Angle detection method and angle detection device

Abstract

The present application provides angle detection method and angle detection device, including: three magnetic sensors, detecting the magnetic flux change caused by the rotation of the rotating shaft, and signal processing unit, processing the signal output from the three magnetic sensors. The signal processing unit acquires the sensor signal output from the three magnetic sensors, selects the intersection of two sensor signals in the three sensor signals, and the zero crossing point of the three sensor signals and the reference signal level, generates a first function θ (Δx) representing the straight line connecting the adjacent intersection points and the zero crossing points, calculates the deviation of the mechanical angle θ calculated based on the first function θ (Δx) and the mechanical angle θe obtained from the encoder arranged on the rotating shaft as the first angle error for multiple points on the straight line, and generates the first angle error function used to calculate the first angle error corresponding to any point on the straight line based on the first angle error calculated for multiple points on the straight line.

Description

Technical Field

[0001] This invention relates to an angle detection method and an angle detection device. Background Technology

[0002] Previously, structures including optical encoders and resolvers, which are known to accurately control the rotational position of motors, were used. However, absolute angle position sensors are large and expensive. Therefore, Patent Document 1 discloses a position estimation method that does not use absolute angle position sensors, but instead uses three inexpensive and small magnetic sensors to estimate the rotational position of the motor.

[0003] Existing technical documents

[0004] Patent documents

[0005] Patent Document 1: Japanese Patent No. 6233532 Summary of the Invention

[0006] The problem that the invention aims to solve

[0007] In the position estimation method described in Patent Document 1, the mechanical angle of the rotating shaft can be estimated with high accuracy using three inexpensive and small magnetic sensors, but sometimes the market demands higher accuracy.

[0008] Technical means to solve the problem

[0009] One embodiment of the angle detection method of the present invention is an angle detection method that detects the mechanical angle of a rotating shaft, and includes: a step of acquiring signals output from three magnetic sensors that detect the change in magnetic flux caused by the rotation of the rotating shaft as sensor signals, the three sensor signals having a phase difference of 120° between them; a step of selecting, within one cycle of the mechanical angle, the intersection point where two of the three sensor signals intersect each other, and the zero-crossing point where each of the three sensor signals intersects with a reference signal level; and a step of generating a linear function θ(Δx) representing a straight line connecting adjacent intersection points to the zero-crossing point. Δx is the length from the starting point of the straight line to any point on the straight line, and θ is the mechanical angle corresponding to any point on the straight line; for multiple points on the straight line, the step of calculating the deviation between the mechanical angle θ calculated based on the linear function θ(Δx) and the mechanical angle θe obtained from the encoder set on the rotating shaft as a first angular error; the step of storing the first angular error calculated for multiple points on the straight line as a learning value; and the step of generating a first angular error function based on the first angular error calculated for multiple points on the straight line to calculate the first angular error corresponding to any point on the straight line.

[0010] One embodiment of the angle detection device of the present invention is an angle detection device that detects the mechanical angle of a rotating shaft, and includes: three magnetic sensors for detecting changes in magnetic flux caused by the rotation of the rotating shaft; and a signal processing unit for processing the signals output from the three magnetic sensors. The signal processing unit performs the following processing: acquiring the signals output from the three magnetic sensors as sensor signals, the three sensor signals having a phase difference of 120° between them; selecting, within one cycle of the mechanical angle, the intersection point where two of the three sensor signals intersect each other, and the zero-crossing point where each of the three sensor signals intersects with a reference signal level; generating a representation of the intersection points where the three adjacent signals intersect each other. A linear function θ(Δx) is defined on the straight line connecting the point to the zero intersection point, where Δx is the length from the starting point of the straight line to any point on the straight line, and θ is the mechanical angle corresponding to any point on the straight line. For multiple points on the straight line, the deviation between the mechanical angle θ calculated based on the linear function θ(Δx) and the mechanical angle θe obtained from the encoder installed on the rotating shaft is calculated as a first angle error. The first angle error calculated for multiple points on the straight line is stored as a learning value. Based on the first angle error calculated for multiple points on the straight line, a first angle error function is generated to calculate the first angle error corresponding to any point on the straight line.

[0011] The effects of the invention

[0012] According to the embodiments of the present invention, an angle detection method and an angle detection device are provided that can improve the estimation accuracy (detection accuracy) of the mechanical angle of a rotating shaft. Attached Figure Description

[0013] Figure 1 This is a block diagram schematically illustrating the structure of an angle detection device according to one embodiment of the present invention.

[0014] Figure 2 This is a diagram showing an example of the waveforms of the U-phase sensor signal Hu, the V-phase sensor signal Hv, and the W-phase sensor signal Hw.

[0015] Figure 3 yes Figure 2 An enlarged view of the U-phase sensor signal Hu, V-phase sensor signal Hv, and W-phase sensor signal Hw contained in a pole pair region.

[0016] Figure 4 This is a diagram showing an example of the waveforms of sensor signals Hu, Hv, and Hw, which contain in-phase signals as noise components.

[0017] Figure 5This is a diagram showing an example of the waveforms of sensor signals Hiu0, Hiv0, and Hiw0 obtained after performing the first correction process.

[0018] Figure 6 This is a diagram showing an example of the waveforms of sensor signals Hiu1, Hiv1, and Hiw1 obtained after performing the second correction process.

[0019] Figure 7 This is a diagram showing an example of the waveforms of sensor signals Hiu2, Hiv2, and Hiw2 obtained after performing the third correction process.

[0020] Figure 8 This is a graph showing the relationship between the waveforms of sensor signals Hiu1, Hiv1, and Hiw1 and the angle error before the third correction process.

[0021] Figure 9 This is a graph showing the relationship between the waveforms of sensor signals Hiu2, Hiv2, and Hiw2 obtained after the third correction process and the angle error.

[0022] Figure 10 This is a graph showing the results obtained from evaluating the encoder error relative to the master lead angle using the basic patented method in a real-machine verification.

[0023] Figure 11 This is a flowchart illustrating the learning process performed by the processing unit 21 of the angle detection device 1 in this embodiment in an offline processing manner.

[0024] Figure 12 This diagram illustrates a method for calculating the deviation between the estimated mechanical angle θ and the true mechanical angle θe for multiple points on a segment, using this deviation as the first angular error.

[0025] Figure 13 This is a diagram illustrating an example of the first angular error calculated for multiple points across 12 segments contained in a polar pair region.

[0026] Figure 14 This is a graph showing the results of evaluating the encoder error relative to the master lead angle, using only the first angle error correction to correct the estimated mechanical angle θ.

[0027] Figure 15 This is a graph showing the results of evaluating the encoder error relative to the master lead angle when correcting the estimated mechanical angle θ using both the first and second angle errors.

[0028] [Explanation of Symbols]

[0029] 1: Angle detection device

[0030] 10: Sensor group

[0031] 11, 12, 13: Magnetic sensors

[0032] 20: Signal Processing Department

[0033] 21: Processing Department

[0034] 22: Storage Department

[0035] 100: Motor

[0036] 110: Rotor shaft

[0037] 120: Sensing magnet

[0038] 200: Encoder Detailed Implementation

[0039] Hereinafter, an embodiment of the present invention will be described in detail with reference to the accompanying drawings.

[0040] Figure 1 This is a block diagram schematically illustrating the structure of an angle detection device 1 according to one embodiment of the present invention. Figure 1 As shown, the angle detection device 1 is a device for detecting the mechanical angle (rotation angle) of the rotor shaft 110, which serves as the rotation axis of the motor 100. In this embodiment, the motor 100 is, for example, an internal rotor type three-phase brushless direct current (DC) motor. The motor 100 includes a rotor shaft 110 and a sensing magnet 120.

[0041] The sensing magnet 120 is a circular plate-shaped magnet mounted on the rotor shaft 110. The sensing magnet 120 rotates synchronously with the rotor shaft 110. The sensing magnet 120 has P pole pairs (P being an integer greater than or equal to 1). In this embodiment, as an example, the sensing magnet 120 has four pole pairs. Furthermore, a pole pair refers to an N pole and a S pole pair. That is, in this embodiment, the sensing magnet 120 has four pairs of N poles and S poles, for a total of eight poles.

[0042] The angle detection device 1 includes a sensor group 10 and a signal processing unit 20. Figure 1 The diagram is omitted, but a circuit board is mounted on the motor 100, and the sensor group 10 and signal processing unit 20 are disposed on the circuit board. The sensing magnet 120 is disposed in a position that does not interfere with the circuit board. The sensing magnet 120 may be disposed inside the housing of the motor 100, or it may be disposed outside the housing.

[0043] Sensor group 10 includes three magnetic sensors: 11, 12, and 13. Magnetic sensors 11, 12, and 13 are arranged on a circuit board facing the sensing magnet 120 and at predetermined intervals along the rotation direction of the sensing magnet 120. In this embodiment, magnetic sensors 11, 12, and 13 are arranged at 30° intervals along the rotation direction of the sensing magnet 120. Magnetic sensors 11, 12, and 13 are analog output type magnetic sensors including magnetoresistive elements, such as Hall elements or linear Hall integrated circuits (ICs).

[0044] If the rotor shaft 110 rotates, the sensing magnet 120 rotates synchronously with the rotor shaft 110. The three magnetic sensors 11, 12 and 13 respectively detect the change in magnetic flux caused by the rotation of the rotor shaft 110, i.e. the rotation of the sensing magnet 120, and output an analog signal representing the detection result of the change in magnetic flux to the signal processing unit 20.

[0045] The electrical angle period of each analog signal output from magnetic sensors 11, 12, and 13 is equivalent to 1 / P of the mechanical angle period. In this embodiment, the number of pole pairs P of the sensing magnet 120 is "4", therefore, the electrical angle period of each analog signal is equivalent to 1 / 4 of the mechanical angle period, which is equivalent to 90° in mechanical angle terms. Furthermore, the analog signals output from magnetic sensors 11, 12, and 13 have a phase difference of 120° between them.

[0046] Hereinafter, the analog signals output from the three magnetic sensors 11, 12, and 13 to the signal processing unit 20 will be referred to as sensor signals. Furthermore, in the following description, the sensor signal output from the magnetic sensor 11 will sometimes be referred to as the U-phase sensor signal Hu, the sensor signal output from the magnetic sensor 12 will be referred to as the V-phase sensor signal Hv, and the sensor signal output from the magnetic sensor 13 will be referred to as the W-phase sensor signal Hw.

[0047] The signal processing unit 20 is a signal processing circuit that processes the sensor signals output from the three magnetic sensors 11, 12, and 13. Based on the U-phase sensor signal Hu output from magnetic sensor 11, the V-phase sensor signal Hv output from magnetic sensor 12, and the W-phase sensor signal Hw output from magnetic sensor 13, the signal processing unit 20 estimates the mechanical angle of the rotor shaft 110, which serves as the rotation axis. The signal processing unit 20 includes a processing unit 21 and a storage unit 22.

[0048] The processing unit 21 is, for example, a microprocessor such as a microcontroller unit (MCU). The U-phase sensor signal Hu output from the magnetic sensor 11, the V-phase sensor signal Hv output from the magnetic sensor 12, and the W-phase sensor signal Hw output from the magnetic sensor 13 are respectively input to the processing unit 21. The processing unit 21 is connected to the storage unit 22 in a communicative manner via a communication bus (not shown). The processing unit 21 executes at least the following two processes according to a program pre-stored in the storage unit 22.

[0049] As offline processing, the processing unit 21 performs learning processing to acquire the learning data required for estimating the mechanical angle of the rotor shaft 110. Offline processing refers to processing performed before the angle detection device 1 is shipped from the manufacturing plant or before it is actually used in a customer-side system. In the learning processing, the processing unit 21 bases its learning on the sensor signals Hu, Hv, and Hw output from the magnetic sensors 11, 12, and 13, and the encoder 200 (see reference 1). Figure 1 The encoder 200 acquires learning data by outputting the signal AS. The encoder 200 is only positioned on the rotor shaft 110 during learning processing. The output signal AS of the encoder 200 represents the mechanical angle of the rotor shaft 110. The encoder 200 can be either an incremental encoder or an absolute encoder.

[0050] Furthermore, as an online processing step, the processing unit 21 performs angle estimation processing for estimating the mechanical angle of the rotor shaft 110 based on the sensor signals Hu, Hv, and Hw output from the magnetic sensors 11, 12, and 13, and the learning data obtained through learning processing. This online processing refers to the processing performed when the angle detection device 1 is actually installed in a system on the customer's side.

[0051] The storage unit 22 includes a non-volatile memory that stores programs, various setting data, and learning data required for the processing unit 21 to perform various processes, and a volatile memory used as a temporary storage location for data when the processing unit 21 performs various processes. The non-volatile memory is, for example, an electrically erasable programmable read-only memory (EEPROM) or flash memory. The volatile memory is, for example, random access memory (RAM).

[0052] Before describing the learning process and angle estimation process performed by the processing unit 21 of the angle detection device 1 configured as described above, a brief description of the position estimation method disclosed in Japanese Patent No. 6233532 will be provided to facilitate understanding of the present invention. In the following description, the position estimation method disclosed in Japanese Patent No. 6233532 will sometimes be referred to as the basic patent method. For details of the basic patent method, please refer to Japanese Patent No. 6233532. Furthermore, for ease of explanation, the following will use… Figure 1 The components shown illustrate the basic patented method.

[0053] First, the learning process performed by the processing unit 21 in the basic patent method will be explained.

[0054] While the sensing magnet 120 rotates together with the rotor shaft 110, the processing unit 21 acquires the signals output from the magnetic sensors 11, 12, and 13 as sensor signals Hu, Hv, and Hw. Specifically, the processing unit 21 incorporates an A / D converter, which converts the U-phase sensor signal Hu, V-phase sensor signal Hv, and W-phase sensor signal Hw into digital signals at a predetermined sampling frequency, thereby acquiring the digital values ​​of the U-phase sensor signal Hu, V-phase sensor signal Hv, and W-phase sensor signal Hw.

[0055] Furthermore, during the learning process, the rotor shaft 110 can be rotated by energizing the motor 100 via a motor control device (not shown). Alternatively, the rotor shaft 110 can be connected to a rotating machine (not shown), and the rotor shaft 110 can be rotated by the rotating machine.

[0056] Figure 2 This is a diagram illustrating an example of the waveforms of the U-phase sensor signal Hu, the V-phase sensor signal Hv, and the W-phase sensor signal Hw. (See diagram for example.) Figure 2 As shown, the electrical angle of each of the sensor signals Hu, Hv, and Hw, with one cycle, corresponds to 1 / 4 of the mechanical angle cycle, i.e., a mechanical angle of 90°. Figure 2 In this context, the period from time t1 to time t5 corresponds to one mechanical angle cycle (360°). Figure 2 In this context, the periods from time t1 to time t2, from time t2 to time t3, from time t3 to time t4, and from time t4 to time t5 each correspond to a mechanical angle of 90°. Furthermore, the sensor signals Hu, Hv, and Hw have a phase difference of 120° with each other.

[0057] Based on the digital values ​​of sensor signals Hu, Hv, and Hw, the processing unit 21 selects, within one cycle of the mechanical angle, the intersection point where two of the three sensor signals intersect each other, and the zero-crossing point where each of the three sensor signals intersects with a reference signal level. The reference signal level is, for example, a ground level. When the reference signal level is ground, the digital value of the reference signal level is "0".

[0058] like Figure 2 As shown, based on the selection result of the zero-crossing point, the processing unit 21 divides the mechanical angle 1 period into four pole pair regions associated with the pole pair number. Figure 2 In this context, "No.C" indicates the pole pair number. For example... Figure 1 As shown, the four pole pairs of the sensing magnet 120 are pre-assigned with pole pair numbers. For example, the pole pairs located in the range of 0° to 90° of the machine angle are assigned the pole pair number "0". The pole pairs located in the range of 90° to 180° of the machine angle are assigned the pole pair number "1". The pole pairs located in the range of 180° to 270° of the machine angle are assigned the pole pair number "2". The pole pairs located in the range of 270° to 360° of the machine angle are assigned the pole pair number "3".

[0059] For example, when using sensor signal Hu as a reference, processing unit 21 identifies the zero-crossing point of sensor signal Hu obtained at a sampling time (time t1) with a mechanical angle of 0° as the starting point of the pole pair region associated with pole pair number "0". Furthermore, processing unit 21 identifies the zero-crossing point of sensor signal Hu obtained at a sampling time (time t2) with a mechanical angle of 90° as the ending point of the pole pair region associated with pole pair number "0". That is, processing unit 21 determines the interval between the zero-crossing point obtained at time t1 and the zero-crossing point obtained at time t2 as the pole pair region associated with pole pair number "0".

[0060] The processing unit 21 also identifies the zero-crossing point of the sensor signal Hu obtained at the sampling time (time t2) with a mechanical angle of 90° as the starting point of the pole pair region associated with pole pair number "1". Furthermore, the processing unit 21 identifies the zero-crossing point of the sensor signal Hu obtained at the sampling time (time t3) with a mechanical angle of 180° as the ending point of the pole pair region associated with pole pair number "1". That is, the processing unit 21 determines the interval between the zero-crossing point obtained at time t2 and the zero-crossing point obtained at time t3 as the pole pair region associated with pole pair number "1".

[0061] The processing unit 21 also identifies the zero-crossing point of the sensor signal Hu obtained at the sampling time (time t3) with a mechanical angle of 180° as the starting point of the pole pair region associated with pole pair number "2". Furthermore, the processing unit 21 identifies the zero-crossing point of the sensor signal Hu obtained at the sampling time (time t4) with a mechanical angle of 270° as the ending point of the pole pair region associated with pole pair number "2". That is, the processing unit 21 determines the interval between the zero-crossing point obtained at time t3 and the zero-crossing point obtained at time t4 as the pole pair region associated with pole pair number "2".

[0062] The processing unit 21 also identifies the zero-crossing point of the sensor signal Hu obtained at the sampling time (time t4) with a mechanical angle of 270° as the starting point of the pole pair region associated with pole pair number "3". Furthermore, the processing unit 21 identifies the zero-crossing point of the sensor signal Hu obtained at the sampling time (time t5) with a mechanical angle of 360° as the ending point of the pole pair region associated with pole pair number "3". That is, the processing unit 21 determines the interval between the zero-crossing point obtained at time t4 and the zero-crossing point obtained at time t5 as the pole pair region associated with pole pair number "3".

[0063] like Figure 2 As shown, based on the selection results of intersection points and zero intersection points, the processing unit 21 divides the four pole pair regions into 12 regions associated with region numbers. Figure 2 In this context, "No. A" represents the district number associated with each district. For example... Figure 2 As shown, the 12 zones included in each of the four polar regions are associated with the zone numbers from "0" to "11".

[0064] Figure 3 yes Figure 2 The image shows a magnified view of the sensor signals Hu, Hv, and Hw contained within a single pole pair region. Figure 3 In this context, the reference value (reference signal level) of the amplitude is "0". Figure 3 In this example, a positive amplitude value represents the magnetic field strength of the N pole. Conversely, a negative amplitude value represents the magnetic field strength of the S pole.

[0065] exist Figure 3 In the diagram, points P1, P3, P5, P7, P9, P11, and P13 are zero-crossing points selected from the digital values ​​of sensor signals Hu, Hv, and Hw contained within a single polar pair region. Furthermore, in... Figure 3In the diagram, points P2, P4, P6, P8, P10, and P12 are intersection points selected from the digital values ​​of sensor signals Hu, Hv, and Hw contained within a single polar pair region. For example... Figure 3 As shown, the processing unit 21 determines the interval between adjacent zero-crossing points and intersection points as a region.

[0066] Processing unit 21 determines the interval between zero-crossing point P1 and intersection point P2 as the area associated with area number "0". Processing unit 21 determines the interval between intersection point P2 and zero-crossing point P3 as the area associated with area number "1". Processing unit 21 determines the interval between zero-crossing point P3 and intersection point P4 as the area associated with area number "2". Processing unit 21 determines the interval between intersection point P4 and zero-crossing point P5 as the area associated with area number "3". Processing unit 21 determines the interval between zero-crossing point P5 and intersection point P6 as the area associated with area number "4". Processing unit 21 determines the interval between intersection point P6 and zero-crossing point P7 as the area associated with area number "5".

[0067] Processing unit 21 determines the interval between zero-crossing point P7 and intersection point P8 as the area associated with area number "6". Processing unit 21 determines the interval between intersection point P8 and zero-crossing point P9 as the area associated with area number "7". Processing unit 21 determines the interval between zero-crossing point P9 and intersection point P10 as the area associated with area number "8". Processing unit 21 determines the interval between intersection point P10 and zero-crossing point P11 as the area associated with area number "9". Processing unit 21 determines the interval between zero-crossing point P11 and intersection point P12 as the area associated with area number "10". Processing unit 21 determines the interval between intersection point P12 and zero-crossing point P13 as the area associated with area number "11".

[0068] Furthermore, in the following description, for example, a zone assigned the zone number "0" will be referred to as "Zone 0", and a zone assigned the zone number "11" will be referred to as "Zone 11".

[0069] like Figure 2 As shown, consecutive numbers throughout the entire period of mechanical angle 1 cycle are used as segment numbers and associated with each zone number. Figure 2 In this context, "No. B" indicates the segment number associated with each zone number. Furthermore, a segment is a term referring to a straight line connecting adjacent intersections to a zero-intersection point. In other words, the straight line connecting the start and end points of each zone is called a segment. Figure 3 In this example, zone 0 starts at the zero-intersection point P1 and ends at the intersection point P2. Therefore, the segment corresponding to zone 0 is the straight line connecting the zero-intersection point P1 and the intersection point P2. Similarly, in Figure 3For example, the starting point of zone 1 is intersection point P2, and the ending point of zone 1 is zero intersection point P3. Therefore, the segment corresponding to zone 1 is the straight line connecting intersection point P2 and zero intersection point P3.

[0070] like Figure 2 As shown, in the polar pair region associated with polar pair number "0", for area number "0" to area number "11", segment number "0" to segment number "11" are associated. In the polar pair region associated with polar pair number "1", for area number "0" to area number "11", segment number "12" to segment number "23" are associated. In the polar pair region associated with polar pair number "2", for area number "0" to area number "11", segment number "24" to segment number "35" are associated. In the polar pair region associated with polar pair number "3", for area number "0" to area number "11", segment number "36" to segment number "47" are associated.

[0071] Furthermore, in the following description, for example, a segment assigned segment number "0" will be referred to as "segment number 1", and a segment assigned segment number "11" will be referred to as "segment number 11".

[0072] Processing unit 21 generates a linear function θ(Δx) representing each segment. Δx is the length (numerical value) from the starting point of the segment to any point on the segment, and θ is the mechanical angle corresponding to any point on the segment. Figure 3 In this context, for example, the starting point of the segment corresponding to zone 0 is the zero intersection point P1, and the ending point of the segment corresponding to zone 0 is the intersection point P2. Similarly, in... Figure 3 For example, the starting point of the segment corresponding to zone 1 is intersection point P2, and the ending point of the segment corresponding to zone 1 is zero intersection point P3.

[0073] For example, the linear function θ(Δx) representing a segment is expressed by the following equation (1). In equation (1), "i" is the segment number, which is an integer from 0 to 47. In the following description, the linear function θ(Δx) represented by equation (1) is sometimes referred to as the machine angle estimate, and the machine angle θ calculated by equation (1) is referred to as the machine angle estimate.

[0074] θ(Δx)=k[i]×Δx+θres[i]…(1)

[0075] In equation (1) above, k[i] is the coefficient referred to as the normalization coefficient. In other words, k[i] is the coefficient representing the slope of segment i. The normalization coefficient k[i] is expressed in equation (2) below. In equation (2) below, ΔXnorm[i] is the deviation of the numerical values ​​between the start and end points of segment i. Figure 3In this context, for example, ΔXnorm[i] of the segment corresponding to zone 0 represents the deviation of the numerical value between the zero intersection point P1 and the intersection point P2. Similarly, in Figure 3 In the example, ΔXnorm[i] of the segment corresponding to zone 1 is the deviation of the numerical value between the intersection point P2 and the zero intersection point P3.

[0076] k[i]=θnorm[i] / ΔXnorm[i]…(2)

[0077] In equation (2) above, θnorm[i] is the deviation of the mechanical angle between the start and end points of segment i, as expressed in equation (3) below. In equation (3) below, t[i] is the time between the start and end points of segment i, t[0] is the time between the start and end points of segment 0, and t

[47] is the time between the start and end points of segment 47. Figure 3 In the case of the segment corresponding to the 0th zone, for example, t[0] is the time between the zero intersection point P1 and the intersection point P2.

[0078] θnorm[i]={t[i] / (t[0]+…+t

[47] )}×360[degM]…(3)

[0079] In equation (1) above, θres[i] is a constant called the angle reset value (the intercept of the linear function θ(Δx)) for segment i. When segment number "i" is "0", the angle reset value θres[i] is expressed by equation (4). When segment number "i" is any one of "1" to "47", the angle reset value θres[i] is expressed by equation (5). Furthermore, θnorm[i] can be obtained not from t[i] as described above, but from the true value of the mechanical angle (e.g., the mechanical angle represented by the output signal of the encoder mounted on rotor shaft 110).

[0080] θres[i]=0[degM]…(4)

[0081] θres[i]=Σ(θnorm[i-1])…(5)

[0082] The processing unit 21 acquires the correspondence between pole pair numbers, area numbers, and segment numbers, the characteristic data of each area, and the mechanical angle estimation formula of each segment by performing the learning process described above, and stores the acquired data as learning data in the storage unit 22. Furthermore, the characteristic data of each area includes the magnitude relationship and sign of the digital values ​​of the sensor signals Hu, Hv, and Hw contained in each area. Moreover, the normalization coefficient k[i] and the angle reset value θres[i] constituting the mechanical angle estimation formula of each segment are stored as learning data in the storage unit 22.

[0083] Next, the angle estimation process performed by the processing unit 21 in the basic patent method will be explained.

[0084] The processing unit 21 acquires sensor signals Hu, Hv, and Hw output from magnetic sensors 11, 12, and 13. Specifically, the processing unit 21 uses an A / D converter to digitally convert the U-phase sensor signal Hu, V-phase sensor signal Hv, and W-phase sensor signal Hw at a predetermined sampling frequency, thereby acquiring the digital values ​​of the U-phase sensor signal Hu, V-phase sensor signal Hv, and W-phase sensor signal Hw.

[0085] Then, the processing unit 21 determines the current zone number and pole pair number based on the digital values ​​of sensor signal Hu, sensor signal Hv, and sensor signal Hw obtained at this sampling time. For example, in Figure 3 In this context, it is assumed that point PHu on the waveform of the U-phase sensor signal Hu, point PHv on the waveform of the V-phase sensor signal Hv, and point PHw on the waveform of the W-phase sensor signal Hw are the digital values ​​of each sensor signal Hu, sensor signal Hv, and sensor signal Hw obtained at the current sampling time. The processing unit 21 determines the current region (region number) by comparing the magnitude relationship and sign of the digital values ​​of points PHu, PHv, and PHw with the feature data of each region included in the learning data stored in the storage unit 22. Figure 3 In the example, area 9 is determined as the current area. Furthermore, the method for determining the pole pair number is not described in this specification. For the method for determining the pole pair number, please refer to Japanese Patent No. 6233532. Assuming that the pole pair number "2" is determined for this sampling time, for example...

[0086] Then, processing unit 21 determines the current segment number based on the determined current zone number and pole pair number. For example, processing unit 21 determines the current segment number using the formula "segment number = 12 × pole pair number + zone number". As described above, it is assumed that zone number "9" is determined as the current zone number, and pole pair number "2" is determined as the current pole pair number. In this case, processing unit 21 determines segment number "33" as the current segment number (see reference). Figure 2 ).

[0087] The processing unit 21 reads the normalization coefficient k[i] and angle reset value θres[i] corresponding to the determined segment number "i" from the learning data stored in the storage unit 22, and calculates the estimated mechanical angle value θ using the mechanical angle estimation formula expressed in equation (1) above. Here, as Δx substituted into the mechanical angle estimation formula, the digital value of the sensor signal corresponding to the determined segment is used. For example, as described above, when the segment number "33" is determined to be the current segment number, the processing unit 21 reads the normalization coefficient k

[33] and angle reset value θres

[33] from the storage unit 22, and calculates the estimated mechanical angle value θ using the digital value of point PHv (refer to...). Figure 3 Substituting Δx into the mechanical angle estimation formula, the estimated mechanical angle θ under this sampling time is calculated.

[0088] The above is the basic estimation procedure for the mechanical angle in the basic patent method that forms the basis of this invention.

[0089] In the basic patented method, correction processing is performed on sensor signals Hu, Hv, and Hw to improve the estimation accuracy of the machine angle (the accuracy of the estimated machine angle value θ). For example, ... Figure 2 As shown, the amplitude values ​​of the sensor signals Hu, Hv, and Hw may not be consistent. Furthermore, for example, as... Figure 4 As shown, there are cases where each sensor signal Hu, sensor signal Hv, and sensor signal Hw contains in-phase signals (DC signals and third harmonic signals, etc.) as noise components. Figure 4 This is a diagram illustrating an example of the waveforms of sensor signals Hu, Hv, and Hw, which contain in-phase signals as noise components. Figure 4 In the diagram, the vertical axis represents numerical values, and the horizontal axis represents electrical angles.

[0090] Therefore, after the processing unit 21 in the basic patent method obtains the digital values ​​of sensor signal Hu, sensor signal Hv and sensor signal Hw when performing learning processing and angle estimation processing, it first performs a first correction process based on the following formulas (6), (7) and (8) to remove in-phase signals from sensor signal Hu, sensor signal Hv and sensor signal Hw.

[0091] Hiu0=Hu-(Hv+Hw) / 2…(6)

[0092] Hiv0=Hv-(Hu+Hw) / 2…(7)

[0093] Hiw0=Hw-(Hu+Hv) / 2…(8)

[0094] In equation (6), Hiu0 is the digital value of the U-phase sensor signal obtained by performing the first correction process on the U-phase sensor signal Hu. In equation (7), Hiv0 is the digital value of the V-phase sensor signal obtained by performing the first correction process on the V-phase sensor signal Hv. In equation (8), Hiw0 is the digital value of the W-phase sensor signal obtained by performing the first correction process on the W-phase sensor signal Hw. Figure 5 This is a diagram illustrating an example of the waveforms of sensor signals Hiu0, Hiv0, and Hiw0 obtained after performing the first correction process. Figure 5 In the diagram, the vertical axis represents numerical values, and the horizontal axis represents electrical angles.

[0095] After performing the first correction process, the processing unit 21 in the basic patent method performs a second correction process on the sensor signal Hiu0, the sensor signal Hiv0 and the sensor signal Hiw0 based on the following formulas (9) to (14) to make the amplitude values ​​consistent.

[0096] Hiu1(ppn)=au_max(ppn)×Hiu0(ppn)+bu…(9)

[0097] Hiu1(ppn)=au_min(ppn)×Hiu0(ppn)+bu…(10)

[0098] Hiv1(ppn)=av_max(ppn)×Hiv0(ppn)+bv…(11)

[0099] Hiv1(ppn)=av_min(ppn)×Hiv0(ppn)+bv…(12)

[0100] Hiw1(ppn)=aw_max(ppn)×Hiw0(ppn)+bw…(13)

[0101] Hiw1(ppn)=aw_min(ppn)×Hiw0(ppn)+bw…(14)

[0102] The processing unit 21 uses the information stored in the storage unit 22 to perform a second correction process on the digital value of the positive side of the U-phase sensor signal Hiu0 according to the above formula (9). Furthermore, the processing unit 21 uses the information stored in the storage unit 22 to perform a second correction process on the digital value of the negative side of the U-phase sensor signal Hiu0 according to the above formula (10).

[0103] The processing unit 21 uses the information stored in the storage unit 22 to perform a second correction process on the digital value of the positive side of the V-phase sensor signal Hiv0 according to the above formula (11). Furthermore, the processing unit 21 uses the information stored in the storage unit 22 to perform a second correction process on the digital value of the negative side of the V-phase sensor signal Hiv0 according to the above formula (12).

[0104] The processing unit 21 uses the information stored in the storage unit 22 to perform a second correction process on the positive digital value of the W-phase sensor signal Hiw0 according to the above formula (13). Furthermore, the processing unit 21 uses the information stored in the storage unit 22 to perform a second correction process on the negative digital value of the W-phase sensor signal Hiw0 according to the above formula (14).

[0105] In equations (9) and (10), Hiu1 is the digital value of the U-phase sensor signal obtained by performing a second correction process on the U-phase sensor signal Hiu0. In equations (11) and (12), Hiv1 is the digital value of the V-phase sensor signal obtained by performing a second correction process on the V-phase sensor signal Hiv0. In equations (13) and (14), Hiw1 is the digital value of the W-phase sensor signal obtained by performing a second correction process on the W-phase sensor signal Hiw0. Figure 6 This is a diagram illustrating an example of the waveforms of sensor signals Hiu1, Hiv1, and Hiw1 obtained after performing the second correction process. Figure 6 In the diagram, the vertical axis represents numerical values, and the horizontal axis represents electrical angles.

[0106] Furthermore, in equations (9) to (14), ppn is the pole pair number from 0 to 3. In equations (9), (11), and (13), au_max(ppn), av_max(ppn), and aw_max(ppn) are positive gain correction values ​​relative to the positive side of the digital value of the electrical angle 1-cycle quantity corresponding to each magnetic pole pair pre-stored in the storage unit 22. In equations (10), (12), and (14), au_min(ppn), av_min(ppn), and aw_min(ppn) are negative gain correction values ​​relative to the negative side of the digital value of the electrical angle 1-cycle quantity corresponding to each magnetic pole pair pre-stored in the storage unit 22. In equations (9) to (14), bu, bv, and bw are offset correction values ​​for each phase stored in the storage unit 22. Furthermore, au_max(ppn), av_max(ppn), aw_max(ppn), au_min(ppn), av_min(ppn), and aw_min(ppn) are the correction values ​​for each pole pair. Therefore, there are 12 positive-side gain correction values ​​(3 phases × 4 pole pairs). Similarly, there are 12 negative-side gain correction values.

[0107] After performing the second correction process, the processing unit 21 in the basic patent method performs a third correction process on the sensor signals Hiu1, Hiv1, and Hiw1 to linearize a portion (segmented signal) of the sensor signal corresponding to each segment. Figure 3 In the case where, for example, the segment corresponding to zone 0 is zone 0, the segmentation signal corresponding to said zone 0 is the portion of the U-phase sensor signal Hu that connects the zero-crossing point P1 and the intersection point P2. Similarly, in Figure 3 In the case where, for example, the segment corresponding to zone 1 is zone 1, the segmentation signal corresponding to zone 1 is the signal in the W-phase sensor signal Hw that connects the intersection point P2 and the zero intersection point P3.

[0108] The processing unit 21 performs a third correction process on the sensor signals Hiu1, Hiv1, and Hiw1 by using values ​​pre-stored in the storage unit 22 as coefficients to change the scale of each sensor signal. This third correction process linearizes the approximate S-shaped shape of the segmented signals corresponding to each segment. Here, the values ​​stored in the storage unit 22 are pre-designed values. The third correction process uses these pre-designed values ​​and performs calculations using quadratic, cubic, or trigonometric functions.

[0109] As an example, the processing unit 21 performs a third correction process on the sensor signal Hiu1, the sensor signal Hiv1, and the sensor signal Hiw1 based on the following equations (15) to (17). In the following equations (15) to (17), a and b are coefficients pre-stored in the storage unit 22.

[0110] Hiu2=b×tan(a×Hiu1)…(15)

[0111] Hiv2=b×tan(a×Hiv1)…(16)

[0112] Hiw2=b×tan(a×Hiw1)…(17)

[0113] In equation (15), Hiu2 is the digital value of the U-phase sensor signal obtained by performing a third correction process on the U-phase sensor signal Hiu1. In equation (16), Hiv2 is the digital value of the V-phase sensor signal obtained by performing a third correction process on the V-phase sensor signal Hiv1. In equation (17), Hiw2 is the digital value of the W-phase sensor signal obtained by performing a third correction process on the W-phase sensor signal Hiw1. Figure 7This is a diagram illustrating an example of the waveforms of sensor signals Hiu2, Hiv2, and Hiw2 obtained after performing the third correction process. Figure 7 In the diagram, the vertical axis represents numerical values, and the horizontal axis represents electrical angles.

[0114] As described above, in the basic patent method, the first correction process reduces the in-phase noise contained in the sensor signals Hu, Hv, and Hw. Furthermore, in the basic patent method, the second correction process corrects the mutual deviations between the sensor signals. Here, mutual deviations refer to, for example, deviations in the amplitude values ​​and offset components of each sensor signal. Moreover, in the basic patent method, the third correction process linearizes the curved portion of the waveform of each sensor signal. In particular, by performing the second correction process, the length of a portion of the sensor signal corresponding to a segment (the segmented signal) is uniformized; therefore, in the third correction process, it is easy to apply the same calculation process to all segmented signals. Therefore, by performing the second correction process before the third correction process, the curved portion of the waveform can be further linearized.

[0115] As a result, in the basic patent method, the signal part (segmented signal) required for the calculation of the estimated mechanical angle θ based on the above equation (1) is further linearized, which can reduce the difference between the estimated mechanical angle θ and the true mechanical angle (e.g., the mechanical angle represented by the output signal of the encoder installed on the rotor shaft 110), thus enabling high-precision mechanical angle estimation.

[0116] Figure 8 This is a graph showing the relationship between the waveforms of sensor signals Hiu1, Hiv1, and Hiw1 and the angle error before the third correction process. Figure 9 This is a graph showing the relationship between the waveforms of sensor signals Hiu2, Hiv2, and Hiw2 obtained after the third correction process and the angle error. Figure 8 and Figure 9 In this context, the angle error is the value obtained by subtracting the true value of the mechanical angle (e.g., the mechanical angle represented by the output signal of the encoder installed on the rotor shaft 110) from the estimated mechanical angle θ calculated by the above formula (1).

[0117] like Figure 8 As shown, when the estimated mechanical angle θ is calculated based on sensor signals Hiu1, Hiv1, and Hiw1 before the third correction process, the angle error is approximately ±0.5 [deg]. On the other hand, as... Figure 9As shown, when the estimated mechanical angle θ is calculated based on the sensor signals Hiu2, Hiv2, and Hiw2 obtained after the third correction process, the angle error is approximately ±0.1 [deg]. As described above, in the basic patent method, by linearizing the segmented signals corresponding to each segment, the angle error between the estimated mechanical angle θ and the true mechanical angle can be reduced.

[0118] like Figure 10 As shown, the inventors of this application, using the basic patent method, verified the encoder error relative to the master lead angle using a real machine. The result obtained was that the angle error between the estimated mechanical angle θ and the true mechanical angle was approximately within ±0.06 [deg]. However, the angle error may increase or decrease due to changes in the operating environment such as temperature, and therefore, there may be situations where the operating environment causes the angle error to be greater than the stated error.

[0119] The purpose of this invention is to further reduce the angular error between the estimated mechanical angle θ and the true mechanical angle compared to the basic patented method, thereby improving the accuracy of mechanical angle detection of the rotating shaft.

[0120] Hereinafter, in order to solve the aforementioned technical problem, the learning processing and angle estimation processing performed by the processing unit 21 of the angle detection device 1 in this embodiment will be described.

[0121] First, the learning process performed by the processing unit 21 of the angle detection device 1 in this embodiment will be explained.

[0122] Figure 11 This is a flowchart illustrating the learning process performed by the processing unit 21 of the angle detection device 1 in this embodiment in an offline processing manner. The processing unit 21 performs the learning process in an offline processing manner to obtain the learning data required to estimate the mechanical angle of the rotor shaft 110.

[0123] like Figure 11 As shown, when the sensing magnet 120 rotates together with the rotor shaft 110, the processing unit 21 acquires the signals output from the three magnetic sensors 11, 12, and 13 as sensor signals Hu, Hv, and Hw (step S1). Specifically, the processing unit 21 has an integrated A / D converter. The processing unit 21 uses the A / D converter to digitally convert the U-phase sensor signal Hu, V-phase sensor signal Hv, and W-phase sensor signal Hw at a predetermined sampling frequency, thereby acquiring the digital values ​​of the U-phase sensor signal Hu, V-phase sensor signal Hv, and W-phase sensor signal Hw.

[0124] Next, based on the digital values ​​of sensor signals Hu, Hv, and Hw, the processing unit 21 selects the intersection point where two of the three sensor signals intersect each other and the zero-crossing point where each of the three sensor signals intersects with the reference signal level within a mechanical angle period of 1 (step S2). Then, the processing unit 21 generates a linear function θ(Δx) representing the straight line (segment) connecting adjacent intersection points and zero-crossing points, i.e., the mechanical angle estimation formula for each segment (step S3).

[0125] The processing in steps S1 to S3 is the same as the learning processing in the basic patent method, so detailed explanation is omitted. Furthermore, for the sensor signals Hu, Hv, and Hw obtained through step S1, the first correction processing, second correction processing, and third correction processing in the basic patent method can be performed, and the processing after step S2 is performed based on the sensor signals Hiu2, Hiiv2, and Hiw2 obtained after the third correction processing.

[0126] Processing unit 21, by executing steps S1 to S3, acquires the correspondence between pole pair numbers, zone numbers, and segment numbers, the characteristic data of each zone, and the mechanical angle estimation formula for each segment, and stores the acquired data as learning data in storage unit 22. Furthermore, the characteristic data of each zone includes the magnitude relationship and positive / negative signs of the digital values ​​of sensor signals Hu, Hv, and Hw contained in each zone. Moreover, the normalization coefficient k[i] and angle reset value θres[i] constituting the mechanical angle estimation formula for each segment are stored as learning data in storage unit 22.

[0127] Next, for multiple points on the segment, the processing unit 21 calculates the deviation between the estimated mechanical angle θ calculated based on the mechanical angle estimation formula and the mechanical angle θe obtained from the output signal AS of the encoder 200 installed on the rotor shaft 110 as the first angle error (step S4). Hereinafter, the mechanical angle θe obtained from the encoder 200 will sometimes be referred to as the true mechanical angle. For example, as... Figure 12 As shown, in this embodiment, the processing unit 21 calculates the deviation between the estimated mechanical angle θ and the true mechanical angle θe for nine points on the segment, and uses this deviation as the first angle error θerr. Hereinafter, the nine points on the segment will sometimes be referred to as points of interest.

[0128] exist Figure 12 In this context, "k" represents the numbering from "0" to "8" assigned to the nine points of interest within the segment. Hereinafter, the point of interest assigned the number "k" will sometimes be referred to as point of interest number k. For example... Figure 12As shown, the nine points of interest on the segment include the start and end points of the segment. Point of interest 0 corresponds to the start of the segment, and point of interest 8 corresponds to the end of the segment. The length from point of interest 0 to point of interest 8 is equivalent to the numerical deviation between the start and end points of segment i, i.e., ΔXnorm[i]. In this embodiment, the intervals between the multiple points of interest on the segment are equal. In other words, in this embodiment, the segment is divided into eight equal parts by nine equally spaced points of interest.

[0129] The processing unit 21 calculates the estimated mechanical angle θ corresponding to the k-th point of interest by substituting "k×ΔXnorm[i] / 8" into Δx in the above equation (1), and calculates the deviation between the true mechanical angle θe obtained from the encoder 200 at the same sampling time as the k-th point of interest and the calculated estimated mechanical angle θ as the first angle error θerrk. Furthermore, in Figure 12 In the above, the first angle error θerr0 is the first angle error calculated by the method for point of interest 0. The first angle error θerr1 is the first angle error calculated by the method for point of interest 1. The first angle error θerr7 is the first angle error calculated by the method for point of interest 7. The first angle error θerr8 is the first angle error calculated by the method for point of interest 8. The processing unit 21 sequentially reads all the segment-related learning data from segment 0 to segment 47, and calculates the deviation between the estimated mechanical angle value θ and the true mechanical angle value θe for the nine points of interest in each segment as the first angle error θerr.

[0130] By performing the processing described in step S4 above, eight first angular errors θerr can be obtained for each of the 12 segments contained in a pole pair region. Therefore, as Figure 13 As shown, for one pole pair region, a total of 96 first angle errors θerr can be obtained. Therefore, a total of 384 first angle errors θerr can be obtained for the four pole pair regions as a whole. As described above, the processing unit 21 stores the first angle errors θerr calculated for the nine points of interest in each segment as learning values ​​in the storage unit 22 according to each segment (step S5).

[0131] Next, the processing unit 21 generates a first angle error function (step S6) based on the first angle error θerr calculated for the nine points of interest in each segment, to calculate the first angle error θerr corresponding to any point in the segment. The first angle error function is expressed by the following equation (18). In the following equation (18), "x" has the same meaning as Δx, and is the length (numerical value) from the starting point of segment i to any point (refer to...). Figure 12 Moreover, in equation (18), Xnorm and ΔXnorm[i] have the same meaning.

[0132] [Number 1]

[0133]

[0134] Next, for multiple points on the segment, the processing unit 21 calculates the deviation between the estimated mechanical angle θ calculated based on the mechanical angle estimation formula and the first angle error θer1 calculated based on the first angle error function represented by the above formula (18), and the true mechanical angle value θe obtained from the encoder 200, as the second angle error (step S7). The multiple points on the segment used in step S7 can be the same as the nine points of interest used in step S4, or they can be different points. Through the processing in step S7, the angle error (second angle error) can be further learned while using the first angle error obtained in step S4. The processing unit 21 stores the second angle error calculated for multiple points on the segment as a learning value in the storage unit 22 according to each segment (step S8).

[0135] Finally, the processing unit 21 generates a second angle error function (step S9) based on the second angle error calculated for multiple points on each segment, to calculate the second angle error corresponding to any point on the segment. The second angle error function is represented by the same formula as above (18).

[0136] Processing unit 21, by performing steps S1 to S3 of the learning process described above, acquires the correspondence between pole pair numbers, area numbers, and segment numbers, characteristic data of each area, and mechanical angle estimation formulas for each segment, and stores the acquired data as learning data in storage unit 22. Furthermore, processing unit 21, by performing steps S4 to S9 of the learning process described above, generates a first angle error function for calculating the first angle error corresponding to any point on each segment, and a second angle error function for calculating the second angle error corresponding to any point on each segment, and stores these required learning values ​​as learning data in storage unit 22.

[0137] Next, the angle estimation process performed by the processing unit 21 in this embodiment will be explained.

[0138] The processing unit 21 acquires sensor signals Hu, Hv, and Hw output from magnetic sensors 11, 12, and 13. Specifically, the processing unit 21 uses an A / D converter to digitally convert the U-phase sensor signal Hu, V-phase sensor signal Hv, and W-phase sensor signal Hw at a predetermined sampling frequency, thereby acquiring the digital values ​​of the U-phase sensor signal Hu, V-phase sensor signal Hv, and W-phase sensor signal Hw.

[0139] Then, the processing unit 21 determines the current zone number and pole pair number based on the digital values ​​of sensor signal Hu, sensor signal Hv, and sensor signal Hw obtained at this sampling time. For example, in Figure 3 In this context, it is assumed that point PHu on the waveform of the U-phase sensor signal Hu, point PHv on the waveform of the V-phase sensor signal Hv, and point PHw on the waveform of the W-phase sensor signal Hw are the digital values ​​of each sensor signal Hu, sensor signal Hv, and sensor signal Hw obtained at the current sampling time. The processing unit 21 determines the current region (region number) by comparing the magnitude relationship and sign of the digital values ​​of points PHu, PHv, and PHw with the feature data of each region included in the learning data stored in the storage unit 22. Figure 3 In the example, area 9 is determined as the current area. Moreover, as the pole pair number for this sampling time, it is assumed, for example, that pole pair number "2" has been determined.

[0140] Then, processing unit 21 determines the current segment number based on the determined current zone number and pole pair number. For example, processing unit 21 determines the current segment number using the formula "segment number = 12 × pole pair number + zone number". As described above, it is assumed that zone number "9" is determined as the current zone number, and pole pair number "2" is determined as the current pole pair number. In this case, processing unit 21 determines segment number "33" as the current segment number (see reference). Figure 2 ).

[0141] The processing unit 21 reads the normalization coefficient k[i] and the angle reset value θres[i] corresponding to the determined segment number "i" from the learning data stored in the storage unit 22, and calculates the estimated mechanical angle value θ using the mechanical angle estimation formula expressed by equation (19). Here, Δx, which is substituted into the mechanical angle estimation formula, is the digital value of the sensor signal corresponding to the determined segment. In equation (19), θer1 is the first angle error obtained by substituting Δx into the first angle error function obtained through the learning process, and θer2 is the second angle error obtained by substituting Δx into the second angle error function obtained through the learning process.

[0142] θ(Δx)=k[i]×Δx+θres[i]-(θer1+θer2)…(19)

[0143] For example, as described above, when the segment number "33" is determined to be the current segment number, the processing unit 21 reads the normalization coefficient k

[33] and the angle reset value θres

[33] from the storage unit 22, and sets the digital value of point PHv (refer to) Figure 3Substitute Δx into the first angle error function and the second angle error function to calculate the first angle error θer1 and the second angle error θer2 corresponding to the digital value of point PHv. Substitute these Δx, first angle error θer1 and second angle error θer2 into the mechanical angle estimation formula represented by the above formula (19) to calculate the mechanical angle estimation value θ under this sampling time.

[0144] Figure 14 This is a graph showing the results of evaluating the encoder error relative to the master lead angle, using only the first angle error correction to correct for the estimated mechanical angle θ. In other words, Figure 14 This is a graph showing the result of the angle error between the estimated value θ of the machine angle calculated by the following formula (20) and the true value θe of the machine angle obtained by actual machine verification. Figure 15 This is a graph showing the results of evaluating the encoder error relative to the master lead angle, using both the first and second angle errors to correct for the estimated mechanical angle θ. In other words, Figure 15 The graph shows the result of the actual machine verification of the angle error between the estimated value θ of the machine angle calculated by the above formula (19) and the true value θe of the machine angle.

[0145] θ(Δx)=k[i]×Δx+θres[i]-θer1…(20)

[0146] like Figure 14 As shown, the inventors of this application, using the angle estimation method of the present invention, verified the encoder error relative to the main lead angle in a real-world test. The results showed that even when only the first angle error was used to correct the estimated mechanical angle θ, the angle error between the estimated mechanical angle θ and the true mechanical angle θe was within ±0.03 [deg]. Moreover, as... Figure 15 As shown, the inventors of this application used the angle estimation method of the present invention to verify the encoder error relative to the main lead angle in a real machine. The result obtained was that, when the estimated mechanical angle θ was corrected by using the first angle error and the second angle error, the angle error between the estimated mechanical angle θ and the true mechanical angle θe was within ±0.02 [deg].

[0147] Based on these results, the present invention enables the angle error to be less than that of the basic patented method.

[0148] As described above, the angle detection device 1 in this embodiment includes: three magnetic sensors 11, 12, and 13 to detect changes in magnetic flux caused by the rotation of the rotor shaft 110; and a signal processing unit 20 to process the signals output from the three magnetic sensors. The processing unit 21 of the signal processing unit 20 performs the following processing: acquiring the signals output from the three magnetic sensors as sensor signals Hu, Hv, and Hw (step S1), wherein the three sensor signals Hu, Hv, and Hw have a phase difference of 120° electrical angle; selecting, within a mechanical angle period of 1 cycle, the intersection points where two of the three sensor signals Hu, Hv, and Hw intersect each other, and the zero-crossing points where the three sensor signals Hu, Hv, and Hw intersect with a reference signal level (step S2); and generating a straight line (segment) representing the connection between adjacent intersection points and zero-crossing points. The processing of a linear function θ(Δx) for multiple points on the segment (step S3), where Δx is the length from the starting point of the segment to any point on the segment, and θ is the mechanical angle corresponding to any point on the segment; the processing of calculating the deviation between the mechanical angle θ calculated based on the linear function θ(Δx) and the mechanical angle θe obtained from the encoder 200 set on the rotor shaft 110 as a first angle error for multiple points on the segment (step S4); the processing of storing the first angle error calculated for multiple points on the segment as a learning value (step S5); and the processing of generating a first angle error function to calculate the first angle error corresponding to any point on the segment based on the first angle error calculated for multiple points on the segment (step S6).

[0149] According to the embodiment described above, compared with the basic patent method disclosed in Japanese Patent No. 6233532, the angular error between the estimated mechanical angle θ and the true mechanical angle θe can be further reduced, thereby improving the accuracy of mechanical angle detection of the rotating shaft.

[0150] In this embodiment, the processing unit 21 further includes the following processes: for multiple points on the segment, a process of calculating the deviation between the value obtained by subtracting the first angle error calculated based on the first angle error function from the mechanical angle θ calculated based on the linear function θ(Δx) and the mechanical angle θe obtained from the encoder 200 as a second angle error (step S7); a process of storing the second angle error calculated for multiple points on the segment as a learning value (step S8); and a process of generating a second angle error function for calculating the second angle error corresponding to any point on the segment based on the second angle error calculated for multiple points on the segment (step S9).

[0151] Therefore, since both the first angle error and the second angle error are used to correct the estimated mechanical angle θ, the angle error between the estimated mechanical angle θ and the true mechanical angle θe can be further reduced compared to the case where only the first angle error is used to correct the estimated mechanical angle θ.

[0152] In this embodiment, the intervals between multiple points on the segment are equal.

[0153] Therefore, the first angle error and the second angle error at any point on the segment can be calculated with good accuracy by linear interpolation, for example, expressed by equation (18). By correcting the estimated mechanical angle θ by using at least one of these first angle errors and the second angle errors, the accuracy of the estimated mechanical angle θ can be improved.

[0154] (Modified Example)

[0155] This invention is not limited to the embodiments described herein, and the various structures described in this specification may be appropriately combined within a range that does not contradict each other.

[0156] For example, in the described embodiment, the case where the estimated mechanical angle θ is corrected using equation (19) with both the first and second angle errors is illustrated. However, even when only the first angle error is used to correct the estimated mechanical angle θ (using equation (20)), there is no significant difference in the angle error between the estimated mechanical angle θ and the true mechanical angle θe. Therefore, the estimated mechanical angle θ can also be corrected using equation (20) with only the first angle error. In this case, the estimated mechanical angle θ can be... Figure 11 The processing steps S7 to S9 in the learning process shown are deleted.

[0157] In the described embodiment, the processing unit 21 calculates the deviation between the estimated mechanical angle θ and the true mechanical angle θe for nine points of interest on the segment as a first angle error θerr. However, the number of points of interest arranged on the segment is not limited to nine. Furthermore, in the described embodiment, the segment is divided into eight equal parts by nine equally spaced points of interest. However, the intervals between multiple points of interest arranged on the segment may not be equal. The number of segment divisions is also not limited to eight.

[0158] In the described embodiment, a case is illustrated where a sensing magnet 120 is used as a position detection magnet, i.e., a magnet that rotates synchronously with the rotor shaft 110 of the motor 100. However, a rotor magnet mounted on the rotor of the motor 100 can also be used as a position detection magnet. The rotor magnet is also a magnet that rotates synchronously with the rotor shaft 110 and has multiple pole pairs.

[0159] In the described embodiment, the sensor group 10 is illustrated as including three magnetic sensors 11, 12, and 13. However, the number of magnetic sensors is not limited to three; any number of N (where N is a multiple of 3) is acceptable. Furthermore, in the described embodiment, the sensing magnet 120 is illustrated as having four pole pairs. However, the number of pole pairs of the sensing magnet 120 is not limited to four. Similarly, when a rotor magnet is used as the magnet for position detection, the number of pole pairs of the rotor magnet is not limited to four.

Claims

1. An angle detection method for detecting the mechanical angle of a rotating shaft, characterized in that, include: The step of acquiring the signals output from three magnetic sensors that detect the change in magnetic flux caused by the rotation of the rotating shaft as sensor signals, wherein the three sensor signals have a phase difference of 120° with each other. The steps include selecting the intersection point where two of the three sensor signals intersect each other and the zero-crossing point where each of the three sensor signals intersects with the reference signal level within a mechanical angle period of 1 cycle; The step of generating a linear function θ(Δx) representing a straight line connecting adjacent intersection points to the zero intersection point, wherein Δx is the length from the starting point of the straight line to any point on the straight line, and θ is the mechanical angle corresponding to any point on the straight line; For multiple points on the straight line, the step of calculating the deviation between the mechanical angle θ calculated based on the linear function θ(Δx) and the mechanical angle θe obtained from the encoder set on the rotating shaft is taken as the first angle error; The step of storing the first angle error calculated for multiple points on the line as a learning value; and The step of generating a first angle error function to calculate the first angle error corresponding to any point on the line, based on the first angle error calculated for multiple points on the line; The angle detection method also includes: For multiple points on the straight line, the step of calculating the deviation between the mechanical angle θ calculated based on the linear function θ(Δx) and the first angle error calculated based on the first angle error function, and the mechanical angle θe obtained from the encoder, is taken as the second angle error; The step of storing the second angle error calculated for multiple points on the line as a learning value; and The step of generating a second angle error function to calculate the second angle error corresponding to any point on the line, based on the second angle error calculated for multiple points on the line.

2. The angle detection method according to claim 1, characterized in that, The points on the straight line are spaced equally.

3. An angle detection device for detecting the mechanical angle of a rotating shaft, characterized in that, include: Three magnetic sensors detect changes in magnetic flux caused by the rotation of the rotating shaft; and The signal processing unit processes the signals output from the three magnetic sensors. The signal processing unit performs the following processing: The signals output from the three magnetic sensors are acquired as sensor signals, and the three sensor signals have a phase difference of 120° with each other. Within a mechanical angle 1 cycle, select the intersection point where two of the three sensor signals intersect each other, and the zero-crossing point where each of the three sensor signals intersects with the reference signal level; Generate a linear function θ(Δx) representing the line connecting the adjacent intersection points to the zero intersection point, where Δx is the length from the starting point of the line to any point on the line, and θ is the mechanical angle corresponding to any point on the line. For multiple points on the straight line, the deviation between the mechanical angle θ calculated based on the linear function θ(Δx) and the mechanical angle θe obtained from the encoder set on the rotating shaft is calculated as the first angle error; The first angle error calculated for multiple points on the line is stored as a learning value; and Based on the first angle error calculated for multiple points on the line, a first angle error function is generated to calculate the first angle error corresponding to any point on the line. The signal processing unit further performs the following processing: For multiple points on the straight line, the deviation between the mechanical angle θ calculated based on the linear function θ(Δx) and the first angle error calculated based on the first angle error function, and the mechanical angle θe obtained from the encoder, is taken as the second angle error. The second angle error calculated for multiple points on the line is stored as a learning value; and Based on the second angle error calculated for multiple points on the line, a second angle error function is generated to calculate the second angle error corresponding to any point on the line.

4. The angle detection device according to claim 3, characterized in that, The points on the straight line are spaced equally.

Images