Motor device and map setting method

Abstract

To provide a motor control device capable of reducing torque ripple and a setting method of a map.SOLUTION: A motor control device 20 includes an open-loop processing unit 60 that sets a harmonic voltage for correcting a command voltage vd* of a d-axis and a command voltage vq* of a q-axis as a correction voltage vdh of the d-axis and a correction voltage vqh of the q-axis. The open-loop processing unit 60 has a harmonic flux calculation unit 62 that calculates harmonic flux λdh of the d-axis and harmonic flux λqh of the q-axis based on a rotation angle θe and a rotation speed ωe of a motor 10 and a correction voltage calculation unit 64 that calculates the correction voltage vdh of the d-axis and the correction voltage vqh of the q-axis based on the harmonic flux λdh of the d-axis and harmonic flux λqh of the q-axis calculated by the harmonic flux calculation unit 62. The motor control device 20 operates an inverter circuit INV based on the command voltage vd* of the d-axis and the command voltage vq* of the q-axis that are corrected by the correction voltage vdh of the d-axis and the correction voltage vqh of the q-axis.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention is Equipment This relates to a method for setting the location and map. [Background technology]

[0002] Patent Document 1 discloses an electric power steering device. The electric power steering device assists the driver's steering operation by applying motor torque of a motor to a steering mechanism of a vehicle. A control device that controls a motor mounted in the electric power steering device calculates a command current that is a target of an actual current flowing through the motor based on a detected steering torque. The control device calculates a correction current for correcting the command current based on a rotation angle of the motor. The correction current is for reducing torque ripple caused by spatial harmonics. The control device then controls the motor torque output by the motor by executing current feedback control that causes the actual current to follow the command current corrected using the correction current. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2003-137110 A Summary of the Invention [Problem to be solved by the invention]

[0004] When the command current is corrected using the correction current, the response speed of the feedback loop of the current feedback control must be made sufficiently fast in order to obtain the effect of reducing torque ripple caused by spatial harmonics. However, the faster the response speed of the feedback loop of the current feedback control is, the more difficult it becomes to make the actual current follow the command current. For this reason, the response speed of the feedback loop of the current feedback control must be set to a level at which the actual current can follow the command current, and there is a risk that the response speed of such a feedback loop of the current feedback control is not sufficient to obtain the effect of reducing torque ripple. In this case, there is a risk that the motor will generate noise and vibration due to the generation of torque ripple.

[0005] The object of the present invention is to Equipment The purpose of the present invention is to reduce torque ripple in the setting of the control unit and the map. [Means for solving the problem]

[0006] A model that solves the above problems Equipment The location is The steering device is mounted on a steering device of a vehicle and is configured to operate a motor of the steering device so as to apply torque to the steering device. A motor control device that controls an inverter circuit that applies voltage to a motor and a rotation angle sensor for detecting a rotation angle of the motor. And, The motor control device includes: The motor drive system includes a command voltage calculation unit that calculates a command voltage as a manipulation amount for current feedback control to make an actual current flowing through the motor follow a command current, an operation processing unit that operates the inverter circuit based on the command voltage calculated by the command voltage calculation unit so that the voltage applied to the motor becomes the command voltage, and an open loop processing unit that sets a harmonic voltage for correcting the command voltage as a correction voltage, and the open loop processing unit Report Rotation angle and forward Report The inverter circuit includes a harmonic flux calculation unit that calculates a harmonic flux based on a rotation speed, which is a change in a rotation angle, and a correction voltage calculation unit that calculates the correction voltage based on the harmonic flux calculated by the harmonic flux calculation unit. The operation processing unit operates the inverter circuit based on the command voltage corrected by the correction voltage calculated by the correction voltage calculation unit. The motor control device further includes a map setting unit that sets a map indicating a relationship between the rotation angle and the rotation speed and the harmonic magnetic flux, the map setting unit detects a rotation speed fluctuation indicating an amount of change in the rotation speed obtained from the rotation angle detected by the rotation angle sensor, estimates a torque ripple by multiplying the rotation speed fluctuation by a predetermined coefficient indicating a moment of inertia, generates the map for calculating the harmonic magnetic flux based on the torque ripple, and sets the map so that the harmonic magnetic flux calculation unit can refer to it. do.

[0007] According to the above configuration, the open loop processing unit calculates a correction voltage for correcting the command voltage using the harmonic magnetic flux calculated based on the rotation angle and the rotation speed. When calculating the harmonic magnetic flux, the harmonic magnetic flux is calculated using not only the rotation angle but also the rotation speed. Since the torque ripple caused by the spatial harmonics can be estimated based on the fluctuation of the rotation speed, the correction voltage for reducing the torque ripple can be calculated based on the rotation speed. The target of the correction to reduce the torque ripple is not the command current used in the current feedback control, but the command voltage. Therefore, even if the response speed of the feedback loop of the current feedback control is set to a level that allows the actual current to follow the command current, the torque ripple can be reduced by correcting the command voltage. Furthermore, according to the above configuration, the map setting unit does not detect the torque ripple from the torque fluctuation, but estimates the torque ripple from the rotational speed fluctuation. Information required to generate a map for calculating the harmonic magnetic flux is the rotational speed fluctuation that can be measured through a rotational angle sensor. This rotational speed fluctuation can be obtained through a rotational angle sensor provided in the motor control device. This rotational angle sensor is a sensor that is provided in the first place in order to detect the rotational angle required to apply a voltage to the motor. Therefore, when the map setting unit generates the map, the configuration required to generate the map can be reduced compared to a case in which a separate sensor such as a torque meter for detecting the torque ripple is provided.

[0008] The above mode Equipment In the apparatus, it is preferable that the open loop processing unit sets a harmonic voltage applied to the motor when an actual current flowing through the motor is a fundamental current as the correction voltage.

[0009] In particular, in the high rotation speed region, the torque ripple caused by spatial harmonics due to the deviation of the flux linkage from the fundamental wave becomes less noticeable, while the torque ripple caused by spatial harmonics due to magnetic saturation when the actual current becomes large tends to become prominent. According to the above configuration, since the command voltage is corrected by the correction voltage, the actual current flowing through the motor can be brought closer to the fundamental wave current, and the torque ripple caused by spatial harmonics can be reduced even in the high rotation speed region.

[0010] The above mode EquipmentIn the apparatus, it is preferable that the harmonic magnetic flux calculation unit is provided with a four-dimensional map that defines a relationship between an actual current on a d-axis, an actual current on a q-axis, the rotation angle, and the rotation speed, and the harmonic magnetic flux on the d-axis and the harmonic magnetic flux on the q-axis, and the harmonic magnetic flux on the d-axis and the harmonic magnetic flux on the q-axis are calculated by referring to the four-dimensional map based on the actual current on the d-axis, the actual current on the q-axis, the rotation angle, and the rotation speed.

[0011] According to the above configuration, since the rotation speed can be input to the four-dimensional map, it is possible to set a four-dimensional map that takes into consideration the decrease in amplitude of harmonic magnetic flux according to the rotation speed, the phase delay of harmonic magnetic flux according to the rotation speed, etc. The harmonic magnetic flux calculation unit can calculate a suitable harmonic magnetic flux based on the actual d-axis current, the actual q-axis current, the rotation angle, and the rotation speed, compared to the case where a three-dimensional map is referenced that defines the relationship between the actual d-axis current, the actual q-axis current, and the rotation angle, and the harmonic magnetic flux on the d-axis and the harmonic magnetic flux on the q-axis.

[0014] A model that solves the above problems Equipment The method of setting the map used in the device includes a measurement step of measuring the change in the rotation angle and the rotation speed, and a step of calculating the change in the rotation speed measured in the measurement step. Indicate quantity a fluctuation detection step of detecting a rotation speed fluctuation; By multiplying by a predetermined coefficient indicating the moment of inertia The method includes an estimation step of estimating a torque ripple, and a map generation step of generating a map for calculating the harmonic magnetic flux based on the torque ripple estimated in the estimation step, the map defining a relationship between the rotation angle and the rotation speed and the harmonic magnetic flux.

[0015] According to the above method, in the fluctuation detection step, the change in the rotation angle and the rotation speed measured in the measurement step is detected. Indicate quantityA rotation speed fluctuation is detected. There is a predetermined relationship between the torque ripple and the rotation speed fluctuation. That is, the torque ripple is a value obtained by multiplying the time derivative of the rotation speed fluctuation by a predetermined coefficient indicating the moment of inertia. Therefore, in the estimation step, the torque ripple can be estimated from the rotation speed fluctuation detected in the fluctuation detection step. Then, a harmonic magnetic flux for reducing the torque ripple can be calculated based on the torque ripple estimated in the estimation step. This makes it possible to obtain the relationship between the rotation angle and rotation speed and the harmonic magnetic flux, and therefore a map for calculating the harmonic magnetic flux can be generated. Effect of the Invention

[0016] The present invention Equipment According to the method of setting the position and map, the torque ripple can be reduced. [Brief description of the drawings]

[0017] [Figure 1] FIG. 2 is a diagram showing a schematic configuration of a motor device. [Diagram 2] FIG. 2 is a perspective view showing a schematic configuration of a motor device. [Diagram 3] FIG. 4 is a diagram showing a schematic configuration of a harmonic magnetic flux calculation unit. [Figure 4] (a) and (b) are diagrams showing distortion of phase voltage. [Diagram 5] A chart showing the steps for setting up a map. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0018] An embodiment of a motor control device will be described with reference to the drawings. Fig. 1 shows the configuration of a motor device 1 in this embodiment. The motor device 1 includes a motor 10, an inverter circuit INV, and a motor control device 20. The motor device 1 is mounted, for example, in an electric power steering device of a vehicle. The motor 10 is a three-phase motor. For example, a surface permanent magnet synchronous motor (SPMSM) with 10 poles and 12 slots is used for the motor 10. The motor 10 generates an assist torque for steering the steered wheels in response to a steering operation by a user.

[0019] As shown in Fig. 2, the motor device 1 is a so-called motor control unit in which a motor 10, an inverter circuit INV, and a motor control device 20 are integrated. The motor 10, the inverter circuit INV, and the motor control device 20 are accommodated inside a housing of the motor device 1. The motor control device 20 also includes a rotation angle sensor 12. The rotation angle sensor 12 is provided on a control board on which various electronic components such as a microcomputer that constitutes the motor control device 20 are mounted. The rotation angle sensor 12 detects the rotation angle of a rotating shaft 10a of the motor 10.

[0020] 1, each terminal of the motor coil of the motor 10 is connected to the battery 14 via an inverter circuit INV. The inverter circuit INV is a circuit that opens and closes between the positive electrode of the battery 14 and each of the three terminals of the motor coil of the motor 10, and also opens and closes between the negative electrode of the battery 14 and each of the three terminals of the motor coil of the motor 10.

[0021] The inverter circuit INV is composed of six switching elements. Metal-oxide-semiconductor field-effect transistors (MOSFETs) are used as the switching elements. The switching elements connected to the terminals of the motor coil of the U phase of the motor 10 are labeled "u", those connected to the terminals of the motor coil of the V phase are labeled "v", and those connected to the terminals of the motor coil of the W phase are labeled "w". The upper arm of the switching element MOSFET is labeled "p", that is, the positive electrode side of the battery 14. The lower arm of the switching element is labeled "n", that is, the negative electrode side of the battery 14. In the following, "u, v, w" are collectively represented as "￥", and "p, n" are collectively represented as "#". The inverter circuit INV includes a switching element S￥p that opens and closes between the positive electrode of the battery 14 and the terminals of the motor coil of the motor 10, and a switching element S￥n that opens and closes between the negative electrode of the battery 14 and the terminals of the motor coil of the motor 10. The switching element S￥p and the switching element S￥n are connected in series. Each of the switching elements S￥# has a diode D￥#.

[0022] A current sensor 16 is provided on the connection line between each terminal of the motor coil of the motor 10 and the inverter circuit INV. The current sensor 16 detects an actual current obtained from the current value of each phase of the motor 10 flowing through the connection line. For ease of explanation, the current sensors for each phase are illustrated as one sensor in FIG. 1. In the following description, the actual current of each phase flowing through the connection line is represented as ￥phase actual current i￥.

[0023] The motor control device 20 will now be described. The motor control device 20 controls the motor 10. The motor control device 20 operates an inverter circuit INV that applies a voltage to the motor 10.

[0024] The motor control device 20 includes a dq conversion unit 22, a rotational speed calculation unit 23, a command current setting unit 24, deviation calculation units 26, 28, current feedback control units 30, 32, a non-interference control unit 34, a uvw conversion unit 36, a PWM processing unit 38, a dead time generation unit 40, an open loop processing unit 60, and voltage correction units 66, 68.

[0025] The dq converter 22 converts the three-phase actual currents iu, iv, iw detected by the current sensor 16 into an actual current id on the d-axis and an actual current iq on the q-axis. The rotation speed calculation unit 23 calculates the rotation speed ωe by time differentiating the rotation angle θe detected by the rotation angle sensor 12. The rotation angle θe detected by the rotation angle sensor 12 is an electrical angle, and the rotation speed ωe is an electrical angular velocity.

[0026] The command current setting unit 24 sets the d-axis command current id* and the q-axis command current iq* based on the torque command value Trq*, which is a command value for the torque generated by the motor 10, and the rotation speed ωe. The command current setting unit 24 sets the q-axis command current iq* so that its absolute value increases as the absolute value of the torque command value Trq* increases. On the other hand, the command current setting unit 24 sets the d-axis command current id* so that its absolute value is greater than zero when the absolute value of the rotation speed ωe is equal to or greater than a predetermined speed. The command current setting unit 24 sets the absolute value so that its absolute value increases as the rotation speed ωe increases in a region where the absolute value of the rotation speed ωe is equal to or greater than a predetermined speed. The d-axis command current id* is used to perform well-known field-weakening control.

[0027] The deviation calculation unit 26 calculates the deviation by subtracting the d-axis actual current id from the d-axis command current id*. The deviation calculation unit 28 calculates the deviation by subtracting the q-axis actual current iq from the q-axis command current iq*. The current feedback control unit 30 executes current feedback control of the d-axis actual current id so that the d-axis actual current id follows the d-axis command current id*, thereby calculating the d-axis voltage as the manipulated variable for current feedback control. That is, the current feedback control unit 30 calculates the d-axis voltage so as to eliminate the deviation between the d-axis command current id* and the actual current id. The current feedback control unit 32 executes current feedback control of the q-axis actual current iq so that the q-axis actual current iq follows the q-axis command current iq*, thereby calculating the q-axis voltage as the manipulated variable for current feedback control. That is, the current feedback control unit 32 calculates the q-axis voltage so as to eliminate the deviation between the q-axis command current iq* and the q-axis actual current iq. In this embodiment, the current feedback control units 30 and 32 execute PI control using a proportional element and an integral element as the current feedback control.

[0028] The decoupling control unit 34 corrects the d-axis voltage and q-axis voltage calculated by the current feedback control units 30, 32 using a decoupling term and an induced voltage compensation term, and outputs the corrected voltage as the final d-axis command voltage vd* and q-axis command voltage vq*. Here, the decoupling term is an open-loop operation amount based on the d-axis actual current id and the q-axis actual current iq. Also, the induced voltage compensation term is an open-loop operation amount based on the rotation speed ωe. These are well known, so further description will be omitted. The command voltage calculation unit described in the claims corresponds to the current feedback control units 30, 32 and the decoupling control unit 34.

[0029] The uvw conversion unit 36 ​​converts the d-axis command voltage vd* and the q-axis command voltage vq* into three-phase command voltages vu*, vv*, and vw*. The PWM processing unit 38 generates three-phase PWM signals gu, gv, and gw based on the three-phase command voltages vu*, vv*, and vw*. The dead time generation unit 40 generates an operation signal g¥# for the switching element S¥# based on the PWM signal g¥, and outputs it to the inverter circuit INV. The PWM signal g¥ specifies the on-period of the switching element S¥p of the upper arm and the on-period of the switching element S¥n of the lower arm. A dead time is added to the PWM signal g¥# so that the switching element S¥p of the upper arm or the switching element S¥n of the lower arm is turned off before the other is turned on from off. The operation processing unit described in the claims corresponds to the uvw conversion unit 36, the PWM processing unit 38, and the dead time generation unit 40.

[0030] The open-loop processing unit 60 receives the d-axis real current id, the q-axis real current iq, the rotation angle θe, and the rotation speed ωe. The open-loop processing unit 60 calculates the d-axis correction voltage vdh for correcting the d-axis command voltage vd* and calculates the q-axis correction voltage vqh for correcting the q-axis command voltage vq* based on the d-axis real current id, the q-axis real current iq, the rotation angle θe, and the rotation speed ωe. The d-axis correction voltage vdh and the q-axis correction voltage vqh are intended to suppress deviation of the real current flowing through the motor 10 from the fundamental wave due to spatial harmonics caused by magnetic saturation of the inductance of the stator coil when the absolute value of the real current flowing through the motor 10 becomes large. Here, the d-axis correction voltage vdh and the q-axis correction voltage vqh are open-loop control amounts.

[0031] Specifically, the open loop processing unit 60 includes a harmonic magnetic flux calculation unit 62 and a correction voltage calculation unit 64 . As shown in Fig. 3, the harmonic flux calculation unit 62 receives the d-axis real current id, the q-axis real current iq, the rotation angle θe, and the rotation speed ωe. The harmonic flux calculation unit 62 calculates the harmonic fluxes λdh, λqh based on the d-axis real current id, the q-axis real current iq, the rotation angle θe, and the rotation speed ωe. The harmonic flux calculation unit 62 includes a four-dimensional map M that defines the relationship between the d-axis real current id, the q-axis real current iq, the rotation angle θe, and the rotation speed ωe and the d-axis harmonic flux λdh and the q-axis harmonic flux λqh. The four-dimensional map M is stored in a storage unit of the harmonic flux calculation unit 62. The storage unit is a predetermined storage area of ​​a memory (not shown). The four-dimensional map M is, for example, a three-dimensional map of the d-axis real current id, the q-axis real current iq, and the rotation angle θe, which is prepared for each rotation speed ωe. This four-dimensional map M is generated by obtaining the harmonic magnetic fluxes λdh, λqh for each set of the d-axis real current id, the q-axis real current iq, the rotation angle θe, and the rotation speed ωe when a fundamental current is steadily applied to the motor 10. A method for setting the four-dimensional map M will be described later. In the four-dimensional map M, the greater the absolute value of the d-axis real current id and the greater the absolute value of the q-axis real current iq, the greater the absolute values ​​of the d-axis harmonic magnetic flux λdh and the q-axis harmonic magnetic flux λqh. The angle range of the rotation angle θe input to the four-dimensional map M is, for example, "0 to 60°". This setting is aimed at reducing the sixth harmonics and their constant multiple harmonics.

[0032] FIG. 4(a) shows the transition of phase voltages vu, vv, and vw, which are voltages applied to each terminal of the motor coil of the motor 10 when a fundamental wave current is passed through the motor 10. As shown in the figure, in this case, the phase voltages vu, vv, and vw have distortion with respect to the fundamental wave. FIG. 4(b) shows the quantified distortion rate of the phase voltage v￥ in FIG. 4(a). Here, the distortion rate is quantified as a percentage of the effective value of the harmonic with respect to the effective value of the fundamental wave. As shown in the figure, the distortion rates are large in the 3rd, 5th, 7th, 11th, and 13th harmonic components in the fixed coordinate system of uvw. Here, the 3rd harmonic component does not actually contribute to torque ripple, etc. On the other hand, the 3rd and 5th harmonic components become the 6th harmonic components in the rotating coordinate system, and the 11th and 13th harmonic components become the 12th harmonic components in the rotating coordinate system. For this reason, in this embodiment, the 6th harmonic and its multiple harmonic components are targeted. Here, the sixth harmonic has a period of 60°, and its multiple harmonics have periods that are divisors of 60°. Therefore, the domain of the rotation angle θe as an input variable of the map is 60°.

[0033] As shown in FIG. 3, the harmonic magnetic flux calculation unit 62 refers to a four-dimensional map M to perform map calculation of the d-axis harmonic magnetic flux λdh and the q-axis harmonic magnetic flux λqh based on the d-axis actual current id, the q-axis actual current iq, the rotation angle θe, and the rotation speed ωe.

[0034] 1, the correction voltage calculation unit 64 calculates the d-axis correction voltage vdh and the q-axis correction voltage vqh using the following formula (c1) based on the d-axis harmonic magnetic flux λdh and the q-axis harmonic magnetic flux λqh calculated by the harmonic magnetic flux calculation unit 62 and the rotation speed ωe. Note that the derivation of formula (c1) is described in the section <Regarding derivation of correction voltage>.

[0035]

number

[0036] The motor control device 20 further includes a map setting unit 100. The map setting unit 100 receives the actual current id of the d-axis, the actual current iq of the q-axis, the rotation angle θe, and the rotation speed ωe. The map setting unit 100 generates the above-mentioned four-dimensional map M based on the actual current id of the d-axis, the actual current iq of the q-axis, the rotation angle θe, and the rotation speed ωe. The map setting unit 100 outputs the generated four-dimensional map M to the harmonic magnetic flux calculation unit 62 of the open-loop processing unit 60, thereby setting the four-dimensional map M. A method for setting the four-dimensional map M will be described in the section <Regarding a method for setting the four-dimensional map M>.

[0037] <Derivation of correction voltage> The above formula (c1) will be derived below. First, the voltage equation on the dq axes is the following formula (c2).

[0038]

number

[0039]

number

[0040]

number

[0041] <How to set up 4D map M> The four-dimensional map M is set before the motor device 1 is mounted in an electric power steering device of a vehicle. For this reason, for example, the four-dimensional map M is set before the motor device 1 is shipped from a factory.

[0042] In setting the four-dimensional map M, first, the motor device 1, which is the target of setting the four-dimensional map M, is installed in a predetermined map setting area in a factory. This map setting area may be any place in the factory as long as power can be supplied to the motor device 1. This is because the motor device 1 has a rotation angle sensor 12 for detecting information required for setting the four-dimensional map M. Then, as shown in FIG. 2, after the motor device 1 is installed in the map setting area, the motor device 1 is connected to a map setting control device 110. The map setting control device 110 has a battery that supplies power when setting the four-dimensional map M. When setting the four-dimensional map M, the battery of the map setting control device 110 functions as the battery 14 in FIG. 1. Then, after preparations for setting the four-dimensional map M are completed, the map setting control device 110 outputs a signal to the motor device 1 to start setting the four-dimensional map M. When the motor control device 20 of the motor device 1 receives a signal to start setting the four-dimensional map M, it executes setting of the four-dimensional map M according to the procedure shown in the flowchart of FIG. 5.

[0043] As shown in FIG. 5, the motor control device 20 executes a measurement step (step S1). In the measurement step, the motor control device 20 measures the d-axis actual current id, the q-axis actual current iq, the rotation angle θe, and the rotation speed ωe when a fundamental current is steadily applied to the motor 10. The motor 10 rotates when a fundamental current is steadily applied, that is, when a constant d-axis command current id* and a constant q-axis command current iq* are applied to the motor 10. In this case, if no torque ripple is generated, the motor 10 rotates with a constant load, and therefore the rotation of the motor 10 accelerates at a constant acceleration. The motor 10 rotates within an angle range corresponding to the angle range in which the rotating shaft 10a can rotate when the motor device 1 is mounted on an electric power steering device. The map setting unit 100 receives inputs of a change in the rotation angle θe detected by the rotation angle sensor 12 and a change in the rotation speed ωe calculated by the rotation speed calculation unit 23. Furthermore, changes in the actual current id on the d-axis and the actual current iq on the q-axis are input to the map setting unit 100. The map setting unit 100 prepares measurement data for each set of the actual current id on the d-axis, the actual current iq on the q-axis, the rotation angle θe, and the rotation speed ωe.

[0044] After the measurement step is completed, the map setting unit 100 executes a fluctuation detection step (step S2). In the fluctuation detection step, the map setting unit 100 detects a rotation speed fluctuation, which is a fluctuation in the rotation speed ωe, based on a change in the rotation speed ωe. If no torque ripple is generated, the motor 10 rotates with a constant load, and therefore the absolute value of the rotation speed ωe changes at a constant rate. However, if a torque ripple is generated, the relationship of the rotation speed ωe of the motor 10, in which the absolute value of the rotation speed ωe changes at a constant rate, is lost. For this reason, a discrepancy occurs between the rotation speed ωe when no torque ripple is generated and the rotation speed ωe when a torque ripple is generated.

[0045] In the fluctuation detection step, the map setting unit 100 stores, for each rotation speed fluctuation, the amplitude and phase of the rotation speed fluctuation as information indicating the rotation speed fluctuation. The amplitude of the rotation speed fluctuation is the deviation between the rotation speed ωe when no torque ripple occurs and the rotation speed ωe when a torque ripple occurs. The phase of the rotation speed fluctuation is the position of the rotation angle θe at which the rotation speed fluctuation occurs.

[0046] After the fluctuation detection step is completed, the map setting unit 100 executes an estimation step (step S3). Here, the torque ripple is a value obtained by multiplying the time derivative of the rotation speed fluctuation by a predetermined coefficient indicating the moment of inertia. Therefore, a predetermined relationship exists between the torque ripple and the rotation speed fluctuation. Therefore, in the estimation step, the map setting unit 100 estimates the torque ripple for each set of the actual current id of the d-axis, the actual current iq of the q-axis, the rotation angle θe, and the rotation speed ωe based on the rotation speed fluctuation detected in the fluctuation detection step.

[0047] After the estimation step is completed, the map setting unit 100 executes a map generation step (step S4). In the map generation step, the map setting unit 100 generates a four-dimensional map M for calculating harmonic magnetic fluxes based on the torque ripple estimated in the estimation step. Specifically, the map setting unit 100 calculates the d-axis harmonic magnetic flux λdh and the q-axis harmonic magnetic flux λqh at which the torque ripple is estimated to be minimum. The map setting unit 100 calculates the d-axis harmonic magnetic flux λdh and the q-axis harmonic magnetic flux λqh for each set of the d-axis actual current id, the q-axis actual current iq, the rotation angle θe, and the rotation speed ωe. The map setting unit 100 generates the four-dimensional map M by collecting these calculation results. After generating the four-dimensional map M, the map setting unit 100 outputs the generated four-dimensional map M to the harmonic magnetic flux calculation unit 62 of the open-loop processing unit 60. The harmonic magnetic flux calculation unit 62 stores the four-dimensional map M generated by the map setting unit 100 in a storage unit and sets the four-dimensional map M. In this manner, the four-dimensional map M that defines the relationship between the actual current id of the d-axis, the actual current iq of the q-axis, the rotation angle θe, and the rotation speed ωe, and the harmonic magnetic flux λdh of the d-axis and the harmonic magnetic flux λqh of the q-axis is set in the harmonic magnetic flux calculation unit 62.

[0048] After the four-dimensional map M is set in the harmonic magnetic flux calculation unit 62, the connection between the motor device 1 and the map setting control device 110 is released. The motor device 1 in which the four-dimensional map M has been set in this manner is shipped from the factory and mounted on the electric power steering device of the vehicle.

[0049] The operation of this embodiment will now be described. The open loop processing unit 60 uses the d-axis harmonic magnetic flux λdh and the q-axis harmonic magnetic flux λqh to calculate the d-axis command voltage vd* and the q-axis command voltage vq*. When calculating the d-axis harmonic magnetic flux λdh and the q-axis harmonic magnetic flux λqh, the d-axis harmonic magnetic flux λdh and the q-axis harmonic magnetic flux λqh are calculated using not only the rotation angle but also the rotation speed ωe. Since the torque ripple caused by the spatial harmonics can be estimated based on the rotation speed fluctuation, the d-axis correction voltage vdh and the q-axis correction voltage vqh for reducing the torque ripple can be calculated based on the rotation speed ωe. Here, the d-axis correction voltage vdh and the q-axis correction voltage vqh are determined so as to minimize the torque ripple when a fundamental wave current is passed through the motor 10. For this reason, by superimposing the d-axis correction voltage vdh calculated by the open-loop processing unit 60 on the d-axis command voltage vd* calculated by the non-interference control unit 34, and by superimposing the q-axis correction voltage vqh calculated by the open-loop processing unit 60 on the q-axis command voltage vq* calculated by the non-interference control unit 34, it is possible to make the actual current flowing through the motor 10 closer to the fundamental current. Since the actual current flowing through the motor 10 can be made closer to the fundamental current, the d-axis actual current id and the q-axis actual current iq input to the open-loop processing unit 60 also become closer to the fundamental current.

[0050] The effects of this embodiment will be described. (1) The object of correction to reduce torque ripple is not the d-axis command current id* and q-axis command current iq* used in current feedback control, but the d-axis command voltage vd* and q-axis command voltage vq*. Therefore, even if the response speed of the feedback loop of the current feedback control is set to a level that allows the actual current to follow the command current, the torque ripple can be reduced by correcting the d-axis command voltage vd* and the q-axis command voltage vq*.

[0051] (2) In particular, in the high rotation speed region, the torque ripple caused by the spatial harmonics due to the deviation of the interlinkage magnetic flux from the fundamental wave is not noticeable, and the torque ripple caused by the spatial harmonics due to the magnetic saturation when the actual current is large is likely to be noticeable. In particular, the torque ripple caused by the spatial harmonics due to the magnetic saturation may match the resonance frequency of the mechanical configuration of the motor device 1 and the mechanical configuration of the electric power steering device in which the motor device 1 is mounted, and may cause abnormal noise. In this embodiment, the d-axis command voltage vd* and the q-axis command voltage vq* are corrected by the d-axis correction voltage vdh and the q-axis correction voltage vqh, so that the actual current flowing through the motor 10 can be brought close to the fundamental wave current, and the torque ripple caused by the spatial harmonics can be reduced even in the high rotation speed region.

[0052] (3) Since the rotation speed ωe can be input to the four-dimensional map M, it is possible to set the four-dimensional map M taking into consideration the decrease in amplitude of the harmonic magnetic flux according to the rotation speed ωe, the phase delay of the harmonic magnetic flux according to the rotation speed ωe, etc. The harmonic magnetic flux calculation unit 62 can calculate a suitable harmonic magnetic flux based on the d-axis actual current id, the q-axis actual current iq, the rotation angle θe, and the rotation speed ωe, as compared to the case of referring to a three-dimensional map that defines the relationship between the d-axis actual current id, the q-axis actual current iq, and the rotation angle θe and the d-axis harmonic magnetic flux λdh and the q-axis harmonic magnetic flux λqh, for example.

[0053] (4) The map setting unit 100 does not detect the torque ripple from the torque fluctuation, but estimates the torque ripple from the rotational speed fluctuation. Information required to generate the four-dimensional map M for calculating the harmonic magnetic flux is the rotational speed fluctuation that can be measured through the rotational angle sensor 12. This rotational speed fluctuation can be obtained through the rotational angle sensor 12 provided in the motor control device 20. This rotational angle sensor 12 is a sensor that is provided originally to detect the rotational angle θe that is required to apply a voltage to the motor 10. For this reason, when the map setting unit 100 generates the four-dimensional map M, the configuration required to generate the four-dimensional map M can be reduced compared to the case where a separate sensor such as a torque meter for detecting the torque ripple is provided.

[0054] (5) When measuring the change in the rotation speed ωe of the motor 10 in the measurement step, the motor 10 rotates by itself. Therefore, when the map setting unit 100 generates the four-dimensional map M, the configuration required for generating the four-dimensional map M can be reduced compared to when a separate external motor is provided for rotating the rotating shaft 10a of the motor 10.

[0055] (6) In the fluctuation detection step, the rotation speed fluctuation is detected based on the change in the rotation speed of the motor 10 measured in the measurement step. There is a predetermined relationship between the torque ripple and the rotation speed fluctuation. That is, the torque ripple is a value obtained by multiplying the time derivative of the rotation speed fluctuation by a predetermined coefficient indicating the moment of inertia. Therefore, in the estimation step, the torque ripple can be estimated from the rotation speed fluctuation detected in the fluctuation detection step. Then, based on the torque ripple estimated in the estimation step, the d-axis harmonic magnetic flux λdh and the q-axis harmonic magnetic flux λqh for reducing the torque ripple can be calculated. This makes it possible to obtain the relationship between the rotation angle θe and the rotation speed ωe and the d-axis harmonic magnetic flux λdh and the q-axis harmonic magnetic flux λqh, and therefore it is possible to generate a four-dimensional map M for calculating the d-axis harmonic magnetic flux λdh and the q-axis harmonic magnetic flux λqh.

[0056] The above embodiment may be modified as follows. In addition, the following other embodiments may be combined with each other to the extent that there is no technical contradiction. When calculating the d-axis harmonic magnetic flux λdh and the q-axis harmonic magnetic flux λqh, the harmonic magnetic flux calculation unit 62 may use the d-axis command current id* instead of the d-axis actual current id and the q-axis command current iq* instead of the q-axis actual current iq. There may also be a correlation between the torque command value Trq* and the q-axis command current id*, or a correlation between the rotation speed ωe and the d-axis command current id*. In this case, when calculating the d-axis harmonic magnetic flux λdh and the q-axis harmonic magnetic flux λqh, the harmonic magnetic flux calculation unit 62 may use the torque command value Trq* instead of the q-axis actual current iq and the rotation speed ωe instead of the d-axis actual current id.

[0057] When calculating the d-axis harmonic magnetic flux λdh and the q-axis harmonic magnetic flux λqh, the harmonic magnetic flux calculation unit 62 may use parameters such as the rotational acceleration, which is the amount of change in the rotational speed ωe, in addition to the d-axis actual current id, the q-axis actual current iq, the rotation angle θe, and the rotational speed ωe.

[0058] When calculating the d-axis harmonic magnetic flux λdh and the q-axis harmonic magnetic flux λqh, the harmonic magnetic flux calculation unit 62 may use values ​​in a fixed two-phase coordinate system or a fixed three-phase coordinate system instead of values ​​in a rotating coordinate system.

[0059] When calculating the d-axis harmonic magnetic flux λdh and the q-axis harmonic magnetic flux λqh, the harmonic magnetic flux calculation unit 62 may use a rotation angle that is a mechanical angle instead of the rotation angle θe that is an electrical angle. Also, when calculating the d-axis harmonic magnetic flux λdh and the q-axis harmonic magnetic flux λqh, the harmonic magnetic flux calculation unit 62 may use a rotation speed that is the amount of change in the rotation angle that is a mechanical angle instead of the rotation speed ωe that is the amount of change in the rotation angle θe that is an electrical angle. The same applies to the correction voltage calculation unit 64.

[0060] In the above embodiment, the d-axis correction voltage vdh and the q-axis correction voltage vqh are calculated to reduce the sixth-order torque ripple. However, the torque ripple of other orders, such as the twelfth-order torque ripple, may be reduced.

[0061] Although the four-dimensional map M is generated by the map setting unit 100 of the motor control device 20, it may be generated by a control device other than the motor control device 20, such as a map setting control device 110. In this case, the motor control device 20 does not need to include the map setting unit 100. The other control device outputs the generated map to the motor control device 20, and the harmonic magnetic flux calculation unit 62 stores the acquired map in a storage unit.

[0062] An external motor separate from the motor 10 may be attached to the rotating shaft 10a of the motor 10. In this case, in the measurement step, the rotating shaft 10a is not rotated by the motor 10 itself, but the rotating shaft 10a can be rotated by the external motor.

[0063] The motor control device 20 does not have to be equipped with the rotation angle sensor 12. In this case, for example, an external rotation angle sensor may be connected to the motor control device 20. In the measurement step, the external rotation angle sensor detects changes in the rotation angle θe and the rotation speed ωe. Also, for example, a position sensor such as an encoder that detects the position of the rotating shaft 10a may be connected. The position sensor detects the rotation angle θe by detecting the position of the rotating shaft 10a. In the measurement step, the position sensor detects changes in the rotation angle θe and the rotation speed ωe.

[0064] Even if the motor control device 20 is equipped with the rotation angle sensor 12, an external rotation angle sensor or position sensor may be connected to the motor control device 20. In this case, in the measurement step, changes in the rotation angle θe and the rotation speed ωe are detected by the external rotation angle sensor or position sensor instead of the rotation angle sensor 12.

[0065] In the measurement step, the four-dimensional map M is set before the motor device 1 is mounted on the electric power steering device of the vehicle, but the four-dimensional map M may be set after the motor device 1 is mounted on the electric power steering device of the vehicle. In this case, the measurement step is performed within an angle range in which the rotating shaft 10a of the motor 10 can rotate when the steering wheel of the electric power steering device is steered from a state in which it is steered maximally to the right to a state in which it is steered maximally to the left.

[0066] In the estimation step, the map setting unit 100 may estimate the torque ripple based on other parameters, such as the actual current, in addition to the rotation speed fluctuation. In the map generating step, the map setting unit 100 may generate the four-dimensional map M based on other parameters, such as the actual current, in addition to the torque ripple.

[0067] The motor control device 20 may include an abnormality detection unit. In this case, the abnormality detection unit constantly monitors the rotation speed fluctuation, and detects that an abnormality has occurred in the motor 10 when the rotation speed fluctuation exceeds a threshold value indicating the maximum value of the rotation speed fluctuation. The threshold value is set to the maximum value of the rotation speed fluctuation detected when setting the four-dimensional map M, for example. Note that possible abnormalities in the motor 10 include, for example, demagnetization or falling off of a permanent magnet provided in the rotor, deterioration of the motor coil, etc.

[0068] The motor 10 is not limited to an SPMSM, but may be an interior magnet synchronous motor. It is not limited to a permanent magnet synchronous motor, but may be, for example, a wound field type synchronous motor or a reluctance motor.

[0069] The inverter circuit INV may be, for example, a three-level inverter circuit, or may be a circuit having the same circuit configuration as a DC-DC converter connected to each terminal of the motor coil of the motor 10.

[0070] The motor device 1 is not limited to being mounted in an electric power steering device, but may be mounted in a steer-by-wire device, for example, or may be used as a generator.

[0071] The technical ideas that can be understood from the above-mentioned embodiment and other embodiments will be described below together with their effects. (A) A motor control device in which the command current, the command voltage, the harmonic magnetic flux, and the correction voltage are values ​​in a rotating coordinate system, and the harmonic magnetic flux calculation unit calculates the correction voltage based on a differential calculation of the harmonic magnetic flux. According to the above configuration, by using the rotating coordinate system, the fundamental wave component can be treated as a DC component, and therefore the correction voltage can be easily calculated. [Explanation of symbols]

[0072] 1...Motor device 10…Motor 20...Motor control device INV: Inverter circuit 60...Open loop processing section 62...Harmonic magnetic flux calculation section 64...Compensation voltage calculation unit 100…Map setting section vd*,vq*: d-axis and q-axis command voltages vdh, vqh: d-axis and q-axis compensation voltages λdh, λqh: d-axis and q-axis harmonic magnetic flux M...4-dimensional map

Claims

1. A motor device that is mounted on a steering device of a vehicle and includes a motor control device that controls an inverter circuit that applies a voltage to a motor provided in the steering device to operate the motor so as to apply a torque to the steering device, and a rotation angle sensor that detects a rotation angle of the motor, The motor control device includes: a command voltage calculation unit that calculates a command voltage as a manipulated variable for current feedback control to make an actual current flowing through the motor follow a command current; an operation processing unit that operates the inverter circuit based on the command voltage calculated by the command voltage calculation unit so that a voltage applied to the motor becomes the command voltage; an open-loop processing unit that sets a harmonic voltage for correcting the command voltage as a correction voltage; The open loop processing unit includes: a harmonic magnetic flux calculation unit that calculates harmonic magnetic flux based on the rotation angle and a rotation speed that is an amount of change in the rotation angle; a correction voltage calculation unit that calculates the correction voltage based on the harmonic magnetic flux calculated by the harmonic magnetic flux calculation unit, The operation processing unit operates the inverter circuit based on the command voltage corrected by the correction voltage calculated by the correction voltage calculation unit, the motor control device further includes a map setting unit that sets a map indicating a relationship between the rotation angle and the rotation speed, and the harmonic magnetic flux; The map setting unit detects rotational speed fluctuation indicating an amount of change in the rotational speed obtained from the rotation angle detected by the rotation angle sensor, estimates a torque ripple by multiplying the rotational speed fluctuation by a predetermined coefficient indicating a moment of inertia, generates the map for calculating the harmonic magnetic flux based on the torque ripple, and sets the map so that the harmonic magnetic flux calculation unit can refer to it.

2. 2. The motor device according to claim 1, wherein the open loop processing unit sets a harmonic voltage applied to the motor as the correction voltage when an actual current flowing through the motor is a fundamental current.

3. the harmonic magnetic flux calculation unit is provided with a four-dimensional map that defines a relationship between a d-axis actual current, a q-axis actual current, the rotation angle, and the rotation speed, and the d-axis harmonic magnetic flux and the q-axis harmonic magnetic flux; 3. The motor device according to claim 1, wherein the harmonic magnetic flux calculation unit refers to the four-dimensional map to calculate the harmonic magnetic flux on the d-axis and the harmonic magnetic flux on the q-axis based on the actual current on the d-axis, the actual current on the q-axis, the rotation angle, and the rotation speed.

4. A method for setting a map used in the open loop processing unit of the motor device according to any one of claims 1 to 3, comprising: a measuring step of measuring changes in the rotation angle and the rotation speed; a fluctuation detection step of detecting a rotation speed fluctuation indicating an amount of change in the rotation speed measured in the measurement step; an estimation step of estimating a torque ripple by multiplying the rotation speed fluctuation detected in the fluctuation detection step by a predetermined coefficient indicating a moment of inertia; a map generating step of generating a map for calculating the harmonic magnetic flux based on the torque ripple estimated in the estimating step, A method for setting a map, the map defining a relationship between the rotation angle and the rotation speed and the harmonic magnetic flux.

Images