Control device for a rotating electric machine and electric power steering device

Abstract

The present application provides a kind of control device of rotary electric machine and electric power steering device, in the absence of using inductance and rotor cross-link magnetic flux information, regardless of the increase or decrease of rotational angular velocity, can appropriately increase the increase of d-axis current in negative direction, appropriately carry out weak magnetic flux.This control device of rotary electric machine (10) in the case where q-axis current command value (Iqo) is positive, based on the deviation (ΔIq_erroff) of offset q-axis current command value (Iqoffo) obtained from q-axis current command value (Iqo) minus positive q-axis offset value (ΔIqoff) and q-axis current detection value (Iqs), make d-axis current command value (Ido) change.

Description

Technical Field

[0001] The present application relates to a control device for a rotating electric machine and an electric power steering device. Background Art

[0002] In permanent magnet synchronous motors, the interlinkage flux of the permanent magnets generates an induced voltage proportional to the rotor's angular velocity. During high-speed rotation, if the difference between the maximum applied voltage and the induced voltage decreases, the desired q-axis current cannot flow through the windings, resulting in a decrease in output torque. Therefore, flux-weakening control is typically implemented to reduce the induced voltage by increasing the d-axis current in the negative direction during high-speed rotation. This generates flux in the windings that weakens the interlinkage flux of the permanent magnets.

[0003] There are various methods for flux weakening control. Patent Document 1 discloses a configuration in which proportional control or integral control is performed based on a deviation between a q-axis current command value and a q-axis current detection value to increase or decrease the d-axis current, thereby performing flux weakening control.

[0004] Prior art literature

[0005] Patent Literature

[0006] Patent Document 1: Japanese Patent No. 3559258 Summary of the Invention

[0007] Technical problem to be solved by the invention

[0008] In flux-weakening control, feedforward setting of optimal d-axis and q-axis current command values ​​requires high-precision information on the inductance and rotor interlinkage flux, which contribute to the generation of the induced voltage. However, without high-precision information on the inductance and rotor interlinkage flux, optimal d-axis and q-axis current command values ​​cannot be set feedforward. Furthermore, if the inductance and rotor interlinkage flux fluctuate over time or due to temperature characteristics, the accuracy of setting the d-axis and q-axis current command values ​​deteriorates.

[0009] On the other hand, the technology of Patent Document 1 increases or decreases the d-axis current command value through proportional or integral control based on the q-axis current deviation between the q-axis current command value and the q-axis current detection value. Therefore, the d-axis and q-axis current command values ​​are calculated without using information about the inductance and rotor interlinkage flux. However, as the angular velocity increases, the induced voltage increases, and the q-axis current is limited to a voltage limit ellipse based on the induced voltage. When the angular velocity decreases, the q-axis current deviation increases, causing the d-axis current to increase in the negative direction, enabling a higher degree of flux weakening. On the other hand, as the angular velocity decreases, the induced voltage decreases, and the q-axis current is no longer limited by the voltage limit ellipse. When the q-axis current deviation disappears, the q-axis current deviation becomes zero. Therefore, the increase in the d-axis current in the negative direction cannot be reduced, and proper flux weakening cannot be achieved. Furthermore, in this state, the d-axis current command value may fluctuate in unexpected directions due to, for example, noise components in the q-axis current.

[0010] Therefore, the purpose of the present application is to provide a control device for a rotating electric machine and an electric power steering device, which can appropriately increase or decrease the increase in the d-axis current in the negative direction and appropriately perform flux weakening regardless of the increase or decrease in the rotational angular velocity without using information on the inductance and the cross-linked magnetic flux of the rotor.

[0011] Technical means for solving technical problems

[0012] A control device for a rotating electric machine according to the present application controls a rotating electric machine having a stator with multi-phase windings and a rotor with permanent magnets via a power converter. The control device for a rotating electric machine includes:

[0013] a current detecting unit configured to detect a current flowing through the multi-phase winding;

[0014] a current coordinate conversion unit that converts the current detection value into a d-axis current detection value and a q-axis current detection value on a dq-axis rotating coordinate system consisting of a d-axis defined in a direction of a magnetic pole position of the rotor and a q-axis defined in a direction 90 degrees ahead of the d-axis in electrical angle, based on a rotation angle of the rotor;

[0015] a current command value calculation unit that calculates a d-axis current command value and a q-axis current command value;

[0016] a voltage command value calculation unit that changes a d-axis voltage command value and a q-axis voltage command value so that the d-axis current detection value approaches the d-axis current command value and the q-axis current detection value approaches the q-axis current command value, and converts the d-axis voltage command value and the q-axis voltage command value into multi-phase voltage command values ​​based on the rotation angle; and

[0017] a switching control unit that turns on and off a plurality of switching elements included in the power converter based on the multi-phase voltage command values;

[0018] The current command value calculation unit changes the d-axis current command value based on a deviation between an offset q-axis current command value obtained by subtracting a positive q-axis offset value from the q-axis current command value and the q-axis current detection value when the q-axis current command value is positive.

[0019] The electric power steering device involved in this application includes:

[0020] Control devices for rotating electrical machines;

[0021] the power converter;

[0022] the rotating electrical machine; and

[0023] A driving force transmission mechanism transmits the driving force of the rotating electric machine to a steering device of a vehicle.

[0024] Effects of the Invention

[0025] According to the control device for a rotating electric machine and an electric power steering device of the present application, when the q-axis current is upper-limited by the voltage limit ellipse of the induced voltage, the negative increment of the d-axis current command value is increased or decreased so as to move the d-axis current command value to the intersection of the voltage limit ellipse and a line that is lower than the q-axis current command value by the q-axis offset value. In this case, since the q-axis current detection value is lower than the q-axis current command value by the q-axis offset value, the q-axis voltage command value can be kept constant at the upper limit value, thereby maximizing the voltage utilization rate of the voltage command value. Even if the q-axis offset value is greater than the amplitude of the noise component of the q-axis current detection value, the q-axis voltage command value can be kept constant at the upper limit value, thereby reducing abnormal sounds and noises in the rotating electric machine. Therefore, based on the deviation between the offset q-axis current command value obtained by subtracting the q-axis offset value from the q-axis current command value and the q-axis current detection value, the d-axis current command value is changed. As a result, without using information on the inductance and the interlinked magnetic flux of the rotor, it is possible to automatically follow the voltage limit ellipse that changes due to the rotational angular velocity, so that the d-axis current and the q-axis current change toward the intersection of the voltage limit ellipse and the straight line that is lower than the q-axis current command value by the q-axis offset value, thereby appropriately performing flux weakening. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a schematic configuration diagram of a rotating electrical machine, a power converter, and a control device according to the first embodiment.

[0027] Figure 2This is a schematic block diagram of the control device according to the first embodiment.

[0028] Figure 3 This is a hardware configuration diagram of the control device according to the first embodiment.

[0029] Figure 4 It is a diagram for explaining the flux-weakening control according to the first embodiment.

[0030] Figure 5 It is a diagram for explaining flux weakening control according to a comparative example.

[0031] Figure 6 It is a diagram for explaining flux weakening control according to a comparative example.

[0032] Figure 7 This is a block diagram of a d-axis current command value changing unit according to the first embodiment.

[0033] Figure 8 This is a diagram explaining the control operation when Iqo>0 according to the first embodiment.

[0034] Figure 9 This is a diagram explaining the control operation when Iqo>0 according to the first embodiment.

[0035] Figure 10 This is a diagram explaining the control operation when Iqo>0 according to the first embodiment.

[0036] Figure 11 This is a diagram for explaining the control operation when Iqo<0 according to the first embodiment.

[0037] Figure 12 This is a diagram for explaining the control operation when Iqo<0 according to the first embodiment.

[0038] Figure 13 This is a timing chart illustrating the control operation according to the first embodiment.

[0039] Figure 14 This is a block diagram for explaining the setting of the proportional gain according to the first embodiment.

[0040] Figure 15 This is a Bode plot of an open-loop transfer function from the q-axis current offset deviation to the q-axis current detection value according to the first embodiment.

[0041] Figure 16 1 is a Bode plot of an open-loop transfer function from the q-axis current offset deviation to the q-axis current detection value according to the comparative example.

[0042] Figure 17This is a diagram illustrating offset value setting data for setting the q-axis offset value according to the second embodiment.

[0043] Figure 18 This is a block diagram of a d-axis current command value changing unit according to the third embodiment.

[0044] Figure 19 This is a block diagram of a current command value calculation unit according to the fourth embodiment.

[0045] Figure 20 This is a block diagram of a q-axis current command value changing unit according to the fourth embodiment. DETAILED DESCRIPTION

[0046] 1. Implementation Method 1

[0047] A control device 10 for a rotating electrical machine according to Embodiment 1 (hereinafter simply referred to as the control device 10 ) will be described with reference to the drawings. Figure 1 This is a schematic diagram of the configuration of the rotating electric machine 1, power converter 4, and control device 10 according to this embodiment. In this embodiment, the rotating electric machine 1 serves as the driving force source for the electric power steering device 100, and the rotating electric machine 1, power converter 4, and control device 10 constitute the electric power steering device 100.

[0048] 1-1. Rotating electric machine 1

[0049] The rotating electrical machine 1 includes a stator and a rotor arranged radially inward of the stator. The stator is provided with a multi-phase winding (in this example, three-phase windings Cu, Cv, and Cw for the U-phase, V-phase, and W-phase). The stator is provided with permanent magnets, and the rotating electrical machine 1 is configured as a permanent magnet synchronous rotating electrical machine. It is configured as a surface magnet type in which permanent magnets are provided on the outer circumference of the rotor. Alternatively, it can be configured as an embedded magnet type in which permanent magnets are provided inside the rotor. The three-phase winding can be connected in either a star or a delta configuration.

[0050] The rotor includes a rotation sensor 2 for detecting its rotation angle. Rotation sensor 2 employs a resolver, encoder, MR sensor, or the like. The output signal of rotation sensor 2 is input to control device 10. As described later, a sensorless configuration can be employed in which the angle is estimated based on current information without rotation sensor 2.

[0051] 1-2. Power converter 4

[0052] An inverter is used as the power converter 4. Alternatively, a power converter other than an inverter, such as a matrix converter, may be used as the power converter 4.

[0053] Inverter 4 is provided with three series circuits (branches) corresponding to each of the three phases. These series circuits are connected in series with a positive-side switching element SP connected to the positive side of DC power supply 3 and a negative-side switching element SN connected to the negative side of DC power supply 3. The connection point between two switching elements in the series circuit of each phase is connected to the winding of the corresponding phase.

[0054] Specifically, in the U-phase series circuit, the U-phase positive-side switching element SPu and the U-phase negative-side switching element SNu are connected in series, with the connection point between the two switching elements connected to the U-phase winding Cu. In the V-phase series circuit, the V-phase positive-side switching element SPv and the V-phase negative-side switching element SNv are connected in series, with the connection point between the two switching elements connected to the V-phase winding Cv. In the W-phase series circuit, the W-phase positive-side switching element SPw and the W-phase negative-side switching element SNw are connected in series, with the connection point between the two switching elements connected to the W-phase winding Cw. A smoothing capacitor 5 is connected between the positive and negative sides of the DC power supply 3.

[0055] Switching elements include IGBTs (Insulated Gate Bipolar Transistors) with diodes connected in antiparallel, MOSFETs (Metal Oxide Semiconductor Field Effect Transistors), and bipolar transistors with diodes connected in antiparallel. The gate terminal of each switching element is connected to the control device 10 via a gate drive circuit or the like. Each switching element is turned on or off by switching signals GPu to GNw output from the control device 10.

[0056] The DC power supply 3 outputs a DC voltage Vdc to the inverter 4. In this embodiment, the DC voltage Vdc is set to 12V. The DC power supply 3 can be any device that outputs a DC voltage Vdc, such as a battery, a DC-DC converter, a diode rectifier, or a PWM rectifier. The DC power supply 3 is equipped with a voltage sensor for detecting the DC voltage Vdc. The output signal of the voltage sensor can be input to the control device 10. The control device 10 can use the detected DC voltage Vdc for control.

[0057] A current sensor 6 is provided for detecting the current flowing through each phase winding. The current sensor 6 is a current sensor such as a shunt resistor or a Hall element. The output signal of the current sensor 6 is input to the control device 10.

[0058] In this embodiment, current sensor 6 can also be provided in a series circuit of two switching elements in each phase. A U-phase resistor Ru, a V-phase resistor Rv, and a W-phase resistor Rw are connected in series to the negative side of the switching element SN on the negative side of each phase. The potential difference between the two ends of the three-phase resistors Ru, Rv, and Rw is detected by amplifiers 21, 22, and 23, and this potential difference is input to the control device 10.

[0059] Alternatively, the current sensor 6 may be provided on the wire connecting the series circuit of the two switching elements of each phase to the coil of each phase. Alternatively, the current sensor may be provided on the wire connecting the inverter 4 and the DC power supply 3, and the current of each phase winding may be detected using the well-known "busbar 1 shunt method."

[0060] 1-3. Electric Power Steering Device 100

[0061] The electric power steering device 100 includes a rotary electric machine control device 10 , an inverter 4 , a rotary electric machine 1 , and a driving force transmission mechanism 101 that transmits driving force of the rotary electric machine 1 to a steering device 102 of a vehicle.

[0062] The rotating shaft of the rotor of the rotating electric machine 1 is connected to a steering system 102 for wheels 103 via a driving force transmission mechanism 101. For example, the electric power steering system 100 includes a steering wheel 104, which is turned left or right by the driver; a shaft 105 connected to the steering wheel 104 and transmitting the steering torque of the steering wheel 104 to the steering system 102 of the wheels 103; a torque sensor 106 mounted on the shaft 105 and detecting the steering torque Ts of the steering wheel 104; and a driving force transmission mechanism 101, such as a worm gear mechanism, connecting the rotating shaft of the rotating electric machine 1 to the shaft 105. The output signal of the torque sensor 106 is input to the control device 10 (input circuit 92).

[0063] 1-4. Control device 10

[0064] The control device 10 controls the rotating electrical machine 1 via the inverter 4. Figure 2 As shown in FIG. 1 , the control device 10 includes a rotation detection unit 31, a current detection unit 32, a current coordinate conversion unit 33, a current command value calculation unit 34, a voltage command value calculation unit 35, and a switch control unit 36. Each function of the control device 10 is realized by a processing circuit provided by the control device 10. Specifically, the control device 10 is as follows. Figure 3 As shown, the processing circuit includes: an operation processing device 90 (computer) such as a CPU (Central Processing Unit); a storage device 91 for exchanging data with the operation processing device 90; an input circuit 92 for inputting external signals to the operation processing device 90; and an output circuit 93 for outputting signals from the operation processing device 90 to the outside.

[0065] The processing unit 90 may include an ASIC (Application Specific Integrated Circuit), an IC (Integrated Circuit), a DSP (Digital Signal Processor), an FPGA (Field Programmable Gate Array), various logic circuits, and various signal processing circuits. Furthermore, the processing unit 90 may include multiple processing units of the same or different types to share the execution of various processes. The storage device 91 may include a RAM (Random Access Memory) configured to read and write data from the processing unit 90, or a ROM (Read Only Memory) configured to read data from the processing unit 90. The input circuit 92 is connected to various sensors and switches, such as the rotation sensor 2, the current sensor 6, and the torque sensor 106, and includes an A / D converter, etc., which inputs the output signals of these sensors and switches to the processing unit 90. The output circuit 93 is connected to an electric load such as a gate drive circuit that drives the switching element to be turned on and off, and includes a drive circuit that outputs a control signal from the arithmetic processing device 90 to the electric load.

[0066] Furthermore, the functions of the control units 31 to 36 and the like included in the control device 10 are realized by the arithmetic processing unit 90 executing software (programs) stored in a storage device 91, such as a ROM, in cooperation with other hardware components of the control device 10, such as the storage device 91, the input circuit 92, and the output circuit 93. Furthermore, setting data such as the internal resolution and control gain used by the control units 31 to 36 and the like is stored as part of the software (programs) in the storage device 91, such as the ROM. The functions of the control device 10 will be described in detail below.

[0067] 1-4-1. Basic Control

[0068] <Rotation Detection Unit 31>

[0069] The rotation detection unit 31 detects the magnetic pole position θ (rotation angle θ of the rotor) and the rotational angular velocity ω of the rotor under the electrical angle. In the present embodiment, the rotation detection unit 31 detects the magnetic pole position θ (rotation angle θ) and the rotational angular velocity ω of the rotor based on the output signal of the rotation sensor 2. In the present embodiment, the magnetic pole position is set in the direction of the N pole of the permanent magnet provided on the rotor. The rotational angular velocity ω is calculated by differentiating the rotation angle θ. In addition, the rotation detection unit 31 can be configured to estimate the rotation angle (magnetic pole position) based on current information obtained by superimposing harmonic components on the current command value, etc., without using a rotation sensor (so-called sensorless method).

[0070] <Current Detection Unit 32>

[0071] The current detection unit 32 detects the currents Ius, Ivs, and Iws flowing through the three-phase windings based on the output signal of the current sensor 6. Based on the output signal of the current sensor 6, the current detection unit 32 detects the current Ius flowing through the U-phase winding, the current Ivs flowing through the V-phase winding, and the current Iws flowing through the W-phase winding. Alternatively, the current sensor 6 may be configured to detect the winding currents of two phases, and the winding current of the remaining phase may be calculated based on the detected values ​​of the winding currents of the two phases. For example, the current sensor 6 may detect the winding currents Ivs and Iws of the V-phase and W-phase windings, and the winding current Ius of the U-phase winding may be calculated as Ius = -Ivs - Iws.

[0072] <Current Coordinate Converter 33>

[0073] The current coordinate converter 33 converts the three-phase winding current detection values ​​Ius, Ivs, and Iws into the d-axis current detection value Ids and the q-axis current detection value Iqs based on the rotation angle θ. In this embodiment, the current coordinate converter 33 converts the three-phase winding current detection values ​​Ius, Ivs, and Iws into the d-axis current detection value Ids and the q-axis current detection value Iqs by performing three-phase-to-two-phase conversion and rotational coordinate conversion based on the rotation angle θ, as shown in the following equation.

[0074] [Mathematical formula 1]

[0075]

[0076] The d-axis is defined as the direction of the magnetic pole (N pole) of the magnet, and the q-axis is defined as the direction 90 degrees ahead of the d-axis in electrical angle.

[0077] <Voltage Command Value Calculation Unit 35>

[0078] The voltage command value calculation unit 35 includes a current control unit 351, a q-axis voltage limiter 352, and a voltage coordinate converter 353. The current control unit 351 varies the d-axis voltage command value Vdo and the q-axis voltage command value Vqo so that the d-axis current detection value Ids approaches the d-axis current command value Ido, and the q-axis current detection value Iqs approaches the q-axis current command value Iqo. The calculation of the d-axis and q-axis current command values ​​Ido and Iqo by the current command value calculation unit 34 will be described later. For example, the voltage command value calculation unit 35 performs proportional-integral control as shown in the following equation.

[0079] [Mathematical formula 2]

[0080]

[0081]

[0082] Here, Kd and Kq are proportional gains, Td and Tq are integration times, and s is the Laplace operator.

[0083] In addition, feedforward control can be performed to prevent the d-axis current and the q-axis current from interfering with each other. That is, "-ω×Lq×Iqo" can be added to the d-axis voltage command value Vdo to obtain the value of Added to the q-axis voltage command value Vqo. Lq is the q-axis inductance, Ld is the d-axis inductance, It is the interlinked magnetic flux between the magnetomotive force of the magnet and the winding.

[0084] The q-axis voltage limiter 352 limits the q-axis voltage command value Vqo based on the DC voltage Vdc and the d-axis voltage command value Vdo so that the three-phase voltage command values ​​Vuo, Vvo, and Vwo do not exceed the maximum applied voltage Vdc / km corresponding to the DC voltage Vdc. For example, as shown in the following equation, the q-axis voltage limiter 352 applies upper and lower limits to the q-axis voltage command value Vqo so that the q-axis voltage command value Vqo does not exceed the upper limit value VqlmtH and the lower limit value VqlmtL calculated based on the maximum applied voltage Vdc / km and the d-axis voltage command value Vdo. The resulting upper and lower limit values ​​are then calculated as the final q-axis voltage command value Vqo.

[0085] [Mathematical formula 3]

[0086]

[0087]

[0088]

[0089]

[0090] Here, Km is a coefficient corresponding to the voltage utilization factor, and is set according to the presence or absence of modulation such as superposition of the third harmonic as shown in the following equation.

[0091] [Formula 4]

[0092] 1) Without modulation,

[0093]

[0094] 2) In the case of modulation, ···(4)

[0095]

[0096] With this configuration, the d-axis voltage command value Vdo is preferentially varied along the voltage limit circle corresponding to the maximum applied voltage Vdc / Km, while the q-axis voltage command value Vqo is subordinately varied. This allows the d-axis voltage command value Vdo to be preferentially varied, and the d-axis current Id to be preferentially varied, corresponding to the configuration that preferentially varies the d-axis current command value Ido during flux-weakening control, as described later.

[0097] The q-axis voltage limiter 352 can perform low-pass filtering on the d-axis voltage command value Vdo used to limit the q-axis voltage command value Vqo. If the d-axis voltage command value Vdo vibrates, the q-axis voltage command value Vqo also vibrates. By using a low-pass filtered value for the d-axis voltage command value Vdo, the vibration of the q-axis voltage command value Vqo can be suppressed, thereby reducing vibration and noise in the rotating electrical machine.

[0098] The voltage coordinate converter 353 converts the d-axis and q-axis voltage command values ​​Vdo and Vqo into three-phase voltage command values ​​Vuo, Vvo, and Vwo based on the rotation angle θ. In this embodiment, the voltage coordinate converter 353 converts the d-axis and q-axis voltage command values ​​Vdo and Vqo into three-phase voltage command values ​​Vuo, Vvo, and Vwo by performing fixed coordinate conversion and two-phase to three-phase conversion based on the rotation angle θ, as shown in the following equation.

[0099] [Formula 5]

[0100]

[0101] Furthermore, in order to improve voltage utilization, the voltage coordinate conversion unit 353 may apply known modulation such as two-phase modulation or third harmonic superposition to the three-phase voltage command values ​​Vuo, Vvo, and Vwo.

[0102] <Switch Control Unit 36>

[0103] The switching control unit 36 ​​turns on and off the plurality of switching elements included in the inverter 4 based on the three-phase voltage command values ​​Vuo, Vvo, and Vwo. The switching control unit 36 ​​uses a known carrier comparison PWM or space vector PWM.

[0104] When using carrier comparison PWM, the switching control unit 36 ​​compares the carrier with each of the three-phase voltage command values ​​Vuo, Vvo, and Vwo and switches multiple switching elements on and off based on the comparison results. The carrier is a triangular wave that oscillates with an amplitude of half the DC voltage (Vdc / 2) around zero during the PWM period Tc. For each phase, when the carrier is lower than the voltage command value, the switching control unit 36 ​​turns on the switching signal GP for the positive-side switching element, turning the positive-side switching element on. When the carrier CA exceeds the voltage command value, the switching control unit 36 ​​turns off the switching signal GP for the positive-side switching element, turning the positive-side switching element off. Conversely, for each phase, when the carrier is lower than the voltage command value, the switching control unit 36 ​​turns off the switching signal GN for the negative-side switching element, turning the negative-side switching element off. When the carrier CA exceeds the voltage command value, the switching control unit 36 ​​turns on the switching signal GN for the negative-side switching element, turning the negative-side switching element on. In addition, for each phase, a short-circuit prevention period (dead time) in which both the positive and negative switching elements are turned off may be provided between the on-period of the positive switching element and the on-period of the negative switching element.

[0105] When using space vector PWM, the switching control unit 36 ​​generates a voltage command vector based on the three-phase voltage command values ​​Vuo, Vvo, and Vwo, determines the output time distribution of the seven basic voltage vectors in the PWM cycle based on the voltage command vector, and generates switching signals that turn each switching element on and off in the PWM cycle based on the output time distribution of the seven basic voltage vectors.

[0106] 1-4-2. Current Command Value Calculation Unit 34

[0107] The current command value calculation unit 34 calculates the d-axis current command value Ido and the q-axis current command value Iqo. In this embodiment, the voltage command value calculation unit 34 includes a basic current command value calculation unit 341, a d-axis current command value variation unit 342, a d-axis current command value restriction unit 343, and a q-axis current command value restriction unit 344.

[0108] 1-4-2-1. Basic Current Command Value Calculation Unit 341

[0109] The basic current command value calculation unit 341 calculates the basic current command value Idob for the d-axis and the basic current command value Iqob for the q-axis. In this embodiment, the basic current command value calculation unit 341 detects the driver's steering torque Ts based on the output signal of the torque sensor 106. Then, as shown in the following equation, the basic current command value calculation unit 341 sets the basic current command value Iqob for the q-axis based on the steering torque Ts and sets the basic current command value Idob for the d-axis to 0. In other words, Id=0 control is performed. In Id=0 control, the basic current command value Idob for the d-axis is set to 0. Id=0 control is suitable for the surface magnet type rotating electric machine of this embodiment. In a surface magnet type rotating electric machine, the d-axis inductance Ld and the q-axis inductance Lq are approximately equal, and the torque varies proportionally with the q-axis current Iq.

[0110] [Formula 6]

[0111] I qob =K a T s

[0112] I dob =0···(6)

[0113] Here, Ka is a constant, but may vary depending on the steering torque Ts, the vehicle's running speed, etc. The q-axis basic current command value Iqob may be set based on a known compensation control corresponding to the steering situation.

[0114] In embedded-magnet rotating electrical machines, the d-axis and q-axis basic current command values ​​Idob and Iqob can be set using other control methods, such as maximum torque current control, instead of Id=0 control. In maximum torque current control, the d-axis and q-axis basic current command values ​​Idob and Iqob are calculated to maximize the generated torque for the same current.

[0115] 1-4-2-2. Flux-weakening control

[0116] <Problems with Flux-Weakening Control Without Using Inductance and Interlinkage Flux Information>

[0117] The voltage equation of the rotating electrical machine is as shown in equation (1).

[0118] [Formula 7]

[0119]

[0120] Here, Vd is the applied voltage of the d-axis, Vq is the applied voltage of the q-axis, Id is the current of the d-axis, Iq is the current of the q-axis, s is the Laplace operator, R is the winding resistance, is the linkage flux of the rotor magnet, Ld is the d-axis inductance, and Lq is the q-axis inductance.

[0121] The term multiplied by the rotational angular velocity ω in equation (7) is the term of the induced voltage generated in the winding as shown in the following equation. The induced voltage Vdi of the d-axis and the induced voltage Vqi of the q-axis increase as the rotational angular velocity ω increases.

[0122] [Formula 8]

[0123]

[0124] The induced voltage Vi is as shown in the following equation. When the induced voltage Vi approaches the maximum applicable voltage Vdc / Km, the amount of winding current that can be passed decreases, so the torque of the rotating electrical machine decreases.

[0125] [Formula 9]

[0126]

[0127] Therefore, from equation (9), it can be seen that generally, by increasing the d-axis current Id in the negative direction, the rotor linkage flux is generated to cancel the rotor linkage flux. The flux weakening control reduces the induced voltage Vi and increases the winding current.

[0128] Next, as shown in the following equation, the winding current is capped by the maximum current value Imax of the winding current that can flow. Within the current limit circle shown in equation (10), the d-axis current Id and the q-axis current Iq need to be controlled.

[0129] [Formula 10]

[0130] I d 2 +I q 2 ≤I max 2 ···(10)

[0131] Furthermore, as shown in the following equation, the d-axis current Id and the q-axis current Iq are limited so that the induced voltage Vi is within the range of the voltage limit ellipse that coincides with the maximum applicable voltage Vdc / Km.

[0132] [Mathematical formula 11]

[0133]

[0134] like Figure 4As shown, at a certain rotational angular velocity ω in the flux-weakening control region, the d-axis current Id and q-axis current Iq that produce the desired torque are at the intersection of the voltage limit ellipse and the q-axis current command value Iqo. When the q-axis current command value Iqo is limited by the current limit circle, the intersection of the voltage limit ellipse and the current limit circle is also at the intersection.

[0135] However, in order to set such optimal d-axis and q-axis current command values ​​in a feedforward manner as in the conventional art, the d-axis and q-axis inductances Ld and Lq related to the voltage limit ellipse and the rotor linkage flux are required. However, when the inductance Ld, Lq and the rotor cross-linked flux cannot be obtained, In the case of high-precision information, it is impossible to set the optimal d-axis and q-axis current command values ​​in a feedforward manner. When the current command values ​​for the d-axis and q-axis fluctuate due to changes over time or temperature characteristics, the setting accuracy of the current command values ​​for the d-axis and q-axis deteriorates.

[0136] On the other hand, in the technology of Patent Document 1, the d-axis current command value is increased or decreased by proportional control or integral control based on the q-axis current deviation ΔIq_err between the q-axis current command value and the q-axis current detection value. In the technology of Patent Document 1, if integral control is performed, Figure 5 As shown in FIG, it is considered that the d-axis current command value Ido can be changed along the voltage limit ellipse to the intersection of the voltage limit ellipse and the q-axis current command value Iqo before the q-axis current deviation ΔIq_err becomes zero. However, this is considered to be because the voltage limit ellipse narrows due to an increase in the rotational angular velocity ω, or the q-axis current command value Iqo increases due to an increase in the target torque, etc., so the d-axis current command value Ido and the q-axis current command value Iqo are located outside the voltage limit ellipse. When the d-axis current and the q-axis current are limited by the voltage limit ellipse, they can function better. On the other hand, as Figure 6 As shown, when the d-axis current command value Ido and the q-axis current command value Ido are located inside the voltage limit ellipse due to the expansion of the voltage limit ellipse caused by a decrease in the rotational angular velocity, or the decrease in the q-axis current command value Iqo caused by a decrease in the target torque, the d-axis current and the q-axis current are not restricted by the voltage limit ellipse. Therefore, the q-axis current deviation ΔIq_err remains at 0, and the d-axis current and the q-axis current do not change toward the intersection of the voltage limit ellipse and the q-axis current command value Iqo, but instead change in an unexpected direction due to the noise component and error component of the q-axis current.

[0137] In this way, by performing integral control based only on the q-axis current deviation ΔIq_err and changing the d-axis current command value Ido, the inductances Ld, Lq and the interlinkage flux are not used. Therefore, it is impossible to determine whether the current command values ​​of the d-axis and q-axis are outside or inside the voltage limit ellipse. In response to changes in operating conditions such as the increase or decrease of the rotational angular velocity ω and the increase or decrease of the target torque, it is impossible to set the optimal current command values ​​of the d-axis and q-axis for the weak flux control that maximizes the torque.

[0138] Therefore, a control device is sought that can achieve the desired effect in flux weakening control without using the inductance Ld, Lq and the interlinkage flux of the rotor. Under the condition of information, regardless of the increase or decrease of the rotation angular velocity ω and the target torque, the d-axis current Id and the q-axis current Iq can be controlled near the intersection of the voltage limit ellipse and the q-axis current command value Iqo, and the magnetic flux is weakened appropriately to increase the torque.

[0139] <d-axis current command value changing unit 342>

[0140] Figure 7 2 shows a block diagram of the d-axis current command value varying unit 342. When the q-axis current command value Iqo is positive, the d-axis current command value varying unit 342 varies the d-axis current command value Ido based on a deviation ΔIq_erroff (hereinafter referred to as q-axis current offset deviation ΔIq_erroff) between the offset q-axis current command value Iqoffo and the q-axis current detection value Iqs, obtained by subtracting a positive q-axis offset value ΔIqoff from the q-axis current command value Iqo.

[0141] [Mathematical formula 12]

[0142] 1)I qo >0,

[0143] I qoffo =I qo -ΔI qoff , ΔI qoff >0···(12)

[0144] ΔI q_erroff =I qoffo -I qs

[0145] Figure 7 The order of addition and subtraction operations of Iqo, Iqs, and ΔIqoff is listed as an example. Regardless of the order of addition and subtraction operations, when Iqo>0, ΔIq_erroff=Iqo-Iqs-ΔIqoff holds true. When Iqo<0 as described later, the structure in which ΔIq_erroff=Iqo-Iqs+ΔIqoff holds true is of course within the scope of the present invention.

[0146] With this configuration, when the q-axis current is upper-limited by the voltage limit ellipse, the negative increment of the d-axis current command value Ido is adjusted to move the d-axis current command value Ido to the intersection of the voltage limit ellipse and a line that is lower than the q-axis current command value Iqo by the q-axis offset value ΔIqoff. When the q-axis current command value Iqo is limited by the current limit circle, the d-axis current command value Ido is moved to the intersection of the voltage limit ellipse and a circle that is lower than the q-axis offset value ΔIqoff. At this point, since the q-axis current detection value Iqs is lower than the q-axis current command value Iqo by the q-axis offset value ΔIqoff, the q-axis voltage command value Vqo can be kept within the upper limit value VqlmtH based on the voltage limit circle, thereby maintaining the voltage utilization rate at its maximum value. Furthermore, when the q-axis offset value ΔIqoff is larger than the amplitude of the noise component of the q-axis current detection value Iqs, even if a noise component is generated, the q-axis voltage command value Vqo can be kept within the upper limit value VqlmtH, thereby reducing abnormal sounds and noises of the rotating electrical machine.

[0147] For example, Figure 8 As shown in FIG. 1 , in the voltage saturation state of the voltage command value, when the rotational angular velocity ω increases, the induced voltage increases and the q-axis current decreases. Therefore, in order to maintain the q-axis offset value ΔIqoff, the d-axis current command value Ido increases in the negative direction and moves to the intersection of the voltage limit ellipse, which becomes narrower due to the increase in the rotational angular velocity ω, and the straight line that is lower than the q-axis current command value Iqo by the q-axis offset value ΔIqoff. On the other hand, as Figure 9 As shown in FIG. 1 , in the voltage saturation state of the voltage command value, when the rotational angular velocity ω decreases, the induced voltage decreases and the q-axis current increases. Therefore, in order to maintain the q-axis offset value ΔIqoff, the d-axis current command value Ido increases in the positive direction and moves to the intersection of the voltage limit ellipse, which expands due to the decrease in the rotational angular velocity ω, and the straight line that is lower than the q-axis current command value Iqo by the q-axis offset value ΔIqoff. Therefore, the inductances Ld and Lq and the rotor linkage flux are not used. By only performing feedback control with a q-axis offset value ΔIqoff, the voltage limiting ellipse that changes due to the rotational angular velocity ω can be automatically followed, and the d-axis current Id and the q-axis current Iq can be controlled to the intersection of the voltage limiting ellipse and the straight line that is lower than the q-axis current command value Iqo by the q-axis offset value ΔIqoff, thereby appropriately performing flux weakening and increasing torque.

[0148] On the other hand, Figure 10As shown, when the value obtained by subtracting the q-axis offset value ΔIqoff from the q-axis current command value Iqo is inside the voltage limit ellipse, the q-axis current follows the q-axis current command value Iqo. Therefore, in order to generate the q-axis offset value ΔIqoff, the d-axis current command value Ido increases in the positive direction, and the increase amount of the d-axis current command value Ido in the negative direction is 0, and field weakening control is not performed.

[0149] When the q-axis current command value Iqo is positive, the d-axis current command value change unit 342 decreases the d-axis current command value Ido when the detected q-axis current value Iqs is lower than the offset q-axis current command value Iqoffo (ΔIq_erroff > 0), and increases the d-axis current command value Ido when the detected q-axis current value Iqs exceeds the offset q-axis current command value Iqoffo (ΔIq_erroff < 0).

[0150] The d-axis current command value change unit 342 calculates the d-axis current command value change amount ΔIdo based on the q-axis current offset deviation ΔIq_erroff, adds the d-axis current command value change amount ΔIdo to the d-axis basic current command value Idob, and calculates the d-axis current command value Ido. For example, the d-axis current command value change unit 342 performs proportional integral control based on the q-axis current offset deviation ΔIq_erroff and calculates the d-axis current command value change amount ΔIdo.

[0151] [Mathematical formula 13]

[0152] 1) When I qo > 0,

[0153]

[0154] I do = I dob + ΔI do ···(13)

[0155] Here, Kpid is the proportional gain for calculating the d-axis current command value set to a positive value, Tiid is the integral time for calculating the d-axis current command value, and s is the Laplace operator. The setting method for the proportional gain Kpid and the integral time Tiid for calculating the d-axis current command value will be described later. In addition, any feedback control such as integral control, proportional control, or proportional integral derivative control can be used instead of proportional integral control. <00,00460>

[0156] <When the q-axis current command value Iqo < 0>

[0157] On the other hand, when the q-axis current command value Iqo is negative, the d-axis current command value changing unit 342 changes the d-axis current command value Ido based on the deviation ΔIq_erroff (q-axis current offset deviation) between the offset q-axis current command value Iqoffo obtained by adding a positive q-axis offset value ΔIqoff to the q-axis current command value Iqo and the q-axis current detection value Iqs.

[0158] [Mathematical formula 14]

[0159] 2)I q0 When <0,

[0160] I qoffo =I qo +ΔI qoff , ΔI qoff >0···(14)

[0161] ΔI q_erroff =I qoffo -I qs

[0162] According to this configuration, when the q-axis current is upper-limited by the voltage limit ellipse, the negative increase in the d-axis current command value Ido is adjusted to move the d-axis current command value Ido to the intersection of the voltage limit ellipse and a line obtained by increasing the q-axis offset value ΔIqoff from the q-axis current command value Iqo. When the q-axis current command value Iqo is limited by the current limit circle, the d-axis current command value Ido is moved to the intersection of the voltage limit ellipse and a circle obtained by increasing the q-axis offset value ΔIqoff from the current limit circle. In this case, since the q-axis current is higher than the q-axis current command value Iqo by the q-axis offset value ΔIqoff, the q-axis voltage command value Vqo can be kept constant at the lower limit value VqlmtL based on the voltage limit circle, thereby maintaining the voltage utilization rate at its maximum value. Furthermore, if the q-axis offset value ΔIqoff is greater than the amplitude of the noise component of the q-axis current detection value Iqs, even if noise is present, the q-axis voltage command value Vqo can be kept constant at the lower limit value VqlmtL, thereby reducing abnormal sounds and noises in the rotating electric machine.

[0163] For example, Figure 11 As shown in FIG. 1 , in the voltage saturation state of the voltage command value, when the rotational angular velocity ω increases, the induced voltage increases, and the negative q-axis current increases. Therefore, in order to maintain the q-axis offset value ΔIqoff, the d-axis current command value Ido increases in the negative direction and moves to the intersection of the voltage limit ellipse, which becomes narrower due to the increase in the rotational angular velocity ω, and the straight line that increases the q-axis offset value ΔIqoff from the q-axis current command value Iqo. On the other hand, as Figure 12As shown in FIG. 1 , in the voltage saturation state of the voltage command value, when the rotational angular velocity ω decreases, the induced voltage decreases, and the negative q-axis current decreases. Therefore, in order to maintain the q-axis offset value ΔIqoff, the d-axis current command value Ido increases in the positive direction and moves to the intersection of the voltage limit ellipse, which expands due to the decrease in the rotational angular velocity ω, and the straight line that increases the q-axis offset value ΔIqoff from the q-axis current command value Iqo. Therefore, the inductances Ld and Lq and the rotor linkage flux are not used. By only performing feedback control with a q-axis offset value ΔIqoff, the voltage limit ellipse that changes due to the rotational angular velocity ω can be automatically followed, and the d-axis current Id and the q-axis current Iq can be controlled to be the intersection of the voltage limit ellipse and a straight line with the q-axis offset value ΔIqoff added to the q-axis current command value Iqo, and flux weakening is appropriately performed to increase the absolute value of the negative torque.

[0164] The d-axis current command value changing unit 342 decreases the d-axis current command value Ido when the q-axis current command value Iqo is negative and the q-axis current detection value Iqs is higher than the offset q-axis current command value Iqoffo (ΔIq_erroff<0), and increases the d-axis current command value Ido when the q-axis current detection value Iqs is lower than the offset q-axis current command value Iqoffo (ΔIq_erroff>0).

[0165] When the q-axis current command value Iqo is a negative value, the d-axis current command value changing unit 342 performs proportional-integral control based on the q-axis current offset deviation ΔIq_erroff to calculate the d-axis current command value change amount ΔIdo as shown in the following equation.

[0166] [Mathematical formula 15]

[0167] 2)I qo When <0,

[0168]

[0169] I do =I dob +ΔI do ···(15)

[0170] <Upper and lower limits of the d-axis current command value Ido>

[0171] As shown in the following equation, d-axis current command value limiting unit 343 sets an upper limit on d-axis current command value Ido using an upper limit value IdlmtH and a lower limit value IdlmtL. Upper limit value IdlmtH is set to the base d-axis current command value Idob. Lower limit value IdlmtL is set to a negative limit value to prevent irreversible demagnetization of the rotor's permanent magnets.

[0172] [Formula 16]

[0173] I dlmtH =I dob , I dlmtL <0

[0174] 1)I do >I dlmtH In the case of

[0175] I do =I dlmtH

[0176] 2)I dlmtH >I do >I dlmtL In the case of

[0177] I do =I do ···(16)

[0178] 3)I dlmtL >I do In the case of

[0179] I do =I dlmtL

[0180] In a region where the flux-weakening control does not need to be performed, such as when the rotational angular velocity ω is equal to or lower than the base rotational angular velocity, Ido=Idob may be forcibly set.

[0181] <q-Axis Current Command Value Limiter 344 >

[0182] The q-axis current command value limiting unit 344 limits the q-axis current command value Iqo based on the maximum current value Imax and the d-axis current command value Ido so that the current supplied to the three-phase winding does not exceed the maximum current value Imax that can be supplied to the three-phase winding. For example, as shown in the following equation, the q-axis current command value limiting unit 344 applies upper and lower limits to the q-axis base current command value Iqob so that the q-axis base current command value Iqob does not exceed the upper limit value IqlmtH and lower limit value IqlmtL calculated based on the maximum current value Iamx and the d-axis current command value Ido. The q-axis current command value Iqo is then calculated using these upper and lower limits. This limiting process limits the d-axis and q-axis current command values ​​Iqo and Ido to within the current limit circle of the maximum current value Imax.

[0183] [Mathematical formula 17]

[0184]

[0185] 1)I qob >I qlmtH In the case of

[0186] I qo =I qlmtH

[0187] 2)I qlmtH >I qob >I qlmtL In the case of

[0188] I qo =I qob ···(17)

[0189] 3)I qlmtL >I qob In the case of

[0190] I qo =I qlmtl .

[0191] With this configuration, when the q-axis base current command value Iqob is limited to the current limit circle corresponding to the maximum current value Imax, the d-axis current command value Ido can be preferentially varied along the current limit circle corresponding to the maximum current value Imax, while the q-axis current command value Iqo can be varied as a subordinate function. This allows the d-axis current command value Ido to be preferentially varied during flux-weakening control, optimizing the flux-weakening effect.

[0192] <Control Action>

[0193] Reference Figure 13The control operation is described in the timing diagram of FIG. 1. When the q-axis basic current command value Iqob is a positive fixed value smaller than the maximum current value Imax, the rotational angular velocity ω gradually increases until time t03 and gradually decreases after time t03.

[0194] By time t01, the rotational angular velocity ω is low, causing the voltage limit ellipse to expand. The q-axis and d-axis current command values ​​Iqo and Ido lie within the voltage limit ellipse, and the q-axis current detection value Iqs follows the q-axis current command value Iqo. Consequently, the q-axis current offset deviation ΔIq_erroff becomes the negative of the q-axis offset value ΔIqoff, and the d-axis current command value change ΔIdo decreases to zero, setting the d-axis current command value Ido to zero, disabling flux-weakening control.

[0195] On the other hand, before time t01, as the rotational angular velocity ω gradually increases, the induced voltage gradually increases. In order to maintain the q-axis current detection value Iqs at the q-axis current command value Iqo, the q-axis voltage command value Vqo gradually increases through current feedback control.

[0196] At time t01, the induced voltage increases due to the increase in rotational angular velocity ω, narrowing the voltage limit ellipse. The q-axis and d-axis current command values ​​Iqo and Ido align with the voltage limit ellipse, and the q-axis voltage command value Vqo reaches the voltage limit circle. Furthermore, from time t01 to time t02, the q-axis and d-axis current command values ​​Iqo and Ido move outside the voltage limit ellipse, becoming constrained by the voltage limit ellipse. The q-axis current detection value Iqs gradually decreases from the q-axis current command value Iqo. However, the decrease in the q-axis current relative to the command value does not exceed the q-axis offset value ΔIqoff. The q-axis current offset deviation ΔIq_erroff remains negative, and the d-axis current command value change ΔIdo continues to increase to zero.

[0197] After time t02, if Figure 13 As shown by the dashed line in the middle, if the d-axis current command value change ΔIdo remains at zero, the decrease in the q-axis current relative to the command value exceeds the q-axis offset value ΔIqoff, and the q-axis current offset deviation ΔIq_erroff becomes positive. Therefore, in order to maintain the q-axis current offset deviation ΔIq_erroff at zero, the d-axis current command value change ΔIdo gradually decreases from zero, and the d-axis current command value Ido gradually decreases from zero. At this time, since the q-axis current detection value Iqs is lower than the q-axis current command value Iqo by the q-axis offset value ΔIqoff, the q-axis voltage command value Vqo can be kept within the upper limit value VqlmtH based on the voltage limit circle, thereby maintaining the voltage utilization rate at the maximum value. Thus, by increasing the d-axis current command value Ido in the negative direction, the flux weakening control is carried out. Figure 13 Compared with the case of ΔIdo=0 indicated by the dotted line, the q-axis current can be increased, and the torque can be increased. Therefore, when the rotational angular velocity ω increases, the inductance Ld, Lq and the rotor interlinkage flux are not used. By only performing feedback control with a q-axis offset value ΔIqoff, the system can automatically follow the gradually narrowing voltage limit ellipse, control the d-axis current Id and the q-axis current Iq to the intersection of the voltage limit ellipse and the straight line that is lower than the q-axis current command value Iqo by the q-axis offset value ΔIqoff, appropriately perform flux weakening, and increase torque.

[0198] After time t03, the rotation angular velocity ω gradually decreases, the induced voltage gradually increases, and the voltage limit ellipse gradually expands. Figure 13 As shown by the dotted line, if the d-axis current command value change ΔIdo remains at the value at time t03, the decrease in the q-axis current relative to the command value becomes less than the q-axis offset value ΔIqoff, and the q-axis current offset deviation ΔIq_erroff becomes negative. Therefore, to maintain the q-axis current offset deviation ΔIq_erroff at zero, the d-axis current command value change ΔIdo gradually increases, and the d-axis current command value Ido gradually increases. At this time, since the q-axis current detection value Iqs is lower than the q-axis current command value Iqo by the q-axis offset value ΔIqoff, the q-axis voltage command value Vqo is kept within the upper limit value VqlmtH based on the voltage limit circle, thereby maintaining the voltage utilization rate at its maximum. Consequently, even when the rotational angular velocity ω decreases, the inductances Ld and Lq and the rotor's interlinkage flux are not used. By only performing feedback control with a q-axis offset value ΔIqoff, the system can automatically follow the gradually expanding voltage limit ellipse, control the d-axis current Id and the q-axis current Iq to the intersection of the voltage limit ellipse and the straight line that is lower than the q-axis current command value Iqo by the q-axis offset value ΔIqoff, and appropriately perform flux weakening to increase torque.

[0199] At time t04, the d-axis current command value change ΔIdo reaches zero, setting the d-axis current command value Ido to zero, and ending flux-weakening control. From time t04 to time t05, as the rotational angular velocity ω increases, the induced voltage gradually decreases, the voltage limit ellipse gradually expands, and the decrease in the q-axis current relative to the command value gradually decreases. Then, at time t05, the q- and d-axis current command values ​​Iqo and Ido coincide with the voltage limit ellipse, and the q-axis voltage command value Vqo reaches the voltage limit circle. After time t05, the q- and d-axis current command values ​​Iqo and Ido lie within the voltage limit ellipse, and the q-axis current detection value Iqs tracks the q-axis current command value Iqo. As the induced voltage decreases, the q-axis voltage command value Vqo decreases.

[0200] <Setting the q-axis offset value ΔIqoff>

[0201] As described above, the d-axis current command value Ido is preferentially changed to adjust the q-axis current detection value Iqs to be lower or higher than the q-axis current command value Iqo by the q-axis offset value ΔIqoff. Therefore, through feedback control of the q-axis current, the q-axis voltage command value Vqo is kept within the upper limit value VqlmtH or lower limit value VqlmtL based on the voltage limit circle, maintaining the voltage utilization rate at its maximum. Meanwhile, since torque is proportional to the q-axis current, if the q-axis current detection value Iqs is lower or higher than the q-axis current command value Iqo by the q-axis offset value ΔIqoff, the absolute value of the torque is reduced by the q-axis offset value ΔIqoff. To minimize the reduction in the absolute value of torque, it is desirable to reduce the q-axis offset value ΔIqoff. However, if the q-axis current detection value Iqs exceeds or falls below the q-axis current command value Iqo due to fluctuations or noise components in the q-axis current detection value Iqs or the q-axis current command value Iqo, the q-axis voltage command value Vqo cannot always adhere to the upper limit value VqlmtH or lower limit value VqlmtL of the voltage limit circle, and the voltage utilization factor cannot always be maintained at the maximum value. Therefore, the q-axis offset value ΔIqoff is preferably set to a value greater than the fluctuations or noise components in the q-axis current detection value Iqs or the q-axis current command value Iqo.

[0202] If the absolute value of the q-axis current command value Iqo is small, increasing the q-axis offset value ΔIqoff causes the sign of the q-axis current detection value Iqs to differ from the sign of the q-axis current command value Iqo, reversing the sign of the torque. To prevent this, the d-axis current command value varying unit 342 sets the q-axis offset value ΔIqoff to a value smaller than the absolute value of the q-axis current command value Iqo, as shown in the following equation.

[0203] [Mathematical formula 18]

[0204] ΔI qoff <|I qo |<···(18)

[0205] As described later using equation (26), the response of the change ΔIqs in the q-axis current detection value per control cycle to the change ΔIds in the d-axis current detection value per control cycle is expressed as follows. Therefore, if the change ΔIqs in the q-axis current detection value is smaller than the q-axis offset value ΔIqoff, the q-axis voltage command value Vqo can be kept constant within the upper limit value VqlmtH or the lower limit value VqlmtL based on the voltage limit circle.

[0206] [Mathematical formula 19]

[0207]

[0208] Here, the control period is the control period for calculating the d-axis current command value Ido. The rotational angular velocity ω can be assumed to be the maximum rotational angular velocity ωmax of the rotating motor. The differential inductance term Lq×s is ignored because it always changes depending on the operating conditions and has a minimal impact on the overall formula. In addition, the current feedback control responds quickly enough to changes in the d-axis current command value, so the change in the d-axis current detection value ΔIds can be replaced by the change in the d-axis current command value ΔIdodT per control period. Therefore, Equation (19) can be changed to the following equation.

[0209] [Mathematical formula 20]

[0210]

[0211] The amount of change ΔIdodT in the d-axis current command value for each control cycle can be known in advance. Therefore, the d-axis current command value changing unit 342 may set the q-axis offset value ΔIqoff so as to satisfy equation (20).

[0212] Furthermore, as shown in the following equation, the q-axis offset value ΔIqoff is preferably set to a value greater than the amplitude ΔIqns of the noise component superimposed on the q-axis current detection value Iqs. This ensures that the q-axis voltage command value Vqo remains within the upper limit value VqlmtH or lower limit value VqlmtL based on the voltage limit circle, even when noise components are superimposed on the q-axis current.

[0213] [Mathematical formula 21]

[0214] ΔI qns <ΔI qoff ···(twenty one)

[0215] The noise component of the q-axis current detection value Iqs causes fluctuations in the q-axis current offset deviation ΔIq_erroff, which in turn causes fluctuations in the d-axis current command value Ido. In particular, fluctuations in the proportional control term increase, necessitating consideration of amplification using the proportional gain Kpid. Fluctuations in the d-axis current command value Ido cause the q-axis current command value Iqo to fluctuate proportionally through the current limit circle.

[0216] Therefore, as shown in the following equation, the q-axis offset value ΔIqoff is preferably set to a value larger than the value obtained by multiplying the amplitude ΔIqns of the noise component superimposed on the q-axis current detection value Iqs by the proportional gain Kpid for calculating the d-axis current command value.

[0217] [Mathematical formula 22]

[0218] Kpid ΔI qns <ΔI qoff ···(twenty two)

[0219] To reduce the effects of q-axis current noise, the q-axis current detection value Iqs used to calculate the q-axis current offset error ΔIq_erroff can be subjected to low-pass filtering to reduce noise. The cutoff frequency of the low-pass filtering process can be set to ensure that the responsiveness of the calculation process for the d-axis current command value Ido is not degraded.

[0220] As described above, for example, the d-axis current command value changing unit 342 only needs to set an upper limit on the q-axis offset value ΔIqoff, which is set to satisfy one or more of the equations (20), (21), and (22), by the absolute value of the q-axis current command value Iqo as shown in equation (19).

[0221] <Setting the proportional gain Kpid for calculating the d-axis current command value>

[0222] The proportional gain Kpid in equations (13) and (15) can be set to a positive value, but a preferred setting method that takes responsiveness into consideration is described below. The d-axis current command value changing unit 342 changes the proportional gain Kpid used for calculating the d-axis current command value inversely proportional to the rotational angular velocity ω. In this embodiment, as shown in the following equation, the value obtained by dividing the target response angular frequency ωido by the rotational angular velocity ω is set as the proportional gain Kpid. The target response angular frequency ωido is the target response angular frequency of the feedback control system, which changes the d-axis current command value Ido according to the q-axis current offset deviation ΔIq_erroff, thereby reducing the absolute value of the q-axis current offset deviation ΔIq_erroff.

[0223] [Mathematical formula 23]

[0224]

[0225] The target response angular frequency ωido can be set to a value greater than R / Lq. If set in this way, the time constant for the q-axis current offset deviation ΔIq_erroff to converge is shorter than the circuit time constant Lq / R. Consequently, in the flux-weakening control region, the d-axis and q-axis current command values ​​Ido and Iqo can be changed relatively quickly relative to changes in the rotational angular velocity and required torque, thereby varying the output torque. When a rotating electric machine is used as an auxiliary driving force source for the electric power steering device 100, as in the present embodiment, for example, when the target response angular frequency ωido is set to a value between 250 [rad / s] and 1200 [rad / s], a good steering feel can sometimes be achieved.

[0226] The integration time Tiid in equations (13) and (15) can be set to, for example, the circuit time constant Lq / R. Furthermore, the integral gain is Kpid / Tiid. Therefore, according to equation (23), the integral gain also varies inversely with the rotational angular velocity ω. For example, the integral gain is set to the value obtained by multiplying the target response angular frequency ωido by the winding resistance R, divided by the rotational angular velocity ω and the inductance Lq.

[0227] If the formula of the q-axis voltage Vq is extracted from the voltage equation of formula (7), Vq is replaced by Vqo, and Id and Iq are replaced by Ids and Iqs, respectively, the following formula is obtained.

[0228] [Mathematical formula 24]

[0229]

[0230] If equation (24) is solved for the q-axis current detection value Iqs, the following equation is obtained.

[0231] [Mathematical formula 25]

[0232]

[0233] When the flux weakening control is executed, the q-axis voltage command value Vqo becomes consistent with the upper limit value VqlmtH or the lower limit value VqlmtL based on the voltage limit circle. Therefore, the control of the q-axis current detection value Iqs based on the operation of the q-axis voltage command value Vqo is not considered, but the control of the q-axis current detection value Iqs based on the operation of the d-axis current detection value Ids is considered. Therefore, if Vqo and VqlmtH in formula (25) are ignored, , then Equation (25) becomes as follows.

[0234] [Mathematical formula 26]

[0235]

[0236] According to equation (26), the transfer function Gp(s) from the d-axis current detection value Ids to the q-axis current detection value Iqs becomes as shown below.

[0237] [Mathematical formula 27]

[0238]

[0239] From this, it can be seen that the amount of change in the q-axis current detection value Iqs caused by the manipulation of the d-axis current detection value Ids increases in proportion to the rotational angular velocity ω.

[0240] If the transfer function Gp(s) is used, the control system is as follows Figure 14The transfer function Gc(s) from the q-axis current offset deviation ΔIq_erroff to the d-axis current command value Ido is expressed as follows based on equations (13) and (23). The integration time Tiid is set to Lq / R.

[0241] [Mathematical formula 28]

[0242]

[0243] The transfer function Gd(s) from the d-axis current command value Ido to the d-axis current detection value Ids is expressed by the following equation: Where ωids is the target response angular frequency of the d-axis current feedback control.

[0244] [Mathematical formula 29]

[0245]

[0246] It is preferred that the target response angular frequency ωids of the current feedback control of the d-axis be set sufficiently higher than the target response frequency ωido of the current command value of the d-axis. ωids uses the proportional gain Kd of the d-axis in equation (2) and becomes Kd / Ld. Therefore, the proportional gain Kd of the d-axis can be set to a value greater than ωido×Ld. For example, it is preferred that the proportional gain Kd of the d-axis be set to a value greater than 3×ωido×Ld, and it is more preferred that the proportional gain Kd of the d-axis be set to a value greater than 5×ωido×Ld. By setting the proportional gain Kd of the d-axis in this way, it can be considered that Gd(S)≈1.

[0247] [Mathematical formula 30]

[0248]

[0249] K d >ω ido L d

[0250] K d >3ω ido L d ···(30)

[0251] K d >5ω ido L d

[0252] As described above, the open-loop transfer function Gop(s) from the q-axis current offset deviation ΔIq_erroff to the q-axis current detection value Iqs is expressed by the following equation: Here, it is assumed that the d-axis inductance Ld and the q-axis inductance Lq are substantially equal, and Gd(s)≈1.

[0253] [Mathematical formula 31]

[0254]

[0255] Equation (31) is a simple integral characteristic and does not depend on the rotational angular velocity ω. When the Bode diagram is drawn, as Figure 15 In this way, the slope of the gain is a constant value of -20 dB / dec, and is 0 dB at an angular frequency = ωido.

[0256] As a result, as shown in the following equation, the closed-loop transfer function Gfb(s) of the change in the q-axis current detection value Iqs relative to the change in the q-axis basic current command value Iqob becomes a first-order delay with a time constant equal to the inverse of the target response angular frequency ωido. Thus, in flux-weakening control, by varying the proportional gain Kpid used to calculate the d-axis current command value inversely with the rotational angular velocity ω as shown in equation (23), the response of the change in the q-axis current detection value Iqs relative to the change in the q-axis basic current command value Iqob can be set to a first-order delay with a time constant equal to the inverse of the target response angular frequency ωido that is constant regardless of the rotational angular velocity ω. Thus, by setting the target response angular frequency ωido, the desired torque responsiveness can be achieved. Consequently, the steering feel of the electric power steering device 100 is improved.

[0257] [Mathematical formula 32]

[0258]

[0259] Unlike equation (23), the following describes a case where the proportional gain Kpid used to calculate the d-axis current command value is set to a fixed value, rather than changing according to the rotational angular velocity ω. In this case, in equation (31), the ω term in Gp(s) is not offset by the 1 / ω term in Gc(s) using the proportional gain Kpid. Therefore, as shown in the following equation, the open-loop transfer function Gop(s) exhibits a characteristic proportional to ω.

[0260] [Mathematical formula 33]

[0261]

[0262] Here, K is a constant. If you want to obtain the same characteristics as Equation (31) under ω = ωmd, set K = 1 / ωmd. Figure 16As shown in the Bode plots for ω = 0.5 × ωmd, ω = 1 × ωmd, and ω = 2 × ωmd, when the rotational angular velocity ω changes from ωmd, the response changes by a factor of ω / ωmd. When ω = 1 × ωmd, the response of the open-loop transfer function Gop(s) becomes ωido, thus achieving the desired response. However, when ω = 2 × ωmd, the response of Gop(s) is doubled. While this improves responsiveness, the feedback amount of the noise component included in the q-axis current detection value Iqs is doubled, potentially increasing abnormal noise from the rotating electric machine. On the other hand, when ω = 0.5 × ωmd, the response of Gop(s) is 0.5 times, deteriorating the responsiveness of the q-axis current and, consequently, the responsiveness of the torque. Consequently, the torque response fluctuates according to the rotational angular velocity ω, potentially deteriorating the steering feel of the electric power steering device 100.

[0263] 2. Implementation Method 2

[0264] A control device 10 according to Embodiment 2 will be described. Description of components identical to those in Embodiment 1 will be omitted. The basic configurations of the rotating electrical machine 1, power converter 4, and control device 10 according to this embodiment are identical to those in Embodiment 1, but the method for setting the q-axis offset value ΔIqoff differs from that in Embodiment 1.

[0265] As in embodiment 1 Figure 13 As shown, when the rotational angular velocity ω increases from time t02 to time t03, as indicated by the dotted line, the q-axis current detection value Ids decreases without limit while the d-axis current command value change ΔIdo remains at zero, while the q-axis current offset deviation ΔIq_erroff increases without limit. Consequently, the responsiveness of the d-axis current command value change ΔIdo calculated based on the q-axis current offset deviation ΔIq_erroff is independent of the increase in the rotational angular velocity ω and corresponds to the set value of the proportional gain Kpid, thus posing no particular problem.

[0266] On the other hand, when the rotational angular velocity ω decreases from time t03 to time t04, as shown by the dashed line, the q-axis current detection value Ids increases only to the q-axis current command value Iqo when the d-axis current command value change ΔIdo remains at the value at time t03, and the absolute value of the q-axis current offset deviation ΔIq_erroff is capped by the q-axis offset value ΔIqoff. Therefore, when the decrease in the rotational angular velocity ω is significant, the absolute value of the q-axis current offset deviation ΔIq_erroff is capped by the q-axis offset value ΔIqoff. Consequently, the responsiveness of the d-axis current command value change ΔIdo calculated based on the q-axis current offset deviation ΔIq_erroff becomes lower than the responsiveness corresponding to the set value of the proportional gain Kpid. Furthermore, the response frequency of the d-axis current command value variation ΔIdo to changes in the q-axis current detection value Iqs is lower than the response frequency of the q-axis voltage command value Vqo. Therefore, the increase in the d-axis current command value variation ΔIdo is delayed relative to the decrease in the induced voltage and the increase in the q-axis current detection value Iqs caused by a decrease in the rotational angular velocity ω, allowing the q-axis current detection value Iqs to rise more easily. On the other hand, setting the response frequency of the d-axis current command value variation ΔIdo too high degrades the stability of the control system, thus having a limit. Specifically, if the q-axis offset value ΔIqoff is set excessively low when the rotational angular velocity ω decreases, the responsiveness of the d-axis current command value variation ΔIdo deteriorates as the decrease in the rotational angular velocity ω increases. Decreasing the responsiveness of the d-axis current command value variation ΔIdo also deteriorates the responsiveness of the q-axis current and, consequently, the torque.

[0267] Therefore, in the present embodiment, the d-axis current command value changing unit 342 sets the q-axis offset value ΔIqoff to a larger value when the rotational angular velocity ω decreases than when the rotational angular velocity ω increases.

[0268] According to this configuration, when the rotational angular velocity ω decreases, the q-axis offset value ΔIqoff, which serves as the upper limit of the absolute value of the q-axis current offset deviation Iq_erroff, is increased. When the decrease in the rotational angular velocity ω is large, the deterioration in the responsiveness of the d-axis current command value change ΔIdo can be suppressed, thereby preventing deterioration in the responsiveness of the q-axis current and the responsiveness of the torque.

[0269] In this embodiment, the d-axis current command value varying unit 342 refers to the relationship between the change in rotational angular velocity per unit time Δω and the q-axis offset value ΔIqoff by referring to pre-set offset value setting data, and calculates the q-axis offset value ΔIqoff corresponding to the current change in rotational angular velocity Δω. As shown in the following equation, the d-axis current command value varying unit 342 subtracts the rotational angular velocity ω per unit time Δt from the current rotational angular velocity ω(t) (t-Δt) to calculate the change in rotational angular velocity per unit time Δω. For example, the unit time Δt is set to a natural integer multiple of the calculation period of the rotational angular velocity ω. The change in rotational angular velocity per unit time Δω corresponds to the rotational angular acceleration.

[0270] [Mathematical formula 34]

[0271] Δω=ω(t)-ω(t-Δt)···(34)

[0272] Offset value setting data such as Figure 17 Set as shown. Figure 17 In the example shown, when the angular velocity change Δω exceeds the threshold ωth, the q-axis offset value ΔIqoff is set to a low set value ΔIqoffL. When the angular velocity change Δω is less than the threshold ωth, the q-axis offset value ΔIqoff is set to a high set value ΔIqoffH, which is greater than the low set value ΔIqoffL. The threshold ωth is set to 0 or a negative value. The high set value ΔIqoffH and the low set value ΔIqoffL are fixed values. Therefore, even if the angular velocity change Δω fluctuates above or below the threshold ωth, q-axis current fluctuations, and thus torque fluctuations, can be suppressed.

[0273] The angular velocity ω is calculated by differentiating the rotation angle θ and is therefore susceptible to fluctuation. Therefore, regardless of the constant speed, the change in angular velocity Δω tends to fluctuate around zero. Setting the threshold ωth to a negative value allows the q-axis offset value ΔIqoff to be maintained at a low set value, IqoffL, even when the change in angular velocity Δω fluctuates around zero, thereby suppressing fluctuations in the d-axis and q-axis currents. Alternatively, a low-pass filtered angular velocity ω may be used.

[0274] 3. Implementation Method 3

[0275] The control device 10 according to the third embodiment will be described. Descriptions of the same components as those in the first embodiment will be omitted. The basic configurations of the rotating electrical machine 1, power converter 4, and control device 10 according to this embodiment are the same as those in the first embodiment, but the configuration of the d-axis current command value varying unit 342 differs partially from that in the first embodiment. Figure 18A block diagram of the d-axis current command value changing unit 342 according to the present embodiment is shown.

[0276] First, the concept of control design will be explained. Taking the formula for the q-axis voltage Vq from the voltage equation in equation (7), the current state is expressed as follows.

[0277] [Mathematical formula 35]

[0278]

[0279] In the execution state of weak flux control, when the current command value Iqo of the q-axis is positive, the voltage command value Vqo of the q-axis is upper-limited by the upper limit value VqlmtH based on the voltage limit circle, so the d-axis current only operates ΔId, and the state in which the voltage command value Vqo of the q-axis reaches the upper limit value VqlmtH based on the voltage limit circle is as shown in the following formula.

[0280] [Mathematical formula 36]

[0281]

[0282] If equation (35) is subtracted from equation (36), the manipulated variable ΔIdffo of the d-axis current for making the deviation between Vq and VqlmtH 0 is expressed as follows: When the q-axis current command value Iqo is negative, the lower limit value VqlmtL is also used for derivation.

[0283] [Mathematical formula 37]

[0284] 1)I qo >0,

[0285]

[0286] 2)I qo When <0,

[0287]

[0288] Thus, in order to make the q-axis voltage command value Vqo consistent with the upper limit value VqlmtH or lower limit value VqlmtL of the q-axis voltage command value, the d-axis current command value Ido can be corrected by the feedforward correction amount ΔIdffo calculated by equation (37). The proportional gain multiplied by the deviation changes inversely with the rotational angular velocity ω. Thus, in this embodiment, the d-axis current command value changing unit 342 corrects the d-axis current command value Ido based on the deviation between the q-axis voltage command value Vqo and the upper limit value VqlmtH or lower limit value VqlmtL of the q-axis voltage command value. This correction is performed in the execution area of ​​the weak magnetic flux control. The feedforward correction amount ΔIdffo can also be subjected to low-pass filtering to reduce the noise component.

[0289] In order to change the q-axis voltage command value Vqo to the upper limit value VqlmtH or the lower limit value VqlmtL, the d-axis current command value Ido may be changed in a feedforward manner to improve responsiveness.

[0290] The d-axis current command value changing unit 342 calculates the d-axis current command value Ido by adding the d-axis current command value change amount ΔIdo and the feedforward correction amount ΔIdffo to the d-axis basic current command value Idob as shown in the following equation.

[0291] [Mathematical formula 38]

[0292] I do =I dob +ΔI do +ΔI dffo ···(38)

[0293] The d-axis current command value limiting unit 343 performs upper and lower limit operations on the d-axis current command value Ido calculated by equation (38) using an upper limit value IdlmtH and a lower limit value IdlmtL, similarly to the first embodiment.

[0294] 4. Implementation Method 4

[0295] A control device 10 according to Embodiment 4 will be described. Descriptions of components identical to those in Embodiment 1 will be omitted. The basic configurations of the rotating electrical machine 1, power converter 4, and control device 10 according to this embodiment are the same as those in Embodiment 1. However, this embodiment differs from Embodiment 1 in that the current command value calculation unit 34 includes a q-axis current command value variation unit 345. Figure 19 A block diagram showing the current command value calculation unit 34 according to this embodiment is shown. Figure 20 A block diagram of the q-axis current command value changing unit 345 is shown.

[0296] The q-axis current command value changing unit 345 calculates a q-axis current command correction value ΔIqo based on the q-axis current deviation obtained by subtracting the q-axis current detection value Iqs from the q-axis current command value Iqo. The q-axis current command correction value ΔIqo is added to the pre-corrected q-axis current command value Iqobf to calculate the q-axis current command value Iqo. The q-axis current command value Iqo used in current feedback control is used for calculating the q-axis current deviation, but the value calculated in the previous calculation cycle is used. The q-axis current command value Iqo, which is limited by the current limit circle as shown in equation (17), is used for the pre-corrected q-axis current command value Iqobf.

[0297] Because the q-axis current detection value Iqs is reduced or increased by the q-axis offset value ΔIqoff relative to the q-axis current command value Iqo, the absolute value of the output torque decreases by an amount corresponding to the q-axis offset value ΔIqoff. Therefore, as described above, by increasing or decreasing the q-axis current command value Iqo by the amount by which the q-axis current detection value Iqs decreases or increases relative to the q-axis current command value Iqo, the absolute value of the output torque can be maintained. Since the q-axis current command correction value ΔIqo is calculated based on the q-axis current deviation, the q-axis current being limited by the voltage limit ellipse can be automatically detected and corrected without requiring any conditional evaluation.

[0298] The q-axis current command value changing unit 345 calculates the q-axis current deviation as the q-axis current command correction value ΔIqo by low-pass filtering the q-axis current deviation. For example, a primary delay filter may be used as the low-pass filter. Alternatively, other low-pass filters such as a moving average may be used. Furthermore, since the q-axis current detection value Iqs is a oscillating element, the q-axis current deviation may be calculated using the low-pass filtered q-axis current detection value Iqs, rather than low-pass filtering the q-axis current deviation.

[0299] [Mathematical formula 39]

[0300]

[0301] I qo =T qobf +ΔI qo

[0302] <Transfer Example>

[0303] The rotating electric machine 1 can be used as a driving force source for various devices other than the electric power steering device 100. For example, the rotating electric machine 1 can be used as a driving force source for wheels.

[0304] The stator may also be provided with a multi-phase winding (for example, two-phase or four-phase winding) other than three-phase winding.

[0305] The stator may also be provided with multiple groups (for example, two groups) of three-phase windings, and respective parts of the power converter and the control device are provided corresponding to each group of three-phase windings.

[0306] Although this application describes various exemplary embodiments and examples, the various features, methods, and functions described in one or more embodiments are not limited to the application of specific embodiments and can be applied to the embodiments individually or in various combinations. Therefore, it can be considered that countless variations not illustrated are also included in the technical scope disclosed in this application specification. For example, it is set to include the case where at least one component is deformed, added, or omitted, and the case where at least one component is extracted and combined with the components of other embodiments.

[0307] Label Description

[0308] 1 Rotating electric machine, 4 Power converter, 10 Rotating electric machine control device, 31 Rotation detection unit, 32 Current detection unit, 33 Current coordinate conversion unit, 34 Current command value calculation unit, 35 Voltage command value calculation unit, 36 Switch control unit, 100 Electric power steering device, 101 Driving force transmission mechanism, Ido D-axis current command value, Idob D-axis basic current command value, Ids D-axis current detection value, Imax Maximum current value, Iqo Q-axis current command value, Iqob Q-axis basic current command value, Iqoffo Offset Q-axis current command value, Iqs Q-axis current detection value, Vdc DC voltage, Vdo D-axis voltage command value, Vqo Q-axis voltage command value, ΔIdo D-axis current command value change, ωido Target response angular frequency

Claims

1. A control device for a rotating electrical machine, which controls a rotating electrical machine having a stator provided with a multi-phase winding and a rotor provided with a magnet via a power converter, the control device for the rotating electrical machine comprising: a current detecting unit configured to detect a current flowing through the multi-phase winding; a current coordinate conversion unit that converts the current detection value into a d-axis current detection value and a q-axis current detection value on a dq-axis rotating coordinate system consisting of a d-axis defined in a direction of the magnetic pole position of the rotor and a q-axis defined in a direction 90 degrees ahead of the d-axis in electrical angle, based on the rotation angle of the rotor; a current command value calculation unit that calculates a d-axis current command value and a q-axis current command value; a voltage command value calculation unit that changes a d-axis voltage command value and a q-axis voltage command value so that the d-axis current detection value approaches the d-axis current command value and the q-axis current detection value approaches the q-axis current command value, and converts the d-axis voltage command value and the q-axis voltage command value into multi-phase voltage command values ​​based on the rotation angle; as well as a switching control unit that turns on and off a plurality of switching elements included in the power converter based on the multi-phase voltage command values; The current command value calculation unit changes the d-axis current command value based on a deviation between an offset q-axis current command value obtained by subtracting a positive q-axis offset value from the q-axis current command value and the q-axis current detection value when the q-axis current command value is positive.

2. The control device for a rotating electrical machine according to claim 1, wherein: The current command value calculation unit decreases the d-axis current command value when the q-axis current command value is positive and the q-axis current detection value is lower than the offset q-axis current command value, and increases the d-axis current command value when the q-axis current detection value exceeds the offset q-axis current command value.

3. The control device for a rotating electrical machine according to claim 1 or 2, wherein: The current command value calculation unit limits the q-axis current command value based on the maximum current value and the d-axis current command value so that the current supplied to the multi-phase winding does not exceed a range of a maximum current value that can be supplied to the multi-phase winding.

4. The control device for a rotating electrical machine according to claim 1 or 2, wherein: The voltage command value calculation unit limits the q-axis voltage command value based on the DC voltage and the d-axis voltage command value so that the voltage command value applied to the winding does not exceed a range of a maximum applied voltage corresponding to the DC voltage supplied to the power converter.

5. The control device for a rotating electrical machine according to claim 4, wherein: The voltage command value calculation unit performs low-pass filtering on the d-axis voltage command value used in limiting the q-axis voltage command value.

6. The control device for a rotating electrical machine according to claim 1 or 2, wherein: The current command value calculation unit changes the d-axis current command value based on a deviation between an offset q-axis current command value obtained by adding the positive q-axis offset value to the q-axis current command value and the q-axis current detection value when the q-axis current command value is negative.

7. The control device for a rotating electrical machine according to claim 6, wherein: The current command value calculation unit decreases the d-axis current command value when the q-axis current command value is negative and the q-axis current detection value exceeds the offset q-axis current command value, and increases the d-axis current command value when the q-axis current detection value is lower than the offset q-axis current command value.

8. The control device for a rotating electrical machine according to claim 1 or 2, wherein: The current command value calculation unit sets the q-axis offset value to a value smaller than an absolute value of the q-axis current command value.

9. The control device for a rotating electrical machine according to claim 1 or 2, wherein: The current command value calculation unit sets the q-axis offset value to a larger value when the rotational angular velocity of the rotor decreases than when the rotational angular velocity of the rotor increases.

10. The control device for a rotating electrical machine according to claim 1 or 2, wherein: The current command value calculation unit satisfies the following conditions when the maximum rotational angular velocity of the rotating electric machine in electrical angle is ωmax, the inductance of the rotating electric machine is L, the winding resistance is R, the amount of change in the d-axis current detection value per control cycle is ΔIdodT, and the q-axis offset value is ΔIqoff. The q-axis offset value is set in the manner of ΔIqoff>ωmax·L / R·ΔIdodT.

11. The control device for a rotating electrical machine according to claim 1 or 2, wherein: The current command value calculation unit sets the q-axis offset value to a value larger than an amplitude of a noise component superimposed on the q-axis current detection value.

12. The control device for a rotating electrical machine according to claim 1 or 2, wherein: The current command value calculation unit calculates the d-axis current command value by multiplying the deviation between the offset q-axis current command value and the q-axis current detection value by a proportional gain. The current command value calculation unit sets the q-axis offset value to a value larger than a value obtained by multiplying the amplitude of a noise component superimposed on the q-axis current detection value by the proportional gain.

13. The control device for a rotating electrical machine according to claim 1 or 2, wherein: The current command value calculation unit performs low-pass filtering on the q-axis current detection value used for calculating the d-axis current command value.

14. The control device for a rotating electrical machine according to claim 1 or 2, wherein: The current command value calculation unit corrects the d-axis current command value based on a deviation between the q-axis voltage command value and a limit value of the q-axis voltage command value.

15. The control device for a rotating electrical machine according to claim 1 or 2, wherein: The current command value calculation unit calculates a correction value based on a deviation obtained by subtracting the q-axis current detection value from the q-axis current command value, and calculates the q-axis current command value by adding the correction value to the q-axis current command value before correction.

16. The control device for a rotating electrical machine according to claim 15, wherein: The current command value calculation unit calculates, as the correction value, a value obtained by performing low-pass filtering on the deviation obtained by subtracting the q-axis current detection value from the q-axis current command value.

17. An electric power steering device, characterized in that: include: The control device for a rotating electrical machine according to any one of claims 1 to 16; the power converter; the rotating electric machine; as well as A driving force transmission mechanism transmits the driving force of the rotating electric machine to a steering device of a vehicle.

Images