JP7055001B2 - Motor control method and motor control device - Google Patents

Description

The present invention relates to a motor control method and a motor control device.

In a motor, when the rotational speed is controlled, the inductance of the motor may fluctuate due to magnetic non-linearity and magnetic saturation. Therefore, as in the technique disclosed in Patent Document 1, the magnetic flux chain crossings of each winding with respect to the current value are obtained in advance by analysis such as the non-linear finite element method of the motor, and these current values and the magnetic flux chain crossings are used. There is known a technique for controlling the torque of a motor or the like by using a data table corresponding to the above. With this technique, control parameters can be optimized and motor control with higher accuracy can be realized.

Japanese Unexamined Patent Publication No. 2008-141835

Here, among the motors, there is a motor called a variable magnetic flux type (or variable leakage flux type) having a structure in which a magnetic barrier such as a physical gap is provided in the rotor. In this variable magnetic flux type motor, a part of the magnetic flux from the permanent magnet to the stator of the rotor is configured to be blocked by a magnetic barrier, and the amount of magnetic flux blocked by this magnetic barrier is the motor. It changes according to the applied current to. Therefore, since the amount of magnetic flux from the rotor to the stator changes according to the magnitude of the applied current, there is a feature that the amount of change in inductance is relatively large.

In this variable magnetic flux type motor, the inductance is greatly changed by the applied current due to the provision of the magnetic barrier. Therefore, the calculation accuracy of the inductance is not sufficiently high even if the analysis method disclosed in Patent Document 1 is used. Therefore, even if the torque, speed, position, etc. are controlled by using the calculated inductance, the calculation accuracy is high. Rotation control may not be possible. Therefore, there is a demand for a technique for further improving the accuracy of rotation control.

The present invention has been made focusing on such a problem, and an object of the present invention is to provide a motor control method capable of improving the accuracy of motor rotation control, and a motor control device.

One aspect of the motor control method of the present invention is to change the magnetic flux from the permanent magnet to the stator according to the current applied from the inverter to the motor by a magnetic barrier provided in the rotor provided with the permanent magnet. This is a possible variable magnetic flux type motor control method. In this control method, the interlinkage magnetic flux of the motor is estimated by the first feedback control of the applied current, and the first target voltage is set so that the estimated interlinkage magnetic flux becomes the target interlinkage magnetic flux according to the target torque. By the second feedback control, which is different from the first feedback control, the second target voltage is calculated so that the applied current becomes the target current according to the target torque, and the first target voltage and the second target voltage are calculated. The applied current is controlled by controlling the inverter according to the sum of the above, and the gain used for the first feedback control and the second feedback control is the target current from the current operating point indicated by the applied current. It is obtained at the calculation timing determined by the route to the indicated target operating point.

According to one aspect of the present invention, the accuracy of motor rotation control can be improved.

FIG. 1 is a cross-sectional view of the variable magnetic flux type motor of the first embodiment. FIG. 2 is a configuration diagram showing a part of the motor. FIG. 3A is a block diagram of the motor and its control device. FIG. 3B is a block diagram of the torque control unit. FIG. 3C is a block diagram of the magnetic flux estimation unit. FIG. 4 is a diagram showing the gain calculation timing. FIG. 5A is a diagram showing the gain calculation timing in the first modification. FIG. 5B is a diagram showing an example of a change in inductance. FIG. 5C is a diagram showing an example of a change in inductance. FIG. 6A is a diagram showing an example of the gain calculation timing in the second modification. FIG. 6B is a diagram showing another example of the gain calculation timing. FIG. 7 is a diagram showing the relationship between the torque of the motor and the time in the third modification. FIG. 8A is a block diagram of the motor of the fourth modification and the control device thereof. FIG. 8B is a block diagram of the torque control unit. FIG. 9A is a block diagram of the torque control unit of the second embodiment. FIG. 9B is a block diagram of the magnetic flux estimation unit. FIG. 10 is a block diagram of the motor and the control device of the third embodiment. FIG. 11 is a diagram showing an operating point when the target current vibrates. FIG. 12 is a diagram showing an operating point when the target current vibrates. FIG. 13 is a diagram showing an operating point when the target current vibrates in the fifth modification. FIG. 14 is a diagram showing an operating point when the target current vibrates.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.

(First Embodiment)
Generally, a motor called a variable magnetic flux type or a variable leakage flux type is configured so that the amount of magnetic flux from the rotor to the stator can be changed due to the physical configuration of the rotor or the like. It is known that the amount of change in this magnetic flux changes according to the current applied to the motor, and the inductance can be changed by controlling the applied current, so that the optimum operating point according to the required load is realized. be able to.

FIG. 1 is a cross-sectional view of the variable magnetic flux type electric motor (rotary electric machine) 100 of the first embodiment perpendicular to the axial direction, and is a diagram showing a quarter of the entire configuration. For the remaining three-quarters of the overall configuration, the partial configuration shown in FIG. 1 is continuously repeated. The motor 100 of the present embodiment has an annular stator 1 and a rotor 2 that is concentric with the stator 1 and has an air gap 13 between the stator 1 and the stator 1. , A plurality of permanent magnets 3 fitted to the rotor 2 are provided to form an electric motor or a generator.

The stator 1 includes a ring-shaped stator core 11, a plurality of teeth 8 protruding from the stator core 11 toward the inner peripheral side, and a slot 9 which is a space between adjacent teeth 8. A stator winding 10 is wound around the teeth 8. The stator core 11 is formed of, for example, an electromagnetic steel plate which is a soft magnetic material.

The rotor 2 has a rotor core 12. The rotor core 12 is formed in a cylindrical shape by a so-called laminated steel plate structure, which is formed by laminating a large number of electromagnetic steel plates formed by punching a metal steel plate having high magnetic permeability in an annular shape in the axial direction. ing. Further, in the vicinity of the peripheral portion of the rotor core 12 facing the stator 1, a plurality of permanent magnets 3 are equidistant to each other along the circumferential direction, and the polarities of the permanent magnets 3 adjacent to each other are different. It is provided to be polar. As inferred from the partial configuration shown in FIG. 1, the rotor core 12 according to the variable magnetic flux type motor 100 of the present embodiment has an eight-pole structure in which eight permanent magnets 3 are provided along the circumferential direction. Has.

Further, the rotor 2 is a magnetic barrier (also referred to as a flux barrier) which is a space portion formed by punching an electromagnetic steel plate forming the rotor core 12 between the magnetic poles formed by the adjacent permanent magnets 3. ) It has 4 and 5. The magnetic barriers 4 and 5 have a higher magnetic resistance than the magnetic steel sheet. Therefore, the magnetic barriers 4 and 5 act as a magnetic flux barrier against the magnetic flux in the magnetic circuit in which the permanent magnet 3 is formed on the rotor 2. As shown in the figure, the magnetic barrier 4 is formed in a substantially triangular shape having a narrower width on the rotation center side than on the outer peripheral side of the rotor 2. The magnetic barrier 5 is formed in a substantially U shape on the rotation center side of the magnetic barrier 4 so as to cover the magnetic barrier 4. However, the shapes of the magnetic barriers 4 and 5 are not limited to the shapes shown in the figure as long as they have the technical effects described later. Further, since the magnetic barriers 4 and 5 are formed on the rotor 2 as shown in the figure, the magnetic flux generated from the permanent magnet 3 constituting one magnetic pole is formed by another permanent magnet 3 adjacent to the rotor 2. A magnetic flux bypass path 6 is formed as a path for leakage to the magnetic pole side. Further, the rotor 2 is provided with a narrow bridge-shaped portion 7 connected to the magnetic flux bypass path 6 on the outer peripheral side of the magnetic barrier 4.

The permanent magnet 3 is fixed to the rotor core 12 by being fitted into a gap formed in the corresponding portion of the rotor core 12. Further, in the permanent magnet 3, the radial direction of the rotor 2 is the magnetization direction. Here, in the present embodiment, the geometric center of the permanent magnet 3 is defined as the d-axis, and the position electrically orthogonal to the d-axis is defined as the q-axis. Since the variable magnetic flux type motor 100 of this embodiment has an 8-pole structure, the position of 22.5 degrees from the d-axis at the mechanical angle is defined as the q-axis. The magnetic barriers 4 and 5 described above are formed on the q-axis.

The above is the basic configuration of the variable magnetic flux type motor 100 of the first embodiment. Here, before the detailed description of the present embodiment, the basic principle of the variable magnetic flux type motor 100 which is the object of the present invention will be described.

The variable magnetic flux type motor 100, which is the object of the present invention, is a part of the magnetic flux (leakage magnetic flux) emitted from the permanent magnet 3 included in the rotor 2 due to the action of the load current (stator current) applied to the stator winding 10. ) Is characterized in that the magnetic path can be changed. Due to this feature, the motor 100 can passively change the stator interlinkage magnetic flux by controlling the stator current, so that the magnetic force of the permanent magnet 3 included in the rotor 2 is apparently variable by current control. Has the property of being able to. Due to such characteristics, as described above, the variable flux type motor 100 which is the object of the present invention is also called a variable leakage flux type.

FIG. 2 is a configuration diagram showing a part of the motor 100, and is a diagram for explaining an outline of operation in the variable magnetic flux type. The shape and arrangement of each configuration of the variable magnetic flux type motor 100 shown in FIG. 2 are general ones shown for explaining the outline of operation.

The motor 100 includes a stator 1 and a rotor 2 arranged so as to form an air gap 13 between the stator 1. The rotor 2 includes a permanent magnet 3, a magnetic barriers 4 and 5, and a magnetic flux bypass path 6 connected via the magnetic barriers 4 and 5 provided between two adjacent permanent magnets. Has. The direction indicated by the arrow on the left side of the figure indicates the rotation direction of the rotor 2.

The magnetic path direction 15 shown in FIG. 2 indicates the direction in which the leakage flux flows when no current is applied to the stator winding 10 included in the stator 1. As shown in FIG. 2, in the rotor, a part of the magnetic flux generated from the permanent magnet 3 leaks to the adjacent different electrode side through the magnetic flux bypass path 6.

Therefore, the amount of magnetic flux of the main magnetic flux component to the stator 1 side emitted from the permanent magnet 3, that is, the stator interlinkage magnetic flux is relatively reduced, so that the magnetic force of the permanent magnet 3 is apparently weakened. This makes it possible to reduce the iron loss generated by the stator interlinkage magnetic flux. This effect is particularly effective in improving the efficiency of the motor 100 in a low load and high speed region.

On the other hand, the magnetic path direction 20 indicates the direction in which the magnetic flux flows when a current is applied to the stator winding 10. As the stator current flows, the magnetic flux generated from the permanent magnet 3 is attracted to the stator side in the rotation direction of the rotor 2 as indicated by the magnetic path direction 20. Therefore, when the current is not applied to the stator winding 10, the magnetic flux leaked in the rotor 2 as the leakage flux can be efficiently converted into the stator interlinkage magnetic flux. As a result, the motor 100 can output a high torque equivalent to that in a state where the magnetic flux emitted from the permanent magnet 3 does not leak.

In this way, the magnetic flux bypass path 6 is provided between the magnetic poles configured in the rotor 2, and the stator interlinkage magnetic flux is passively changed by changing the leakage flux in the rotor by controlling the stator current. The motor having the above is called a variable magnetic flux type (variable leakage flux type).

The above is the basic principle of the variable magnetic flux type motor 100 which is the object of the present invention. Hereinafter, the details of the control device for controlling the variable magnetic flux type motor 100 will be described.

FIG. 3A is a block diagram of the motor 100 and the control device 200.

The control device 200 uses an applied current I (d-axis applied current I _d ) for rotation control based on the target torque T ^* calculated by the target torque calculation unit 250 based on the operation of the accelerator, brake, or the like by the driver. , And the q-axis applied current I _q ) is calculated, and the applied current I is output to the motor 100. The control device 200 is composed of a microcomputer including a central processing unit (CPU), a read-only memory (ROM), a random access memory (RAM), and an input / output interface (I / O interface). The control device 200 stores a program, and by executing the stored program, the control described later is executed.

The torque control unit 210 has a target torque T ^* , an interlinkage magnetic flux λ estimated by the magnetic flux estimation unit 230 (d-axis interlinkage magnetic flux λ _d , and q-axis interlinkage magnetic flux λ _q ), and the motor 100. The applied current I of is input. The torque control unit 210 calculates the target voltage V ^* (d-axis target voltage V _d ^* and q-axis target voltage V _q ^* ) from these input values. The torque control unit 210 is required to calculate the gain timing TM, and the gain is output to the magnetic flux estimation unit 230. The details of the processing in the torque control unit 210 will be described later with reference to FIG. 3B.

The inverter 220 applies the applied current I to the motor 100 by switching control according to the input target voltage V ^* . It is assumed that the motor 100 is driven by three phases, and the three-phase alternating currents I _u , I _v , and I _w are output from the inverter 220.

The magnetic flux estimation unit 230 inputs the applied current I, the target voltage V ^* , the inductance L (d-axis inductance L _d , and the q-axis inductance L _q ), and the magnet magnetic flux ψ _a at the timing when the calculation timing TM is received. Therefore, the interlinkage magnetic flux λ is estimated and output to the torque control unit 210. The details of the processing in the magnetic flux estimation unit 230 will be calculated later using FIG. 3C.

In the inductance estimation unit 240, the applied current I to the motor 100 is input, the inductance L in the motor 100 and the magnetic flux ψ _a are calculated based on the applied current I, and the calculation results are the torque control unit 210 and the magnetic flux. It is output to the estimation unit 230.

An ammeter 101 is attached to the motor 100, and the current applied to the motor 100 is measured by the ammeter 101. Although the applied current measured by the current meter 101 is a three-phase current, conversion from the three-phase to the dq axis is performed, and the d-axis current I _d and the q-axis current I _q are used as the magnetic flux estimation unit 230 and the magnetic flux estimation unit 230. , Is output to the inductance estimation unit 240.

Here, in the variable magnetic flux type motor 100, the inductance and the magnet magnetic flux ψ _a change greatly according to the applied current. Therefore, the inductance estimation unit 240 calculates the inductance L and the magnet magnetic flux ψ _a based on the applied current I by using a map or the like stored in advance. Then, the inductance estimation unit 240 outputs these calculated values to the torque control unit 210 and the magnetic flux estimation unit 230.

FIG. 3B is a block diagram showing details of the torque control unit 210.

The input target torque T ^* is input to the target interlinkage magnetic flux map 211 and the target current map 212, and the calculation results in each map are finally added to obtain the target voltage V ^* (d-axis target voltage V _d). ^* And the q-axis target voltage V _q ^* ) are obtained.

In the target interlinkage magnetic flux map 211, the target interlinkage magnetic flux λ ^* (d-axis target interlinkage magnetic flux λ _d ^* ) is such that the motor 100 has an inductance that realizes the target torque T ^* input from the target torque calculation unit 250. , And the q-axis target interlinkage magnetic flux λ _q ^* ).

In the subtractor 213, the interlinkage magnetic flux λ (d-axis interlinkage magnetic flux λ _d and q-axis) estimated by the magnetic flux estimation unit 230 from the target interlinkage magnetic flux λ ^* obtained by the target interlinkage magnetic flux map 211. The interlinkage magnetic flux λ _q ) is subtracted. Then, the subtraction result is output as a target magnetic flux applied voltage Vλ ^* (d-axis target magnetic flux applied voltage V _d λ ^* and q-axis target magnetic flux applied voltage V _q λ ^* ) via the λ / V conversion unit 214. ..

In the target current map 212, the target current I ^* (d-axis target current I _d ^* and q-axis target current I _q ^* ) required to realize the target torque T ^* in the motor 100 is obtained. The target current I ^* calculated by the target current map 212 is output to the subtractor 215 in the previous stage of the current PI control unit 216 and the calculation timing generation unit 219.

In the subtractor 215, the applied current I (d-axis current I _d and q-axis current I _q ) detected by the ammeter 101 is subtracted from the target current I ^* , and the subtraction result is the current PI control unit. It is input to 216.

The subtraction value in the subtractor 215 and the gain G (d-axis gain G _d and q-axis gain G _q ) obtained by the gain map 218 are input to the current PI control unit 216. The gain G includes a proportional control gain and an integrated control gain, and the current PI control unit 216 controls by performing proportional control gain using the gain G and feedback control using the integrated control gain. The voltage based on the result is output as the target current applied voltage V _i ^* (d-axis target current applied voltage V _di ^* and q-axis target current applied voltage V _qi ^* ).

The adder 217 has a target magnetic flux applied voltage Vλ ^* (d-axis target magnetic flux applied voltage V _d λ ^* and a q-axis target magnetic flux applied voltage V _q λ ^* ) and a target current applied voltage V _i ^* (d-axis target current). The target voltage V ^* (d-axis target voltage V _d ^* and q-axis target voltage V _q ^* ) is calculated by adding the applied voltage V _di ^* and the q-axis target current applied voltage V _qi ^* ), respectively. And output.

The gain map 218 is a calculation timing TM generated by the calculation timing generation unit 219 according to the current I applied to the motor 100 using a map stored in advance, and is used for feedback control in the current PI control unit 216. Used to calculate the gain G. Since the gain G is calculated for each calculation timing TM, the gain G calculated by a certain calculation timing TM will be used until the next calculation timing.

In the variable magnetic flux type motor 100, the inductance changes greatly according to the applied current. Further, in general, in the motor 100, the responsiveness in the feedback control of the applied current changes according to the inductance. That is, in the variable magnetic flux type motor 100, the responsiveness of the feedback control greatly differs depending on the applied current.

The gain (proportional control gain and integral control gain) used for feedback control in the current PI control unit 216 has a shorter time until the current reaches a steady state, but may oscillate and diverge. .. On the other hand, the smaller the gain, the smaller the risk of divergence, but the longer it takes to reach a steady state. Therefore, it is assumed that each gain calculation unit stores in advance as a gain map the gain that causes the steady state at the shortest without divergence according to the current I applied to the motor 100.

In the calculation timing generation unit 219, the current operating point determined by the input d-axis current I _d and the q-axis current I _q , the d-axis target current I _d ^* , and the q-axis target current I _q ^* are used. The gain calculation timing is calculated from the path between the target operating point and the determined target operating point. Then, the calculation timing generation unit 219 outputs the calculation timing TM indicating this timing to the gain map 218 and the magnetic flux estimation unit 230. The operation state such as the control cycle, the load condition, and the rotation speed of the motor 100 is input to the calculation timing generation unit 219, and the calculation timing may be calculated using these operation states. The details of the gain calculation timing will be described later.

The target interlinkage magnetic flux map 211, the subtractor 213, and the λ / V conversion unit 214, together with the magnetic flux estimation unit 230, constitute a first calculation unit for setting a first target voltage for controlling the interlinkage magnetic flux. do. The target current map 212, the subtractor 215, and the current PI control unit 216 constitute a second calculation unit that sets a second target voltage for controlling the current.

Here, in general motor control, a fixed value is often used for the gain. On the other hand, in the variable magnetic flux type motor 100, the inductance changes greatly according to the applied current, so if a fixed value is used as the gain used for feedback control, the followability deteriorates, and overshoot and stabilization occur. It may take some time. Therefore, by sequentially calculating the gain at the calculation timing instructed by the calculation timing generation unit 219, it is possible to shorten the time until stabilization without divergence by the feedback control.

FIG. 3C is a diagram showing details of the magnetic flux estimation unit 230.

The q-axis current I _q input from the ammeter 101 is input to the subtractor 231q and the q-axis gain map 232q.

The subtractor 231q subtracts the q-axis predicted current I _qp output from the divider 247d from the q-axis current I _q , and outputs the subtraction result to the q-axis magnetic flux PI control unit 233q.

In the q-axis gain map 232q, the q-axis gain G _q used for feedback control in the q-axis magnetic flux PI control unit 233q is calculated using the q-axis current I _q at the timing of receiving the calculation timing TM.

In the q-axis magnetic flux PI control unit 233q, the subtraction value in the subtractor 231q provided in the previous stage and the q-axis gain G _q obtained by the q-axis gain map 232q are input, and PI control using these input values (PI control ( Feedback control) is performed. Then, the control result is converted into a voltage and output as a q-axis target voltage V _q ^* .

In the adder 234q, the q-axis target voltage Vq ^* and the _q -axis voltage Vq output from the torque control unit 210 are added, and the addition result is output to the integrator 235q. In the integrator 235q, the d-axis interlinkage magnetic flux λ _d is obtained by integrating the addition result obtained by the adder 234q, and is output to the torque control unit 210 and the subtractor 236q. In the subtractor 236q, the magnet magnetic flux ψ _a output from the magnetic flux estimation unit 230 is subtracted from the d-axis interlinkage magnetic flux λ _d . In the divider 237q, the d-axis predicted current I _dp is obtained by dividing the subtraction result in the subtractor 236q by the _d -axis inductance Ld output from the magnetic flux estimation unit 230.

Here, the processing from the integrator 235q to the divider 237q will be described. Since it is known that the interlinkage magnetic flux can be obtained by integrating the voltage values, the d-axis interlinkage magnetic flux λ _d can be obtained by integrating the _q -axis voltage Vq output from the adder 234q with the integrator 235q. Be done. Further, since the relationship shown by the following equation (1) modified from the equation (2) is known, the result of subtracting the magnet magnetic flux ψ _a from the d-axis interlinkage magnetic flux λ _d in the subtractor 236q is obtained. The d-axis predicted current I _dp can be obtained by dividing by the d-axis inductance L _d in the divider 237q. It should be noted that I _d in the following equation corresponds to the d-axis predicted current I _dp obtained by the above configuration.

For the d-axis, the subtractor 231d, the d-axis gain map 232d, the d-axis magnetic flux PI control unit 233d, the adder 234d, the integrator 235d, and the divider 237d similarly use the q-axis magnetic flux λ _q and the d-axis interlinkage magnetic flux λ q. , The q-axis predicted current I _qp can be calculated.

In the calculation of the d-axis predicted current I _dp , the configuration corresponding to the subtractor 236q for subtracting the magnet magnetic flux ψ _a is not provided. This is because the magnet magnetic flux ψ _a does not contribute to the q-axis interlinkage magnetic flux as shown in the following equation (4), which is a modification of the q-axis interlinkage magnetic flux λ _q .

FIG. 4 is a diagram showing the calculation timing by the calculation timing generation unit 219.

In this figure, the d-axis current I _d is shown on the horizontal axis, the q-axis current I _q is shown on the vertical axis, and the operating point of the motor 100 is shown. From P0, which is the initial operating point where both the d-axis current I _d and the q-axis current I _q are zero, to P ^* , which is the target operating point indicated by the d-axis target current I _d ^* and the q-axis target current I _q ^* . And the path when changing the applied current is shown. Then, in this example, the gain is calculated at a plurality of calculation timings from P1 to P7 in the path from the initial operating point P0 to the target operating point P ^* . It should be noted that the calculation timings from P1 to P7 do not have to be determined at one time, and the sequential calculation timing may be obtained so that, for example, P2 is determined in P1 and then P3 is determined in P2.

Specifically, for example, at the initial operating point P0, a gain is calculated from the current value, and feedback control is performed from the initial operating point P0 to the operating point P1 which is the next calculation timing using the calculated gain. It will be. At the operating point P1, the gain is calculated from the current value at the operating point P1, and feedback control is performed from the operating point P1 to the operating point P2, which is the timing for calculating the next gain. By calculating the sequential gain in this way, the control accuracy of the feedback control can be improved.

In the present embodiment, the gain is calculated at seven operating points on the path, but the gain is not limited to this. The sequential gain may be calculated at a plurality of operating points. Further, the calculation timing generation unit 219 can appropriately determine the gain calculation timing, the gain calculation timing, and the calculation cycle according to the operation state such as the control cycle, the load condition, and the rotation speed of the motor 100.

According to the first embodiment, the following effects can be obtained.

The motor 100 controlled by the control method of the first embodiment is a variable magnetic flux type, has a rotor 2 provided with magnetic barriers 4 and 5, and has a rotor 2 according to an applied current. It is possible to change the magnetic flux toward the stator 1 from the permanent magnet 3 possessed by the magnet. In order to obtain the applied current I applied to the motor 100 via the inverter 220 in order to control the motor 100, the torque control unit 210 obtains the target voltage V ^* .

In the magnetic flux estimation unit 230, in the d-axis magnetic flux PI control unit 233d, the d-axis current Id is controlled by the first feedback control in which the _d -axis predicted current I _dp is feedback-input, and the control result is transmitted to the integrator 235d. The q-axis interlinkage magnetic flux λ _q is estimated by integrating. Similarly, in the q-axis magnetic flux PI control unit 233q, the q-axis current I _q is controlled by the first feedback control in which the q-axis predicted current I _qp is input back, and the control result is integrated by the integrator 235q. Then, the d-axis interlinkage magnetic flux λ _d is estimated.

Then, in the torque control unit 210, the d-axis interlinkage magnetic flux λ _d and the q-axis interlinkage magnetic flux λ _q estimated by the subtractor 213 are the target interlinkage magnetic flux λ ^* defined in the target interlinkage magnetic flux map 211. The target magnetic flux applied voltage Vλ ^* (first target voltage) is obtained so as to be (d-axis target magnetic flux λ _d ^* and q-axis target magnetic flux λ _q ^* ).

On the other hand, in the current PI control unit 216, the applied current I is determined in the target current map 212 by the second feedback control in which the applied current I (d-axis current I _d and the q-axis current I _q ) is input back. The target current applied voltage V _i ^* (second target voltage) such that the target current I ^* (d-axis target current I _d ^* and q-axis target current I _q ^* ) is calculated.

Then, in the adder 217, the target voltage applied voltage Vλ ^* (first target voltage) and the target current applied voltage V _i ^* (second target voltage) are added up to obtain the target voltage V ^* (d-axis target voltage V _d). ^* And, it is output to the inverter 220 as the q-axis target voltage V _q ^* ).

The calculation timing generation unit 219 sets a plurality of operating points P1 to P7 on the path as calculation timings in the path from the initial operating point P0 to the target operating point P ^* indicated by the target current. Then, the current PI control unit 216 of the torque control unit 210, the q-axis magnetic flux PI control unit 233q of the magnetic flux estimation unit 230, the d-axis magnetic flux PI control unit 233d, and the like are sequentially generated by the calculation timing generation unit 219. The gain is calculated at the calculation timing. Feedback control is performed using the gain calculated sequentially in this way, and finally, the currents I _d and I _q (I _u , I _v , I _w ) applied to the motor 100 are obtained.

The variable magnetic flux type motor 100 has a feature that the inductance changes greatly according to the applied current. Further, in general, in the motor 100, the responsiveness in the feedback control of the applied current differs depending on the inductance. That is, in the variable magnetic flux type motor 100, the responsiveness of the feedback control greatly differs depending on the input current, so it is necessary to calculate the gain G used for the feedback control according to the applied current.

Therefore, when the motor 100 controls the change from the initial operating point P0 to the target operating point P ^* , a plurality of calculation timings are set in the change path, and the gain is calculated sequentially at these calculation timings. Then, using the sequentially calculated gain, the first feedback control of the current in the current PI control unit 216, the estimation of the magnetic flux by the q-axis magnetic flux PI control unit 233q, and the p-axis magnetic flux PI control unit 233p, and the like are performed. It is performed by feedback control. By doing so, the timely gain is calculated even if the inductance changes significantly, and the feedback control is performed appropriately. Therefore, the final calculated target voltage V ^* (d-axis target voltage V _d ^* , In addition, at the q-axis target voltage V _q ^* ), divergence can be suppressed and the stabilization time can be shortened.

According to the control method of the motor 100 of the first embodiment, the calculation timing generation unit 219 determines the gain calculation timing according to the operation state such as the control cycle, the load condition, and the rotation speed of the motor 100. By doing so, for example, when the amount of change in the inductance of the motor 100 is large, the optimum feedback control can be realized by shortening the gain calculation interval. On the other hand, when the amount of change in the inductance of the motor 100 is small, it is possible to suppress an unnecessary change in gain by lengthening the calculation interval.

According to the control method of the motor 100 of the first embodiment, the gain map 218, the q-axis gain map 232q, and the d-axis gain map 232d have current values indicating the current calculated timing operating points at the gain calculation timing. Calculate the gain based on this. By doing so, the gain can be sequentially calculated with relatively simple control, so that the entire processing can be simplified.

(First modification)
In the above-mentioned example of the first embodiment, an example in which the gain is corrected at a predetermined calculation cycle based on the current of the current operating point at the gain calculation timing has been described. In this embodiment, a modified example of the method for setting the gain calculation timing will be described.

FIG. 5A is a diagram showing a first modification of the gain calculation timing. In this figure, only the path of change of the operating point is shown in one dimension.

According to this example, the calculation timing generation unit 219 sets the operating points from P1 to P6 as the gain calculation timing in the path from the initial operating point P0 to the target operating point P ^* . Then, at the calculation timing of each gain, an intermediate operating point between the operating point of the current calculation timing and the operating point at the next calculation timing is obtained, and the gain is calculated using the current indicating the intermediate operating point.

Specifically, for example, at the initial operating point P0, the calculation timing generation unit 219 sets an intermediate operating point between the initial operating point P0 of the current calculation timing and the operating point P1 at the next calculation timing on the path. It is generated between the initial operating point P0 and the operating point P1. Then, the intermediate operating point is transmitted to the gain map 218, the q-axis gain map 232q, and the d-axis gain map 232d by using the calculation timing TM. In this way, the gain at the intermediate operating point is calculated in the gain map 218, the q-axis gain map 232q, and the d-axis gain map 232d, and the gain is used from the initial operating point P0 to the operating point P1. Feedback control can be performed.

Here, at the timing of calculating the gain, a case where the inductance at the next calculation timing is smaller than the inductance of the current operating point will be examined.

FIG. 5B is a diagram showing the relationship between the current and the inductance when the inductance becomes small. This figure shows the operating point at the current calculation timing and the operating point at the next calculation timing. Then, it is shown that the inductance L becomes smaller from the current operating point to the operating point at the next calculation timing.

Considering the section from the current operating point to the operating point of the next calculation timing, at the current operating point, which is the start timing of this section, the inductance is relatively large, but it is based on the intermediate operating point. Since the relatively small gain calculated in the above is used for the feedback control, the responsiveness is low. On the other hand, at the operating point of the next calculation timing, which is the end timing of this section, a relatively large gain calculated based on the intermediate operating point is used even though the inductance is relatively small. , High responsiveness. As described above, in the section where the initial operating point P0 changes to the operating point P1, the stability is expected to be improved by gradually increasing the responsiveness.

Next, at the timing of calculating the gain, a case where the inductance at the next calculation timing is larger than the inductance of the current operating point will be examined.

FIG. 5C is a diagram showing the relationship between the current and the inductance when the inductance becomes large. This figure shows that the inductance L increases from the current operating point to the operating point at the next calculation timing.

In the section from the current operating point to the operating point of the next calculation timing, the comparison calculated based on the intermediate operating point is performed at the current operating point, which is the start timing, even though the inductance is relatively small. Since a large gain is used, the responsiveness is high. On the other hand, at the operating point of the next calculation timing, which is the end timing of this section, a relatively small gain calculated based on the intermediate operating point is used even though the inductance is relatively large. , The responsiveness becomes low. As described above, in the section where the operating point changes from P0 to P1, the responsiveness gradually decreases, so that the responsiveness is high at the beginning of the section and the divergence is likely to occur, but the divergence is expected to be suppressed at the end.

According to the first modification, the following effects can be obtained.

According to the control method of the motor 100 of the first modification, the calculation timing generation unit 219 sets an intermediate operating point between the current operating point and the operating point at the next calculation timing as a path between the two in the gain calculation timing. Generated in. Then, the gain map 218, the q-axis gain map 232q, and the d-axis gain map 232d calculate the gain based on the current indicating the intermediate operating point thereof. By doing so, even if the inductance of the motor 100 changes significantly until the next gain calculation timing, feedback is provided by using the gain calculated using the current at the intermediate operating point. It is possible to suppress divergence and prolonged convergence time due to control.

According to the control method of the motor 100 of the first modification, as shown in FIG. 5B, the inductance of the motor 100 decreases from the current operating point toward the operating point at the next calculation timing. In such a case, the stability can be improved by gradually increasing the responsiveness in the section from the current operating point to the operating point of the next calculation timing.

According to the control method of the motor 100 of the first modification, as shown in FIG. 5C, the inductance of the motor 100 increases from the current operating point to the operating point at the next calculation timing. In such a case, divergence can be suppressed by gradually lowering the responsiveness in the section from the current operating point to the operating point at the next calculation timing.

(Second modification)
In the examples so far in the first embodiment, an example in which the gain calculation cycle is constant has been described. In the second modification, an example of changing the calculation cycle will be described.

6A and 6B are diagrams showing changes in operating points in the second modification.

In the case shown in FIG. 6A, the difference in the q-axis current I _q between the initial operating point P0 and the target operating point P ^* is larger than in the case shown in FIG. 6B. Here, since the q-axis current I _q affects the torque of the motor 100, the case shown in FIG. 6A is from the initial operating point P0 to the target operating point P ^* as compared with the case shown in FIG. 6B. The change in torque of the motor 100 between them is small.

Therefore, when the change in torque is relatively large as shown in FIG. 6A, a large change in gain is expected. Therefore, the gain calculation interval is shortened between the initial operating point P0 and the target operating point P ^*. Set. On the other hand, when the change in torque is relatively small as shown in FIG. 6B, the calculation interval is set long.

As a result, when the torque fluctuation is large and it is necessary to increase the gain calculation frequency, the gain calculation frequency is increased by shortening the gain calculation interval, and the followability in feedback control can be improved. .. On the other hand, when the fluctuation of the torque is small and it is not necessary to calculate a large amount of gain, the gain calculation frequency becomes low by lengthening the gain calculation interval, and unnecessary processing can be omitted.

According to the second modification, the following effects can be obtained.

According to the control method of the motor 100 of the second modification, when the change in the q-axis current is large in the path from the initial operating point P0 to the target operating point P ^* , the calculation timing generation unit 219 performs the gain calculation timing. Increase the interval between. By doing so, the gain can be calculated frequently when the torque fluctuation is large, so that the followability can be improved. On the other hand, when the change in the d-axis current is small and the change in the torque is small, it is not necessary to change the gain frequently. Therefore, by lengthening the gain calculation interval, unnecessary processing can be omitted.

(Third modification example)
In the second modification, an example of changing the gain calculation cycle has been described. In the third modification, an example in which a plurality of gain calculation cycles are set variably will be described.

FIG. 7 is a diagram showing the relationship between the torque of the motor 100 and the elapsed time when controlled by the control device 200. The vertical axis shows the torque, and the torque at the initial operating point P0 and the target operating point P ^* is shown as T0 and T ^* . The times at which the torques T0 and T ^* are obtained are shown as t0 and te.

According to this figure, when the value of the torque is small, the rate of change is large, and when the value is large, the rate of change is small. Therefore, with respect to the torque, T1 to T8 are set so that the intervals from the torques T0 to T ^* are evenly spaced. Then, the time corresponding to the torques T1 to T8 is set as t1 to t8.

For example, in the section where the torque changes from T0 to T1, the rate of change of the torque is relatively large, so that the gain calculation interval from the time t0 to t1 is relatively short. On the other hand, in the section from the torque change from T8 to the target torque Tf, the torque change rate is relatively small, so that the gain calculation interval from the time t8 to the target time tf is relatively long.

In this way, the calculation timing generation unit 219 uses the relationship between torque and time as shown in FIG. 7 to set t1 to t8, which are gain calculation timings, so that the amount of change in torque is evenly spaced. do. By doing so, when the torque change rate is large, the gain calculation frequency increases, so that the followability can be improved, and when the torque change rate is small, the gain calculation frequency decreases. Unnecessary processing can be omitted.

According to the third modification, the following effects can be obtained.

According to the control method of the motor 100 of the third modification, the calculation timing generation unit 219 determines the gain when the torque change rate is relatively large in the path from the initial operating point P0 to the target operating point P ^* . Shorten the calculation interval. By doing so, the gain is updated frequently, so that the followability can be improved. On the other hand, when the rate of change in torque is relatively small, it is not necessary to change the gain frequently. Therefore, by lengthening the calculation interval, unnecessary processing can be omitted.

(Fourth modification)
In the previous examples of the first embodiment, the case where the target torque T ^* is input to the target interlinkage magnetic flux map 211 of the torque control unit 210 has been described, but the present invention is not limited to this. As a fourth modification, the target interlinkage magnetic flux calculation unit 211A may be provided in place of the target interlinkage magnetic flux map 211, and further input may be made to the target interlinkage magnetic flux calculation unit 211A.

FIG. 8A is a block diagram of the control device 200 in this modification. According to this figure, as compared with the configuration shown in FIG. 3A, the point where the magnet magnetic flux ψ _a , the d-axis inductance L _d , and the q-axis inductance L _q are input from the inductance estimation unit 240 to the torque control unit 210. Is different.

FIG. 8B is a block diagram of the torque control unit 210. In this modification, as compared with the configuration shown in FIG. 3B, the target interlinkage magnetic flux calculation unit 211A is provided in place of the target interlinkage magnetic flux map 211, and the target interlinkage magnetic flux calculation unit 211A has a target torque T. In addition to ^* , the magnetic flux ψ _a , the d-axis inductance L _d , and the q-axis inductance L _q are input from the inductance estimation unit 240. Here, it is known that the following equation holds for the target torque T ^* .

Substituting the equations (2) and (4) into the equation (5) gives the following equation.

By using this equation, the target interlinkage magnetic flux calculation unit 211A can calculate the d-axis target interlinkage magnetic flux λ _d ^* and the q-axis target interlinkage magnetic flux λ _q ^* , so that a map or the like is provided. It is not necessary and the configuration can be simplified.

(Second Embodiment)
In the previous examples of the first embodiment, an example in which a gain calculation unit is provided in front of the current PI control unit 216, the q-axis magnetic flux PI control unit 233q, and the d-axis magnetic flux PI control unit 233d that perform feedback control. However, it is not limited to this. The gain may be calculated in the control unit that performs feedback control. The input / output between the configurations in the control device 200 of the present embodiment shall be the same as the input / output shown in FIG. 8A.

FIG. 9A shows a block diagram of the torque control unit 210 of the present embodiment. In this example, as compared with the block diagram shown in FIG. 3B, the gain map 218 is deleted, the calculation timing TM is input from the calculation timing generation unit 219 to the current PI control unit 216, and the magnetic flux is estimated. The difference is that the inductance L (d-axis inductance L _d and q-axis inductance L _q ) is input from the unit 230. The current PI control unit 216 calculates a gain based on the input inductance L according to the calculation timing TM, and performs PI control (feedback control) based on the gain.

As described above, in the motor 100, since the responsiveness in the feedback control of the applied current changes according to the inductance, the current PI control unit 216 stores a map in which the inductance L and the gain correspond to each other. Gains can be generated using this map.

FIG. 9B shows a block diagram of the magnetic flux estimation unit 230. In this example, as compared with the block diagram shown in FIG. 3C, the q-axis gain map 232q and the d-axis gain map 232d are deleted, and the q-axis magnetic flux PI control unit 233q of the magnetic flux estimation unit 230 and the q-axis magnetic flux PI control unit 233q and The calculation timing TM is input from the torque control unit 210 (calculation timing generation unit 219) to the d-axis magnetic flux PI control unit 233d, and the inductance L (d-axis inductance L _d and q-axis inductance L _q ) is input from the magnetic flux estimation unit 230. ) Has been entered. The q-axis magnetic flux PI control unit 233q and the d-axis magnetic flux PI control unit 233d calculate a gain based on the input inductance L according to the calculation timing TM, and PI control (feedback control) based on the gain. Is done.

According to the second embodiment, the following effects can be obtained.

According to the control method of the motor 100 of the second embodiment, the gain is obtained in the current PI control unit 216, the q-axis magnetic flux PI control unit 233q, and the d-axis magnetic flux PI control unit 233d without separately providing a block for generating the gain. Is also calculated, so that the configuration can be simplified. Further, in addition to the embodiment in which the gain is calculated based on the d-axis current I _d and the q-axis current I _q as in the first embodiment, the gain is set in a different embodiment using inductance to stabilize the feedback control. Therefore, the degree of freedom in design can be improved.

According to the second embodiment, the following effects can be obtained.

According to the control method of the second embodiment, the current PI control unit 216, the q-axis magnetic flux PI control unit 233q, and the d-axis magnetic flux PI control unit 233d gain gains based on the inductance calculated by the inductance estimation unit 240. Is calculated. Here, since the responsiveness of the motor 100 changes according to the inductance, the gain can be obtained also based on the inductance. Therefore, in addition to the embodiment in which the gain is calculated based on the d-axis current I _d and the q-axis current I _q as in the first embodiment, the gain is set in a different embodiment using inductance to stabilize the feedback control. Therefore, the degree of freedom in design can be improved.

(Third Embodiment)
In the first and second embodiments, an example in which the inductance estimation unit 240 measures the inductance by using the feedback input from the motor 100 has been described. In the second embodiment, an example in which the operating point is slightly changed in order to measure the inductance will be described.

FIG. 10 is a block diagram of the torque control unit 210 of the present embodiment. In the present embodiment, as compared with the configuration shown by FIG. 3A in the first embodiment, in the torque control unit 210, the calculation timing TM is input from the calculation timing generation unit 219 to the target current map 212. different. When the target current map 212 receives the calculation timing TM from the calculation timing generation unit 219, the target current map 212 vibrates the d-axis target current I _d ^* with an amplitude of ΔI _d for a predetermined time. By doing so, the current applied to the motor 100 also oscillates.

FIG. 11 shows an operating point when the d-axis target current I _d ^* vibrates.

According to this figure, it is assumed that the lower limit of the d-axis current is _id1 and the upper limit is _id2 for the operating point that vibrates at a certain gain calculation timing. Here, when the d-axis current is _id1 , the q-axis voltage is V _q1 and the q-axis current is i _q1 . When the q-axis voltage is V _q2 and the q-axis current is i _q2 when the _d -axis current is id2, the following equation is established.

However, it is assumed that the resistance of the motor 100 is indicated by Ra and the differential operator is indicated by ρ. Further, the magnet magnetic flux in the motor 100 is ψ _a , and the interlinkage magnetic flux is λ.

By using these equations, the inductance estimation unit 240 can estimate the _d -axis inductance Ld and the q-axis inductance Lq without using a map as in the first embodiment.

According to the third embodiment, the following effects can be obtained.

According to the control method of the motor 100 of the third embodiment, the current value indicating the operating point at the calculation timing is oscillated with a predetermined amplitude. In such a case, for example, for two operating points, that is, the operating point corresponding to the upper limit current of vibration and the operating point corresponding to the lower limit current, the current value corresponding to each operating point and the voltage. Can be used to determine the inductance of the motor 100. Therefore, unlike the first embodiment, it is not necessary to have the gain map 218, the q-axis gain map 232q, and the d-axis gain map 232d, so that the capacity of the storage memory can be reduced.

According to the control method of the motor 100 of the third embodiment, the _d -axis target current Id ^* of the target current is vibrated. By doing so, the q-axis target current I _q ^* , which affects the torque of the motor 100, does not change, so that the impedance can be estimated without affecting the rotation control of the motor 100.

In the present embodiment, the _d -axis current Id, which has little effect on the torque on the motor 100, is oscillated, but the present invention is not limited to this. The q-axis current I _q may be oscillated, or the d-axis current indicating the operating point of the calculation timing in order to change both the d-axis current I _d and the q-axis current I _q as shown in FIG. Further, the current may be changed into a circular locus centered on the q-axis current and having r as a radius.

In such a case, the d-axis inductance L _d and the q-axis inductance Lq can be estimated by using the following equations and the equations (10) and (11) instead of the equation (7). In this example, since the q-axis current also changes, the d-axis voltages V _d1 and V _d2 are further used.

(Fifth modification)
In the example of the first embodiment, an example in which the current indicating the operating point at the calculation timing is vibrated with a predetermined amplitude has been described. In the sixth modification, an example in which this amplitude is changed for each operating point will be described.

FIG. 13 shows a plurality of operating points that the d-axis target current I _d ^* vibrates.

According to this figure, the amplitude is set to increase in the order of Δ _d1 , Δ _d2 , and Δ _d3 as the operating points P1, P2, and P3, which are the gain calculation timings, change. The amplitude shall be determined at a constant rate according to the magnitude of the d-axis current at the operating point. That is, when the absolute value of the d-axis target current I _d ^* is relatively small as in the operating point P1, the vibration is performed with a relatively small amplitude Δ _d1 . On the other hand, when the absolute value of the d-axis target current I _d ^* is relatively large as in the operating point P3, the vibration is performed with a relatively large amplitude Δ _d3 . By doing so, it is possible to vibrate at each operating point with an amplitude of a magnitude corresponding to the current, so that vibration of an unnecessary magnitude can be suppressed.

According to the fifth modification, the following effects can be obtained.

According to the control method of the motor 100 of the fifth modification, when the d-axis current increases as the operating points change to P1, P2, and P3, the vibration amplitudes at the operating points are Δ _d1 and Δ _d2 . , Δ _d3 and so on. By doing so, the amplitude can be set according to each operating point, so that vibration with an unnecessarily large amplitude is suppressed, and the vibration for measuring the inductance is the rotation control of the motor 100. The effect on the

In this modification, the _d -axis current Id, which has little effect on the torque of the motor 100, is vibrated, but the present invention is not limited to this. The q-axis current I _q may be oscillated, or the d-axis current indicating the operating point of the calculation timing in order to change both the d-axis current I _d and the q-axis current I _q as shown in FIG. Further, the current may be changed into a circular locus centered on the q-axis current and having r as a radius. In such a case, the radius r changes according to the magnitude of the d-axis current.

Although the embodiments of the present invention have been described above, the above embodiments are only a part of the application examples of the present invention, and the technical scope of the present invention is limited to the specific configuration of the above embodiments. not.

100 Motor 200 Control device 210 Torque control unit 211 Target interlinkage magnetic flux map 211A Target interlinkage magnetic flux calculation unit 212 Target current map 216, 233d, 233q Current PI control unit 218, 232d, 232q Gain map 219 Calculation timing generation unit 220 Inverter 230 Magnetic flux estimation unit 250 Target torque calculation unit

Claims (11)

A variable magnetic flux type motor control method that can change the magnetic flux from the permanent magnet to the stator according to the current applied from the inverter to the motor by means of a magnetic barrier provided on the rotor equipped with a permanent magnet. There,
The interlinkage magnetic flux of the motor is estimated by the first feedback control of the applied current, and the first target voltage is calculated so that the estimated interlinkage magnetic flux becomes the target interlinkage magnetic flux according to the target torque.
By the second feedback control different from the first feedback control, the second target voltage is calculated so that the applied current becomes the target current corresponding to the target torque.
By controlling the inverter according to the sum of the first target voltage and the second target voltage, the applied current is controlled.
The gain used for the first feedback control and the second feedback control is calculated at a calculation timing defined in the path from the current operating point indicated by the applied current to the target operating point indicated by the target current. , Motor control method required for each. In the motor control method according to claim 1,
A motor control method in which the inductance of the motor is obtained by using the current of the operating point in the path, and the gain is calculated based on the inductance of the motor. In the motor control method according to claim 2,
A motor control method in which the applied current is vibrated by vibrating the target current, and the inductance of the motor is calculated using a plurality of values of the vibrated applied current. In the motor control method according to claim 3,
A method for controlling a motor, wherein the target current to be vibrated is a d-axis current. In the motor control method according to claim 3 or 4,
A motor control method in which the amplitude of the vibration is determined according to the operating point. The motor control method according to any one of claims 1 to 5.
The calculation timing is a plurality of motor control methods defined according to the operating state of the motor. In the motor control method according to claim 6,
A motor control method in which the calculation timing interval is set so that the larger the torque change amount in the path, the longer the interval. In the motor control method according to claim 6,
A motor control method in which the calculation timing interval is set so that the amount of torque change is equal at the interval. The motor control method according to any one of claims 1 to 8.
The calculation timing is a plurality of motor control methods determined based on the current at the current operating point. The motor control method according to any one of claims 1 to 9.
A motor control method that generates an intermediate operating point between the current operating point and the operating point at the next calculated timing at the calculated timing, and calculates the gain based on the current at the intermediate operating point. .. A variable magnetic flux type motor control device capable of changing the magnetic flux from the permanent magnet of the rotor toward the stator according to the current applied to the motor by a magnetic barrier provided on the rotor. ,
The first feedback control of the applied current estimates the interlinkage magnetic flux of the motor, and the first target voltage is calculated so that the estimated interlinkage magnetic flux becomes the target interlinkage magnetic flux according to the target torque. Calculation unit and
A second calculation unit that calculates a second target voltage so that the applied current becomes a target current corresponding to the target torque by a second feedback control different from the first feedback control.
An adder that calculates the sum of the first target voltage calculated by the first calculation unit and the second target voltage calculated by the second calculation unit .
An inverter controlled according to the sum calculated by the adder, and
A calculation timing setting unit that determines the calculation timing in the path from the current operating point indicated by the applied current to the target operating point indicated by the target current.
A control device having a gain setting unit for obtaining a gain used for the first feedback control and the second feedback control at a calculation timing determined by the calculation timing setting unit.

Images