CN103322922B - The optical heterodyne interference method of nonlinearity erron is eliminated based on fft algorithm - Google Patents

Abstract

The optical heterodyne interferometry based on the FFT algorithm to eliminate nonlinear errors belongs to the field of optical heterodyne interference. This method is to obtain at least 3 signals on the optical path with nonlinear error interference components and make the respective signals of at least 3 of them The periodic non-linear error interference component, the phase angle on the frequency component of the signal to be eliminated is different within a corresponding circular angle range or within a corresponding circular angle range after excluding integer multiples of the circular angle, and in the analog circuit Filters are used in the signal channel to filter the signal. After the analog-to-digital conversion, FFT is used to judge the phase of no less than 3 signals respectively. After the displacement is calculated by the displacement calculation part, subsequent processing is performed to eliminate or suppress interference. Element. The invention can eliminate nonlinear error or specific harmonic component of interference, the adjustment process is very easy, the system performance is improved, and the cost is reduced.

Description

Optical Heterodyne Interferometry Based on FFT Algorithm to Eliminate Nonlinear Errors

technical field

The invention belongs to the field of optical heterodyne interferometry (Heterodyne Interferometry) measurement, and specifically relates to an optical heterodyne interferometry method based on an FFT (fast Fourier transform) algorithm that eliminates nonlinear error fundamental frequency without fine adjustment.

Background technique

The laser heterodyne interferometry system can be used to measure physical quantities such as displacement and length, and is one of the best nanometer measurement methods. The system converts the measured displacement into the frequency or phase change of the heterodyne signal, and then measures this change. Since the frequency of the heterodyne signal is much lower than the optical frequency, the photoelectric signal is easy to process and can be electronically fine-tuned. Points to achieve higher measurement resolution, the resolution can reach picometer (pm) or better.

However, there are generally nonlinearity problems in optical heterodyne interferometry. These factors can be the main source of error in displacement measurement, making the accuracy generally only on the order of nanometers to tens of nanometers. The reason is that beams of different frequencies cannot good separation. These periodic nonlinear error problems have always been an obstacle to the technological development in this field.

Over the years, some improved methods have been continuously invented at home and abroad, but there are also some limitations or problems. For example, Badami and Pattersom use rate meters and spectrum analyzers to directly measure and compensate nonlinear error components in mobile systems; TaeBongEom and TaeYoungChoi et al. The method of ellipse fitting after phase signal integration is used to compensate the nonlinear error. These methods can compensate nonlinear errors when there is no other auxiliary interferometer, but the system is complex, and requires a large number of fluctuation signal periods and a relatively large number of calculations, which affects the real-time performance of the measurement. A new method is disclosed in the Chinese invention patent with the publication number CN101660924, which can effectively solve this problem, but if you want to get very accurate results, you also need very precise optical adjustment, which is usually raletively hard.

Contents of the invention

In order to solve the technical problems of the existing laser heterodyne interferometry system, such as complex system, large amount of calculation, poor real-time performance, and inaccurate results, the present invention provides an optical heterodyne interferometry based on resonant filtering and FFT algorithm to eliminate nonlinear errors, It can eliminate the specific harmonic components of nonlinear error or interference, and eliminate the need for precision optical adjustment devices and precise adjustment processes required to eliminate nonlinear errors, making this adjustment very easy, while greatly improving system performance and reducing costs.

The AC signal obtained by the output end of the photodetector of the laser heterodyne interferometry arm is an electrical signal with a different frequency optical frequency difference, and the additional phase difference is in a precise proportional relationship with the measured displacement. But it contains the so-called nonlinear error, which is mainly distributed on the first harmonic of the Fourier spectrum (generally, it can also be called the fundamental frequency according to the nature of the Fourier series, which is a kind of spatial frequency). The method of the present invention can eliminate the fundamental frequency component of the nonlinear error; in fact, the method of the present invention is effective for any certain frequency, and does not need special precise adjustment to the optical system.

In the signal optical path with nonlinear error interference components in the laser heterodyne interference system, at least 3 signals (channels) are obtained, by setting at least 3 interference light photoelectric detection parts on the optical arm carrying the detected signal Obtain these signals, and make the phase angles of the first harmonic (or other frequency components) of the nonlinear error contained in at least three of them vary within a corresponding circle, so that sufficient information is obtained, so that The fundamental frequency (or any frequency) component in the nonlinear error can be eliminated through subsequent weighting and superposition processing. There are many ways to realize that the phase angle of the first harmonic (or other frequency components) is different in a corresponding circle, for example, the phase angle of the corresponding first harmonic can be changed by adjusting the optical path, and the first harmonic has been disclosed in the prior art. The harmonic error measurement method can easily obtain the amplitude and initial phase angle of the first harmonic. In general, the phase angles of the signal errors of different optical paths are usually unequal. If the phase angles are equal, it can be known by the method of measuring the phase angles in the prior art, and they can be made unequal by adjusting the optical path. Furthermore, in general, in the optical system we have built randomly, there will not be any two-way nonlinear error phase angles that are exactly the same within a corresponding circle, but if the adjustment of the optical system is completed, two of the three-way non-linear errors Linearity error The phase angle of the first harmonic (or other frequency components to be eliminated) is not much different, which will affect the amplitude of the useful signal after weighted superposition, thereby affecting the signal-to-noise ratio of the output signal. At this time, we can roughly adjust it The difference between the phase angle of the frequency component to be eliminated in the 1-2 channel relative to the corresponding signal phase angle (of the frequency component of the nonlinear error to be eliminated) in other channels increases it until sufficient (i.e. sufficient). Now there are many methods to detect the phase and amplitude of the first harmonic of nonlinear error (or other frequency components to be eliminated), such as "Measurement method of laser heterodyne interference nonlinear error" "Journal of Beijing University of Technology" by Chen Hongfang et al. No. 6, 2010; Zhong Zhi et al. "A Novel Measurement Method for Nonlinear Error of Laser Heterodyne Interferometry", Optoelectronic Laser, 2005, etc. The method of adjusting the phase angle can be realized by adjusting the optical path difference between the two light rays participating in the interference.

Based on the periodicity of harmonic functions, it is of course possible to exclude several integer multiples of circular angles included in the above-mentioned phase difference as usual (no similar reminders will be given below).

The following method is used for subsequent processing: first assume that there are only 3 signals, and compare the phases of the 3 signals with the reference signal until 3 displacements are solved. At this time, the phase of the nonlinear error in each channel is different.

It can be seen from the following proof that there is a non-zero weighting coefficient array that can make the amplitude of the fundamental frequency or other frequency components of the nonlinear error zero after superposition, while retaining useful signals.

Among the three displacements solved above, the displacement measurement result of the i-th road can be expressed as:

Wherein, the first term x at the right end of the equal sign is the real physical displacement, and the second term is the primary component of the nonlinear error. Of course, the content of the present invention is illustrated here with the first harmonic and does not lose generality; a is a coefficient, Expressing the magnitude of the nonlinear error is generally a small amount, but it will not be approximated in the subsequent derivation. is the initial phase angle of the nonlinear error, and λ is the light wavelength. Here the wavenumber in the nonlinear error term Indicates the first harmonic component and is reflected by a single reflection of the measuring prism. For the second harmonic case, this term becomes

The three results are weighted and superimposed, and normalized into:

The new set of ki is just for neatness in form, and can be regarded as a new weighting coefficient. Our goal is to find a set of ki such that: X=x.

From the above conditions we can get the following linear equations:

k ₁ +k ₂ +k ₃ =b

where b≠0

The coefficient matrix of variable ki is a square matrix of order 3 with full rank, and the system of equations about ki has non-zero solutions. That is, you can make

In summary, it has been proved that after such weighted superposition, the first harmonic component of the nonlinear error has been eliminated.

if in Any two of them differ by 180°, which will make the problem simple and beneficial. This situation generally requires precise optical adjustment to obtain, and only two signals with a difference of 180° in nonlinear components are enough to eliminate the non-linearity of a specific frequency. The linear error component has been described in detail in the invention patent mentioned in the background technology.

So far we have solved the nonlinear error problem in laser heterodyne interferometry in the 3-dimensional linear space of electronics.

For the situation with more than 3 signals (channels), we can reserve at least 3 signals with different phase angles in advance, and then superimpose the remaining signals with any one of the reserved 3 signals, and turn them into non- The phase angle difference at a certain frequency of the linearity error is not zero in 3 ways. In this way, the problem that has been solved above is solved, that is, the situation of 3 channels, and has been proved. Of course, the above-mentioned first simplification into 3-way processing may not be adopted, but may be completed together in the weighted superposition process.

In practice, the correct weighting coefficient array can be obtained by the following methods to eliminate the error component: (1) Linearly superimpose the two lines, because in the 2-dimensional space of this nonlinear harmonic component, the two lines of signals have formed A set of bases of the space, so any vector can be generated by adjusting the superposition coefficient, we let the generated vector and the third signal (vector) be in the opposite direction on a nonlinear harmonic component, and then the generated signal Superimposed linearly with the third signal (vector), adjusting the weighting coefficient can completely subtract a certain harmonic component of the nonlinearity to zero, while the useful signal will still be retained. (2) The weighting coefficients of at least 2 channels can be repeatedly and alternately scanned by means of electrical adjustment, and the non-linear error value in the superimposed output can be measured at the same time, so that it can be continuously reduced until it is zero. (3) Using the most basic method, after measuring the phase angle and amplitude in each channel, it is directly calculated by the analytical method. The idea in (1) can also be used here, but it is not a decomposition adjustment, but a step-by-step calculation , which will not be explained in detail here. That is to say, after discovering that the first harmonic error can cancel each other through weighted superposition, thereby removing the above-mentioned principle of the error component, according to the relevant content recorded in the present invention, an appropriate weighted superposition method can be selected, and the first harmonic error can be realized. eliminate. The above-mentioned various methods for obtaining weighting coefficients are only enumerated for the convenience of understanding, and do not completely limit the technical solution.

In order to avoid the corresponding optical fine-tuning, eliminate or suppress the nonlinear error, and help to retain the useful signal and choose a reasonable parameter, as a preferred solution, at least three periodic nonlinear errors in the three-way signal are to be detected The difference between the phase angles on the frequency components of the eliminated signals is not less than 60 degrees. This is because in practice we can always easily adjust or set so that the difference angle between any two phases is not too small. Both theory and experiment show that 60° as the lower limit of this angle is a reasonable value, which is good for the signal-to-noise ratio after processing, and the useful signal output will not be lower than the amplitude of a single channel after weighted superposition, and at the same time As long as you blindly try to randomly fix the components on the optical path that can affect the phase difference between the three paths 3 to 5 times, there is roughly one chance to meet this condition, thereby achieving mechanical "adjustment-free" and saving the corresponding precision. The adjustment device makes the adjustment or setting work simple and easy. Whether this condition has been met can be judged according to the above-mentioned method of measuring the initial phase angle of the first harmonic error, so as to stop the adjustment. It is also possible to eliminate a certain nonlinear harmonic component by using methods (1) and (2) similar to the actual elimination of error components mentioned above, and then check the amplitude of the useful signal. If it is too low, it means that the "60 degree condition" is not met. ” or similar criteria. Check whether the adjustment is complete and the error has been eliminated can be judged by, for example, comparing the results with a more precise measuring instrument.

This method of eliminating precision mechanical adjustment or optical fine adjustment by replacing optical "precision adjustment" with electronic "precision adjustment" can greatly reduce the cost, and the accuracy requirements for optical adjustment can be reduced by more than 2-3 orders of magnitude. The accuracy of the whole system at the specific frequency of the nonlinear error will be significantly improved.

Obviously, the basic principles of this method and its own advantages and significance have nothing to do with whether integrated optical methods, including optical fiber methods, or bulk wave methods are used. This method is valid for all periodic errors in optical heterodyne interference systems, as long as we can obtain at least three signals whose interference components have different phases.

For the judgment of the phases of no less than three signals, digital processing is used, and then weighting and superposition processing are performed after the displacement is calculated in the displacement calculation part according to the results of these processings. The optical heterodyne interference signal is a special weak signal. The instantaneous amplitude noise and phase noise in the light wave ray generated by the radiation are very large, but the FFT algorithm can obtain accurate calculation results, and can directly carry out subsequent weighted superposition, etc. deal with. In a single digital part, digital processing processes such as digital phase interpretation, displacement calculation and weighted superposition are completed continuously at one time, so that the system has the advantages of high efficiency, fast speed, low cost, small size and simple structure.

Based on the above, the present invention provides a kind of optical heterodyne interferometry that can eliminate or reduce the first harmonic or other frequency components of periodic nonlinear errors, improve measurement accuracy, and avoid corresponding precise optical adjustments, that is, based on FFT algorithm to eliminate The optical heterodyne interferometry of nonlinear error, the method is: on the optical path with nonlinear error interference components, obtain at least 3 measurement signals and 1 reference signal through light splitting, and make the respective of at least 3 measurement signals The phase angle of the periodic nonlinear error on the frequency component to be eliminated is different within a corresponding circular angle range; at least 3 measurement signals and 1 reference signal are respectively converted into corresponding at least 3 measurement analog electrical signals and 1 channel of reference analog electrical signal; performing the first filtering process and analog-to-digital conversion on each of the at least 3 channels of measured analog electrical signals, correspondingly obtaining at least 3 channels of measured digital electrical signals; performing 1 channel of reference analog electrical signal The second filtering process and analog-to-digital conversion correspond to obtaining 1 channel of reference digital electrical signals; respectively input the at least 3 channels of measurement digital electrical signals and 1 channel of reference digital electrical signals into the corresponding FFT phase interpretation module, and calculate and obtain at least 3 channels of measurement signals Phase angle and 1-way reference signal phase angle; said at least 3-way measurement signal phase angle and 1-way reference signal phase angle respectively perform phase comparison and phase difference period counting, and convert it into displacement to obtain at least 3-way displacement measurement results, At this time, the displacement measurement results of each path contain nonlinear errors; the obtained at least 3 displacement measurement results are subjected to a real number weighted operation, and then the addition operation can obtain the final measurement result; the first filtering process adopts a single-stage resonant filter .

The above-mentioned single-stage resonant filter includes parallel or series inductance-capacitance resonant circuits.

The phase difference of at least 3 periodic nonlinear errors among the at least 3 measurement signals mentioned above is not less than 60°.

The aforementioned first filtering process and the second filtering process use different types of filters.

The above-mentioned first filtering process and the second filtering process use the same type of filter.

The above step of obtaining at least 3 measurement signals and 1 reference signal through light splitting includes: the output A-channel optical signal is sequentially passed through at least 4 optical splitting devices to obtain A1, A2...An optical signals respectively, wherein n≥4; and The output A-channel optical signal with frequency difference B-channel optical signal passes through the optical splitting device to obtain the B1-channel optical signal, and the remaining optical signals of the B-channel are reflected by the reflecting device and then pass through the optical splitting device to obtain B2...Bn optical signals, where n≥ 4. The A1 optical signal and the B1 optical signal interfere to form a reference signal, and the A2...An optical signal and the B2...Bn optical signal respectively interfere to form n-1 measurement signals.

In fact, the laser heterodyne interference signal is a typical weak signal, especially in this precision system that eliminates nonlinear errors, the signal purity is required. The filter in the analog electrical signal is a very important link in the whole system. The role is unique. Considering that the useful signal of the system is a single point frequency component, the filter composed of parallel or series inductance-capacitor resonant circuits in analog electronics works exactly at the natural frequency of a single point. This kind of resonance circuit composed of The resonant characteristics and narrow bandwidth of the filter can greatly suppress various noises, nonlinear intermodulation and spectral components outside the band pass required by the sampling theorem, and this is exactly the goal to be achieved. The main characteristics of this resonant circuit are smooth, clear, and simple. The circuit itself is noise-free, anti-radiation, and easy to control, so no matter whether there is a low-pass filter (such as Chebyshev filter) set in the system to ensure the sampling theorem , it is very desirable to add this resonant circuit to the analog electrical signal channel to filter the signal, so as to obtain high-quality signals for subsequent weighted superposition processing. This filter can also reduce the low-pass set by the sampling theorem filter requirements. This filtering solution can make the whole system accurate, stable, reliable, easy to adjust and produce. Therefore, we can use a filter composed of parallel or series inductance-capacitor resonant circuits in the analog electrical signal channel to filter the signal so that it can be processed by subsequent electronics such as weighting and superposition.

The quality factor of the RF inductance-capacitance resonant circuit can reach tens or even hundreds. From the perspective of filtering, it is enough to make the system achieve a precision better than nanometers and meet the needs of most users.

Add resistance or equivalent resistance in the resonant circuit to adjust the quality factor for filtering, so that the quality factor can be adjusted and stabilized. Add resistors (including adjustable resistors, potentiometers, or equivalent methods) to a filter circuit composed of parallel or series inductor-capacitor resonant circuits with a relatively high quality factor (Q value), so that the quality factor is mainly controlled by the resistor It is determined by the equal concentration parameters, rather than being determined by the distribution parameters, so as to obtain the desired and stable Q value. The Q value has many physical meanings. If the Q value is too high here, it will also affect the highest movement speed under the precision requirement.

For the filter composed of the above-mentioned parallel or series inductance-capacitance resonant circuit, a weak coupling method is adopted between the output terminal of the previous stage and the input terminal of the subsequent stage to ensure the stability and sufficient quality factor of the resonant circuit. These weakly coupled methods are realized by at least one of the following methods, 1) capacitive coupling, 2) inductive coupling, 3) resistive coupling, and 4) transformer coupling in the form of windings.

The present invention adopts the FFT algorithm in the digital processing for the interpretation of the phases of no less than three signals after the filter formed by the inductance-capacitance resonant circuit connected in parallel or in series and its analog-to-digital conversion, and then the displacement is calculated by the displacement calculation part After that, weighting and stacking are performed. FFT can accurately calculate the phase difference, and can also directly obtain the moving speed of the system under test through the frequency variable, which is very important sometimes. FFT software and hardware are mature and reliable, and the research and development are perfect. In addition, the broad-spectrum analysis of FFT is also suitable for the movement speed of large measured objects, and can also be used for the evaluation and diagnosis of the system itself, and can also be shared with other systems. It is very suitable for applications where people can Direct maintenance or telemetry and remote intervention, occasions with high reliability and easy maintenance requirements, such as radiation environments such as radiation and space. This overall structure can obtain accurate and stable technical indicators, and has a simple structure and good reliability.

The last step in the method of the present invention is the superposition operation. The whole system maintains the longest respective discrete channel parts, which facilitates the detection of these discrete channel parts. For example, if the input of the three channels is short-circuited and the same signal is input, the output results of the three channels can be compared and detected to evaluate the working status, which is conducive to rapid detection and remote detection.

Another technical solution of the present invention is: an optical heterodyne interference device adopting the above-mentioned optical heterodyne interferometry, comprising: a first beam splitting device, which splits the A-path optical signal to obtain A1, A2...An-path optical signals, Wherein n≥4; the second optical splitting device splits the B-channel optical signal having a frequency difference with the output A-channel optical signal to obtain the B1-channel optical signal and other optical signals, and the remaining optical signals are reflected by the second optical signal after being reflected by the reflecting device. The light splitting device divides the light to obtain the B2...Bn path optical signal, where n≥4; the interference forming device makes the A1 path optical signal and the B1 path optical signal interfere to form a reference signal, and the A2...An path optical signal and the B2.. .Bn optical signals correspond to interference to form n-1 measurement signals respectively; the signal detection device converts the above-mentioned 1-channel reference signal and n-1 measurement signal into 1-channel reference analog electrical signal and n-1 measurement analog electrical signal signal; the first filtering and conversion module performs the first filtering process and analog-to-digital conversion on n-1 channels of measured analog electrical signals, and correspondingly obtains n-1 channels of measured digital electrical signals; the second filtering and conversion module performs 1 channel of measurement Carry out the second filtering process and analog-to-digital conversion with reference to the analog electrical signal, and obtain 1 channel of reference digital electrical signal correspondingly; the FFT phase interpretation module calculates and obtains the phase angle of n-1 channel of measured digital electrical signal and the phase angle of 1 channel of reference digital electrical signal; The displacement calculation module calculates and obtains n-1 displacement measurement results according to the n-1 measurement digital electrical signal phase angle and the 1 reference digital electrical signal phase angle; the weighted superposition module calculates the obtained n-1 displacement measurement results Real weighted operations and addition operations.

The present invention omits some explanations and representations of the prior art, such as omitting the FFT system, adopting theorem, circuit board distributed filter capacitance of the power supply system, the amplifier provided for the system gain, etc., the general basic part Discussion and block diagrams etc.

The beneficial effects of the present invention are: on the basis of the usual laser heterodyne interferometry system, the present invention creatively divides the measurement optical path into multiple paths, and carries out electronic adjustment, signal processing, weighted superposition and other processes respectively, and uses electronic precision adjustment Long, to complement the short of optical precision adjustment, eliminating the need for complex optical fine adjustment or other non-electronic adjustments, instead of using electronic adjustment and signal processing methods to eliminate nonlinear errors and reduce the complexity of optical adjustment , reducing costs and improving efficiency. At the same time, it effectively eliminates or greatly reduces the periodic interference components such as nonlinear errors, thereby improving the measurement accuracy or reducing interference. Devices, quality factor adjustment resistors, etc. improve the measurement accuracy of the entire system, have strong and stable anti-interference capabilities, are conducive to large-scale production, and have low production costs.

Description of drawings

Fig. 1 is a schematic diagram of the optical part of the optical heterodyne interferometry based on the FFT algorithm of the present invention to eliminate nonlinear errors.

FIG. 2 is a schematic diagram of a filter circuit composed of a one-stage parallel inductor-capacitor resonant circuit in the present invention.

Fig. 3 is a schematic diagram of the principle of optical heterodyne interferometry based on FFT algorithm to eliminate nonlinear error in the present invention.

detailed description

The present invention will be described in further detail below in conjunction with the accompanying drawings.

The system implementing the optical heterodyne interferometry based on the FFT algorithm of the present invention to eliminate nonlinear errors can be divided into an optical part and a subsequent electronic signal processing part. The principle of the optical part is shown in Figure 1, and the present invention adds the part in the dotted line box in the figure on the basis of an original optical part.

First, the working principle of the optical part of the original heterodyne interferometry system is described: the laser light is emitted by the laser 101, enters the acousto-optic device 103 through the first part of the mirror 102, is frequency-shifted, and then exits to the second part of the mirror 114. The light is reflected back by the measuring prism 115 and is incident on the third part of the reflecting mirror 111 , and the reflected part interferes with another beam of light passing through the first polarizer 131 . The light reflected by the first partial reflector 102 (which provides reference light for the following interference parts) passes through the first reflector 125, and then is reflected by the fourth partial reflector 124, the fifth partial reflector 123, and the sixth partial reflector. After the mirror 122 and the second reflecting mirror 121, the reflected light of the third partial reflecting mirror 111 enters the first polarizing plate 131 together, and the interference light emitted by it is converted by the first photodetector 141 into electric signal. This electrical signal is used as a measurement signal, which already contains the displacement information obtained from the measurement prism 115, and also contains nonlinear error components.

The light reflected by the second partial reflector 114 merges with the light reflected by the fourth partial reflector 124 and is incident on the second polarizer 134. Part of it is converted into an electrical signal, which is the reference signal.

The above-mentioned laser 101 , first partial reflector 102 and acousto-optic device 103 constitute a dual-frequency laser generator with slightly different frequencies, which may also be replaced by other solutions.

The present invention adds the part inside the dotted line box in Fig. 1, which mainly includes two new measuring optical arms/optical paths and corresponding circuit parts.

Together with the original first optical arm composed of the third partial reflector 111 in the figure, the first polarizer 131, the first photodetector 141, etc., there are three optical arms, or called three measuring optical paths, and Thus, measurement signal output 1 , measurement signal output 2 and measurement signal output 3 are obtained. These two new measurement optical arms are completely identical in structure to the first optical arm. Adding the reference signal, there are a total of four signal outputs, and these four signals constitute the output of the optical part, which is transmitted to the subsequent electronic signal processing part, as shown in Figure 3.

In the electronic signal processing part, the signals are respectively passed through the filter formed by the parallel or series inductance-capacitance resonant circuit shown in Figure 2 (the filtering parts for the first channel to the fourth channel are respectively 211 and 221 in Figure 3 , 231, 241); the analog-to-digital conversion part (for the first channel to the fourth channel are respectively 212, 222, 232, 242 in the figure); the part of calculating the phase angle in the system described in Fig. 3 is called the FFT part (for The first to fourth channels are 213, 223, 233, 243 in the figure); then the three-way measurement data are respectively in the displacement calculation part (for the first to third channels are 214, 224, 234 in the figure) and reference Phase comparison and cycle counting of phase difference are carried out on the signal, and converted into displacement, the corresponding displacement measurement is completed, and the measurement results of each channel are obtained; of course, the results also contain their own nonlinear errors that destroy the accuracy of the system. Then the measurement results of the three channels are respectively sent to the weighted superposition operation part 301 for real number weighting operation, and then the addition operation can be used to obtain the final measurement result, and the fundamental frequency of the nonlinear error has been eliminated or greatly suppressed.

Claims (10)

1. eliminate the optical heterodyne interference method of nonlinearity erron based on fft algorithm, it is characterized in that: the method comprises:

In the light path that there is nonlinearity erron interference component, obtain at least 3 road measuring-signals and 1 tunnel reference signal by light splitting, and make the phasing degree of the respective periodicity nonlinearity erron of wherein at least 3 road measuring-signals in the frequency content that will be eliminated different within the scope of a corresponding angle of circumference;

At least 3 road measuring-signals and 1 tunnel reference signal are separately converted to corresponding at least 3 drive test amount analog electrical signals and 1 tunnel reference analog electrical signal;

The first filtering and analog to digital conversion are carried out respectively in each road of described at least 3 drive test amount analog electrical signals, and correspondence obtains at least 3 drive test amount digital electric signals; The second filtering and analog to digital conversion are carried out with reference to analog electrical signal in 1 tunnel, and correspondence obtains 1 road reference number electric signal;

Described at least 3 drive test amount digital electric signals are inputted corresponding FFT phase place reading module respectively with 1 road reference number electric signal, calculates at least 3 drive test amount signal phase angles and 1 reference signal phasing degree, tunnel;

The cycle count of phase compare and phase differential is carried out at described at least 3 drive test amount signal phase angles and 1 reference signal phasing degree, tunnel respectively, and is converted into displacement, obtains at least 3 tunnel displacement measurement, now contains nonlinearity erron in the displacement measurement of each road;

At least 3 tunnel displacement measurement obtained are carried out real number ranking operation, then carries out additive operation and can obtain final measurement;

Described first filtering adopts single-stage resonance filter.

2. eliminate the optical heterodyne interference method of nonlinearity erron as claimed in claim 1 based on fft algorithm, it is characterized in that: described single-stage resonance filter comprises the inductance capacitance resonant tank of parallel connection or series connection.

3. eliminate the optical heterodyne interference method of nonlinearity erron as claimed in claim 1 based on fft algorithm, it is characterized in that: described single-stage resonance filter adopts weak coupling mode to be coupled with between its prime output terminal and rear class input end.

4. eliminate as claimed in claim 1 the optical heterodyne interference method of nonlinearity erron based on fft algorithm, it is characterized in that: in the resonant tank of described single-stage resonance filter, add resistance or equivalent resistance adjustment quality factor, to make quality factor adjustable and stable.

5. eliminate as claimed in claim 3 the optical heterodyne interference method of nonlinearity erron based on fft algorithm, it is characterized in that: the method for described weak coupling is capacitive couplings or inductive coupling method or resistively couple method or the transformer coupled method of winding type.

6. eliminate the optical heterodyne interference method of nonlinearity erron as claimed in claim 1 based on fft algorithm, it is characterized in that: the phase differential of at least 3 periodicity nonlinearity errons in described at least 3 road measuring-signals is not less than 60 °.

7. eliminate the optical heterodyne interference method of nonlinearity erron as claimed in claim 1 based on fft algorithm, it is characterized in that: described first filtering and the second filtering adopt different types of wave filter.

8. eliminate the optical heterodyne interference method of nonlinearity erron as claimed in claim 1 based on fft algorithm, it is characterized in that: described first filtering and the second filtering adopt the wave filter of identical type.

9. the optical heterodyne interference method of nonlinearity erron is eliminated as claimed in claim 1 based on fft algorithm, it is characterized in that: the step being obtained at least 3 road measuring-signals and 1 tunnel reference signal by light splitting is comprised: the A road light signal of output obtains A1, A2 respectively by least 4 light-dividing devices successively ... An road light signal, wherein n >=4; The B road light signal with the A road light signal exported with frequency difference obtains B1 road light signal by light-dividing device, obtains B2 successively after all the other light signals of B road are reflected by reflection unit by light-dividing device ... Bn road light signal, wherein n >=4; A1 road light signal and B1 road light signal interfere formation reference signal, A2 ... An road light signal and B2 ... the corresponding interference forms n-1 road measuring-signal to Bn road light signal respectively.

10. adopt a laser heterodyne interferometry device for the optical heterodyne interference method as described in claim 1-9, it is characterized in that: comprising:

First light-dividing device, obtains A1, A2 by the light signal light splitting of A road ... An road light signal, wherein n >=4;

Second light-dividing device, obtains B1 road light signal and remaining light signal thereof by the B road light signal light splitting that the A road light signal with output has frequency difference, and all the other light signals described are obtained B2 by the second light-dividing device light splitting after reflection unit reflection ... Bn road light signal, wherein n >=4;

Interfere forming apparatus, make A1 road light signal and B1 road light signal interfere the reference signal formed, A2 ... An road light signal and B2 ... the corresponding interference forms n-1 road measuring-signal to Bn road light signal respectively;

Signal detecting device, is converted to 1 tunnel with reference to analog electrical signal, n-1 drive test amount analog electrical signal respectively by above-mentioned 1 tunnel reference signal, n-1 road measuring-signal;

First filtering and modular converter, carry out the first filtering and analog to digital conversion respectively to n-1 drive test amount analog electrical signal, and correspondence obtains n-1 drive test amount digital electric signal;

Second filtering and modular converter, carry out the second filtering and analog to digital conversion to 1 tunnel with reference to analog electrical signal, correspondence obtains 1 road reference number electric signal;

FFT phase place reading module, calculates n-1 drive test amount digital electric signal phasing degree and 1 reference number electrical signal phase angle, road;

Displacement computing module, according to n-1 drive test amount digital electric signal phasing degree and 1 reference number electrical signal phase angle, road, calculates n-1 road displacement measurement;

Weighted stacking module, carries out real number ranking operation and additive operation by the n-1 road displacement measurement obtained.