is (Δθ _zero / 2). In the case of S26, V0 is first selected and the angle ratio is (Δθ _zero / 2).

Next, the phase region is specified again (S28), and the process proceeds to S30 or S32 depending on the result. For example, in the case of the synchronization number timing 1 in the phase region I, the process proceeds to S30, and in the case of the synchronization number timing 2 in the phase region I, the process proceeds to S32. In the case of S30, the maximum angle vector V _max is selected as the second switching vector, and its angle ratio is Δθ _max . Next, the minimum angle vector V _min is selected as the third switching vector, and its angle ratio is Δθ _min . In the case of S32, V _min is selected as the second switching vector, and its angle ratio is Δθ _min . Next, V _max is selected as the third switching vector, and its angular ratio is Δθ _max .

Then, the phase region is identified again (S34), and the process proceeds to S36 or S38 depending on the result. For example, in the case of the synchronization number timing 1 in the phase region I, the process proceeds to S36, and in the case of the synchronization number timing 2 in the phase region I, the process proceeds to S38. In the case of S36, V0 is selected and the angle ratio is (Δθ _zero / 2). In the case of S38, V7 is selected and the angle ratio is (Δθ _zero / 2).

FIG. 14 shows the selection and switching order of the switching vectors for the phase region I. The switching phases between the switching vectors are indicated by θ _comp1 , θ _comp2 , and θ _comp3 . This switching phase can be determined as follows according to the angle ratio.

That is, at the synchronization number timing 1, the first zero vector is V7 and the last zero vector is V0, but at the synchronization number timing 2, the first zero vector is V0 and the last zero vector is V7. It is. Then, θ _comp1 , which is the first switching phase of each half cycle control, is set by Expression (15) using (φ _V + θ _e ) that is the phase of voltage command vector V and control angle period Δθ.

_Further , θ _comp2 and θ _{comp3 in} the case of the synchronization number timing 1 are set by Expression (16) using the angle ratio Δθ _min and the angle ratio Δθ _max .

Similarly, θ _comp2 and θ _{comp3 in} the case of the synchronization number timing 2 are also set by Expression (17) using the angle ratio Δθ _min and the angle ratio Δθ _max .

  In the above description, the synchronization number timings 1 and 2 of the phase region I have been described. However, the switching phase can be similarly set for the other phase regions. That is, the equation (16) used for the synchronization number timing 1 of the phase region I is the synchronization number timing 2 of the phase region II, the synchronization number timing 1 of the phase region III, the synchronization number timing 2 of the phase region IV, and the phase region V The synchronization number timing 1 and the phase number VI synchronization number timing 2 can be used as they are. Further, the equation (17) used for the synchronization number timing 2 of the phase region I is the synchronization number timing 1 of the phase region II, the synchronization number timing 2 of the phase region III, the synchronization number timing 1 of the phase region IV, and the phase region V The synchronization number timing 2 and the synchronization number timing 1 of the phase region VI can be used as they are.

As described above, the switching phase varies depending on the angle ratio Δθ _min and the angle ratio Δθ _max , so that the switching phase varies depending on the magnitude of the voltage command vector V. FIG. 15 shows an angle ratio Δθ _min and an angle ratio Δθ _{max of} two basic voltage vectors for each region obtained by dividing the phase region I into two when the synchronization number K = 6 and half-cycle control, and this angle ratio is set as a voltage command. It is a figure shown corresponding to the switching order of a vector.

In FIG. 15, the voltage command vector remains inside the hexagonal side 94 that defines the phase region I. FIG. 16 shows the time when the voltage command vector has just reached the hexagonal side 94. Since this hexagonal side 94 is a boundary line that enters the overmodulation control mode beyond this, the voltage command vector in FIGS. 15 and 16 is driven in the sine wave control mode. FIG. 16 shows the limit of the sine wave control mode, and the zero vector disappears. That is, Δθ = (Δθ _max + Δθ _min ).

  As shown in FIGS. 15 and 16, when an angle ratio is set for each switching vector in a predetermined switching order, a PWM signal is generated based on the angle ratio. For example, at the synchronization number timing 1 in the phase region I in FIG. 15, the switching vectors are switched in the order of V7-V3-V1-V0. When these are arranged in eight switching patterns, the order is (1,1,1)-(1,1,0)-(1,0,0)-(0,0,0).

Here, when the change of the switching state of the U-phase arm 22 is aligned with the angle ratio, one state of the angle ratio (Δθ _zero / 2) −one state of the angle ratio Δθ _{max (1)} −angle ratio Δθ _{min (} The 1 state of ₁₎ -the angle ratio (Δθ _zero / 2) is 0 state. Similarly, when the change of the switching state of the V-phase arm 24 is aligned with the angle ratio, one state of the angle ratio (Δθ _zero / 2) −one state of the angle ratio Δθ _{max (1)} −angle ratio Δθ _{min (} 0 state of ₁₎ - a 0 state of the angle ratio (Δθ _zero / 2). When the changes in the switching state of the W-phase arm 26 are arranged together with the angle ratio, the one state of the angle ratio (Δθ _zero / 2) —the _zero state of the angle ratio Δθ _{max (1)} —the angle ratio Δθ _{min (1)} The zero state is the zero state-angle ratio (Δθ _zero / 2).

  If this is arranged for each phase region, the switching pattern of each phase arm of the inverter circuit 20 is obtained for each phase. This is the PWM signal obtained for the phase. Since the switching pattern can be determined by the phase regardless of time, it is not necessary to prepare a large timer even if the frequency of the fundamental wave signal of the inverter circuit 20 is low.

FIG. 17 is a diagram showing a result of generating the PWM signal in the sine wave control mode, that is, the voltage command vector in the state shown in FIGS. Here, the horizontal axis represents phase, and the vertical axis represents voltage. For example, the section in which the phase is 0 degree to 30 degrees corresponds to the region of the synchronization number timing 1 in the phase region I. However, as described above, the switching state of the U-phase arm is 1 state-1 state-1 State-0 state, the switching state of the V phase arm is 1 state-1 state-0 state-0 state, and the switching state of the W phase arm is 1 state-0 state-0 state-0 state . Similarly, the PWM signal indicating the switching state of each phase arm is also generated in the other phase sections.

  When the PWM signal is generated in this way, the process returns to FIG. 7 again, the interrupt angle which is the next interrupt phase is calculated, and a series of processes in the control angle cycle Δθ is completed (S20).

  The above is the explanation in the sine wave control mode. FIG. 18 shows a state in which the magnitude of the voltage command vector is further larger than the state of FIG. FIG. While the basic voltage vector component of the voltage command vector remains inside the envelope circle 96, the overmodulation control mode is set. However, when the basic voltage vector component of the voltage command vector reaches the envelope circle 96, the rectangular wave control mode is set. Become.

In the state of FIG. 18, the value of (Δθ _min + Δθ _max ) using Δθ _min and Δθ _max in calculating the angle ratio exceeds Δθ, so that the zero vector disappears as described in FIG. Then, either (Δθ _min + Δθ _max ) = Δθ or any one of the calculated angle ratios of Δθ _{min that} is the calculated angle ratio and Δθ _max that is the calculated angle ratio is directly used as one of the basics. The angle ratio of the voltage vector is the angle ratio of the other basic voltage vector, and the angle ratio of one of the basic voltage vectors is subtracted from 1. For example, when the calculated Δθ _min is prioritized, Δθ _min is left as it is, and Δθ _max = (Δθ−Δθ _min ). When this is mathematically expressed, the following equation (18) is obtained.

Of course, when the calculated Δθ _max is prioritized, Δθ _max is left as it is, and Δθ _min = (Δθ−Δθ _max ). When this is mathematically expressed, the following equation (19) is obtained. It can be determined in advance which of Δθ _min and Δθ _max is prioritized.

The switching order in the overmodulation control mode is the one in which the zero vector is omitted in the sine wave control mode. In the case of the phase region I with the synchronization number K = 6 in the half cycle control of FIG. 18, the switching order is V3-V1-V1-V3. The angle ratio between V1 and V3 is set while maintaining (Δθ _min + Δθ _max ) = Δθ by the one-side priority method described above. In this manner, when the angle ratio is set for each basic voltage vector in the predetermined switching order, a PWM signal in the overmodulation control mode is generated based on this. FIG. 19 is a diagram illustrating a result of generating a PWM signal in the overmodulation control mode.

  FIG. 20 is a diagram illustrating a state where the magnitude of the voltage command vector is further larger than the state of FIG. 18 and the basic voltage vector component of the voltage command vector has reached the envelope circle 96. This state is a rectangular wave control mode.

In the state of FIG. 20, one of Δθ _{min and} Δθ _{max in} the angle ratio calculation is the same as Δθ. Therefore, the larger calculated angle ratio of Δθ _{min that} is the calculated angle ratio and Δθ _max that is the calculated angle ratio is set to its basic voltage so that (Δθ _min + Δθ _max ) = Δθ. The angle ratio of the vector is set, and the angle ratio of the other basic voltage vector is set to zero. When this is mathematically expressed, the following equation (20) is obtained.

The switching order in the rectangular wave control mode is the arrangement order of the basic voltage vectors on the phase plane. When the synchronization number K = 6 and the phase region I is in the half cycle control of FIG. 20, the switching order is V1-V3. Then, as described above, the angle ratio between V1 and V3 is set to the larger one of Δθ _min calculated for each basic voltage vector and Δθ _max calculated. Actually, since the larger angle ratio is 1, V1 or V3 is applied over the entire period of the control angle period Δθ. In this manner, when 1 is set as the angle ratio for each basic voltage vector in a predetermined switching order, a PWM signal in the overmodulation control mode is generated based on this. FIG. 21 is a diagram illustrating a result of generating a PWM signal in the rectangular wave control mode.

  As described with reference to FIGS. 17, 19, and 21, by using the angle ratio, it is possible to consistently generate a PWM signal in the same method for the sine wave control mode, the overmodulation control mode, and the rectangular wave control mode.

  In the above description, the synchronization number K = 6 has been described in the half-cycle control. Even if the synchronization number K is other than 6, the PWM signal can be generated using the angle ratio similarly. FIG. 22 to FIG. 26 are examples of the synchronization number K = 9 in the half cycle control. In this case, the control angle period Δθ is 180 degrees / 9 = 20 degrees. FIGS. 22 and 23 are diagrams corresponding to FIGS. 4 and 5 in which the number of synchronizations K = 6, and show how the interrupt phase is set at intervals of 20 degrees. The predicted phase is 1.5 times 20 degrees, and the phase is 30 degrees ahead of the interrupt phase.

  FIG. 24 is a diagram corresponding to FIG. 17 in which the synchronization number K = 6, and is a diagram illustrating a state of PWM signal generation in the state of the sine wave control mode. FIG. 25 is a diagram corresponding to FIG. 19 in which the number of synchronizations K = 6, and is a diagram illustrating how PWM signals are generated in the overmodulation control mode. FIG. 26 is a diagram corresponding to FIG. 21 with the number of synchronizations K = 6, and is a diagram showing a state of PWM signal generation in the state of the rectangular wave control mode. In this way, by using the angle ratio, even if the synchronization number K is changed, the PWM signal can be generated consistently in the same manner for the sine wave control mode, the overmodulation control mode, and the rectangular wave control mode.

Next, linear correction will be described. The fundamental wave amplitude corresponding to the PWM signal generated as described above does not necessarily change linearly with the voltage command amplitude corresponding to the voltage command vector from which it was generated. For example, in the overmodulation control mode, the increase in the fundamental wave amplitude is less than the increase in the voltage command amplitude. The linear correction is to correct a deviation in linearity between the fundamental wave amplitude and the voltage command amplitude. The correction is performed by correcting the voltage command amplitude by correcting the d-axis voltage command value V _d ^* and the q-axis voltage command value V _q ^* .

Specifically, a voltage amplitude characteristic map indicating the relationship between the voltage command amplitude and the fundamental wave amplitude is stored in advance in the storage unit 48 as a linear correction value. Then, the linear correction unit 42 of the control block 50 reads the voltage amplitude characteristic map, calculates a voltage command amplitude necessary for matching the fundamental wave amplitude to the voltage command amplitude, and based on the calculated voltage command amplitude, The d-axis voltage command value V _d and the q-axis voltage command value V _q after linear correction are calculated. By doing in this way, it can be set as the fundamental wave amplitude according to voltage command amplitude.

  27 to 30 are voltage amplitude characteristic maps used as linear correction values. The vertical axis represents the fundamental wave amplitude, and the horizontal axis represents the voltage command amplitude necessary for matching the fundamental wave amplitude to the voltage command amplitude. It has been taken. These linear correction values are stored in the storage unit 48 in association with the synchronization number K.

  The half cycle control has been described in detail above. The half-cycle control is performed in the period from the valley of the triangular wave to the peak in the conventional PWM signal generation in the triangular wave from the acquisition timing of the data such as the phase current to the calculation of the fundamental wave signal indicating the voltage command value. Along with the half-cycle control, even in the conventional PWM signal generation in a triangular wave, full-cycle control is performed from the trough of the triangular wave to the calculation of the fundamental wave signal indicating the voltage command value from the timing of data acquisition such as phase current Has been done. Needless to say, the PWM signal generation using the angle ratio is applicable to this full cycle control. In the following, PWM signal generation using the angle ratio in full cycle control will be described focusing on differences from half cycle control. Here, following the description of mainly the synchronization number K = 6 in the half cycle control, the synchronization number K = 6 in the full cycle control will be mainly described with reference to FIGS. 31 to 38, and the synchronization number K = 9. The case will be briefly described with reference to FIGS. 39 and 40. FIG.

  The control angle period Δθ is twice different between the half-cycle control and the full-cycle control. FIGS. 31 and 32 are diagrams corresponding to FIGS. 4 and 5 of the half-cycle control and showing the state of the interrupt phase in the full-cycle control. In this case, since the control angle period Δθ is 360 degrees / 6 = 60 degrees, an interrupt phase is set every 60 degrees.

  FIG. 33 is a diagram corresponding to FIG. 6 of the half cycle control, and is a diagram illustrating the state of the predicted phase in the full cycle control. In the full cycle control, two prediction phases, ie, a first prediction phase and a second prediction phase are used. The first predicted phase is set to a phase advanced by 1.25 × Δθ from the interrupt phase, and the second predicted phase is set to a phase advanced by 1.75 × Δθ from the interrupt phase.

  FIG. 34 is a flowchart corresponding to FIG. 7 of the half cycle control, and is a flowchart for explaining the procedure of generating the PWM signal in the full cycle control. Since S40 has the same contents as S10 described in relation to FIG. 7, detailed description thereof is omitted.

  S42 is a processing procedure corresponding to S12 described in relation to FIG. That is, the voltage command values Vα and Vβ of the αβ coordinate system in the first predicted phase that is the predicted angle after 1.25 cycles from the interrupt phase are calculated, and the first predicted angle that is the predicted angle after 1.75 cycles from the interrupt phase. Voltage command values Vα and Vβ in the αβ coordinate system in two predicted phases are calculated.

Specifically, the αβ coordinate conversion unit 44 uses the following equations (21) and (22) to set the d-axis voltage command value V _d and the q-axis voltage command value V _q after linear correction to the first The α-axis voltage command value Vα and the β-axis voltage command value Vβ in the predicted phase and the second predicted phase are converted. This is performed in order to calculate a voltage average value at the next interruption. Expressions (21) and (22) are both in the case of normal rotation of the rotating electrical machine 12, but in the case of reverse rotation, the sign relating to Δθ may be changed from + to −.

  S44 is a processing procedure corresponding to S14 described in relation to FIG. 7, and since the content is the same as S14, detailed description thereof is omitted. In this case, the first predicted phase is the phase of V1. Since the second predicted phase is 105 degrees from the phase of V1, the predicted phase region is the phase region II.

S46 and S48 are processes for calculating Δθ _min , Δθ _max, and Δθ _zero in the processing procedure corresponding to S16 described in relation to FIG. 7. Here, S46 and S48 for the first predicted phase are calculated. The process is divided into S48 for the predicted phase. The subscripts _fh and _sh are used as abbreviations indicating first meaning first and second meaning second.

That is, in S46, Δθ _min-fh , Δθ _max-fh, and Δθ _zero-fh in the first predicted phase are calculated by Expression (23) corresponding to Expression (14) used in S16.

Expression (23) is a general expression, and is used in the sine wave control mode. However, in the case of the overmodulation control mode, Expression (18) and (19) can be selectively used depending on which of Δθ _min and Δθ _max is prioritized. Corresponding equations (24) and (25) are used.

In the case of the rectangular wave control mode, Expression (26) corresponding to Expression (20) is used.

Similarly, in S48, Δθ _min-sh , Δθ _max-sh, and Δθ _zero-sh in the second predicted phase are calculated by Expression (27) corresponding to Expression (23) in the first predicted phase.

Expression (27) is a general expression and is used in the sine wave control mode. In the case of the overmodulation control mode, the expression in the first prediction phase indicates which of Δθ _min-sh and Δθ _max-sh has priority. Expressions (28) and (29) corresponding to (24) and (25) are used.

In the case of the rectangular wave control mode, Expression (30) corresponding to Expression (26) in the first predicted phase is used.

FIG. 35 is a diagram corresponding to FIG. 11 and shows voltage command vectors at 1.25 Δθ and 1.75 Δθ when the number of synchronizations is K = 6 and the interrupt phase is matched with the basic voltage vector of V1 in full cycle control. It is a figure which shows the mode of ( _{DELTA) ( theta) min-fh} , ( _DELTA) (theta) _max-fh , ( _DELTA) (theta) _min-sh , (DELTA) (theta) _max-sh about each of these. As described above, since the first predicted phase is 75 degrees from the phase of V1, and the second predicted phase is 105 degrees from the phase of V1, it is predicted that the voltage command vector V exists at that phase.

Here, if the magnitude of Δθ is the magnitude of Vmin and Vmax, the Vmin component of the voltage command vector in the first predicted phase becomes the angle ratio Δθ _min-fh , and the Vmax component becomes the angle ratio Δθ _max-fh . The remainder obtained by subtracting (Δθ _min-fh + Δθ _max-fh ) from the magnitude of Vmin or Vmax is Δθ _zero-fh . Similarly, the Vmin component of the voltage command vector in the second predicted phase is the angle ratio Δθ _min-sh , and the Vmax component is the angle ratio Δθ _max-sh . The remainder obtained by subtracting (Δθ _min-sh + Δθ _max-sh ) from the magnitude of Vmin or Vmax is Δθ _zero-sh .

  Referring back to FIG. 34, S50 is a processing procedure corresponding to S18 described in relation to FIG. Then, the switching phase between the switching vectors is calculated, and the PWM signal is generated.

  FIG. 36 is a diagram corresponding to FIG. 12 and shows the switching order of switching vectors in full cycle control in association with each phase region. Since the content of each item is the same as that described in FIG. 12, detailed description thereof is omitted.

FIG. 37 is a flowchart corresponding to FIG. Here, first, V7 is selected, and the angle ratio is set to (Δθ _zero-fh / 2) (S54). Next, the phase region is specified (S56), and the process proceeds to S58 or S60 depending on the result. For example, in the case of the phase region I, the process proceeds to S58, and in the case of the phase region II, the process proceeds to S60. In the case of S58, the maximum angle vector V _max-fh in the first predicted phase is selected as the second switching vector, and the angle ratio is Δθ _max-fh . Next, the minimum angle vector V _min-fh in the first predicted phase is selected as the third switching vector, and its angle ratio is Δθ _min-fh . In the case of S60, V _min-fh is selected as the second switching vector, and its angle ratio is Δθ _min-fh . Next, V _max-fh is selected as the third switching vector, and its angular ratio is Δθ _max-fh .

Next, V0 is selected as the fourth switching vector, and the angle ratio is set to (Δθ _zero-fh / 2) + (Δθ _zero-sh / 2) (S62). Then, the phase region is specified again (S64), and the process proceeds to S66 or S68 depending on the result. For example, in the case of the phase region I, the process proceeds to S68, and in the case of the phase region II, the process proceeds to S66. In the case of S66, the maximum angle vector V _max-sh in the second predicted phase is selected as the fifth switching vector, and the angle ratio is Δθ _max-sh . Next, the minimum angle vector V _min-sh in the second predicted phase is selected as the sixth switching vector, and its angle ratio is Δθ _min-sh . In the case of S68, V _min-sh is selected as the fifth switching vector, and its angle ratio is Δθ _min-sh . Next, V _max-sh is selected as the sixth switching vector, and its angular ratio is Δθ _max-sh . Finally, V7 is selected as the seventh switching vector, and its angle ratio is (Δθ _zero-fh / 2) (S70).

FIG. 38 is a diagram corresponding to FIG. 14 and shows switching vector selection and switching order for the phase region I. The switching phases between the switching vectors are θ _comp1 , θ _comp2 , θ _comp3 , θ _comp4. , Θ _comp5 , θ _comp6 . This switching phase can be determined as follows according to the angle ratio.

That is, V7 is used as the first zero vector, and θ _comp1 which is the first switching phase therefrom is (φ _V + θ _e ), the control angle period Δθ, and the angle ratio Δθ _{zero- Using fh} , it is set by equation (31).

Also, θ _comp2 and θ _comp3 have the minimum angle vector V1 and the maximum angle vector V3 in the first prediction phase, and therefore, the angle ratio Δθ _{min-fh of} V1 and the angle ratio Δθ _max-fh of V3 are And set by equation (32).

Thereafter, the zero vector V0 is used, and the next θ _comp4 is set by the equation (33) using the angle ratios Δθ _zero-fh and Δθ _zero-sh .

Also, θ _comp5 and θ _comp6 have the minimum angle vector V1 and the maximum angle vector V3 in the second prediction phase, and therefore, the angle ratio Δθ _{min-sh of} V1 and the angle ratio Δθ _max-sh of V3 are And set by equation (34). Zero vector V7 is used after θ _comp6 .

  The switching phase is similarly applied to the phase regions III and V as described with reference to FIG.

For phase regions II, IV, and VI, θ _comp1 is the same as equation (31), but θ _comp2 and θ _comp3 are set by equation (35).

Further, θ _comp4 for the phase regions II, IV, and VI is the same as that in the equation (33), but θ _comp5 and θ _comp6 are set by the equation (36).

  When the PWM signal is generated in this way, the process returns to FIG. 34 again, the interrupt angle that is the next interrupt phase is calculated, and a series of processes in the control angle period Δθ is completed (S52).

  In the above description, the synchronization number K = 6 has been described in the full cycle control. However, even if the synchronization number K is other than 6, the PWM signal can be generated using the angle ratio similarly. FIGS. 39 and 40 are examples of the synchronization number K = 9 in the full cycle control. In this case, the control angle period Δθ is 360 degrees / 9 = 40 degrees. FIGS. 39 and 40 are diagrams corresponding to FIGS. 31 and 32 in which the number of synchronizations K = 6, and show how the interrupt phase is set at intervals of 40 degrees. Note that the predicted phase is 1.5 times 40 degrees, and the phase advanced 60 degrees from the interrupt phase.

In the above, each phase region is one region, and the phase between the minimum angle vector and the maximum angle vector is 60 degrees. Accordingly, the angle ratio is often set to 1/2, 1/4, etc., based on this 60 degrees. For example, the predicted phase may be 30 degrees from the minimum angle vector. In this case, it is necessary to accurately distinguish the equiphase from the minimum angle vector and the maximum angle vector of each phase region. Although the phase is set based on the electrical angle θ _e detected by the resolver 14, the resolution of the resolver 14 is not necessarily a multiple of 6. Therefore, it may occur that the equiphase cannot be accurately distinguished from the minimum angle vector and the maximum angle vector in each phase region.

This is a problem of the accuracy of the electrical angle sensor. In order to prevent this problem from affecting the calculation of the angle ratio, in the half cycle control, the phase region can be divided into two as shown in FIG. In FIG. 41, the phase region I is divided into a region IA of 30 degrees ± (1/4) Δθ in phase from the minimum angle vector and a region IB other than that. (1/4) Δθ is preferably (1/2 ⁿ ) Δθ, such as (1/8) Δθ. The power of 2 is used to reduce the amount of calculation by bit shift, and is set to a numerical value that can determine whether the prediction angle after 1.5 cycles, which is the prediction phase, is the region IA or the region IB. do it.

The calculation of the angle ratio in the region IB can use the equations (14), (18), (19), and (20) already described. The following formula (37) can be used to calculate the angle ratio in the region IA.

Here, for example, considering that the predicted phase is 30 degrees from the minimum angle vector, the angle ratios Δθ _min and Δθ _max are essentially the same value. As described above, depending on the accuracy of the resolver 14, Δθ _min and Δθ _max calculated based on the detected value of the resolver 14 may not be the same value. For such a case, in the region IA, the calculation angle ratio of one of the two basic voltage vectors is directly used as the angle ratio of the one basic voltage vector, and the other basic voltage is used. The angle ratio of the vector is preferably equal to the angle ratio of one of the basic voltage vectors. This is the reason why Δθ _max is not shown in Equation (37). It should be noted that the above equation can be expressed as Equation (38).

Equations (37) and (38) are for half-cycle control, but in the case of full-cycle control, it is necessary to distinguish the first predicted phase and the second predicted phase. In the case of the first prediction phase, the following equations (39) and (40) can be used, and in the case of the second prediction phase, equations (41) and (42) can be used.

Regarding the phase setting, in addition to being able to accurately distinguish the equiphase from the minimum angle vector and the maximum angle vector in each phase region, it is also possible to accurately distinguish the phase of the minimum angle vector and the phase of the maximum angle vector. is necessary. Therefore, as shown in FIG. 42, the phase region can be divided into three types. In FIG. 42, the phase region I is divided into a region IA of ± (1/4) Δθ in phase from the minimum angle vector and the maximum angle vector, and 30 ° ± (1/4) Δθ in phase from the minimum angle vector. Area I-B and other area I-C. (1/4) Δθ is preferably (1/2 ⁿ ) Δθ, such as (1/8) Δθ. The reason for the power of 2 is to reduce the amount of calculation by bit shift, and whether the prediction angle after 1.5 cycles, which is the prediction phase, is the region IA, the region IB, or the region IC. What is necessary is just to set to the numerical value which can be judged.

The calculation of the angle ratio in the region I-B and the region I-C can use the equations (14), (18), (19), (20), (37), and (38) already described. The following formula (43) can be used to calculate the angle ratio in the region IA.

Here, for example, considering that the prediction phase is at the minimum angle vector or the maximum angle vector, the angle ratio Δθ _min or Δθ _max is originally zero. As described above, depending on the accuracy of the resolver 14, Δθ _min and Δθ _max calculated based on the detected value of the resolver 14 may not be zero. For such a case, in region IA, the calculated angular ratio of one of the two basic voltage vectors is directly used as the angle ratio of one of the basic voltage vectors, and the angle of the other basic voltage vector The ratio should be zero. This is why Δθ _max is not shown in the equation (43). It should be noted that the above is mathematically expressed as equation (44).

  In the above description, the synchronization number K is assumed to be constant. For example, when the control mode is changed among the sine wave control mode, the overmodulation control mode, and the rectangular wave control mode, the synchronization number K is changed. Is performed. When the synchronization number K changes, the control angle period Δθ changes, so that the interrupt phase changes as it is. Since the interrupt phase is synchronized with the basic voltage vector, it is preferable that the interrupt phase does not change much even if the synchronization number K changes. Therefore, as shown in FIGS. 43 to 46, the synchronization number is changed when the interrupt phase matches the phase of the basic voltage vector determined in advance according to the synchronization number K. This function is executed by the synchronization number switching unit 64 of the synchronization control device 54.

  FIG. 43 shows an example of the case where the number of synchronizations K = 6 in half-cycle control, and the number of synchronizations K is switched during the interrupt phase indicated by a dotted circle. FIG. 44 shows an example of the case where the number of synchronizations K = 9 in the half cycle control, and the number of synchronizations K is switched during the interrupt phase indicated by a broken circle. As shown here, in the half-cycle control, the synchronization number K is switched in an interrupt phase that coincides with one of the six basic voltage vectors.

  FIG. 45 shows an example in which the synchronization number K = 6 for full-cycle control, and the synchronization number K is switched during the interruption phase indicated by a dotted circle. FIG. 46 shows an example in which the synchronization number K = 9 for full-cycle control, and the synchronization number K is switched during the interrupt phase indicated by a dotted circle. As shown here, in the full cycle control, among the six basic voltage vectors, only the U-phase arm is turned on and the lower element is turned off V1, and only the V-phase arm is turned on and the lower element is turned off. The switching of the synchronization number K is performed in an interrupt phase corresponding to any one phase of V4 in which only the V2 and W-phase arms are turned on and V4 is turned off.

In the above description, the first interrupt phase is described as being matched with any one of the six basic voltage vectors. However, depending on the synchronization number K, the subsequent interrupt phases for each control angle period Δθ are not necessarily six basic. It happens that it does not match any one of the voltage vectors. Even in this case, for example, in the half-cycle control, the predicted phase after 1.5 Δθ needs to be (θ _e + φ _V ) = 90 degrees. This is because the peak of the fundamental wave signal of the inverter circuit 20 is desired to be a predicted phase. Therefore, a condition is set for setting the interrupt phase depending on the synchronization number K.

FIG. 47 shows a case where the number of synchronizations K = 9 in the half cycle control. FIG. 48 shows a case where the number of synchronizations = 15 in the half cycle control. In these, as the number of synchronization timings are alternately distinguished from 1 and 2, they are indicated by square marks and triangle marks, respectively. The interrupt phase for the predicted phase after 1.5 Δθ to be (θ _e + φ _V ) = 90 degrees is one where the phase of 1.5Δθ is delayed from the predicted phase, so both FIG. 47 and FIG. This is when the synchronization number timing is 2. As described above, since the condition is added to the interrupt phase, the basic voltage vector to be matched with the first interrupt phase varies depending on the synchronization number K.

  The rotating electrical machine control system according to the present invention can be used for a system that controls a rotating electrical machine in a sine wave control mode, an overmodulation control mode, and a rectangular wave modulation mode.

  DESCRIPTION OF SYMBOLS 10 Rotating electrical machine control system, 12 Rotating electrical machine, 14 Resolver, 16, 18 Current detection part, 20 Inverter circuit, 22 U-phase arm, 24 V-phase arm, 26 W-phase arm, 30 Operation target command value, 32 Current command generation part 36 subtraction unit 38 current control unit 42 linear correction unit 44 αβ coordinate conversion unit 48 storage unit 50 control block 52 PWM control device 54 synchronization control device 56 PWM signal generation device 60 synchronization number setting unit 62 interrupt phase synchronization unit, 64 synchronization number switching unit, 66 voltage command vector generation unit, 68 region determination unit, 70 angle ratio calculation unit, 72 switching phase calculation unit, 80 angle / speed conversion unit, 82 dq coordinate conversion unit , 90 hexagon, 92 inscribed circle, 94 sides, 96 envelope circle.

Claims (7)

A three-phase inverter circuit connected to the three-phase rotating electrical machine;
The eight switching patterns for on / off of the switching elements constituting each phase arm of the three-phase inverter circuit are associated with two kinds of zero vectors and six kinds of basic voltage vectors, and these six kinds of basic voltage vectors are mutually connected. A PWM signal for each phase arm is generated using a space vector method in which the six phase regions partitioned by sequentially arranging them on the phase plane with a phase difference of 60 degrees are phase regions I to phase regions VI. A control device;
The control device includes
With respect to each of the phase region I to the phase region VI, the switching order of the switching vectors for generating the PWM signal is defined as the switching switching order with the two basic voltage vectors and the two zero vectors defining one phase region as the switching vectors. A storage unit for storing in advance;
The control angle period Δθ as the period on the phase plane for generating the PWM signal is divided by 2 K of the phase of one phase of the voltage command vector in the three-phase inverter circuit by the synchronization number K as an integer in the case of full-cycle control. A control angle cycle setting unit that sets the angle to 2π / K, and in the case of half-cycle control, π / K that is 1/2 of that,
An interrupt phase synchronizer that synchronizes the interrupt phase with any one of six basic voltage vectors on the phase plane, with the phase of the starting point of the control angle period as the interrupt phase;
The phase current of the three-phase rotating electrical machine and the electrical angle of the rotor are acquired at the timing of the interrupt phase, the d-axis current value and the q-axis current value are calculated based on the acquired phase current, and the d-axis command current value and d The d-axis voltage command value is obtained based on the deviation between the axis current value, the q-axis voltage command value is obtained based on the deviation between the q-axis command current value and the q-axis current value, and the d-axis voltage command value is obtained. A voltage command vector generation unit that generates a voltage command vector based on the q-axis voltage command value and the electrical angle of the rotor;
As a timing for generating the PWM signal using the generated voltage command vector, it is predicted that the voltage vector command exists in the predicted phase advanced by nΔθ from the interrupt phase, and which region of the phase region I to VI the predicted phase belongs to An area determination unit for determining
An angle ratio calculation unit that calculates a phase ratio to which the predicted phase belongs as a predicted phase area, and calculates a contribution ratio of the voltage command vector to a switching vector that defines the predicted phase area as an angle ratio with respect to the control angle period Δθ;
The switching switching order in the predicted phase region is read, and in accordance with the switching order, the switching phase between the switching vectors in the predicted phase region is sequentially calculated at the angle ratio calculated by the angle ratio calculating means, and the PMW signal in the phase region is calculated. A switching phase calculation unit to generate,
A rotating electrical machine control system comprising: In the rotating electrical machine control system according to claim 1,
The angle ratio calculation unit
The DC voltage applied to the three-phase inverter circuit when the V _dc, the radius of the hexagonal inscribed circle on the phase plane defined from the phase region I in phase region VI is V _dc / (2) ^1/2 Since the radius of the envelope circle connecting each vertex of the hexagon is (2/3) ^1/2 V _dc , the voltage at the origin of the hexagon on the phase plane defined by the phase region I to the phase region VI is When the magnitude of the voltage command vector is within the hexagon as the origin of the command vector, it corresponds to the sine wave control mode, the magnitude of the voltage command vector exceeds the hexagon, and the basic voltage vector component of the voltage command vector is an envelope. A control mode determination means for determining that it corresponds to the over-modulation control mode when it is inside the circle, and corresponds to the rectangular wave control mode when the basic voltage vector component of the voltage command vector extends outside the envelope circle;
When the control mode is the overmodulation control mode, the angle ratio of the two zero vectors of the switching vectors in the predicted phase region are both zero, and any of the calculated values of the angle ratio of the voltage command vector with respect to the two basic voltage vectors The calculated angle ratio of one of the basic voltage vectors is directly used as the angle ratio of the one of the basic voltage vectors, and the angle ratio of the other basic voltage vector is a value obtained by subtracting the angle ratio of one of the basic voltage vectors from 1. Modulation angle ratio calculating means;
When the control mode is the rectangular wave control mode, the angle ratios of the two zero vectors in the switching vector in the predicted phase region are both zero, and the angle ratios of the voltage command vectors with respect to the two basic voltage vectors are both ½. A rectangular wave angle ratio calculating means to:
A rotating electrical machine control system comprising: In the rotating electrical machine control system according to claim 2,
The area determination unit
A rotary electric machine characterized in that n = 1.5 is set for an expected phase in the case of half-cycle control, and two expected phases are set in the case of full-cycle control with n = 1.25 and n = 1.75. Control system. In the rotating electrical machine control system according to claim 3,
The angle ratio calculation unit
When the predicted phase is the same phase from the two basic voltage vectors constituting the predicted phase region, the calculated angular ratio of one of the two basic voltage vectors is directly used as one of the basic voltage vectors. And the angular ratio of the other basic voltage vector is equal to the angular ratio of the one basic voltage vector. In the rotating electrical machine control system according to claim 4,
The angle ratio calculation unit
When the predicted phase is within a predetermined phase range from one of the two basic voltage vectors constituting the predicted phase region, one of the basic voltage vectors is directly calculated. The angle ratio of the other basic voltage vector is zero, and when the other basic voltage vector is within a predetermined phase range, the calculated angle ratio of the other basic voltage vector is directly used as the other basic voltage. A rotating electrical machine control system, characterized in that the angle ratio of a vector is one and the angle ratio of one basic voltage vector is zero. In the rotating electrical machine control system according to claim 1,
A rotating electrical machine control system comprising a synchronization number switching unit that changes a synchronization number when an interrupt phase coincides with a phase of a basic voltage vector determined in advance according to the synchronization number K. In the rotating electrical machine control system according to claim 6,
The storage unit
A linear correction value for correcting the d-axis voltage command value and the q-axis voltage command value is set for each synchronization number K so that the fundamental wave amplitude corresponding to the generated PWM signal matches the voltage command amplitude corresponding to the voltage command vector. Remember,
The voltage command vector generator
A rotating electrical machine control system that reads a linear correction value corresponding to a synchronization number K, corrects a d-axis voltage command value and a q-axis voltage command value, and generates a voltage command vector.

Images