Detection signal noise suppression method of laser wind finding radar

Abstract

The invention relates to the technical field of signal processing, in particular to a detection signal noise suppression method of a laser wind finding radar. Based on spatial physical redundancy of laser radar multi-range gate observation, the technical problem that a real wind field and platform noise are difficult to separate under the frequency domain aliasing condition is solved by utilizing the physical characteristics that platform rigid body motion is a full-field common modulus and atmospheric turbulence is a spatial differential modulus. By calculating a spatial frequency standard deviation sequence, the fuzziness of pure time dimension analysis is relieved by utilizing the redundancy of spatial information. According to the technical means, a motion inversion algorithm can adaptively adjust a trajectory smoothing weight coefficient according to environmental turbulence intensity, inertial constraints are automatically reduced at a strong turbulence moment to retain high-frequency wind speed characteristics, and constraints are enhanced at a weak turbulence moment to smooth observation noise; therefore, the method achieves the accurate stripping of the common-mode motion track of the platform under a complex meteorological condition, effectively inhibits the noise of a detection signal, and improves the wind speed inversion precision.

Description

Technical Field

[0001] This invention relates to the field of signal noise processing technology, and specifically to a method for suppressing noise in the detection signal of a laser wind-measuring radar. Background Technology

[0002] Laser wind radar inverts wind field velocity by emitting a laser beam and receiving the backscattered echoes from atmospheric aerosols. When the radar is mounted on a dynamic platform such as a ship, buoy, or airborne vessel, the platform's high-frequency rigid body motion introduces an additional Doppler frequency shift into the echo signal. To achieve long-range detection, radar typically employs incoherent accumulation techniques to improve the signal-to-noise ratio (SNR) of weak signals. However, during the accumulation period, the platform's rapid movement causes the Doppler frequency of the echo signal to drift at different times. This non-stationarity results in energy dispersion and broadening of the accumulated signal spectrum, preventing the SNR from effectively improving with accumulation time and severely limiting the radar's long-field detection capability in dynamic environments.

[0003] In practice, the inventors discovered the following defects in the aforementioned prior art: Existing technologies typically rely on inertial measurement units (IMUs) for motion compensation. However, in low-cost applications or under conditions of severe shaking, the measurement residuals of the IMUs still remain in the signal. Extracting motion parameters from the radar echo signal itself is an effective supplementary method, but in real atmospheric environments, the wind field itself undergoes turbulent changes. The rapid wind speed fluctuations caused by turbulence and the velocity changes caused by platform motion are highly overlapping in the frequency domain. Traditional low-pass filtering methods struggle to distinguish between the two, easily misinterpreting real wind shear as platform noise for filtering, or failing to completely eliminate platform interference, leading to a decrease in the accuracy of the final wind speed inversion. Summary of the Invention

[0004] To address the technical problem of existing technologies' inability to distinguish between rapid wind speed fluctuations caused by turbulence and velocity changes caused by platform motion, which exhibit highly mixed frequency domain characteristics, the present invention aims to provide a method for suppressing noise in the detection signal of a laser wind-measuring radar. The specific technical solution adopted is as follows: This invention proposes a method for suppressing noise in the detection signal of a laser wind-measuring radar, the method comprising: The raw echo signal of the radar over the entire observation period is obtained, and the power spectral density matrix is ​​obtained by short-time slicing and spectral transformation of the raw echo signal. Reference range gates are selected based on the carrier-to-noise ratio of each initial range gate across all short-time slices; frequency fluctuation matrices are obtained based on the frequency fluctuations of each reference range gate in each short-time slice; a hypothetical platform frequency shift grid is constructed based on the frequency fluctuation matrix; and spatial frequency standard deviation sequences are determined based on the additive motion characteristics of the platform and the spatial non-uniform motion characteristics of turbulence, according to the frequency distribution of all reference range gates in each short-time slice. Based on the hypothetical platform frequency shift grid, a cost function for the dynamic programming algorithm is defined, which includes a frequency shift mutation penalty. Based on the spatial frequency standard deviation sequence, the trajectory smoothing weight coefficient in the dynamic programming algorithm is determined. Using the trajectory smoothing weight coefficient, dynamic programming recursion and backtracking are performed in the hypothetical platform frequency shift grid according to the frequency shift mutation penalty to search for the globally optimal platform common mode frequency shift estimation sequence. The measured spectrum data is extracted from the power spectral density matrix, and corrected using the platform common-mode frequency shift estimation sequence to output high-precision wind speed data free from platform motion interference.

[0005] Furthermore, the power spectral density matrix is ​​obtained in the following ways: For the data within each short-time slice, windowing is used to suppress spectral sidelobe leakage; Perform a fast Fourier transform on the windowed data, calculate the square of its modulus, and obtain the Doppler power spectral density; all Doppler power spectral densities constitute the power spectral density matrix.

[0006] Furthermore, the method for obtaining the reference distance gate includes: Calculate the average carrier-to-noise ratio for each initial distance gate across all short-time slices; Sort all initial range gates in descending order of average carrier-to-noise ratio (CNR) values, and select the top-ranked initial range gates to form a reference range gate set.

[0007] Furthermore, the frequency fluctuation matrix is ​​obtained in the following ways: For each reference range gate in the set of reference range gates, the Doppler center frequency is extracted from the spectral data of each short-time slice to obtain the instantaneous frequency value; The frequency mean of the reference range gate over the entire observation period is calculated. The instantaneous frequency value is subtracted from the frequency mean to obtain the frequency fluctuation value after removing the DC component. The frequency fluctuation values ​​of all short-time slices and the reference range gate are combined to construct the frequency fluctuation matrix.

[0008] Furthermore, the method for obtaining the hypothetical platform frequency shift grid includes: Windowing is applied to the data within each short-time slice, and a fast Fourier transform is performed on the windowed data to determine the frequency resolution. Based on the platform's physical motion limits calibrated by the inertial navigation unit, the maximum possible frequency shift boundary is determined; The frequency resolution is set to the step size, and multiple discrete states are divided within the maximum possible frequency shift boundary interval. A hypothetical platform frequency shift grid is constructed, which contains the hypothetical platform common-mode frequency shift of multiple discrete states.

[0009] Furthermore, the method for obtaining the spatial frequency standard deviation sequence includes: Calculate the normalized signal-to-noise ratio weight for each reference range gate in the reference range gate set; Based on normalized weights, the frequency-weighted mean of all reference distance gates for each short-time slice is calculated. Then, the weighted spatial standard deviation of each short-time slice is calculated. The weighted spatial standard deviations of all short-time slices form a spatial frequency standard deviation sequence.

[0010] Furthermore, the method for obtaining the trajectory smoothing weight coefficient includes: Calculate the global standard deviation of all elements in the frequency fluctuation matrix and set it as the reference turbulence baseline value; The trajectory smoothing weight coefficient is obtained by calculating the fraction using the ratio of the elements of the spatial frequency standard deviation sequence to the reference turbulence baseline value and the preset minimum positive sum value as the denominator, and the preset basic inertial constant as the numerator.

[0011] Furthermore, the method for obtaining the platform common-mode frequency shift estimation sequence includes: Based on the frequency shift abruptness penalty and the assumed platform frequency shift grid, the Viterbi algorithm is used for recursion and backtracking to obtain the globally optimal platform common mode frequency shift estimation sequence.

[0012] Furthermore, the method for obtaining the frequency shift mutation penalty includes: The frequency shift abrupt penalty is obtained by calculating the trajectory smoothing weight coefficient and the square of the hypothetical platform common-mode frequency shift difference between adjacent short-time slices.

[0013] Furthermore, the method for obtaining the high-precision wind speed data after removing platform motion interference includes: Extract the target spectrum data from the power spectral density matrix; perform frame-by-frame correction using the target spectrum data from the platform common-mode frequency shift estimation sequence; obtain the motion-compensated cumulative spectrum based on the corrected spectrum data, and perform peak search based on the cumulative spectrum to determine the frequency index corresponding to the maximum capability peak; and calculate the final wind speed data using the Doppler calculation formula.

[0014] The present invention has the following beneficial effects: This invention leverages the spatial physical redundancy of multi-range gate observations by lidar, utilizing the physical characteristics of platform rigid body motion as the global common modulus and atmospheric turbulence as the spatial differential modulus. This solves the technical challenge of separating the real wind field from platform noise under frequency domain aliasing conditions. By calculating the spatial frequency standard deviation sequence, this invention utilizes the redundancy of spatial information to eliminate the ambiguity of purely temporal dimension analysis. This technique enables the motion inversion algorithm to adaptively adjust the trajectory smoothing weight coefficient according to the intensity of environmental turbulence. During periods of strong turbulence, it automatically reduces inertial constraints to preserve high-frequency wind speed characteristics, while during periods of weak turbulence, it strengthens constraints to smooth observation noise. This achieves accurate separation of the platform's common-mode motion trajectory under complex meteorological conditions, effectively suppressing detection signal noise and improving wind speed inversion accuracy. Attached Figure Description

[0015] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a flowchart of a method for suppressing noise in the detection signal of a laser wind measuring radar, provided as an embodiment of the present invention. Detailed Implementation

[0017] To further illustrate the technical means and effects adopted by the present invention to achieve the intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a laser wind-measuring radar detection signal noise suppression method proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0018] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0019] The following describes in detail, with reference to the accompanying drawings, a specific scheme for noise suppression of the detection signal of a laser wind-measuring radar provided by the present invention.

[0020] Example 1: This invention proposes a method for suppressing noise in the detection signal of a laser wind-measuring radar. Please refer to [link / reference]. Figure 1 The diagram illustrates a flowchart of a method for suppressing noise in a laser wind-measuring radar according to an embodiment of the present invention. The method includes: Step S101: Obtain the raw echo signal for the entire radar observation period, and perform short-time slicing and spectral transformation on the raw echo signal to obtain the power spectral density matrix.

[0021] The application scenario of this invention can be specifically a scenario of laser wind measurement on dynamic platforms such as ships, buoys or airborne aircraft. The original signal can be collected by the laser wind radar, and the original signal can be preprocessed. After signal processing, the original echo signal is obtained and input into the memory for retrieval.

[0022] The original echo signal is a continuous long-term observation signal. In order to capture the high-frequency transient motion characteristics of the platform, it is necessary to decompose the continuous long-term observation signal into a series of short-term observation units.

[0023] Acquire the raw echo signal of the radar within a single beam dwell period. Set the time length and sliding step size of the short-time slices. Divide the raw echo signal into multiple overlapping short-time slices.

[0024] In one implementation of this invention, the time length T of the short-time slice is typically 1 / 10 to 1 / 20 of the platform motion cycle, for example, 50ms, and the sliding step size t is typically set to 50% of T to maintain data continuity.

[0025] Perform a Fast Fourier Transform on the time-domain data of each short-time slice, calculate the square of its modulus, and obtain the Doppler power spectral density. All Doppler power spectral densities constitute the power spectral density matrix.

[0026] Step S102: Select reference range gates based on the carrier-to-noise ratio of each initial range gate in all short-time slices; obtain the frequency fluctuation matrix based on the frequency fluctuation of each reference range gate in each short-time slice; construct the hypothetical platform frequency shift grid based on the frequency fluctuation matrix; determine the spatial frequency standard deviation sequence according to the frequency distribution of all reference range gates in each short-time slice.

[0027] The initial range gate refers to all spatial sampling points divided by the radar along the entire detection path, including both near-field strong signal regions and far-field weak signal regions. The initial range gate is filtered to obtain the reference range gate.

[0028] Among them, the reference range gate is a subset of high carrier-to-noise ratio selected from all range gates, usually located in the near-field region. These range gates, due to their strong echo signals and low susceptibility to noise, can reflect frequency fluctuations caused by platform motion and atmospheric turbulence with high fidelity. They are used as the benchmark observation source for inverting the common-mode motion trajectory of the platform to decouple and compensate for motion errors in weak far-field signals.

[0029] Due to the attenuation characteristics of laser light during atmospheric transmission, the echo signal-to-noise ratio (SNR) varies significantly among different range gates. Low SNR signals are heavily influenced by shot noise and cannot accurately reflect the fine motion of the platform, while strong echo signals and weak signals are modulated synchronously by the rigid body motion of the platform. Therefore, it is necessary to select high SNR signals as the reference for inverting platform motion. The Doppler power spectral density matrix is ​​traversed to calculate the average carrier-to-noise ratio (CNR) of each range gate across all short-time slices. All range gates are then sorted from largest to smallest based on their average CNR values. The top-ranked range gates are selected to form a reference range gate set.

[0030] The frequency fluctuation matrix is ​​a spatiotemporal two-dimensional data matrix used to characterize the instantaneous AC frequency variation characteristics in the radar's near-field high signal-to-noise ratio region after removing the DC component of the average wind speed. Each row of the matrix represents a short-time slice, and each column represents a preferred reference range gate.

[0031] The absolute Doppler frequency measured by radar includes a DC component generated by the average wind speed and an AC component generated by turbulence and platform motion. The platform motion manifests only as time-varying AC fluctuations. To separate the motion characteristics, it is necessary to remove the DC frequency offset of each range gate.

[0032] For each reference range gate in the reference range gate set, the Doppler center frequency is extracted from the spectral data of each short-time slice to obtain the instantaneous frequency value. The frequency mean of this reference range gate over the entire observation period is calculated: the instantaneous frequency value is subtracted from the frequency mean to obtain the frequency fluctuation value after removing the DC component. The frequency fluctuation values ​​of all short-time slices and reference range gates are combined to construct a frequency fluctuation matrix. This matrix contains only AC variation information reflecting platform motion and atmospheric turbulence, and serves as the input data for subsequent motion inversion algorithms.

[0033] In this scheme, it is assumed that the platform frequency shift grid is a preset state space constructed for discretized optimal path search.

[0034] Since the Doppler frequency shift caused by platform motion is physically a continuously varying analog quantity, directly performing nonlinear trajectory optimization on observation data containing random noise and turbulence disturbances in the continuous domain is a mathematical problem with extremely high computational complexity, or even unsolvable. Therefore, the continuous frequency shift space is first discretized into a hypothetical platform frequency shift grid based on the platform's maximum dynamic range. Through this operation, the complex trajectory inversion is transformed into a computable discrete state search problem, and a necessary normalized scale is provided for the adaptive inertial constraint mechanism, thus ensuring the feasibility and robustness of the motion inversion algorithm.

[0035] Among them, the spatial frequency standard deviation sequence is a time series constructed to quantify the spatial non-uniformity of the atmospheric wind field at each moment.

[0036] Since the flight time of the laser pulse is much shorter than the variation period of the platform motion and turbulence, it can be assumed that within the same short-time slice, the platform motion exerts an equal Doppler frequency shift (i.e., common-mode interference) on all range gates. According to the principle of variance translation invariance, the dispersion of the full-field reference range gate frequency distribution is not affected by the common-mode frequency shift, but depends only on the spatial inhomogeneity of atmospheric turbulence. Therefore, calculating the spatial standard deviation can effectively isolate the atmospheric turbulence intensity from the platform's mixed motion.

[0037] In one implementation of this invention, the reference distance gate selection range is typically the top 5 to 10 distance gates ranked by average carrier-to-noise ratio, to ensure statistical robustness.

[0038] In one implementation of this invention, the Doppler center frequency extraction method commonly uses the spectral peak search method or the centroid method.

[0039] In one implementation of this invention, the maximum dynamic range of the platform can be determined by obtaining the maximum possible frequency shift boundary from the platform's physical motion limits calibrated by the inertial navigation unit.

[0040] Step S103: Based on the hypothetical platform frequency shift grid, define the cost function of the dynamic programming algorithm, which includes a frequency shift mutation penalty; based on the spatial frequency standard deviation sequence, determine the trajectory smoothing weight coefficient in the dynamic programming algorithm; using the trajectory smoothing weight coefficient, perform dynamic programming recursion and backtracking in the hypothetical platform frequency shift grid according to the frequency shift mutation penalty to search for the globally optimal platform common mode frequency shift estimation sequence.

[0041] Among them, the frequency shift mutation penalty is a cost function term used in the dynamic programming algorithm of this scheme to constrain the smoothness of the trajectory, which aims to reflect the physical inertia of the object's motion.

[0042] In order to recover the physical trajectory of a platform from noisy observation data, it is necessary to establish a cost evaluation system that includes constraints on motion continuity.

[0043] The motion continuity constraint cannot be fixed: under weak turbulence, the constraint should be strengthened to suppress noise, while under strong turbulence, the constraint should be relaxed to allow the algorithm to track actual wind speed changes. Furthermore, to address the problem of discontinuity in single-point estimation, a dynamic programming algorithm is used to find a globally optimal path with the minimum cumulative cost over the entire time axis. This path is the platform motion trajectory obtained through inversion.

[0044] Among them, the trajectory smoothing weight coefficient is the core control parameter used in this scheme to dynamically adjust the inertial constraint strength of the motion inversion algorithm.

[0045] The purpose of calculating the trajectory smoothing weight coefficient is to dynamically adjust the constraint of the motion inversion algorithm on frequency abrupt changes based on the real-time perceived environmental turbulence intensity, in order to balance noise smoothing and high-frequency feature tracking. This endows the algorithm with environmental adaptability, solves the problem of difficulty in separating real wind shear and platform noise under frequency domain aliasing, and ensures accurate inversion of the platform motion trajectory under complex conditions such as strong turbulence or violent shaking, thereby significantly improving compensation accuracy and the signal-to-noise ratio of weak far-field signals.

[0046] Among them, the platform common-mode frequency shift estimation sequence is the final motion inversion result output after spatiotemporal joint solution.

[0047] The purpose of computing the common-mode frequency shift estimation sequence of the platform is to accurately extract the additive Doppler frequency shift component caused solely by the rigid body motion of the platform from the mixed echoes affected by atmospheric turbulence. Its beneficial effect is to provide a high-precision motion compensation benchmark, enabling the system to eliminate the spectral broadening and drift caused by the dynamic platform through inverse frequency reconfiguration without the need for external sensors. This achieves coherent focusing of signal energy and significantly improves the radar's far-field detection range and wind speed inversion accuracy in dynamic environments.

[0048] Using the trajectory smoothing weight coefficient, dynamic programming recursion and backtracking are performed in the assumed platform frequency shift grid according to the frequency shift mutation penalty to search for the globally optimal platform common mode frequency shift estimation sequence.

[0049] Step S104: Extract the spectrum data to be measured from the power spectral density matrix, correct it using the platform common-mode frequency shift estimation sequence, and output high-precision wind speed data with platform motion interference removed.

[0050] The target spectrum data is extracted from the power spectral density matrix after time-frequency transformation to construct the target spectrum matrix to be corrected. This data is modulated by the non-stationary motion of the dynamic platform, resulting in diffuse and broadened peak energy in the frequency domain, leading to an extremely low signal-to-noise ratio.

[0051] Using the platform common-mode frequency shift estimation sequence as a compensation benchmark, the Doppler frequency shift caused by the platform motion at each moment is calculated, and the target spectrum matrix is ​​then subjected to frame-by-frame inverse frequency shift. This operation aims to construct a servo coordinate system in the frequency domain, correcting and aligning the signal energy that originally drifted over time to the equivalent stationary base frequency position.

[0052] Coherent accumulation of the calibrated spectral data is performed, and the signal phase alignment characteristics are used to achieve linear energy focusing and signal-to-noise ratio improvement. Based on the accumulated high signal-to-noise ratio spectrum, the center frequency of the spectral peak is searched to obtain line-of-sight wind speed data after removing platform motion interference components, thereby achieving high-precision far-field detection in dynamic environments.

[0053] After processing steps S101 to S104, high-precision wind speed data, free from platform motion interference, is obtained. During wind speed data calculation, the collected data is first sliced ​​and transformed into the time-frequency domain. Considering the significant differences in echo signal-to-noise ratio (SNR) between different range gates, a reference range gate is selected from the initial range gates using the carrier-to-noise ratio (CNR). Since the absolute Doppler frequency measured by radar includes a DC component generated by the average wind speed and an AC component generated by turbulence and platform motion, a frequency fluctuation matrix is ​​obtained based on the frequency fluctuation characteristics of the reference range gate in each short-time slice. Before executing the dynamic programming algorithm, because the dynamic programming algorithm cannot directly perform path optimization in a continuous infinite numerical space, and the subsequent adaptive weight calculation lacks a unified quantification standard, the continuous frequency shift space is discretized into a hypothetical platform frequency shift grid based on the platform's maximum dynamic range, thus limiting a reasonable path optimization space to ensure the rationality of the scheme. The dispersion of the full-field reference range gate frequency distribution is not affected by the common-mode frequency shift, but only depends on the spatial inhomogeneity of atmospheric turbulence. Therefore, calculating the spatial standard deviation can effectively isolate atmospheric turbulence intensity from the platform's mixed motion. To recover the physically consistent platform trajectory from noisy observation data, a cost evaluation system with motion continuity constraints was established. A dynamic programming algorithm was used to find the globally optimal path with the minimum cumulative cost over the entire time axis. Finally, the collected spectral data was corrected using the platform common-mode frequency shift estimation sequence to obtain high-precision wind speed data free from platform motion interference.

[0054] Preferably, in some possible implementations of the embodiments of the present invention, the power spectral density matrix is ​​obtained in the following ways: Because finite-length time-domain truncation of non-integer period signals causes discontinuous abrupt changes at the boundaries of FFT periodic extension, spectral sidelobe leakage occurs. This phenomenon significantly increases the sidelobe level and widens the main lobe, causing strong signal sidelobes to mask adjacent weak signals and severely reducing the system's dynamic range. Therefore, a time-domain windowing method is needed to smooth and attenuate signal edges to eliminate boundary abrupt changes, thereby effectively suppressing sidelobe leakage and improving the detection capability of weak signals.

[0055] For the data within each short-time slice, a Hanning window or a Hamming window is applied to suppress spectral sidelobe leakage. A Fast Fourier Transform is performed on the windowed data, and the square of its modulus is calculated to obtain the Doppler power spectral density. All Doppler power spectral densities constitute the power spectral density matrix.

[0056] Preferably, in some possible implementations of the embodiments of the present invention, the method for obtaining the reference distance gate includes: Due to the non-uniformity of atmospheric aerosol distribution and the geometric attenuation effect of laser transmission, the echo signals of lidar at different detection distances exhibit significantly different carrier-to-noise ratios. Echo signals in the far-field or low-aerosol regions are weak and severely contaminated by shot noise and speckle noise, resulting in significant random errors in the extracted instantaneous frequencies. These frequencies cannot accurately reflect the fine motion characteristics of the platform, while the echo signals from all range gates are physically modulated by the platform's rigid body motion in a common-mode synchronous manner.

[0057] Step S201: Calculate the average carrier-to-noise ratio for each initial distance gate across all short-time slices.

[0058] The average carrier-to-noise ratio (CNR) screening reference distance gate is used to select near-field data with minimal noise contamination and highest signal quality from the full-field echoes as the observation benchmark. Only high CNR signals can faithfully reflect the fine frequency fluctuations of the platform motion, minimizing the interference of measurement noise on turbulent frequency characteristics, thereby ensuring the physical authenticity and solution accuracy of subsequent motion inversion algorithms.

[0059] Step S202: Sort all initial range gates in descending order of average carrier-to-noise ratio, and select the top-ranked initial range gates to form a reference range gate set.

[0060] Sort the data by average carrier-to-noise ratio from largest to smallest to ensure that the reference source used as input to the algorithm is always the highest quality and most reliable data throughout the entire observation period.

[0061] The reference range gate is obtained by filtering the initial range gate and is usually located in the near-field region. These range gates, due to their strong echo signals and low susceptibility to noise, can reflect frequency fluctuations caused by platform motion and atmospheric turbulence with high fidelity. They are used as the benchmark observation source for inverting the common-mode motion trajectory of the platform to decouple and compensate for motion errors in weak far-field signals.

[0062] This embodiment sorts the data based on the average carrier-to-noise ratio over the entire observation period and selects several high-confidence reference range gates. By utilizing the high signal-to-noise ratio of near-field strong echo signals to replace the full-field data for motion feature extraction, the impact of observation noise on frequency estimation accuracy is significantly reduced. This ensures that the subsequently constructed frequency fluctuation matrix can faithfully preserve the true details of the platform's motion, thus laying a low-noise data foundation for high-precision motion trajectory inversion.

[0063] It should be noted that, in one specific embodiment of this example, the reference distance gate is obtained by traversing the power spectral density matrix and calculating each distance gate. In all The average carrier-to-noise ratio (CNR) within each short-time slice. All range gates are sorted from highest to lowest average CNR. The gates with the highest CNR values ​​are selected. A distance gate ( The values ​​typically range from 5 to 10 to ensure statistical robustness, forming the reference distance gate set. Elements in the set That is, the selected reference distance gate index, where To reference the index of the distance gate in the set, .

[0064] It should be noted that for long observation periods or when the platform attitude changes drastically, a segmented screening strategy can be adopted, such as re-evaluating the CNR every time time H elapses, to adapt to changes in the strong echo region. The value of H can be selected based on the actual degree of platform attitude change or the total length of the observation period. In this embodiment, the value is 5 minutes.

[0065] Preferably, in some possible implementations of the embodiments of the present invention, the frequency fluctuation matrix is ​​obtained by means of: The frequency fluctuation matrix is ​​a spatiotemporal two-dimensional data matrix used to characterize the instantaneous AC frequency variation characteristics in the radar's near-field high signal-to-noise ratio region after removing the DC component of the average wind speed. Each row of the matrix represents a short-time slice, and each column represents a preferred reference range gate.

[0066] The absolute Doppler frequencies measured by radar include a DC component generated by the average wind speed and an AC component generated by turbulence and platform motion, with the platform motion manifesting only as time-varying AC fluctuations. To separate the motion characteristics, the DC frequency bias of each range gate needs to be removed. This decouples the DC background wind field from the AC disturbance signal in the frequency domain, eliminates the bias introduced by wind shear, and allows the residual fluctuation quantities in the matrix to accurately characterize the time-varying features of the platform's rigid body motion and atmospheric turbulence. This provides precise zero-mean observation input for subsequent trajectory inversion based on spatial common-mode characteristics.

[0067] Specifically, for each reference range gate in the reference range gate set, the Doppler center frequency is extracted from the spectral data of each short-time slice to obtain the instantaneous frequency value. The frequency mean of this reference range gate over the entire observation period is calculated: the instantaneous frequency value is subtracted from the frequency mean to obtain the frequency fluctuation value after removing the DC component. The frequency fluctuation values ​​of all short-time slices and reference range gates are combined to construct a frequency fluctuation matrix. This matrix contains only AC variation information reflecting platform motion and atmospheric turbulence, and serves as the input data for subsequent motion inversion algorithms.

[0068] It should be noted that, in one specific embodiment of this example, the frequency fluctuation matrix is ​​obtained as follows: For each reference range gate in the reference range gate set In each short-time slice Spectrum data The Doppler center frequency is extracted. Common extraction methods include the spectral peak search method or the centroid method, to obtain the instantaneous frequency value. .

[0069] Calculate the reference distance gate over the entire observation period. frequency mean within .

[0070] Instantaneous frequency value Subtract the frequency mean This yields the frequency fluctuation value after removing the DC component. All A short slice, A frequency fluctuation matrix is ​​constructed by combining the frequency fluctuation values ​​of each reference distance gate. .

[0071] The frequency fluctuation matrix The dimension is Elements in the matrix Characterized the first The reference distance gate at the 1st Doppler frequency fluctuations within a short-time slice.

[0072] Preferably, in some possible implementations of the embodiments of the present invention, it is assumed that the method for obtaining the platform frequency shift grid includes: In this scheme, it is assumed that the platform frequency shift grid is a preset state space constructed for discretized optimal path search.

[0073] Since the Doppler frequency shift caused by platform motion is physically a continuously changing analog quantity, directly performing nonlinear trajectory optimization on observation data containing random noise and turbulence disturbances in the continuous domain is a mathematical problem with extremely high computational complexity, or even unsolvable. Furthermore, the dynamic programming approach used in this scheme is essentially a globally optimal path search algorithm based on discrete state space. Therefore, this embodiment, based on the maximum dynamic range and frequency resolution of the radar system, discretizes the continuous Doppler frequency shift interval into a finite number of state nodes, constructing a hypothetical platform frequency shift grid. This transforms the continuous function fitting problem into a shortest path search problem in graph theory. While ensuring the frequency quantization accuracy meets application requirements, it significantly reduces the algorithm's search dimensionality and computational overhead, making it possible to obtain the globally optimal platform motion trajectory with limited computational resources.

[0074] It should be noted that, in one specific implementation of this embodiment, the platform frequency shift grid is obtained as follows: the maximum possible frequency shift boundary is determined based on the platform's physical motion limits calibrated by the inertial navigation unit. Frequency resolution determined by FFT Let the step size be the interval. Inner division A discrete state is used to construct a hypothetical platform frequency shift grid. .

[0075] The physical motion limit of the platform refers to the maximum instantaneous line-of-sight velocity boundary that the platform can achieve under the constraints of its own dynamic characteristics and external environmental loads (such as sea state). This parameter is usually determined based on the platform's structural design specifications, seakeeping indicators, or the statistical distribution of historical measured data under typical operating conditions. For example, the heave velocity limit of an ocean buoy under severe sea conditions is usually set within ±5 m / s.

[0076] The maximum possible frequency shift boundary is based on the Doppler effect principle, mapping the aforementioned physical velocity limit to a search space constraint value in the frequency domain. This boundary defines the effective state space of the motion inversion algorithm, ensuring that the solution results conform to physical reality. For conventional 1.5μm band shipborne or buoy radar systems, the value of this frequency shift boundary is usually set between ±2MHz and ±10MHz, with the specific value depending on the platform type and the maximum expected sea state.

[0077] Preferably, in some possible implementations of the embodiments of the present invention, the method for obtaining the spatial frequency standard deviation sequence includes: Among them, the spatial frequency standard deviation sequence is a time series constructed to quantify the spatial non-uniformity of the atmospheric wind field at each moment.

[0078] Because the Doppler frequency shift in radar echoes simultaneously incorporates both platform rigid body motion and atmospheric turbulence, and these two phenomena are highly overlapping in their spectral characteristics over a single time dimension (both being high-frequency random fluctuations), traditional time-domain filtering or frequency-domain low-pass methods cannot remove platform motion interference while preserving the true high-frequency wind field characteristics. Based on physical facts, platform rigid body motion, as an additive common modulus, only causes an overall shift in the frequency distribution without changing its dispersion, while atmospheric turbulence, as a spatially non-uniform quantity, directly leads to an increase in the dispersion of the frequency distribution.

[0079] Therefore, this embodiment utilizes the observation redundancy of the spatial dimension to calculate the spatial frequency standard deviation sequence of the full-field reference range gate frequency distribution at each moment. This operation constructs a turbulence intensity quantification index independent of the time dimension, successfully achieving independent perception and decoupling of atmospheric turbulence under unknown platform motion conditions, providing a key environmental prior criterion for adaptive adjustment of inertial constraint weights in subsequent dynamic programming algorithms.

[0080] Calculate the normalized signal-to-noise ratio weight for each reference range gate in the reference range gate set. Based on normalized weights Calculate the frequency-weighted mean of all reference distance gates for each short-time slice, then calculate the frequency-weighted spatial standard deviation of each short-time slice. The frequency-weighted spatial standard deviations of all short-time slices form the frequency-weighted spatial standard deviation sequence, i.e., the elements in the frequency-weighted spatial standard deviation sequence. All the elements are combined to form the frequency-weighted spatial standard deviation sequence.

[0081] It should be noted that, in this embodiment of the invention, the maximum-minimum normalization process can be used to perform maximum-minimum normalization on the signal ratio weights of all reference distance gates, so that their value range is between [0,1].

[0082] It should be noted that, in one specific embodiment of this example, the spatial frequency standard deviation sequence is obtained as follows: For frequency fluctuation matrix The Row data (of which) Based on normalized weights Calculate the weighted average of the reference distance gate frequency at that moment. .

[0083] Subsequently, the weighted spatial standard deviation at that moment is calculated, which is the element in the spatial frequency standard deviation sequence. .

[0084] For all Repeat the above calculation for each short-time slice to generate a complete spatial frequency standard deviation sequence. The larger the mean of the sequence, the worse the spatial consistency of the wind field at the current moment (the stronger the turbulence); the smaller the mean, the more stable the wind field.

[0085] Preferably, in some possible implementations of the embodiments of the present invention, the trajectory smoothing weight coefficient is obtained by means of: Among them, the trajectory smoothing weight coefficient is the core control parameter used in this scheme to dynamically adjust the inertial constraint strength of the motion inversion algorithm.

[0086] Dynamic radar platforms exhibit non-stationary motion characteristics under varying sea states and weather conditions: in stable sea states, the platform moves slowly, and observation noise dominates, requiring enhanced smoothing constraints; however, in severe sea states (strong turbulence), the platform often experiences high-frequency, violent shaking, and forced smoothing can lead to hysteresis errors in the algorithm and the loss of high-frequency motion details. Traditional fixed-parameter filters cannot simultaneously meet the needs of these two opposing scenarios.

[0087] Therefore, an adaptive mapping mechanism based on the standard deviation of spatial frequency is established to calculate the trajectory smoothing weight coefficients in the dynamic programming algorithm in real time. This enables dynamic adjustment of the inertial constraint term in the state transition cost function: automatically reducing the weight to relax the constraint during strong turbulence, allowing the algorithm to track rapidly changing platform trajectories; and increasing the weight to suppress observation noise during weak turbulence. This achieves an optimal trade-off between denoising capability and dynamic tracking capability in non-stationary environments.

[0088] It should be noted that, in one specific implementation of this embodiment, the trajectory smoothing weight coefficient is obtained by first calculating the reference turbulence baseline value: The reference turbulence baseline value is a statistical normalization factor used to standardize the turbulence intensity metric.

[0089] Calculate the global standard deviation of all elements in the frequency fluctuation matrix and set it as the reference turbulence baseline value. .

[0090] Further utilize spatial frequency standard deviation sequences and reference turbulence baseline value Perform the calculation: in, The basic inertial constant (its value typically ranges from 10 to 100, depending on the sampling rate of the radar system and the mechanical inertial characteristics of the platform). To adjust the sensitivity index (usually taken as...) ), It is a very small positive number to prevent division by zero errors (usually taken as...). This formula makes local turbulence... Significantly higher than the benchmark value At that time, weight It automatically decreases, thereby reducing the penalty for frequency mutations.

[0091] Preferably, in some possible implementations of the embodiments of the present invention, the method for obtaining the platform common-mode frequency shift estimation sequence includes: Among them, the platform common-mode frequency shift estimation sequence is the final motion inversion result output after spatiotemporal joint solution.

[0092] Because the non-stationary motion of the dynamic platform introduces time-varying Doppler frequency shift noise into the radar echo signal, the signal spectrum after long-term accumulation undergoes energy broadening and peak blurring, severely reducing the signal-to-noise ratio and detection range of weak far-field signals. Furthermore, the platform motion modulation experienced by different range gates is common-mode, requiring a precise time-series trajectory describing this common-mode motion for unified correction of the entire field data. Therefore, a dynamic programming algorithm is used to search for the globally optimal platform common-mode frequency shift estimation sequence in the discrete state space. This sequence serves as a high-precision parameterized representation of the platform's instantaneous radial velocity, not only removing environmental turbulence interference from the mixed signal but also providing a precise inverse compensation benchmark for subsequent processing. This makes it possible to construct a following coordinate system in the frequency domain and eliminate Doppler non-stationarity caused by platform motion.

[0093] In one specific implementation of this invention, the platform common mode estimation sequence is obtained as follows: First calculate the frequency shift fitting residual. : Among them, the frequency shift fitting residual is a measure of the performance at the 1st epoch. At any given moment, assuming the platform's common-mode frequency shift is... The degree of deviation from the actual observed data. The calculation method is as follows: .

[0094] The frequency shift fitting residual and the frequency shift abrupt change penalty are input into the Viterbi algorithm, and forward recursive calculation is performed. The specific steps are as follows: 1. Establish the cumulative cost matrix (dimension) ) and path backtracking pointer matrix (dimension) ).

[0095] 2. Initialize the cumulative cost matrix: For each state ( The cumulative cost depends only on the fitting residuals: .

[0096] 3. Recursive calculation: For each state at the current time step Iterate through all possible states from the previous time step. ( ), find the predecessor state that minimizes the cumulative cost plus the transition cost of the previous time step.

[0097] Calculate the minimum path cost: .

[0098] Record the optimal predecessor state index: .

[0099] After completing the forward recursive calculation, a global path backtracking is performed, the specific process of which is as follows: When the recursion reaches the last short-time observation window At that time, a backtracking operation is performed to extract the optimal path.

[0100] At any moment Find the terminating state index with minimum cumulative cost : .

[0101] Using the path backtracking pointer matrix ,from Reverse derivation to The optimal state index at each time step is obtained sequentially. : .

[0102] Finally, the optimal state index sequence Mapping back to the hypothetical platform frequency shift grid The corresponding frequency value in the index, i.e., for each time point m, is determined using the index value. exist Numerical values ​​are acquired during the process to obtain corresponding frequency data, which then form the final platform common-mode frequency shift estimation sequence. .

[0103] Preferably, in some possible implementations of the embodiments of the present invention, the method for obtaining the frequency shift mutation penalty includes: Among them, the frequency shift mutation penalty is a cost function term used in the dynamic programming algorithm of this scheme to constrain the smoothness of the trajectory, which aims to reflect the physical inertia of the object's motion.

[0104] In order to recover the physical trajectory of a platform from noisy observation data, it is necessary to establish a cost evaluation system that includes constraints on motion continuity.

[0105] The motion continuity constraint cannot be fixed: under weak turbulence, the constraint should be strengthened to suppress noise, while under strong turbulence, the constraint should be relaxed to allow the algorithm to track actual wind speed changes. Furthermore, to address the problem of discontinuity in single-point estimation, a dynamic programming algorithm is used to find a globally optimal path with the minimum cumulative cost over the entire time axis. This path is the platform motion trajectory obtained through inversion.

[0106] In one specific implementation of this invention, a frequency shift abruptness penalty is applied. The calculation method can be as follows: in, For trajectory smoothing weighting coefficients, These represent the hypothetical platform common-mode frequency shift states at the current and previous times, respectively.

[0107] Preferably, in some possible implementations of the embodiments of the present invention, the method of extracting the spectrum data to be measured from the power spectral density matrix, correcting it using the platform common-mode frequency shift estimation sequence, and outputting high-precision wind speed data free from platform motion interference is as follows: Extract target data from the power spectral density matrix. This can be done based on a user-specified detection range or any range gate index across the entire field, excluding the reference range gate. Extract the distance to the door in all Spectral data within a short time slice are used to construct the target spectral matrix to be corrected. .

[0108] The dimension of this matrix is , its first row vector Representing the The instantaneous power spectrum within a short time slice. At this time, due to the platform motion, the spectral peaks of the same wind speed target have drastic frequency drifts between different time slices, and direct accumulation will lead to the dispersion of spectral peak energy.

[0109] Estimating the sequence using the common-mode frequency shift of the platform The target spectrum is corrected frame by frame. Because This characterizes the additive Doppler frequency shift noise introduced by the platform motion, therefore the correction operation is the inverse process of subtracting this frequency shift.

[0110] For each short-time slice ( Read the corresponding platform common-mode frequency shift value. Frequency resolution utilized Calculate the discrete frequency index offset corresponding to this moment. : in This indicates the rounding operation.

[0111] Next, regarding the first spectral vector of the row Perform a cyclic shift operation, treating spectral energy exceeding the boundary as noise truncation or using zero-padding for shifting. Index the spectrum as... Data moved to new location The calculation formula is: For out of index range The part uses cyclic boundary handling (i.e., modulo). (operation).

[0112] For all The above operation is repeated for each row of data to generate a corrected spectral matrix. Through this step, the signal energy that originally drifted over time is realigned to a constant frequency index corresponding to its true line-of-sight wind speed.

[0113] The frequency-corrected spectral matrix along the time dimension Perform an arithmetic mean to generate the motion-compensated cumulative spectrum. : .

[0114] During this accumulation process, the amplitude of the signal spectrum peak increases linearly with the number of accumulations, while the amplitude of background noise increases at a lower rate due to its random phase characteristics. Therefore, the accumulated spectrum after motion compensation... It exhibits a significant improvement in signal-to-noise ratio.

[0115] Finally, for Perform a peak search to determine the frequency index corresponding to the maximum energy peak. The final line-of-sight wind speed was calculated using the Doppler formula. : in The operating wavelength of the lidar (typically 1000 nm) This wind speed is the high-precision wind measurement result after removing interference from platform motion.

[0116] In summary, this invention, based on the spatial physical redundancy of multi-range gate observations by lidar, utilizes the physical characteristics of platform rigid body motion as the global common modulus and atmospheric turbulence as the spatial difference modulus to construct a spatial frequency standard deviation sequence as an independent measure of environmental turbulence intensity. Furthermore, an adaptive dynamic programming inversion operation is implemented, adjusting the trajectory smoothing weight coefficient in real time according to the turbulence intensity to calculate a platform common-mode frequency shift estimation sequence that balances physical inertia and high-frequency dynamic characteristics in a discrete frequency shift grid. Finally, this sequence is used to perform inverse frequency domain reconfiguration and coherent accumulation of far-field echoes. Through these technical means, this scheme effectively solves the decoupling problem between the real wind field and platform noise under time-frequency aliasing conditions, achieving the elimination of Doppler spectral broadening caused by platform motion without relying on external attitude sensors, and significantly improving the signal-to-noise ratio and detection accuracy of weak far-field signals in dynamic environments.

[0117] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0118] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for suppressing noise in the detection signal of a laser wind-measuring radar, characterized in that, The method includes: The raw echo signal of the radar over the entire observation period is obtained, and the power spectral density matrix is ​​obtained by short-time slicing and spectral transformation of the raw echo signal. Reference range gates are selected based on the carrier-to-noise ratio of each initial range gate across all short-time slices; frequency fluctuation matrices are obtained based on the frequency fluctuations of each reference range gate in each short-time slice; a hypothetical platform frequency shift grid is constructed based on the frequency fluctuation matrix; and spatial frequency standard deviation sequences are determined based on the additive motion characteristics of the platform and the spatial non-uniform motion characteristics of turbulence, according to the frequency distribution of all reference range gates in each short-time slice. Based on the hypothetical platform frequency shift grid, a cost function for the dynamic programming algorithm is defined, which includes a frequency shift mutation penalty. Based on the spatial frequency standard deviation sequence, the trajectory smoothing weight coefficient in the dynamic programming algorithm is determined. Using the trajectory smoothing weight coefficient, dynamic programming recursion and backtracking are performed in the hypothetical platform frequency shift grid according to the frequency shift mutation penalty to search for the globally optimal platform common mode frequency shift estimation sequence. The measured spectrum data is extracted from the power spectral density matrix, and corrected using the platform common-mode frequency shift estimation sequence to output high-precision wind speed data free from platform motion interference.

2. The method for suppressing noise in the detection signal of a laser wind-measuring radar according to claim 1, characterized in that, The power spectral density matrix is ​​obtained in the following ways: For the data within each short-time slice, windowing is used to suppress spectral sidelobe leakage; Perform a fast Fourier transform on the windowed data, calculate the square of its modulus, and obtain the Doppler power spectral density; all Doppler power spectral densities constitute the power spectral density matrix.

3. The method for suppressing noise in the detection signal of a laser wind-measuring radar according to claim 1, characterized in that, The methods for obtaining the reference distance gate include: Calculate the average carrier-to-noise ratio for each initial distance gate across all short-time slices; Sort all initial range gates in descending order of average carrier-to-noise ratio (CNR) values, and select the top-ranked initial range gates to form a reference range gate set.

4. The method for suppressing noise in the detection signal of a laser wind-measuring radar according to claim 1, characterized in that, The frequency fluctuation matrix is ​​obtained in the following ways: For each reference range gate in the set of reference range gates, the Doppler center frequency is extracted from the spectral data of each short-time slice to obtain the instantaneous frequency value; The frequency mean of the reference range gate over the entire observation period is calculated. The instantaneous frequency value is subtracted from the frequency mean to obtain the frequency fluctuation value after removing the DC component. The frequency fluctuation values ​​of all short-time slices and the reference range gate are combined to construct the frequency fluctuation matrix.

5. The method for suppressing noise in the detection signal of a laser wind-measuring radar according to claim 1, characterized in that, The methods for obtaining the hypothetical platform frequency shift grid include: Windowing is applied to the data within each short-time slice, and a fast Fourier transform is performed on the windowed data to determine the frequency resolution. Based on the platform's physical motion limits calibrated by the inertial navigation unit, the maximum possible frequency shift boundary is determined; The frequency resolution is set to the step size, and multiple discrete states are divided within the maximum possible frequency shift boundary interval. A hypothetical platform frequency shift grid is constructed, which contains the hypothetical platform common-mode frequency shift of multiple discrete states.

6. The method for suppressing noise in the detection signal of a laser wind-measuring radar according to claim 1, characterized in that, The spatial frequency standard deviation sequence is obtained through the following methods: Calculate the normalized signal-to-noise ratio weight for each reference range gate in the reference range gate set; Based on normalized weights, the frequency-weighted mean of all reference distance gates for each short-time slice is calculated. Then, the weighted spatial standard deviation of each short-time slice is calculated. The weighted spatial standard deviations of all short-time slices form a spatial frequency standard deviation sequence.

7. The method for suppressing noise in the detection signal of a laser wind-measuring radar according to claim 1, characterized in that, The methods for obtaining the trajectory smoothing weight coefficients include: Calculate the global standard deviation of all elements in the frequency fluctuation matrix and set it as the reference turbulence baseline value; The trajectory smoothing weight coefficient is obtained by calculating the fraction using the ratio of the elements of the spatial frequency standard deviation sequence to the reference turbulence baseline value and the preset minimum positive sum value as the denominator, and the preset basic inertial constant as the numerator.

8. The method for suppressing noise in the detection signal of a laser wind-measuring radar according to claim 1, characterized in that, The methods for obtaining the platform common-mode frequency shift estimation sequence include: Based on the frequency shift abruptness penalty and the assumed platform frequency shift grid, the Viterbi algorithm is used for recursion and backtracking to obtain the globally optimal platform common mode frequency shift estimation sequence.

9. The method for suppressing noise in the detection signal of a laser wind-measuring radar according to claim 1, characterized in that, The frequency shift mutation penalty is obtained in the following ways: The frequency shift abrupt penalty is obtained by multiplying the trajectory smoothing weight coefficient by the square of the hypothetical platform common-mode frequency shift difference between adjacent short-time slices.

10. The method for suppressing noise in the detection signal of a laser wind-measuring radar according to claim 1, characterized in that, The methods for obtaining high-precision wind speed data after removing platform motion interference include: Extract the target spectrum data from the power spectral density matrix; perform frame-by-frame correction using the target spectrum data from the platform common-mode frequency shift estimation sequence; obtain the motion-compensated cumulative spectrum based on the corrected spectrum data, and perform peak search based on the cumulative spectrum to determine the frequency index corresponding to the maximum capability peak; and calculate the final wind speed data using the Doppler calculation formula.

Images