JP2013029529A - Signal processing device and measuring device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a signal processing device which is inexpensive and has high reliability.SOLUTION: There is provided the signal processing device which obtains a combination of a wavenumber and a phase associated with displacement of a body to be measured based upon a plurality of period signals corresponding to the displacement, the signal processing device including: first means of obtaining the phase from the plurality of period signals each time the plurality of period signals are sampled; second means of obtaining a variation quantity of the phase in a sampling interval by carrying out regression computation on a combination of a wavenumber and a phase based upon the phase obtained by the first means each time sampling is performed; third means of predicting a combination in second sampling by adding the variation quantity to the combination in first sampling; fourth means of obtaining a prediction error of the combination predicted by the third means from the phase of the second sampling obtained by the first means; and fifth means of obtaining the combination of the second sampling by adding the prediction error to the combination predicted by the third means.

Description

  The present invention relates to a signal processing device, and more particularly to a signal processing device used in a measuring device that measures the position or angle of an object to be measured.

  An encoder as a measuring device generally measures the amount of light transmitted through a scale provided with a light transmitting portion and a light shielding portion at a constant pitch. In addition, a laser interferometer as a measuring device generally divides a laser beam into two light beams, one of which is reflected by a mirror provided at a movable part, and the other is reflected by a mirror provided at a fixed part. And the intensity of these interference lights is measured.

  In any measuring apparatus, a sinusoidal signal whose phase changes in accordance with the displacement of the position or angle of the object to be measured is output. These measuring devices are generally approximated by a sine function and a cosine function and output a two-phase signal whose phases are different from each other by 90 degrees, or a device that outputs three-phase signals whose phases are different from each other by 120 degrees, etc. This device outputs a plurality of signals having different phases. In general, a signal processing device performs processing such as arctangent calculation on these signals to obtain a signal phase to obtain fine position information, and counts the wave number of the signal to obtain a coarse position. Get a signal.

  Conventionally, the wave number of a signal is generally counted by converting a sine signal and a cosine signal into a binary signal by a comparator and determining the moving direction from the rising and falling edges of one signal and the level of the other signal. And the wave number was counted with an up / down counter.

  FIG. 5 is an example of a conventional signal processing device (measurement device). Reference numeral 60 denotes an encoder which is detection means for detecting a position or angle and outputting a predetermined signal. The encoder 60 outputs a signal proportional to cos (nθ) and sin (nθ) according to the rotation angle θ of the shaft. Here, n is a signal cycle per rotation of the encoder 60, and the signal cycle is generally several hundred to several thousand.

  The output signal from the encoder 60 is amplified to an appropriate level by the operational amplifiers 2-1 and 2-2. These amplified signals are adjusted in impedance by operational amplifiers 3-1 and 3-2 having a first voltage follower configuration, and then pass through low-pass filters 4-1 and 4-2, and then an A / D converter 5-1. The digital signal is converted by 5-2. The output of the digital signal is input to the digital signal processor 6 (signal processing device), and the phase (phase angle) is calculated by arctangent calculation.

  By the way, the encoder 60 outputs hundreds to thousands of similar signals per one rotation of the shaft, and in practice, it is necessary to identify which signal period corresponds to the current angle. For this reason, the output signal from the encoder 60 is amplified by the operational amplifiers 2-1 and 2-2, then passes through the operational amplifiers 7-1 and 7-2 having the second voltage follower configuration, and the comparators 8-1 and 8-2. Is binarized. The wave number is counted by the up / down counter 11 using these binary signals.

  For counting the wave number, both binary signals are sampled in synchronization with the clock by D-FFs 9-1 and 9-2 (delay flip-flops). The output (A phase) of the binarized signal from the D-FF 9-1 is further input to the other D-FF 9-3.

  When the output of the first stage D-FF 9-1 is L and the output of the second stage D-FF 9-3 is H, that is, the rising edge of the A phase signal, and the B phase signal is H. In some cases, the AND gate 10-1 increments the up / down counter 11. When the output of the first stage D-FF 9-1 is H and the output of the second stage D-FF 9-3 is L, that is, the fall of the A phase signal, and the B phase signal is H. In some cases, the AND gate 10-2 decrements the up / down counter 11.

  The output of the up / down counter 11 is input to the digital signal processor 6, and phase information expanded by digits is formed therein by combining with the phase information.

  As described above, a large number of parts are required to perform the digit extension of the phase information with high accuracy. For example, operational amplifiers 7-1 and 7-2 having a voltage follower configuration are inserted in front of the comparators 8-1 and 8-2. If this is omitted, a signal that is positively fed back in order to give hysteresis to the operational amplifier will flow backward through the input resistance of the comparator, and will eventually adversely affect the A / D converters 5-1 and 5-2. .

  If the first stage D-FFs 9-1 and 9-2 that receive the binarized signal are omitted, an extremely narrow signal may be input to the up / down counter 11 depending on the operation timing of the clock signal and the comparator. Yes, it causes malfunction.

  In recent years, A / D converters and signal processing devices that calculate phase angles have been speeded up, and digit expansion is possible without using a comparator or the like.

  FIG. 6 is a block diagram of a conventional signal processing apparatus used for digit expansion. The encoder signal that has been A / D converted and input to the signal processing device is converted into a phase by the phase calculator 70. Inside the signal processing device, a register 80 is provided for holding digit-extended phase information.

  When the phase is newly calculated by the phase calculator 70, the subtracter 85 calculates the amount of phase displacement by subtracting the phase information held in the register 80 from the calculated new phase information. The adder 87 adds the value obtained by digit-expanding the phase displacement amount to the phase-extended phase information held in the register 80. In this way, the value held in the register 80 is updated to the latest digit-extended phase information.

  Patent Document 1 discloses an optical encoder that outputs angle data by combining upper data obtained by counting the number of slits and lower data obtained by electrically dividing one sine wave. It is disclosed.

JP 2003-35569 A

  When the phase information is updated using the above-described method, there is a maximum displacement amount that can be expanded. That is, the amount of displacement is limited to ½ of the maximum value of the phase information, and if the shaft rotates at a speed exceeding this value, it will cause a malfunction.

  For example, in the case of an encoder that outputs a position signal of 1000 cycles per rotation and the maximum rotation speed is 6,000 rpm (100 rps), the maximum frequency of the signal is 100 kHz. In this case, if phase calculation and digit expansion are performed at the same speed using an A / D converter with a sampling frequency of 200 kHz, the amount of change in the phase signal in each operation cycle is limited to ½ cycle or less, resulting in malfunction. Does not occur.

  Such an apparatus has the advantage that the number of parts is reduced and a small apparatus can be generally constructed at low cost. However, there is a reliability problem that miscount occurs when the maximum allowable speed is exceeded due to unexpected disturbance.

  In addition, when using a high-resolution encoder, the maximum speed is more severely limited. When an arithmetic unit capable of high-speed processing using a high-speed A / D converter is used to alleviate this, the cost increases. There was a problem of becoming.

  The present invention has been made to avoid the above-described problems, and provides an inexpensive and highly reliable signal processing apparatus.

  A signal processing device according to one aspect of the present invention is a signal processing device that obtains a combination of a wave number and a phase related to the displacement based on a plurality of periodic signals corresponding to the displacement of the object to be measured. A sampling interval is obtained by performing a regression operation on a combination of a wave number and a phase based on a phase obtained by the first means for each sampling and a phase obtained from the plurality of periodic signals for each sampling of the signal. Obtained by the second means for obtaining the phase change amount in the first sampling, a third means for adding the change amount to the combination in the first sampling to predict the combination in the second sampling, and the first means. The fourth means for obtaining the prediction error of the combination predicted by the third means from the phase at the second sampling, and the prediction by the third means. Has been said by adding the prediction error to the combination and a fifth means for obtaining said combination in said second sampling.

  Other objects and features of the present invention are illustrated in the following examples.

  According to the present invention, an inexpensive and highly reliable signal processing apparatus can be provided.

It is a block diagram of the signal processing apparatus in a present Example. It is a block diagram of the regression calculator in a present Example. It is a schematic block diagram of the encoder in a present Example. It is the (a) side view and (b) front view of the laser interferometer in a present Example. It is an example of the conventional signal processing apparatus. It is a block diagram of the conventional signal processing apparatus.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In each figure, the same members are denoted by the same reference numerals, and redundant description is omitted.

First, the configuration of the measuring apparatus in the present embodiment will be described. The measuring apparatus according to the present embodiment is a measuring apparatus that measures the position (position or angle) of an object to be measured, a detection unit that detects a signal corresponding to the position of the object to be measured, and the object to be measured based on the signal And a signal processing device for determining the position of. Examples of the measurement apparatus according to this embodiment include an encoder, a laser interferometer, and a resolver. In the present embodiment, particularly, typical configurations of an encoder and a laser interferometer will be described.
(Configuration of encoder 200)
First, the configuration of an encoder used as an example of a measuring device will be described. FIG. 3 is a schematic diagram of the encoder 200 in the present embodiment.

  The encoder 200 is an optical linear encoder, and measures a linear mechanical displacement amount of an object to be measured. However, the present embodiment is not limited to this, and can also be applied to a rotary encoder that measures the angle of an object to be measured.

  As shown in FIG. 3, the encoder 200 includes a movable scale 90, a fixed scale 120, a light emitting element (light emitting diode) 140, and a light receiving element (photodiode) 150.

  The movable scale 90 is configured to be linearly movable with the object to be measured. On the other hand, the fixed scale 120 is fixed. The encoder 200 has a configuration in which a movable scale 90 and a fixed scale 120 are disposed between the light emitting element 140 and the light receiving element 150. The movable scale 90 is provided with a slit 100 having a constant width in order to measure the distance moved.

  The fixed scale 120 is disposed to face the movable scale 90 and has fixed slits 130 having the same pitch. The right half and the left half of the fixed scale 120 are provided with openings at positions that are 90 degrees out of phase, that is, at positions that differ by a quarter of the scale pitch. A light receiving element 150 is provided on the back surface of the fixed scale 120, that is, the surface opposite to the surface on which the movable scale 90 is disposed. The light receiving element 150 has two light receiving portions corresponding to the positions of the right half and the left half of the fixed scale, and outputs a cosine signal and a sine signal having phases different from each other by 90 degrees.

  The light emitting element 140 is provided on the back surface of the movable scale 90, that is, the surface opposite to the surface on which the fixed scale 120 is disposed. In order to measure the displacement length of the object to be measured, the light emitting element 140 is always lit. The light of the light emitting element 140 is transmitted or blocked as the movable scale 90 moves.

As described above, the cosine signal and the sine signal obtained by the detection unit of the encoder 200 are supplied to the signal processing apparatus 1 described later, and the position or angle of the measurement object is measured.
(Configuration of laser interferometer 300)
Next, the configuration of the laser interferometer 300 used as the measuring apparatus of this embodiment will be described. 4A is a side view of the laser interferometer 300, and FIG. 4B is a front view of the laser interferometer 300.

  In the laser interferometer 300, a 0.85 μm surface emitting (VCSEL) laser having a stable laser wavelength λ is used as a highly coherent single mode semiconductor laser LD (semiconductor laser LD). The light beam from the semiconductor laser LD becomes collimated light (parallel light) by the collimator lens COL1. Then, the lens LNS1 is focused and illuminated at a position P1 on the focal plane of the lens LNS2 via the half mirror NBS.

  The light beam from the position P1 is emitted from the lens LNS2 as a parallel light beam whose optical axis is slightly oblique. Further, a polarization beam splitter (light splitting means) PBS is used to separate the polarized light component into two light beams. The reflected light (S-polarized light) from the polarizing beam splitter PBS is incident on the reference mirror M1 (reference surface), and the transmitted light (P-polarized light) from the polarizing beam splitter PBS is incident on the mirror M2 (measurement object).

  Then, the respective reflected lights are combined through the polarization beam splitter PBS, condensed and illuminated at the position P2 on the focal plane of the lens LNS2, and returned to the original optical path by the reflection film M0 provided in the vicinity thereof. The reflected light from the position P2 is emitted as a parallel light beam from the lens LNS2, separated into two light beams by the polarization beam splitter PBS, illuminates the reference mirror M1 with the reflected light (S-polarized light), and transmitted light (P-polarized light). The mirror M2 (measurement object) is illuminated.

  Each reflected light collects and illuminates the position P1 of the focal plane of the lens LNS2 via the polarization beam splitter PBS.

  From there, the luminous flux is extracted to the light source side. (S-polarized light makes two reciprocations between the reference mirror M1 and the beam splitter PBS, and P-polarized light makes two reciprocations between the mirror M2 (object to be measured) and the beam splitter PBS). These light beams are extracted to the light receiving system side by a non-polarizing beam splitter (half mirror) NBS, transmitted through the quarter-wave plate QWP, and converted into linearly polarized light whose polarization azimuth rotates in accordance with a change in phase difference.

  This light beam is split into three light beams by a beam splitting element GBS via a condenser lens CON and an aperture AP. The three light beams are incident on a polarizing element array 3CH-POL that is arranged with the transmission axes shifted from each other by 60 degrees. The light that has passed through the polarizing element array 3CH-POL is incident on the light receiving portion of the three-divided light receiving element PDA.

As described above, the detection means of the laser interferometer 300 detects three interference signals UVW whose phases based on the out-of-plane displacement of the mirror M2 (object to be measured) are shifted by 120 degrees. The interference signal UVW detected by the detection means is input to the signal processing device 1.
(Configuration of signal processing apparatus 1)
Next, the signal processing apparatus in the present embodiment will be described. FIG. 1 is a block diagram of a signal processing apparatus 1 in the present embodiment. The signal processing device 1 is a signal processing device used in a measuring device that measures the position or angle of an object to be measured. In addition, since an angle is included in the concept of position, in the following description, “position or angle” is collectively referred to as “position”.

  As shown in FIG. 1, the signal processing apparatus 1 includes a phase calculator 12, registers 16, 17, 18, a regression calculator 19, adders 21, 22, and subtractors 23, 24.

  The phase calculator 12 calculates a signal phase (measurement phase) by performing processing such as arctangent calculation on the signal output according to the position of the object to be measured at a predetermined sampling interval. For example, the phase calculator 12 determines the position of the object to be measured based on the cosine signal and sine signal detected by the encoder 200 or the three interference signals detected by the laser interferometer 300. The phase of the corresponding signal is calculated.

  The register 16 is a predicted position holding unit for holding (storing) the predicted position (predicted digit extended phase information) of the measurement object. As will be described later, the predicted position of the object to be measured is obtained by adding the speed information (phase displacement amount at one sampling interval) calculated by the regression calculator 19 to the current position stored in the register 17. .

  Here, assuming that the sampling next to the first sampling is the second sampling, the predicted position of the measured object is the predicted position of the measured object at the time of the second sampling. In addition, the current position of the measurement object is the position of the measurement object at the time of the first sampling.

  The position of the object to be measured (displacement with respect to the origin) can be represented by a value (wave number) corresponding to a multiple of one period of the periodic signal and a value (phase) less than one period of the periodic signal. For this reason, the register 16 includes a wave number part that holds the wave number and a phase part that holds the phase, and the predicted position is represented by a combination of the wave number part and the phase part.

  Each of the wave number part and the phase part has a bit number of 10 bits, for example. However, the number of bits of the wave number portion and the phase portion is not limited to this, and can be appropriately changed according to the required accuracy. When calculation with higher accuracy is required, the number of bits of the wave number portion and the phase portion can be configured by 16 bits, for example. Hereinafter, in the present embodiment, the wave number portion may be referred to as an upper bit and the phase portion may be referred to as a lower bit.

  The measurement phase calculated by the phase calculator 12 is represented by the same number of bits as the phase portion (lower bits) of the register 16. For this reason, when new phase information is obtained by the phase calculator 12, the subtracter 23 subtracts the value held in the lower bits (phase part) of the register 16 from the measured phase calculated by the phase calculator 12. This value is a difference (prediction error) between the measurement phase and the predicted position held in the register 16.

  As described above, the subtractor 23 is an error calculating means for subtracting the predicted position of the object to be measured from the measurement phase at the time of the second sampling obtained by the phase calculator 12 to obtain a prediction error. More specifically, the subtractor 23 calculates a prediction error by subtracting the phase portion of the predicted position from the measurement phase.

  The prediction error calculated by the subtracter 23 is sign-extended (digit expansion), and is added to the prediction position held in the register 16 by the adder 21. The value calculated by the adder 21 represents the current position of the object to be measured. This value is held (stored) in the register 17. As described above, the adder 21 is a position calculation unit that obtains the position of the object to be measured at the time of the second sampling by adding the prediction error expanded by digits to the predicted position. More specifically, the adder 21 obtains the position of the object to be measured at the time of the second sampling by adding the prediction error expanded by digits to the wave number part and the phase part of the predicted position.

  The register 17 is a current position holding unit for holding (storing) the current position (current digit extended phase information) of the object to be measured. Similarly to the register 16, the register 17 includes a wave number part (upper bits) and a phase part (lower bits). The value held immediately before the new phase information is held in the register 17 is the previous position of the object to be measured. This previous position indicates the position of the object to be measured at the time of the previous sampling. Before the new phase information is stored in the register 17, the previous position is saved in the register 18.

  The register 18 is a previous position holding unit for holding (storing) the previous position (previous digit extended phase information) of the object to be measured. Similarly to the registers 16 and 17, the register 18 includes a wave number part (upper bit) and a phase part (lower bit). As described above, the register 18 holds the value (previous position) held in the register 17 immediately before the new phase information is stored in the register 17 as it is.

  The difference between the current position held in the register 17 and the previous position held in the register 18 is the amount of phase displacement. The subtracter 24 subtracts the previous position held in the register 18 (wave number part and phase part) from the current position held in the register 17 (wave number part and phase part). This subtraction value is input to the regression calculator 19.

  The regression calculator 19 is a regression calculation means for calculating the speed of the measured object by regressing the phase of the signal corresponding to the position of the measured object with respect to time. Such a phase of the signal is obtained at a predetermined sampling interval (time interval). For example, a Kalman filter is used as the regression calculator 19. The Kalman filter is a dynamic linear regression calculation device that calculates speed by multiplying a measurement phase by an exponential weighting factor. However, the present embodiment is not limited to this, and regression means other than the Kalman filter may be used.

  The regression calculator 19 calculates speed information (phase displacement amount at one sampling interval) of the measurement object by regression calculation. The regression calculator 19 calculates noise together with the speed information. This noise can be suppressed by using the information on the current position held in the register 17, but the description in the present embodiment is omitted.

  FIG. 2 is a block diagram of a Kalman filter, which is an example of the regression calculator 19 in the present embodiment.

  DP is the difference between the current position held in the register 17 and the previous position held in the register 18. The PDP is a register for storing the phase difference DP. The register PDP shifts right by a predetermined P bits and stores the phase difference DP.

  A value obtained by subtracting the value of the register FX from the value of the register PDP is stored in the register Q. Next, a value obtained by shifting the value of the register Q to the right by P bits is stored in the register PQ. Further, a value obtained by subtracting the value of the register PQ from the value of the register Q is stored in the register FX.

  A value obtained by adding the value of the register VX to the value of the register PQ is stored in the register B. The value stored in the register B indicates the slope (regression coefficient) in the regression equation, and the register B holds a value corresponding to the speed information (phase displacement speed). As will be described later, this speed information is used to calculate the predicted position of the object to be measured.

  A value obtained by shifting the value of register B to the right by P bits is stored in register PB. Next, a value obtained by subtracting the contents of the register PB from the value of the register B is stored in the register VX. A value C obtained by subtracting the value of the register VX from the value of the register FX becomes a constant term in the regression equation. This constant term can also be used for the purpose of removing noise.

  As shown in FIG. 1, the speed information calculated by the regression calculator 19 is added to the current position held in the register 17 by the adder 22. The adder 22 adds the speed calculated by the regression calculator 19 to the position (current position) of the measurement object at the time of the first sampling, thereby obtaining the predicted position of the measurement object at the time of the second sampling. It is a prediction calculation means. The second sampling is the next sampling after the first sampling.

  The velocity information output from the regression calculator 19 is a phase displacement amount at one sampling interval. Therefore, by adding this speed information to the position (phase) of the object to be measured at the first sampling, the position (phase) of the object to be measured at the next second sampling can be predicted. This predicted position is held in the register 16.

  In FIG. 1, the register 16 for holding the predicted position is provided, but the present embodiment is not limited to this configuration. It suffices if the output line of the adder 22 is provided in the portion corresponding to the register 16, and the register 16 need not be provided.

  As described above, the value held in the phase part (lower order bits) of the register 16 among the predicted positions held in the register 16 is used to obtain a prediction error. That is, by subtracting the predicted position (phase part) of the register 16 from the measured phase calculated by the phase calculator 12, an error (prediction error) between the measured phase that is the actual value and the predicted position that is the predicted value. Desired.

  In the conventional signal processing apparatus, when calculating the position of the measurement object, the difference between the measurement phase and the previous digit extended phase information (previous position) is calculated. At this time, if the phase displacement amount (speed) at one sampling interval is larger than ½ of the maximum value of the phase information, a malfunction occurs. For this reason, the range in which the displacement amount of the object to be measured per sampling can be calculated is limited to ½ of the maximum value of the phase information.

  On the other hand, the signal processing apparatus 1 according to the present embodiment calculates the difference between the measurement phase and the predicted digit extended phase information (predicted position). Since the predicted position is obtained using the velocity information calculated from the regression calculator 19, the error with respect to the measurement phase does not increase. For this reason, even when the measurement object is displaced at a higher speed, the position of the measurement object can be detected. In the signal processing apparatus 1 of the present embodiment, when calculating the difference between the measurement position and the predicted position, the prediction error is limited to ½ period or less, but is not limited to the speed.

  As described above, according to the present embodiment, when the digit extension of the phase signal is performed inside the signal processing apparatus, the speed limitation can be greatly relaxed with a simple configuration. Thus, according to this embodiment, it is possible to provide an inexpensive and highly reliable signal processing apparatus.

  The embodiment of the present invention has been specifically described above. However, the present invention is not limited to the matters described as the above-described embodiments, and can be appropriately changed without departing from the technical idea of the present invention.

DESCRIPTION OF SYMBOLS 1 Signal processing apparatus 12 Phase calculator 16, 17, 18 Register 19 Regression calculator 21, 22 Adder 23, 24 Subtractor 200 Encoder 300 Laser interferometer

Claims (4)

A signal processing device that obtains a combination of wave number and phase related to the displacement based on a plurality of periodic signals according to the displacement of the object to be measured,
First means for obtaining a phase from the plurality of periodic signals for each sampling of the plurality of periodic signals;
Second means for obtaining a change amount of the phase at a sampling interval by performing a regression operation on a combination of wave number and phase based on the phase obtained by the first means for each sampling;
A third means for predicting the combination in the second sampling by adding the amount of change to the combination in the first sampling;
A fourth means for obtaining a prediction error of the combination predicted by the third means from the phase at the second sampling obtained by the first means;
And a fifth means for obtaining the combination in the second sampling by adding the prediction error to the combination predicted by the third means.   The fourth means subtracts the phase of the combination predicted by the third means from the phase of the second sampling obtained by the first means to obtain the prediction error. The signal processing apparatus according to claim 1.   The signal processing apparatus according to claim 1, wherein the second means includes a Kalman filter. A measuring device for measuring the displacement of an object to be measured,
Detecting means for detecting a signal corresponding to the displacement and outputting a plurality of periodic signals;
A signal processing apparatus according to claim 1, wherein the signal processing apparatus according to claim 1 obtains a combination of a wave number and a phase related to the displacement based on the plurality of periodic signals.