JP7588376B2 - Motor Control Device - Google Patents

Description

The present invention relates to a motor control device equipped with a rotation angle detection device that accurately calibrates and detects the magnetic pole position (rotation angle) of a synchronous motor that is vector controlled based on a d-axis current command and a q-axis current command .

Control devices for synchronous motors such as brushless DC motors use a dq-axis coordinate control system to pass current through the appropriate excitation phase winding according to the magnetic pole position of the rotor, generating the desired torque. In such synchronous motors, it is important to accurately calibrate the origin of the electrical angle in order to accurately control the motor output. There are two types of synchronous motors: those equipped with a magnetic pole position sensor such as an encoder to detect the magnetic pole position of the rotor, and those without a magnetic pole position sensor.

In the case of a synchronous motor that does not have a magnetic pole position sensor, it is necessary to perform a magnetic pole position detection process to detect the initial position of the magnetic pole (hereinafter referred to as the "magnetic pole initial position") each time the synchronous motor is turned on (started), and to control the rotation of the synchronous motor based on the magnetic pole position based on this magnetic pole initial position. However, if there is variation in the detection accuracy of this magnetic pole position detection process when the power is turned on, there will also be variation in the torque constant when the synchronous motor is driven, and it may not be possible to generate maximum torque. In particular, when driving at high speeds that require field weakening control, if there is a deviation in the magnetic pole position, it is not possible to flow an appropriate d-axis current, which results in a shortage of drive voltage applied to the synchronous motor, resulting in unstable control.

To address this issue, there is a method in which a magnetic pole correction value, which is the amount of deviation between the magnetic pole reference position and the reference position of the synchronous motor's rotation sensor, is stored in advance in memory when the synchronous motor is started for the first time or when the motor sensor is replaced or maintained. After the magnetic pole position detection process is actually performed when the synchronous motor is turned on, the magnetic pole position used for rotation control is corrected based on the magnetic pole correction value corresponding to the reference position when the synchronous motor is driven, so that the synchronous motor can always be controlled based on the same magnetic pole position when driven. According to this method, a magnetic pole correction value corresponding to the sensor reference position is preset in memory, and the magnetic pole initial position is corrected, so that thereafter, the synchronous motor can always be controlled based on the magnetic pole correction value corresponding to the sensor reference position.

Another method is to align the reference point of the encoder with the reference point of the motor stator that has been DC excited and locked, and then calculate the amount of deviation from the initial magnetic pole estimation result and the encoder information, store the amount of deviation in the encoder's memory, and then make corrections.

In such a method using a magnetic pole correction value, when the reference position of the rotation sensor of the synchronous motor is detected for the first time after the magnetic pole position detection process when the synchronous motor is powered on, the magnetic pole position based on the magnetic pole initial position corresponding to this position is set as the magnetic pole correction value. However, since there is variation in detection accuracy in the magnetic pole position detection process as described above, in order to reduce the effect on the setting of the magnetic pole correction value caused by this variation in detection accuracy, it is necessary to perform the magnetic pole position detection process multiple times and calculate the average value, which is a problem that takes time to calculate. In addition, with the method using the magnetic pole correction value, there is no clear indicator as to how many times the magnetic pole position detection process should be performed, and there is a problem that the error contained in the magnetic pole correction value cannot be sufficiently removed.

Patent No. 5174205

As a method for solving the above-mentioned problem, a detection device described in Japanese Patent No. 5174205 (Patent Document 1) has been proposed. That is, the detection device for detecting the magnetic pole position of a synchronous motor includes a generating means for generating a magnetic pole correction value based on the difference between a forward rotation d-axis voltage command value and a reverse rotation d-axis voltage command value for rotating the synchronous motor when a predetermined d-axis current command value is given to rotate the synchronous motor forward and reverse after detecting the magnetic pole initial position when the synchronous motor is powered on, and a correcting means for correcting the magnetic pole initial position based on the magnetic pole correction value and the sensor reference position, which is the reference position of the sensor of the synchronous motor, and controls the rotation of the synchronous motor based on the corrected magnetic pole initial position.

However, the detection device described in Patent Document 1 performs many calculations, such as arctan calculation of the current value, which results in high calculation costs. In addition, the feedback control unit is in operation, which makes it difficult to obtain an accurate correction amount.

The present invention has been made in view of the above-mentioned circumstances, and an object of the present invention is to provide a motor control device equipped with a rotation angle detection device which has a low computational cost in construction and is capable of easily determining and calibrating an accurate correction amount.

The present invention relates to a motor control device which includes a voltage command value calculation unit which outputs a two-phase voltage command value which is a sum of a three-phase voltage command value calculated based on a dq-axis current command value or a command value calculated based on the dq-axis current command value and a command value for dq-axis non-interference, a back electromotive force compensation unit which calculates a back electromotive force compensation voltage which compensates for a back electromotive force generated in a motor, an adder which adds the back electromotive force compensation voltage to the three-phase voltage command value or the two-phase voltage command value to output a control voltage command value, and an origin correction unit which corrects an output angle from a motor angle detector with a calibration phase angle and outputs the corrected motor angle, and which drives and controls the motor based on the control voltage command value, the calibration phase angle is achieved by including a processing unit which sets an output of the voltage command value calculation unit to zero when the calibration phase angle is calculated, and a phase adjustment unit which changes the back EMF compensation voltage based on an angle obtained by changing the phase with respect to the corrected motor angle when the motor is rotated clockwise or counterclockwise at a predetermined speed using an external servo mechanism, and calculates the calibration phase angle based on changes in dq-axis currents of the motor, after the calibration phase angle is calculated, calculating the control voltage command value based on the corrected motor angle in which the output angle is corrected by the calibration phase angle and the dq-axis current command values, and controlling the motor to drive based on the control voltage command value .

According to the rotation detection device of the present invention, since a back electromotive force compensation unit that compensates for the motor back electromotive force is inserted, the motor back electromotive force compensation can be performed in both dq-axis and three-phase dimensions. In addition, in the error adjustment mode in which the origin of the motor rotation angle is calibrated, the phase that is the advance angle or delay angle is changed while rotating the motor at a predetermined speed, and the phase angle when the d-axis current (or the magnitude of the dq-axis current vector) is minimized is obtained for the CW direction rotation and the CCW direction rotation, and the average value is calculated as the correction value for the origin correction (calibration), so that it is possible to detect an accurately calibrated rotation angle. The back electromotive force compensation unit that compensates for the motor back electromotive force is also used in the normal control mode. The difference in calculation processing between the back electromotive force compensation unit for the error adjustment mode and the back electromotive force compensation unit for the normal control mode is small. Therefore, issues such as an increase in program capacity and an increase in the time required for program verification can be addressed with a small increase in capacity and a short verification time.

11 is a conceptual diagram showing how dq-axis currents flow due to back electromotive force compensation when there is an error from the origin. FIG. 11 is a block diagram showing a configuration example (normal control mode) of a form in which a back electromotive force compensation voltage is applied to a three-phase control voltage command value; FIG. 11 is a block diagram showing a configuration example (error adjustment mode) of a form in which a back electromotive force compensation voltage is applied to a three-phase control voltage command value; FIG. FIG. 13 is a wiring diagram showing a part of a configuration example in an error adjustment mode. FIG. 13 is a block diagram showing a configuration example in an error adjustment mode. FIG. 4 is a flow chart illustrating an example of the operation of the present invention. FIG. 13 is a block diagram showing a configuration example (normal control mode) of a form in which a back electromotive force compensation voltage is applied to a dq axis control voltage command value. 13 is a block diagram showing a configuration example (error adjustment mode) of a form in which a back electromotive force compensation voltage is applied to a dq axis control voltage command value; FIG.

FIG. 1 is a conceptual diagram showing the flow of dq-axis currents due to back electromotive force compensation when there is an error from the origin. At this time, it is assumed that there is no voltage command value based on the dq-axis current command value. d and q respectively indicate the d-axis and q-axis on the motor shaft, and _dd and _qd respectively indicate the d-axis and q-axis in control. Also, _qccw and _qcw respectively indicate the q-axis on the motor shaft when considering the time lapse from the time of angle detection until the back electromotive force compensation value is reflected in the motor terminal. And, "CW" indicates the clockwise rotation of the motor, and "CCW" indicates the counterclockwise rotation of the motor. If there is an error Δθ in the angle from the reference angle, the dq-axis current flows, but since the error Δθ is small, the ratio of the d-axis current is larger than the q-axis current. Also, there is a delay time from the time of angle detection until the back electromotive force compensation voltage is generated at the motor terminal, and therefore, the dq-axis current differs depending on the rotation direction.

When the back electromotive force compensation functions correctly, the d-axis and q-axis current values during rotation are zero (in the absence of PI compensation). However, if the phase of the back electromotive force compensation is misaligned, the compensation is incomplete, and the d-axis and q-axis current values are not zero. In this case, it is necessary to adjust the phase α (error from the origin), which is the advance or delay angle, so that the d-axis and q-axis current values become zero. If the origin of the detected electrical angle is misaligned from the true value, the optimum value of the phase α will be different depending on the direction of rotation. The phase angle _αcw of the back electromotive force compensation in the CW direction that makes the d-axis current zero, and the phase angle _αccw of the back electromotive force compensation in the CCW direction that makes the d-axis current zero are found, and the phase angle average value _αm is taken as the angle error, which corresponds to the electrical angle offset calibration amount (error correction amount from the origin).

The currents used to search for the optimum value of phase α for leading or retarding can be various, such as dq-axis current values and phase current amplitudes. As the change in response to the operation of phase α is large, it is preferable to measure the d-axis current, but the q-axis current may also be used. In other words, it is possible to find a phase α such that at least one of the dq-axis current values is zero. It is also possible to find a phase α such that the square root of the sum of the squares of the d-axis current and the q-axis current (the magnitude of the dq-axis current vector) or the sum of the squares is minimized.

Figure 1 shows the resultant vector voltage (thick dashed line) of the back EMF compensation voltage (thick solid line) and the actual back EMF (dotted line). Since the resultant vector voltage that contributes to the motor current appears to be applied to the motor terminals, the dq-axis currents have a magnitude that corresponds to the dq-axis voltage components. The error from the reference point (origin) is considered small because it is adjusted to some extent when the motor is shipped, but in Figure 1 it is shown as a large angle error for ease of illustration and understanding. Also, the d-axis has a larger contribution as a component of the resultant vector, but if the error is large this can be reversed.

2 shows an example of the configuration of a motor control system (normal control mode) in which back electromotive force compensation voltages ( _vu1 *, _vv1 *, _vw1 *) are applied to three-phase control voltage command values _vu , _vv , _vw , and an origin correction unit 70 stores a correction value (phase _αm ) previously determined in the error adjustment mode. The d-axis current command value idref and the q-axis current command value iqref are added and input to subtraction units 12d and 12q, respectively, and the measured d-axis current id and q-axis current iq are subtracted from them, respectively. The current deviations resulting from the subtraction are input to PI (Proportion-Integral) control units 11d and 11q, respectively. The voltage command values _vd and _vq from the PI control units 11d and 11q are input to a two-phase/three-phase conversion unit 10 and converted to three-phase control voltage command values _vu , _vv , _vw . The control voltage command values _vu , _vv , and _vw are input to the adders 31u, 31v, and 31w, respectively, and are added to the back electromotive force compensation voltages _vu1 *, _vv1 *, and _vw1 * from the back electromotive force compensation unit 41. The control voltage command values resulting from the additions of the adders 31u, 31v, and 31w are added and input to the subtracters 32u, 32v, and 32w, respectively, and the back electromotive force _vu2 *, _vv2 *, and _vw2 * from the actual back electromotive force unit 40 are subtracted, and the result is input to the motor winding models 30u, 30v, and 30w, respectively. The subtracters 32u, 32v, and 32w and the actual back electromotive force unit 40 are models of the back electromotive force actually generated in the motor. If the output of the winding model is a current I and the terminal voltage is V, the voltage equation is expressed by the following formula 1.
(Equation 1)
V=L・dI/dt+R・I
When the above equation 1 is subjected to Laplace transformation, the following equation 2 is obtained.
(Equation 2)
V(s)=L・s・I(s)+R・I(s)

The transfer function (output current/input voltage) is expressed as "1/(L·s+R)=I(s)/V(s)". Note that the resistance R includes the resistance of the drive circuit.

The motor generates actual motor back electromotive voltages _vu2 *, _vv2 *, and _vw2 * according to the following formula 3, and the actual motor back electromotive voltages _vu2 *, _vv2 *, and _vw2 * are subtracted and input to subtraction units 32u, 32v, and 32w on the paths of the three-phase control voltage command values. Note that in formula 3, "Ker" is the motor back electromotive voltage constant, "θr" is the actual rotation angle, and "ωr" is the actual angular velocity.

モータ巻線モデル３０ｕ、３０ｖ、３０ｗからの３相の相電流はモータモデル２０に入力されると共に、３相／２相変換部５０に入力される。モータモデル２０の伝達関数は“Ｋｔ／（Ｊｓ＋ｄ）”であり、“Ｋｔ”はモータトルク定数（Ｎｍ／Ａ）、“Ｊ”はモータ慣性モーメント、“ｄ”は粘性係数である。モータモデル２０の出力は積分（ｓ）のモータモデル（ｐ/s）２１で実回転角度（電気角）θrとして出力される。また、モータモデル（ｐ）２４からは、実角速度ωrが出力される。実回転角度θrは、角度検出器２３を介して回転角度θsとして、原点補正用の減算部７１に加算入力される。角度検出器２３の取り付け誤差及び原点補正部の調整誤差が、原点誤差Δθを生じさせる。

The three-phase currents from the motor winding models 30u, 30v, and 30w are input to the motor model 20 and also to the three-phase/two-phase conversion unit 50. The transfer function of the motor model 20 is "Kt/(Js+d)", where "Kt" is the motor torque constant (Nm/A), "J" is the motor moment of inertia, and "d" is the viscosity coefficient. The output of the motor model 20 is output as an actual rotation angle (electrical angle) θr by the motor model (p/s) 21 of integration (s). In addition, an actual angular velocity ωr is output from the motor model (p) 24. The actual rotation angle θr is added to and input as a rotation angle θs via the angle detector 23 to a subtraction unit 71 for origin correction. An installation error of the angle detector 23 and an adjustment error of the origin correction unit cause an origin error Δθ.

Note that "p" in both motor models 21 and 24 indicates the number of motor pole pairs. In motor model 21, the motor mechanical angle is converted to electrical angle by multiplying it by p, and in motor model 24, the motor mechanical speed is converted to electrical speed by multiplying it by p.

The rotation angle θs added and input to the subtraction unit 71 is subtracted by a correction signal CR (phase angle _{αm )} from the origin correction unit 70 calculated in the error adjustment mode. The rotation angle θ (θ=θs- _αm ) corrected by subtracting the correction signal CR (phase angle αm) is input to the differentiation unit 22, the back electromotive force compensation unit 41, the two-phase/three-phase conversion unit 10, and the three-phase/two-phase conversion unit 50. The angular velocity (motor rotation speed) ω obtained by differentiation in the differentiation unit 22 is input to the back electromotive force compensation unit 41. The back electromotive force compensation unit 41 calculates back electromotive force compensation voltages _vu1 *, _vv1 *, and _vw1 * based on the rotation angle θ and the angular velocity ω according to the following equation 4, and the back electromotive force compensation voltages _vu1 *, _vv1 *, and _vw1 * are input to the addition units 31u, 31v, and 31w of the paths of the three-phase control voltage command values. In addition, "g·ω" in the formula 4 is an angle for adjusting the phase delay due to the time required from the time of angle detection until the generation of the motor terminal voltage.

加算部３１ｕ、３１ｖ、３１ｗの加算結果がそれぞれ減算部３２ｕ、３２ｖ、３２ｗに加算入力される。また、３相／２相変換部５０で得られた２相のｄ軸電流id及びｑ軸電流iqは、それぞれ減算部１２ｄ及び１２ｑに減算入力される。

The sums of the adders 31u, 31v, and 31w are respectively added and input to subtractors 32u, 32v, and 32w. The two-phase d-axis current id and q-axis current iq obtained by the three-phase/two-phase converter 50 are respectively subtracted and input to subtractors 12d and 12q.

In this configuration, in the normal control mode, the correction value (phase α _m ) CR from the origin correction unit 70 is subtracted from the actual rotation angle θs from the angle detector 23 to control the motor, and the rotation angle θ from the subtraction unit 71 is input to the back electromotive force compensation unit 41, the three-phase/two-phase conversion unit 50, the two-phase/three-phase conversion unit 10, and the differentiation unit 22. The motor is vector-controlled according to the d-axis current command value idref and the q-axis current command value iqref, and in the normal control, a method is generally adopted in which the current deviation is PI-controlled so as to reduce the current deviation, and the three-phase control voltage command values are applied to the motor. The actual back electromotive force v _u2 *, v _v2 *, and v _w2 * of the actual back electromotive force unit 40 are generated in a direction that reduces the control voltage command value, and therefore appear as a load (disturbance) in the PI control. In the present invention, in order to offset (compensate) this disturbance, the back electromotive force compensation is applied (added) to the control voltage command value in a feedforward manner. That is, the back electromotive force compensation voltages _vu1 *, _vv1 *, and _vw1 * from the back electromotive force compensation unit 41 are input to the adders 31u, 31v, and 31w, respectively, to compensate for the decreases in the actual back electromotive force voltages _vu2 *, _vv2 *, and _vw2 *.

In error adjustment mode, the existing back EMF compensation block is used with minor modifications for error adjustment mode. Details will be given later. Since it can be used with minor modifications, software assets (storage capacity) can be used effectively. Other methods require additional logic to be implemented, which requires additional storage capacity.

Next, the operation of the error adjustment mode will be explained with reference to Figure 3.

The back electromotive voltage phase and the phase of the output of the motor rotation angle sensor (resolver, MR sensor, etc.) are adjusted to a certain extent when the motor is shipped, and the error adjustment mode is started when the motor is initially connected to the ECU (Engine Control Unit). Therefore, in the error adjustment mode, the motor is not connected to the controlled object, and can be rotated without load by an external servo mechanism 111 via an external measurement device 100 as described later. In the error adjustment mode, the phase adjustment unit 60 shown in FIG. 3 obtains the phase α at which the d-axis current id is minimum in the clockwise direction CW and the counterclockwise direction CCW, calculates the average value of the minimum phase α _CW in the clockwise direction CW and the minimum phase α _CCW in the counterclockwise direction CCW, and stores the calculated average phase α _m in the ECU (origin correction unit 70) as a correction value CR for correcting the origin. In normal control, the rotation angle θs is controlled by the rotation angle θ obtained by origin correction (calibration) using the correction value CR (phase angle α _m ) obtained and stored in advance in the error adjustment mode.

In the error adjustment mode, as shown in FIG. 4, change-over switches SW1 to SW3 are provided on the output side of two-phase/three-phase conversion unit 10, and in the normal control mode, they are turned ON to pass the three-phase control voltage command values _vu , _vv , and _vw , and in the error adjustment mode, the change-over switches SW1 to SW3 are turned OFF by a switching signal SWS from external measurement device 100, and the three-phase control voltage command values _vu , _vv , and _vw are grounded to zero. In addition to this switching method, it is also possible to use a configuration in which the subtraction units 12d and 12q and the PI control units 11d and 11q are all disabled and the three-phase control voltage command values _vu , _vv , and _vw are grounded to zero, or a configuration in which the control constants of the PI control or the like are changed to set the voltage command values _vd and _vq to zero (the proportional gain, integral gain, and past integral value are set to zero), or a configuration in which the output of the two-phase/three-phase conversion unit 10 is forcibly set to zero. In short, it is sufficient if the three-phase control voltage command values _vu , _vv , and _vw can be set to zero.

In the error adjustment mode, the delay angle g·ω of the back electromotive force compensation unit 41 in equation (4) is replaced with the phase α from the phase adjustment unit 60. The error adjustment mode in Fig. 3 is premised on the first adjustment, and when it is performed again, a predetermined value (e.g., zero or the current value of the origin correction unit 70) is reset as the new correction value CR in response to an instruction from the phase adjustment unit 60. Then, the average value _αm of the phase advance angle or phase delay angle calculated by a procedure described in detail below is added to the reset correction value CR, and the result is stored in the origin correction unit 70 as the final correction amount.

The configuration of the entire adjustment system in the error adjustment mode is shown in Figure 5, where an external measurement device 100 operated by an operator for measurement is connected by wire or wirelessly to a phase adjustment unit 60 and a servo rotation instruction unit 110 via a communication unit 101. The servo rotation instruction unit 110 is a mechanism that rotates a motor at a predetermined speed via an external servo mechanism 111. Although Figure 5 shows a configuration in which a control target is connected to the motor, in the initial error adjustment mode, the control target is not connected to the motor.

The phase adjustment unit 60 includes a communication unit 61 and can communicate with the external measurement device 100 via the communication unit 101. The phase adjustment unit 60 includes a phase change unit 62 that continuously changes the phase by an advance angle, a phase output unit 63 that outputs the changing phase α, a memory 64 that appropriately stores the phase α from the phase output unit 63, an average value calculation unit 65 that calculates an average value _αm of the stored phases, and a current determination unit 67 that determines the minimum value of the d-axis current id. The phase α that minimizes the d-axis current is obtained because the back electromotive force constant used in the back electromotive force compensation does not completely match the actual back electromotive force constant, and therefore it is assumed that a certain amount of current flows.

In such a configuration, a method for determining the correction value CR (phase α _m ) in the error adjustment mode will be described with reference to the flowchart in FIG.

When the error adjustment mode is entered, there are two procedures: an almost automatic calibration and a procedure involving an operator. In the procedure involving an operator, the servo rotation instruction unit 110 is started via the external measurement device 100, and the motor is rotated in the clockwise direction CW at a predetermined speed (for example, 1500 rpm) by the external servo mechanism 111 (step S10). Then, the phase to be the advance angle is instructed from the external measurement device 100 to the phase adjustment unit 60 via the communication units 101 and 61 (step S11). In the phase adjustment unit 60, the phase change unit 62 changes the phase, and the changing phase α is output from the phase output unit 63, and the phase α is input to the back electromotive force compensation unit 41. The d-axis current id and the q-axis current iq are input to the phase adjustment unit 60, and in this example, the magnitude of the d-axis current id is judged by the current judgment unit 67. Then, the operator observes the result of the current determination unit 67 via the communication units 101 and 61 to determine when the d-axis current id is minimum (step S12), and stores the phase α _CW at that time in the memory 64 (step S13).

The back electromotive force compensation unit 41 calculates back electromotive force compensation voltages _vu1 **, _vv1 **, and _vw1 ** according to the input phase α using the following equation 5, and adds them to the adders 31u, 31v, and 31w, respectively, to perform back electromotive force compensation.

その後、作業者は同様に、外部測定装置１００を介して、サーボ回転指示部１１０及び外部サーボ機構１１１によりモータを反時計方向ＣＣＷに同じ所定速度（例えば１５００ｒｐｍ）で回転させ（ステップＳ２０）、進角となる位相を外部測定装置１００から通信部１０１及び６１を介して位相調整部６０に指示する（ステップＳ２１）。位相調整部６０では、位相変化部６２が位相を変化させ、変化する位相αが位相出力部６３から出力され、位相αは逆起電圧補償部４１に入力される。逆起電圧補償部４１で、上述と同様な数５に従った逆起電圧補償電圧ｖ_ｕ１＊＊、ｖ_ｖ１＊＊、ｖ_ｗ１＊＊が演算され、それぞれ加算部３１ｕ、３１ｖ、３１ｗに加算されて逆起電圧補償が実施される。そして、作業者は、ｄ軸電流idが最小となるときを、電流判定部６７の結果を観測して判定し（ステップＳ２２）その時の位相α_ＣＣＷをメモリ６４に記憶する（ステップＳ２３）。

Thereafter, the operator similarly rotates the motor in the counterclockwise direction CCW at the same predetermined speed (e.g., 1500 rpm) by the servo rotation instruction unit 110 and the external servo mechanism 111 via the external measurement device 100 (step S20), and instructs the phase to be the advance angle from the external measurement device 100 to the phase adjustment unit 60 via the communication units 101 and 61 (step S21). In the phase adjustment unit 60, the phase change unit 62 changes the phase, and the changing phase α is output from the phase output unit 63, and the phase α is input to the back electromotive force compensation unit 41. In the back electromotive force compensation unit 41, the back electromotive force compensation voltages v _u1 **, v _v1 **, and v _w1 ** are calculated according to the same formula 5 as above, and are added to the adders 31u, 31v, and 31w, respectively, to perform back electromotive force compensation. Then, the operator observes the result of the current determination unit 67 to determine when the d-axis current id is minimum (step S22), and stores the phase α _CCW at that time in the memory 64 (step S23).

Next, the average value calculation unit 65 reads out the phases α _CW and α _CCW stored in the memory 64, calculates the average value α _m (step S30), and stores it as a correction value in the origin correction unit 70 from the external measurement device 100 via a CAN or the like (step S31).

The above is a procedure mediated by an operator, but it is also possible to perform the calibration almost automatically. In this case, after connecting the external servo mechanism 111 to the motor, the operator instructs the phase adjustment unit 60 via the communication unit 101 to transition to error adjustment mode. The phase adjustment unit 60 instructs the rotation of the external servo mechanism 111 in steps S10 and S20 above via the communication unit 101, and executes the processing of the other steps within the phase adjustment unit 60.

7 shows a form in normal control. It shows a form in which back electromotive force compensation voltages Vd and Vq for the error adjustment mode are applied on the dq axis, and the thick lines show the dq axis signals. That is, the d axis current id and the q axis current iq are input to the dq axis non-interference unit 14 and are subtracted and input to the subtraction unit 12. The motor rotation speed ω is input to the dq axis non-interference unit 14, the voltage command value _vn1 from the dq axis non-interference unit 14 is input to the adder 16 for the error adjustment mode, the current deviation between the current command value (idref, iqref) from the subtraction unit 12 and the dq axis current ( _id , _iq ) is input to the PI control unit 13, the voltage command value _vn2 from the PI control unit 13 is input to the adder 15, and the dq axis control voltage command value from the adder 15 is input to the two-phase/three-phase conversion unit 10. The configuration after the two-phase/three-phase conversion unit 10 is the same as in the case of the three-phase compensation described above. In normal control, the back electromotive force compensation voltages Vd and Vq (v _c1 ) for the error adjustment mode are forcibly set to zero.

Here, we will explain the dq axis two-phase back electromotive force compensation (decoupling control).

If the d-axis armature voltage is Vdd, the q-axis armature voltage is Vqq, the d-axis armature winding resistance is Rd, the q-axis armature winding resistance is Rq, the speed electromotive force induced on the q-axis is φω (flux linkage φ, motor angular velocity ω), the self-inductance of the d-axis controlled object (motor) is Ld, the self-inductance of the q-axis controlled object (motor) is Lq, and the Laplace operator is s, the d-axis controlled object (motor) is expressed as "1/(Ld・s+Rd)" and the q-axis controlled object (motor) is expressed as "1/(Lq・s+Rq)", and the d-axis armature voltage Vdd and the q-axis armature voltage Vqq are expressed by the following equation 6.

数６を解くと、下記数７が成立する。

By solving Equation 6, the following Equation 7 is established.

数７のｄ軸電機子電圧Ｖｄｄの右辺第１項はモータの逆関数、第２項は非干渉成分、数７のｑ軸電機子電圧Ｖｑｑの右辺第１項はモータの逆関数、第２項は非干渉成分、第３項は逆起電圧成分である。ｄ軸電機子電圧Ｖｄｄから“－ωＬ_qi_q”を減算した電圧降下後有効電圧によってd軸電流が発生し、ｄ軸電機子電圧Ｖｑｑから“ωＬ_ｄi_d＋ωφ”を減算した電圧降下後有効電圧によってq軸電流が発生する（特許第６６５８９９５号参照）。従って、ｄｑ軸非干渉化部１４による非干渉化制御としては、これを相殺する方向に補正する。

The first term on the right side of the d-axis armature voltage Vdd in Equation 7 is an inverse function of the motor, the second term is a non-interference component, the first term on the right side of the q-axis armature voltage Vqq in Equation 7 is an inverse function of the motor, the second term is a non-interference component, and the third term is a back electromotive force component. The d-axis current is generated by the effective voltage after voltage drop obtained by subtracting "-ωL _{qi q} " from the d-axis armature voltage Vdd, and the q-axis current is generated by the effective voltage after voltage drop obtained by subtracting "ωL _d i _d + ωφ" from the d-axis armature voltage Vqq (see Japanese Patent No. 6658995). Therefore, the non-interference control by the dq-axis non-interference unit 14 is to correct in a direction to cancel this out.

In the error adjustment mode, instead of the dq axis decoupling unit 14, a back electromotive force v _c1 (Vd, Vq) shown in the following equation 8 is generated by a back electromotive force compensating unit 41A to cancel the back electromotive force.

このような構成において、通常制御モードにおいては前述の３相補償の場合と同様な動作であり、モータ出力角度θを原点補正部７０からの補正信号ＣＲ（位相α_ｍ）で補正して制御する。

In this configuration, in the normal control mode, the operation is similar to that in the case of the above-mentioned three-phase compensation, and the motor output angle θ is corrected and controlled by the correction signal CR (phase α _m ) from the origin correction unit 70 .

In the error adjustment mode, as shown in FIG. 8, in the same manner as described above, the voltage command value _vn2 which is the output of the PI control unit 13 and the voltage command value _vn1 which is the output of the dq axis non-interference unit 14 are each set to zero, and the phase _αm of the average value is found by the same operation as in FIG. 6.

The back electromotive force compensation voltage unit 41A for the error adjustment mode on the dq axis is different from the back electromotive force compensation voltage unit 41 on the three phases, but the processing of the phase adjustment unit 60 is common. The calculation formula and zeroing processing of other signals in the back electromotive force compensation voltage unit 41A are simple, and the increase in program capacity and processing load are small. The dq axis non-interference unit 14 does not have information related to the phase, but the back electromotive force compensation voltage unit 41A is designed to directly adjust the phase, which has the advantage of making calibration easier.

When measuring in error adjustment mode after incorporating the motor into the controlled object, the controlled object becomes a load because it is connected to the motor, and a large-capacity external servo mechanism is required.

10: 2-phase/3-phase conversion unit 11d, 11q, 13: PI control unit 12, 12d, 12q: subtraction unit 14: dq-axis non-interference unit 20: motor model 21: motor model (integration)
22 Differentiation section 23 Angle detector 40 Actual back electromotive force section 41, 41A Back electromotive force compensation section 50 Three-phase/two-phase conversion section 60 Phase adjustment section 62 Phase change section 63 Phase output section 64 Memory 65 Average value calculation section 70 Origin correction section 100 External measurement device 101 Communication section 110 Servo rotation instruction section 111 External servo mechanism

Claims (2)

a voltage command value calculation unit that outputs three-phase voltage command values calculated based on the dq-axis current command values or two-phase voltage command values that are the sum of command values calculated based on the dq-axis current command values and a command value for dq-axis non-interference;
a back electromotive force compensation unit that calculates a back electromotive force compensation voltage that compensates for a back electromotive force generated in the motor;
an adder unit that adds the back electromotive force compensation voltage to the three-phase voltage command value or the two-phase voltage command value to output a control voltage command value;
an origin correction unit that corrects an output angle from a motor angle detector with a calibration phase angle and outputs the corrected motor angle;
a motor control device that drives and controls the motor based on the control voltage command value,
a processing unit that sets an output of the voltage command value calculation unit to zero when the calibration phase angle is calculated;
a phase adjustment unit that changes the back electromotive force compensation voltage based on an angle obtained by changing a phase with respect to the corrected motor angle when the motor is rotated in a clockwise or counterclockwise direction at a predetermined speed using an external servo mechanism , and that determines the calibration phase angle based on a change in dq-axis current of the motor;
Equipped with
a motor control device which calculates , after determining the calibration phase angle, the control voltage command value based on the corrected motor angle obtained by correcting the output angle by the calibration phase angle and the dq-axis current command value, and drives and controls the motor based on the control voltage command value . The phase adjustment unit,
a phase change unit that changes a phase of the corrected motor angle ;
a current determination unit that determines a minimum value of at least one of the d-axis current and the q-axis current or a minimum value of the magnitude of the d-axis current;
a memory for storing a minimum phase angle when the minimum value is determined;
an average value calculation unit that calculates an average value of a minimum phase angle during clockwise rotation and a minimum phase angle during counterclockwise rotation as the calibration phase angle;
2. The motor control device according to claim 1, comprising:

Images