JP5502605B2 - Motor control device - Google Patents

Description

  The present invention relates to a motor control device that drives a motor used in industrial equipment or electric vehicles such as electric assist bicycles, electric motorcycles, and electric vehicles.

Conventionally, as a control method for driving a brushless motor, a 120-degree energization method having a low circuit configuration and high efficiency has been widely used. In the 120-degree energization method, the magnetic pole position of the rotor is detected using a magnetic pole position detection element, the switching element of the inverter circuit is turned on / off, current is passed through the windings of each phase, and the motor is driven. However, when switching the energized phase, torque ripple may be generated, causing vibration and noise, and this phenomenon is particularly noticeable in the low-speed rotation region.
To solve this problem, the upper and lower stages of the switching elements included in the inverter circuit in synchronization with the rotor position (the upper and lower stages of a pair of switching elements connected in series) are alternately turned ON / OFF to generate a sinusoidal modulated wave signal There is a control method (so-called sine wave drive method) that generates a current flowing through the motor winding and makes the current flowing in the motor winding a sine wave. However, an expensive encoder or resolver is required for rotor position detection, and the fence circuit for inputting the magnetic pole position detection signal to the microcomputer requires a large number of parts. There is a problem of becoming.
To solve these problems, a control method has been proposed in which an inexpensive Hall IC is used for magnetic pole position detection, and a pseudo sine wave-like modulated wave signal is generated from the magnetic pole position detection signal of the Hall IC to drive a sine wave. (For example, refer to Patent Document 1).

Japanese Patent Laid-Open No. 2004-120841

However, the three magnetic pole position detection signals for the U phase, the V phase, and the W phase have only six patterns including the rising and falling edges of each phase. Thus, for example, in the case of 12 slots motor 4 pole pairs rotor, a resolution of only 1 number of rotation per 6 × 4 of the motor. For this reason, a 6 × 4 pattern magnetic pole position detection signal generates a sinusoidal modulation wave signal, which causes problems such as a rough sine wave waveform and increased distortion.
In particular, in the low-speed rotation region of the motor, the update of the magnetic pole position detection signal is slow (the output interval of the magnetic pole position detection signal is long), so that a rough sine wave extends further on the time axis, and the sine of the modulation wave signal There is a problem that deviation and distortion from the waveform increase.
The present invention has been made in view of the above-described circumstances, and provides a motor control device that can improve generation of a sinusoidal modulation wave signal in a low-speed rotation region and can generate a modulation wave signal with less sine wave distortion. The purpose is to do.

In order to achieve the above object, the present invention provides a plurality of magnetic sensors provided on a rotor of a motor, and a plurality of magnetic sensors that output a magnetic pole position detection signal in accordance with a change in the magnetic pole position accompanying the rotation of the rotor. The magnetic sensor detects a magnetic pole position of the rotor based on a change and an arrangement position of each magnetic pole position detection signal of the magnetic sensor, generates a sine wave-shaped modulated wave signal based on the magnetic pole position, and In the motor control apparatus that performs sine wave driving, magnetic pole position interpolation means for interpolating the magnetic pole position while the magnetic pole position detection signal changes is provided, and the magnetic pole position interpolation means is required for the change of the magnetic pole position detection signal. The amount of movement per time obtained from the time and the amount of movement of the magnetic pole position when the magnetic pole position detection signal changes, and the elapsed time since the change of the magnetic pole position detection signal The magnetic pole position is interpolated based on the change of the magnetic pole position detection signal after the elapsed time exceeds the predetermined time indicating the extremely low speed immediately before the rotation of the motor is stopped, immediately after the start, and immediately after the rotation direction is reversed. The magnetic pole position is interpolated by increasing the magnetic pole position at every previous time required .

  According to the present invention, even if the change pattern of the magnetic pole position detection signal is small, the magnetic pole position during the change is interpolated by the magnetic pole position interpolation means, so that the modulated wave signal having a waveform close to a sine waveform can be obtained. Can be generated. In particular, even in a low-speed rotation region in which the magnetic pole position detection signal changes for a long time, the magnetic pole position between them is interpolated for a while, so that the sinusoidal modulation waveform is prevented from becoming extremely rough. Further, even when an inexpensive sensor such as a Hall IC is used for the magnetic sensor, a modulated wave signal close to a sine wave can be generated and the motor can be driven in a pseudo sine wave.

  According to the present invention, the magnetic pole position during the change of the magnetic pole position detection signal can be interpolated according to the rotational speed of the rotor (motor) at that time, and the modulated wave with little distortion regardless of the rotational speed of the motor. A signal can be generated.

  According to the present invention, even when the motor rotates at a very low speed and the time until the magnetic pole position detection signal changes becomes long, when the predetermined time has passed since the change of the magnetic pole position detection signal The degree of increase of the magnetic pole position can be suppressed, and switching to a modulated wave signal corresponding to the low speed rotation of the motor can be performed. In particular, since the magnetic pole position is increased at every previous time required for the change of the magnetic pole position detection signal, the amount of increase in the magnetic pole position with respect to the elapsed time can be suppressed relatively abruptly. In the case where the rotational speed of the motor is drastically reduced, a modulated wave signal corresponding to this can be generated.

  In the motor control device, the present invention further includes a pulse generation circuit that generates a series of pulse signals in which changes in the magnetic pole position detection signals of the magnetic sensors are reflected, the rising edge of the pulse signal of the pulse generation circuit, The magnetic pole position is detected based on the falling edge, and a sine wave-shaped modulated wave signal is generated based on the magnetic pole position, and the magnetic pole position interpolation means determines the magnetic pole position based on the rising / falling edge of the pulse signal. It is characterized by interpolation.

  According to the present invention, the pulse signal reflects a change in the magnetic pole position detection signal of each of the plurality of magnetic sensors, so that the cycle is shorter than that of one magnetic pole position detection signal. Then, by detecting the magnetic pole position based on the rise and fall of the pulse signal, the detection interval of the magnetic pole position is shortened, so that the sine wave distortion of the modulation wave signal can be more efficiently suppressed.

  In the motor control device according to the present invention, the magnetic pole position interpolation unit may determine the magnetic pole position in each period from the rising edge of the pulse signal to the next rising edge, or in each period from the falling edge to the next falling edge. It is characterized by interpolation.

  According to the present invention, it is possible to remove an interpolation error caused by various error components such as a magnetic sensor error and a detection magnet error. Thereby, even when an inexpensive Hall IC or the like is used for the magnetic sensor, the magnetic pole position can be accurately interpolated.

According to the present invention, even if the change pattern of the magnetic pole position detection signal is small, the magnetic pole position during the change is interpolated by the magnetic pole position interpolation means, so that the modulated wave signal having a waveform close to a sine waveform can be obtained. Can be generated. In particular, even in a low-speed rotation region in which the magnetic pole position detection signal changes for a long time, the magnetic pole position between them is interpolated for a while, so that the sinusoidal modulation waveform is prevented from becoming extremely rough. Further, even when an inexpensive sensor such as a Hall IC is used for the magnetic sensor, a modulated wave signal close to a sine wave can be generated and the motor can be driven in a pseudo sine wave.
Further, in the present invention, the magnetic pole position interpolation means has a movement amount per unit time obtained from a time required for the change of the magnetic pole position detection signal and a movement amount of the magnetic pole position when the magnetic pole position detection signal changes. The magnetic pole position can be interpolated according to the rotational speed of the rotor (motor) at that time by adopting a configuration in which the magnetic pole position is interpolated based on the elapsed time after the magnetic pole position detection signal changes. Thus, a modulated wave signal with little distortion can be generated regardless of the rotational speed of the motor.
In the present invention, the magnetic pole position interpolating means is provided at every previous time required for the change of the magnetic pole position detection signal after the elapsed time from the change of the magnetic pole position detection signal exceeds a predetermined time. By interpolating by increasing the magnetic pole position by a certain amount, even if the motor rotates at a very low speed and the time until the magnetic pole position detection signal changes becomes longer, the magnetic pole position is detected. When a predetermined time has passed since the change of the signal, the degree of increase in the magnetic pole position is suppressed, and it is possible to switch to a modulated wave signal corresponding to the low-speed rotation of the motor. In particular, since the magnetic pole position is increased at every previous time required for the change of the magnetic pole position detection signal, the amount of increase in the magnetic pole position with respect to the elapsed time can be suppressed relatively abruptly. In the case where the rotational speed of the motor is drastically reduced, a modulated wave signal corresponding to this can be generated.
The present invention further includes a pulse generation circuit that generates a series of pulse signals reflecting changes in the magnetic pole position detection signals of the magnetic sensors, and based on rising and falling edges of the pulse signals of the pulse generation circuit. The magnetic pole position is detected, a sine wave-like modulated wave signal is generated based on the magnetic pole position, and the magnetic pole position interpolation means is configured to interpolate the magnetic pole position between the rise and fall of the pulse signal. By detecting the magnetic pole position based on the rise and fall of the pulse signal having a shorter cycle than one magnetic pole position detection signal, the detection interval of the magnetic pole position is shortened, and the sine wave distortion of the modulation wave signal is further increased. It can be suppressed efficiently.
Further, in the present invention, the magnetic pole position interpolation means is configured to interpolate the magnetic pole position in each period from the rising edge to the next rising edge of the pulse signal or in each period from the falling edge to the next falling edge. Thus, interpolation errors caused by various error components such as a magnetic sensor error and a detection magnet error can be removed. Thereby, even when an inexpensive Hall IC or the like is used for the magnetic sensor, the magnetic pole position can be accurately interpolated.

It is a figure showing typically composition of a motor drive system concerning an embodiment of the present invention. It is a block diagram which shows DC brushless motor. It is the arrow line view seen from the II-II line in FIG. It is a top view which shows the structure of a stator. It is a figure which shows the structure of a magnetic detection circuit and a drive circuit. (A) is a figure which shows the output waveform of a 1st magnetic sensor, (B) is a figure which shows the output waveform of a noise filter, (C) is a figure which shows the output waveform of a waveform shaping circuit. (A) is a waveform diagram showing a three-phase rotor position detection pulse, (B) is a waveform diagram showing a rotor speed detection pulse, and (C) is a waveform diagram showing a sine wave signal. It is a table | surface figure which shows the truth table of a pulse generation circuit. (A) is explanatory drawing which shows a 3-phase rotor position detection pulse with an error component, (B) is explanatory drawing which shows the influence by an error component at the time of speed detection by a 1st method based on a rotor speed detection pulse. (C) is an explanatory view showing the influence of an error component when speed is detected by the first method. It is a block diagram which shows the motor control system of a motor drive system. It is a figure which shows the relationship between a rotor speed detection pulse signal and a magnetic pole position. It is a figure which shows a sine wave amplitude signal. It is a flowchart of a magnetic pole position interpolation process. It is a figure which each shows a rotor position detection pulse, the magnetic pole position which a magnetic pole position interpolator outputs, and the modulated wave signal of a sine waveform. It is a figure which shows the relationship between a motor rotation speed and an angle interpolation error. It is a block diagram which shows the motor control system of the motor drive system which concerns on the modification of this invention.

Hereinafter, embodiments of the present invention will be described with reference to the drawings. In this embodiment, a DC brushless motor (hereinafter simply referred to as “motor”) is illustrated as an electric motor used in an electric vehicle.
FIG. 1 is a diagram schematically illustrating a configuration of a motor drive system 1 according to the present embodiment.
As shown in the figure, the motor drive system 1 controls a DC power source E such as a vehicle-mounted battery, a motor 10 that is driven by electric power of the DC power source E, and a power source for driving the vehicle, and driving of the motor 10. And a control device 2 for performing the operation. The control device 2 controls the waveform of the motor drive current based on the command speed value that is the command value of the rotation speed of the motor 10, and controls the DC current of the DC power source E under the control of the drive circuit 4. And an inverter circuit 6 for converting to an AC motor driving current and outputting it to the motor 10. The inverter circuit 6 is a circuit in which six switching elements Tr1 to Tr6 such as power MOSFETs and IGBTs are combined, and each switching element Tr1 to Tr6 is turned on / off by a PWM drive signal from the drive circuit 4. The command speed value is generated by a computer on the vehicle side and input to the drive circuit 4 of the control device 2.

FIG. 2 is a configuration diagram illustrating the motor 10 according to the present embodiment.
In the present embodiment, as the motor 10, as illustrated in FIG. 2, an inner rotor type motor 10 having a rotor 12 and a stator 14 is illustrated. An outer rotor type motor can also be implemented.
The rotor 12 includes a cylindrical casing 16 and a rotor shaft 18 that extends in the axial direction at the center of the casing 16. As shown in FIG. 3, a permanent magnet 20 is disposed inside the housing 16 along the inner wall of the housing 16, and is configured as an 8-pole (4-pole pair) motor 10, for example. A detection magnet 22 is installed in a portion of the rotor 12 that faces the stator 14. The permanent magnet 20 is installed for rotating the rotor 12, and the detection magnet 22 is installed for detecting the speed of the rotor 12.

The detection magnet 22 is magnetized by alternately arranging N poles and S poles so as to correspond to the magnetic pole positions of the motor 10. Therefore, the detection magnet 22 has eight boundaries (the first boundary 24a to the eighth boundary 24h) with respect to the boundary between the N pole and the S pole.
On the other hand, as shown in FIG. 4, the stator 14 has twelve slots (first slot 26a to twelfth slot 261) at equal intervals along the circumference and, for example, in the clockwise direction, the first slot 26a and the second slot. The second slot 26b, the third slot 26c,... Are arranged in this order. Among them, the first slot 26a, the fourth slot 26d, the seventh slot 26g, and the tenth slot 26j are wound with the first phase winding, and the second slot 26b, the fifth slot 26e, the eighth slot 26h, and the A second phase winding is wound around the 11th slot 26k, and a third phase winding is wound around the third slot 26c, the sixth slot 26f, the ninth slot 26i and the twelfth slot 26l.
Further, the stator 14 is provided with a hole 28 into which the end of the rotor shaft 18 is inserted at the center thereof, and a mounting plate having a U-shape, for example, is formed so as to surround the hole 28 around the hole 28. 30 is installed. Further, around the hole 28, there are a first connection terminal portion 32a for the first phase winding, a second connection terminal portion 32b for the second phase winding, and a third connection for the third phase winding. A connection terminal portion 32c is provided.
Note that the winding connection method is not limited to the Y connection shown in the figure, and other connection methods such as a star connection and a delta connection can also be implemented.

Six magnetic sensors (first magnetic sensor 34 a to sixth magnetic sensor 34 f) are arranged on the surface (for example, the upper surface) of the mounting plate 30 facing the detection magnet 22 of the rotor 12.
Specifically, when the reference line m passing through the center position of the hole 28 and passing through the centers of the fifth slot 26e and the eleventh slot 26k is considered, the upper surface of the mounting plate 30 is on the reference line m. In addition, the first magnetic sensor 34a is installed at a position on the fifth slot 26e side, and the first magnetic sensor 34a is spaced from the first magnetic sensor 34a along the circumference of the hole 28 by, for example, 30 ° clockwise. The second magnetic sensor 34b is installed, and the third magnetic sensor 34c is installed at a position 30 ° away from the second magnetic sensor 34b along the circumference of the hole 28.
Similarly, a fourth magnetic sensor 34d is installed at a position on the reference line m on the upper surface of the mounting plate 30 and on the eleventh slot 26k side, and the circle of the hole 28 extends from the fourth magnetic sensor 34d. For example, a fifth magnetic sensor 34e is installed at a position spaced 30 ° in the counterclockwise direction along the circumference, and a sixth magnetic sensor is disposed at a position spaced 30 ° along the circumference of the hole 28 from the fifth magnetic sensor 34e. 34f is installed.

  Among the first magnetic sensor 34a to the sixth magnetic sensor 34f, the first magnetic sensor 34a to the third magnetic sensor 34c are for detecting the forward rotation speed of the rotor 12, and the first magnetic sensor 34a is the first magnetic sensor 34a. The second magnetic sensor 34b corresponds to the second phase, and the third magnetic sensor 34c corresponds to the third phase. Therefore, in other words, the first magnetic sensor 34a to the third magnetic sensor 34c are arranged in the order of the first magnetic sensor 34a, the second magnetic sensor 34b, and the third magnetic sensor 34c along the forward rotation direction of the rotor 12. And 30 ° apart from each other along the forward rotation direction of the rotor 12.

  Similarly, the fourth magnetic sensor 34d to the sixth magnetic sensor 34f are for detecting the reverse rotation speed of the rotor 12, the fourth magnetic sensor 34d corresponds to the first phase, and the fifth magnetic sensor 34e is the first one. Corresponding to the three phases, the sixth magnetic sensor 34f corresponds to the second phase. Therefore, in other words, the fourth magnetic sensor 34d to the sixth magnetic sensor 34f are arranged in the order of the fourth magnetic sensor 34d, the fifth magnetic sensor 34e, and the sixth magnetic sensor 34f along the reverse direction of the rotor 12. They are arranged and arranged 30 ° apart from each other along the reverse direction of the rotor 12.

  The first magnetic sensor 34a to the sixth magnetic sensor 34f are each configured by a Hall IC. The Hall IC is a magnetic sensor in which a Hall element and a logic circuit are integrated into an IC, and detects the magnetic pole (N pole or S pole) of the detection magnet 22 of the rotor 12 and the strength of the magnetic pole by the electromagnetic phenomenon of the Hall element. . Therefore, when the rotor 12 rotates, an analog voltage signal proportional to the magnetic flux density is output from the Hall element. The logic circuit shapes the analog voltage signal from the Hall element and outputs a digital waveform corresponding to the polarity of the magnetic field, for example, a high-level digital waveform at the N pole and a low-level digital waveform at the S pole. In this embodiment, for example, an open collector output type Hall IC having an output transistor is used.

A magnetic detection circuit 52 (FIG. 1) that detects the magnetic pole position of the motor 10 includes the first magnetic sensor 34a to the sixth magnetic sensor 34f. As shown in FIGS. 1 and 5 (described later), the magnetic detection circuit 52 includes a first magnetic sensor unit 64a including a first magnetic sensor 34a to a third magnetic sensor 34c and outputting a first digital waveform of three phases. And a second magnetic sensor unit 64b including a fourth magnetic sensor 34d to a sixth magnetic sensor 34f and outputting a three-phase second digital waveform. These first and second digital waveforms are input to the drive circuit 4 as magnetic pole position detection signals.
As shown in FIG. 1, the drive circuit 4 includes a magnetic pole position detection signal processing circuit 54, a pulse generation circuit 56, and a CPU 58. The CPU 58 operates at least a rotor position detecting means 60 and a rotor speed detecting means 62 as software.

FIG. 5 is a diagram showing the configuration of the drive circuit 4 together with the magnetic detection circuit 52. In addition, in the same figure, illustration is abbreviate | omitted about the structural member for producing | generating the signal input into the inverter circuit 6. FIG.
As shown in the figure, the magnetic pole position detection signal processing circuit 54 includes a digital waveform from the magnetic detection circuit 52, that is, a three-phase first digital waveform from the first magnetic sensor unit 64a and a second magnetic sensor unit 64b. A pull-up resistor 66 for stabilizing the second digital waveform of the three phases, a noise filter 68 (low-pass filter) for suppressing high-frequency components (noise), a Schmitt trigger function, and the noise filter 68 Based on a waveform shaping circuit 70 for shaping the waveform into a pulse waveform and a control signal Sc indicating normal rotation or reverse rotation of the rotor 12 from the CPU 58, the first digital waveform of three phases or the second digital waveform of three phases is switched and selected. And output as a three-phase rotor position detection pulse Sa (first-phase rotor position detection pulse Sa1 to third-phase rotor position detection pulse Sa3). And a selection circuit 72 that.

  The pulse generation circuit 56 is a circuit that generates a series of pulse signals that reflect each rising edge and each falling edge of the three-phase rotor position detection pulse Sa from the selection circuit 72. That is, the pulse generation circuit 56 performs exclusive OR of the first-phase rotor position detection pulse Sa1 and the second-phase rotor position detection pulse Sa2 among the three-phase rotor position detection pulses Sa from the selection circuit 72. The first logic circuit 74 for outputting, and the second logic circuit 76 for outputting the exclusive OR of the third-phase rotor position detection pulse Sa3 and the output of the first logic circuit 74 as the rotor speed detection pulse Sb.

The rotor position detection means 60 realized by the CPU 58 detects the position of the rotor 12 based on the three-phase rotor position detection pulse Sa from the selection circuit 72, and the rotor speed detection means 62 realized by the CPU 58 is a pulse generation circuit 56. The rotational speed (forward rotation speed or reverse rotation speed) of the rotor 12 is detected based on the rotor speed detection pulse Sb from. Further, the CPU 58 feedback-controls the rotational speed of the motor 10 based on the rotational speed detected by the rotor speed detecting means 62 and the speed command value input from the vehicle side. The configuration related to the motor control will be described later.
For example, when only the rotor position and the normal rotation speed at the time of the forward rotation of the rotor 12 are detected, the second magnetic sensor unit 64b (the fourth magnetic sensor 34d to the sixth magnetic sensor 34f) related to the reverse rotation of the rotor 12 is omitted. In response to this, some of the components of the pull-up resistor 66, the components of the noise filter 68, and the components of the waveform shaping circuit 70 can be omitted, and the selection circuit 72 can be omitted.

6A and 6B are diagrams for explaining the operation of the magnetic detection circuit 52 and the drive circuit 4, FIG. 6A is a diagram showing an output waveform of the first magnetic sensor 34a, and FIG. 6B is an output waveform of the noise filter. FIG. 6C is a diagram illustrating an output waveform of the waveform shaping circuit 70.
For example, the digital waveform output from the first magnetic sensor 34a will be described. For example, the magnetic pole facing the first magnetic sensor 34a as the rotor 12 rotates forward is, for example, N pole → S pole → N pole → S pole. Since the order changes, the output transistor of the first magnetic sensor 34a is turned off, for example, in a period facing the N pole, and turned on in a period facing the S pole. Since the pull-up resistor 66 is connected to the output of the first magnetic sensor 34a, the output voltage is raised to near the power supply voltage Vcc when the output transistor is OFF. In the output of the first magnetic sensor 34a, a high frequency component (noise) is suppressed by the noise filter 68 at the subsequent stage. However, the output waveform of the first magnetic sensor 34a is a first-order lag waveform in which the rise and fall correspond to the time constant of the noise filter 68, as shown in FIG.

The output of the first magnetic sensor 34a that has risen and fallen by the noise filter 68 is shaped into a pulse waveform by the Schmitt trigger function of the waveform shaping circuit 70 at the subsequent stage. Specifically, the waveform shaping circuit 70 is set to a low level when the output (output voltage) of the noise filter 68 becomes equal to or higher than the first threshold voltage Vtp, and when it becomes equal to or lower than the second threshold voltage Vtn (<Vtp). To a high level. The waveform shaping circuit 70 also prevents chattering.
On the other hand, the selection circuit 72 outputs the three-phase first digital waveform corresponding to the first magnetic sensor unit 64a output from the waveform shaping circuit 70 and the three-phase second digital waveform corresponding to the second magnetic sensor unit 64b, Based on the control signal Sc indicating normal rotation or reverse rotation of the rotor 12 from the rotor position detecting means 60 operated by the CPU 58, switching is selected and output as a three-phase rotor position detection pulse Sa. An example of a three-phase rotor position detection pulse Sa when the rotor 12 is rotating forward is shown in FIG.

The three-phase rotor position detection pulse Sa is used to detect forward rotation or reverse rotation of the rotor 12 and further to detect the position of the magnetic pole of the rotor 12. Of course, the speed of the rotor 12 can be detected at a timing of 4 cycles (pulse cycle Ta) per rotation of the rotor 12, but the detection accuracy is low.
In the pulse generation circuit 56, an exclusive OR of the first phase rotor position detection pulse Sa 1 and the second phase rotor position detection pulse Sa 2 is output from the first logic circuit 74, and the second logic circuit 76 outputs the second logic circuit 76. An exclusive OR of the three-phase rotor position detection pulse Sa3 and the output of the first logic circuit 74 is output. That is, the rotor speed detection pulse Sb is generated by a 3-input exclusive OR function.

  A truth table of the pulse generation circuit 56 is shown in FIG. In this truth table, if all of the first-phase rotor position detection pulse Sa1 to the third-phase rotor position detection pulse Sa3 are “0” or “1”, it is an abnormal output due to a disconnection or short circuit of the harness. The CPU 58 performs error processing such as motor stop.

The drive circuit 4 detects the rotational speed (forward rotation speed and reverse rotation speed) of the rotor 12 based on the output from the pulse generation circuit 56. Since the pulse period of the rotor speed detection pulse Sb output from the pulse generation circuit 56 is 1/3 of the pulse period Ta of the rotor position detection pulse Sa, it is based on the rotor speed detection pulse Sb output from the pulse generation circuit 56. By detecting the speed of the rotor 12, the detection accuracy can be improved as compared with the case where the three-phase rotor position detection pulse Sa is used.
As the detection timing, as a first method, it is conceivable that the timing at which the pulse waveform changes, that is, the period from the fall to the next rise and the period from the rise to the next fall are used as the speed detection period. . In this case, since 24 speed detection cycles arrive during one rotation of the rotor 12, the speed of the rotor 12 can be detected with high accuracy.

However, the start point of each speed detection period is often influenced by various error components. Examples of error components include a detection magnet error, a magnetic sensor error, a noise filter error, and a CPU error.
The detection magnet error includes an error in the magnetizing range of the detection magnet 22 and an installation error with respect to the rotor 12, and an arrangement error of the actual first boundary 24a to the eighth boundary 24h with respect to the ideal first boundary 24a to the eighth boundary 24h. Refers to ingredients. The magnetic sensor error includes a reading error of the first magnetic sensor 34a to the third magnetic sensor 34c and an attachment error with respect to the stator 14 when detecting the normal rotation speed of the rotor 12, and when detecting the reverse rotation speed of the rotor 12. The reading error of the fourth magnetic sensor 34d to the sixth magnetic sensor 34f and the mounting error with respect to the stator 14 are included. The noise filter error includes a time constant error caused by variations in circuit constants of circuit elements. The CPU error includes a quantization error when converting an analog signal to a digital signal. Of these, the CPU error depends on the performance of the CPU 58 itself, and is therefore excluded from the error components described above.

As shown in FIG. 6B, the time constant error is generated when the output waveform of the noise filter 68 becomes a first order lag waveform due to the CR time constant.
That is, the pulse waveform (see FIG. 6C) output from the waveform shaping circuit 70 at the subsequent stage ideally has a falling time point of the output waveform of the noise filter 68 (see FIG. 6B). The rise time is almost simultaneously with the rise time, and the rise time is almost simultaneously with the fall time of the output waveform of the noise filter 68.
However, since the output waveform of the noise filter 68 is a first-order lag waveform, the output (output voltage) of the noise filter 68 starts from the rising point of the output waveform of the noise filter 68 at the falling point of the output waveform of the waveform shaping circuit 70. This is delayed by a time until the time when the voltage becomes equal to or higher than the first threshold voltage Vtp, and this delay time Δt1 becomes a time constant error.

The rise time of the output waveform of the waveform shaping circuit 70 is delayed by the time from the fall time of the output waveform of the noise filter 68 to the time when the output (output voltage) of the noise filter 68 becomes equal to or lower than the second threshold voltage Vtn. However, this delay time Δt2 is short enough to be ignored. Therefore, the output waveform of the waveform shaping circuit 70 can be treated as having no time constant error at the rise time.
From this, as shown in FIG. 9A, each falling edge of the three-phase rotor position detection pulse Sa is affected by a noise filter error.

As described above, the detection magnet error includes an error in the magnetization range of the detection magnet 22 and an installation error, and therefore, the detection magnet error is affected by the detection magnet error when each magnetic sensor detects a change in the magnetic pole. That is, each falling edge and each rising edge of the three-phase rotor position detection pulse Sa are affected by the detection magnet error.
Specifically, when the first magnetic sensor 34a is used as the reference position, for example, in the first-phase rotor position detection pulse Sa1, the first falling from the reference position is the position error of the first boundary 24a in the detection magnet 22. The next first rising edge is affected by the position error (second detecting magnet error) of the second boundary 24b in the detecting magnet 22, and the next second falling edge is influenced by the (first detecting magnet error). Due to the influence of the position error (third detection magnet error) of the third boundary 24 c in the detection magnet 22, the next second rise is the position error (fourth detection magnet error of the fourth boundary 24 d in the detection magnet 22). ), The next third fall is affected by the position error (fifth detection magnet error) of the fifth boundary 24e in the detection magnet 22 and the next third fall. The rise is affected by the position error (sixth detection magnet error) of the sixth boundary 24f in the detection magnet 22, and the next fourth fall is the position error (seventh detection of the seventh boundary 24g in the detection magnet). It is influenced by the magnet error. The same applies to the second-phase rotor position detection pulse Sa2 and the third-phase rotor position detection pulse Sa3.

  As described above, the magnetic sensor error includes the reading error of the first magnetic sensor 34a to the third magnetic sensor 34c and the mounting error with respect to the stator 14 when detecting the forward rotation speed of the rotor 12. When the sensor 34a is used as a reference, each rising edge and each falling edge of the second-phase rotor position detection pulse Sa2 corresponding to the second magnetic sensor 34b is a magnetic sensor error (second magnetic sensor error) caused by the second magnetic sensor 34b. The rise and fall of the third-phase rotor position detection pulse Sa3 corresponding to the third magnetic sensor 34c is affected by the third magnetic sensor 34c (the third magnetic sensor error). It is influenced by.

Therefore, the rotor speed detection pulse Sb output from the pulse generation circuit 56 is affected by the various error components described above.
For example, as shown in FIG. 9B, at time t1, it is affected by the noise filter error, the third magnetic sensor error, and the seventh detection magnet error, at time t2, it is affected by the second magnetic sensor error, and at time t3. The same applies to the influence of the noise filter error and the first detection magnet error.
Therefore, when the speed of the rotor 12 is detected using the first method described above, that is, the period from the falling edge of the rotor speed detection pulse Sb to the next rising edge and the period from the rising edge to the next falling edge, respectively, as speed detection cycles, As shown in FIG. 9B, it is influenced by the error components of all the three-phase rotor position detection pulses Sa included in the time point t1 to the time point t23.

Therefore, as a second method, the speed of the rotor 12 is detected using the period from the falling edge of the rotor speed detection pulse Sb to the next arbitrary falling edge, and the period from the rising edge to the next arbitrary rising edge as the speed detection period. Is preferred. In this case, in order to increase the detection accuracy, it is preferable to satisfy the speed detection period <the pulse period Ta of the rotor position detection pulse Sa.
For example, when the speed of the rotor 12 is detected using the period from the fall of the rotor speed detection pulse Sb to the next fall (t0 → t2, t2 → t4, t4 → t6,...) As the speed detection cycle, FIG. C), the second magnetic sensor error at time t2, the third magnetic sensor error at time t4, the second detection magnet error at time t6, the second detection magnet error and the second magnetism at time t8. Sensor error, second detection magnet error and third magnetic sensor error at time t10, fourth detection magnet error at time t12, fourth detection magnet error and second magnetic sensor error at time t14, first detection at time t16 4 detection magnet error and third magnetic sensor error, sixth detection magnet error at time t18, sixth detection magnet error and second magnetic sensor error at time t20, sixth detection magnet error and third magnetic at time t22 Affected by sensor error Is a Runomi, it is possible to minimize the impact of the error component. This leads to further improvement in the detection accuracy of the speed of the rotor 12.

The period from the fall of the rotor speed detection pulse Sb used as the speed detection period to the next arbitrary fall is one period from the fall in addition to the period from the fall to the next fall described above. It is also possible to adopt a period (t0 → t4, t4 → t8, t8 → t12,...) Until the trailing edge.
Further, as a period from the rise of the rotor speed detection pulse Sb used as the speed detection period to the next arbitrary rise, periods from the rise to the next rise (t1 → t3, t3 → t5, t5 → t7,... Or a period from a rising edge to every other rising edge (t1 → t5, t5 → t9, t9 → t13,...) May be adopted.
In any case, it is preferable to adopt a period in which the influence of the error component is minimized as the speed detection period.
As described above, the speed detection circuit 50 according to the present embodiment can minimize the influence of the error component, and can detect the speed of the rotor 12 with high accuracy.

That is, in a motor system that performs speed control, the detection accuracy of the rotational speed of the motor 10 (rotor 12) is extremely important. Therefore, it is conceivable to use an expensive encoder or resolver for detecting the number of rotations of the motor 10 in a motor system that performs speed control, although it is highly accurate. On the other hand, the Hall IC used in the present embodiment is inexpensive, but has a problem that the resolution is low and generally the detection error of the rotational speed of the motor 10 is large.
On the other hand, the speed of the rotor 12 can be detected with high accuracy by providing the pulse generation circuit 56 that generates the rotor speed detection pulse Sb based on the three-phase rotor position detection pulse Sa as in the present embodiment. . Furthermore, the detection accuracy can be further improved by selecting a period from the falling edge of the rotor speed detection pulse Sb to the next arbitrary falling edge or a period from the rising edge to the next arbitrary rising edge as the speed detection period. . In particular, by selecting the period from the fall of the rotor speed detection pulse Sb to the next fall as the speed detection period, the influence of the error component can be minimized, and as a result, the speed detection accuracy of the rotor 12 can be improved. The rotor position detection pulse Sa can be improved three times as long as the pulse detection period Ta is the speed detection period, and the error component can be reduced to about 1/10.
As a result, even if a Hall IC is used as the magnetic sensor, the speed control of the motor 10 can be performed with high accuracy, and the speed command value from the CPU 58 can be followed with high accuracy with almost no error.
Furthermore, the detection accuracy can be further improved by selecting a period from the falling edge of the rotor speed detection pulse Sb to the next arbitrary falling edge or a period from the rising edge to the next arbitrary rising edge as the speed detection period. . In particular, by selecting the period from the fall of the rotor speed detection pulse Sb to the next fall as the speed detection period, the influence of the error component can be minimized, and as a result, the speed detection accuracy of the rotor 12 can be improved. The rotor position detection pulse Sa can be improved three times as long as the pulse detection period Ta is the speed detection period, and the error component can be reduced to about 1/10.

Next, the drive control system of the motor 10 will be described.
FIG. 10 is a block diagram showing a motor drive control system of the motor drive system 1.
In this motor drive control system, a command speed ω _rm ^{*, which} is a speed command value, is input to the control device 2 from a computer mounted on the vehicle, and feedback control is performed so that the motor 10 is driven at the command speed ω _rm ^*. Is called. That is, as shown in FIG. 10, the motor drive control system roughly includes a position / speed detector 82, a speed controller 84, a drive signal generator 86, and an angle interpolator 88. These units are realized by the CPU 58 provided in the drive circuit 4.
The position / speed detector 82 detects the magnetic pole position θ _{re of} the rotor 12 and the rotational speed ω _rm of the rotor 12, outputs the rotational speed ω _rm to the speed controller 84, and outputs the magnetic pole position θ _re to the angle interpolator. The rotor position detecting means 60 and the rotor speed detecting means 62 described above are included.

The speed control unit 84 calculates a voltage command value based on the command speed ω _rm ^* input from the computer included in the vehicle and the rotational speed ω _rm detected by the position / speed detection unit 82, and the drive signal generation unit 86. Output to. More specifically, the speed control unit 84 calculates a torque command value from the command speed ω _rm ^* and the rotation speed ω _rm, and a duty that can be searched by the torque command value and the command speed ω _rm ^* ( (Duty) map 91 and advance angle map 92 are provided.
The speed controller 90 has a proportional (P) controller and an integral (I) controller. Based on the deviation between the command speed ω _rm ^* and the rotational speed ω _rm , the speed controller 90 compensates for the error by PI control and generates a torque. Calculate the command value. The duty map 91 is a map that preliminarily defines the mutual relationship between the command speed ω _rm ^* , the duty of a PWM drive signal to be described later, and the torque. By searching the duty map 91 using the torque command value and the command speed ω _rm ^* , the duty at which the torque of the torque command value is obtained is determined. The advance angle map 92 is a map that predetermines the mutual relationship between the command speed ω _rm ^* , the advance angle, and the torque. That is, by searching the advance angle map 92 by the torque command value and the command speed omega _rm ^*, the optimum advance angle with respect to the torque command value and the command speed omega _rm ^* is obtained. Then, a voltage command value including these lead angle and duty indication values is input to the drive signal generator 86.

The drive signal generation unit 86 includes a waveform generator 93 and a PWM drive signal generation unit 94. Waveform generator 93, U-phase based on the magnetic pole position theta _re of the rotor 12 from the angle interpolation unit 88 to be described later, together with the V-phase, and each phase of the sine amplitude signal sin [theta _re the W phase are inputted, the speed control unit 84 The voltage command value is input from the above, and the signal amplitude of the sine amplitude signal sin θ _re of each phase is adjusted by the voltage command value, and the PWM drive signal generation unit is set as a sine wave modulation wave signal Vua ^* , Vva ^* , Vwa ^* , respectively. Output to 94. The PWM drive signal generation unit 94 generates a PWM drive signal based on the modulation wave signals Vua ^* , Vva ^* , and Vwa ^* of each sine wave. Specifically, the PWM drive signal generation unit 94 generates a PWM drive signal by a triangular wave comparison asynchronous method, and each modulated wave signal Vua ^* , Vva ^* , Vwa ^{* of} each sine wave and a triangular wave that is a carrier wave. The voltage values are compared to generate a PWM drive signal and output to the inverter circuit 6.

The angle interpolation unit 88 generates a sine amplitude signal sin θ _re based on the magnetic pole position θ _re of the rotor 12 output from the position / velocity detection unit 82 and outputs the sine amplitude signal sin θ _re to the drive signal generation unit 86. 95 and a trigonometric function calculator 96. As shown in FIG. 7C, the sine amplitude signal sin θ _re is a sine wave amplitude signal having a period synchronized with the rotor position detection pulse Sa.
In the present embodiment, since the pulse period of the rotor speed detection pulse Sb is 1/3 (electrical angle 60 degrees) of the pulse period Ta of the rotor position detection pulse Sa, the rise and fall of the rotor speed detection pulse Sb. The sine amplitude signal sin θ _re is generated at each timing, and thereby the resolution is increased as compared with the case where the sine amplitude signal sin θ _re is generated based on the rise and fall of the rotor position detection pulse Sa. Of course, the electrical angle varies depending on the structure of the motor 10 (the structure of the rotor and stator, the number of slots, etc.).
The magnetic pole position interpolator 95 outputs the magnetic pole position θ _re to the trigonometric function calculator 96 at the rising and falling timings of the rotor speed detection pulse Sb. That is, when the motor 10 is rotating in a certain direction, the magnetic pole position θ _re is 0 degree → 60 degrees → 120 degrees → 180 degrees → 240 degrees → 300 each time the rotor speed detection pulse Sb rises and falls. It changes cyclically from 0 degrees to 360 degrees at intervals of 60 degrees, such as degrees → 360 degrees (0 degrees) → 60 degrees. The magnetic pole position interpolator 95 sets the timing at which the rotor speed detection pulse Sb first rises or falls at the start of rotation of the motor 10 or after detection of reversal of the rotation direction to the magnetic pole position θ _re = 0 degrees, and then Each time the rotor speed detection pulse Sb rises or falls, the magnetic pole position θ _re output to the trigonometric function calculator 96 is increased by 60 degrees.

The trigonometric function calculator 96 calculates a sine value (sin θ _re ) of the magnetic pole position θ _re every time the magnetic pole position θ _re is input, and outputs the sine amplitude signal sin θ _re to the waveform generator 93. At this time, the trigonometric function calculator 96 updates the value of the sine amplitude signal sin θ _re every time the magnetic pole position θ _re is input, and the output signal waveform of the trigonometric function calculator 96 is a so-called stepped signal waveform. It becomes. Therefore, if the magnetic pole position interpolator 95 outputs the magnetic pole position θ _re only at every electrical angle of 60 degrees, the sine amplitude signal sin θ output from the trigonometric function calculator 96 as shown by a one-dot chain line in FIG. _re is a rough signal Sj with a large deviation from the sine waveform. In particular, when the rotational speed ω _rm of the motor 10 is slow, the pulse period Tc of the rotor speed detection pulse Sb becomes long. Therefore, the sine amplitude signal sin θ _re becomes a waveform obtained by extending the rough signal Sj on the time axis. The deviation from is remarkable.
Therefore, in this embodiment, the magnetic pole position interpolator 95 interpolates the magnetic pole position θ _re at a constant time interval and outputs it to the trigonometric function calculator 96 in addition to the rise and fall of the rotor speed detection pulse Sb. The resolution of the sine amplitude signal sin θ _re is increased.
In this embodiment, as shown in FIG. 12, the sine wave signal of one cycle is divided into 512, that is, the angle resolution is set to 512/360 (deg), and the magnetic pole position interpolator 95 interpolates the angle with this angle resolution. The graph of the sine amplitude signal sin θ _re output for a while by the trigonometric function calculator 96 is made to be sufficiently approximate to a sine wave.

By the way, when the carrier frequency of the triangular wave used for the carrier wave for generating the PWM control signal is fc and the command voltage frequency that is the frequency of the voltage command value input from the speed control unit 84 to the waveform generator 93 is fl, the triangular wave comparison is asynchronous. If the motor 10 is driven by the PWM control of the equation and the sine wave modulation wave signals Vua ^* , Vva ^* and Vwa ^* generated as PWM drive signals are not commanded at a sufficient speed, the waveform will be broken. For this reason, the waveform quality is maintained by setting fc / fl ≧ N (= 9). (Here, N = 9 is obtained from a value experiment, an empirical value, or the like as a value for obtaining a practical frequency for the carrier frequency fc.)
That is, when the set maximum rotational speed of the motor 10 is X [rpm], the command voltage frequency fl = X / 60 × (number of poles / 2) [Hz] is obtained, so that the carrier frequency fc of the carrier wave = fc = If Y [kHz] is set so as to satisfy the relationship of fc / fl ≧ N (= 9), the waveform does not collapse. In the present embodiment, the maximum set rotational speed X is 3500 [rpm], and fc = 16 kHz is used as the carrier frequency fc.
The magnetic pole position interpolator 95 is interrupted every carrier frequency fc (that is, every 62.5 μsec), and the magnetic pole position interpolator 95 performs the magnetic pole position interpolation processing every time this interrupt occurs.

The magnetic pole position interpolation process is a process of interpolating the magnetic pole position θ _re between the rising edge and the falling edge of the rotor speed detection pulse Sb. Since the rotor speed detection pulse Sb rises and falls (ie, 1/2 pulse period) corresponds to an electrical angle of 60 degrees, the interpolation angle α is set to 60 from the rise to the fall of the rotor speed detection pulse Sb. By gradually increasing (forward rotation) or decreasing (reverse rotation) according to the elapsed time, the interpolation angle α is _added to the magnetic pole position θ _re output at the previous rise or fall of the rotor speed detection pulse Sb. Interpolation is performed by adding.

FIG. 13 is a flowchart of magnetic pole position interpolation processing.
As shown in the figure, when an interrupt occurs (step S1), the magnetic pole position interpolator 95 first determines whether or not the signal level has changed as the rotor speed detection pulse Sb rises or falls ( Step S2). If there is a change in the signal level (step S2: YED), 30 degrees is set as the initial value of the interpolation angle α (step S3). This initial value is a value determined based on the number of slots of the motor 10. That is, when the number of slots of the motor 10 is 12, the slots are arranged at intervals of 30 degrees, and therefore the first magnetic sensor 34a to the sixth magnetic sensor 34f are also arranged at intervals of 30 degrees. Therefore, a magnetic pole position detection signal is output every 30 degrees of electrical angle from any of the first magnetic sensor 34a to the sixth magnetic sensor 34f. Since the rotor speed detection pulse Sb is generated by the pulse generation circuit 56 based on the magnetic pole position detection signal, the magnetic pole position θ _re has already moved by an electrical angle of 30 degrees when the rotor speed detection pulse Sb rises and falls. This electrical angle of 30 degrees is set as an initial value. When the number of slots is M and a magnetic sensor is arranged for each slot, (360 degrees / slot number M) is set as an initial value.
When a magnetic sensor is disposed in each of the U-phase, V-phase, and W-phase slots,
(360 degrees / number of slots M) × 2> (360 degrees / number of magnetic poles)> (360 degrees / number of slots M)
The relationship holds.

Next, the magnetic pole position interpolator 95 determines whether or not the rotation direction of the motor 10 is reversed (step S4). When the rotation direction of the motor 10 is rotating in the same direction (step S4: NO), on condition that an interrupt counter value Cn described later has not reached the count upper limit value Cmax (step S5: NO), “Normal interpolation” is set to the interpolation flag Qf for designating the calculation algorithm of the interpolation angle α (step S6). On the other hand, when the rotation direction of the motor 10 is reversed (step S4: YES), or when the interrupt counter value Cn has reached the count upper limit value Cmax (step S5: YES), an algorithm for calculating the interpolation angle α is set. In order to change from the normal algorithm, “very low speed interpolation” is set in the interpolation flag Qf (step S7).
Here, since the output order of the magnetic pole position detection signals of the magnetic sensor 34a to the sixth magnetic sensor 34f is uniquely determined for each rotation direction of the motor 10, the magnetic pole position detection signal is output in step S4. Based on the current magnetic sensor 34 and the output order of the magnetic pole position detection signals of each magnetic sensor 34 until the previous time, it is determined whether or not the rotation direction of the motor 10 is reversed. The order of the magnetic sensor 34 that outputs the magnetic pole position detection signal is updated and recorded in the RAM or the like every time the signal level of the rotor speed detection pulse Sb changes in step S2.

The interrupt counter value Cn is a value of a counter that counts the number of interrupts generated while the signal level of the rotor speed detection pulse Sb changes. Every time the signal level of the rotor speed detection pulse Sb changes once, the magnetic pole position θ _re changes by an electrical angle of 60 degrees, so that the final count value Cend (change in the previous signal level) of the interrupt counter value Cn during this time The amount of movement of the magnetic pole position θ _re per one occurrence of interruption (the amount of movement per interruption occurrence time) is obtained by dividing the electrical angle by 60 degrees by the time required for. The magnetic pole position interpolator 95 finally calculates the current interrupt counter value Cn counted between the previous change of the rotor speed detection pulse Sb and the current signal level in preparation for the calculation of the interpolation angle α performed at the next interrupt occurrence. The count value Cend is set and recorded (step S8).

Here, the change interval (cycle) of the signal level of the rotor speed detection pulse Sb becomes longer as the rotation speed of the motor 10 is slower, and the interrupt counter value Cn increases in proportion to this, so the change in the signal level. When the interval becomes very long, an overflow occurs in the interrupt counter value Cn. Specifically, in this embodiment, since the interrupt is generated at 16 kHz, when the interrupt counter value Cn is incremented every 62.5 μsec and the interrupt counter is configured using a 10-bit timer, 62 Since .5 .mu.sec.times.1024.apprxeq.64 msec is the upper limit value of the count, overflow occurs when the signal level change interval exceeds this. That is, for example, when the rotation speed of the motor 10 is as slow as 30 rpm, the time of one cycle of the U phase is (30/60) × 4 and 2 Hz (500 msec), and thus overflow occurs. When an overflow occurs, the interpolation angle α calculated based on the interrupt counter value Cn is distorted. Therefore, in the present embodiment, based on the number of bits of the timer constituting the interrupt counter, the above-described count upper limit value Cmax is set in a range where overflow does not occur, and the interrupt counter value Cn exceeds the count upper limit value Cmax. In this case, it is assumed that the motor 10 is rotating at an extremely low speed, and the algorithm for calculating the interpolation angle α is changed from a normal algorithm. For this reason, as described above, the magnetic pole position interpolator 95 determines whether or not the interrupt counter value Cn has reached the count upper limit value Cmax in step S5, and if not (step S5: NO). Then, “normal interpolation” is set in the interpolation flag Qf (step S6), and if the count upper limit Cmax is exceeded (step S5: YES), “very low speed interpolation” is set in the interpolation flag Qf. Yes.
The timing at which the rotation of the motor 10 becomes extremely low includes before the motor 10 is stopped, immediately after starting, and immediately after the rotation direction is reversed.

Next, the magnetic pole position interpolator 95 uses a very low speed parameter lowpath used for the interpolation angle α when the motor 10 is rotating at a very low speed.
Low speed parameter lowpath
= (Current interrupt counter value) + 1
Is calculated and recorded (step S9). In this equation, the current interrupt counter value Cn indicates a count value (corresponding to an elapsed time) until the signal of the rotor speed detection pulse Sb changes.

  The magnetic pole position interpolator 95 sets the interpolation flag Qf, records the final count value Cend, and sets the very low speed parameter lowpath as described above, and then clears the interrupt counter value Cn to zero (step S10) The processing procedure is returned to step S1.

When the signal level of the rotor speed detection pulse Sb has not changed at the time of occurrence of an interrupt (step S1) (step S2: NO), the interpolation angle α of the magnetic pole position _θre is calculated. That is, the magnetic pole position interpolator 95 first increments the interrupt counter value Cn by “1” (step S11), and as a result, determines whether or not the interrupt counter value Cn has reached the count upper limit value Cmax. (Step S12). When the count upper limit value Cmax is reached (step S12: YES), “interpolation for extremely low speed” is set in the interpolation flag Qf to indicate that the motor 10 is rotating at an extremely low speed (step S13).

Then, the magnetic pole position interpolator 95 calculates the interpolation angle α of the magnetic pole position θ _re with an algorithm according to the interpolation flag Qf (step S14). Specifically, when the interpolation flag Qf is “normal interpolation” (step S14: normal interpolation), the interpolation angle α is set to the following in order to temporarily increase or decrease the interpolation angle α in proportion to the interrupt counter value Cn. Calculation is performed using an equation (step S15).
Interpolation angle α =-(initial value of interpolation angle α)
+ (Current interrupt counter value Cn × 60 degrees / final count value Cend)
As a result, the interpolation angle α is calculated in which the interpolation angle α increases for a while from −30 degrees (or +30 degrees) in proportion to the interrupt counter value Cn.

On the other hand, when the interpolation flag Qf is “very low speed interpolation” (step S14: very low speed interpolation), the interpolation counter α is not simply increased in proportion to the interrupt counter value Cn, but the interrupt counter value is increased. When the remainder obtained by dividing Cn by the very low speed parameter lowpath becomes zero, that is, when the interrupt counter value Cn is increased by the very low speed parameter lowpath, the interpolation angle α is incremented by 1 ( Step S16).
As a result, the increase in the interpolation angle α becomes slower with respect to the increase in the interrupt counter value Cn, and the interpolation angle α can be gradually changed according to the low speed rotation of the motor 10. Also very low speed parameter lowpath indicates the previous elapsed time required for the change of the signal level of the rotor speed detection pulse Sb, every elapsed time for increasing the magnetic pole position theta _re, the magnetic pole position theta _re with respect to the elapsed time The increase is suppressed relatively rapidly. As a result, for example, when the rotational speed of the motor 10 is abruptly reduced just before the motor 10 is stopped, the magnetic pole position θ _re can be interpolated accordingly. Even if the interrupt counter value Cn overflows, the interrupt counter value Cn fluctuates cyclically, so that the interrupt counter value Cn fluctuates by the very low speed parameter lowpath regardless of the presence or absence of overflow. Each time the interpolation angle α can be increased.

Next, the magnetic pole position interpolator 95 sets the interpolation angle α to −30 degrees, for example, when the interpolation angle α becomes −30 degrees or less due to an overflow of the interrupt counter or the like (step S17: YES). When the angle exceeds 30 degrees (step S19: YES), the interpolation angle α is set to 30 degrees (maximum value that can be taken as the interpolation angle α) (step S20).
The magnetic pole position interpolator 95 calculates the interpolated magnetic pole position θ _re by adding the interpolation angle α to the magnetic pole position θ _re output at the previous rise or fall of the rotor speed detection pulse Sb, and the trigonometric function calculator After outputting to 96 (step S21), the process returns to step S1.
For details on the calculation of the interpolated magnetic pole position _θre ,
Interpolated magnetic pole position θ _re =
Magnetic pole position θ _re when the signal level of rotor speed detection pulse Sb changes
+ (Interpolation angle α + advance) × rotation direction coefficient Pdir
However, in the case of forward rotation, Pdir = 1, and in the case of reverse rotation, Pdir = −1.
Is calculated by

By the above magnetic pole position interpolation processing, the magnetic pole position θ _re is interpolated also during timings other than the rising / falling timing of the rotor speed detection pulse Sb, and as shown in FIG. 14, the rotor 12 makes one rotation (electrical angle 360 degrees). In the meantime, the magnetic pole position θ _re is output to the magnetic pole position interpolator 95 with an angular resolution of 360/512. As a result, the graph waveform of the sine amplitude signal sin θ _re output from the trigonometric function calculator 96 is also close to a sine wave, and the sinusoidal modulation wave signal generated from the sine amplitude signal sin θ _re is brought close to a sine wave. Can do.

Here, as described above, various error components are included in the detected magnet errors by the first magnetic sensor 34a to the sixth magnetic sensor 34f, and therefore based on the outputs of the first magnetic sensor 34a to the sixth magnetic sensor 34f. An angle error is also included in the interpolation angle α of the magnetic pole position _θre . As shown in FIG. 15, this angle error depends on the number of rotations of the motor 10 and becomes larger at higher speeds. Therefore, in order to eliminate this angular error, the magnetic pole position θ _re is not interpolated between the rising and falling edges of the rotor speed detection pulse Sb, but in the same way as when the rotor speed is detected, the rising edge of the rotor speed detection pulse Sb. By setting each period from the fall to the next fall or each period from the rise to the next rise, the angle error of the interpolation angle α can be removed. In particular, the influence of the error component can be minimized by interpolating the magnetic pole position θ _{re in} each period from the fall of the rotor speed detection pulse Sb to the next fall.

As described above, according to the control device 2 of the motor 10 of the present embodiment, the magnetic pole position θ _re while the signal level of the rotor speed detection pulse Sb changes is interpolated by the magnetic pole position interpolator 95. Regardless of the period of the rotor speed detection pulse Sb, the waveform generator 93 can generate a modulated wave signal having a waveform close to a sine waveform.
In particular, as the motor 10 rotates at a lower speed, the period during which the signal level of the rotor speed detection pulse Sb changes becomes longer. However, since the magnetic pole position _θre is interpolated for a while, the sinusoidal modulation waveform becomes extremely rough. It is prevented from becoming. Further, by using an inexpensive sensor such as a Hall IC as a magnetic sensor, a modulated wave signal close to a sine wave can be generated to drive the motor 10 with a pseudo sine wave.

Further, according to the present embodiment, the magnetic pole position interpolator 95 includes the time (final count value Cend) previously required to change the signal level of the rotor speed detection pulse Sb and the magnetic pole position θ when the signal level changes. _The magnetic pole position θ based on the movement amount per hour (electrical angle 60 degrees / final count value Cend) obtained from the _re movement amount and the elapsed time (interrupt counter value Cn) after the signal level changes. _{The re} is interpolated. With this configuration, the magnetic pole position θ _re while the signal level changes can be interpolated according to the rotational speed of the motor 10 at that time, and a modulated wave signal with little distortion is generated regardless of the rotational speed of the motor 10. it can.

Further, according to the present embodiment, the magnetic pole position interpolator 95 has the elapsed time (interrupt counter value Cn) after the signal level of the rotor speed detection pulse Sb has changed exceeds a predetermined time (count upper limit value Cmax). thereafter, (in this embodiment a time) a predetermined amount for each previous time required for the change of the signal level (Lowpath) was configured to interpolate by increasing the magnetic pole position theta _re each.
With this configuration, even when the motor 10 rotates at a very low speed and the time until the signal level of the rotor speed detection pulse Sb changes becomes long, a predetermined time (count upper limit value) from the previous change time. when Cmax) is exceeded, the degree of increase in the magnetic pole position theta _re is suppressed, can be switched to a modulation wave signal corresponding to the low speed rotation of the motor 10. In particular, since the magnetic pole position θ _re is increased at every previous time required to change the signal level, the increase amount of the magnetic pole position θ _re with respect to the elapsed time can be suppressed relatively abruptly. When the rotational speed of the motor 10 decreases rapidly as in the case of a time, a modulated wave signal corresponding to this can be generated.

In addition, according to the present embodiment, the pulse generation circuit 56 that generates a series of pulse signals reflecting changes in the magnetic pole position detection signals of the magnetic sensor 34 is provided. The magnetic pole position θ _re is detected based on the falling edge, and a sine wave-like modulated wave signal is generated based on the magnetic pole position θ _re , and the magnetic pole position interpolator 95 includes a magnetic pole between the rising edge and the falling edge of the pulse signal. The position θ _re is interpolated.
With this configuration, the pulse signal reflects changes in the magnetic pole position detection signals of the plurality of magnetic sensors 34, and therefore the cycle is shorter than that of one magnetic pole position detection signal (in this embodiment, three phases). Since a pulse signal is generated from the magnetic pole position detection signal for one minute, the period becomes 1/3). And by detecting the magnetic pole position θ _re based on the rise and fall of this pulse signal, the detection interval of the magnetic pole position θ _re can be shortened, and the sine wave distortion of the modulation wave signal can be more efficiently suppressed. Can do.

Further, according to the present embodiment, the magnetic pole position interpolator 95 interpolates the magnetic pole position θ _{re in} each of the period from the rising edge of the pulse signal to the next rising edge or in each period from the falling edge to the next falling edge. The configuration. According to this configuration, the angle error of the interpolation angle α caused by various error components such as an error of the magnetic sensor 34 and a detected magnet error can be removed. Thereby, even when an inexpensive Hall IC or the like is used for the magnetic sensor 34, the magnetic pole position _θre can be accurately interpolated.

In addition, embodiment mentioned above shows the one aspect | mode of this invention to the last, and a deformation | transformation and application are arbitrarily possible within the scope of the present invention.
For example, in the above-described embodiment, the magnetic pole position interpolator 95 detects the magnetic pole position θ _re based on the change in the signal level of the rotor speed detection pulse Sb. However, the present invention is not limited to this, and the rotor shown in FIG. A configuration may be _adopted in which the magnetic pole position _θre is detected based on each rising / falling of the position detection pulse Sa. In this case, the magnetic pole position _θre is interpolated in each period from the rise / fall of one of the rotor position detection pulses Sa to the rise / fall of another rotor position detection pulse Sa. The magnetic pole position interpolation processing at that time is the same as the flowchart shown in FIG.

Further, for example, in the above-described embodiment, the motor drive control by the command speed ω _rm ^* is exemplified as the motor drive system 1, but is not limited thereto. That is, as shown in FIG. 16, a motor drive system 100 that drives the motor 10 based on the command torque T ^* may be used. In this figure, the same members as those in FIG.

In the figure, the command torque T ^* is determined according to a torque map that defines the relationship between the rotational speed ω _rm of the motor 10 and the throttle opening and the target torque command value. That is, the rotational speed ω _{rm of the} motor 10 is input from the position / speed detection unit 82 to the computer or the like provided in the vehicle, and the throttle opening is input, and the computer or the like receives a torque command value based on the torque map. Is input to the torque controller 190 of the motor drive system 100 as the command torque T ^* .
The torque controller 190 adjusts the output torque T so that the current output torque T becomes the command torque T ^* , and outputs it to the duty map 91 and the advance angle map 92. The duty map 91 and the advance map 92 also _{receive the} rotational speed ω _rm , whereby the optimum duty and the advanced angle for the command torque T ^* and the rotational speed ω _rm are input to the drive signal generator 86. By outputting, drive control of the motor 10 based on the command torque T ^* is performed.

  Moreover, although embodiment mentioned above demonstrated embodiment using an inverter, voltage conversion circuits, such as a DC / DC converter circuit which raises / lowers the voltage of the battery E between the battery E and the inverter circuit 6, were incorporated. It is also possible to apply to a circuit.

1, 100 Motor drive system 2 Control device (motor control device)
DESCRIPTION OF SYMBOLS 4 Drive circuit 6 Inverter circuit 12 Rotor 14 Stator 20 Permanent magnet 22 Detection magnet 34, 34a-34f Magnetic sensor 50 Speed detection circuit 52 Magnetic detection circuit 54 Magnetic pole position detection signal processing circuit 56 Pulse generation circuit 58 CPU
Reference Signs List 60 rotor position detection means 60 trigonometric function calculator 62 rotor speed detection means 82 position / speed detection section 84 speed control section 86 drive signal generation section 88 angle interpolation section 93 waveform generator 94 PWM drive signal generation section 95 magnetic pole position interpolator 96 Trigonometric function calculator Cend Final count value Cmax Count upper limit value Cn Interrupt counter value Sa, Sa1 to Sa3 Rotor position detection pulse Sb Rotor speed detection pulse Vua * to Vwa * Modulation wave signal Lowpath Parameter for very low speed sinθ _re Sine amplitude signal θ _re Magnetic pole position ω _rm Rotational speed ω _rm * Command speed

Claims (3)

A plurality of magnetic sensors are provided on the rotor of the motor, and a plurality of magnetic sensors that output a magnetic pole position detection signal in accordance with a change in the magnetic pole position accompanying the rotation of the rotor are disposed at positions facing the rotor
A magnetic pole position of the rotor is detected based on a change and an arrangement position of each magnetic pole position detection signal of the magnetic sensor, a sine wave-shaped modulation wave signal is generated based on the magnetic pole position, and a sine wave drive of the motor is performed. In the motor control device to perform,
Magnetic pole position interpolation means for interpolating the magnetic pole position while the magnetic pole position detection signal is changed,
The magnetic pole position interpolation means includes
The amount of time required to change the magnetic pole position detection signal, the amount of movement per time obtained from the amount of movement of the magnetic pole position when the magnetic pole position detection signal changes, and the time elapsed since the change of the magnetic pole position detection signal And interpolating the magnetic pole position based on time,
After the elapsed time exceeds the predetermined time indicating the extremely low speed immediately before the rotation of the motor is stopped, immediately after the start, and immediately after the reversal of the rotation direction, the magnetic pole is detected at every previous time required for the change of the magnetic pole position detection signal. A motor control device that interpolates the magnetic pole position by increasing the position . A pulse generation circuit that generates a series of pulse signals reflecting changes in the respective magnetic pole position detection signals of the magnetic sensor, and detects the magnetic pole position based on the rise and fall of the pulse signal of the pulse generation circuit; And generating a sinusoidal modulated wave signal based on the magnetic pole position,
The motor control apparatus according to claim 1 , wherein the magnetic pole position interpolating unit interpolates the magnetic pole position based on a rise / fall of the pulse signal. The magnetic pole position interpolator, respectively from the rise of the pulse signal of the period until the next rising or falling in claim 2, wherein the interpolating of each magnetic pole position of the period until the next falling The motor control apparatus described.

Images