CN1272600C - Phase differential nonsensitive counting method and device for node tracking mohr interference fringe signals - Google Patents

Abstract

The present invention relates to a counting subdivision method for cross point tracking type moire (interference) fringe signals which are not sensitive to phase difference. The present invention is characterized in that 1) segmentation is carried out for a signal period in a floating mode according to the polarity and the cross points of two paths of signals; 2) the peak value and the zero level value of the two paths of signals are tracked; 3) the cross points of the two paths of signals are tracked in a floating mode; 4) counting is carried out; 5) subdivision is carried out; 6) a large number is corrected; 7) the direction of the displacement of a moire fringe is changed. A device comprises an operational amplifier, an analog-digital conversion chip, a digital signal processor, a single-chip microcomputer, a decoder and a liquid crystal display, wherein the operational amplifier is connected with the digital signal processor by the analog-digital conversion chip; the digital signal processor is simultaneously connected with the single-chip microcomputer and the decoder; the single-chip microcomputer is connected with the liquid crystal display. The defects of the DC level shift of signals, the unequal amplitude values of the two paths of signals, the influence of the counting and the subdivision caused by the nonlinearity of the signals and no orthogonality of the signals can be simultaneously overcome by the present invention; besides, the counting errors can be eliminated. Without the orthogonality of the signals, accurate counting subdivision and accurate direction recognition can be realized.

Description

Intersection tracking moiré interference fringe signal counting and subdivision method and device insensitive to phase difference

Technical field:

The present invention relates to a precision measurement technology, in particular to a method for correct direction identification, counting and subdivision of optical moiré (interference) fringe photoelectric signals deviated from an orthogonal state, and a device designed by the method .

Background technique:

Ultra-precision machining technology is a new technology developed to meet the needs of modern cutting-edge technology. It is an extremely important foundation for technologies such as aviation, aerospace, electronics, atomic energy, machinery, and instrumentation. It plays a very important role in the manufacture of modern weapons. . The fringe measurement technology using Moiré (interference) fringes to realize precise displacement measurement is the most widely used and mature technology in ultra-precision machining. Moiré (interference) fringes are optical fringes that indicate alternate dark and white. Since the Edinburgh laboratory of the British Ferranti Company established the first working prototype using Moiré Fringe (Moiré Fringe) to measure displacement in 1953, many optical systems that use Moiré (interference) fringes to precisely measure displacement have emerged. The principle is: when the measured target produces a certain displacement, the optical system will generate a series of moiré (interference) fringes in proportion, and the measuring digital display circuit will count the whole cycle signal, and count the signal that cannot constitute a whole cycle. Subdivision, and then the precise displacement is obtained from the count value and the subdivision value. For example, when a signal period represents 1 μm, if the signal count value is 500 and the subdivision segment value is 0.32 signal period, then the total displacement is 500.32 μm. However, in the prior art, the counting of Moiré (interference) fringes and the high-multiple subdivision are based on two orthogonal signals. When the two signals deviate from the orthogonal state, not only will it cause subdivision errors, Reduce the subdivision multiple, and the situation that the displacement of less than one signal period can not be correctly identified. In addition, the existing conventional counting method can count correctly when the moiré fringes move in one direction, but when the movement of the moiré fringes is more complicated, the phenomenon of missing or over counting will occur.

Invention content:

The technical problem to be solved by the present invention is to overcome the defects of the above-mentioned prior art and provide an intersection tracking type Moire (interference) fringe signal counting and subdivision method which is insensitive to phase difference.

The technical problem of the present invention is solved by the following technical solutions. It is characterized in that its method comprises the following steps:

(1) Carry out floating segmentation of the signal period according to the polarity and intersection point of the two signals:

According to the relative amplitude and polarity of the two-way signals, a signal cycle is divided into 8 segments, and then two units are set in the digital signal processor as intersection registers, and the initial value is 0; during the measurement process, real-time judgment of the two-way The polarity and amplitude of the signal sampling data, calculate the section where the current sampling point is located, and compare it with the section where the previous sampling point is cached, each time when the sampling point enters section 2 from section 1 Or when entering section 6 from section 5, store the larger amplitude value of the two signals in an intersection register, and call this value intersection point I; when the sampling point enters section 4 or When entering section 8 from section 7, the one with the larger amplitude of the two signals is stored in another intersection register, and this value is called intersection point II; when the phase difference of the two signals changes, the intersection point The amplitudes of I and intersection point II also change accordingly, so as to realize the floating segmentation of a signal cycle;

(2) Set a unit in the digital signal processor as a reversible counter, and assign an initial value of 0, and then compare the section where the current sampling point is located with the section where the previous sampling point is cached, and the sampling point Count up once when entering section 2 from section 1 or entering section 6 from section 5, and the moiré (interference) fringes move forward; the sampling point enters section 3 from section 4 or enters section 8 from section 8 At segment 7, a countdown is performed, and the moiré (interference) fringes move in reverse, and the digital signal processor counts twice for each cycle of signal change;

(3) When starting the measurement, collect two-channel data, normalize them respectively, judge the section where the measurement starting point is located according to its polarity and size, and store the smaller one of the current sampling data into the starting point register During the measurement process, collect two-channel data, normalize them respectively, judge the section they are in, and cache the smaller one of the data into the termination point register; at the same time, track the two-channel signal Peak value and valley value, track the intersection point of the two signals and cache it in the corresponding register; map the obtained measurement start point, end point, intersection point I and intersection point II to the displacement data table pre-stored in the memory of the single-chip microcomputer respectively, and obtain its amplitude The displacement corresponding to the value, and then according to the section where the start point and the end point are located, combined with the intersection point I and intersection point II, calculate the displacement value L that is less than one signal cycle; if the correction value in the reversible counter is N, one signal cycle represents If the displacement equivalent is d, the current total measured displacement value _Ltotal is calculated as follows: _Ltotal =N·d/2+L.

The present invention has the following technical effects:

1. Through the floating tracking of the intersection of the two input signals, it can overcome the segment discrimination error caused by the phase difference change of the two signals in the traditional subdivision method;

2. There is no need for signal quadrature, as long as the two moiré (interference) fringe photoelectric signals are not completely in phase or out of phase, accurate counting subdivision and correct direction discrimination (including correct direction discrimination for less than one signal period displacement) can be realized , breaking through the limitation that the two signals of the traditional counting subdivision method must be orthogonal, and has stronger environmental adaptability and anti-interference ability;

3. It can simultaneously eliminate the influence of signal DC level drift, unequal amplitude of two signals, signal non-linearity and non-orthogonal signal on subdivision, and achieve higher subdivision accuracy.

4. The counting dynamic range can reach 0Hz-500KHz, and can process axial vibration displacement signals, with strong anti-interference;

5. High reliability and high subdivision multiple.

Description of drawings:

Fig. 1 is two-way non-orthogonal signal absolute value waveform and segmental figure of the present invention;

Fig. 2 is a structural block diagram of the device of the present invention.

Detailed ways:

The present invention is realized through the following steps, and it comprises:

1. According to the polarity and intersection point of the two signals, the signal cycle is divided into floating segments:

As shown in Figure 1, in order to correctly identify and subdivide the input signal, a signal cycle is divided into 8 segments according to Table 1. When the two signals are orthogonal, the intersection points are equal in amplitude, and the phases corresponding to each segment are equal; when the two signals are not orthogonal, the intersection point I (point A, C or E in Figure 1) and The position of intersection II (point B or D in Figure 1) changes and each segment stretches or compresses accordingly. However, the sum of the lengths of segments 1, 2, 3, and 4 is always equal to the sum of the lengths of segments 5, 6, 7, and 8, which correspond to half a signal cycle and do not change with the phase difference of the two signals.

Table 1 Segmentation of full-period signals Sin1 Polarity Sin2 Polarity |Sin1|-|Sin2| Selected signals for segmentation 1 + + <0 |Sin1| 2 + + >0 |Sin2| 3 + - >0 |Sin2| 4 + - <0 |Sin1| 5 - - <0 |Sin1| 6 - - >0 |Sin2| 7 - + >0 |Sin2| 8 - + <0 |Sin1|

2. Tracking of the peak value and zero level value of the two signals:

Before the measurement starts, set a self-test program so that the moiré fringes have a displacement greater than one cycle. At the same time, two input signals are collected at high speed, and their peak and valley values are respectively judged and buffered. Calculate the zero-level value and maximum value of the current two input signals, which are used for normalization of the sampled data in the first cycle at the beginning of the measurement. During the measurement process, it is judged in real time whether the signal has moved a cycle according to the counting of large numbers, and at the same time, it judges and caches the peak and valley values of the two signals in the cycle for normalization of the sampled data in the next cycle. The initial state of the signal is determined by self-inspection, and the peak and valley values of the two input signals are tracked in real time during the measurement process, which ensures the floating of the zero-level value of the signal and overcomes the influence of the DC level drift of the signal on counting and subdivision.

3. Floating tracking of intersection of two signals:

Set two units in the digital signal processor as intersection registers I and II, and assign an initial value of 0. During the measurement process, the polarity and amplitude of the two-way sampling data are judged in real time, and the section where the current sampling point is located is calculated, and compared with the section where the previous sampling point is cached. Each time when the sampling point enters section 2 from section 1 or enters section 6 from section 5, the larger amplitude value of the two signals is stored in the intersection register I; when the sampling point enters section 2 from section 3 When segment 4 or segment 7 enters segment 8, the larger amplitude of the two signals is stored in the intersection register II. When the phase difference of the two signals changes, the amplitudes of intersection point I and intersection point II also change accordingly. The real-time floating tracking of the intersection of the two input signals overcomes the section judgment error caused by the phase difference change, and ensures accurate signal subdivision and direction identification when the signal deviates from the orthogonal state.

4. Count:

Set a unit in the digital signal processor as a reversible counter, and assign an initial value of 0, and then compare the section where the current sampling point is with the section where the previous sampling point is cached. When the sampling point enters section 2 from section 1 or enters section 6 from section 5, count up once, and the moiré fringes move forward; when the sampling point enters section 3 from section 4 or enters section 8 In section 7, count down once, and the Moiré fringes move in reverse. In this way, every time the signal changes a cycle, two counts are performed. Determine whether the signal enters segment 2 twice from segment 1 or enters segment 3 twice from segment 4, and if so, calculate and update the zero level value from the buffered signal peak and valley values. In this way, while counting, the zero-level value of the signal and the floating tracking of the intersection point of the two signals are realized, which eliminates the counting error caused by the DC level drift of the signal, and also avoids the error of section discrimination when the signal deviates from the orthogonal state. And it can overcome the influence of high-frequency disturbance near the zero crossing point of the signal on the counting, so as to ensure accurate counting of large numbers.

5. Segmentation:

5.1 When starting the measurement, collect two-channel data, normalize them respectively, and judge the section where the measurement starting point is located according to its polarity and size; store the smaller one of the normalized sampling data into the starting point in the register;

5.2 During the measurement process, collect two-channel data, normalize them respectively, judge the section they are in, and cache the smaller one of the normalized data into the termination point register. Simultaneously track the peak and valley values of the two signals, track the intersection of the two signals and cache them in the register;

5.3 Look up the obtained measurement start point, end point, intersection point I and intersection point II respectively (that is, according to the subdivision multiple, set an appropriate unit in the microprocessor in advance, and set a signal cycle according to the corresponding relationship between the amplitude and the displacement The calculated data are stored in order to form a table), and the displacement value corresponding to the amplitude is obtained, wherein the displacement value L _First corresponding to the amplitude of the intersection point I; the displacement value L _Second corresponding to the amplitude value of the intersection point II; the amplitude value of the starting point The corresponding displacement value is L _Ori ; the displacement value corresponding to the termination point amplitude is L _Stp .

Assuming that the displacement equivalent represented by one signal period is d, when the starting point is in section 1 and the ending point is in other sections, the displacement L of less than one signal period is calculated as follows:

①. When the end point is in section 1

Positive displacement when L _Stp > L _Ori : L = L _Stp - L _Ori

Reverse displacement when L _Ori > L _Stp : L = L _Ori - L _Stp

②. When the end point is in section 2

For forward displacement: L＝|L _First -L _Ori |+|L _First -L _Stp |

For reverse displacement: L＝d-(|L _First -L _Ori |+|L _First -L _Stp |)

③. When the end point is in section 3

For positive displacement: L＝|L _First -L _Ori |+L _First +L _Stp

For reverse displacement: L＝L _Ori +d/2+L _Second +|L _Secondt -L _Stp |)

④. When the end point is in section 4

For positive displacement: L＝|L _First -L _Ori |+d/4+|L _Second -L _Stp |

For reverse displacement: L＝L _Ori +d/2+L _Stp

⑤. When the end point is in section 5

For positive displacement: L＝|L _First -L _Ori |+d/4+L _Second +L _Stp

For reverse displacement: L＝L _Ori +d/4+L _Second +|L _Second -L _Stp |

⑥. When the end point is in section 6

For positive displacement: L＝|L _First -L _Ori |+d/2+|L _First -L _Stp |

For reverse displacement: L＝L _Ori +2·L _Second +L _Stp

⑦. When the end point is in section 7

For positive displacement: L＝|L _First -L _Ori |+d/2+L _First +L _Stp

For reverse displacement: L＝L _Ori +L _Second +|L _Second -L _Stp |

⑧. When the end point is in section 8

For forward displacement: L＝d-(L _Ori +L _Stp )

For reverse displacement: L＝L _Ori +L _Stp

When the starting point is in other sections, there are similar calculation formulas, which will not be listed one by one.

The starting point and the ending point take the smaller one of the two normalized data, so that subdivision can avoid the section with poor signal linearity; when calculating the displacement of less than one signal period, the look-up table method can be used to effectively overcome the The subdivision error caused by the nonlinearity of the signal.

6. Correction of large number count value

The above-mentioned counting of large numbers may be overcounted in some cases. For example, for forward movement, if the starting point is at segment 1 or 8 and the ending point is at segment 2 or 3, an extra number will be counted; if If the starting point is at segment 1 or 8 and the ending point is at segment 6 or 7, two more numbers will be counted. Therefore, a corresponding correction must be made to the count value of the large number. For the forward movement, there are 64 situations to be considered according to the positions of the start point and the end point; similarly, for the reverse movement, there are also 64 situations to be considered. After careful analysis, we have given the "Karnaugh map" used to correct the count value when the signal moves forward and reverse, as shown in Table 2 and Table 3.

Table 2 "Karnaugh map" of the corrected count value when the signal is moving forward

Note: In Table 1, "counting value minus 1" is represented by "1"; "counting value minus 2" is represented by "2"; for odd-numbered segments "start point amplitude > end point amplitude, count value minus 2; For an even-numbered segment, "start point amplitude < end point amplitude, the count value will be reduced by 2; otherwise, it will not be reduced", use "-" to indicate; if "do not correct the count value", use "X" to indicate.

Table 3 "Karnaugh map" of the corrected count value when the signal moves in the negative direction

Note: In Table 2, "counting value minus 1" is represented by "1"; "counting value minus 2" is represented by "2"; for odd segment "start point amplitude < end point amplitude, count value minus 2; For the even-numbered segment, "starting point amplitude > ending point amplitude, the count value minus 2, otherwise no reduction" or "otherwise no reduction" is represented by "-"; if "the count value is not corrected", it is represented by "X".

7. Distinguishing direction of moiré (interference) fringe displacement

If the large number count value is not 0, the displacement direction is determined by the positive or negative of the count value. The count value is positive, and the displacement is positive; the count value is negative, and the displacement is negative.

When the count value of the large number is 0, the displacement direction is judged by the section where the measurement start point and end point are located, as shown in Table 4. If the starting point and the ending point are in the same section, the displacement direction is judged according to the monotonicity of the signal.

              Table 4 Judgment of the displacement direction when the count of large numbers is zero

Note: "+" in Table 4 means positive displacement; "-" means reverse displacement

According to the section where the starting point and the ending point are located, combined with the intersection point I and the intersection point II, the displacement value L of less than one signal period is calculated; if the corrected large number counting result N is calculated, the current total measured displacement value L is _calculated The formula is

_Ltotal =N·d/2+L

The device that the present invention designs according to said method is as shown in Figure 2, and it is by operational amplifier 1,2, analog-to-digital conversion chip 3,4, digital signal processor 5, single-chip microcomputer 6, decoder 7, liquid crystal display 8,40M Clock 9 etc. constitute. Wherein the operational amplifiers 1 and 2 are connected with the analog-to-digital conversion chips 3 and 4 respectively, and the analog-to-digital conversion chips 3 and 4 are AD9225, and its highest sampling rate can reach 20M/s, and they are all connected with the signal processor 5; Processor 5 is the core component of the device of the present invention, and its operation speed is up to 80MHz. It is connected with single-chip microcomputer 6 and decoder 7 simultaneously;

After the moiré (interference) fringe photoelectric signals Sin1 and Sin2 with a positioning difference in two paths pass through the operational amplifiers 1 and 2, they are sent to the analog-to-digital conversion chips 3 and 4, and the analog-to-digital signals are synchronized under the control of the digital signal processor 5 Conversion, the conversion frequency is 5M/s. After receiving the serial interrupt signal, the digital signal processor 5 calls the interrupt processing subroutine to obtain the sampling data of the two signals. The digital signal processor 5 performs the above-mentioned processing and judgment on the sampled data, calculates the section and amplitude of the current sampling point, and tracks the peak and valley values of the two signals, and tracks the amplitudes of intersection I and II. The single-chip microcomputer 6 obtains the data calculated by the digital signal processor 5 through the internal direct memory access (IDMA) mode, and obtains the displacement values corresponding to the measurement start point, the end point, the intersection point I and the intersection point II amplitude by the way of looking up the table, and distinguishes direction, calculate the displacement value of less than one signal period formed by the measurement starting point and the current sampling point, and correct the large number count value, and finally send the total displacement value and displacement direction to the liquid crystal display. Due to the adoption of high-speed analog-to-digital conversion chips 3 and 4 and digital signal processor 5 on the hardware, the dynamic range of the system is very wide, which can be from 0Hz to 500KHz. The maximum data output of the device can reach 40k/s, which can fully meet the needs of high-speed testing in machining.

In addition, depending on the required subdivision multiple and the number of digits of the selected analog-to-digital conversion chip (ADC), the phase difference limits of the two signals that can be correctly processed by this counting subdivision method are different, as shown in Table 5. If the phase difference of the two signals exceeds the limit value, it will be seriously affected by the nonlinearity of the signal, and the required resolution cannot be realized.

        Table 5 The phase difference limit of the two signals determined by the subdivision multiple and the number of bits of the ADC

If a higher subdivision multiple (such as 2000 times) is required, and at the same time, it is necessary to be able to deal with the situation where the two signals deviate from the orthogonal state in a large range (such as the phase difference between the two signals is 1°), it is necessary to use an ADC of 14 bits or more. . However, for actual moire fringe signals, by adjusting the optical path, the phase difference of the two signals can generally be guaranteed to be within 90°±60°, so it is completely possible to use a low-bit analog-to-digital conversion chip without being affected by the nonlinearity of the signal.

After actual testing, under the condition of using 10Bit analog-to-digital converter and subdivision multiple of 1000, the device of the present invention can correctly identify, count and subdivide two signals with a phase difference of 20°, which can fully meet the requirements of Moore. Streak test needs.

The present invention dynamically tracks the intersection point of the two input signals, and performs direction identification, counting and subdivision according to the section where the signal sampling point is located, and can simultaneously eliminate the DC level drift of the signal, the unequal amplitude of the two signals, and the The non-orthogonal signal and the nonlinearity of the signal affect the counting and subdivision, and can process the axial vibration displacement signal, which has stronger environmental adaptability and anti-interference ability.

Claims (2)

1, a kind of to the insensitive intersection point tracking mode of phase differential Moire fringe signal-count divided method, it is characterized in that it comprises the steps:

(1) polarity and the intersection point by two paths of signals carries out floating type segmentation to the signal period:

According to the amplitude relative size and the polarity of two paths of signals, a signal period is divided into 8 sections, two unit are set then as the intersection point register in digital signal processor, and initialize 0; In the measuring process, the polarity and the amplitude of real-time judge two paths of signals sampled data, calculate the residing section of current sampling point, and compare with the residing section of previous sampled point that buffer memory gets off, at every turn when sampled point enters section 2 by section 1 or enters section 6 by section 5, one that amplitude in the two paths of signals is bigger deposits in the intersection point register, and this value is called intersection I; When sampled point entered section 4 by section 3 or enters section 8 by section 7, one that amplitude in the two paths of signals is bigger deposited in another intersection point register, and this value is called intersection I I; When the phasic difference of two paths of signals changed, the amplitude of intersection I and intersection I I also correspondingly changed, thereby realized the floating type segmentation to a signal period;

(2) unit is set as up-down counter in digital signal processor, and initialize 0, the residing section of previous sampled point that the residing section of current sampling point and buffer memory are got off compares then, do when sampled point enters section 2 by section 1 or enters section 6 by section 5 and once add counting, the Moire fringe forward moves; Do when sampled point enters section 3 by section 4 or enters section 7 by section 8 and once subtract counting, Moire fringe oppositely moves, the every variation one-period of signal, and digital signal processor carries out counting twice;

When (3) beginning to measure, gather two channel datas, respectively it carried out normalization, judge the section of measuring the starting point place according to its polarity and size, with in the current sampled data less one deposit in the starting point register; In the measuring process, gather two channel datas, respectively it is carried out normalization, judge its residing section, and less in its a data buffer memory is advanced in the terminating point register; Simultaneously, follow the tracks of the peak value and the valley of two paths of signals, the intersection point of tracking two paths of signals also is cached in the corresponding registers; Measurement starting point, terminating point and the intersection I, the intersection I I that obtain are mapped to the displacement data table that is stored in advance in the single-chip microcomputer internal memory respectively, obtain the displacement of its amplitude correspondence, again according to the section at starting point and terminating point place, in conjunction with intersection I and intersection I I, calculate the shift value L of a signal period of less than; If the modified value in the up-down counter is N, the displacement equivalent of a signal period representative is d, then current total Displacement Measurement value L _AlwaysComputing formula be: L _Always=Nd/2+L.

2, according to claim 1 to the insensitive intersection point tracking mode of phase differential moire frange signal count fine separating method, it is characterized in that: starting point in section 1, terminating point when other section, the following calculating of displacement L of a signal period of less than:

When 1) terminating point is in section 1

L _Stp＞L _OriThe time positive-displacement: L=L _Stp-L _Ori

L _Ori＞L _StpShi Fanxiang displacement: L=L _Ori-L _Stp

When 2) terminating point is in section 2

During positive-displacement: L=|L _First-L _Ori|+| L _First-L _Stp|

Oppositely during displacement: L=d-(| L _First-L _Ori|+| L _First-L _Stp|)

When 3) terminating point is in section 3

During positive-displacement: L=|L _First-L _Ori|+L _First+ L _Stp

Oppositely during displacement: L=L _Ori+ d/2+L _Second+ | L _Second-L _Stp|)

When 4) terminating point is in section 4

During positive-displacement: L=|L _First-L _Ori|+d/4+|L _Second-L _Stp|

Oppositely during displacement: L=L _Ori+ d/2+L _Stp

When 5) terminating point is in section 5

During positive-displacement: L=|L _First-L _Ori|+d/4+L _Second+ L _Stp

Oppositely during displacement: L=L _Ori+ d/4+L _Second+ | L _Second-L _Stp|

When 6) terminating point is in section 6

During positive-displacement: L=|L _First-L _Ori|+d/2+|L _First-L _Stp|

Oppositely during displacement: L=L _Ori+ 2L _Second+ L _Stp

When 7) terminating point is in section 7

During positive-displacement: L=|L _First-L _Ori|+d/2+L _First+ L _Stp

Oppositely during displacement: L=L _Ori+ L _Second+ | L _Second-L _Stp|

When 8) terminating point is in section 8

During positive-displacement: L=d-(L _Ori+ L _Stp)

Oppositely during displacement: L=L _Ori+ L _Stp

L in the above-mentioned formula _FirstBe the shift value of intersection I amplitude correspondence, L _SecondBe the shift value of intersection I I amplitude correspondence, L _OriBe the shift value of starting point amplitude correspondence, L _StpShift value for terminating point amplitude correspondence; Starting point is when other section, and computing formula is identical.

Images