Motor control device

Abstract

This motor control device comprises: a current detector (9) that detects a current flowing through a three-phase motor; a coordinate converter (11) that uses the current value to obtain a q-axis current value and a d-axis current value; a q-axis voltage command generation unit (50a) that inputs the q-axis current value and a q-axis current command value to generate a q-axis voltage command value; and a d-axis voltage command generation unit (50b) that inputs the d-axis current value and a d-axis current command value to generate a d-axis voltage command value. The q-axis voltage command generation unit (50a) has: a q-axis current control error detection subtractor (2a) for determining a q-axis current error from the q-axis current value and the q-axis current command value; and a q-axis single frequency component limited repetition controller (200a) for calculating a q-axis current correction value obtained by estimating, from the q-axis current error and an electrical angle, a frequency component that is N times the electrical angle and for adding the q-axis current correction value to the q-axis current error. The q-axis voltage command generation unit (50a) obtains the q-axis voltage command value using the added value.

Description

Motor control device

[0001] The present disclosure relates to a motor controller for controlling a three-phase motor.

[0002] Patent Document 1 describes a motor control device for controlling a three-phase motor. This motor control device samples torque command data only during constant-speed operation and performs averaging processing based on the number of repetitions taking into account the servo gain and the operating range of the motor's rotation angle or travel distance. This determines torque ripple such as cogging and reduces the torque ripple evenly across the motor's operating range.

[0003] Japanese Patent No. 5187172

[0004] However, the technique disclosed in Patent Document 1 has a condition of constant speed, and if this condition is not limited, there is a problem in that torque ripple at a predetermined multiple of the motor electrical angle cannot be suppressed. Examples of cases where the condition is not limited include when the motor is jerking or accelerating / decelerating.

[0005] The present disclosure has been made to solve the above-mentioned problems, and aims to provide a motor control device for a three-phase motor that suppresses torque ripple at a predetermined multiple of the motor's electrical angle or mechanical angle, regardless of the conditions of use.

[0006] A motor control device according to the present disclosure includes a current detector that detects currents of at least two phases out of currents of U-phase, V-phase, and W-phase flowing through a three-phase motor; a coordinate converter that uses the current values ​​detected by the current detector to determine a q-axis current value and a d-axis current value; a q-axis voltage command generation unit that receives the q-axis current value and the q-axis current command value to generate a q-axis voltage command value and outputs it to a dq-three-phase coordinate converter; a d-axis voltage command generation unit that receives the d-axis current value and the d-axis current command value to generate a d-axis voltage command value and outputs it to the dq-three-phase coordinate converter; and a rotation detector that detects the electrical angle of the three-phase motor. The q-axis voltage command generation unit has a q-axis current control error detection subtractor that determines the q-axis current error from the q-axis current value and the q-axis current command value, and a q-axis single frequency component limited repetitive controller that calculates a q-axis current correction value by estimating a frequency component that is N times the electrical angle from the q-axis current error and the electrical angle, and adds the q-axis current correction value to the q-axis current error, and determines the q-axis voltage command value using the value obtained by the addition.

[0007] According to the present disclosure, in a motor control device for controlling a three-phase motor, torque ripples occurring due to various causes can be accurately suppressed without being limited to the use condition of a constant speed state.

[0008] FIG. 1 is a block diagram showing an example of the configuration of a motor control device for a three-phase motor according to a first embodiment. FIG. 2 is a block diagram of a q-axis single frequency component only repetitive controller according to the first embodiment. FIG. 3 is a detailed block diagram of a single frequency component detector according to the first embodiment. FIG. 4 is a time-series waveform showing an example in which the value of the feedback gain K of the single frequency component detector according to the first embodiment is changed in synchronization with a speed pattern. FIG. 5 is a frequency characteristic diagram showing an example of setting the feedback gain K of the single frequency component detector according to the first embodiment. FIG. 6 is a diagram showing open-loop characteristics when the single frequency component only repetitive controller according to the first embodiment is applied to a current control system. FIG. 7 is a diagram showing open-loop characteristics when the single frequency component only repetitive controller according to the first embodiment is applied to a speed control system. FIG. 8 is an experimental result showing the torque ripple suppression effect according to the first embodiment. FIG. 9 is an experimental result of frequency analysis showing the torque ripple suppression effect according to the first embodiment. FIG. 10 is a block diagram showing an example of the configuration of a motor control device for a three-phase motor according to a second embodiment. FIG. 11 is a block diagram showing the configuration of the single frequency component detector in the motor control device shown in FIG. 3. FIG. 12 is a block diagram showing that the q-axis single frequency component only repetitive controller according to the first embodiment can be equivalently converted to the configuration of the q-axis single frequency component only repetitive controller according to the second embodiment. Fig. 10 is a block diagram showing the configuration of a motor control device according to embodiment 3. Fig. 11 is a block diagram showing the configuration of a motor control device according to embodiment 4. Fig. 12 is a diagram showing an example of hardware resources of a motor control device. Fig. 13 is a diagram showing another example of hardware resources of a motor control device.

[0009] The following detailed description will be given with reference to the drawings. Duplicate descriptions will be simplified or omitted as appropriate. In each drawing, the same reference numerals indicate the same or corresponding parts.

[0010] Embodiment 1. Figure 1 is a block diagram showing an example of the configuration of a motor control device for a three-phase motor in embodiment 1. Note that this three-phase motor may be, for example, a motor for driving an elevator car, a motor used in electric power steering, or a motor used to power machine tools. Naturally, the present invention may also be used in motors for other applications.

[0011] 1, a motor control device 100 performs negative feedback control so that the q-axis current and the d-axis current follow set values. The motor control device 100 includes a q-axis current control error detection subtractor 2a, a q-axis single-frequency component limited repetition controller 200a, a q-axis control processing unit 5a, and a q-axis adder 6a, which are provided for q-axis current control, and a d-axis current control error detection subtractor 2b, a d-axis single-frequency component limited repetition controller 200b, a d-axis control processing unit 5b, and a d-axis subtractor 6b, which are provided for d-axis current control, as well as a dq-to-three-phase coordinate converter 7, a three-phase PWM inverter 8, a current detector 9, an AD converter 10 for the current detector 9, a three-phase-to-dq coordinate converter 11, an angular velocity calculator 12, a q-axis decoupling unit 13a, a d-axis decoupling unit 13b, and a rotation detector 14.

[0012] The q-axis current control error detection subtractor 2a, the q-axis single frequency component limited repetition controller 200a, the q-axis control processing unit 5a, and the q-axis adder 6a correspond to the q-axis voltage command generation unit 50a, which receives the q-axis current value and the q-axis current command value and performs processing to output the q-axis voltage command value.

[0013] The d-axis current control error detection subtractor 2b, the d-axis single frequency component limited repetition controller 200b, the d-axis control processing unit 5b, and the d-axis subtractor 6b correspond to the d-axis voltage command generation unit 50b, which inputs the d-axis current value and the d-axis current command value and performs processing to output the d-axis voltage command value.

[0014] In this way, two control loops for the d-axis current and the q-axis current are closed, and a control system that follows the current command values ​​for the d-axis and the q-axis is realized.

[0015] Next, each component will be described in detail. Three-phase motor 1 has windings wound around each of the U, V, and W phases. Three-phase PWM inverter 8 supplies power to drive three-phase motor 1. That is, using power supplied from a power source, it converts a voltage command value supplied from dq-three-phase coordinate conversion unit 7 into power for driving three-phase motor 1 and supplies the converted power.

[0016] In FIG. 1 , vu* indicates the U-phase drive current command voltage supplied from the dq-three-phase coordinate converter 7. vv* indicates the V-phase drive current command voltage supplied from the dq-three-phase coordinate converter 7. vw* indicates the W-phase drive current command voltage supplied from the dq-three-phase coordinate converter 7. iu indicates the current flowing through the U-phase of the three-phase motor 1 supplied from the three-phase PWM inverter 8. iv indicates the current flowing through the V-phase of the three-phase motor 1 supplied from the three-phase PWM inverter 8. iw indicates the current flowing through the W-phase of the three-phase motor 1 supplied from the three-phase PWM inverter 8.

[0017] The current detectors 9 detect the value of the current flowing through each phase of the three-phase motor 1. That is, the U-phase current detector 9a detects the value of the current flowing through the U-phase of the three-phase motor 1. The V-phase current detector 9b detects the value of the current flowing through the V-phase. The W-phase current detector 9c detects the value of the current flowing through the W-phase.

[0018] The AD converter 10 of the current detector 9 AD converts the detected value of the current flowing through each phase of the three-phase motor 1. That is, the U-phase AD converter 10a of the U-phase current detector 9a AD converts the value of the current flowing through the U-phase of the three-phase motor 1. The V-phase AD converter 10b of the V-phase current detector 9b AD converts the value of the current flowing through the V-phase. The W-phase AD converter 10c of the W-phase current detector 9c AD converts the value of the current flowing through the W-phase.

[0019] 1, ius denotes the current detection value of the current flowing through the U phase, which is detected by the U-phase current detection unit 9a and AD converted, ivs denotes the current detection value of the current flowing through the V phase, which is detected by the V-phase current detection unit 9b and AD converted, and iws denotes the current detection value of the current flowing through the W-phase, which is detected by the W-phase current detection unit 9c and AD converted.

[0020] The rotation detector 14 detects the electrical angle of the three-phase motor 1. In FIG.

[0021] The angular velocity calculation unit 12 calculates the electrical angular velocity by differentiating the electrical angle θre with respect to time. In FIG.

[0022] The three-phase-dq coordinate converter 11 calculates the d-axis current value id and the q-axis current value iq based on the current values ​​ius, ivs, iws of each phase and the electrical angle θre detected by the rotation detector 14.

[0023] The q-axis decoupling unit 13a and the d-axis decoupling unit 13b have the function of canceling mutual interference between the d-axis current and the q-axis current. That is, the q-axis decoupling unit 13a in q-axis control calculates the d-axis current value id and the electrical angular velocity ωre and outputs a correction value that cancels interference with the d-axis current. The d-axis decoupling unit 13b in d-axis control calculates the q-axis current value iq and the electrical angular velocity ωre and outputs a correction value that cancels interference with the q-axis current.

[0024] The subtractor 2a for detecting q-axis current control error calculates the error between the q-axis current command value and the q-axis current value iq calculated by the three-phase-dq coordinate converter 11. In Fig. 1, iq* indicates the q-axis current command value, and er_iq indicates the q-axis current error calculated by the subtractor 2a for detecting q-axis current control error.

[0025] The q-axis single frequency component limited repetitive controller 200a installed in the q-axis system has an adder 3a and a single frequency component detector 4a. The single frequency component detector 4a receives the electrical angle θre and the output of the adder 3a, and outputs a q-axis current command correction value nf_sens_q obtained by estimating and extracting a frequency component that is a predetermined multiple of θre of the output of the adder 3a. The adder 3a also adds the q-axis current error er_iq and the q-axis current command correction value nf_sens_q.

[0026] In FIG. 1, the q-axis single frequency component limited repetitive controller 200a is installed between the q-axis current control error detection subtractor 2a and the q-axis control processing unit 5a, but it may be installed at any position in the q-axis voltage command generating unit 50a.

[0027] The q-axis control processing unit 5a receives the output of the adder 3a, which is the output of the q-axis single frequency component limited repetitive controller 200a, and outputs the results of compensation processing so that the q-axis current control is stabilized within a predetermined control band.

[0028] The q-axis adder 6a outputs a q-axis voltage command value vq* obtained by adding the output of the q-axis control processing unit 5a and the correction value output from the q-axis non-interference unit 13a.

[0029] As in the q-axis current control, the subtractor 2b for detecting d-axis current control error calculates the error between the d-axis current command value and the d-axis current value id calculated by the three-phase-dq coordinate converter 11. In Fig. 1, id* indicates the d-axis current command value, and er_id indicates the d-axis current error calculated by the subtractor 2b for detecting d-axis current control error.

[0030] The d-axis single-frequency component limited repetitive controller 200b installed in the d-axis system has an adder 3b and a single-frequency component detector 4b. The single-frequency component detector 4b receives the electrical angle θre and the output of the adder 3b, and outputs a d-axis current command correction value nf_sens_d obtained by estimating and extracting a frequency component that is a predetermined multiple of θre of the output of the adder 3b. The adder 3b also adds the d-axis current error er_id and the d-axis current command correction value nf_sens_d.

[0031] In FIG. 1, the d-axis single frequency component limited repetitive controller 200b is installed between the d-axis current control error detection subtractor 2b and the d-axis control processing unit 5b, but it may be installed at any position in the d-axis voltage command generating unit 50b.

[0032] The d-axis control processing unit 5b receives as input the output of the adder 3b, which is the output of the d-axis single frequency component limited repetitive controller 200b installed for d-axis control, and outputs the results of compensation processing so that the d-axis current control is stabilized within a specified control band.

[0033] The d-axis subtractor 6b outputs a d-axis voltage command value vd* obtained by subtracting the correction value output from the d-axis non-interference unit 13b in the d-axis control from the output of the d-axis control processing unit 5b.

[0034] The dq-three-phase coordinate conversion unit 7 calculates voltage command values ​​vu*, vv*, vw* for each phase of the three-phase motor 1 based on the d-axis voltage command value vd*, the q-axis voltage command value vq*, and the electrical angle θre.

[0035] With the above configuration, two control loops for the d-axis current and the q-axis current are closed, and a control system that follows the current command values ​​for the d-axis and q-axis is realized.

[0036] Next, the function of the q-axis single frequency component limited repetitive controller 200a will be described in detail.

[0037] 2 is a block diagram of a q-axis single frequency component limited repetition controller 200a. In fact, FIG. 2 shows only the portion of the q-axis single frequency component limited repetition controller 200a shown in FIG.

[0038] Here, the q-axis single frequency component limited repetitive controller 200a is configured to include an adder 3a and a single frequency component detector 4a, as already described. The electrical angle θre and the output of the adder 3a are input to the single frequency component detector 4a, and the adder 3a outputs a q-axis current command correction value nf_sens_q obtained by estimating and extracting a frequency component that is a predetermined multiple of θre from the output of the adder 3a. The adder 3a adds the q-axis current error er_iq and the q-axis current command correction value nf_sens_q.

[0039] Here, the single frequency component detector 4a is a mechanism that detects in real time disturbances such as torque ripples of a predetermined multiple of the electrical angle in the q-axis current control loop and generates the disturbances. The q-axis single frequency component limited repetitive controller 200a is configured as a positive feedback loop that includes a disturbance generating mechanism, similar to the so-called repetitive control in Patent Document 1.

[0040] 3 is a detailed block diagram of single frequency component detector 4a, which mainly includes a magnification setting unit 4a1, an electrical angle multiplier 4a2, a cosine calculation unit 4a3, a sine calculation unit 4a4, an internal subtractor 4a5, a first multiplier 4a6, a first magnification integrator 4a7, a second multiplier 4a8, a third multiplier 4a9, a second magnification integrator 4a10, a fourth multiplier 4a11, and an internal adder 4a12.

[0041] The electrical angle θre is multiplied by the magnification N set by the magnification setting unit 4a1 in the electrical angle multiplier 4a2 to obtain Nθre. N is generally set to an integer, but is not limited to this and may be a fraction. This value is converted into a cosine value cos(Nθre) in the cosine calculation unit 4a3, and similarly converted into a sine value sin(Nθre) in the sine calculation unit 4a4.

[0042] 2, the input to the single frequency component detector 4a is the sum of the q-axis current error er_iq and the q-axis current command correction value nf_sens_q, which is the output of the single frequency component detector 4a. This is shown in equation (1).

[0043]

[0044] The output from the single frequency component detector 4a is expressed by equation (2).

[0045]

[0046] The internal subtractor 4a5 subtracts the output (output) from the input (input) and outputs an error er_sens_q. The first multiplier 4a6 multiplies the error er_sens_q by a cosine value cos(Nθre) and outputs the result. Similarly, the third multiplier 4a9 multiplies the error er_sens_q by a sine value sin(Nθre) and outputs the result.

[0047] First multiplier integrator 4a7 multiplies the output of first multiplier 4a6 by feedback gain K, performs time integration, and outputs the result as -α'. Similarly, second multiplier integrator 4a10 multiplies the output of third multiplier 4a9 by K, performs time integration, and outputs the result as -β'. K is a predetermined multiplier, and details will be described later.

[0048] The second multiplier 4a8 multiplies -α' by the cosine value cos(Nθre) and outputs the result. Similarly, the fourth multiplier 4a11 multiplies -β' by the sine value sin(Nθre) and outputs the result. The internal adder 4a12 adds the outputs of the second multiplier 4a8 and the fourth multiplier 4a11 and outputs the result as the q-axis current command correction value nf_sens_q.

[0049] Thus, the single frequency component detector 4a, excluding the portion at the bottom that negatively feeds back the output (OUTPUT) to the input (INPUT), is configured to estimate a frequency component that is N times the electrical angle θre contained in the input when the value of K is 1. Furthermore, by negatively feeding back this estimated value to the input, the single frequency component detector 4a is configured to converge so that the error between the actual value of the frequency component signal that is N times the electrical angle θre contained in the input and the estimated value becomes zero. Therefore, it is desirable to set the feedback gain K of the first multiplying integrator 4a7 to a value greater than 1 to improve convergence responsiveness. Note that the first embodiment will be described assuming that the value of K is set to a value sufficiently greater than 1.

[0050] The estimation characteristics of the frequency component N times the electrical angle θre in the single frequency component detector 4a will be described below with reference to Fig. 3. The q-axis current command correction value nf_sens_q of the single frequency component detector 4a is expressed by equation (3).

[0051]

[0052] Furthermore, real_dis, which is a frequency component N times the electrical angle θre to be detected, is defined as shown in equation (4).

[0053]

[0054] Here, it is assumed that the q-axis current error er_iq is approximately equal to the frequency component real_dis that is N times the electrical angle θre to be detected. As a result, the estimated error er_sens_q of the single frequency component detector 4a is expressed as in equation (5).

[0055]

[0056] Here, −α′ and −β′ are expressed by equations (6) and (7), respectively.

[0057]

[0058]

[0059] Here, in equations (6) and (7), there is a double angle term of Nθre in the estimation coefficients. If this is ignored, equation (6) becomes equation (8), which ultimately becomes equation (9), and equation (7) becomes equation (10), which ultimately becomes equation (11).

[0060]

[0061]

[0062]

[0063]

[0064] From these equations (9) and (11), it can be seen that the two estimation coefficients α′β′ basically have the detection characteristics of a first-order LPF (low-pass filter) with a cutoff frequency of 0.5K [rad / sec] relative to the true value. Equation (12) is the characteristic equation of the first-order LPF.

[0065]

[0066] Then, by substituting the results of equations (9), (11), and (12) into equation (3), equation (13) is obtained.

[0067]

[0068] From this, it can be seen that the LPF shown in equation (12) is an index that determines the estimable band of the single frequency component detector 4a. Furthermore, if the single frequency component detector 4a is configured as shown in FIG. 3, it can function as a sensor that detects frequency components that are N times the electrical angle contained in the input. The detectable frequency band of the single frequency component detector 4a as a sensor is the cutoff frequency of the LPF. To broaden the detectable band and improve responsiveness, the cutoff frequency can be increased, and the value of the feedback gain K can be increased.

[0069] However, in reality, there is a double angle term, so blindly increasing the value of K will cause problems. Specifically, if K is increased too much, the double angle component will be amplified, which is a trade-off. The reason for this will be explained below.

[0070] Actual estimated characteristics without ignoring the double angle term are expressed as in equations (14) and (15).

[0071]

[0072]

[0073] In equation (14) formulating the cosine estimation coefficient α', the second term on the right side contains a term for the estimation error of the sine coefficient, which is unrelated to cosine. Similarly, in equation (15) formulating the sine estimation coefficient β', the second term on the right side contains a term for the estimation error of the cosine coefficient, which is unrelated to sine.

[0074] These two estimation coefficients α′ and β′ are defined as mutual interference terms, and since these are terms that should be nullified, the key to design is to make the functions related to these terms small, i.e., the functions shown in Equation (16) and Equation (17).

[0075]

[0076]

[0077] These are first-order LPF characteristics. The cutoff frequency is expressed as a trigonometric function, so it fluctuates periodically from moment to moment within the range of -1 to 1. The DC gain of this LPF characteristic is maximized when the LPF numerator sin2Nθre reaches its maximum value of 1, which is when, for example, equation (18) holds.

[0078]

[0079] In this worst case, the LPF characteristics are those of a first-order LPF with a cutoff frequency of 0.5K [rad / sec] for both the cosine estimation coefficient α' and the sine estimation coefficient β'. This is the same characteristic as the LPF that determines the estimable band of the single frequency component detector 4a shown in equation (12).

[0080] The trigonometric function term to be nullified is a double angle of the frequency component to be detected, and therefore can be attenuated at least as long as the cutoff frequency of the LPF shown in equation (12) is lower than this. In other words, if the feedback gain K is set to a value equal to or lower than the double angle frequency of the trigonometric function term, it is possible to suppress the double angle component that is unnecessary for at least the frequency to be detected. This relational expression is shown in equation (19).

[0081]

[0082] Here, ω0 [rad / sec] is the frequency to be detected and suppressed by the system, and equation (20) holds.

[0083]

[0084] In this application for suppressing the N-fold angle component of the electrical angle, ω0 is Nf [rad / sec] (f is the electrical angle frequency) as shown in equation (21), and equations (22) and (23) result, so the recommended upper limit value of K is 2Nf [rad / sec].

[0085]

[0086]

[0087]

[0088] In actual operation, a three-phase motor is used as a power source for a specific device, and in many cases, its rotation speed starts from zero, reaches a specific rotation speed, and then returns to zero. In this case, the recommended upper limit value for K should be set to change accordingly, since Nf changes from moment to moment depending on the rotation speed.

[0089] Figure 4 shows an example in which the value of feedback gain K is changed in synchronization with the speed pattern. In the figure, (a) is a time-series waveform of the motor rotation speed. (b) is a time-series waveform of the upper limit of feedback gain K, time-aligned with the waveform in (a). Nf is proportional to the rotation speed, and as mentioned above, the recommended upper limit of K is 2Nf [rad / sec]. Therefore, it is desirable to define the recommended upper limit of K in a proportional relationship synchronized with the rotation speed, as shown in Figure 4. In this case, if the upper limit of the motor rotation speed is ωmax [rad / sec], the recommended upper limit of feedback gain K will be 4ωmax [rad / sec].

[0090] Next, the setting value of the feedback gain K will be described. FIG. 5 shows the estimated characteristics of the single frequency component detector 4a, and more specifically, is a frequency characteristic diagram showing an example of the setting of the feedback gain K in the LPF shown in equation (12). As mentioned above, this frequency characteristic is a first-order LPF characteristic, with a cutoff frequency of 0.5K [rad / sec] and a DC gain of 0 [dB]. From equation (22), the cutoff frequency of 0.5K [rad / sec] is set to a band lower than 2Nf [rad / sec]. Basically, this setting is desirable.

[0091] However, it is preferable to further broaden the estimation bandwidth and quickly respond to momentary changes in the frequency ω0 [rad / sec] to be detected and suppressed, thereby outputting accurate values ​​in real time. For this reason, as shown in Figure 5, it is desirable to set the lower limit of the cutoff frequency 0.5K [rad / sec] to the estimated disturbance frequency Nf [rad / sec]. This setting allows the estimation characteristics of the single frequency component detector 4a to be both fast and accurate, preventing the inclusion of unnecessary information.

[0092] Next, we will explain from the perspective of control loop stability why a single-frequency component repetitive controller is introduced not into the speed control system but into the current control system, which is the inner loop of speed control. A single-frequency component repetitive controller applies the internal model principle. Therefore, the configuration of a single-frequency component repetitive controller is equivalent to repetitive control in that it has a disturbance generating mechanism within the positive feedback loop. This makes it possible to neutralize and suppress to approximately zero the steady-state error caused by torque ripple disturbances of a predetermined multiple of the electrical angle.

[0093] In the first embodiment, the amplitude and phase of disturbance components at a frequency that is a predetermined multiple of the electrical angle are accurately detected in real time by the single frequency component detector 4a, which is configured as a negative feedback loop for the disturbance generating mechanism. This makes it possible to constantly suppress the control deviation caused by disturbances at a frequency that is a predetermined multiple of the electrical angle to almost zero, even during acceleration or jerk, when the period of the motor electrical angle changes from moment to moment, which was not possible with conventional repetitive control using a memory that stores periodic components.

[0094] Figures 6 and 7 show the open loop characteristics when a single frequency component limited repetitive controller is applied to a control system. Figure 6 shows the case when it is applied to a current control system, and Figure 7 shows the case when it is applied to a speed control system. To simplify the discussion, the open loop characteristics are assumed to be ideal characteristics with a first-order phase margin of 90 degrees. In motor control, the current control loop and the speed control loop are in a relationship where the current control loop is contained within the speed control loop as an inner loop. The crossover frequency of the current control loop is set to be one or two orders of magnitude higher than that of the speed control loop. The crossover frequency in the open loop characteristics of the speed control loop is set to ω ｖ [rad / sec], and the crossover frequency in the current control loop is ω ｉ [rad / sec], then equation (24) holds.

[0095]

[0096] 6 and 7, a significant peak is seen in the gain characteristics at the N-fold period component (ω0) of the motor's electrical angle, but this is because the open loop gain ideally becomes infinite due to the effect of repetitive control. The phase characteristics near ω0 [rad / sec] are such that, compared to the phase rotation characteristic of the original control loop of -90 [deg] up to ω0, the phase advances to 90 [deg], becomes 0 [deg] at ω0, and after ω0 the phase rotates to -90 [deg] before returning to 0 [deg].

[0097] 6 and 7, the white arrow at the top indicates the region where the N-fold period component (ω0) of the motor's electrical angle changes. The right end of the white arrow indicates the maximum value of ω0, which is the case when the motor rotation speed is at its highest. The dotted line shows the open-loop characteristics at this time, and the solid line shows an example when the motor rotation speed is a predetermined value below the maximum speed.

[0098] In the open loop characteristics when the single frequency component limited repetitive controller is applied to the speed control loop of FIG. 7, the crossover frequency ω ｖ [rad / sec] is designed to be a relatively low frequency. Therefore, it is included in the change region of the N-fold period component (ω0) of the motor's electrical angle, as indicated by the white arrow at the top. In this case, the motor functions stably at low speeds. However, as the rotation speed increases, the phase rotation due to the effect of the single frequency component limited repetitive controller reaches -90 degrees, and when the frequency reaches the crossover frequency, the phase margin disappears and the speed control loop oscillates. Therefore, it is clear that applying a single frequency component limited repetitive controller to the speed control loop is not appropriate due to stability issues.

[0099] On the other hand, in the open loop characteristics when a single frequency component limited repetitive controller is applied to the current control loop shown in FIG. 6, the crossover frequency ω ｉ [rad / sec] is in a band sufficiently higher than the change region of the N-fold period component (ω0) of the motor's electrical angle indicated by the white arrow at the top. Therefore, the phase rotation by applying the single frequency component limited repetitive controller is ｉAs shown in the figure, even when the motor rotation speed reaches the maximum speed, the crossover frequency ω of the current control loop is ｉ Therefore, even in an operating mode in which the motor rotation speed varies from zero to maximum speed, by applying a single frequency component limited repetitive controller to the current control loop, it is possible to effectively suppress the N-fold period component (ω0) of the motor electrical angle without impairing the stability of the current control loop.

[0100] In this way, the case where the q-axis single frequency component restricted repetitive controller 200a is applied to the q-axis current control loop has been described. The case where the d-axis single frequency component restricted repetitive controller 200b is applied to the d-axis current control loop also has the same structure. Here, the same logical development produces the same effect, so the description thereof will be omitted.

[0101] Next, we will explain the results of an experiment using the configuration shown in Figure 1 on a motor load test device. Figure 8 shows the results of an experiment demonstrating the torque ripple suppression effect of the first embodiment. Specifically, the magnetic circuit impedance of only one phase of a three-phase motor was doubled to intentionally create a system in which torque ripple, which is a vibration disturbance of the doubled component of the electrical angle, was generated. The results were obtained by applying the q-axis single frequency component limited repetition controller 200a and the d-axis single frequency component limited repetition controller 200b to both the q-axis current control loop and the d-axis current control loop. Note that the value of N in the single frequency component detectors 4a and 4b was set to 2 because the torque ripple to be suppressed is a doubled component of the electrical angle.

[0102] In Figure 8, the top waveform (a) represents the motor rotation speed. The middle waveform (b) represents the d-axis current. The bottom waveform (c) represents the q-axis current. In each of these figures, the left side represents the case where the q-axis single frequency component limited repetitive controller 200a and the d-axis single frequency component limited repetitive controller 200b are not applied, and the right side represents the case where they are applied.

[0103] The experimental conditions were as follows: the motor was rotated from zero speed to a constant speed, and the load torque was changed stepwise. Because the q-axis current is proportional to the torque waveform, the temporal change in the load torque can be determined. Under these experimental conditions, the speed waveform fluctuated significantly at the timing of the torque change, but converged to a constant value (2.4 m / s).

[0104] In the d-axis current waveform (b) and the q-axis current waveform (c), the waveform on the left has a double-angle component of the electrical angle superimposed, making the waveform appear thicker, and it can be seen that the amplitude tends to increase as the load torque increases. In the waveform on the right, which uses the q-axis single-frequency component limited repetitive controller 200a and the d-axis single-frequency component limited repetitive controller 200b, the superimposed double-angle component of the electrical angle is suppressed, resulting in the thickness of the d-axis current waveform and the q-axis current waveform becoming thinner. Furthermore, even when the speed is not constant due to a sudden change in torque, the effect remains unchanged, confirming that the double-angle component of the electrical angle is effectively suppressed.

[0105] However, because the waveform in Figure 8 is time-series data, disturbances of frequencies other than the electrical angle double angle segments are mixed in and superimposed. Random noise components are also superimposed. This makes it difficult to clarify the effect. Therefore, the effect will be confirmed by frequency analysis using FFT (Fast Fourier Transform).

[0106] Figure 9 shows the results of an experiment using frequency analysis that demonstrates the torque ripple suppression effect. Specifically, it shows the results of FFT analysis of the data in the region where the speed remains constant and where the d-axis and q-axis current waveforms are thickest in the time-series data from the experiment shown in Figure 8. Specifically, the time-series data is from 18 seconds to 20 seconds.

[0107] In Figure 9, the upper (a) waveform shows the d-axis current waveform obtained by FFT analysis. The lower (b) waveform shows the q-axis current waveform obtained by FFT analysis. In both figures, the dotted lines indicate the cases without application of the q-axis single frequency component limited repetitive controller 200a and the d-axis single frequency component limited repetitive controller 200b, and the solid lines indicate the cases with application. The peak of just under 70 Hz in the waveform without application is a component superimposed on the current control deviation caused by torque ripple of twice the electrical angle.

[0108] Comparing the dotted line and the solid line reveals that applying the q-axis single-frequency component-only repetition controller 200a and the d-axis single-frequency component-only repetition controller 200b eliminated the frequency amplitude and improved the noise level. The suppression effect was -50 dB. Furthermore, not only were the double-angle components of the electrical angle suppressed, but frequency components in the surrounding bands were also suppressed. It was also confirmed that there was no change or a slight attenuation tendency at the second peak of 140 Hz or higher. This confirms that applying the q-axis single-frequency component-only repetition controller 200a and the d-axis single-frequency component-only repetition controller 200b not only provides an effective solution, but also does not result in the amplification of components in a specific band.

[0109] As described above, the motor control device described in the first embodiment can accurately estimate, in real time, moment-to-moment changes in both the amplitude and phase of torque ripple that is a predetermined multiple of the motor's electrical angle. There is no need to limit the application conditions, or to perform a predetermined learning operation or function to activate it. Therefore, regardless of whether the motor is jerking or accelerating or decelerating, torque ripple that is a predetermined multiple of the motor's electrical angle can be effectively suppressed without performing a learning operation.

[0110] In the first embodiment, the current values ​​of the U-phase, V-phase, and W-phase are detected by the U-phase current detector 9a, the V-phase current detector 9b, and the W-phase current detector 9c, but it is sufficient to detect at least two of these. If only two current values, for example, the U-phase and the V-phase, are detected, the current value of the remaining phase, the W-phase, must be calculated.

[0111] In this first embodiment, the case where there is a single torque ripple frequency component has been described, but the present disclosure can also be applied to cases where there are multiple torque ripple frequency components, and similar effects can be achieved. In this case, the q-axis single frequency component only repetition controller 200a and the d-axis single frequency component only repetition controller 200b are prepared in the number corresponding to the frequency component of the torque ripple to be suppressed, and inserted in series or parallel into the current control loop, as in the configuration shown in Figure 1. Note that the variable setting for each of the multiple q-axis single frequency component only repetition controllers 200a and d-axis single frequency component only repetition controllers 200b is simply set to the value N of the single frequency component detectors 4a, 4b, which corresponds to the frequency component of the torque ripple to be suppressed.

[0112] 10 is a block diagram showing an example of the configuration of a motor control device for a three-phase motor in embodiment 2. The configurations of a q-axis single frequency component limited repetitive controller 200a and a d-axis single frequency component limited repetitive controller 200b differ from those of embodiment 1.

[0113] That is, in the q-axis single frequency component limited repetitive controller 200a of the first embodiment, the electrical angle θre and the output of the adder 3a are input to the single frequency component detector 4a, and the q-axis current command correction value nf_sens_q, which is the output of the single frequency component detector 4a, is added to er_iq by the adder 3a.

[0114] 10, in a q-axis single frequency component limited repetitive controller 200a according to the second embodiment, the electrical angle θre and the q-axis current error er_iq from the q-axis current control error detection subtractor 2a are input to a single frequency component detector 4a'. Then, the q-axis current command correction value nf_sens_q output from the single frequency component detector 4a' is added to er_iq by an adder 3a'.

[0115] Fig. 11 is a block diagram showing a single frequency component detector 4a' according to embodiment 2. Fig. 12 is a block diagram showing the configuration of the single frequency component detector 4a in the motor control device shown in Fig. 3, but with dotted lines added. The portion with the dotted lines excluding the internal subtractor 4a5 that negatively feeds back the output of the single frequency component detector 4a to its input.

[0116] The portion enclosed by the dotted line in Fig. 12 corresponds to single frequency component detector 4a' in Fig. 11. That is, single frequency component detector 4a' is mainly composed of magnification setting unit 4a1, electrical angle multiplier 4a2, cosine calculation unit 4a3, sine calculation unit 4a4, first multiplier 4a6, first magnification integrator 4a7, second multiplier 4a8, third multiplier 4a9, second magnification integrator 4a10, fourth multiplier 4a11, and internal adder 4a12.

[0117] 13 is a block diagram showing that the q-axis single-frequency component restricted repetition controller 200a of the first embodiment can be equivalently converted to the configuration of the q-axis single-frequency component restricted repetition controller 200a of the second embodiment. The upper part (a) shows an excerpt of the input / output portion of the q-axis single-frequency component restricted repetition controller 200a of the motor control device of the first embodiment shown in FIG. 1. The lower part (b) shows an excerpt of the input / output portion of the q-axis single-frequency component restricted repetition controller 200a of the motor control device of the second embodiment shown in FIG.

[0118] In (a) showing the first embodiment, the q-axis single frequency component only repetition controller 200a has a positive feedback loop configuration. Therefore, the negative feedback portion of the single frequency component detector 4a and the positive feedback loop configuration of the q-axis single frequency component only repetition controller 200a cancel each other out. As a result, the configuration is equivalently converted to the configuration shown in (b) showing the second embodiment.

[0119] Therefore, although the internal configuration of the q-axis single frequency component limited repetitive controller 200a is different in (a) and (b), it is equivalent, and therefore the behavior is exactly the same, and the function, performance, and effect are also the same.

[0120] The configuration of embodiment 2 cancels out and omits the positive feedback processing of the repetitive control and the negative feedback processing of the single frequency component detector, which has the effect of simplifying the processing when realizing a single frequency component limited repetitive controller through software processing.

[0121] As described above, the motor control device described in the second embodiment, like the first embodiment, can accurately estimate, in real time, changes in both the amplitude and phase of torque ripple that is a predetermined multiple of the motor's electrical angle. There is no need to limit the application conditions, or to perform a predetermined learning operation or function to operate it. Therefore, regardless of whether the motor is jerking or accelerating or decelerating, torque ripple that is a predetermined multiple of the motor's electrical angle can be effectively suppressed without performing a learning operation.

[0122] Embodiment 3. Figure 14 is a block diagram showing the configuration of a motor control device according to embodiment 3. In Figure 14, in the motor control device according to embodiment 1 in Figure 1, only the q-axis current control system has a single frequency component limited repetitive controller.

[0123] The torque generated by the motor is mainly defined as a predetermined multiplication of the q-axis current.

[0124] It is true that the torque ripple suppression effect of a predetermined multiple of the motor electrical angle in embodiment 3 is affected by the residual interference term of the d-axis current control, compared to embodiment 1. However, if this residual amount is less than the reference value of the motor torque ripple, the effect is small.

[0125] Therefore, by applying the single frequency component limited repetitive controller only to the q-axis current control, omitting the application of the d-axis current control which is not directly related to torque, the current control processing system can be made simpler.

[0126] As described above, the motor control device described in embodiment 3, like embodiment 1, can accurately estimate, in real time, changes in both the amplitude and phase of torque ripple that is a predetermined multiple of the motor's electrical angle. There is no need to limit the application conditions, or to perform a predetermined learning operation or function to operate it. Therefore, regardless of whether the motor is jerking or accelerating or decelerating, torque ripple that is a predetermined multiple of the motor's electrical angle can be effectively suppressed without performing a learning operation.

[0127] Embodiment 4. Figure 15 is a block diagram showing the configuration of a motor control device according to embodiment 4. In Figure 15, in the motor control device according to embodiment 2 in Figure 10, only the q-axis current control system has a single frequency component limited repetitive controller.

[0128] The torque generated by the motor is mainly defined as a predetermined multiplication of the q-axis current.

[0129] It is true that the torque ripple suppression effect of a predetermined multiple of the motor electrical angle in embodiment 4 is affected by the residual influence of the interference term in d-axis current control, compared to embodiment 2. However, if this residual amount is less than the reference value of the motor torque ripple, the effect is small.

[0130] Therefore, by applying the single frequency component limited repetitive controller only to the q-axis current control, omitting the application of the d-axis current control which is not directly related to torque, the current control processing system can be made simpler.

[0131] As described above, the motor control device described in embodiment 4, like embodiment 1, can accurately estimate torque ripple that is a predetermined multiple of the motor's electrical angle by tracking moment-to-moment changes in both amplitude and phase in real time. There is no need to limit application conditions, or to perform a predetermined learning operation or function to activate it. Therefore, torque ripple that is a predetermined multiple of the motor's electrical angle can be effectively suppressed without performing a learning operation, regardless of whether the motor is jerking or accelerating or decelerating.

[0132] 16 is a diagram showing an example of hardware resources of the motor control device 100. The motor control device 100 includes, as hardware resources, a processor 101 and a memory 102. Note that there may be multiple processors 101 and multiple memories 102.

[0133] In the first to fourth embodiments, the motor control device performs calculations and stores information. The storage of this information is performed in memory 102. Furthermore, calculation processing is performed by processor 101 using software, firmware, or a combination of software and firmware written as a program stored in memory 102.

[0134] The processor 101 is also called a CPU (Central Processing Unit), central processing unit, arithmetic unit, microprocessor, microcomputer, or DSP. The memory 102 may be a semiconductor memory, a magnetic disk, a flexible disk, an optical disk, a compact disk, a minidisk, or a DVD. Usable semiconductor memories include RAM, ROM, flash memory, EPROM, and EEPROM.

[0135] FIG. 17 is a diagram showing another example of hardware resources of the motor control device 100. In the example of FIG. 17, the motor control device 100 includes a processing circuit including a processor 101, a memory 102, and dedicated hardware 103. FIG. 17 shows an example in which some of the functions of the motor control device 100 are implemented by the dedicated hardware 103. It is also possible to implement all of the functions of the motor control device 100 by the dedicated hardware 103. The dedicated hardware 103 can be a single circuit, a composite circuit, a programmed processor, a parallel programmed processor, an ASIC, an FPGA, or a combination of these.

[0136] The motor control device according to the present disclosure can be applied to control three-phase motors such as various drive motors for elevators, automobiles, and industrial machines.

[0137] 1 Three-phase motor, 2a q-axis current control error detection subtractor, 2b d-axis current control error detection subtractor, 3a adder, 3b adder, 4a single frequency component detector, 4a1 magnification setting unit, 4a2 electrical angle multiplier, 4a3 cosine calculation unit, 4a4 sine calculation unit, 4a5 internal subtractor, 4a6 first multiplier, 4a7 first magnification integrator, 4a8 second multiplier, 4a9 third multiplier, 4a10 second magnification integrator, 4a11 fourth multiplier, 4a12 internal adder, 4b single frequency component detector, 5a q-axis control processing unit, 5b d-axis control processing unit, 6a q-axis adder, 6b d-axis subtractor, 7 dq-three-phase coordinate conversion unit, 8 three-phase PWM inverter, 9a U-phase current detection unit, 9b V-phase current detection unit, 9c W-phase current detection unit, 10a U-phase AD conversion unit, 10b V-phase AD conversion unit, 10c W-phase AD conversion unit, 11 three-phase-dq coordinate conversion unit, 12 angular velocity calculation unit, 13a q-axis decoupling unit, 13b d-axis decoupling unit, 14 rotation detector, 50a q-axis voltage command generation unit, 50b d-axis voltage command generation unit, 100 motor control device, 101 processor, 102 memory, 103 dedicated hardware, 200a q-axis single frequency component limited repetitive controller, 200b d-axis single frequency component limited repetitive controller

Claims

1. A current detector that detects the currents of at least two of the U-phase, V-phase, and W-phase currents flowing through a three-phase motor, A coordinate converter that uses the current value detected by the current detector to determine the q-axis current value and the d-axis current value, A q-axis voltage command generation unit that receives the q-axis current value and the q-axis current command value as input, generates a q-axis voltage command value, and outputs it to a dq-three-phase coordinate converter, A d-axis voltage command generation unit that receives the d-axis current value and the d-axis current command value as input, generates a d-axis voltage command value, and outputs it to the dq-three-phase coordinate converter, A three-phase inverter supplies power to the three-phase motor based on the three-phase voltage command values ​​output from the dq-three-phase coordinate converter, The three-phase motor is equipped with a rotation detector that detects the electrical angle, The q-axis voltage command generation unit includes a subtractor for detecting q-axis current control errors that calculates the q-axis current error from the q-axis current value and the q-axis current command value, The system includes a q-axis single-frequency component limited repeating controller that calculates a q-axis current correction value by estimating a frequency component that is N times the electrical angle from the q-axis current error and the electrical angle, and adds the q-axis current correction value to the q-axis current error. Using the added value, the q-axis voltage command value is determined. The q-axis single-frequency component limited repeating controller has a single-frequency component detector that internally feeds back estimated values ​​to the input as negative feedback. A motor control device characterized by the following features.

2. The q-axis single-frequency component limited repeating controller has an adder, The q-axis current correction value is output from the single-frequency component detector. The adder adds the q-axis current error and the q-axis current correction value and outputs the result. The motor control device according to claim 1, characterized in that the output from the adder is input to the single-frequency component detector.

3. The motor control device according to claim 2, characterized in that the electrical angle is input to the single-frequency component detector.

4. The single-frequency component detector is An electrical angle multiplier that multiplies the aforementioned electrical angle by N to generate an electrical angle corresponding to the torque ripple frequency, A cosine calculation unit that calculates the cosine value of the N-folded electrical angle, A sine calculation unit that calculates the sine value of the N-folded electric angle, An internal subtractor that detects the error between the output of the adder and the output of the single-frequency component detector. A first multiplier that multiplies the error by the output of the cosine calculation unit, A first multiplier integrator that integrates the output of the first multiplier with respect to the amplification factor, A second multiplier that multiplies the output of the first multiplier integrator and the output of the cosine calculation unit, A third multiplier that multiplies the error by the output of the sine function unit, A second multiplier integrator that integrates the output of the third multiplier by the amplification factor, A fourth multiplier that multiplies the output of the second multiplier integrator and the output of the sine arithmetic unit, It comprises an internal adder that adds the output of the second multiplier and the output of the fourth multiplier, The motor control device according to claim 3, characterized in that the output of the internal adder is set to the q-axis current correction value.

5. The q-axis single-frequency component limited repeating controller has an adder, The q-axis current correction value is output from the single-frequency component detector. The adder adds the q-axis current error and the q-axis current correction value and outputs the result. The motor control device according to claim 1, characterized in that the q-axis current error is input to the single-frequency component detector.

6. The motor control device according to claim 5, characterized in that the electrical angle is input to the single-frequency component detector.

7. The single-frequency component detector is An electrical angle multiplier that multiplies the aforementioned electrical angle by N to generate an electrical angle corresponding to the torque ripple frequency, A cosine calculation unit that calculates the cosine value of the N-folded electrical angle, A sine calculation unit that calculates the sine value of the N-folded electric angle, A first multiplier that multiplies the q-axis current error by the output of the cosine calculation unit, A first multiplier integrator that integrates the output of the first multiplier with respect to the amplification factor, A second multiplier that multiplies the output of the first multiplier integrator and the output of the cosine calculation unit, A third multiplier that multiplies the q-axis current error by the output of the sine function unit, A second multiplier integrator that integrates the output of the third multiplier by the amplification factor, A fourth multiplier that multiplies the output of the second multiplier integrator and the output of the sine arithmetic unit, It comprises an internal adder that adds the output of the second multiplier and the output of the fourth multiplier, The motor control device according to claim 6, characterized in that the output of the internal adder is set to the q-axis current correction value.

8. The motor control device according to 4 or 7, characterized in that the amplification factors of the first multiplier integrator and the second multiplier integrator are greater than 1 and less than 4 times the torque ripple frequency of the three-phase motor.

9. The d-axis voltage command generation unit includes a subtractor for detecting d-axis current control errors that calculates the d-axis current error from the d-axis current value and the d-axis current command value, The device includes a d-axis single-frequency component limited repeating controller that calculates a d-axis current correction value by estimating a frequency component that is N times the electrical angle from the d-axis current error and the electrical angle, and adds the d-axis current correction value to the d-axis current error. The motor control device according to any one of claims 1 to 7, characterized in that the d-axis voltage command value is determined using the summed value.