JP2012005318A - Motor controller and electric power steering device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a motor controller for precisely detecting occurrence of a conduction failure in a power supply passage without detecting rotation angle speed based on a real rotation angle.SOLUTION: A conduction failure detection portion judges whether an induced voltage square sum Esq_αβ exceeds a prescribed threshold value E1 or not (step 602), and judges whether a motor rotation angle speed estimation value ωm_e is lower than a prescribed threshold value ω1 or not (step 603). When contradiction is detected that the motor rotation angle speed estimation value ωm_e is in a minimum region showing a non-rotating state (ωm_e<ω1, step 603:YES) although the induced voltage square sum Esq_αβ is in a maximum region corresponding to a maximum speed rotating state (Esq_αβ>E1, step 602:YES), it is judged that abnormality showing occurrence of the conduction failure exists in one of phases (step 604).

Description

  The present invention relates to a motor control device and an electric power steering device.

  In general, a motor control device that controls the operation of a motor through the supply of drive power has a function of detecting the occurrence of energization failure in the power supply path. That is, in a motor control device that controls a brushless motor that rotates based on three-phase (U, V, W) driving power, when the current value of a phase that should be in an energized state is a value that indicates a non-energized state. It can be determined that an energization failure (disconnection state) has occurred in the phase (see, for example, Patent Document 1). Whether or not the detection target phase is “a phase that should be in an energized state” can be determined based on, for example, a current command value, a voltage command value (or duty), or the like of the phase. Further, by adding a condition for the motor rotation angular velocity and eliminating the high-speed rotation region where the current cannot flow into the motor coil due to the influence of the induced voltage, that is, the motor current is minimized, the current conduction failure can be detected more accurately. (For example, refer to Patent Document 2).

International Publication No. 2006/112033 Pamphlet JP 2007-244028 A JP 2010-11709 A JP 2010-29031 A

  However, some motor control devices that control brushless motors supply the drive power without detecting the actual rotation angle of the motor (see, for example, Patent Documents 3 and 4). For this reason, for those performing so-called resolveless control, the speed condition determination cannot be performed using the actual rotational angular velocity, and there is a problem that improvement in detection accuracy by adding the speed condition cannot be expected. However, there was still room for improvement.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to accurately detect the occurrence of energization failure in the power supply path without detecting the actual rotational angular velocity. A motor control device and an electric power steering device are provided.

  In order to solve the above problems, the invention according to claim 1 is a motor control signal output means for outputting a motor control signal, and a three-phase driving power based on a power supply voltage that is operated by the input of the motor control signal. And a drive circuit for supplying a motor to the motor, and an abnormality detection means for detecting the occurrence of an energization failure in the current supply path to the motor, wherein the motor control signal output means corresponds to the amount of change in the motor rotation angle for each calculation cycle. The motor control signal is calculated by calculating a motor rotation angle on the control by calculating the addition angle to be added and integrating the addition angle, and executing current feedback control in a rotating coordinate system according to the motor rotation angle on the control. In the motor control device that outputs the three-phase current values and the phase voltage values to the respective axis current values and the respective axis voltage values of the two-phase fixed coordinate system. And an induced voltage estimation means for calculating each axis induced voltage estimated value of the two-phase fixed coordinate system based on each axis current value and each axis voltage value, and a motor rotation angle estimated value based on each axis induced voltage estimated value A rotation angular velocity estimation means for calculating a motor rotation angular velocity estimation value based on a change amount of the motor rotation angle estimation value, and the abnormality detection means has a region in which the sum of squares of the respective shaft induced voltage estimation values is a maximum region. When the estimated motor rotation angular velocity is in the minimum region, it is determined that there is an abnormality indicating the occurrence of the energization failure.

  That is, when an energization failure occurs in any of the phases, the voltage value of each phase becomes very close to what is called phase-fixed energization performed by fixing the energization pattern by current feedback control. At this time, any one of the current values of each phase has a theoretical value of “0” due to the occurrence of an energization failure. For this reason, each axis induced voltage estimated value in the two-phase fixed coordinate system becomes a substantially constant value, and as a result, the motor rotation angle estimated value based on each axis induced voltage estimated value is also obtained during phase fixing energization by the energization pattern. It becomes a substantially constant value in the vicinity of the generated rotation angle (electrical angle) (fixed state). Then, when the estimated motor rotation angular velocity is calculated based on the estimated motor rotation angle, a contradiction arises between the estimated motor rotation angular velocity and the sum of squares of the respective shaft induced voltage estimates. Therefore, as in the above configuration, by monitoring the contradiction between the estimated motor rotation angular velocity and the sum of squares of each axis induced voltage estimation value, even when performing the resolverless control without using the actual rotation angular velocity of the motor, The occurrence of energization failure in the power supply path can be detected with high accuracy.

  According to a second aspect of the present invention, the motor control signal output means calculates the addition angle based on the estimated motor rotation angular velocity, and the abnormality detection means indicates the occurrence of the energization failure. The gist is to confirm the abnormality when the abnormality continues.

  That is, when the estimated motor rotation angle is fixed and the estimated motor rotation angular velocity is substantially “0”, the addition angle calculated based on the estimated motor rotation angular velocity is also a small value. And since the change of the virtual motor rotation angle on the control becomes small, the state where the estimated value of the motor rotation angle is fixed also continues. Therefore, according to the said structure, generation | occurrence | production of a conduction failure can be detected more accurately.

The gist of the invention described in claim 3 is an electric power steering apparatus provided with the motor control device described in claim 1 or claim 2.
According to the above configuration, it is possible to accurately detect the occurrence of an energization failure even when performing the resolverless control without using the actual rotational angular velocity. As a result, it is possible to provide an electric power steering apparatus that is excellent in steering feeling by quickly stopping the motor drive after failing in energization and achieving fail safe.

  According to the present invention, it is possible to provide a motor control device and an electric power steering device that can accurately detect the occurrence of an energization failure in a power supply path without performing a rotation angular velocity detection based on an actual rotation angle. it can.

The schematic block diagram of an electric power steering device (EPS). The block diagram which shows the electric constitution of EPS. The circuit diagram of a drive circuit. The schematic block diagram of a 1st control part. The schematic block diagram of a 2nd control part. The block diagram which shows schematic structure of a disturbance observer. The flowchart which shows the process sequence of rotation angular velocity estimation. The flowchart which shows the process sequence of addition angle adjustment calculation. The schematic block diagram of the electric current command value calculating part by the side of a 2nd control part. The flowchart which shows the process sequence which detects selection of the specific phase used as a detection object phase, and generation | occurrence | production of the electricity supply failure in the said specific phase. The flowchart which shows the process sequence of normal electricity supply failure determination. The flowchart which shows the process sequence at the time of switching the mode of electricity supply failure determination. The flowchart which shows the process sequence of alternative electricity supply failure determination. The flowchart which shows the process sequence of abnormality continuation determination. The flowchart which shows the process sequence at the time of specifying electricity supply failure generation | occurrence | production.

DESCRIPTION OF EXEMPLARY EMBODIMENTS Hereinafter, an embodiment of the invention will be described with reference to the drawings.
As shown in FIG. 1, in the electric power steering apparatus (EPS) 1 of this embodiment, a steering shaft 3 to which a steering 2 is fixed is connected to a rack shaft 5 via a rack and pinion mechanism 4. The rotation of the steering shaft 3 accompanying the steering operation is converted into a reciprocating linear motion of the rack shaft 5 by the rack and pinion mechanism 4. The steering shaft 3 of this embodiment is formed by connecting a column shaft 3a, an intermediate shaft 3b, and a pinion shaft 3c. Then, the reciprocating linear motion of the rack shaft 5 accompanying the rotation of the steering shaft 3 is transmitted to a knuckle (not shown) via tie rods 6 connected to both ends of the rack shaft 5, whereby the steering angle of the steered wheels 7. That is, the traveling direction of the vehicle is changed.

  Further, the EPS 1 includes an EPS actuator 10 as a steering force assisting device that applies an assist force for assisting a steering operation to the steering system, and an ECU 11 as a control unit that controls the operation of the EPS actuator 10. .

  The EPS actuator 10 of the present embodiment is configured as a so-called column-type EPS actuator in which a motor 12 that is a drive source is drivingly connected to a column shaft 3 a via a speed reduction mechanism 13. In the present embodiment, the motor 12 employs a brushless motor that rotates based on three-phase (U, V, W) driving power. The EPS actuator 10 is configured to apply an assist force based on the motor torque to the steering system by decelerating the rotation of the motor 12 and transmitting it to the column shaft 3a.

  On the other hand, a torque sensor 14 is connected to the ECU 11, and the ECU 11 detects a steering torque τ transmitted through the steering shaft 3 based on an output signal of the torque sensor 14. Further, the ECU 11 of the present embodiment receives the left and right wheel speeds Wr, Wl detected by the wheel speed sensor 15 and the vehicle speed V detected based on the wheel speeds Wr, Wl. Then, the ECU 11 of the present embodiment calculates a target assist force to be applied to the steering system based on each of these state quantities, and supplies drive power to generate a motor torque corresponding to the target assist force. The operation of the EPS actuator 10 using the motor 12 as a drive source, that is, the assist force applied to the steering system is controlled (power assist control).

Next, the electrical configuration of the EPS of this embodiment will be described.
FIG. 2 is a control block diagram of the EPS of this embodiment. As shown in the figure, the ECU 11 supplies three-phase drive power to the motor 12 based on the microcomputer 17 serving as motor control signal output means for outputting a motor control signal and the motor control signal output from the microcomputer 17. And a drive circuit 18.

  Each control block shown below is realized by a computer program executed by the microcomputer 17. The microcomputer 17 detects each state quantity at a predetermined sampling period, and generates a motor control signal by executing each arithmetic processing shown in the following control blocks at every predetermined period.

  As shown in FIG. 3, the drive circuit 18 of the present embodiment is formed by connecting a plurality of FETs 18 a to 18 f as switching elements. Specifically, each series circuit of FETs 18a and 18d, FETs 18b and 18e, and FETs 18c and 18f is connected in parallel. In the present embodiment, N-channel MOSFETs are used for these FETs 18a to 18f. And each connection point of the switching element pair connected in series, that is, each connection point 19u, 19v, 19w of FET18a, 18d, FET18b, 18e, FET18c, 18f is respectively via power line 20u, 20v, 20w. The motor 12 is connected to each phase motor coil 12u, 12v, 12w.

  As described above, the drive circuit 18 of the present embodiment connects the three switching arms 18u, 18v, and 18w corresponding to each phase in parallel with a pair of switching elements connected in series as a basic unit (switching arm). This is configured as a known PWM inverter. And it is the structure which act | operates by the input of the said motor control signal, and supplies the three-phase drive electric power based on the applied power supply voltage V_pig to a motor.

  That is, the motor control signal output from the microcomputer 17 defines the on / off state of each phase switching element constituting the drive circuit 18 (duty of each phase switching arm). The FETs 18a to 18f are turned on / off in response to the motor control signal, and the energization pattern for each phase motor coil 12u, 12v, 12w is switched, so that the power supply voltage V_pig applied to the drive circuit 18 is three. It is converted into driving power of phases (U, V, W) and output to the motor 12.

  Further, the ECU 11 is provided with a current sensor 21 (21u, 21v, 21w) for detecting each phase current value Iu, Iv, Iw of the motor 12. Specifically, the current sensors 21u, 21v, 21w are connected to the low potential side (ground side, lower side in FIG. 3) of the switching arms 18u, 18v, 18w corresponding to the phases of the motor 12, respectively. Each is formed by connecting a shunt resistor. And the microcomputer 17 of this embodiment is based on the output signal (voltage between terminals of shunt resistance) of each of these current sensors 21u, 21v, 21w, the phase current value Iu, which flows through each phase motor coil 12u, 12v, 12w, It is configured to detect Iv and Iw.

  Further, the microcomputer 17 of the present embodiment detects the actual rotation angle (motor rotation angle θm) of the motor 12 based on the output signal of the motor resolver 23 (see FIGS. 1 and 2). In the present embodiment, the motor resolver 23 outputs a two-phase sine wave signal (sine signal S_sin and cosine signal S_cos) whose amplitude changes according to the motor rotation angle θm (electrical angle) as the sensor signal. A winding type resolver is used. The microcomputer 17 according to the present embodiment executes the current feedback control based on the phase current values Iu, Iv, Iw and the motor rotation angle θm of the motor 12 to output the motor control to the drive circuit 18. Generate a signal.

  More specifically, as shown in FIG. 2, the motor control unit 24 of the microcomputer 17 has phase voltage command values Vu *, Vv *, Vw to be applied to each phase of the motor 12 by executing current control in the rotating coordinate system. * A first control unit 25 and a second control unit 26 for calculating (Vu **, Vv **, Vw **), and a PWM conversion unit 27 for converting the phase voltage command value into a motor control signal are provided. ing. The microcomputer 17 according to the present embodiment is configured to output the motor control signal generated by the motor control unit 24 to the drive circuit 18.

  As shown in FIG. 4, the first control unit 25 calculates a current command value corresponding to the target assist force based on the steering torque τ and the vehicle speed V acquired by the ECU 11 as described above. It has. Further, the first control unit 25 includes a d / q conversion unit 32, and the d / q conversion unit 32 detects the phase current values Iu, Iv, and Iw by the motor resolver 23. The d-axis current value Id and the q-axis current value Iq are calculated by mapping on the orthogonal coordinates of the rotation coordinate system (d / q coordinate system) according to the angle θm, that is, the actual rotation angle of the motor 12. And the 1st control part 25 performs phase feedback command value Vu *, Vv *, Vw * which shows the voltage which should be impressed to each phase of motor 12 by performing current feedback control in the d / q coordinate system. It is configured to calculate.

  That is, the current command value calculation unit 31 calculates the q-axis current command value Iq * as the current command value. Specifically, the current command value calculation unit 31 calculates a q-axis current command value Iq * that generates a larger assist force as the input steering torque τ is larger and the vehicle speed V is smaller. . The d-axis current command value Id * is fixed to “0” (Id * = 0). The d-axis current command value Id * and the q-axis current command value Iq *, together with the d-axis current value Id and the q-axis current value Iq output from the d / q conversion unit 32, the corresponding subtractors 33d and 33q. Is input.

  Next, the current deviations ΔId and ΔIq of the respective axes calculated by the subtracters 33d and 33q are input to the corresponding F / B control units (feedback control units) 34d and 34q, respectively. Each of the F / B control units 34d and 34q executes d / q by executing a feedback control calculation based on the input current deviations ΔId and ΔIq and a predetermined feedback gain (proportional: P, integral: I). A d-axis voltage command value Vd * and a q-axis voltage command value Vq *, which are voltage command values in the coordinate system, are calculated.

  Specifically, each of the F / B control units 34d and 34q sets a proportional component obtained by multiplying the input current deviations ΔId and ΔIq by a proportional gain, and an integral value of the current deviations ΔId and ΔIq, respectively. The integral component obtained by multiplying the integral gain is calculated. Then, the d-axis voltage command value Vd * and the q-axis voltage command value Vq * are generated by adding the proportional component and the integral component.

  Next, the d-axis voltage command value Vd * and the q-axis voltage command value Vq * are mapped onto three-phase (U, V, W) AC coordinates in the d / q inverse conversion unit 35. The first control unit 25 is configured to output the phase voltage command values Vu *, Vv *, Vw * obtained by the reverse conversion executed by the d / q reverse conversion unit 35 to the PWM conversion unit 27. ing.

  On the other hand, as shown in FIG. 5, the second control unit 26 calculates the addition angle θa (θa ′) corresponding to the motor rotation angle change amount for each calculation cycle, and the addition angle θa. And a control angle calculation unit 42 that calculates a control angle θc as a virtual motor rotation angle in terms of control by accumulating every calculation cycle. And the 2nd control part 26 performs phase feedback command value Vu **, Vv **, Vw ** by performing electric current feedback control in the rotation coordinate system ((gamma) / (delta) coordinate system) according to the control angle (theta) c. It is configured to calculate.

  More specifically, the steering torque τ and the vehicle speed V acquired by the ECU 11 as described above are input to the addition angle calculation unit 41 of the present embodiment. Further, the microcomputer 17 of the present embodiment includes a steering angle estimation calculation unit 43 that estimates the steering angle θs generated in the steering 2 based on the left and right wheel speeds Wr and Wl detected by the wheel speed sensor 15. The estimated steering angle θs is input to the addition angle calculation unit 41 (see FIG. 2). Note that the steering angle estimation calculation unit 43 of the present embodiment converts the steering angle (steering angle θt) of the steered wheels obtained by substituting the left and right wheel speeds Wr and Wl into the well-known arithmetic expressions shown below. Thus, the steering angle θs is estimated.

θt = (2 × WB × (W1−Wr)) / (RW × (W1 + Wr)) × (180 / π)
... (1)
In the equation (1), “WB” is the wheel base length of the vehicle, and “RW” is the tread length of the vehicle.

  Furthermore, the addition angle calculation unit 41 includes a target torque calculation unit 45 that calculates a target torque τ * corresponding to the target value of the steering torque τ based on the steering angle θs generated in the steering 2 and the vehicle speed V. The target torque τ * calculated by the target torque calculation unit 45 is input to the subtractor 46 together with the steering torque τ. Then, the addition angle calculation unit 41 of this embodiment calculates the addition angle θa based on the torque deviation Δτ obtained by subtracting the target torque τ * from the steering torque τ.

  That is, in EPS 1 that applies assist force based on motor torque to the steering system, the target torque τ * is a parameter corresponding to the motor torque that the motor 12 should generate, and the steering torque τ corresponds to the actual torque of the motor 12. It is a parameter to do. Then, the addition angle calculation unit 41 of the present embodiment executes torque feedback control based on the torque deviation Δτ, thereby adding an addition angle corresponding to the torque deviation between the target torque to be generated by the motor 12 and the actual torque. Calculate θa.

  Specifically, the torque deviation Δτ calculated by the subtractor 46 is input to the F / B control unit 47. Then, the F / B control unit 47 calculates a proportional component obtained by multiplying the torque deviation Δτ by a proportional gain, and an addition value of an integral component obtained by multiplying the integral value of the torque deviation Δτ by an integral gain. Calculation is performed as the first change component dθτ of the motor rotation angle in each calculation cycle.

  In the present embodiment, the second control unit 26 is provided with a rotation angular velocity estimation calculation unit 50 that estimates the motor rotation angular velocity, and the addition angle calculation unit 41 includes the rotation angular velocity estimation calculation unit 50. The estimated motor rotation angular velocity (motor rotation angular velocity estimation value ωm_e) is input as the second change component dθω of the motor rotation angle in each calculation cycle. Then, the addition angle calculation unit 41 of the present embodiment calculates the addition angle θa by using the first change component dθτ based on the torque deviation Δτ and the second change component dθω based on the estimated motor rotation angular velocity value ωm_e. .

  More specifically, the second controller 26 includes phase voltage command values Vu *, Vv *, Vw * (Vu **, Vv **, Vw *) used when the PWM converter 27 generates a motor control signal. An internal command value corresponding to *), that is, Duty is input. Further, the ECU 11 of the present embodiment is provided with a voltage sensor 51 as voltage detection means for detecting the power supply voltage V_pig applied to the drive circuit 18 (see FIG. 2). The second control unit 26 includes a phase voltage calculation unit 52 that calculates the phase voltage values Vu, Vv, and Vw of the motor 12 based on the detected power supply voltage V_pig and the duty.

  In the second control unit 26, the phase voltage values Vu, Vv, Vw and the phase current values Iu, Iv, Iw of the motor 12 detected by the current sensor 21 are further converted to α / β conversion as conversion means. The unit 53 converts the α-axis voltage value Vα and β-axis voltage value Vβ, and the α-axis current value Iα and β-axis current value Iβ of the two-phase fixed coordinate system (α / β coordinate system), respectively. Then, the rotational angular velocity estimation calculation unit 50 as the motor rotational angular velocity estimating means converts the motor voltage and motor current indicated by the α-axis voltage value Vα and β-axis voltage value Vβ and the α-axis current value Iα and β-axis current value Iβ. Based on this, the motor rotational angular velocity estimated value ωm_e is calculated.

  More specifically, the rotational angular velocity estimation calculation unit 50 of this embodiment includes a disturbance observer 54 that estimates an induced voltage generated in the motor 12 as a disturbance based on the motor model.

  As shown in FIG. 6, in the block diagram, the motor 12 is represented by a motor model M1 that generates a motor current (Iα, Iβ) based on the motor voltage (Vα, Vβ) and the induced voltage (Eα, Eβ). The Therefore, the induced voltage estimated value Eα_e is obtained by the inverse motor model M2 having the motor current (Iα, Iβ) as an input, and the subtractor 55 having the output of the inverse motor model M2 and the motor voltage (Vα, Vβ) as inputs. , Eβ_e as a disturbance can be formed. For example, if the motor model M1 is “1 / (R + pL)”, the reverse motor model M2 is “R + pL” (where R: armature winding resistance, L: inductance, p: differential operator). Then, the rotational angular velocity estimation calculation unit 50 of the present embodiment is based on the respective axis induced voltage estimated values (induced voltage estimated values Eα_e, Eβ_e) of the two-phase fixed coordinate system output from the disturbance observer 54 as the induced voltage estimating means. Thus, the estimated motor rotation angular velocity value ωm_e is calculated.

That is, the induced voltages (Eα, Eβ) in the α / β coordinate system are expressed by the following equations (2) and (3), respectively. In each equation, “Ke” is an induced voltage constant, and “ω” is a motor rotation angular velocity.
Eα = −Ke × ω × sinθ (2)
Eβ = Ke × ω × cosθ (3)
Furthermore, the following equation (4) is obtained by solving these equations (2) and (3) with respect to the angle “θ”. In the formula, “arctan” is “tangent inverse function: arc tangent”.

θ = arctan (−Eα / Eβ) (4)
Therefore, the motor rotation angle (motor rotation angle estimated value θm_e) can be estimated from the induced voltage estimated values Eα_e and Eβ_e output from the disturbance observer 54. Then, the rotational angular velocity estimation calculation unit 50 of the present embodiment calculates the motor rotational angular velocity estimated value ωm_e by differentiating the motor rotational angle estimated value θm_e, that is, by obtaining the amount of change per unit time.

  Specifically, as shown in the flowchart of FIG. 7, when the rotational angular velocity estimation calculation unit 50 estimates the induced voltage of the motor 12 by the disturbance observer 54 (Eα_e, Eβ_e, step 101), first, the induced voltage estimation is performed. Filter processing is performed on the values Eα_e and Eβ_e (LPF: low-pass filter, step 102). Next, the rotation angular velocity estimation calculation unit 50 calculates the motor rotation angle estimated value θm_e from the induced voltage estimated values Eα_e and Eβ_e by using the above equation (4) (rotation angle estimation, step 103), and further. Then, the estimated motor rotation angle estimation value θm_e is differentiated to calculate the motor rotation angular speed estimation value ωm_e (rotation angular velocity estimation, step 104). The estimated motor rotation angular velocity value ωm_e is output to the addition angle calculation unit 41 as the second change component dθω of the motor rotation angle in each calculation cycle (step 105).

  As shown in FIG. 5, in the addition angle calculation unit 41 of the present embodiment, the first change component dθτ of the motor rotation angle based on the torque deviation Δτ calculated by the F / B control unit 47 and the rotation angular velocity estimation calculation unit The second change component dθω of the motor rotation angle based on the estimated motor rotation angular velocity value ωm_e calculated at 50 is input to the addition angle adjustment calculation unit 58. In this embodiment, the rotational angular velocity estimation calculation unit 50 calculates the square sum (induced voltage square sum Esq_αβ) of the induced voltage estimated values Eα_e and Eβ_e output from the disturbance observer 54 (Esq_αβ = (Eα_e) ^ 2+ (Eβ_e) ^ 2, where “^ 2” indicates a square), and the induced voltage square sum Esq_αβ is output to the addition angle adjustment calculation unit 58. Then, according to the value of the induced voltage square sum Esq_αβ, the addition angle calculation unit 41 of the present embodiment performs the second change based on the first change component dθτ based on the torque deviation Δτ and the estimated motor rotation angular velocity ωm_e. The calculation form of the addition angle θa using the component dθω is changed.

  More specifically, the addition angle adjustment calculation unit 58 of the present embodiment compares the input induced voltage square sum Esq_αβ with a predetermined threshold value (E0). When the induced voltage square sum Esq_αβ exceeds the threshold value (E0), the addition value of the first change component dθτ based on the torque deviation Δτ and the second change component dθω based on the motor rotational angular velocity estimated value ωm_e is added. When the angle θa is equal to or smaller than the threshold value (E0), the first change component dθτ based on the torque deviation Δτ is set as the addition angle θa.

  In other words, the motor rotation angular velocity having one calculation cycle as a basic unit has an equivalent meaning to the motor rotation angle change amount per one calculation cycle. The estimation of the induced voltage based on the motor current and the motor voltage using the disturbance observer 54 as described above ensures higher accuracy in the high-speed rotation region where the induced voltage increases.

  In view of this point, in the present embodiment, by comparing the induced voltage square sum Esq_αβ and the threshold value (E0), the rotational state of the motor 12 is calculated from the motor rotation angular velocity estimated value ωm_e as the second change component dθω of the motor rotation angle. It is determined whether it exists in the high-speed rotation area | region where the estimation precision which can be utilized as this is ensured. The second variation component dθω based on the estimated motor rotational angular velocity value ωm_e is used only when the required estimation accuracy is in a high-speed rotation region.

  Next, the addition angle θa output from the addition angle adjustment calculation unit 58 is input to the addition angle restriction unit 59. Then, the addition angle calculation unit 41 of the present embodiment outputs the addition angle θa (θa ′) after the addition angle limitation processing (described later) is performed in the addition angle limitation unit 59 to the control angle calculation unit 42. It is the composition to do.

  That is, as shown in the flowchart of FIG. 8, the addition angle calculation unit 41 of the present embodiment acquires the vehicle speed V, the steering angle θs, and the steering torque τ as the state quantities used for the calculation of the first change component dθτ. (Step 201) First, based on the steering angle θs and the vehicle speed V, the target torque τ * is calculated (Step 202). Next, the addition angle calculation unit 41 calculates the torque deviation Δτ by taking the difference between the target torque τ * and the steering torque τ (step 203). Then, the first change component dθτ is calculated by executing a torque feedback control calculation based on the torque deviation Δτ (step 204).

  Further, when the addition angle calculation unit 41 obtains the second change component dθω based on the motor rotation angular velocity estimation value ωm_e calculated by the rotation angular velocity estimation calculation unit 50 as described above (step 205), subsequently, the rotation angular velocity estimation The induced voltage square sum Esq_αβ is obtained from the arithmetic unit 50 (step 206). Then, the addition angle calculation unit 41 determines whether or not the induced voltage square sum Esq_αβ exceeds the threshold value E0 (step 207).

  Next, when it is determined in this step 207 that the induced voltage square sum Esq_αβ exceeds the threshold value E0 (Esq_αβ> E0, step 207: YES), the addition angle calculation unit 41 continues to already generate the induced voltage square sum. It is determined whether or not an excess flag indicating that Esq_αβ exceeds the threshold value E0 is set (step 208). If the excess flag is not set (step 208: NO), the excess flag is set (step 209), and the value of the first change component dθτ calculated in step 201 is cleared (dθτ = 0, step 210).

  If the excess flag has already been set in step 208 (step 208: YES), the processing in steps 209 and 210 is not executed. When it is determined in step 207 that the induced voltage square sum Esq_αβ exceeds the threshold value E0 (step 207: YES), the first based on the torque deviation Δτ regardless of the excess flag. The addition angle θa is calculated by adding the change component dθτ and the second change component dθω based on the estimated motor rotational angular velocity value ωm_e (step 211).

  On the other hand, when it is determined in step 207 that the induced voltage square sum Esq_αβ is equal to or less than the threshold value E0 (Esq_αβ ≦ E0, step 207: NO), the addition angle calculation unit 41 also determines whether the excess flag is set. (Step 212), if it is set (step 212: YES), the excess flag is reset (step 213). If the excess flag is not set (step 212: NO), the process of step 213 is not executed. Then, the first change component dθτ calculated in step 204 is calculated as the addition angle θa (step 214).

  That is, the first change component dθτ based on the torque deviation Δτ is a value corresponding to the magnitude of the deviation between the actual rotation angle of the motor 12 and the virtual motor rotation angle in control. Therefore, the value is less affected by the motor rotation state than the second change component dθω based on the estimated motor rotation angular velocity value ωm_e.

  In consideration of this point, in the present embodiment, as described above, when the motor rotation state is in the low speed region (Esq_αβ ≦ E0, Step 207: NO), the first change component dθτ is set as the addition angle θa. The first change component dθτ is cleared in the first calculation cycle (step 207: YES and step 208: NO) in which the addition angle θa is calculated using the second change component dθω based on the estimated motor rotational angular velocity value ωm_e. (Step 209) is because the first change component dθτ reflects the state of the previous calculation cycle in which the second change component dθω was not used. In the present embodiment, this makes it possible to calculate the addition angle with high accuracy regardless of the motor rotation state.

  Further, when the addition angle calculator 41 calculates the addition angle θa by adding the first change component dθτ and the second change component dθω in step 211, the addition angle θa (the absolute value thereof) is the second value. The limiting process of the addition angle θa is executed so as not to exceed the value obtained by adding the predetermined threshold value b to the change component dθω (absolute value thereof) (first limiting process: | θa | ≦ | dθω | + b, step 215). . In step 214, if the first change component dθτ is the addition angle θa, the processing in step 215 is not executed. And the addition angle | corner calculating part 41 of this embodiment further performed the 2nd restriction | limiting process (2nd restriction | limiting process: | (theta) a | <= B, step 216) so that the value may not exceed the predetermined threshold value b. The subsequent addition angle θa (θa ′) is output to the control angle calculator 42.

  On the other hand, as shown in FIG. 5, the control angle calculation unit 42 holds the previous value of the control angle θc calculated in the previous calculation cycle in a storage area (not shown), and sets the addition angle θa to the previous value. A new control angle θc is calculated by addition. Then, by updating the previous value stored in the storage area with the new control angle θc, the control angle θc is calculated by adding the addition angle θa for each calculation cycle. Yes.

  In the second control unit 26, the control angle θc as the virtual motor rotation angle for control calculated in this way is the two-phase fixed coordinate system (α / β coordinate) output by the α / β conversion unit 53. System) and α-axis current value Iα and β-axis current value Iβ. Then, the γ / δ conversion unit 60 maps the α-axis current value Iα and the β-axis current value Iβ onto the rotation coordinate system according to the control angle θc, that is, the orthogonal coordinate of the γ / δ coordinate system, thereby The γ-axis current value Iγ and the δ-axis current value Iδ are calculated as actual current values in the γ / δ coordinate system.

The second control unit 26 includes a current command value calculation unit 61 that calculates a γ-axis current command value Iγ * and a δ-axis current command value Iδ * as the current command value of the γ / δ coordinate system.
More specifically, the current command value calculation unit 61 of the present embodiment receives the torque deviation Δτ and the target torque τ * calculated by the addition angle calculation unit 41. Then, the current command value calculation unit 61 calculates the γ-axis current command value Iγ * and the δ-axis current command value Iδ * based on the torque deviation Δτ and the target torque τ *.

  As shown in FIG. 9, the current command value calculation unit 61 of this embodiment calculates an increase / decrease value (γ-axis current increase / decrease value η) of the γ-axis current command value Iγ * in each calculation cycle based on the torque deviation Δτ. A γ-axis current increase / decrease value calculation unit 62, and an integration control unit 63 that integrates the γ-axis current increase / decrease value η for each calculation cycle. The current command value calculation unit 61 is configured to output the control output of the integration control unit 63 as the γ-axis current command value Iγ * and output “0” as the δ-axis current command value Iδ *.

  More specifically, the γ-axis current increase / decrease value calculation unit 62 of the present embodiment is provided with two maps (62a, 62b) in which the torque deviation Δτ and the γ-axis current increase / decrease value η are associated. In the present embodiment, the first map 62a is formed corresponding to the case where the sign (direction) of the target torque τ * is positive (τ *> 0). In the first map 62a, the γ-axis current increase / decrease value η is set so as to increase linearly as the torque deviation Δτ increases. On the other hand, the second map 62b is formed corresponding to the case where the sign of the target torque τ * is negative (τ * <0). In the second map 62b, the γ-axis current increase / decrease value η is set to linearly decrease as the torque deviation Δτ increases.

  In the present embodiment, the first map 62a and the second map 62b both have a value of the torque deviation Δτ within a predetermined range (−A <Δτ <A) centered on a point where the torque deviation Δτ is zero. Regardless of this, a dead zone is set such that the γ-axis current increase / decrease value η is “0”. Then, the γ-axis current increase / decrease value calculation unit 62 of the present embodiment switches the map to be referred to according to the sign of the target torque τ *, and based on the input torque deviation Δτ, the γ-axis current in each calculation cycle. The increase / decrease value η is calculated.

  On the other hand, the integration control unit 63 holds the control output in the previous calculation cycle, that is, the previous value of the γ-axis current command value Iγ * in a storage area (not shown), and adds the γ-axis current increase / decrease value η to the previous value. Is added to calculate a new γ-axis current command value Iγ *. Then, by updating the previous value held in the storage area with the new γ-axis current command value Iγ *, the γ-axis current command value obtained by integrating the γ-axis current command value Iγ * is calculated for each calculation cycle. It is configured to execute the calculation of Iγ *.

  As shown in FIG. 5, the γ-axis current command value Iγ * calculated by the current command value calculation unit 61 in this way is input to the corresponding subtracter 64a together with the γ-axis current value Iγ. Similarly, the δ-axis current command value Iδ * is also input to the corresponding subtracter 64b together with the δ-axis current value Iδ. The current deviations ΔIγ and ΔIδ calculated by the subtracters 64a and 64b are input to the corresponding F / B control units 65a and 65b, respectively.

  Next, each F / B control unit 65a, 65b executes a feedback control calculation based on the current deviations ΔIγ, ΔIδ and a predetermined feedback gain (proportional: P, integral: I), thereby obtaining a γ / δ coordinate system. The γ-axis voltage command value Vγ * and the δ-axis voltage command value Vδ * which are the voltage command values are calculated. The feedback control calculation performed by each of the F / B controllers 65a and 65b is the same as that of the F / B controllers 34d and 34q on the first controller 25 side. Is omitted.

  Further, the γ-axis voltage command value Vγ * and the δ-axis voltage command value Vδ * are mapped onto the three-phase (U, V, W) AC coordinates in the two-phase / three-phase converter 66. The second control unit 26 is configured to output the phase voltage command values Vu **, Vv **, and Vw ** generated by the two-phase / three-phase conversion unit 66 to the PWM conversion unit 27. ing. For details of the principle of resolverless control executed by the second control unit 26 as described above, refer to, for example, the descriptions in Patent Document 3 and Patent Document 4 above.

  As shown in FIG. 2, the microcomputer 17 of the present embodiment includes a rotation angle abnormality detection unit 68 that detects an abnormality in the motor rotation angle θm detected by the motor resolver 23 as an actual rotation angle. Specifically, the rotation angle abnormality detection unit 68 of this embodiment determines whether or not the sum of squares of the sine signal S_sin and the cosine signal S_cos output from the motor resolver 23 is within an appropriate range. Based on the determination result, an abnormality in the motor rotation angle θm, which is the actual rotation angle, is detected. For details of such rotation angle abnormality detection, refer to, for example, the description of JP-A-2006-177750.

  Furthermore, in the present embodiment, the result of abnormality detection by the rotation angle abnormality detection unit 68 is input to the motor control unit 24 as a rotation angle abnormality detection signal S_rsf. When the motor rotation angle θm detected by the motor resolver 23 as the actual rotation angle is not abnormal, the motor control unit 24 of the present embodiment, the phase voltage command value Vu * calculated by the first control unit 25, When a motor control signal is output based on Vv * and Vw * and an abnormality occurs in the motor rotation angle θm, the phase voltage command values Vu **, Vv **, A motor control signal is output based on Vw **.

  That is, as described above, the second control unit 26 does not use the motor rotation angle θm detected by the motor resolver 23 as the actual rotation angle, but uses the control angle θc that is a virtual motor rotation angle for control. The phase voltage command values Vu **, Vv **, and Vw ** are calculated. In this embodiment, by generating a motor control signal based on the phase voltage command values Vu **, Vv **, and Vw ** calculated by the second control unit 26, the motor rotation that is the actual rotation angle. Even after an abnormality is detected in the angle θm, the motor control can be continued stably.

(Energization failure detection)
Next, the manner of detecting energization failure in this embodiment will be described.
As shown in FIG. 2, the microcomputer 17 of the present embodiment has an energization failure that detects an energization failure that has occurred in the power supply path for supplying drive power to the motor 12, that is, a state in which no current flows through any of the phases. A detection unit 71 is provided. In addition, as a mode of the energization failure, the power lines 20u, 20v, 20b, and 12w connecting the drive circuit 18 (the switching arms 18u, 18v, and 18w of each phase) and the phase motor coils 12u, 12v, and 12w, For example, a disconnection failure of 20w or an open failure (open fixing failure) of each of the FETs 18a to 18f constituting the drive circuit 18 can be mentioned. In addition, the motor control unit 24 is inputted with a determination result by the energization failure detection unit 71 as an energization failure detection signal S_pde. Then, the motor control unit 24 of the present embodiment indicates that the motor control unit 24 should stop the motor 12 when the above-described energization failure is detected by the energization failure detection signal S_pde. Is configured to promptly make the fail safe.

  More specifically, in the present embodiment, the energization failure detection unit 71 detects each phase current value Iu, Iv, Iw of the motor 12 detected by the current sensor 21 and an actual rotation angle by the motor resolver 23. The actual motor rotation angular velocity (actual rotation angular velocity ωm) calculated based on the motor rotation angle θm is input. In the present embodiment, the actual rotational angular velocity ωm is calculated by differentiating the motor rotational angle θm. Further, the energization failure detection unit 71 has internal command values corresponding to phase voltage command values Vu *, Vv *, Vw * (Vu **, Vv **, Vw **) used when generating a motor control signal. That is, the duty (Du, Du, Dw) and the power supply voltage V_pig applied to the drive circuit 18 detected by the voltage sensor 51 are input. And the electricity supply failure detection part 71 as an abnormality detection means detects generation | occurrence | production of the electricity supply failure for every phase based on the mutual relationship of these each state quantity.

  Specifically, the energization failure detection unit 71 determines one phase of “U, V, W” as a specific phase (X phase) and determines whether or not there is an abnormality in the specific phase. And when there is no abnormality in the said phase, it becomes the structure which detects the generation | occurrence | production of the electrical conduction failure sequentially for every phase by changing the specific phase made into the determination object.

  That is, as shown in the flowchart of FIG. 10, the energization failure detection unit 71 first determines whether or not there is an abnormality in the U phase with the U phase as a specific phase (step 301), and the abnormality is confirmed. In (Step 301: YES), the occurrence of an energization failure in the U phase is detected (Step 302). On the other hand, when there is no abnormality in the U phase (step 301: NO), it is then determined whether there is an abnormality in the V phase with the V phase as the specific phase (step 303), and when the abnormality is confirmed (step 303) 303: YES), the occurrence of an energization failure in the V phase is detected (step 304). Further, when there is no abnormality in the V phase (step 303: NO), it is next determined whether or not there is an abnormality in the W phase with the W phase as a specific phase (step 305), and when the abnormality is confirmed (step) (305: YES), the occurrence of an energization failure in the W phase is detected (step 306). If it is determined in step 305 that there is no abnormality in the W phase (step 305: NO), it is determined that there is no abnormality in each phase (step 307).

  And the electricity supply failure detection part 71 of this embodiment detects the generation | occurrence | production of the electricity supply failure sequentially for every phase by performing the process of this step 301-step 307 with a predetermined calculation period.

  Note that the energization failure detection unit 71 according to the present embodiment compares the power supply voltage V_pig with a predetermined threshold value when executing the above energization failure determination, so that the power supply voltage V_pig has an appropriate value (for example, 9 V or more). It is determined whether or not it has. When the voltage drops such that the power supply voltage V_pig falls below an appropriate value, the energization failure determination shown in steps 301 to 307 is not executed.

  More specifically, as shown in the flowchart of FIG. 11, the energization failure detection unit 71 first performs each energization failure determination for the specific phase (X phase) selected as the detection target phase. The state quantity, that is, the phase current value Ix and Duty (Dx) of the specific phase, and the actual rotational angular velocity ωm of the motor 12 are acquired (step 401).

  Next, the energization failure detection unit 71 determines whether or not the duty (Dx) of the specific phase is a value at which the specific phase should be energized. Specifically, the energization failure detection unit 71 of the present embodiment is such that the Duty (Dx) is clearly a value indicating that the specific phase should be in the energized state, that is, the lower limit value D_lo in the normal control or the upper limit. It is determined whether or not the value is larger than the value D_hi (step 402). When the determination condition of step 402 is satisfied (Dx <D_lo or Dx> D_hi, step 402: YES), the phase current value Ix (absolute value) of the specific phase is a value indicating a non-energized state. It is determined whether or not. Specifically, whether the phase current value Ix of the specific phase is smaller than a predetermined threshold value I1 set on the basis of “0”, which is a theoretical value indicating a non-energized state in consideration of the detection error. It is determined whether or not (step 403).

  Further, when the current failure detection unit 71 determines in step 403 that the phase current value Ix is smaller than the threshold value I1 (| Ix | <I1, step 403: YES), the actual rotational angular velocity ωm (of It is determined whether or not the absolute value is smaller than a predetermined threshold value ω0 set based on the maximum rotational speed of the motor 12 in consideration of the detection error (step 404). If it is determined that the actual rotational angular velocity ωm is smaller than the threshold ω0 (ωm <ω0, step 404: YES), that is, if the motor current is not in the high-speed rotation region where the motor current is minimized due to the influence of the induced voltage, It is determined that an abnormality indicating the occurrence of an energization failure has occurred in the X phase, which is a phase (step 405).

  As described above, the current-carrying failure detection unit 71 of the present embodiment has an abnormality indicating the occurrence of current-carrying failure in the X phase, which is the specific phase, only when all the determination conditions of Steps 402 to 404 are satisfied. Is determined. If any of the above conditions from Step 402 to Step 404 is not satisfied (D_lo ≦ Dx ≦ D_hi, Step 402: NO, or | Ix | ≧ I1, Step 403: NO, or ωm ≧ ω0, Step 404: NO ) Is determined that there is no abnormality in the X phase, which is the specific phase (step 406).

  That is, basically, although the specific phase (X phase) that is the detection target phase is a phase that should be in the energized state (step 402: YES), the phase current value Ix of the specific phase is in the non-energized state. In the case of indicating (step 403: YES), it is possible to determine that an abnormality indicating an energization failure has occurred in the specific phase due to the contradiction. In this embodiment, by adding the speed condition of step 404 to this and excluding the high-speed rotation region in which the motor current is minimized due to the influence of the induced voltage, it is possible to accurately detect the conduction failure. It has become.

  Further, as shown in FIG. 2, the conduction failure detection unit 71 of the present embodiment includes each state quantity used for the conduction failure determination shown in the flowchart of FIG. 11, that is, each phase current value Iu, Iv, Iw, actual rotational angular velocity. The rotation angle abnormality detection signal S_rsf output from the rotation angle abnormality detector 68 is input together with ωm, Duty (Du, Du, Dw), and the power supply voltage V_pig. Further, the motor energization failure detection unit 71 receives the motor rotation angular velocity estimation value ωm_e calculated by the rotation angular velocity estimation calculation unit 50 and the square sum of the induced voltage estimation values Eα_e and Eβ_e, that is, the induced voltage square sum Esq_αβ. . Then, when an abnormality occurs in the motor rotation angle θm detected by the motor resolver 23 as the actual rotation angle, the energization failure detection unit 71 of the present embodiment uses the state quantities as described above. 2 Even when the resolverless control executed by the control unit 26 is executed, the detection of the energization failure can be continued.

  That is, when the motor rotation angle θm detected by the motor resolver 23 is not normal as the actual rotation angle, the accuracy of the energization failure determination using the actual rotation angular velocity ωm calculated based on the motor rotation angle θm is also lowered. become.

  In consideration of this point, the conduction failure detection unit 71 of the present embodiment, when executing the conduction failure determination, as shown in the flowchart of FIG. Based on S_rsf, it is determined whether or not the motor rotation angle θm detected by the motor resolver 23 as the actual rotation angle is normal (step 501). Then, based on the determination result, the execution mode of the energization failure determination is switched.

  That is, when the motor rotation angle θm as the actual rotation angle is normal (step 501: YES), normal energization failure determination using the actual rotation angular velocity ωm as shown in the flowchart of FIG. 11 is executed (step 501). 502). If it is determined that the actual rotation angle is not normal (step 501: NO), an alternative energization failure determination based on the motor rotation angular velocity estimated value ωm_e and the induced voltage square sum Esq_αβ is executed (step 503).

  More specifically, as shown in the flowchart of FIG. 13, the energization failure detection unit 71 of the present embodiment includes the phase current value Ix of the specific phase (X phase) as a state quantity for executing the alternative energization failure determination. The motor rotational angular velocity estimated value ωm_e and the induced voltage square sum Esq_αβ are acquired (step 601).

  Next, the energization failure detection unit 71 first determines whether or not the induced voltage square sum Esq_αβ exceeds a predetermined threshold E1 (step 602). If it is determined in step 602 that the induced voltage square sum Esq_αβ exceeds the predetermined threshold value E1 (step 602: YES), is the motor rotational angular velocity estimated value ωm_e lower than the predetermined threshold value ω1? It is determined whether or not (step 603).

  Here, the threshold value E1 in step 602 is set corresponding to the maximum region indicating that the motor 12 is in the highest speed rotation state in consideration of the error. The threshold E1 is higher than the threshold E0 (see FIG. 8, step 207) for determining whether or not the estimated motor rotation angular velocity value ωm_e can be used as the second change component dθω of the motor rotation angle. Is set. In addition, the threshold value ω1 in step 603 is set in correspondence with the minimum region based on “0”, which is a theoretical value indicating a non-rotation state in consideration of an error. Then, when it is determined in step 603 that the motor rotation angular velocity estimation value ωm_e is lower than the threshold ω1 (step 603: YES), the energization failure detection unit 71 of the present embodiment enters any phase of the motor 12. It is determined that an abnormality indicating the occurrence of an energization failure has occurred (step 604).

  In step 602, when the induced voltage square sum Esq_αβ is equal to or less than the threshold E1 (Esq_αβ ≦ E1, step 602: NO), or when it is determined that the estimated motor rotation angular velocity ωm_e is equal to or greater than the threshold ω1 (ωm_e ≧ In ω1, step 603: NO), it is determined that the motor 12 has no abnormality indicating the occurrence of an energization failure (step 605).

  That is, in the motor drive by current feedback control, when a current failure occurs in any of the three phases, the duty of the current failure occurrence phase increases to the maximum value according to the sign of the phase current (sign (In the case of “+”) or descend to the minimum value (in the case of “−”). In this embodiment, the theoretical maximum value of the duty is “100%” and the minimum value is “0%”. The remaining two-phase duty is the extreme region on the opposite side of the current-carrying failure occurrence phase (when the current-carrying failure occurrence phase is the maximum value, near the minimum value, and when the current-carrying failure occurrence phase is the minimum value) Indicates the value near the maximum value.

  That is, the phase voltage values Vu, Vv, and Vw when the energization failure occurs are extremely close to the so-called phase-fixed energization performed by fixing the energization pattern. At this time, any one of the phase current values Iu, Iv, and Iw has a theoretical value of “0” due to the occurrence of an energization failure.

  For this reason, the induced voltage estimated values Eα_e and Eβ_e calculated by the rotational angular velocity estimation calculating unit 50 become substantially constant values. As a result, the estimated motor rotation angle θm_e based on the induced voltage estimated values Eα_e and Eβ_e is also It becomes a substantially constant value (fixed state) in the vicinity of the rotation angle (electrical angle) generated during phase fixing energization by the energization pattern. Then, by calculating the motor rotation angular velocity estimation value ωm_e based on the fixed motor rotation angle estimation value θm_e, the above contradiction occurs between the motor rotation angular velocity estimation value ωm_e and the induced voltage square sum Esq_αβ. To do.

  More specifically, for example, if a U-phase current failure occurs in the rotation angle region (180 ° to 360 °) in which the U-phase current (sign) is “−”, the duty is reduced to the minimum value. The remaining two-phase duty rises to near the maximum value. That is, a phase-fixed energization state from “V, W phase to U phase”, that is, a state in which a maximum negative voltage is applied to the U phase.

Here, the three-phase current values Iu, Iv, and Iw are converted into an α-axis current value Iα and a β-axis current value Iβ, respectively, by the following equations (5) and (6).
Iα = Iu− (1/2) × Iv− (1/2) × Iw (5)
Iβ = (√3 / 2) × Iv− (√3 / 2) × Iw (6)
Similarly, each phase voltage value Vu, Vv, Vw is converted into an α-axis voltage value Vα and a β-axis voltage value Vβ by the following equations (7) and (8), respectively.

Vα = Vu− (1/2) × Vv− (1/2) × Vw (7)
Vβ = (√3 / 2) × Vv− (√3 / 2) × Vw (8)
Also, according to Kirchhoff's law, the sign of the remaining two-phase phase current values is reversed when a current failure occurs. In the phase-fixed energization state from the “V, W phase to U phase”, the “V-W” line voltage is minimized.

  Therefore, the α-axis current value Iα and the β-axis current value Iβ are both substantially zero when the energization failure occurs, while the α-axis voltage value Vα is maximized and the β-axis voltage value Vβ is substantially zero. This relationship is the same even when the energization failure occurrence phase is the V phase or the W phase.

Further, in the disturbance observer 54 provided in the rotational angular velocity estimation calculation unit 50 of the present embodiment, the motor voltage equation as the basis thereof is expressed by the following equations (8) and (9).
Vα = R × Iα + Eα (9)
Vβ = R × Iβ + Eβ (10)
Therefore, when an energization failure occurs, the output of the disturbance observer 54, that is, the induced voltage estimated values Eα_e and Eβ_e become substantially constant at “Eα = maximum” and “Eβ = 0”, and is calculated based on the induced voltage estimated values Eα_e and Eβ_e. The estimated motor rotation angle value θm_e is fixed in the vicinity of the rotation angle (electrical angle) generated during phase-fixed energization by the energization pattern.

  That is, in the region where the phase current of the U phase is “−” as shown in this example, when the conduction failure occurs in the U phase, the estimated motor rotation angle θm_e is the tangent inverse represented by the above equation (4). From the function, it is fixed near “270 °” where “−Eα / Eβ” (absolute value) becomes infinite. Similarly, in the region where the phase current of the U phase is “+”, when a current failure occurs in the U phase, “−Eα / Eβ” (absolute value thereof) is similarly “90 °”. It sticks in the vicinity. The reason why the estimated motor rotation angle value θm_e does not become a completely constant value but becomes “… close to the vicinity” is that the applied voltage to the V phase and the W phase varies due to the execution of the current feedback control. is there. When the V phase and the W phase are poorly energized phases, the rotation angles “270 °” and “90 °” when the U phase is poorly energized are shifted by 2 / 3π, that is, “30 °”, “210 °”, “150 °”, and “330 °” are rotation angles at which the estimated motor rotation angle θm_e is fixed.

  Then, the motor rotational angular velocity estimated value ωm_e is calculated based on the fixed motor rotational angle estimated value θm_e, thereby causing a contradiction between the motor rotational angular velocity estimated value ωm_e and the induced voltage square sum Esq_αβ.

  That is, as described above, from “Eα = maximum”, although the induced voltage square sum Esq_αβ is in the maximum region corresponding to the highest speed rotation state (Esq_αβ> E1), the estimated motor rotation angular velocity value ωm_e is in the non-rotation state. (Ωm_e <ω1) occurs in the minimum region indicating. Then, the energization failure detection unit 71 of the present embodiment detects any of the motors 12 when such a contradiction is detected by the alternative energization failure determination (see FIG. 13, step 602: YES and step 603: YES). In this phase, it is determined that an abnormality indicating the occurrence of an energization failure has occurred.

  More specifically, the energization failure detection unit 71 of the present embodiment has an abnormality indicating the occurrence of an energization failure based on the contradiction between the estimated motor rotational angular velocity value ωm_e and the induced voltage square sum Esq_αβ as described above. If it is determined that the abnormality is present, it is monitored whether or not the abnormality is continuous. Then, when it is determined that the abnormal state is continuous, the abnormality is confirmed, that is, occurrence of an energization failure is detected.

  That is, as described above, when the motor rotation angular velocity estimation value ωm_e can be used as the second change component dθω of the motor rotation angle, the addition angle calculation unit 41 of the present embodiment uses the motor rotation angular velocity estimation value ωm_e as the motor rotation angular velocity estimation value ωm_e. The addition angle θa is calculated using the second change component dθω based thereon. The threshold E0 of the induced voltage square sum Esq_αβ used when determining whether or not it is usable is set to be lower than the threshold E1 used for the alternative abnormality determination. Accordingly, when the estimated motor rotation angle value θm_e is fixed and the estimated motor rotation angle velocity value ωm_e is substantially “0”, the addition angle θa based on the estimated motor rotation angle velocity value ωm_e also becomes a small value. As the change in the control angle θc, which is a virtual motor rotation angle, becomes smaller, the state where the estimated motor rotation angle θm_e is fixed also continues. And the electricity supply failure detection part 71 of this embodiment can detect generation | occurrence | production of electricity supply failure accurately by detecting the continuation.

  Specifically, as shown in FIG. 14, the energization failure detection unit 71 of the present embodiment performs the energization failure determination (step 701), and the determination result indicates that there is an abnormality ( In step 702: YES, the abnormality counter and the interval counter are incremented (steps 703 and 704).

  When it is determined in step 702 that there is no abnormality (step 702: NO), the abnormality counter increment process in step 703 is not executed. In step 704, the interval counter is incremented.

  Next, the energization failure detection unit 71 determines whether or not the count value n of the abnormality counter exceeds a predetermined threshold value n0 (step 705). When the count value n exceeds the threshold value n0 (n> n0, step 705: YES), an abnormality indicating the occurrence of the energization failure is determined (step 706).

  In step 705, when the count value n of the abnormality counter is equal to or less than the threshold value n0 (n ≦ n0, step 705: NO), whether the count value N of the interval counter subsequently exceeds the predetermined threshold value N0 or not. Is determined (step 707). When the count value N exceeds the threshold value N0 (N> N0, step 707: YES), both of the two counters are cleared (N = 0 and n = 0, step 708).

  In step 707, when the count value N of the interval counter is equal to or less than the threshold value N0 (N ≦ N0, step 707: NO), the process in step 708 is not executed.

  In other words, if the abnormality indicating that the energization is unnecessary continues, even if the abnormality counter is cleared together with the interval counter (N = 0 and n = 0) in a predetermined cycle (N0), the interval counter is updated. Within the period, the abnormality counter (count value n) reaches a level (threshold n0) at which the abnormality can be determined. In the present embodiment, this makes it possible to accurately detect the occurrence of an energization failure with a simple configuration.

  In the present embodiment, as described above, the calculation of the addition angle θa based on the estimated motor rotation angular velocity value ωm_e is performed on the addition value of the second change component dθω based on the estimated motor rotation angular velocity value ωm_e and the torque deviation Δτ. This is performed by adding the first change component dθτ based on it (see FIG. 8, step 211). Therefore, even if the motor rotation angular velocity estimation value ωm_e is zero, the motor rotation angle estimation value θm_e gradually transitions due to the torque deviation Δτ. Accordingly, the threshold value N0 is set corresponding to a time shorter than an estimated time required for the motor rotation angle estimated value θm_e to be released from the fixed state by the first change component dθτ based on the torque deviation Δτ. ing.

  Further, in the present embodiment, the abnormality continuation determination is executed even when the motor rotation angle θm detected as the actual rotation angle is normal, that is, in the situation where the normal energization failure determination using the actual rotation angular velocity ωm is performed. In this case, however, the processing procedure is the same as in the flowchart shown in FIG.

As described above, according to the present embodiment, the following operations and effects can be obtained.
(1) The energization failure detection unit 71 determines whether or not the induced voltage square sum Esq_αβ exceeds a predetermined threshold value E1 (step 602). Further, the energization failure detection unit 71 determines whether or not the motor rotation angular velocity estimated value ωm_e is lower than a predetermined threshold ω1 (step 603). Even though the induced voltage square sum Esq_αβ is in the maximum region corresponding to the highest speed rotation state (Esq_αβ> E1, step 602: YES), the motor rotation angular velocity estimated value ωm_e is in the minimum region indicating the non-rotation state. When a contradiction that exists (ωm_e <ω1, step 603: YES) is detected, it is determined that there is an abnormality indicating the occurrence of an energization failure in any phase (step 604).

  That is, when a current-carrying failure occurs in any of the phases, the duty of the current-carrying failure generation phase increases to the maximum value or decreases to the minimum value depending on the sign of the phase current, and the remaining two-phase Duty indicates the value of the limit region on the opposite side to the current-carrying failure occurrence phase. That is, the phase voltage values Vu, Vv, and Vw at the time of energization failure are very close to what is called phase-fixed energization performed by fixing the energization pattern. At this time, any one of the phase current values Iu, Iv, and Iw has a value of “0” due to the occurrence of an energization failure.

  For this reason, the induced voltage estimated values Eα_e and Eβ_e calculated by the rotational angular velocity estimation calculating unit 50 become substantially constant values. As a result, the estimated motor rotation angle θm_e based on the induced voltage estimated values Eα_e and Eβ_e is also It becomes a substantially constant value (fixed state) in the vicinity of the rotation angle (electrical angle) generated during phase fixing energization by the energization pattern. Then, by calculating the motor rotation angular velocity estimation value ωm_e based on the motor rotation angle estimation value θm_e, a contradiction occurs between the motor rotation angular velocity estimation value ωm_e and the induced voltage square sum Esq_αβ.

  Accordingly, by monitoring the contradiction between the estimated motor rotational angular velocity value ωm_e and the induced voltage square sum Esq_αβ as in the above configuration, accuracy can be achieved even when performing the resolverless control without using the actual rotational angular velocity ωm of the motor 12. Well, it is possible to detect the occurrence of an energization failure in the power supply path.

  (2) The addition angle calculation unit 41 calculates the addition angle θa using the second change component dθω based on the estimated motor rotation angular velocity value ωm_e in a region where the induced voltage square sum Esq_αβ is sufficiently large. On the other hand, when it is determined that the abnormality indicating the occurrence of the conduction failure has occurred based on the contradiction between the estimated motor rotational angular velocity value ωm_e and the induced voltage square sum Esq_αβ, the failure detection unit 71 Monitor for continuity. Then, when it is determined that the abnormal state is continuous, the abnormality is determined.

  That is, when the estimated motor rotation angle value θm_e is fixed and the estimated motor rotation angle velocity value ωm_e becomes substantially “0”, the addition angle θa based on the estimated motor rotation angle velocity value ωm_e also becomes a small value. Then, since the change in the control angle θc, which is a virtual motor rotation angle in the control, becomes small, the state where the estimated motor rotation angle value θm_e is fixed is also continued. Therefore, according to the said structure, generation | occurrence | production of a conduction failure can be detected more accurately.

  (3) Every time it is determined that there is an abnormality (step 702: YES), the energization failure detection unit 71 increments the abnormality counter (count value n) (step 703), and regardless of whether the abnormality is determined, The interval counter (count value N) is incremented in the calculation cycle (step 704). Then, in a predetermined cycle (N> N0, step 707: YES), while clearing both the interval counter and the abnormality counter (N = 0 and n = 0, step 708), the abnormality counter is set to a level where the abnormality can be determined. (N> n0, step 706), the abnormality is confirmed (step 711). Thereby, the continuation of the abnormality can be determined with a simple configuration.

In addition, you may change the said embodiment as follows.
In the above embodiment, the present invention is embodied in the ECU 11 as a motor control device that controls the operation of the motor 12 that is the drive source of the EPS actuator 10. However, the present invention is not limited to this and may be applied to uses other than EPS.

Further, even when applied to EPS, the present invention is not limited to the so-called column type as in each of the above embodiments, and may be applied to, for example, a so-called pinion type or rack assist type EPS.
In the above embodiment, when there is no abnormality in the motor rotation angle θm detected as the actual rotation angle, the motor is based on the phase voltage command values Vu *, Vv *, Vw * calculated by the first control unit 25. Output a control signal. When an abnormality occurs in the motor rotation angle θm, the phase voltage command value Vu **, which is calculated by the second control unit 26 using the control angle θc that is a virtual motor rotation angle for control. The motor control signal is output based on Vv ** and Vw **. However, the present invention is not limited to this, and the present invention may be applied to one that performs resolverless control without detecting the actual rotation angle of the motor from the beginning.

  In the above embodiment, the induced voltage square sum Esq_αβ is compared with the predetermined threshold value E0. When the induced voltage square sum Esq_αβ exceeds the threshold value E0, the sum of the first change component dθτ based on the torque deviation Δτ and the second change component dθω based on the estimated motor rotation angular velocity ωm_e is added to the addition angle θa. When the value is equal to or less than the threshold value E0, the first change component dθτ based on the torque deviation Δτ is set as the addition angle θa. However, the present invention is not limited to this, and it may always be embodied in a configuration in which the second change component dθω based on the estimated motor rotational angular velocity value ωm_e is the addition angle θa. The addition value of the first change component dθτ based on the torque deviation Δτ and the second change component dθω based on the estimated motor rotation angular velocity value ωm_e may be used as the addition angle θa. Note that the first change component dθτ based on the torque deviation Δτ may always be embodied as the addition angle θa. However, from the viewpoint of improving the accuracy of detecting the energization failure, the first change component dθτ is based on the estimated motor rotational angular velocity value ωm_e. If applied to the configuration for calculating the addition angle θa, a more remarkable effect can be obtained.

  In the above embodiment, in the alternative energization failure determination (see FIG. 13), the energization failure occurrence phase is not specified, but the energization is performed based on the estimated motor rotation angle θm_e that is fixed when the energization failure occurs. It is good also as a structure which specifies a defect generation | occurrence | production phase.

  That is, as described above, the rotation angle (electrical angle) to which the estimated motor rotation angle value θm_e is fixed when an energization failure occurs has a value specific to the energization failure occurrence phase (and the sign at the time of occurrence). Therefore, it is possible to identify the energization failure occurrence phase based on the estimated motor rotation angle value θm_e.

  Specifically, as shown in the flowchart of FIG. 15, when the motor rotation angle estimated value θm_e is fixed at (270 °) or “90 °” (near) (step 801: YES), the U phase is changed. It is identified as an energization failure occurrence phase (step 802). When the estimated motor rotation angle value θm_e is fixed at (near) “30 °” or “210 °” (step 803: YES), the V-phase is identified as the energization failure occurrence phase (step 804), If it is fixed at (in the vicinity of) “150 °” or “330 °” (step 805: YES), the W phase is identified as a current-carrying failure occurrence phase (step 806).

-Moreover, the structure which specifies the phase whose value of a phase current is a value (substantially zero) which shows a non-energized state as a conduction failure generation | occurrence | production phase may be sufficient.
In the above embodiment, the energization failure detection unit 71 increments the abnormality counter every time it is determined that there is an abnormality indicating the occurrence of an energization failure, and the interval counter is set in each calculation cycle regardless of the abnormality determination. Increment. Then, while clearing both the interval counter and the abnormality counter at a predetermined cycle, the abnormality is determined by reaching the level at which the abnormality can be determined (see FIG. 14). However, the abnormality continuation determination method for determining whether or not the abnormality indicating the occurrence of the energization failure is continuous is not necessarily limited to this, for example, whether or not the abnormality continues for a predetermined time or more. It may be performed by determining.

  In the above embodiment, in the alternative energization failure determination (see FIG. 13), whether or not the motor current indicates a non-energized state even though it should be in the energized state, that is, whether it should be in the energized state. The decision was not made. However, the present invention is not limited to this, and determinations regarding inconsistencies in such energized states may be combined. Thereby, generation | occurrence | production of a conduction failure can be detected more accurately.

  Specifically, for example, as in the normal energization failure determination (see FIG. 11), “determination as to whether or not to be in an energized state (step 402)” “determination as to whether or not to be in an unenergized state (step 403) "may be executed. And "whether it should be in an energized state" may be configured to use a voltage command value or a current command value in addition to Duty.

  In the above embodiment, the drive circuit 18 is formed by connecting a plurality of switching elements (FETs 18a to 18f, N-channel MOSFETs) having parasitic diodes. However, the present invention is not limited to this. For example, the drive circuit 18 may be embodied by a switching element such as a relay that does not have a parasitic diode.

  In the above embodiment, the steering angle θs is estimated based on the left and right wheel speeds Wr and Wl. However, the actual steering angle θs may be detected using a rotation angle sensor such as a steering sensor. . In the configuration in which the steering angle θs is estimated based on the left and right wheel speeds Wr and Wl as in this embodiment, when the front wheel speed and the rear wheel speed can be detected independently, the front wheel speed The average value of the estimated value based on the above and the estimated value based on the rear wheel speed may be the steering angle θs.

Next, technical ideas that can be grasped from the above embodiments will be described together with effects.
(A) The motor control device characterized in that the abnormality detection means identifies a phase having an abnormality indicating the occurrence of the energization failure based on the estimated motor rotation angle.

  That is, the rotation angle (electrical angle) to which the estimated value of the motor rotation angle is fixed when an energization failure occurs has a value specific to the energization failure occurrence phase (and the sign at the time of occurrence). Therefore, according to the above configuration, it is possible to specify the occurrence of the energization failure with a simple configuration.

  (B) The motor control device characterized in that the abnormality detection means identifies a phase indicating that the phase current value is in a non-energized state as an abnormality occurrence phase. Thereby, it is possible to specify the occurrence of the energization failure with a simple configuration.

  DESCRIPTION OF SYMBOLS 1 ... Electric power steering apparatus (EPS), 10 ... EPS actuator, 11 ... ECU, 12 ... Motor, 12u, 12v, 12w ... Motor coil, 17 ... Microcomputer, 18 ... Drive circuit, 18a-18f ... FET, D ... Parasitic Diode, 18u, 18v, 18w ... Switching arm, 20u, 20v, 20w ... Power line, 21 (21u, 21v, 21w) ... Current sensor, 23 ... Motor resolver, 24 ... Motor controller, 25 ... First controller, 26 ... 2nd control part, 27 ... PWM conversion part, 31 ... Current command value calculating part, 32 ... d / q conversion part, 34d, 34q ... F / B control part, 35 ... d / q reverse conversion part, 41 ... Addition angle calculation unit, 42 ... Control angle calculation unit, 43 ... Steering angle estimation calculation unit, 45 ... Target torque calculation unit, 46 ... Subtractor, 47 ... F / B control unit, 50 ... Rotational angular velocity estimation Calculation unit 51... Voltage sensor 52... Phase voltage calculation unit 53... Α / β conversion unit 54. Disturbance observer 58 58 Addition angle adjustment calculation unit 59 59 Addition angle limiting unit 60. , 61 ... current command value calculation unit, 62 ... γ-axis current increase / decrease value calculation unit, 63 ... integration control unit, 65a, 65b ... F / B control unit, 66 ... 2-phase / 3-phase conversion unit, 68 ... rotation angle abnormality Detecting unit, 71 ... Energization failure detecting unit, Iu, Iv, Iw, Ix ... Phase current value, I1 ... Threshold value, θm ... Motor rotational angle, ωm ... Actual rotational angular velocity, ω0 ... Threshold value, Id ... ... q-axis current value, Id * ... d-axis current command value, Iq * ... q-axis current command value, ΔId, ΔIq ... current deviation, Vu *, Vv *, Vw * ... phase voltage command value, Dx, Du, Dv , Dw ... Duty, τ ... steering torque, τ * ... target torque, Δτ ... torque deviation, η ... γ-axis current increase / decrease value, dθτ ... first change component, Iα ... α-axis current value Iβ ... β-axis current value, Vα ... α-axis voltage value, Vβ ... β-axis voltage value, Eα, Eβ ... induced voltage, Eα_e, Eβ_e ... induced voltage estimated value, Esq_αβ ... induced voltage square sum, E0, E1 ... threshold value, ωm_e: estimated motor rotational angular velocity, ωm_e: estimated motor rotational angular velocity, ω1: threshold, dθω: second change component, θa, θa ′: added angle, b, B: threshold, θc: control angle, Iγ: γ axis Current value, Iδ ... δ-axis current value, Iγ * ... γ-axis current command value, Iδ * ... δ-axis current command value, ΔIγ, ΔIδ ... Current deviation, Vu **, Vv **, Vw ** ... Phase voltage command Value, Vu, Vv, Vw ... Phase voltage value, n, N ... Counter value, n0, N0 ... Threshold value, S_rsf ... Rotation angle abnormality detection signal, S_pde ... Energization failure detection signal.

Claims (3)

Motor control signal output means for outputting a motor control signal, a drive circuit that operates in response to the input of the motor control signal and supplies three-phase drive power based on a power supply voltage to the motor, and energization in a current supply path to the motor An abnormality detection means for detecting the occurrence of a defect, and the motor control signal output means calculates an addition angle corresponding to a motor rotation angle change amount for each calculation cycle, and integrates the addition angle for control. In the motor control device that calculates the motor rotation angle and outputs the motor control signal by executing current feedback control in a rotating coordinate system according to the motor rotation angle on the control,
Conversion means for converting each phase current value and each phase voltage value of the three phases into each axis current value and each axis voltage value of a two-phase fixed coordinate system;
Induced voltage estimation means for calculating each axis induced voltage estimated value of the two-phase fixed coordinate system based on each axis current value and each axis voltage value;
A rotation angular velocity estimation means for calculating a motor rotation angle estimation value based on each axis induced voltage estimation value, and calculating a motor rotation angular velocity estimation value based on a change amount of the motor rotation angle estimation value;
The abnormality detecting means determines that there is an abnormality indicating the occurrence of the energization failure when the sum of squares of the estimated values of the respective shaft induced voltages is in a maximum region and the estimated motor rotation angular velocity is in a minimum region. A motor control device. The motor control device according to claim 1,
The motor control signal output means calculates the addition angle based on the estimated motor rotation angular velocity,
The electric power steering apparatus comprising: the abnormality detecting unit determining the abnormality when the abnormality indicating the occurrence of the energization failure continues.   An electric power steering device comprising the motor control device according to claim 1.

Images