Fsi dynamic ranging method based on higher order linear prediction - least mean square error

Abstract

The application discloses a kind of FSI dynamic ranging methods based on high-order linear prediction-minimum mean square error, mainly solve the problem of slow speed of existing FSI ranging system to the distance to be measured solution, insufficient dynamic measurement capability.Solution includes: 1) construct to generate the optical frequency scanning interference ranging system required for distance solution interference signal;2) set the sweep rate of laser, so that it carries out optical frequency continuous tuning output, after interference, generates the sinusoidal interference signal that changes continuously on time domain after detection;3) the data acquisition device in FSI system carries out real-time acquisition to sinusoidal interference signal, and sample point is transmitted to measurement module;4) measurement module utilizes high-order linear prediction-minimum mean square error cascade algorithm to carry out real-time prediction to sampling data, and utilizes the interference signal frequency obtained to complete the real-time solution of distance.The application can make the distance solution speed of FSI system significantly improve, effectively improve the ranging rate.

Description

Technical Field

[0001] This invention belongs to the field of laser precision measurement technology, and further relates to a dynamic ranging method based on high-order linear prediction-minimum mean square error optical frequency scanning interferometry (FSI), which is used to dynamically track and calculate the frequency of the interference signal in ranging. Background Technology

[0002] Currently, precision measurement, represented by laser precision ranging, has become a crucial technological support for scientific and technological development and engineering applications. Popular research fields and major engineering projects such as space satellite formation, space gravitational wave detection, automated assembly of ultra-large equipment, and autonomous driving all demonstrate an urgent technological demand for laser precision ranging. Among numerous ranging technologies, optical frequency scanning interferometric ranging, as a non-contact, light-disconnectable, and high-precision ranging technology, has already been practically applied in scientific research projects and engineering. Currently, with the continuous advancement of science and technology, the requirements for laser ranging performance are no longer limited to high precision alone; real-time distance measurement and dynamic measurement capabilities are also required. Improving ranging performance poses severe challenges to both the hardware and algorithms of the measurement system. While hardware improvements can enhance ranging performance to a certain extent, they also increase the construction cost of the measurement system, which is not conducive to the application and development of optical frequency scanning interferometric ranging. Compared to improving hardware performance, improving and optimizing measurement algorithms is an economical and effective approach. Continuous improvement and optimization of measurement algorithms have significant value in both controlling system design costs and improving system measurement performance. Therefore, there is an urgent need to develop a ranging scheme that can improve the system's measurement performance while effectively controlling the system's design cost.

[0003] Among existing measurement algorithms, the following are some common measurement techniques: 1) Measurement algorithms based on time-domain phase characteristics. This type of method requires phase difference interception within a large optical frequency scanning bandwidth. On the one hand, it requires the laser to have a large scanning bandwidth, and on the other hand, it requires a certain amount of time accumulation, making it impossible to achieve fast and real-time solution of the measured distance. 2) Spectrum measurement algorithms based on frequency measurement methods such as Fourier transform. When applying this method, the extraction of the spectral peaks of the interference signal is easily affected by the environment and laser performance, making it difficult to achieve accurate extraction of the interference signal frequency. This type of method still requires sacrificing some time for distance solution, making it impossible to achieve real-time distance solution. In addition, when performing dynamic optical path measurement, the change of optical path causes the spectrum of the interference signal to broaden, making it impossible to achieve dynamic measurement. 3) FSI distance measurement algorithm based on Kalman filtering and its improved algorithms. Kalman filtering is designed for Gaussian environments and is only suitable for parameter estimation of linear systems. However, FSI ranging systems are nonlinear systems, and the resulting interference signals are nonlinear, making it impossible to effectively measure distance. Even if improved Kalman filtering algorithms can handle nonlinear signals, they still face problems such as low estimation accuracy, high algorithm complexity, and low efficiency.

[0004] 4) The FSI distance measurement algorithm based on particle filtering is affected by the number of particles. A large number of particles will result in a large amount of computation, while a finite number of particles will inevitably cause particle degradation after several iterations. In addition, the calculation of the likelihood function and the selection of the importance probability density function will directly affect the efficiency and accuracy of the particle filtering algorithm in distance calculation. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by proposing a dynamic ranging method for FSI based on high-order linear prediction and minimum mean square error. This method solves the problems of slow distance calculation speed, insufficient dynamic measurement capability, and high system design cost in current FSI ranging systems. The method cascades and combines a high-order linear prediction algorithm with a minimum mean square prediction algorithm to calculate the frequency of the interference signal at the sampling point in real time, and uses the measured frequency to achieve real-time distance calculation. This invention enables dynamic measurement of targets by the FSI system and effectively improves the system's measurement speed.

[0006] The basic idea behind this invention is as follows: First, an optical frequency scanning interferometric ranging system is constructed to generate the interference signal required for distance calculation. The laser's sweep rate is set so that the laser continuously tunes its output. After passing through the interference optical path, the continuously tuned light source is detected by a detector, generating a continuously varying sinusoidal interference signal in the time domain. The data acquisition device in the FSI system acquires the sinusoidal interference signal in real time and transmits the sampling points to the measurement module. The measurement module uses a high-order linear prediction-minimum mean square error cascade algorithm to process the sampled data in real time, completing the real-time prediction of the interference signal frequency, and using the obtained interference signal frequency to calculate the distance in real time. The high-order linear prediction-minimum mean square error cascade algorithm refreshes the interference signal frequency in real time using the sampling data at the latest sampling moment, enabling the distance calculation speed of the FSI system to reach the system's sampling speed, greatly improving the distance measurement rate of the FSI ranging system.

[0007] The specific steps of this invention to achieve the above objectives are as follows:

[0008] (1) A frequency scanning interferometric FSI ranging system is built using a tunable external cavity semiconductor laser ECDL, collimator, beam splitter, Fabry-Perot FP etalon, photodetector and Michelson interferometric optical path, and a data acquisition device is equipped in the system.

[0009] (2) The tunable external cavity semiconductor laser ECDL of the FSI ranging system generates a tunable light source S whose frequency varies with time t and outputs it to the collimator.

[0010] (3) After the collimator collimates the tuned light source S, it is split into two beams S1 and S2 by a beam splitter prism. S1 enters the Fabry-Perot FP etalon and S2 enters the Michelson interference optical path.

[0011] (4) The Fabry-Perot FP etalon acquires the transmission signal of S1 and uses a photodetector to detect the transmission signal to obtain the FP signal; the photodetector in the Michelson interferometer detects S2 after it enters the optical path and undergoes beam splitting, reflection and convergence interference to obtain the interference signal;

[0012] (5) Use the data acquisition device to sample the FP signal obtained in step (4) and preprocess it to obtain the linear frequency modulation rate β of the laser;

[0013] (6) Use a data acquisition device to sample the interference signal obtained in step (4) in real time;

[0014] (7) The high-order linear prediction filtering algorithm and the minimum mean square error prediction algorithm (LMS) are cascaded and combined to obtain the LMS-HOLP filtering algorithm.

[0015] (8) The data acquisition device transmits the real-time sampled data of the interference signal to the LMS-HOLP filtering algorithm for processing, thereby completing the real-time prediction of the interference signal frequency, as follows:

[0016] (8.1) The high-order linear prediction algorithm uses the first M sample values ​​of the interference signal to provide the optimal initial filtering coefficients for the LMS-HOLP filtering algorithm, and assigns the initial optimal filtering coefficients to the LMS filtering algorithm.

[0017] (8.2) Sample the first M terms of the interference signal The signal is input to the LMS filter to obtain the expected value of the signal at the current time k.

[0018] (8.3) Using the expected value of the signal The estimation error e is obtained by subtracting the actual sampled value from the value at the current time k. k Obtain the estimated error e k The z-transform Q(z) of the impulse response is obtained, its roots are found, and the optimal root is defined. Obtain the instantaneous phase θ0 of the signal;

[0019] (8.4) Solve for the optimal estimate of the frequency of the interference signal corresponding to the current time k according to the following formula:

[0020]

[0021] (9) The optimal estimate of the linear modulation rate β of the laser obtained in step (5.2) and the frequency of the interference signal obtained in step (8.4) Use the distance measurement formula to calculate the distance at time k. The solution;

[0022] (10) The data acquisition device transmits the first M sampled values ​​of the interference signal to the LMS-HOLP filtering algorithm at the next moment to complete the optimal estimate of the frequency of the interference signal corresponding to the current moment k. The replacement is performed to complete the distance calculation for the next time step; finally, the distance value for each sampling time step is calculated to obtain the distance to be measured.

[0023] Compared with the prior art, the present invention has the following advantages:

[0024] First, because this invention uses minimum mean square error for prediction, it estimates the frequency of the signal based on the current sample value and the sample value at past times. It is applicable to the frequency prediction of nonlinear signals and has high frequency prediction accuracy. It overcomes the shortcomings of existing technologies in terms of insufficient frequency prediction capability and low efficiency for nonlinear signals, and is suitable for application in FSI ranging systems.

[0025] Secondly, since the cascaded algorithm proposed in this invention does not require knowing the form of noise contained in the signal, the FSI dynamic ranging method based on high-order linear prediction-minimum mean square error has no special requirements for the measurement environment and does not require knowing the nature of noise in the measurement environment. It has wide applicability and helps to promote the practical application of FSI ranging technology.

[0026] Third, the algorithm used in this invention does not require the establishment of a complex mathematical model, making the method very simple in principle, suitable for porting to hardware systems, and able to improve the online measurement capability of the ranging system, thus possessing extremely high practical application value; at the same time, since the algorithm has few parameters, it does not require manual selection or setting of parameters, and has a high degree of adaptability. Attached image description:

[0027] Figure 1 This is a flowchart illustrating the implementation of the present invention;

[0028] Figure 2 This is a schematic diagram of the overall structure of the FSI ranging system constructed in this invention;

[0029] Figure 3 This is a time-domain distribution diagram of the discrete FP signal in this invention;

[0030] Figure 4 This is a time-domain distribution diagram of the discrete interference signal in this invention;

[0031] Figure 5 This is a schematic diagram of a high-order linear prediction filter.

[0032] Figure 6 This is a schematic diagram of the minimum mean square error prediction filter. Detailed Implementation

[0033] The present invention will now be further described with reference to the accompanying drawings.

[0034] Example 1:

[0035] See attached document Figure 1 The present invention proposes a dynamic ranging method for FSI based on high-order linear prediction and minimum mean square error, which specifically includes the following steps:

[0036] Step 1: Construct an optical frequency scanning interferometric (FSI) ranging system using a tunable external cavity semiconductor laser (ECDL), a collimator, a beam splitter prism, a Fabry-Perot etalon (FP), a photodetector, and a Michelson interferometer optical path, and equip the system with a data acquisition device; wherein: the tunable external cavity semiconductor laser (ECDL) is used to generate a tunable light source with continuously changing frequency; the Michelson interferometer optical path includes an fiber collimator (FC), a beam splitter prism (BS), a corner bevel reflecting prism (RR), a measurement reflecting prism (RM), and a photodetector (PD).

[0037] Step 2: The tunable external cavity semiconductor laser ECDL of the FSI ranging system generates a tunable light source S whose frequency varies with time t and outputs it to the collimator.

[0038] Step 3: After the collimator collimates the tuned light source S, it is split into two beams, S1 and S2, using a beam splitter prism. S1 enters the Fabry-Perot FP etalon and S2 enters the Michelson interference optical path.

[0039] Step 4: The Fabry-Perot FP etalon acquires the transmission signal of S1 and uses a photodetector to detect the transmission signal to obtain the FP signal; the photodetector in the Michelson interferometer detects S2 after it enters the optical path and undergoes beam splitting, reflection and convergence interference to obtain the interference signal;

[0040] The transmission signal generated by light S1 in the Fabry-Perot FP etalon is converted by the FP controller and finally output as an FP signal; S2 is coupled into space through the fiber collimator FC, and light S2 is split into two new beams S by the beam splitter prism BS. 21 and S 22 S 21 As the reference light for the ranging optical path, S 22 The measuring light in the ranging optical path is reflected by the cornerstone reflecting prism RR and the measuring reflecting prism RM respectively, and then converges and interferes at the beam splitter prism BS. The interference signal is detected by the photodetector PD.

[0041] Step 5: The data acquisition device is used to acquire the interference signal in real time, realizing the sampling of the FP signal and the interference signal; the FP signal obtained in step (4) is sampled using the data acquisition device and preprocessed to obtain the linear frequency modulation rate β of the laser; the implementation steps are as follows:

[0042] (5.1) The optical frequency tuning range Δv of the ECDL output tuned light source of the laser is calculated based on the number of resonant peaks of the FP signal:

[0043] Δv=(r-1)·FSR,

[0044] Where r represents the number of resonance peaks of the FP signal, and FSR represents the optical frequency difference between two adjacent resonance peaks;

[0045] (5.2) Using the optical frequency tuning range Δv, the linear frequency modulation rate β of the laser is obtained according to the following formula:

[0046]

[0047] Among them, F sThe sampling rate of the measurement system is represented by N, which represents the number of sampling points corresponding to the FP signal.

[0048] Step 6: Use a data acquisition device to sample the interference signal obtained in step 4 in real time;

[0049] Step 7: Cascade the higher-order linear prediction filter algorithm and the minimum mean square error prediction algorithm (LMS) to obtain the LMS-HOLP filter algorithm. The cascade combination refers to the higher-order linear prediction filter obtaining a set of optimal filter coefficients based on the first M sample values, and then assigning these optimal filter coefficients to the same-order minimum mean square error filter to form a combination.

[0050] Step 8: The data acquisition device transmits the real-time sampled data of the interference signal to the LMS-HOLP filtering algorithm for processing, thereby completing the real-time prediction of the interference signal frequency, as follows:

[0051] (8.1) The high-order linear prediction algorithm uses the first M sample values ​​of the interference signal to provide the optimal initial filtering coefficients for the LMS-HOLP filtering algorithm, and assigns the initial optimal filtering coefficients to the LMS filtering algorithm.

[0052] (8.2) Sample the first M terms of the interference signal The signal is input to the LMS filter to obtain the expected value of the signal at the current time k. Specifically as follows:

[0053]

[0054] Among them, W k This represents the coefficient matrix of the LMS filter at time k.

[0055] (8.3) Using the expected value of the signal The estimation error e is obtained by subtracting the actual sampled value from the value at the current time k. k Obtain the estimated error e k The z-transform Q(z) of the impulse response is obtained, its roots are found, and the optimal root is defined. Obtain the instantaneous phase θ0 of the signal;

[0056] The estimation error e k The z-transform of the impulse response, Q(z), is expressed as follows:

[0057]

[0058] Where M represents the order of the filter, z is the representation of the z-transform, W is the coefficient matrix of the LMS filter, i represents the sequence number of the coefficient in the coefficient matrix W, and W(i) represents the i-th coefficient in the coefficient matrix of the LMS filter.

[0059] The process of finding the root of Q(z) and the optimal root refers to finding the root that is located inside the unit circle and is closest to the unit circle among the obtained roots. This optimal root contains the true frequency of the signal.

[0060] (8.4) Solve for the optimal estimate of the frequency of the interference signal corresponding to the current time k according to the following formula:

[0061]

[0062] Step 9: Based on the optimal estimate of the linear modulation rate β of the laser obtained in Step 5 and the interference signal frequency obtained in Step 8.4. Use the distance measurement formula to calculate the distance at time k. The solution is as follows:

[0063]

[0064] Where c is the speed of light and n is the refractive index of air.

[0065] Step 10: The data acquisition device transmits the first M sampled values ​​of the interference signal to the LMS-HOLP filtering algorithm at the next moment to complete the optimal estimate of the interference signal frequency corresponding to the current moment k. The replacement is performed to complete the distance calculation for the next time step; finally, the distance value for each sampling time step is calculated to obtain the distance to be measured.

[0066] Example 2: Refer to Appendix Figure 2 The overall implementation steps of the method of the present invention are the same as those in Embodiment 1. The FSI ranging optical path in the present invention will now be described in further detail.

[0067] The optical frequency scanning interferometric (FSI) ranging system includes: a tunable external cavity semiconductor laser ECDL, a collimator, a first beam splitter prism BS1, a Fabry-Perot FP etalon, a first photodetector PD1, a Michelson interferometer optical path, and a data acquisition device.

[0068] The tunable external cavity semiconductor laser ECDL is used to generate a tunable light source with continuously varying frequency. After being collimated by a collimator, the tunable light source is split into two beams, S1 and S2, by an optical fiber beam splitter. One beam, S1, directly enters the Fabry-Perot FP etalon. The transmission signal generated by the beam S1 in the Fabry-Perot FP etalon is converted by the FP controller and finally output as an FP signal. The other beam, S2, enters the Michelson interferometer optical path.

[0069] The Michelson interferometer optical path includes a fiber collimator FC, a beam splitter BS, a pyramidal reflecting prism RR, a measurement reflecting prism RM, and a photodetector PD; S2 is coupled out into space through the fiber collimator FC, and light S2 is split into two new beams S by the beam splitter BS. 21 and S 22 S 21 As the reference light for the ranging optical path, S 22 The measuring light in the ranging optical path is reflected by the cornerstone reflecting prism RR and the measuring reflecting prism RM respectively, and then converges and interferes at the beam splitter prism BS. The interference signal is detected by the photodetector PD.

[0070] The data acquisition device DAQ is used to sample FP signals and interference signals;

[0071] By acquiring and processing the FP signal, the laser optical frequency change rate β can be calculated.

[0072] DAQ acquires interference signals in real time, and transmits the sampled data to software on a computer PC for real-time calculation of the interference signal frequency and the measured distance.

[0073] Example 3: The overall implementation steps of the method of the present invention are the same as those of Example 1, wherein in step 7, the high-order linear prediction filtering algorithm and the minimum mean square error prediction algorithm (LMS) are cascaded and combined to obtain the LMS-HOLP filtering algorithm; the principle of using the LMS-HOLP filtering algorithm to realize FSI dynamic ranging will now be explained in more detail.

[0074] (I) Principle of Higher-Order Linear Prediction: Refer to Appendix Figure 5 Let u(n), u(n-1), ..., u(nM) be the discrete-time series of a stationary random process, representing the M sample values ​​at time n and before that time. The essence of prediction is to estimate the sample value u(n) at time n using a set of sample values ​​u(n), u(n-1), ..., u(nM). The optimal filter tap weight vector of the higher-order linear predictive filter (M×1) is W. k The expansion is:

[0075] W k =[w k,1 ,w k,2 ,…,w k,M ] T

[0076] The optimal filter tap weight vector is solved using the Wiener-Hough equation, which requires knowing two quantities:

[0077] 1) The M×M correlation matrix of tap input u(n-1), u(n-2), ..., u(nM)

[0078] 2) The M×1 cross-correlation vector between the tap input and the expected u(n).

[0079] Define the M×1 dimensional tapped input vector U(n-1) as follows:

[0080] U(n-1)=[u(n-1),u(n-2),...u(nM)] T

[0081] The M×M correlation matrix of the input vector U(n-1) is:

[0082]

[0083] Where r(k) is the autocorrelation function of the input process with a delay of k, where k = 0, 1, ..., M-1, (·) H This is the complex conjugate transpose.

[0084] The M×1-dimensional cross-correlation vector between the input vector U(n-1) and the desired vector u(n) is:

[0085]

[0086] The variance of the input u(n) at time n can be expressed as r(0). After obtaining the above three quantities, the optimal tap weight vector W of the higher-order linear prediction filter can be solved using the Wiener-Hough equation. k :

[0087] RW k =r.

[0088] (II) Minimum Mean Square Error Prediction Principle: Refer to Appendix Figure 6 ,in The input data is at time k, y k It is the expected signal. It utilizes adaptive weights W k The weight vector is obtained by minimizing the error signal e from the estimated signal. k get.

[0089] In the adaptive filtering phase, the transverse filter applies a different input vector X. k Response, obtain the desired signal

[0090]

[0091] In the above formula, W k The initial value is given by a higher-order linear prediction filter.

[0092] The output signal of the transverse filter As for the desired signal yk The estimation is obtained, and the estimation error e is obtained. k :

[0093]

[0094] Substituting the error signal e(k) into the adaptive weight control algorithm yields the adaptive adjustment of the filter weights:

[0095]

[0096] In the above formula, the step size of the μ filter's adaptive iteration is... The error performance surface with respect to the filter coefficients W k The gradient is expressed as:

[0097]

[0098] The initial coefficients of the filter are assigned by the higher-order linear prediction filter, and the minimum mean square error filter first calculates the error signal e. k Then, using the error signal e k The coefficients of the filter are updated iteratively.

[0099] (III) Principle of High-Order Linear Minimum Mean Square Error FSI Dynamic Ranging Algorithm:

[0100] Let the form of the sampled interference signal be:

[0101]

[0102] Where v(k) is the unknown noise at time k, Let be the phase of the interference signal sampled at time k, which can be expressed as:

[0103]

[0104] Where f(k) is the frequency of the discrete interference signal at time k.

[0105] The prediction model for the minimum mean square error filter is:

[0106]

[0107] Where M is the length of the LMS filter, i.e. the order of the LMS filter, and W(i) are the corresponding x(ni) filter coefficients. The optimal initial value is assigned by the higher-order linear prediction filter.

[0108] The prediction error with minimum mean square error is defined as:

[0109]

[0110] According to the above equation, the Z-transform of the prediction error impulse response is:

[0111]

[0112] An M-order minimum mean square error filter corresponds to M roots. The root that lies within the unit circle and is closest to the unit circle is considered the optimal root (i.e., contains the true frequency of the signal). This optimal root is defined as... θ0 is the phase of the signal, and the sampling rate is defined as F. s Then the frequency estimate of the signal is:

[0113]

[0114] As sampling continues, the frequency of the interference signal can be continuously refreshed.

[0115] The measured Substituting into the following formula, the distance to be measured can be obtained. Real-time solution:

[0116]

[0117] In the above formula, β is the frequency modulation rate of the laser, c is the speed of light, n is the refractive index of air, and k represents time k.

[0118] The method proposed in this invention can effectively control the system design cost while improving the system measurement performance, which is of great significance for the practical application of FSI measurement systems.

[0119] The parts of this invention not described in detail are common knowledge to those skilled in the art.

[0120] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Obviously, those skilled in the art, after understanding the content and principle of the present invention, may make various modifications and changes in form and detail without departing from the principle and structure of the present invention. However, these modifications and changes based on the concept of the present invention are still within the scope of protection of the claims of the present invention.

Claims

1. A FSI dynamic ranging method based on high order linear prediction and minimum mean square error, characterized in that, It comprises the following steps: (1) An optical frequency scanning interference (FSI) ranging system is built by using a tunable external cavity diode laser (ECDL), a collimator, a beam splitter prism, a Fabry-Perot (FP) etalon, a photodetector and a Michelson interference optical path, and a data acquisition device is arranged in the system; (2) A tunable light source S with a frequency varying with time t is generated by the tunable external cavity diode laser (ECDL) of the FSI ranging system and output to the collimator; (3) After collimation of the tunable light source S by the collimator, the beam splitter prism is used to divide the tunable light source S into two beams of S1 and S2, and S1 enters the Fabry-Perot (FP) etalon and S2 enters the Michelson interference optical path; (4) The FP etalon obtains the transmission signal of S1, and the photodetector is used to detect the transmission signal to obtain an FP signal; the photodetector in the Michelson interference optical path detects S2 after splitting, reflection and convergent interference to obtain an interference signal; (5) The data acquisition device is used to sample the FP signal obtained in step (4) and pre-process the FP signal to obtain a linear frequency modulation rate β of the laser; (6) The data acquisition device is used to sample the interference signal obtained in step (4) in real time; (7) A high-order linear prediction filter algorithm and a least mean square error prediction algorithm (LMS) are cascaded and combined to obtain an LMS-HOLP filter algorithm; (8) The data acquisition device transmits the real-time sampling data of the interference signal to the LMS-HOLP filter algorithm for processing to complete real-time prediction of the frequency of the interference signal, and the following is realized: (8.1) The high-order linear prediction algorithm uses the first M sampling values of the interference signal to provide optimal initial filter coefficients for the LMS-HOLP filter algorithm, and the initial optimal filter coefficients are assigned to the LMS filter algorithm; (8.2) the first M sample values of the interference signal are input to the LMS filter to obtain the signal expectation value corresponding to the current time k (8.3) Signal expectation value The estimated error e is obtained by subtracting the actual sampling value at the current time k k , obtaining the z transform Q(z) of the impulse response of the estimated error e k , finding the optimal root by taking the root and finding the optimal root, and defining the optimal root The instantaneous phase θ0 of the signal is obtained; (8.4) The optimal estimation value of the frequency of the interference signal corresponding to the current time k is solved according to the following formula: (9) the optimal estimate of the frequency of the interference signal obtained in step (8.4) and the chirp rate β of the laser obtained in step (5) The distance at time k is calculated by using the distance measurement formula . (10) The data acquisition device transmits the first M sample values of the interference signal to the LMS-HOLP filtering algorithm at the next moment, to complete the optimal estimation value of the interference signal frequency corresponding to the current moment k , so as to complete the distance calculation at the next moment; and finally, the distance value at each sampling moment is calculated to obtain the to-be-measured distance.

2. The method of claim 1, wherein: The FSI ranging system built in step (1), wherein: the tunable external cavity semiconductor laser ECDL is used to generate a frequency continuously changing tuning light source; the transmission signal of the light S1 generated in the Fabry-Perot FP etalon is converted via the FP controller control and finally outputs the FP signal; the Michelson interference optical path includes a fiber collimator FC, a beam splitter BS, a corner cube reflector RR, a measurement reflector RM and a photodetector PD; S2 is coupled out into space through the fiber collimator FC, and the light S2 is divided into two new beams S 21 and S 22 , wherein S 21 is the reference light of the ranging optical path, S 22 is the measurement light of the ranging optical path, both of which are reflected by the corner cube reflector RR and the measurement reflector RM, converge and interfere at the beam splitter BS, and are detected by the photodetector PD to obtain an interference signal; the data acquisition device is used to collect the interference signal in real time to realize sampling of the FP signal and the interference signal.

3. The method of claim 1, wherein: The linear frequency modulation rate β of the laser obtained in step (5) is realized as follows: (5.1) The optical frequency tuning range Δv of the tunable light source output by the laser ECDL is calculated according to the number value of the resonance peaks of the FP signal: Δv = (r-1)·FSR, wherein r represents the number of resonance peaks of the FP signal, and FSR represents the optical frequency difference corresponding to two adjacent resonance peaks; (5.2) The linear frequency modulation rate β of the laser is solved according to the following formula by using the optical frequency tuning range Δv: where F s represents the sampling rate of the measurement system, and N represents the number of sampling points corresponding to the FP signal.

4. The method of claim 1, wherein: The cascade combination in step (7) means that a high-order linear prediction filter obtains a set of optimal filter coefficients according to the first M sampling values, and the optimal filter coefficients are assigned to a least mean square error filter of the same order, so as to form a combination.

5. The method of claim 1, wherein: The signal expected value at the current time k described in step (8.2) In detail as follows: where W k denotes the coefficient matrix of the LMS filter at time k.

6. The method of claim 1, wherein: The estimate error e of step (8.3) k The z-transform Q(z) of the impulse response of the filter is given by Wherein M represents the order of the filter, z is the representation form of z transform, W is the coefficient matrix of the LMS filter, i represents the serial number of the coefficients in the coefficient matrix W, and W(i) represents the i th coefficient in the coefficient matrix of the LMS filter.

7. The method of claim 6, wherein: The roots of Q(z) are solved and the optimal root is searched in step (8.3), which means that the root located in the unit circle and closest to the unit circle is searched from the obtained roots, and the optimal root contains the real frequency of the signal.

8. The method of claim 1, wherein: The distance measurement formula is used to complete the distance calculation at time k in step (9) as follows: The distance measurement formula is used to complete the distance calculation at time k in step (9) as follows: Wherein c is the speed of light, and n is the air refractive index.

Images