JP2013221827A - Signal processor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a signal processor capable of expanding an opportunity allowing to remove an offset error included in an excitation signal SC and modulated waves Sin, Cos.SOLUTION: An A/D converter 48 samples an excitation signal SC and modulated waves Sin, Cos at a sampling period obtained by dividing one period of the excitation signal SC by N (N is two or more integers) and calculates a correction value ΔV# (#=r, c, s) as the mean value of digital data of excitation signals SCs sampled during one period of the excitation signal SC when determined that the rotation of a rotor 10a is stopped. Correction parts 54a-54c correct the digital data of the excitation signals SCs by the correction value ΔV#.

Description

  The present invention relates to a signal processing apparatus including a modulated wave in which an AC excitation signal is amplitude-modulated according to position information of a position detection target and a demodulating unit that demodulates the position information based on the excitation signal.

  As this type of signal processing apparatus, for example, as can be seen in Patent Document 1 below, for a pair of modulated waves output from a resolver for detecting a rotation angle of a rotating body, an error that deviates by a predetermined value from an actual value. A technique for removing (offset error) is known. Specifically, in this technique, first, the maximum value and the minimum value of the modulated wave are sampled, and the average value of the modulated wave is calculated from the sampled maximum value and minimum value. The offset error is removed from the modulated wave by correcting the modulated wave based on the average value.

Japanese Patent Laying-Open No. 2005-208028

  Here, with the above technique, it is conceivable that the opportunity to correct the modulated wave is limited. This is because the timing at which the modulated wave reaches its maximum and minimum values rarely matches the sampling timing of the modulated wave. In particular, the higher the rotational speed of the rotation angle detection target, the smaller the number of samplings in the specified period, so there is an opportunity to match the timing at which the modulated wave reaches its maximum and minimum values with the sampling timing of the modulated wave. The decrease is remarkable.

  Another reason is that when the rotational speed of the rotating body is low, the offset error cannot be accurately grasped by the average value based on the maximum value and the minimum value. This is because when the rotational speed of the rotating body is low, distortion from the ideal waveform (sine wave) of the modulated wave increases.

  The offset error is not limited to the modulated wave but may be included in the excitation signal.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to provide a signal processing apparatus capable of expanding opportunities for correcting an offset error.

  In order to solve the above problem, the invention according to claim 1 is a modulated wave (Sin, Cos) in which an alternating excitation signal (Sc) is amplitude-modulated in accordance with position information (θ) of the position detection target (10a). And a sampling means (48) for sampling a plurality of the excitation signals during one period of the excitation signal, and a demodulating means for demodulating the position information based on the sampled excitation signal and the modulated wave ( 56, 58, 60, 62, 64, 66) and the correction target parameter which is at least one of the excitation signal and the modulated wave, over the M period (M is a positive integer) of the correction target parameter. Correction value calculation means (70) for calculating offset error correction values (ΔV #; # = r, c, s) from one or a plurality of sets of values sampled by the sampling means; And error correction means (54a, 54b, 54c) for correcting the sampled correction target parameter used for the demodulation based on the correction value calculated by the correction value calculation means.

  In the above invention, by providing the sampling means and the correction value calculation means, it is possible to calculate the correction value in which the offset error is quantified based on each sampling value. For this reason, the opportunity which can correct | amend the parameter for correction | amendment can be expanded compared with the technique described in the said patent document 1, for example.

1 is a system configuration diagram according to a first embodiment. FIG. 6 is a flowchart showing a procedure of offset correction value calculation processing according to the embodiment. The figure which shows the outline | summary of the calculation of the offset correction value at the time of a rotor stop. The figure which shows the outline | summary of the calculation of the offset correction value at the time of rotation of a rotor. The flowchart which shows the procedure of the offset correction value calculation process concerning 2nd Embodiment. The flowchart which shows the procedure of the offset correction value calculation process concerning 3rd Embodiment.

(First embodiment)
Hereinafter, a first embodiment in which a signal processing apparatus according to the present invention is applied to a digital converter of a resolver will be described with reference to the drawings.

  A motor generator 10 shown in FIG. 1 is an in-vehicle main machine, and is mechanically coupled to drive wheels (not shown). Inverter INV mediates transfer of electric power between motor generator 10 and a battery (not shown). The rotation angle θ (electrical angle) of the rotor 10 a of the motor generator 10 is detected by the resolver 20. In the present embodiment, as the resolver 20, a one-phase excitation two-phase output type including one primary coil 22 and two secondary coils 24 and 26 is used. Hereinafter, a configuration for detecting the rotation angle θ will be described.

  In the present embodiment, as a configuration for detecting the rotation angle θ, in addition to the resolver 20, an interface circuit 28 and a microcomputer (microcomputer 30) are provided. The primary side coil 22 of the resolver 20 is mechanically coupled to the rotor 10 a of the motor generator 10. The primary coil 22 is excited by a sinusoidal AC signal (excitation signal Sc). The excitation signal Sc is generated by the PWM generator 32 and the oscillator 34 built in the microcomputer 30, and the secondary delay element 36 and the amplifier circuit 38 built in the interface circuit 28. Specifically, the PWM generator 32 generates a PWM signal based on a clock output from the oscillator 34. The generated PWM signal is input to the secondary delay element 36, and the output voltage of the secondary delay element 36 is amplified by the amplifier circuit 38, whereby the excitation signal Sc is generated. In the present embodiment, the excitation signal Sc has a fixed frequency.

  When the excitation signal Sc is input to the primary coil 22, a magnetic flux is generated in the primary coil 22. The magnetic flux generated in the primary side coil 22 links the pair of secondary side coils 24 and 26. Here, the mutual inductance of each of the secondary side coils 24 and 26 and the primary side coil 22 is configured to periodically change in accordance with the rotation angle θ of the rotor 10a. As a result, the number of magnetic fluxes interlinking the secondary coils 24 and 26 changes periodically. In particular, in this embodiment, the phases of the voltages generated in the secondary side coils 24 and 26 are shifted from each other by “π / 2”. As a result, the output voltages of the secondary coils 24 and 26 become modulated waves obtained by modulating the excitation signal Sc with the modulated waves sin θ and cos θ, respectively. That is, when the excitation signal Sc is “sin ωt”, the modulated waves are “sin θ × sin ωt” and “cos θ × sin ωt”, respectively.

  The output voltage of the amplifier circuit 38 is voltage-converted by the differential amplifier circuit 40. On the other hand, the output voltage of the secondary side coil 24 is converted by the differential amplifier circuit 42, and the output voltage of the secondary side coil 26 is converted by the differential amplifier circuit 44. The output signals of these differential amplifier circuits 40, 42, 44 are input to an analog-digital converter (hereinafter referred to as A / D converter 48) of the microcomputer 30 via a low-pass filter 46 for removing noise. In the figure, among the voltages input to the A / D converter 48, the voltage corresponding to the output voltage of the differential amplifier circuit 40 is indicated by “SC” and the voltage corresponding to the output voltage of the differential amplifier circuit 42. Is indicated by “Sin”, and the one corresponding to the output voltage of the differential amplifier circuit 44 is indicated by “Cos”.

  The A / D converter 48 converts the excitation signal SC into digital data based on the clock output from the oscillator 34 (samples the excitation signal SC). The A / D converter 48 converts the modulated wave Sin into digital data based on the clock (samples the modulated wave Sin). Further, the A / D converter 48 converts the modulated wave Cos into digital data based on the clock (samples the modulated wave Cos).

  In the present embodiment, the differential amplifier circuit 40 is set so that the voltage range that the excitation signal SC and the modulated waves Sin and Cos can take is a predetermined input voltage range (for example, 0 to 5 V) of the A / D converter 48. , 42, 44 and the reference potentials of the excitation signal SC and the modulated waves Sin, Cos are determined. In particular, in the present embodiment, the reference potentials of the excitation signal SC and the modulated waves Sin and Cos should be unified to a specified potential Vb within the input voltage range (for example, 2.5 V which is the median value of the input voltage range). A common power supply 50 is connected to these differential amplifier circuits 40, 42 and 44. Incidentally, the reference potential of the excitation signal SC is, for example, an average value of this signal in one cycle of the excitation signal SC when the excitation signal SC is a sine wave. The reference potential of the modulated waves Sin, Cos is, for example, the average value of the modulated waves Sin, Cos in one rotation period of the rotor 10a when the rotation speed (electrical angular velocity) of the rotor 10a is constant. That is.

  In the present embodiment, a 12-bit specification is used as the A / D converter 48. In the present embodiment, the specified potential Vb is set to correspond to a predetermined digital unit (2048 LSB) of the A / D converter 48.

  The output signal of the A / D converter 48 is input to the central processing unit (CPU 52) where it is processed by software. In FIG. 1, among the software processing performed by the CPU 52, in particular, the calculation processing of the rotation angle θ and the like are shown in a block diagram.

  The correction units 54 a to 54 c correct the excitation signal SC and the modulated waves Sin and Cos sampled by the A / D converter 48. Specifically, the correction unit 54a corrects the excitation signal SC by adding or subtracting the excitation signal correction value ΔVr from the excitation signal SC, and the correction units 54b and 54c use the modulated wave Sin and Cos for the modulated wave. The modulated waves Sin and Cos are corrected by adding and subtracting the correction values ΔVs and ΔVc. In FIG. 1, the corrected excitation signal SC is indicated by reference “REF”, and the corrected modulated wave is indicated by “SIN, COS”. The correction of the excitation signal SC and the modulated waves Sin and Cos will be described in detail later.

  The cosine function multiplier 56 multiplies the output value (modulated wave SIN) of the correction unit 54b by a cosine function cos φ having the calculated value of the rotation angle θ (hereinafter, calculated angle φ) as an independent variable. On the other hand, the sine function multiplier 58 multiplies the output value (modulated wave COS) of the correction unit 54c by a sine function sin φ having the calculated angle φ as an independent variable. The control deviation calculation unit 60 calculates the control deviation ε by subtracting the output value of the sine function multiplier 58 from the output value of the cosine function multiplier 56.

This control deviation ε is expressed by the following equation (eq1) when the proportionality constant determined by the gains of the differential amplifier circuits 40, 42, and 44 and the amplifier circuit 38 is ignored.
ε = sinωt · sinθ · cosφ−sinωt · cosθ · sinφ
= Sinωt · sin (θ−φ) (eq1)
When the control deviation ε is “0”, the actual rotation angle θ matches the calculated angle φ. Here, the removal processing for removing the influence of the sign of the excitation signal Sc from the control deviation ε is performed by synchronous detection.

  That is, the reference REF is input to a detection signal generation unit 62 as a binary detection signal calculation unit, and here, a signal that becomes “1” or “−1” according to the comparison between the reference REF and “0”. Is processed into a detection signal Rd. Specifically, the detection signal generator 62 sets the detection signal Rd to “1” when the reference REF is “0” or more, and sets the detection signal Rd to “−1” when the reference REF is less than “0”. And

  The synchronous detection unit 64 calculates the detected wave amount εc by multiplying the control deviation ε by the detection signal Rd. The detected wave amount εc is “0” when the difference between the actual rotation angle θ and the calculated angle φ is “0”, and the calculated angle φ is greater than the actual rotation angle θ by the sign. Is an amount indicating whether the value is an advance side value or a retard side value.

  The detected wave amount εc is input to the angle calculation unit 66. The angle calculation unit 66 includes a low-pass filter and an integration element. In the present embodiment, in particular, an example including a proportional element, a double integral element, and a phase compensation filter “(bs + 1) / (as + 1)” is illustrated. Here, the double integral element is used in order to prevent a steady deviation from occurring in the calculated angle φ when the rotation angle θ changes at a constant speed.

  The calculated angle φ is input to the control processing unit 68 in addition to the cosine function multiplier 56 and the sine function multiplier 58. The control processing unit 68 generates an operation signal for the inverter INV based on a detection value of a current sensor (not shown) that detects a current flowing through the motor generator 10, a calculated angle φ, and the like, and outputs the operation signal to the inverter INV. Thereby, the control amount (for example, output torque) of motor generator 10 is controlled to the command value (for example, command torque).

  Incidentally, the control processing unit 68 has a function of outputting the first reference signal Sg1, which is a pulse signal, every time it is determined that the reference REF is zero-crossed. The output cycle of the first reference signal Sg1 is a half cycle of the reference REF. The control processing unit 68 also has a function of outputting the second reference signal Sg2, which is a pulse signal, every time it is determined that the calculated angle φ reaches a specified angle (for example, 0 °). The output cycle of the second reference signal Sg2 is the same cycle as one rotation cycle (360 °) of the rotor 10a.

  Incidentally, the excitation signal SC output from the differential amplifier circuit 40 and the modulated waves Sin and Cos output from the differential amplifier circuits 42 and 44 may include an offset error. The offset error is, for example, an error that deviates from the actual value of the excitation signal SC by a predetermined value when described using the excitation signal SC as an example.

  If an offset error is included in the excitation signal SC or the modulated waves Sin and Cos, the calculation accuracy of the calculation angle φ may be reduced. Specifically, for example, when the excitation signal SC includes an offset error, the sign of the detection signal Rd is different from the sign of the excitation signal SC included in the modulated waves SIN and COS. There is a possibility that the calculation accuracy is lowered.

  In order to solve such a problem, in this embodiment, the correction units 54a to 54c remove offset errors included in the excitation signal SC or the like. In the present embodiment, the excitation value correction value ΔVr and the modulation wave correction values ΔVs and ΔVc used in the correction units 54 a to 54 c are calculated by the correction value calculation unit 70 built in the microcomputer 30.

  FIG. 2 shows a procedure of correction value calculation processing executed by the correction value calculation unit 70. This process is repeatedly executed at a predetermined control cycle Tc, for example.

  In this series of processing, in step S10, the digital data R (n) of the excitation signal SC output from the A / D converter 48 and the digital data Sin (n), Cos (n) ( In the figure, these digital data are combined and expressed as A (n)). Here, in the present embodiment, as shown in FIG. 3, the value obtained by dividing one cycle Tref of the excitation signal Sc by the sampling cycle Tad in the A / D converter 48 is N (N is an integer of 2 or more). In addition, the sampling period Tad of the A / D converter 48 is set. In particular, in the present embodiment, the sampling cycle Tad is set to a cycle obtained by dividing one cycle Tref of the excitation signal Sc by 2 to the Lth power (L is a positive integer, L = 4 is illustrated in the figure). Yes. According to such a setting, when the division process is performed in step S20, which will be described later in FIG. 2, the divisor becomes “2 to the L power”, so that the digital data as the dividend is shifted by N bits. Thereby, the time required for the division process can be suitably shortened. Incidentally, FIG. 3 shows the transition of the excitation signal Sc and the modulated waves Sin and Cos and the transition of the first reference signal Sg1 when the rotation of the rotor 10a is stopped.

  Returning to the description of FIG. 2, in the subsequent step S12, it is determined whether or not the rotation of the rotor 10a is stopped. Here, whether or not the rotation is stopped may be determined based on the calculated angle φ, for example.

  If an affirmative determination is made in step S12, the excitation signal correction value ΔVr is calculated based on the excitation signal SC sampled by the A / D converter 48 in steps S14 to S22, and the sampled modulated signal is modulated. Modulated wave correction values ΔVs and ΔVc are calculated based on the waves Sin and Cos. Specifically, if a negative determination is made in step S14, the process proceeds to step S16, and the first reference signal Sg1 is obtained by adding the current digital data A (n) to the previous integral value S (n-1). The integrated value S (n) of the digital data from the previous detection to the current control cycle is calculated.

  In subsequent step S18, the previous sampling count Nad (n-1) is incremented by 1, thereby calculating the sampling count Nad (n) from the first detection of the first reference signal Sg1 to the current control cycle.

  On the other hand, if an affirmative determination is made in step S14, it is determined that the integration of a set of digital data sampled over one period of the excitation signal Sc has been completed, and the process proceeds to step S20. In step S20, the correction value ΔV # (# = r, c, s) is calculated by dividing the integral value S (n) by the number of samplings Nad (n). In step S22, the integral values S (n) and S (n-1) and the AD conversion times Nad (n) and Nad (n-1) are initialized.

  The correction value ΔV # calculated in this way is equal to the offset error as shown in FIG. The calculated correction value ΔV # is output to the correction units 54a to 54c, whereby the excitation signal SC and the modulated waves Sin and Cos sampled by the A / D converter 48 are corrected, and these digital data are used. Offset error is removed.

  If a negative determination is made in step S12, it is determined that the rotor 10a is rotating, and the correction value ΔV # is calculated in steps S24 to S32. Here, the modulation wave correction value ΔV ¥ (¥ = c, s) is calculated based on one set of digital data of the modulation waves Sin and Cos in one rotation period Trv of the rotor 10a. As shown in FIG. 4, during rotation of the rotor 10a, the offset error of the modulated wave can be quantified based on the integrated value of the digital data of the modulated wave in one rotation period Trv, while the excitation signal This is because the offset value of the modulated wave cannot be quantified with the integral value of one cycle of Sc. FIG. 4 is a diagram illustrating the transition of the excitation signal Sc and the modulated wave Sin and the transition of the output state of the second reference signal Sg2 when the rotor 10a is rotated.

  Returning to the description of FIG. 2, when a negative determination is made in step S24, the process proceeds to step S26, and the current digital data Sin (n), Cos (n) is added to the previous integral value S (n-1). Thus, the integrated value S (n) of the digital data of the modulated waves Sin and Cos from the time when the second reference signal Sg2 was detected last time to the current control period is calculated. In step S28, the previous sampling count Nad (n-1) is incremented by 1, thereby calculating the sampling count Nad (n) from the previous detection of the second reference signal Sg2 to the current control cycle. .

  On the other hand, if an affirmative determination is made in step S24, it is determined that integration of a set of digital data sampled over one rotation period Trv of the rotor 10a has been completed, and the process proceeds to step S30. In step S30, the modulation wave correction value ΔV ¥ (¥ = c, s) is calculated by dividing the integral value S (n) by the number of samplings Nad (n). In step S32, the integral values S (n) and S (n-1) and the AD conversion times Nad (n) and Nad (n-1) are initialized.

  Incidentally, the excitation signal correction value ΔVr is digital data of the excitation signal in one cycle Tref of the excitation signal Sc, as described in steps S12 to S22, even in a situation where a negative determination is made in step S12. It is calculated based on one set of

  In addition, when the process of step S18, S22, S28, S32 is completed, this series of processes is once complete | finished.

  According to the embodiment described in detail above, the following effects can be obtained.

  (1) When it is determined that the rotation of the rotor 10a is stopped, the excitation signal correction value is obtained as an average value of the excitation signal SC and the digital data of the modulated waves Sin and Cos over one cycle Tref of the excitation signal Sc. ΔVr and modulated wave correction values ΔVs and ΔVc were calculated. On the other hand, when it is determined that the rotor 10a is rotating, the modulated wave correction values ΔVs and ΔVc are calculated as average values of the digital data of the modulated waves Sin and Cos over one rotation period Trv of the rotor 10a. did. For this reason, for example, compared with the technique described in Patent Document 1, it is not necessary to sample the maximum value and the minimum value of the modulated waves Sin and Cos, so that the sampled excitation signal SC and the modulated wave are sampled. Opportunities for correcting Sin and Cos can be expanded.

  Furthermore, according to the correction method according to the present embodiment, each of the sampled excitation signal SC and the digital data of the modulated waves Sin and Cos is obtained regardless of whether the rotor 10a is rotating or stopped. It can also be corrected individually.

  (2) The sampling period of the digital data is set to a period obtained by dividing one period of the excitation signal Sc by 2 to the Lth power (L is a positive integer). Thereby, the time required for the calculation process of the correction value ΔV # (# = r, c, s) can be suitably shortened.

(Second Embodiment)
Hereinafter, the second embodiment will be described with reference to the drawings with a focus on differences from the first embodiment.

  In the present embodiment, the correction value ΔV # when the rotation of the rotor 10a is stopped is sampled over the M period (M is an integer of 2 or more) of the excitation signal Sc or the M rotation period of the rotor 10a. It is calculated as an average value of the digital data (M sets of digital data).

  FIG. 5 shows a procedure of correction value calculation processing according to the present embodiment. This process is repeatedly executed by the correction value calculation unit 70, for example, at a predetermined control cycle Tc. In the process of FIG. 5, the same processes as those shown in FIG. 2 are given the same reference numerals for the sake of convenience.

  In this series of processing, when it is determined in step S12 that the rotation of the rotor 10a is stopped, in steps S14a to S22, the excitation signal SC and the modulated signal sampled over the M cycles of the excitation signal Sc. A correction value ΔV # is calculated as an average value of the digital data of the waves Sin and Cos. Here, in step S14a, whether or not the M period of the excitation signal Sc has elapsed is determined based on whether or not the number of detection times of the first reference signal Sg1 has become “2 × M”.

  On the other hand, if a negative determination is made in step S12, a correction value ΔV ¥ is obtained as an average value of the digital data of the modulated waves Sin and Cos sampled over the M rotation period of the rotor 10a in steps S24a to S32. (¥ = c, s) is calculated. Here, in step S24a, whether or not the M rotation period of the rotor 10a has elapsed is determined based on whether or not the number of detection times of the second reference signal Sg2 has become "M".

  In addition, when the process of step S18, S22, S28, S32 is completed, this series of processes is once complete | finished.

  According to the calculation method of the correction value ΔV # described above, for example, even when noise is mixed in digital data, it is possible to further obtain an effect that the influence of noise on the calculation of the correction value ΔV # can be reduced. it can.

(Third embodiment)
Hereinafter, the third embodiment will be described with reference to the drawings with a focus on differences from the first embodiment.

  In this embodiment, when it is determined that the operation state of the motor generator 10 is a transient operation state, the calculation method of the modulation wave correction value ΔV ¥ (¥ = c, s) is changed. Specifically, the excitation signal correction value ΔVr is used as the modulation wave correction value ΔV ¥. This is because, under the situation where the operation state of the motor generator 10 is a transient operation state, the offset error cannot be accurately quantified depending on the average value of the digital data of the modulated waves Sin and Cos, but the digital of the excitation signal SC. In view of the average value of the data, it is possible to quantify the common offset error among the offset errors included in each of the excitation signal SC and the modulated waves Sin and Cos. Here, the common offset error is caused by, for example, an individual difference of the power supply 50, and the output potential of the common power supply 50 connected to the differential amplifier circuits 40, 42, 44 becomes the excitation signal SC and the modulated wave Sin. , Cos is generated by deviating from an appropriate potential for setting the reference potential of Cos to the specified potential Vb.

  FIG. 6 shows a procedure of correction value calculation processing according to the present embodiment. This process is repeatedly executed by the correction value calculation unit 70, for example, at a predetermined control cycle Tc under a situation in which it is determined that the rotor 10a is rotating.

  In this series of processes, first, in step S34, it is determined whether or not the absolute value of the change amount Δω of the rotational speed of the rotor 10a exceeds a specified amount Δωth (> 0). This process is a process for determining whether or not the operation state of the motor generator 10 is a transient operation state.

  If an affirmative determination is made in step S34, the process proceeds to step S36, where the excitation signal correction value ΔVr is output to the correction units 54b and 54c, and the modulated wave correction value ΔV shown in FIG. The calculation of ¥ (¥ = c, s) is stopped. That is, when the motor generator 10 is in a transient operation state, the correction values output from the correction value calculation unit 70 to the correction units 54a to 54c are all excitation signal correction values ΔVr.

  If a negative determination is made in step S34, or if the process of step S36 is completed, this series of processes is temporarily terminated.

  Thus, in the present embodiment, when the operation state of the motor generator 10 is a transient operation state, the excitation signal correction value ΔVr is used as the modulation wave correction value ΔV ¥ (¥ = c, s). . For this reason, the correction of the modulated wave can be continued using the excitation signal correction value ΔVr even in a situation where the calculation accuracy of the modulated wave correction value ΔV ¥ is lowered.

(Other embodiments)
Each of the above embodiments may be modified as follows.

  In the first embodiment, the specified potential Vb (for example, 2.5 V) may be set to correspond to “0LSB” of the A / D converter 48. In addition to these settings, the sampling period is set to a period obtained by dividing one period of the excitation signal Sc by the Lth power of 2 (L is a positive integer), so that the excitation signal sampled over one period of the excitation signal Sc. The correction value ΔV # (# = r, c, s) can be calculated as an addition value of the modulated waves Sin and Cos sampled over one rotation period of the SC and the rotor 10a. That is, since the correction value ΔV # can be calculated only by the addition command without performing a division command, the correction value calculation process can be simplified.

  In the second embodiment, the excitation signal correction value ΔVr is calculated as the average value of the sampling values of the excitation signal over a plurality of consecutive periods of the excitation signal. However, the present invention is not limited to this. For example, with respect to the excitation signal, the excitation signal correction value ΔVr may be calculated as an average value of the sampling values of the excitation signals in periods separated from each other. The same applies to the modulated wave.

  -As a binary detection signal calculation means, it is not restricted to what was illustrated to the said 1st Embodiment. For example, the detection signal Rd may be “1” when the reference REF is greater than “0”, and the detection signal Rd may be “−1” when the reference REF is “0” or less. .

  The power source for setting the reference potential of the excitation signal SC and the modulated waves Sin and Cos to the specified potential Vb is not limited to the common power source 50. For example, different power sources may be connected to the differential amplifier circuits 40, 42, 44. Even in this case, when the excitation signal SC and the modulated waves Sin and Cos include an offset error other than the common offset error, the excitation signal SC and the modulated signal are corrected by the correction method described above with reference to FIGS. It is effective to correct the waves Sin and Cos. In this case, the correction method shown in FIG. 6 cannot be adopted.

  The determination means is not limited to determining whether or not the operation state of the motor generator 10 is a transient operation state based on the rotation speed. For example, the absolute value of the torque change amount of the motor generator 10 is a predetermined value ( > 0) or more.

  The control deviation ε may be a value “sin (θ + φ)” calculated as the sum of the output value of the cosine function multiplier 56 and the output value of the sine function multiplier 58. In this case, since the calculation angle φ is calculated as a negative value, it is only necessary to reverse the sign of the calculation angle φ and grasp the actual rotation angle.

  The demodulation of the modulated wave is not limited to the method using the detection signal Rd, and for example, a method using a value obtained by directly multiplying the reference deviation REF by the control deviation ε may be employed.

  -The resolver is not limited to the one-phase excitation two-phase output type. For example, a two-phase excitation two-phase output type may be used. In this resolver, by inputting an alternating excitation signal having the same amplitude and a phase difference of 90 degrees to each of the pair of primary coils, the phase from each of the pair of secondary coils is mutually different. A pair of modulated waves that differ by 90 degrees and correspond to the rotation angle of the rotor are output.

  The delay element of the interface circuit 28 for generating the excitation signal is not limited to the secondary delay element 36 but may be, for example, a primary delay element.

  The sampling period Tad is not limited to a period obtained by dividing one period of the excitation signal by N (N is an integer of 2 or more). For example, the sampling period Tad is not synchronized with the period obtained by dividing one period of the excitation signal by N on the condition that a plurality of sampling values capable of quantifying the offset error are obtained over one period of the excitation signal. It is good.

  The correction target parameter is not limited to both the excitation signal SC and the modulated waves Sin and Cos. For example, it may be at least one (excluding all) of the excitation signal SC and the modulated waves Sin and Cos.

  The position information of the position detection target is not limited to the rotation angle of the rotor 10a. In short, other position information may be used as long as it is position information capable of amplitude-modulating the excitation signal according to the position information. Even in this case, the application of the correction method described above is effective if the position information includes an offset error.

  DESCRIPTION OF SYMBOLS 10a ... Rotor, 20 ... Resolver, 48 ... A / D converter, 54a-54c ... Correction | amendment part, 56 ... Cosine function multiplier, 58 ... Sine function multiplier, 60 ... Control deviation calculation part, 62 ... Detection signal production | generation Reference numeral 64: Synchronous detection unit 66: Angle calculation unit 70: Correction value calculation unit

Claims (7)

A modulated wave (Sin, Cos) in which an AC excitation signal (Sc) is amplitude-modulated in accordance with position information (θ) of the position detection target (10a) and the excitation signal during one cycle of the excitation signal. Sampling means (48) for sampling a plurality of times;
Demodulation means (56, 58, 60, 62, 64, 66) for demodulating the position information based on the sampled excitation signal and the modulated wave;
For a correction target parameter that is at least one of the excitation signal and the modulated wave, one or more sets of values sampled by the sampling means over M periods (M is a positive integer) of the correction target parameter Correction value calculation means (70) for calculating a correction value (ΔV #; # = r, c, s) of the offset error from the set;
Error correction means (54a, 54b, 54c) for correcting the sampled correction target parameter used for the demodulation based on the correction value calculated by the correction value calculation means;
A signal processing apparatus comprising: The correction target parameter includes the excitation signal,
The correction value calculating means calculates a correction value (ΔVr) of an offset error for the excitation signal from one set or a plurality of sets of the sampled excitation signal values over M cycles of the excitation signal;
The signal processing apparatus according to claim 1, wherein the correction unit corrects the excitation signal based on a correction value of an offset error for the excitation signal. The position information of the position detection target is a rotation angle (θ) of the rotating body (10a),
The correction target parameter includes the modulated wave,
The correction value calculation means calculates a correction value (ΔV ¥; ¥) of the offset error for the modulated wave from one or a plurality of sets of the sampled values of the modulated wave over the M rotation period of the rotating body. = C, s)
The signal processing apparatus according to claim 1, wherein the correction unit corrects the modulated wave based on a correction value of an offset error for the modulated wave. The position information of the position detection target is a rotation angle (θ) of the rotating body (10a),
The correction target parameters are both the excitation signal and the modulated wave,
Each of the excitation signal and the modulated wave is set to a reference potential (Vb) by a common power source (50) before sampling by the sampling means,
A judgment means for judging whether or not the rotational state of the rotating body is a transient state;
When the correction means is determined to be in the transient state, the excitation signal is based on the correction value (ΔVr) of the offset error for the excitation signal calculated by the correction value calculation means using the excitation signal. The signal processing apparatus according to claim 1, wherein the modulated wave is corrected. The demodulating means includes
Binary detection signal calculation means (62) for calculating a binary detection signal (Rd) depending on whether the sampled excitation signal is 0 or more or greater than 0. )
5. The signal processing apparatus according to claim 4, wherein the position information is demodulated based on the detection signal and the modulated wave calculated by the binary detection signal calculation means.   The sampling means samples the excitation signal and the modulated wave at a sampling period (Tad) obtained by dividing one period (Tref) of the excitation signal by N (N is an integer of 2 or more). Item 6. The signal processing device according to any one of Items 1 to 5.   The signal processing apparatus according to claim 6, wherein the sampling period is a period obtained by dividing one period of the excitation signal by L to the power of 2 (L is a positive integer).

Images