Rotating electric machine control device

Abstract

To detect a rotational position of a resolver with accuracy by increasing opportunities for correcting an offset value of the resolver.SOLUTION: A device for controlling a rotary electric machine corrects an offset value of a resolver so as to reduce a difference between a power factor command based on a torque command and a control power factor based on a measurement value in a first predetermined region where an output torque and a rotation number of the rotary electric machine are relatively low. Thereby, since the offset value can be corrected when the rotary electric machine drives in the first predetermined region, opportunities for correcting the offset value of the resolver can be increased compared with a device that corrects the offset value only when a rotational speed of the resolver is constant, and thus, a rotational position of the resolver can be detected with accuracy.SELECTED DRAWING: Figure 3

Description

[Technical field]

[0001] The present invention relates to a control device for a rotating electric machine, and more particularly to a control device for a rotating electric machine equipped with a resolver. [Background technology]

[0002] Conventionally, as a control device for this type of rotating electric machine, one has been proposed that calculates a correction parameter for the resolver based on rotational position data acquired at a plurality of specified time points when the rotational speed of the resolver is constant (for example, see Patent Document 1). In this control device, the correction parameter is calculated so as to be applicable to rotational position data of any value acquired when the resolver is rotating at any rotational speed and rotational position data of any value acquired when the resolver is stopped, thereby enabling the rotational position of the resolver to be detected with higher accuracy. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2009-008536 A Summary of the Invention [Problem to be solved by the invention]

[0004] However, in the above-mentioned control device for a rotating electric machine, the correction parameters are calculated when the rotation speed of the resolver is constant, so in order to calculate the correction parameters accurately, it is necessary to keep the rotation speed of the resolver constant with high accuracy. For this reason, when the control device is mounted on an automobile or the like, the rotation speed of the resolver is seldom kept constant during normal use, and there are fewer opportunities to calculate the correction parameters accurately, which may result in the resolver rotation position not being detected with high accuracy.

[0005] A main object of the control device for a rotating electric machine of the present invention is to increase the opportunities for correcting the offset value of a resolver and to enable more accurate detection of the rotational position of the resolver. [Means for solving the problem]

[0006] The control device for a rotating electric machine according to the present invention employs the following means to achieve the above-mentioned main object.

[0007] The control device for a rotating electric machine of the present invention comprises: A control device for a rotating electric machine including a resolver attached to a rotating shaft of the rotating electric machine driven by an inverter, correcting an offset value of the resolver so that a difference between a power factor command based on a torque command and a control power factor based on a measurement value becomes small when an output torque and a rotation speed of the rotary electric machine are in a first predetermined region of a relatively low torque and a relatively low rotation speed. It is characterized by:

[0008] In the control device for a rotating electric machine of the present invention, when the output torque and rotational speed of the rotating electric machine are in a first predetermined region of relatively low torque and relatively low rotation, the offset value of the resolver is corrected so that the difference between the power factor command based on the torque command and the control power factor based on the measurement value becomes small. Therefore, compared to a case where the offset value of the resolver is corrected when the rotational speed of the resolver is constant, the opportunities for correcting the offset value of the resolver can be increased. As a result, the rotational position of the resolver can be detected with higher accuracy. Here, the power factor is the cosine (cos(voltage phase-current phase)) of the difference between the voltage phase and the current phase.

[0009] In the control device for a rotating electric machine of the present invention, when the output torque of the rotating electric machine and the rotation speed of the rotating electric machine are outside the first predetermined region but within a second predetermined region, the offset value of the resolver may be corrected so that a difference between a power factor command based on a torque command and a control power factor based on a measurement value is reduced by changing at least one of the input voltage of the inverter and the carrier frequency of the inverter. In this way, it is possible to further increase the opportunities for correcting the offset value of the resolver. In this case, when the output torque of the rotating electric machine and the rotation speed of the rotating electric machine are outside the first predetermined region but within the second predetermined region and the dead time is known, the input voltage of the inverter and the carrier frequency of the inverter may be set to a voltage and carrier frequency at which the dead time voltage becomes a standard dead time voltage, and the offset value of the resolver may be corrected. In this way, it is possible to reduce calculation errors and correct the offset value with higher accuracy. Here, the standard dead time voltage is the dead time voltage during the dead time calculated from the power factor command or the dead time voltage during the dead time when actually measured. Furthermore, when the output torque of the rotating electric machine and the rotation speed of the rotating electric machine are outside the first predetermined region but within the second predetermined region, at least one of the input voltage of the inverter and the carrier frequency of the inverter may be changed so as to reduce the dead-time voltage, and the offset value of the resolver may be corrected in this state. The reason for reducing the dead-time voltage in this manner is that the dead-time voltage is reduced to reduce the degree of variation in the dead-time voltage, and the smaller the degree of variation in the dead-time voltage, the higher the accuracy of the offset value correction. [Brief description of the drawings]

[0010] [Figure 1] 1 is a diagram showing an outline of the configuration of a drive device 20 incorporating a control device 30 for a rotating electric machine as one embodiment of the present invention. [Diagram 2] 4 is an explanatory diagram for explaining an offset value OF of a resolver 28. FIG. [Diagram 3]5 is a flowchart showing an example of an offset correction value learning process executed by a control device 30 of the drive device 20 of the embodiment. [Figure 4] 1 is an explanatory diagram showing an example of a first learnable area and a second learnable area. FIG. [Diagram 5] FIG. 2 is an explanatory diagram showing an example of a motor voltage, a dead-time voltage, etc., on the d-axis and q-axis. [Figure 6] 2 is a control block diagram showing an example of a functional block of a control device 30 when executing control for correcting an offset value of a resolver 28. FIG. [Figure 7] FIG. 4 is an explanatory diagram showing an example of a power factor command setting map; [Figure 8] FIG. 13 is an explanatory diagram showing an example of a power factor command setting map according to a modified example. [Figure 9] 1 is an explanatory diagram showing an example of the relationship between the magnitude of the motor rotation speed Nm and a calculation error. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] Next, a mode for carrying out the present invention will be described using examples. EXAMPLES

[0012] 1 is a diagram showing an outline of the configuration of a drive device 20 incorporating a control device 30 for a rotating electric machine as one embodiment of the present invention. The drive device 20 includes a DC power supply 22, an inverter 24, a motor 26, a resolver 28, and a control device 30.

[0013] The DC power supply 22 corresponds to a power storage device such as a battery, and may have a boost circuit that boosts and outputs the voltage of DC power from the battery.

[0014] The motor 26 corresponds to various three-phase AC motors such as a synchronous generator motor or an induction motor, and is driven by a three-phase current of uvw phases applied from the inverter 24. The inverter 24 is configured as a well-known inverter circuit.

[0015] The resolver 28 is configured as a known resolver having an elliptical rotor connected to the rotating shaft of the motor 26, and a stator 128 as a magnetic body incorporating an excitation coil to which an AC current is applied as an excitation signal from an oscillator circuit, and two output coils arranged electrically shifted by 90 degrees (so that the phase difference is 90 degrees). The resolver outputs a sine wave sine signal and a cosine wave cos signal detected by the two output coils as resolver signals.

[0016] The control device 30 is configured as a well-known microcomputer centered on a CPU (not shown), and includes a ROM storing programs and the like, a RAM for temporarily storing data, a flash memory, an input circuit, an output circuit, and the like, all of which are not shown in addition to the CPU. The control device 30 receives, via an input circuit, an inverter input voltage Vinv from a voltage sensor 23 attached to a power line connecting the DC power source 22 and the inverter 24, a v-phase current Iv from a current sensor 25v attached to a v-phase power line of the three-phase current applied from the inverter 24 to the motor 26, a w-phase current Iw from a current sensor 25w attached to a w-phase power line, a resolver signal from a resolver 28, a motor torque command, and the like. The input circuit for the resolver signal is configured as a well-known resolver input circuit including an AD converter that converts a sine signal or a cosine signal into a digital signal, and an RD converter that calculates a resolver angle φ from the sine signal and the cosine signal from the AD converter. The control device 20 also outputs a switching control signal for switching each switching element of the inverter 24, a voltage command Vinv* to the DC power supply 22, and the like through an output circuit. In the control device 20, the u-phase current Iu flowing through the u-phase power line of the three-phase AC is calculated by Iu+Iv+Iw=0 using the v-phase current Iv and the w-phase current Iw. In addition, the control device 20 calculates a corrected resolver angle φ using an offset value OF and an offset correction value F for the resolver angle φ, calculates the rotation speed Nm of the motor 26, and generates a switching control signal for switching each switching element of the inverter 24 using a torque command value T*. The offset value OF is a rotation angle error occurring in the d-axis and q-axis in terms of control due to manufacturing errors, installation errors, and the like, as shown in FIG. 2.

[0017] Next, a learning process of the offset correction value F for correcting the offset value OF of the resolver 28 by the control device 30 of the drive device 20 configured as described above will be described. Fig. 3 is a flowchart showing an example of the offset correction value learning process executed by the control device 30 of the drive device 20 of the embodiment. This process is executed repeatedly.

[0018] When the offset correction value learning process is executed, the control device 30 first inputs data required for the process, such as the torque command value T*, the motor speed Nm, the inverter input voltage Vinv from the voltage sensor 23, and the carrier frequency fc of the inverter 24 (step S100). In the embodiment, the torque command value T* is set by the control device 30 based on the motor speed Nm, the accelerator operation amount, and the like. The motor speed Nm is calculated based on the resolver angle φ and the like. The carrier frequency fc of the inverter 24 is input as a carrier frequency in pulse width modulation (PWM) used when the control device 30 generates a switching control signal for the switching element of the inverter 24.

[0019] Next, it is determined whether the input torque command value T* and motor speed Nm are within a first learnable range (step S110). The first learnable range is a range that is determined in advance by experiments or the like as a range in which the offset correction value F can be learned well. FIG. 5 is an explanatory diagram showing an example of a motor voltage, a dead-time voltage, etc. on the d-axis and q-axis. In the figure, in the second quadrant, the solid arrow closest to the d-axis is a vector of the motor voltage, the dashed-dotted arrow from the q-axis is a vector of the current, and the two solid arrows in the fourth quadrant are a standard dead-time voltage DTtyp and a maximum dead-time voltage DTmax. The standard dead-time voltage DTtyp is a dead-time voltage during a dead time calculated from the power factor command Pf* or a dead-time voltage during a dead time when actually measured. The maximum dead-time voltage DTmax is a maximum value in the variation of the dead-time voltage caused by the variation of the dead time. In the second quadrant, the DTtyp motor voltage is a vector obtained by subtracting the standard dead-time voltage DTtyp from the motor voltage, and the DTmax motor voltage is a vector obtained by subtracting the maximum dead-time voltage DTmax from the motor voltage. The calculation error of the voltage phase-current phase is a value (θtyp-θmax) obtained by subtracting the phase θmax of the DTmax motor voltage and current from the phase θtyp of the DTtyp motor voltage and current. The first learnable range can be said to be a range in which this calculation error (θtyp-θmax) is tolerable. FIG. 4 shows an example of the first learnable range and the second learnable range. In the figure, the unhatched white area is the first learnable range, and the hatched area is the second learnable range. As shown in the figure, the first learnable range is set to a relatively low torque and relatively low rotation range, and the second learnable range is set to a slightly higher torque side and a slightly higher rotation speed side of the first learnable range. The second learnable range will be described later.

[0020] When it is determined that the input torque command value T* and motor rotation speed Nm are within the first learnable range, the offset correction value F is learned (step S140), and this process ends. The offset correction value F is learned by a control block illustrated in FIG. 6. The control block in FIG. 6 is an example of a functional block of the control device 30 when executing control to correct the offset value OF of the resolver 28. In this control block, the control device 20 includes, as functional blocks, a power factor command calculation unit 32, a control power factor calculation unit 34, a subtractor 36, and a PI controller 38.

[0021] The power factor command calculation unit 32 calculates the power factor command Pf* using the torque command T*. In the embodiment, the calculation of the power factor command is performed by previously determining the relationship between the torque and the power factor and storing it as a power factor command setting map, and when the torque command T* is given, the torque command T* is applied to the map to derive the power factor obtained, and this is set as the power factor command Pf*. FIG. 6 shows an example of the power factor command setting map. In the power factor command setting map of FIG. 7, the power factor decreases as the torque increases. The power factor is the cosine (cos(voltage phase-current phase)) of the difference between the voltage phase and the current phase, and the power factor command Pf* is an ideal (target) power factor that should act on the motor 26. The power factor command setting map illustrated in FIG. 7 derives a power factor command excluding the dead time voltage, but may derive a power factor command including the dead time voltage. In this case, the power factor command setting map illustrated in FIG. 8 may be used. The power factor command setting map shown in FIG. 8 is a table showing the relationship between the rotation speed of the motor 26, the torque command T*, and the power factor command P The power factor command Pf* is determined in advance. As the rotation speed of the motor 26 increases, the power factor command Pf* increases, although slightly. Also, the relationship between the voltage Vh and the carrier frequency input to the inverter 24, the torque, and the power factor may be determined in advance and used as a power factor command setting map. In this case, the power factor increases, although slightly, as the voltage Vh increases, and the power factor increases, although slightly, as the carrier frequency of the inverter 24 increases.

[0022] The control power factor calculation unit 34 calculates the control power factor Pfc based on the measurement values. For example, the control power factor calculation unit 34 calculates the current phase based on the v-phase current Iv and the w-phase current Iw from the current sensors 25v and 25w, the resolver signal from the resolver 28, the offset value OF, and the like, calculates the voltage phase from the d-axis voltage Vd and the q-axis voltage Vq obtained by the current feedback control, and calculates the control power factor Pfc as the cosine (cos(voltage phase-current phase)) of the difference between the voltage phase and the current phase obtained by the calculation. Here, the calculation of the current phase and the voltage phase uses the corrected resolver angle φ corrected using the offset value OF and the offset correction value F. The voltage phase may be calculated by mapping the self-inductances Ld and Lq of the d-axis and q-axis inductances Ld and Lq in advance, and calculating it according to the following formula (1) using the self-inductances Ld and Lq and the d-axis current id and the q-axis current iq of the armature current obtained from the phase currents. In equation (1), R is the armature resistance, ω is the angular velocity, p is the differential operator, and φ is the armature flux linkage due to the permanent magnet.

[0023]

number

[0024] The subtractor 36 calculates the difference (Pf*-Pfc) between the power factor command Pf* and the control power factor Pfc.

[0025] As shown in the following equation (2), the PI controller 38 calculates and outputs the sum of the proportional term and the integral term for the difference (Pf*-Pfc) between the power factor command Pf* and the control power factor Pfc as an offset correction value F. In equation (2), kp is the gain of the proportional term, and ki is the gain of the integral term.

[0026] F=kp(Pf*-Pfc)+∫ki(Pf*-Pfc)dt (2)

[0027] The offset correction value F thus obtained is used to correct the offset value OF when calculating the resolver angle φ, and a switching control signal to be output to the inverter 24 is generated based on the corrected resolver angle φ.

[0028] Returning to the description of the offset correction value learning process in Fig. 3, when it is determined in step S110 that the input torque command value T* and motor speed Nm are not within the first learnable range, it is determined whether the torque command value T* and motor speed Nm are within the second learnable range (step S120). That is, it is determined whether they are outside the first learnable range but within the second learnable range. The region outside the first learnable range but within the second learnable range is a region in which the calculation error falls within an allowable range if the input voltage Vinv of the inverter 24 or the carrier frequency fc of the inverter 24 is changed.

[0029] FIG. 9 shows an example of the relationship between the magnitude of the motor rotation speed Nm and the calculation error. In the figure, the motor voltage, DTtyp motor voltage, and DTmax motor voltage shown by solid lines indicate the case when the motor rotation speed Nm is small (same as in FIG. 5), and the motor voltage, DTtyp motor voltage, and DTmax motor voltage shown by dashed lines indicate the case when the motor rotation speed Nm is large. It can be seen that the calculation error (θtyp-θmax) when the motor rotation speed Nm is large is smaller than when the motor rotation speed Nm is small. Therefore, since the calculation error (θtyp-θmax) is smaller in the region where the motor rotation speed Nm is larger than the first learnable region, the learning possible region can be expanded to a certain extent. In addition, since the q-axis component VqDT and the d-axis component VdDT of the dead-time voltage are shown by the following equations (3) and (4), it is also possible to reduce the variation in the dead-time voltage by adjusting the carrier frequency fc of the inverter 24 and the inverter input voltage Vinv. Specifically, it is sufficient to reduce the carrier frequency fc of the inverter 24 and the inverter input voltage Vinv. If the carrier frequency fc of the inverter 24 or the inverter input voltage Vinv is reduced, the dead-time voltage is reduced, and the variation is reduced. As a result, the calculation error (θtyp-θmax) is reduced. Therefore, the learning region can be expanded to a certain extent by reducing the carrier frequency fc of the inverter 24 or the inverter input voltage Vinv. From these points, the learning region is the second learnable range. If the dead time is known, the carrier frequency fc and the inverter input voltage Vinv at which the dead-time voltage becomes the standard dead-time voltage DTtyp may be obtained, and the region in which the carrier frequency fc and the inverter input voltage Vinv can be changed to the obtained carrier frequency fc and the inverter input voltage Vinv may be set as the second learnable range. In this case, there is no variation in the dead-time voltage, so the calculation error is 0. The second learnable range can be determined in advance by experiment, calculation, or the like.

[0030] VqDT=√3×DT×fc×Vinv×sin(current phase) (3) VdDT=√3×DT×fc×Vinv×cos(current phase) (4)

[0031] When it is determined in steps S110 and S120 that the voltage is outside the first learnable range but within the second learnable range, the carrier frequency fc of the inverter 24 and the inverter input voltage Vinv are changed to learnable values ​​(step S130). Then, in this state, the offset correction value F is learned (step S140), and this process ends. Note that the inverter input voltage Vinv can be changed by a voltage command from the control device 30 when the DC power supply 22 is configured to adjust the output voltage, for example, when the DC power supply 22 includes a boost converter.

[0032] In the control device 30 incorporated in the drive device 20 of the embodiment described above, when it is determined that the torque command value T* and the motor rotation speed Nm are within a first learnable range in which the offset correction value F can be learned well, the offset correction value F is learned. This makes it possible to increase the opportunities to calculate the offset correction value F of the resolver 28 compared to when the offset correction value F is calculated when the rotation speed of the resolver 28 is constant. As a result, it is possible to detect the rotation position (resolver angle φ) of the resolver 28 with higher accuracy. Moreover, when it is determined that the torque command value T* and the motor rotation speed Nm are outside the first learnable range but within the second learnable range, the carrier frequency fc of the inverter 24 and the inverter input voltage Vinv are changed to learnable values ​​to learn the offset correction value F. This makes it possible to increase the opportunities to calculate the offset correction value F of the resolver 28.

[0033] In the control device 30 incorporated in the driving device 20 of the embodiment, an offset correction value F is learned to correct the offset value OF so that the difference between the power factor command Pf* and the control power factor Pfc becomes small. However, the offset value OF may be learned so that the learned offset correction value F is set as the offset value OF.

[0034] The correspondence between the main elements of the embodiment and the main elements of the invention described in the section on means for solving the problem will be described below. In the embodiment, the inverter 24 corresponds to the "inverter", the motor 26 corresponds to the "rotating electric machine", the resolver 28 corresponds to the "resolver", and the control device 30 corresponds to the "control device".

[0035] The correspondence between the main elements of the Examples and the main elements of the invention described in the Summary of the Problem column does not limit the elements of the invention described in the Summary of the Problem column, since the Examples are examples for specifically explaining the mode for implementing the invention described in the Summary of the Problem column. In other words, the interpretation of the invention described in the Summary of the Problem column should be based on the description in that column, and the Examples are merely a specific example of the invention described in the Summary of the Problem column.

[0036] Although the form for carrying out the present invention has been described above using examples, the present invention is not limited to these examples in any way, and it goes without saying that the present invention can be carried out in various forms without departing from the scope of the gist of the present invention. [Industrial Applicability]

[0037] The present invention can be used in the manufacturing industry of control devices for rotating electrical machines. [Explanation of symbols]

[0038] 20 drive device, 22 DC power supply, 24 inverter, 25v, 25w current sensor, 26 motor, 28 resolver, 30 control device, 32 power factor command calculation unit, 34 control power factor calculation unit, 38 subtractor, 38 PI controller.

Claims

1. A control device for a rotating electric machine including a resolver attached to a rotating shaft of the rotating electric machine driven by an inverter, when the output torque and rotation speed of the rotating electric machine are in a first predetermined region that is determined in advance as a region in which learning of the offset value of the resolver can be performed well, the offset value of the resolver is corrected so that a difference between a power factor command based on a torque command and a control power factor based on a measurement value becomes small. A control device for a rotating electric machine comprising:

2. 2. The control device for a rotating electric machine according to claim 1, When the output torque of the rotating electric machine and the rotation speed of the rotating electric machine are outside the first predetermined region but within a second predetermined region that is determined in advance as a region in which the offset value of the resolver can be learned by reducing the carrier frequency of the inverter or the input voltage of the inverter, at least one of the input voltage of the inverter and the carrier frequency of the inverter is changed, and the offset value of the resolver is corrected so that a difference between a power factor command based on a torque command and a control power factor based on a measurement value becomes small. A control device for a rotating electric machine.

Images