JP2023041534A - magnetic encoder - Google Patents

Abstract

To measure an angle with high accuracy.SOLUTION: A magnetic encoder 1 includes a rotatable disk-shaped magnet 10 having a plurality of magnetic pole pairs, four magnetic sensors 20 arranged on the surroundings of the magnet 10, and a computing device 30 that calculates a mechanical angle, which is a rotation angle of the magnet 10 on the basis of a magnetic flux detected by the four magnetic sensors 20. The plurality of magnetic pole pairs each have a different strength of magnetization.SELECTED DRAWING: Figure 1

Description

The present invention relates to magnetic encoders.

Encoders are often used as position measurement sensors in fields that require precise positioning, such as semiconductor manufacturing, machine tools, and robots.

Encoders include optical encoders and magnetic encoders.

An optical encoder uses a combination of a light-emitting element and a light-receiving element to detect changes in position. Optical encoders are classified into transmissive and reflective types. The transmissive type has the advantage that it is easy to improve signal accuracy and contamination resistance, but it is characterized by the fact that miniaturization is difficult. The reflective type is characterized by its low contamination resistance, although it can be easily miniaturized.

A magnetic encoder detects changes in the position of an object to be measured as changes in magnetic flux. A magnetic encoder basically has a simple configuration consisting of a permanent magnet and a magnetic sensor. Therefore, the magnetic encoder is characterized by being easy to reduce in size and weight and being inexpensive. Magnetic encoders are also very robust and can be used in harsh environments.

For example, Patent Documents 1 and 2 disclose magnetic encoders that detect the position of a piston rod that moves within a cylinder.

JP-A-7-229760 JP-A-7-260512

Encoders are also used to control motors. Systems that control motors are often required to be downsized. Therefore, the magnetic encoder, which can be easily miniaturized, is suitable for use in motor control.

On the other hand, with the increasing complexity of control technology required for motors, there is a demand for highly accurate measurement of the angle of rotation of the motor.

An object of the present invention is to provide a magnetic encoder capable of measuring angles with high accuracy.

A magnetic encoder according to one embodiment of the present invention comprises:
a rotatable disk-shaped magnet having a plurality of magnetic pole pairs;
four magnetic sensors positioned around the magnet;
a computing device that calculates a mechanical angle, which is the rotation angle of the magnet, based on the magnetic fluxes detected by the four magnetic sensors;
The plurality of magnetic pole pairs have different magnetization strengths for each magnetic pole pair.

Further, in the magnetic encoder according to one embodiment of the present invention,
The four magnetic sensors constitute two pairs of magnetic sensors,
One magnetic sensor pair and the other magnetic sensor pair of the two pairs of magnetic sensors are arranged so as to detect changes in magnetic flux due to the rotation of the magnet with a phase difference of 90 degrees. good too.

Further, in the magnetic encoder according to one embodiment of the present invention,
The computing device is
generating quadrature signals based on the magnetic fluxes detected by the four magnetic sensors;
calculating an electrical angle corresponding to the phase of change in magnetic flux accompanying rotation of the magnet based on the quadrature signal;
calculating an offset for associating the electrical angle with the mechanical angle based on the magnetic fluxes detected by the four magnetic sensors and the electrical angle;
The mechanical angle may be calculated based on the electrical angle and the offset.

ただし、式（１）において、Ｊ_ｎは誤差の２乗和を示し、Ｖ_ｉは磁気センサの出力信号を示し、ｆ_ｉはルックアップテーブルを示し、θ_ｅは電気角を示し、Ｐは磁石における磁極対の数を示し、ｎはオフセットを示す。 Further, in the magnetic encoder according to one embodiment of the present invention,
The arithmetic device may calculate n that minimizes the sum of squares of errors J _n in the following equation (1) as the offset.

However, in equation (1), J _n indicates the sum of squares of errors, V _i indicates the output signal of the magnetic sensor, f _i indicates the lookup table, θ _e indicates the electrical angle, and P is the magnet indicates the number of magnetic pole pairs in , and n indicates the offset.

Further, in the magnetic encoder according to one embodiment of the present invention,
The computing unit may generate the quadrature signals by summing magnetic fluxes detected by two magnetic sensors of each magnetic sensor pair in the two sets of magnetic sensors.

Further, in the magnetic encoder according to one embodiment of the present invention,
The computing device may generate the quadrature signals by performing neural network processing on the magnetic fluxes detected by the four magnetic sensors.

Further, in the magnetic encoder according to one embodiment of the present invention,
The computing device is
determining which of the four sections the quadrature signal corresponds to based on the magnitude relationship between the two quadrature signals and the sign of the quadrature signal;
The electrical angle may be calculated by referring to a lookup table corresponding to the interval and comparing the value of the orthogonal signal having the smaller absolute value among the two orthogonal signals with the lookup table.

ただし、式（２）において、ｅは偏差を示し、θ_ｅは電気角を示し、θ_ｘはルックアップテーブルに格納されている角度を示す。 Further, in the magnetic encoder according to one embodiment of the present invention,
The computing device may calculate the electrical angle by controlling the angle _θx so that the deviation e in the following equation (2) becomes 0 using the PLL method.

However, in equation (2), e indicates the deviation, θ _e indicates the electrical angle, and θ _x indicates the angle stored in the lookup table.

According to the magnetic encoder according to one embodiment of the present invention, angles can be measured with high accuracy.

1 is a block diagram showing a configuration example of a magnetic encoder according to one embodiment of the present invention; FIG. It is a figure which shows an example of a structure of a magnet. FIG. 4 is a diagram showing an example of the arrangement of magnetic sensors around a magnet; It is a figure which shows the functional block of an arithmetic unit. It is a figure which shows the signal of a magnetic sensor. FIG. 4 is a diagram showing quadrature signals; FIG. 4 is a diagram showing functional blocks of a converter; It is a figure explaining the determination of the area by the area determination part. It is a figure which shows the conditions by which an area determination part determines an area. It is a figure which shows the functional block of the arithmetic unit based on a 1st modification. FIG. 10 is a diagram showing functional blocks of a converter according to a first modified example; It is a figure which shows the functional block of the arithmetic unit based on a 2nd modification. FIG. 10 is a diagram showing functional blocks of a quadrature signal generator according to a second modification; It is a figure which shows a simulation result.

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing a configuration example of a magnetic encoder 1 according to one embodiment of the present invention. The magnetic encoder 1 is a device that can detect the rotation angle of the magnet 10 .

The magnetic encoder 1 includes a magnet 10, four magnetic sensors 20-0 to 20-3, and an arithmetic device 30. FIG.

The magnet 10 is a disk-shaped permanent magnet. Magnet 10 may be any type of permanent magnet, but may be, for example, a neodymium magnet.

Magnet 10 is rotatable around center 101 . A shaft that rotates as the motor rotates may be connected to the center 101 of the magnet 10, for example. In this case, when the motor rotates, the magnet 10 rotates with the rotation of the motor. Although FIG. 1 shows the magnet 10 rotating clockwise, the magnet 10 may rotate counterclockwise.

Magnet 10 has a plurality of magnetic pole pairs. The configuration of the magnet 10 will be described with reference to FIG. In the example shown in FIG. 2, magnet 10 has four magnetic poles. The magnet 10 is divided into four sector-shaped regions with alternating north and south poles.

In this embodiment, a combination of one N pole and one S pole is called a magnetic pole pair. In the example shown in FIG. 2, magnet 10 has two pole pairs, pole pair 102 and pole pair 103 . Although FIG. 2 shows the case where the magnet 10 has two magnetic pole pairs, the number of magnetic pole pairs that the magnet 10 has is not limited to this. Magnet 10 may have a plurality of magnetic pole pairs, and magnet 10 may have three or more magnetic pole pairs. In this embodiment, an example in which the magnet 10 has two magnetic pole pairs will be described.

The plurality of magnetic pole pairs of the magnet 10 are magnetized with different intensities for each magnetic pole pair. For example, in magnet 10 shown in FIG. 2, pole pair 102 may be less magnetized than pole pair 103 . For example, the maximum energy product of pole pair 102 may be 70% of the maximum energy product of pole pair 103 .

Returning to FIG. 1 again, description will be made. Magnetic sensors 20 - 0 to 20 - 3 are arranged around magnet 10 . The arrangement of the magnetic sensors 20-0 to 20-3 will be described with reference to FIG.

The angle θ shown in FIG. 3 is a clockwise angle with respect to a line drawn upward from the center 101 of the magnet 10 . The magnetic sensor 20-0 is arranged at a position where θ=0 degrees. The magnetic sensor 20-1 is arranged at a position of θ=180 degrees. The magnetic sensor 20-2 is arranged at a position of θ=-45 degrees. The magnetic sensor 20-3 is arranged at a position of θ=135 degrees.

Further, the magnetic sensors 20-0 to 20-3 are arranged at positions equidistant from the center 101 of the magnet 10. FIG.

When there is no particular need to distinguish between the magnetic sensors 20-0 to 20-3, they may be collectively referred to as the magnetic sensor 20 hereinafter.

Magnetic sensor 20 detects the magnetic flux generated by magnet 10 . The magnetic sensor 20 outputs a voltage signal according to the detected magnetic flux. The magnetic sensor 20 may be any sensor capable of detecting magnetic flux, and may be, for example, a Hall element.

ここで、Ｋは磁気センサ２０に固有の値である。磁気センサ２０－０～２０－３においては、Ｋは全て同じ値であってよい。 When the magnetic sensor 20 is a Hall element, the magnetic sensor 20 outputs a voltage _VH according to the magnetic flux density B of the magnetic flux penetrating the magnetic sensor 20 . Voltage _VH is represented by the following equation (3).

Here, K is a value specific to the magnetic sensor 20 . In magnetic sensors 20-0 to 20-3, K may all have the same value.

When the magnet 10 rotates, the magnetic flux detected by the magnetic sensor 20 changes. Since the magnet 10 has two magnetic pole pairs, when the magnet 10 rotates 180 degrees, the phase of the magnetic flux detected by the magnetic sensor 20 changes by 360 degrees. More generally, when the number of magnetic pole pairs that the magnet 10 has is P, when the magnet rotates (360/P) degrees, the phase of the magnetic flux detected by the magnetic sensor 20 changes by 360 degrees.

In the example shown in FIG. 3, the magnetic sensor 20-0 and the magnetic sensor 20-2 are arranged at positions shifted by 45 degrees. When the magnet 10 rotates by 45 degrees, the phase of the magnetic flux detected by the magnetic sensor 20 changes by 90 degrees. Therefore, the magnetic sensors 20-0 and 20-2 detect phase shifts of 90 degrees as changes in the magnetic flux. .

When the magnet 10 rotates clockwise, the magnetic sensor 20-2 detects a phase that leads the magnetic sensor 20-0 by 90 degrees. Therefore, the voltage signal output by the magnetic sensor 20-2 according to the change in the magnetic flux detected by the rotation of the magnet 10 can be expressed as V _cos+ . Also, the voltage signal output by the magnetic sensor 20-0 according to the change in the magnetic flux detected by the rotation of the magnet 10 can be expressed as V _sin+ .

As shown in FIG. 3, the magnetic sensor 20-1 is placed symmetrically with the magnetic sensor 20-0 with respect to the center 101 of the magnet . Therefore, the voltage signal output by the magnetic sensor 20-1 according to the change in the magnetic flux detected by the rotation of the magnet 10 can be expressed as V _sin- . Further, the magnetic sensor 20-3 is arranged at a position symmetrical to the magnetic sensor 20-2 with respect to the center 101 of the magnet 10. FIG. Therefore, the voltage signal output by the magnetic sensor 20-3 according to the change in the magnetic flux detected by the rotation of the magnet 10 can be expressed as V _cos- .

Thus, the four magnetic sensors 20-1 to 20-3 constitute two magnetic sensor 20 pairs. One pair of magnetic sensors 20 is a pair that combines the magnetic sensor 20-0 and the magnetic sensor 20-1. The other pair of magnetic sensors 20 is a pair that combines the magnetic sensors 20-2 and 20-3.

The voltage signal V sin is obtained by adding the voltage signal V _sin+ output _by the magnetic sensor 20-0 and the voltage signal V _sin- output by the magnetic sensor 20-1. The voltage signal V cos is obtained by adding the voltage signal V _cos+ output _by the magnetic sensor 20-2 and the voltage signal V _cos− output by the magnetic sensor 20-3.

Voltage signal V _sin and voltage signal V _cos are quadrature signals that are 90 degrees out of phase. That is, the pair of magnetic sensors 20-0 and 20-1, which is one pair of magnetic sensors 20 of the two pairs of magnetic sensors 20, and the pair of magnetic sensors 20-2, which is the other pair of magnetic sensors 20 The pair of magnetic sensors 20-3 are arranged so as to detect changes in magnetic flux accompanying rotation of the magnet 10 with a 90 degree phase shift.

Returning to FIG. 1 again, description will be made. Arithmetic device 30 includes control unit 31 and storage unit 32 .

Control unit 31 includes at least one processor, at least one dedicated circuit, or a combination thereof. The processor is a general-purpose processor such as a CPU (Central Processing Unit) or GPU (Graphics Processing Unit), or a dedicated processor specialized for specific processing. The dedicated circuit is, for example, an FPGA (Field-Programmable Gate Array) or an ASIC (Application Specific Integrated Circuit). The control unit 31 executes processing related to the operation of the arithmetic device 30 while controlling each part of the arithmetic device 30 .

The storage unit 32 is, for example, a semiconductor memory, a magnetic memory, an optical memory, or the like, but is not limited to these. The storage unit 32 may function, for example, as a main storage device, an auxiliary storage device, or a cache memory. The storage unit 32 stores arbitrary information used for the operation of the arithmetic device 30 . For example, the storage unit 32 may store system programs, application programs, various types of information, and the like. A part of the storage unit 32 may be installed outside the arithmetic device 30 . In that case, part of the storage unit 32 installed outside may be connected to the arithmetic device 30 via an arbitrary interface.

The computing device 30 is connected to four magnetic sensors 20-0 to 20-3. The computing device 30 acquires a voltage signal output by the magnetic sensor 20 according to the detected magnetic flux. Arithmetic device 30 calculates a mechanical angle, which is the rotation angle of magnet 10, based on the magnetic fluxes detected by four magnetic sensors 20-0 to 20-3. Here, the expression "based on the magnetic fluxes detected by the four magnetic sensors 20-0 to 20-3" means "4 magnetic fluxes detected by the four magnetic sensors 20-0 to 20-3. Based on the voltage signals output by the sensors 20-0 to 20-3".

A mechanical angle will be explained. The mechanical angle is the actual rotation angle when the magnet 10 rotates about the center 101 as an axis. For example, the mechanical angle when the magnet 10 rotates 1/4 turn is 90 degrees, the mechanical angle when it rotates 1/2 turn is 180 degrees, and the mechanical angle when it rotates 1 turn is 360 degrees. . Since the mechanical angle represents the absolute angle of rotation of the magnet 10, it is sometimes referred to as the "absolute angle."

On the other hand, there is also an angle called an electrical angle as an angle related to the rotation of the magnet 10 . The magnetic flux detected by the magnetic sensor 20 changes as the magnet 10 rotates, and the angle corresponding to the phase of the change in the magnetic flux is the electrical angle. In the case of the magnet 10 shown in FIG. 1, since the magnet 10 has two magnetic pole pairs, the electrical angle is 180 degrees when the magnet 10 rotates 1/4 turn, and the electrical angle when it rotates 1/2 turn is 360 degrees.

ただし、式（４）において、θｍは機械角を示し、θｅは電気角を示し、Ｐは磁極対の数を示し、ｎはオフセットを示す。 The mechanical angle and the electrical angle are associated as shown in the following formula (4).

However, in equation (4), θm indicates a mechanical angle, θe indicates an electrical angle, P indicates the number of magnetic pole pairs, and n indicates an offset.

The offset n in equation (4) is an integer ranging from 0 to P-1. For example, in the example shown in FIG. 1, the number of magnetic pole pairs is P=2. In this case n is 0 or 1. Unless the offset n is determined, the electrical angle θ _e cannot be uniquely associated with the mechanical angle θ _m . However, once the offset n is determined, the electrical angle θ _e can be uniquely associated with the mechanical angle θ _m .

For example, when the number of magnetic pole pairs is P=2 and the electrical angle θ _e is 90 degrees, the mechanical angle θ _m is 45 degrees or 225 degrees. When the offset n is 0, the mechanical angle _θm is 45 degrees, and when the offset n is 1, the mechanical angle _θm is 225 degrees. If it is possible to specify whether the offset n is 0 or 1, the mechanical angle _θm can be specified.

Calculation of the mechanical angle by the computing device 30 will be described with reference to FIG. Arithmetic device 30 includes a quadrature signal generator 310, a converter 320, an offset calculator 330, and a mechanical angle calculator 340 as functional blocks.

The functions of the quadrature signal generator 310, the converter 320, the offset calculator 330, and the mechanical angle calculator 340 may be executed by the controller 31 shown in FIG.

The computing device 30 acquires the voltage signals output by the magnetic sensors 20-0 to 20-3. Hereinafter, the voltage signal output by the magnetic sensor 20-0 is V _sin+ , the voltage signal output by the magnetic sensor 20-1 is V _sin− , the voltage signal output by the magnetic sensor 20-2 is V _cos+ , and the voltage signal output by the magnetic sensor 20-3 is V sin− . will be described as V _cos− .

An example of V _sin+ , V _sin− , V _cos+ and V _cos− is shown in FIG. 5A. The horizontal axis of the graph shown in FIG. 5A is the mechanical angle of magnet 10 . The vertical axis of the graph shown in FIG. 5A is the amplitude of V _sin+ , V _sin−, V _cos+ and V _cos− .

As shown in FIG. 5A, V _sin+ , V _sin− , V _cos+ and V _cos− are all waveforms with non-constant amplitudes. This is because the plurality of magnetic pole pairs of the magnet 10 have different magnetization strengths for each magnetic pole pair. This is because the amplitude becomes smaller when

The quadrature signal generator 310 adds the voltage signal V _sin+ and the voltage signal V _sin− to generate the voltage signal V _sin , as in Equation (5) below. Also, the voltage signal V _cos+ and the voltage signal V _cos− are added to generate the voltage signal V _cos , as in Equation (6) below.

An example of V _sin and V _cos is shown in FIG. 5B. The horizontal axis of the graph shown in FIG. 5B is the mechanical angle of magnet 10 . The vertical axis of the graph shown in FIG. 5B is the amplitude of V _sin and V _cos .

As shown in FIG. 5B, voltage signal V _sin and voltage signal V _cos are orthogonal signals. That is, the voltage signal V _sin and the voltage signal V _cos are quadrature signals.

The converter 320 acquires the voltage signal V _sin and the voltage signal V _cos , which are quadrature signals, from the quadrature signal generator 310, and calculates the electrical angle _θe .

Calculation of electrical angle θ _e by converter 320 will be described with reference to FIGS.

As shown in FIG. 6, the converter 320 includes an interval determination section 321 and a lookup table reference section 322 as functional blocks. The functions of the section determination section 321 and the lookup table reference section 322 may be executed by the control section 31 shown in FIG.

A voltage signal formed by combining a voltage signal having a smaller absolute value between the voltage signal V _sin and the voltage signal V _cos is referred to as a "linear section extraction signal". A graph V indicated by a thick line in FIG. 7 is a linear section extraction signal.

The electrical angle section is divided into four sections from section I to section IV, as shown in FIG. Sections I to IV correspond to electrical angles as follows.
Interval I: 0 to 45 degrees and 315 to 360 degrees in electrical angle Interval II: 45 to 135 degrees in electrical angle Interval III: 135 to 225 degrees in electrical angle Interval IV: 225 to 315 degrees in electrical angle

The correspondence relationship between the electrical angle and the linear section extraction signal is stored as a lookup table for each section. This lookup table may be stored in advance in the storage unit 32 shown in FIG.

Based on the obtained voltage signal V _sin and voltage signal V _cos , the section determination unit 321 determines to which section of the electrical angle the obtained voltage signal V _sin and voltage signal V _cos correspond.

The section determination unit 321 determines a section based on conditions 1 and 2 shown in FIG. Condition 1 is based on the magnitude relationship between the absolute value of the voltage signal V _sin and the absolute value of the voltage signal V _cos . Condition 2 is based on whether the voltage signal V _sin or the voltage signal V _cos is positive or negative.

As shown in FIG. 8, the section determination unit 321 determines that the section is in section I when the absolute value of the voltage signal V _sin is smaller than the absolute value of the voltage signal V _cos and the voltage signal V _cos is positive. Further, when the absolute value of the voltage signal V _sin is greater than the absolute value of the voltage signal V _cos and the voltage signal V _sin is positive, the section determination unit 321 determines that it is section II. Further, when the absolute value of the voltage signal V _sin is smaller than the absolute value of the voltage signal V _cos and the voltage signal V _cos is negative, the section determination unit 321 determines that it is the section III. Further, when the absolute value of the voltage signal V _sin is larger than the absolute value of the voltage signal V _cos and the voltage signal V _sin is negative, the section determination unit 321 determines that it is the section IV.

The lookup table reference unit 322 refers to the lookup table corresponding to the section acquired from the section determination unit 321, and calculates the electrical angle _θe . For example, if the section acquired from the section determination unit 321 is section II, the lookup table reference unit 322 refers to the lookup table for section II and calculates the electrical angle θ _e based on the value of V _cos .

The offset calculator 330 acquires the voltage signal V _sin+ , the voltage signal V _sin− , the voltage signal V _cos+ and the voltage signal V _cos− from the magnetic sensors 20-0 to 20-3. Offset calculation unit 330 also obtains electrical angle θ _e from converter 320 .

The offset calculator 330 calculates the offset n based on the voltage signal V _sin+ , the voltage signal V _sin− , the voltage signal V _cos+ and the voltage signal V _cos− , and the electrical angle θ _e .

式（７）において、Ｊ_ｎは誤差の２乗和を示す。また、式（７）において、Ｖ_ｉは磁気センサ２０－０～２０－３が出力する電圧信号を示す。ここで、ｉは０～３の整数であり、Ｖ_０はＶ_ｓｉｎ＋を示し、Ｖ_１はＶ_ｓｉｎ－を示し、Ｖ_２はＶ_ｃｏｓ＋を示し、Ｖ_３はＶ_ｃｏｓ－を示す。また、ｆ_ｉは記憶部３２に格納されているルックアップテーブルを示す。ｆ_０はＶ_ｓｉｎ＋と機械角θｍとの関係を対応づけたルックアップテーブルを示し、ｆ_１はＶ_ｓｉｎ－と機械角θｍとの関係を対応づけたルックアップテーブルを示し、ｆ_２はＶ_ｃｏｓ＋と機械角θｍとの関係を対応づけたルックアップテーブルを示し、ｆ_３はＶ_ｃｏｓ－と機械角θｍとの関係を対応づけたルックアップテーブルを示す。Ｐは磁石における磁極対の数を示す。ｎはオフセットを示す。 The offset calculator 330 calculates n that minimizes the sum of squares of errors _Jn in the following equation (7) as the offset n.

In equation (7), _Jn denotes the sum of squares of errors. In equation (7), V _i represents voltage signals output from the magnetic sensors 20-0 to 20-3. Here, i is an integer from 0 to 3, V ₀ indicates V _sin+ , V ₁ indicates V _sin− , V ₂ indicates V _cos+ , and V ₃ indicates V _cos− . Also, f _i indicates a lookup table stored in the storage unit 32 . f ₀ indicates a lookup table that associates the relationship between V _sin+ and mechanical angle θm, f ₁ indicates a lookup table that associates the relationship between V _sin− and mechanical angle θm, and f ₂ indicates V _cos+. and mechanical angle θm, and f ₃ is a lookup table that associates the relationship between V _cos− and mechanical angle θm. P indicates the number of magnetic pole pairs in the magnet. n indicates an offset.

The offset calculator 330 calculates _Jn based on the above equation (7) for all n, and selects the n that minimizes _Jn as the offset n.

Mechanical angle calculator 340 calculates mechanical angle θ _m based on electrical angle θ _e obtained from converter 320 and offset n obtained from offset calculator 330 . The mechanical angle calculator 340 calculates the mechanical angle _θm by substituting the electrical angle _θe and the offset n into the above equation (4). In the example shown in FIG. 1, the number P of magnetic pole pairs is two.

Thus, in the magnetic encoder 1 according to this embodiment, the magnet 10 has a plurality of magnetic pole pairs. Therefore, the magnetic encoder 1 can detect with high accuracy the electrical angle corresponding to the phase of the change in magnetic flux accompanying the rotation of the magnet 10 . In addition, the magnetic pole pairs of the magnet 10 have different magnetization strengths for each magnetic pole pair. Therefore, the magnetic sensor 20 detects magnetic flux densities with different amplitudes when a strongly magnetized magnetic pole pair passes nearby and when a weakly magnetized magnetic pole pair passes nearby. Therefore, the arithmetic device 30 can calculate the offset n. Then, the computing device 30 can calculate the mechanical angle _θm , which is the absolute rotation angle of the magnet 10, based on the electrical angle _θe and the offset n. Therefore, the magnetic encoder 1 can measure the angle of the magnet 10 with high accuracy.

(First modification)
FIG. 9 is a diagram showing functional blocks of an arithmetic device 30a according to the first modification. The arithmetic device 30a includes a quadrature signal generator 310, a converter 320a, an offset calculator 330, and a mechanical angle calculator 340 as functional blocks.

The functions of the quadrature signal generator 310, the converter 320a, the offset calculator 330, and the mechanical angle calculator 340 may be executed by the controller 31 shown in FIG.

Arithmetic device 30a according to the first modification differs from arithmetic device 30 shown in FIG. In the description of the arithmetic device 30a according to the first modification, mainly the points that are different from the arithmetic device 30 shown in FIG. Description is omitted.

The converter 320a acquires the voltage signal V _sin and the voltage signal V _cos , which are quadrature signals, from the quadrature signal generator 310, and calculates the electrical angle _θe using the PLL method. Converter 320a can establish robustness against distortion of the quadrature signals by using the PLL method.

FIG. 10 shows how converter 320a performs the PLL method.

In FIG. 10, “f _sin ” indicates a lookup table that associates the voltage signal V _sin with the electrical angle θ _e . “f _cos ” indicates a lookup table that associates the voltage signal V _cos with the electrical angle θ _e . These lookup tables can directly compensate for the distortion of quadrature signals. These lookup tables may be stored in advance in the storage unit 32 shown in FIG.

In FIG. 10, "e" indicates deviation. f _sin and f _cos can be approximated by sine and cosine waves, respectively, provided that the harmonic components are sufficiently smaller than the fundamental component. At this time, the deviation e can be expressed by the following equation (8).

Converter 320a controls angle θ _x so that deviation e in equation (8) becomes zero. When the deviation e approaches 0, the electrical angle θ _e and the angle θ _x become substantially equal. Thus, converter 320a can calculate electrical angle _θe .

(Second modification)
FIG. 11 is a diagram showing functional blocks of an arithmetic device 30b according to the second modification. The arithmetic device 30b includes a quadrature signal generator 310b, a converter 320, an offset calculator 330, and a mechanical angle calculator 340 as functional blocks.

The functions of the quadrature signal generator 310b, the converter 320, the offset calculator 330, and the mechanical angle calculator 340 may be executed by the controller 31 shown in FIG.

Arithmetic device 30b according to the second modification differs from arithmetic device 30 shown in FIG. In the description of the arithmetic device 30b according to the second modification, mainly the points that are different from the arithmetic device 30 shown in FIG. Description is omitted.

The quadrature signal generator 310b acquires the voltage signals V _sin+ , V _sin− , V cos ₊ and V _cos− from the magnetic sensors 20-0 to 20-3. The quadrature signal generator 310b performs neural network processing on the voltage signals V _sin+ , V _sin− , V _cos+ and V _cos− to generate quadrature signals V _sin and V _cos .

The quadrature signal generator 310b can correct the distortion of the quadrature signal by executing processing using a neural network. For example, if there is variation in the detection sensitivity of the magnetic sensors 20-0 to 20-3, simply add V _sin+ and V _sin- to generate V _sin , and add V _cos+ and V _cos- . to generate V _cos may distort the quadrature signal. The quadrature signal generation unit 310b can correct the distortion of the quadrature signal even when the detection sensitivity of the magnetic sensors 20-0 to 20-3 varies, and generates a quadrature signal without distortion. can do.

As shown in FIG. 12, the quadrature signal generation section 310b includes an LSTM (Long Short Term Memory) section 311, an affine section 312, and a mean square error section 313 as functional blocks. The functions of the LSTM unit 311, the affine unit 312, and the mean squared error unit 313 may be performed by the control unit 31 shown in FIG.

The LSTM unit 311 is a neural network that is an improved recurrent neural network (RNN). LSTM section 311 acquires time-series information of V _sin+ , V _sin− , V _cos+ , and V _cos− . LSTM unit 311 operates to adjust the parameters to reduce the mean squared error.

Affine part 312 is a fully connected layer in which all neurons are connected.

Mean squared error unit 313 calculates the mean squared error between the output of the neural network and the orthogonal signal.

(simulation result)
13A and 13B are diagrams showing results of a simulation performed by the calculation device 30 of the magnetic encoder 1 to calculate the mechanical angle. The horizontal axis is the actual mechanical angle of magnet 10 . The vertical axis on the left side is the mechanical angle _θm calculated by the calculation device 30 in the simulation. The vertical axis on the right side is the offset n calculated by the calculation device 30 in the simulation. The simulation shown in FIG. 13 is the result of executing the simulation with the number of magnetic pole pairs included in the magnet 10 being two.

As shown in FIG. 13, the computing device 30 correctly calculates the offset n as 0 in the range of the actual mechanical angle from 0 to 180 degrees. Further, the calculation device 30 correctly calculates the offset n as 1 in the range of the actual mechanical angle of 180 to 360 degrees.

Further, as shown in FIG. 13, the computing device 30 calculates a mechanical angle that is approximately equal to the actual mechanical angle.

Thus, according to this embodiment, the magnetic encoder 1 includes a rotatable disc-shaped magnet 10 having a plurality of magnetic pole pairs, four magnetic sensors 20 arranged around the magnet 10, and four and an arithmetic device 30 for calculating a mechanical angle, which is the rotation angle of the magnet 10, based on the magnetic flux detected by the magnetic sensor 20. In addition, the magnetic pole pairs of the magnet 10 have different magnetization strengths for each magnetic pole pair. Thus, since the magnet 10 has a plurality of magnetic pole pairs, the magnetic encoder 1 can detect the electrical angle with high accuracy. In addition, since the plurality of magnetic pole pairs have different magnetization strengths for each magnetic pole pair, the magnetic sensor 20 detects when a strongly magnetized magnetic pole pair passes nearby and when a weakly magnetized magnetic pole pair passes nearby. Detect magnetic flux densities with different amplitudes depending on the time. Therefore, the arithmetic device 30 can calculate the offset n. Then, the computing device 30 can calculate the mechanical angle _θm of the magnet 10 based on the electrical angle _θe and the offset n. Therefore, the magnetic encoder 1 can measure the angle of the magnet 10 with high accuracy.

Although the present invention has been described with reference to the drawings and examples, it should be noted that various variations and modifications of the present invention will readily occur to those skilled in the art. Therefore, it should be noted that these variations and modifications are included within the scope of the present invention. For example, the functions included in each component and each step can be rearranged so as not to be logically inconsistent, and multiple components or steps can be combined into one or divided. is. In addition, although the present invention has been mainly described with respect to the apparatus, the present invention can be applied to a method including steps executed by each component of the apparatus, a method executed by a processor provided in the apparatus, a program, or a storage medium recording the program. can also be realized. It should be understood that the scope of the present invention includes these.

For example, in the present embodiment, the magnetic encoder 1 includes four magnetic sensors 20, but the number of magnetic sensors 20 included in the magnetic encoder 1 is not limited to four. The magnetic encoder 1 may have five or more magnetic sensors 20 .

Further, in the present embodiment, the magnetic encoder 1 calculates the mechanical angle of the magnet 10, but the magnetic encoder 1 associates the calculated mechanical angle with the time at which the mechanical angle was measured, The angular velocity of rotation of magnet 10 may also be calculated.

1 magnetic encoder 10 magnet 20 magnetic sensor 30, 30a, 30b arithmetic unit 31 control unit 32 storage unit 101 center 102 magnetic pole pair 103 magnetic pole pair 310, 310b orthogonal signal generation unit 311 LSTM unit 312 affine unit 313 mean square error unit 320, 320a converter 321 section determination unit 322 lookup table reference unit 330 offset calculation unit 340 mechanical angle calculation unit

Claims (8)

a rotatable disk-shaped magnet having a plurality of magnetic pole pairs;
four magnetic sensors positioned around the magnet;
a computing device that calculates a mechanical angle, which is the rotation angle of the magnet, based on the magnetic fluxes detected by the four magnetic sensors;
A magnetic encoder, wherein the plurality of magnetic pole pairs have different magnetization strengths for each magnetic pole pair. In the magnetic encoder according to claim 1,
The four magnetic sensors constitute two pairs of magnetic sensors,
One magnetic sensor pair and the other magnetic sensor pair of the two pairs of magnetic sensors are arranged so as to detect changes in magnetic flux due to the rotation of the magnet with a phase difference of 90 degrees. , magnetic encoder. In the magnetic encoder according to claim 2,
The computing device is
generating quadrature signals based on the magnetic fluxes detected by the four magnetic sensors;
calculating an electrical angle corresponding to the phase of change in magnetic flux accompanying rotation of the magnet based on the quadrature signal;
calculating an offset for associating the electrical angle with the mechanical angle based on the magnetic fluxes detected by the four magnetic sensors and the electrical angle;
A magnetic encoder that calculates the mechanical angle based on the electrical angle and the offset. In the magnetic encoder according to claim 3,
The magnetic encoder, wherein the arithmetic unit calculates n that minimizes the sum of squares of errors J _n in the following equation (1) as the offset.

However, in equation (1), J _n indicates the sum of squares of errors, V _i indicates the output signal of the magnetic sensor, f _i indicates the lookup table, θ _e indicates the electrical angle, and P is the magnet indicates the number of magnetic pole pairs in , and n indicates the offset. In the magnetic encoder according to claim 3 or 4,
A magnetic encoder, wherein the arithmetic unit generates the quadrature signals by adding the magnetic fluxes detected by the two magnetic sensors of each magnetic sensor pair in the two sets of magnetic sensors. In the magnetic encoder according to claim 3 or 4,
A magnetic encoder, wherein the computing device performs neural network processing on the magnetic fluxes detected by the four magnetic sensors to generate the quadrature signals. In the magnetic encoder according to any one of claims 3 to 6,
The computing device is
determining which of the four sections the quadrature signal corresponds to based on the magnitude relationship between the two quadrature signals and the sign of the quadrature signal;
A magnetic encoder that calculates the electrical angle by referring to a lookup table corresponding to the section and comparing the value of the quadrature signal having a smaller absolute value among the two quadrature signals with the lookup table. In the magnetic encoder according to any one of claims 3 to 6,
The arithmetic device uses the PLL method to control the angle _θx so that the deviation e in the following equation (2) becomes 0, thereby calculating the electrical angle.

However, in equation (2), e indicates the deviation, θ _e indicates the electrical angle, and θ _x indicates the angle stored in the lookup table.

Images