High-precision real-time FMCW laser ranging system and method based on ZYNQ system

Abstract

The invention provides a high-precision real-time FMCW laser ranging system and method based on a ZYNQ system, and belongs to the field of fiber interferometer measurement, the system comprises two main interferometers, an auxiliary interferometer, a high-speed signal acquisition unit and a real-time signal processing unit, the real-time signal processing unit comprises a PL and a PS which are connected, the high-speed signal acquisition unit respectively acquires upper frequency sweeping signals and lower frequency sweeping signals of the main interference signals and the auxiliary interference signals and transmits the acquired signals to the PL, and the two main upper frequency sweeping signals, the two main lower frequency sweeping signals, the auxiliary upper frequency sweeping signals and the auxiliary lower frequency sweeping signals form two main-auxiliary upper frequency sweeping signal pairs and two main-auxiliary lower frequency sweeping signal pairs; and the PL performs parallel phase demodulation according to the main-auxiliary up-scanning signal pair and the main-auxiliary down-scanning signal pair at the same time to obtain four paths of echo phase information, and the PS calculates four distance values through least square fitting according to the four paths of echo phase information in sequence. The method has the high-speed, real-time and high-precision distance measurement capability.

Description

Technical Field

[0001] This invention belongs to the field of fiber optic interferometer measurement, specifically relating to a high-precision real-time FMCW laser ranging system and method based on the ZYNQ system. Background Technology

[0002] Fiber optic sensors are high-precision sensors that use optical fibers as the transmission medium. They sense changes in external physical conditions, causing corresponding changes in the optical signal within the fiber. They are widely used in high-resolution optical precision measurements such as distance and vibration measurements. Existing fiber optic ranging methods mainly include pulse methods, phase methods, multi-wavelength methods, optical frequency combs, and frequency-modulated continuous wave (FMCW) methods. Among these, FMCW laser ranging uses interferometry to measure distance. Compared to other ranging methods, it has higher sensitivity and anti-interference capabilities, no blind zone, and is advantageous for on-chip integration. Furthermore, FMCW laser ranging can apply a large frequency modulation bandwidth to the laser carrier, achieving high resolution and measurement accuracy. Therefore, it shows great application potential in civilian and military fields such as precision manufacturing, 3D topography measurement, aerospace technology, and radar detection.

[0003] While current FMCW laser ranging systems have made continuous breakthroughs in ranging accuracy, there are still gaps in real-time performance. Most high-precision ranging systems are based on offline processing or can only achieve online ranging speeds of tens of points per second. Ensuring both real-time performance and high accuracy in ranging systems remains a technical challenge. Summary of the Invention

[0004] The purpose of this invention is to propose a high-precision real-time FMCW laser ranging system and method based on the ZYNQ system, which has the ability to measure distances at high speed, in real time, and with high precision.

[0005] This invention is achieved through the following technical solution:

[0006] The high-precision real-time FMCW laser ranging system based on the ZYNQ system includes two main interferometers, an auxiliary interferometer, and a high-speed signal acquisition unit and a real-time signal processing unit implemented based on the ZYNQ system. The real-time signal processing unit includes a PL and a PS connected together. The main and auxiliary interferometric signals corresponding to the main and auxiliary interferometers each include an up-sweep frequency signal and a down-sweep frequency signal. The high-speed signal acquisition unit acquires the up-sweep frequency signal and the down-sweep frequency signal of the main and auxiliary interferometric signals respectively, and transmits the acquired two main up-sweep signals, two main down-sweep signals, auxiliary up-sweep signals, and auxiliary down-sweep signals to the PL. The two main up-sweep signals, two main down-sweep signals, auxiliary up-sweep signals, and auxiliary down-sweep signals form two main-auxiliary up-sweep signal pairs and two main-auxiliary down-sweep signal pairs. The PL simultaneously performs parallel phase demodulation based on the main-auxiliary up-sweep signal pairs and the main-auxiliary down-sweep signal pairs to obtain four echo phase information. The PS sequentially calculates four distance values ​​based on the four echo phase information through least squares fitting.

[0007] Furthermore, it also includes a laser source. The high-speed signal acquisition unit includes a DAC module, three preamplifiers and an ADC module connected in sequence. The PL includes a modulation signal generation module and a demodulation module. The modulation signal generation module is used to generate a digital modulation signal. The DAC module is connected to the modulation signal generation module to convert the digital modulation signal into an analog modulation signal to modulate the laser source. The three preamplifiers are connected to two main interferometers and an auxiliary interferometer, respectively. The three ADC modules are connected to the demodulation module to transmit the acquired signal to the demodulation module. The analog modulation signal includes a periodic triangular wave or a sine wave.

[0008] Furthermore, the demodulation module has a parallel two-path demodulation process. The demodulation process includes, in sequence, an auxiliary de-envelope module, an auxiliary interval-finding module, an auxiliary signal interpolation module, an auxiliary Hilbert transform module, an auxiliary phase demodulation module, and an auxiliary phase interpolation module; a main de-envelope module; a resampling module connected to the main de-envelope module and the auxiliary phase interpolation module respectively; and, in sequence, a main interval-finding module, a main signal interpolation module, a main Hilbert transform module, a main phase demodulation module, and a main phase interpolation module. The output of the resampling module is connected to the main interval-finding module. An auxiliary up-scan signal or an auxiliary down-scan signal is input to the auxiliary envelope module. Two main up-scan signals or two main down-scan signals are simultaneously input to the main de-envelope module for parallel processing. The main and auxiliary de-envelope modules are used to de-envelope and normalize the input signal. The auxiliary interval finding module is used to extract the integer-cycle signal from the output signal of the auxiliary de-envelope module. The auxiliary signal interpolation module is used to perform linear interpolation on the integer-cycle signal to meet the needs of the auxiliary Hilbert transform module. The auxiliary Hilbert transform module outputs the in-phase and quadrature components of the linearly interpolated signal. The auxiliary phase demodulation module performs auxiliary phase demodulation based on the in-phase and quadrature components. The auxiliary phase interpolation module outputs the auxiliary phase demodulation result with the same length as the signal before the Hilbert transform. The resampling module is used to map the output signal of the main de-envelope module onto the reference phase grid obtained based on the auxiliary phase demodulation result.

[0009] Furthermore, when the main up-scan signal and the main down-scan signal are weak, a main moving average module connected to the main de-envelope module is added to the demodulation process. The main moving average module is used to perform moving average filtering on the input signal to improve the quality of the input signal.

[0010] Furthermore, the main de-envelope module and the auxiliary de-envelope module are implemented in the same way, each calculating the maximum and minimum values ​​of its input signal and the pre-measured light intensity fluctuation curve in the first 1 / 6 interval and the second 1 / 6 interval, respectively, according to the formula... The gain coefficient top_AC and offset top_DC of the input signal are calculated respectively, and top_DC = max2 + max1 - (max02 + max01) * top_AC. Based on the calculated gain coefficient top_AC and offset top_DC, the upper envelope is precisely stretched and translated for alignment. Using the midline of the aligned envelope and the peak-valley difference as a reference, the input signal is de-enveloped and normalized for output. Here, max1 is the maximum value of the input signal in the first 1 / 6 interval, max2 is the maximum value of the input signal in the second 1 / 6 interval, max01 is the upper envelope amplitude extracted from the corresponding position of the maximum value in the first 1 / 6 interval of the light intensity fluctuation curve, and max02 is the upper envelope amplitude extracted from the corresponding position of the maximum value in the second 1 / 6 interval of the light intensity fluctuation curve.

[0011] Furthermore, the auxiliary search interval module and the main search interval module are implemented in the same way. They both scan in parallel from the head and tail of the input signal, write the position index of the input signal when it is greater than the upper threshold or less than the lower threshold into the buffer in sequence, and mark the upper peak or lower peak. At the same time, they record the index of each first threshold value. They continue to detect until two complete three-segment patterns appear in the buffer. They calculate the position of the midpoint of the beginning and end based on the indexes of the two upper peaks and two lower peaks stored in the buffer and output them as the beginning and end addresses. The input signal is truncated based on the beginning and end addresses. The three-segment pattern is either upper peak-lower peak-upper peak or lower peak-upper peak-lower peak.

[0012] Furthermore, the main phase demodulation module and the auxiliary phase demodulation module are implemented in the same way. They both perform arctangent calculation on the Hilbert transform result, compare the difference between two adjacent points in the calculation result, and if the difference exceeds π, they add / subtract 2π from the previous point's phase to eliminate phase wrapping and generate a continuous phase curve. The main phase interpolation module and the auxiliary phase interpolation module are implemented in the same way. They are both used to interpolate the continuous phase curve to the length before the Hilbert transform.

[0013] Furthermore, the resampling module first normalizes the auxiliary phase demodulation result across the entire span and maps the normalized data to an auxiliary index interval that matches the number of output signal points of the main de-envelope module. Then, it searches and locates the main index interval of the de-envelope module output signal corresponding to each index in the auxiliary index interval. Based on the de-envelope module output signal of the main index interval, it uses linear interpolation to restore the amplitude of the de-envelope module output signal at that index position to obtain the resampling sequence.

[0014] Furthermore, the PL also includes a data transmission module. This module converts the four echo phase information channels from the PL clock domain to the PS clock domain via a FIFO, then stores the four echo phase information channels in the BRAM. After storage, it sends an interrupt signal to the PS, informing the PS that it can read the echo phase information. The PS then uses least-squares fitting to obtain the phase slope based on the read echo phase information, and then applies the formula... Calculate the distance value, where x 00 The value is the auxiliary arm length difference calibration value, and n0 and n are the refractive indices of the auxiliary interferometer and the main interferometer, respectively.

[0015] This invention is also achieved through the following technical solutions:

[0016] The ranging method implemented by the high-precision real-time FMCW laser ranging system based on the ZYNQ system as described above includes:

[0017] Step S1: The two main interferometers and the auxiliary interferometer generate triangular wave signals. The high-speed signal acquisition unit acquires the signals of the main and auxiliary interferometers by using up-sweep and down-sweep frequencies, and transmits the acquired two main up-sweep signals, two main down-sweep signals, auxiliary up-sweep signals and auxiliary down-sweep signals to PL.

[0018] Step S2: The two main up-scan signals, the two main down-scan signals, the auxiliary up-scan signal, and the auxiliary down-scan signal form two main-auxiliary up-scan signal pairs and two main-auxiliary down-scan signal pairs. The PL simultaneously performs parallel phase demodulation based on the main-auxiliary up-scan signal pairs and the main-auxiliary down-scan signal pairs to obtain four echo phase information, and transmits it to the PS.

[0019] Step S3 and PS calculate the four distance values ​​sequentially based on the phase information of the four echoes using least squares fitting.

[0020] The present invention has the following beneficial effects:

[0021] 1. The high-speed signal acquisition unit of this invention acquires the up-sweep and down-sweep frequency signals of the main and auxiliary interference signals respectively, and transmits the acquired two main up-sweep signals, two main down-sweep signals, auxiliary up-sweep signals, and auxiliary down-sweep signals to the PL. The two main up-sweep signals, two main down-sweep signals, auxiliary up-sweep signals, and auxiliary down-sweep signals form two main-auxiliary up-sweep signal pairs and two main-auxiliary down-sweep signal pairs. The PL simultaneously performs parallel phase demodulation based on the main-auxiliary up-sweep signal pairs and the main-auxiliary down-sweep signal pairs to obtain four echo phase information. The PS sequentially calculates four distance values ​​based on the four echo phase information through least squares fitting. In this way, by utilizing the parallel processing advantage of the PL part in the ZYNQ system and combining it with the general software execution capability of the PS, real-time and efficient distance measurement is achieved, and the accuracy of distance measurement is effectively improved.

[0022] 2. The resampling module in the demodulation process of this invention uses the phase provided by the auxiliary interferometer as the true horizontal axis to remap the main interferometer signal, which is originally non-uniform in time and affected by frequency sweep nonlinearity and trigger jitter, onto an equiphase grid. This can effectively eliminate distortions caused by time drift and frequency inconsistency, and enable the data to meet the prerequisites for algorithms such as FFT, filtering, and phase demodulation under ideal uniform coordinates, thereby improving the accuracy and stability of distance measurement and spectrum analysis. Attached Figure Description

[0023] The present invention will now be described in further detail with reference to the accompanying drawings.

[0024] Figure 1 This is a schematic diagram of the distance measurement optical path of the present invention.

[0025] Figure 2 This is a schematic diagram of the system of the present invention.

[0026] Figure 3 This is a schematic diagram of the data processing cycle of the present invention.

[0027] Figure 4 This is a schematic diagram of the demodulation process of the present invention.

[0028] Figure 5 This is a schematic diagram before the interference signal envelope is removed.

[0029] Figure 6 This is a schematic diagram after removing the envelope of the interference signal.

[0030] Figure 7 This invention is a top-bottom-top peak type.

[0031] Figure 8 This is the lower-upper-lower peak type of the present invention.

[0032] Figure 9 This is a test scenario diagram for the present invention.

[0033] Figure 10 This is a simulation diagram of the displacement stage ranging results of the present invention. Detailed Implementation

[0034] like Figure 1 The diagram shows the optical path for distance measurement. The basic principle of distance measurement is the frequency-modulated continuous wave beat frequency interference principle. The frequency or phase of the beat frequency signal is analyzed to extract the required distance information. Figure 1 In this process, an arbitrary waveform generator drives a laser source to output linearly frequency-modulated light. After passing through an optical isolator, the light is split into two paths at coupler 1: one path enters the main interferometer, then is introduced into the fiber collimator via a fiber optic circulator. Inside the fiber collimator, a short-path reference light and a long-path measurement light that travels through free space to the target and is reflected back are formed. These two paths converge at the output of the fiber optic circulator, and the main interference signal is collected by detector 1, whose beat frequency is proportional to the target distance L. The other path enters the auxiliary interferometer, where a reference interference signal with a known path difference is formed in the fixed ring delay line between couplers 2 and 3. This reference interference signal is collected by detector 2 to correct the laser frequency modulation nonlinearity and improve ranging accuracy. By synchronously analyzing the two beat frequencies, the frequency modulation process can be linearized, and the distance between the object being measured and the sensor head can be accurately calculated.

[0035] like Figure 2As shown, the high-precision real-time FMCW laser ranging system based on the ZYNQ system includes a laser source, two main interferometers, an auxiliary interferometer, and a high-speed signal acquisition unit and a real-time signal processing unit implemented based on the ZYNQ system. The real-time signal processing unit includes a PL (programmable logic unit) and a PS (processing system) connected via an AXI bus. The main and auxiliary interferometers' corresponding main and auxiliary interferometer signals each include an up-scan frequency signal and a down-scan frequency signal. The high-speed signal acquisition unit acquires the up-scan and down-scan frequency signals of the main and auxiliary interferometer signals respectively, and transmits the acquired two main up-scan signals, two main down-scan signals, auxiliary up-scan signals, and auxiliary down-scan signals to the PL. The two main up-scan signals, two main down-scan signals, auxiliary up-scan signals, and auxiliary down-scan signals form two main-auxiliary up-scan signal pairs and two main-auxiliary down-scan signal pairs. The PL simultaneously performs parallel phase demodulation based on the main-auxiliary up-scan signal pairs and main-auxiliary down-scan signal pairs to obtain four echo phase information. The PS sequentially calculates four distance values ​​based on the four echo phase information through least-squares fitting.

[0036] Specifically, the high-speed signal acquisition unit includes a DAC (digital-to-analog converter) module, three preamplifiers and ADC (analog-to-digital converter) modules connected sequentially, and a PL (power PL) module including a modulation signal generation module and a demodulation module. The modulation signal generation module generates a digital modulation signal. The DAC module is connected to the modulation signal generation module to convert the digital modulation signal into an analog modulation signal to modulate the laser source. The three preamplifiers are respectively connected to the photoelectric conversion modules of the two main interferometers and the auxiliary interferometer. The three ADC modules are respectively connected to the demodulation module to transmit the acquired signal to the demodulation module. The photoelectric conversion module is the detector, which is existing technology. In this embodiment, the analog modulation signal is a periodic triangular wave, and the operation of the modulation signal generation module is existing technology.

[0037] The PL also includes a clock management module, a reset management module, an AD data acquisition module, and a data transmission module. The clock management module generates the required clock through the Clocking Wizard IP core; for example, in this embodiment, the clock is 125MHz. The input clock for the modulation signal generation module is 10MHz, and the read clock for the cross-clock domain buffer in the data acquisition FIFO is 250MHz. The reset management module generates a global reset signal. The modulation signal generation module stores the digital modulation signal in ROM and uses a 10MHz loop to read 4000 data points stored in ROM, generating a triangular wave modulation signal with a modulation frequency of 2.5kHz. The data acquisition module configures the registers of the AD acquisition chip and buffers the acquired main interference signal and auxiliary interference signal in FIFOs for cross-clock domain processing. The data transmission module transmits the demodulated phase to the PS through the AXI CentralDirect Memory Access IP core.

[0038] In this embodiment, as Figure 3 As shown, the sampling rate of the main and auxiliary interference signals is 50MHz, and the signal period is 0.4ms. Considering that the overall processing time of the PL is within 1.4ms, the PL processing period is set to 1.6ms, meaning that the interference signal is processed once every four cycles. Since the processing flow of the upper and lower frequency sweep sections is the same, the PL independently processes the upper and lower frequency sweep distance demodulation tasks through module multiplexing. Therefore, two sets of distance and phase demodulation results can be obtained in one processing cycle, i.e., two distance calculations are achieved. That is, one main interference signal combined with one auxiliary interference signal can achieve a calculation speed of 1250 points / s. By processing two main interference signals in parallel, the calculation speed can reach 2500 points / s.

[0039] like Figure 5 As shown, the system processes two main interference signals and one auxiliary interference signal in parallel. That is, the demodulation module has a parallel two-way demodulation process. The demodulation process includes, in sequence, an auxiliary de-envelope module, an auxiliary interval-finding module, an auxiliary signal interpolation module, an auxiliary Hilbert transform module, an auxiliary phase demodulation module, and an auxiliary phase interpolation module; in sequence, a main moving average module and a main de-envelope module; a resampling module connected to both the main de-envelope module and the auxiliary phase interpolation module; and in sequence, a main interval-finding module, a main signal interpolation module, a main Hilbert transform module, a main phase demodulation module, and a main phase interpolation module. The output of the resampling module is connected to the main interval-finding module. An auxiliary up-scan signal or an auxiliary down-scan signal is input to the auxiliary envelope module. Two main up-scan signals or two main down-scan signals... Simultaneously, the main de-envelope module is input for parallel processing. The main and auxiliary de-envelope modules are used to de-envelope and normalize the input signal. The auxiliary interval finding module is used to extract the integer-cycle signal from the output signal of the auxiliary de-envelope module. The auxiliary signal interpolation module is used to perform linear interpolation on the integer-cycle signal to meet the needs of the auxiliary Hilbert transform module. The auxiliary Hilbert transform module outputs the in-phase and quadrature components of the linearly interpolated signal. The auxiliary phase demodulation module performs auxiliary phase demodulation based on the in-phase and quadrature components. The auxiliary phase interpolation module outputs the auxiliary phase demodulation result with the same length as the signal before the Hilbert transform. The resampling module is used to map the output signal of the main de-envelope module onto the reference phase grid obtained based on the auxiliary phase demodulation result.

[0040] When the main upper and lower scan signals are weak, a main moving average module connected to the main envelope removal module is added to the demodulation process. This main moving average module performs moving average filtering on the input signal to improve its quality. In this embodiment, the main moving average module performs a moving average on the original data. The window length for the moving average is set to five levels: 8, 16, 32, 64, and 128, to cover the main echo signal frequencies. A truncated window is used at the boundaries by default (if the window length is insufficient, only the average of existing elements is calculated). Specifically, for a window of length N, the output sample can be represented as...

[0041]

[0042] The main de-envelope module and the auxiliary de-envelope module serve the same purpose: to de-envelope and normalize the input signal, thereby correcting envelope distortion and standardizing the signal amplitude, preparing for subsequent integer truncation. Taking two main up-scan signals as the input signals to the main de-envelope module as an example, the two main up-scan signals are processed for de-envelope removal in parallel.

[0043] The implementation process of the main and auxiliary envelope removal modules is the same. Both first calculate the maximum and minimum values ​​of the input signal and the pre-measured light intensity fluctuation curve in the first 1 / 6 interval and the second 1 / 6 interval, respectively. Then, they apply the formula... The gain coefficient top_AC of the input signal is calculated. The offset top_AC of the input signal is calculated using the formula top_DC = max2 + max1 - (max02 + max01) * top_AC, where max1 is the maximum value of the input signal in the first 1 / 6 interval, max2 is the maximum value of the input signal in the second 1 / 6 interval, max01 is the upper envelope amplitude extracted from the corresponding position of the maximum value in the first 1 / 6 interval of the light intensity fluctuation curve, and max02 is the upper envelope amplitude extracted from the corresponding position of the maximum value in the second 1 / 6 interval of the light intensity fluctuation curve. Then, based on the calculated gain coefficient top_AC and offset top_DC, the upper envelope is precisely stretched and translated for alignment. Using the aligned envelope centerline and peak-valley difference as a reference, the input signal is de-enveloped and normalized before output. A schematic diagram of the interference signal before and after envelope removal is shown below. Figure 5 and Figure 6 As shown. For processing the lower envelope, simply replace the maximum value in the formula with the minimum value. The process of measuring the light intensity fluctuation curve is existing technology.

[0044] Both the auxiliary and primary interval-finding modules utilize the typical "up-down-up" or "down-up-down" three-segment peak shape at the beginning and end of the input signal to automatically locate a complete interference period interval, such as... Figure 7 and Figure 8As shown, only samples from the start point of the first cycle to the end point of the last cycle are retained to eliminate pseudo-spectral and phase jumps introduced by half-cycles, ensuring that subsequent interpolation, FFT and phase calculation are based on whole-cycle data and obtain the maximum and minimum parameters.

[0045] The implementation process of the auxiliary search interval module and the main search interval module is the same. Both first scan the head and tail of the input signal in parallel. The position index of the input signal when it is greater than the upper threshold or less than the lower threshold is written into the buffer in sequence and the upper peak or lower peak is marked. At the same time, the index of each first threshold value is recorded. The detection continues until two complete three-segment patterns appear in the buffer. Then, the position of the midpoint of the beginning and end is calculated based on the indexes of the two upper peaks and two lower peaks stored in the buffer and output as the beginning and end addresses. Finally, the input signal is truncated according to the beginning and end addresses. The three-segment pattern is either upper peak-lower peak-upper peak or lower peak-upper peak-lower peak.

[0046] Both the main signal interpolation module and the auxiliary signal interpolation module are used to perform linear interpolation on the truncated integer interval signal, uniformly interpolating it from its original arbitrary length to a fixed 8192 points to meet the input length requirements of the subsequent Hilbert transform.

[0047] The implementation process of the main signal interpolation module and the auxiliary signal interpolation module is the same. Both resample an input sequence of length N_cache to a length of 8192. Specifically, the process is as follows: first, a row vector n1 = 1:8192 is generated; then, the normalized position i = n1 / 8192 is calculated for each target output point i, and then mapped to the interpolated coordinates: n1_fix = i * N_cache; for cases where the output point falls outside the beginning or end, the endpoint values ​​are directly copied; otherwise, the classic linear weight K = n1_fix - addr1 is used to weight and accumulate the difference between two adjacent points to complete the interpolation: I_out(i) = I_in(addr1) + (I_in(addr2) - I_in(addr1)) * K, where addr1 refers to the nearest integer sampling address to the left of the new sequence interpolation point in the original input sequence, which is the result of rounding down the floating-point coordinate n1_fix of the target; addr2 refers to the address of the next point immediately adjacent to addr1, addr1+1.

[0048] Both the auxiliary and main Hilbert transform modules perform Hilbert transforms on their input signals. The process is identical for both modules: first, the input signal is subjected to an FFT transform; second, the FFT output is buffered using the internal dual-port RAM; and the first and last five points of the signal are set to 0 to remove DC interference, preventing baseline drift or artifacts caused by zero-frequency leakage in subsequent phase calculations and envelope detection. Next, the amplitude of the positive frequency band is multiplied by 2, the negative frequency band is set to 0, and an IFFT transform is performed. Finally, the real and imaginary parts of the Hilbert transform are obtained from the Xilinx IFFT IP output bus.

[0049] The main phase demodulation module and the auxiliary phase demodulation module are implemented in the same way. They both perform arctangent calculation on the Hilbert transform result to obtain the phase_out. They compare the difference between two adjacent points in the phase_out. If the difference exceeds π, they add / subtract 2π from the previous point's phase to eliminate phase wrapping and generate a continuous phase curve.

[0050] The main phase interpolation module and the auxiliary phase interpolation module are implemented in the same way. They are both used to interpolate the continuous phase curve to the length N_cache before the Hilbert transform, so as to ensure that the point index used by the subsequent algorithm is aligned with the actual sampling time.

[0051] The resampling module first normalizes the auxiliary phase demodulation result over the entire span, and then multiplies the normalized data by the number of points of the main de-envelope module output signal to map the normalized data to an auxiliary index interval (i.e., the true horizontal axis) that matches the number of points of the main de-envelope module output signal. Subsequently, a dual-pointer streaming search is used to locate the main index interval (i.e., the horizontal axis of the de-envelope module output signal) of the de-envelope module output signal to locate the interval where each integer index in the auxiliary index interval is located on the horizontal axis of the de-envelope module output signal. Based on the de-envelope module output signal in that interval, the amplitude of the de-envelope module output signal at that integer index position is restored by linear interpolation. In other words, the main de-envelope module output signal is mapped to the reference phase grid obtained from the auxiliary phase demodulation result to obtain a resampling sequence with equal step size.

[0052] The data transmission module converts the four echo phase information from the PL clock domain to the PS clock domain via FIFO, then stores the four echo phase information in BRAM. After storage, it sends an interrupt signal to the PS, informing the PS that it can read the echo phase information. The PS uses least squares fitting to obtain the phase slope based on the read echo phase information, and then applies the formula... Calculate the distance value, where x 00The calibration value is the length difference of the auxiliary arm. This calibration value is the fixed and known physical length difference or optical path difference between the reference arm and the measuring arm of the auxiliary interferometer. The reference arm and the measuring arm are existing technologies. n0 and n are the refractive indices of the auxiliary interferometer and the main interferometer, respectively.

[0053] In this embodiment, the Zynq XCZU7EV is selected, the DAC chip is AD9783, the ADC chip is AD9643 with a sampling frequency of 250MHz, the clock chip is LMK04828, the communication interface is Gigabit Ethernet, and the system board is equipped with 4GB of DDR4 memory.

[0054] The ranging methods implemented by the high-precision real-time FMCW laser ranging system based on the ZYNQ system include:

[0055] Step S1: The two main interferometers and the auxiliary interferometer generate triangular wave signals. The high-speed signal acquisition unit acquires the signals of the main and auxiliary interferometers by using up-sweep and down-sweep frequencies, and transmits the acquired two main up-sweep signals, two main down-sweep signals, auxiliary up-sweep signals and auxiliary down-sweep signals to PL.

[0056] Step S2: The two main up-scan signals, the two main down-scan signals, the auxiliary up-scan signal, and the auxiliary down-scan signal form two main-auxiliary up-scan signal pairs and two main-auxiliary down-scan signal pairs. The PL simultaneously performs parallel phase demodulation based on the main-auxiliary up-scan signal pairs and the main-auxiliary down-scan signal pairs to obtain four echo phase information, and transmits it to the PS.

[0057] Step S3 and PS calculate the four distance values ​​sequentially based on the phase information of the four echoes using least squares fitting.

[0058] Figure 9 In the test scenario diagram, the main interferometer is the spatial optical path, and the auxiliary interferometer is the fiber optic path. The displacement stage controls the target under test to move in 500-nanometer steps, and the absolute distance measurement results are as follows. Figure 10 As shown, the ranging accuracy (root mean square error) is 0.477 micrometers, and the resolution is 0.467 micrometers. It is evident that this invention achieves high-precision results with a distance measurement accuracy better than 1 micrometer and a resolution better than 500 nanometers.

[0059] The above description is merely a preferred embodiment of the present invention and should not be construed as limiting the scope of the present invention. All equivalent changes and modifications made in accordance with the scope of the patent application and the contents of the specification of the present invention should still fall within the scope of the patent of the present invention.

Claims

1. A high-precision real-time FMCW laser ranging system based on the ZYNQ system, characterized in that: It includes two main interferometers, an auxiliary interferometer, and a high-speed signal acquisition unit and a real-time signal processing unit based on the ZYNQ system. The real-time signal processing unit includes a PL and a PS connected together. The main and auxiliary interferometric signals corresponding to the main and auxiliary interferometers each include an up-sweep frequency signal and a down-sweep frequency signal. The high-speed signal acquisition unit acquires the up-sweep frequency signal and the down-sweep frequency signal of the main and auxiliary interferometric signals respectively, and transmits the acquired two main up-sweep signals, two main down-sweep signals, auxiliary up-sweep signals and auxiliary down-sweep signals to the PL. The two main up-sweep signals, two main down-sweep signals, auxiliary up-sweep signals and auxiliary down-sweep signals form two main-auxiliary up-sweep signal pairs and two main-auxiliary down-sweep signal pairs. The PL simultaneously performs parallel phase demodulation based on the main-auxiliary up-sweep signal pairs and the main-auxiliary down-sweep signal pairs to obtain four echo phase information. The PS sequentially calculates four distance values ​​based on the four echo phase information through least squares fitting.

2. The high-precision real-time FMCW laser ranging system based on the ZYNQ system according to claim 1, characterized in that: It also includes a laser source. The high-speed signal acquisition unit includes a DAC module, three preamplifiers and an ADC module connected in sequence. The PL includes a modulation signal generation module and a demodulation module. The modulation signal generation module is used to generate a digital modulation signal. The DAC module is connected to the modulation signal generation module to convert the digital modulation signal into an analog modulation signal to modulate the laser source. The three preamplifiers are connected to two main interferometers and an auxiliary interferometer, respectively. The three ADC modules are connected to the demodulation module to transmit the acquired signal to the demodulation module. The analog modulation signal includes a periodic triangular wave or a sine wave.

3. The high-precision real-time FMCW laser ranging system based on the ZYNQ system according to claim 2, characterized in that: The demodulation module has a parallel two-path demodulation process. The demodulation process includes, in sequence, an auxiliary envelope removal module, an auxiliary interval seeking module, an auxiliary signal interpolation module, an auxiliary Hilbert transform module, an auxiliary phase demodulation module, and an auxiliary phase interpolation module; a main envelope removal module; a resampling module connected to both the main envelope removal module and the auxiliary phase interpolation module; and, in sequence, a main interval seeking module, a main signal interpolation module, a main Hilbert transform module, a main phase demodulation module, and a main phase interpolation module. The output of the resampling module is connected to the main interval seeking module. An auxiliary up-scan signal or an auxiliary down-scan signal is input to the auxiliary envelope module. Two main up-scan signals or two main down-scan signals are simultaneously input to the main envelope removal module for parallel processing. The auxiliary de-envelope module is used to de-envelope and normalize the input signal. The auxiliary interval finding module is used to extract the integer-cycle signal from the output signal of the auxiliary de-envelope module. The auxiliary signal interpolation module is used to perform linear interpolation on the integer-cycle signal to meet the needs of the auxiliary Hilbert transform module. The auxiliary Hilbert transform module outputs the in-phase and quadrature components of the linearly interpolated signal. The auxiliary phase demodulation module performs auxiliary phase demodulation based on the in-phase and quadrature components. The auxiliary phase interpolation module outputs the auxiliary phase demodulation result with the same length as the signal before the Hilbert transform. The resampling module is used to map the output signal of the main de-envelope module onto the reference phase grid obtained based on the auxiliary phase demodulation result.

4. The high-precision real-time FMCW laser ranging system based on the ZYNQ system according to claim 3, characterized in that: When the main up-scan signal and the main down-scan signal are weak, a main moving average module connected to the main de-envelope module is added to the demodulation process. The main moving average module is used to perform moving average filtering on the input signal to improve the quality of the input signal.

5. The high-precision real-time FMCW laser ranging system based on the ZYNQ system according to claim 3 or 4, characterized in that: The main and auxiliary de-envelope modules are implemented in the same way, each calculating the maximum and minimum values ​​of its input signal and the pre-measured light intensity fluctuation curve in the first 1 / 6 interval and the second 1 / 6 interval, respectively, according to the formula... The input signal gain coefficient top_AC and offset top_DC are calculated respectively, and top_DC = max2 + max1 - (max02 + max01) * top_AC. Based on the calculated gain coefficient top_AC and offset top_DC, the upper envelope is precisely stretched and translated for alignment. Using the aligned envelope midline and peak-valley difference as a reference, the input signal is de-enveloped and normalized for output. Here, max1 is the maximum value of the input signal in the first 1 / 6 interval, max2 is the maximum value of the input signal in the second 1 / 6 interval, max01 is the upper envelope amplitude extracted from the corresponding position of the maximum value in the first 1 / 6 interval of the light intensity fluctuation curve, and max02 is the upper envelope amplitude extracted from the corresponding position of the maximum value in the second 1 / 6 interval of the light intensity fluctuation curve.

6. The high-precision real-time FMCW laser ranging system based on the ZYNQ system according to claim 3 or 4, characterized in that: The auxiliary search interval module and the main search interval module are implemented in the same way. They both scan in parallel from the head and tail of the input signal. The position index of the input signal when it is greater than the upper threshold or less than the lower threshold is written into the buffer in sequence and the upper peak or lower peak is marked. At the same time, the index of each first threshold value is recorded. The detection continues until two complete three-segment patterns appear in the buffer. The position of the midpoint of the beginning and end is calculated based on the indexes of the two upper peaks and two lower peaks stored in the buffer and output as the beginning and end addresses. The input signal is truncated based on the beginning and end addresses. The three-segment pattern is either upper peak-lower peak-upper peak or lower peak-upper peak-lower peak.

7. The high-precision real-time FMCW laser ranging system based on the ZYNQ system according to claim 3 or 4, characterized in that: The main phase demodulation module and the auxiliary phase demodulation module are implemented in the same way. They both perform arctangent calculation on the Hilbert transform result and compare the difference between two adjacent points in the calculation result. If the difference exceeds π, they add / subtract 2π from the previous point's phase to eliminate phase wrapping and generate a continuous phase curve. The main phase interpolation module and the auxiliary phase interpolation module are implemented in the same way. They are both used to interpolate the continuous phase curve to the length before the Hilbert transform.

8. The high-precision real-time FMCW laser ranging system based on the ZYNQ system according to claim 3 or 4, characterized in that: The resampling module first normalizes the auxiliary phase demodulation result across the entire span and maps the normalized data to an auxiliary index interval that matches the number of output signal points of the main de-envelope module. Then, it searches and locates the main index interval of the de-envelope module output signal for each index in the auxiliary index interval. Based on the de-envelope module output signal of the main index interval, it uses linear interpolation to restore the amplitude of the de-envelope module output signal at that index position to obtain the resampling sequence.

9. The high-precision real-time FMCW laser ranging system based on the ZYNQ system according to claim 3 or 4, characterized in that: The PL also includes a data transmission module. This module converts the four echo phase information channels from the PL clock domain to the PS clock domain via a FIFO, then stores the four echo phase information channels in the BRAM. After storage, it sends an interrupt signal to the PS, informing the PS that it can read the echo phase information. The PS then uses least-squares fitting to obtain the phase slope based on the read echo phase information, and then applies the formula... Calculate the distance value, where x 00 The value is the auxiliary arm length difference calibration value, and n0 and n are the refractive indices of the auxiliary interferometer and the main interferometer, respectively.

10. A ranging method implemented by the high-precision real-time FMCW laser ranging system based on the ZYNQ system as described in any one of claims 1 to 9, characterized in that: include: Step S1: The two main interferometers and the auxiliary interferometer generate triangular wave signals. The high-speed signal acquisition unit acquires the signals of the main and auxiliary interferometers by using up-sweep and down-sweep frequencies, and transmits the acquired two main up-sweep signals, two main down-sweep signals, auxiliary up-sweep signals and auxiliary down-sweep signals to PL. Step S2: The two main up-scan signals, the two main down-scan signals, the auxiliary up-scan signal, and the auxiliary down-scan signal form two main-auxiliary up-scan signal pairs and two main-auxiliary down-scan signal pairs. The PL simultaneously performs parallel phase demodulation based on the main-auxiliary up-scan signal pairs and the main-auxiliary down-scan signal pairs to obtain four echo phase information, and transmits it to the PS. Step S3 and PS calculate the four distance values ​​sequentially based on the phase information of the four echoes using least squares fitting.

Images