Speed defuzzification method based on multiband radar Doppler power spectrum joint constraint

Abstract

The invention provides a velocity deblurring method based on multi-band radar Doppler power spectrum joint constraint, which comprises the following steps: acquiring at least two different-band radar Doppler power spectrum data, the maximum non-blurring velocity of a second-band radar being smaller than that of a first-band radar; eliminating invalid distance library data through noise bottom estimation and effective spectrum threshold judgment; deblurring the first wave band radar to obtain a reference speed profile; establishing a cross-band data corresponding relation through space-time matching; determining the fuzzy type of the second band radar library by library; performing spectral domain reconstruction on the power spectrum according to partial folding or complete folding, constructing a candidate spread spectrum, and selecting an optimal folding frequency and a deblurring speed by utilizing cross-band consistency constraint and vertical continuity joint judgment; and outputting the deblurring speed profile and the quality identifier of each wave band. According to the invention, the technical problems of poor speed deblurring accuracy and low robustness in a complex meteorological scene and misjudgment caused by the fact that a single-band method depends on single continuity constraint or average speed are solved.

Description

Technical Field

[0001] This invention relates to the technical field of meteorological radar data quality control and signal processing, and in particular to a velocity deambiguation method based on joint constraints of multi-band radar Doppler power spectrum. Background Technology

[0002] Doppler weather radar obtains radial velocity information in meteorological detection by acquiring the Doppler frequency shift of the backscattered signals from clouds and precipitation particles. However, the maximum detection range and the maximum unambiguous velocity (Nyquist velocity) of the radar are limited by the pulse repetition frequency (PRF) and wavelength. With maximum detection range There is a contradictory relationship: ; When the actual radial velocity exceeds When the range is exceeded, the observations will collapse back into that range, resulting in velocity ambiguity, which severely limits the application of radial velocity data in areas such as strong convection monitoring, wind field inversion, and data assimilation.

[0003] Existing velocity deblurring techniques have the following drawbacks: First, they rely heavily on the average Doppler velocity of single-band radar or single-dimensional continuity constraints. When the particle spectrum is wide and multi-peaked, the average velocity is dominated by strongly scattering particles, resulting in a weighted bias that leads to incorrect folding determination and reduces the representativeness of the true velocity field. Second, in scenarios with discontinuous velocity fields, such as weak echoes, strong wind shear, and isolated echo areas, they are prone to missed or misjudged cases due to a lack of reliable references, and their robustness to complex spectral shapes and weak echo scenarios is insufficient. Third, they do not achieve fuzziness correction at the spectral domain level, but only iterative adjustments at the velocity mean level, which cannot effectively solve the problem of accurate identification of multiple complete folds and partial folds.

[0004] Therefore, there is an urgent need for a velocity defuzzification method based on joint constraints of multi-band radar Doppler power spectrum that can solve the above-mentioned technical problems. Summary of the Invention

[0005] The purpose of this invention is to provide a velocity deambiguation method based on joint constraints of multi-band radar Doppler power spectrum, which solves the problems of insufficient accuracy and robustness of existing technologies in complex meteorological scenarios such as weak echo, multi-peak spectrum, and strong turbulence, as well as the technical problems of misjudgment and deviation caused by traditional single-band methods relying on a single constraint or average velocity.

[0006] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows: This invention provides a velocity deambiguation method based on joint constraints of Doppler power spectrum from multi-band radars, comprising the following steps: S1. Acquiring Doppler power spectrum data and corresponding radar parameters from at least two radars of different bands, wherein the different band radars include at least a first-band radar and a second-band radar, and the maximum unambiguous velocity of the second-band radar is less than the maximum unambiguous velocity of the first-band radar; S2. Estimating the noise floor and determining the effective spectrum threshold for the Doppler power spectrum of each band and each range database, and removing invalid range database data; S3. Performing deambiguation processing on the first-band radar to obtain a reference velocity profile and the corresponding reference expanded spectrum. S4. Perform spatiotemporal matching of the Doppler power spectrum data of the first-band radar and the second-band radar to establish a cross-band data correspondence; S5. Determine the velocity ambiguity type for the range-by-range database of the second-band radar, where the ambiguity type includes partial folding and complete folding; S6. Based on the velocity ambiguity type, reconstruct the spectral domain of the Doppler power spectrum of the second-band radar, construct candidate expanded spectra, and use cross-band consistency constraints and vertical continuity joint decision-making to select the optimal number of folds and deambiguity rate to obtain the deambiguity velocity profile of the second-band radar; S7. Output the deambiguity velocity profiles of each band and the corresponding quality labels.

[0007] Furthermore, the deambiguity processing of the first-band radar includes the following steps: S3-1. Using iterative processing of the range library from cloud top to cloud bottom, detect whether there is partial folding in the Doppler power spectrum of the current range library; S3-2. When it is determined to be partially folded, construct positive candidate expansion spectrum and negative candidate expansion spectrum; S3-3. Calculate the average Doppler velocity of the positive candidate expansion spectrum and the negative candidate expansion spectrum respectively to obtain two candidate velocities; S3-4. Using the deambiguity velocity of the previous effective range library as a reference, select the candidate velocity with the smallest difference from the reference velocity as the optimal deambiguity velocity of the current library, and record the corresponding expansion spectrum as the reference expansion spectrum.

[0008] Furthermore, the criterion for determining the partial folding is: the left endpoint of the effective spectrum. And the right endpoint of the effective spectrum Or, the energy within both boundary windows is simultaneously higher than the noise floor, where The maximum unambiguous velocity for the first-band radar. This is the preset speed tolerance.

[0009] Furthermore, constructing a positive candidate expansion spectrum means shifting the spectral components of the negative velocity region to the positive velocity side, and constructing a negative candidate expansion spectrum means shifting the spectral components of the positive velocity region to the negative velocity side.

[0010] Furthermore, the spatiotemporal matching of the Doppler power spectrum data of the first-band radar and the second-band radar includes time alignment and range resampling: using the time series of the first-band radar as a reference, nearest neighbor matching or linear interpolation matching is performed on the second-band radar, requiring time error... ,in, To maximize the allowable time difference, the power spectrum data of the second-band radar is resampled onto a range grid that is consistent with that of the first-band radar.

[0011] Furthermore, the determination of velocity ambiguity type for the second-band radar range-by-range database includes the following steps: S5-1. Detecting whether it is partially folded according to preset determination conditions, wherein the determination condition is the left endpoint of the effective spectrum. And the right endpoint of the effective spectrum Or, the energy within both boundary windows is simultaneously higher than the noise floor, where The maximum unambiguous velocity for the second-band radar. The preset speed tolerance is used; S5-2. If it is not determined to be a partial fold, calculate the difference between the initial speed of the current library and the defuzzification speed of the previous valid library. ,like If this is the case, then a complete folding candidate judgment is triggered, where, The preset speed mutation threshold is used; S5-3. When the fully folded candidate judgment is triggered, the reference speed of the first-band radar is used. Constructing a candidate set of folding times for second-band radar ,in, The reference velocity for the first-band radar, when If invalid, use the default candidate set.

[0012] Furthermore, the spectral domain reconstruction of the second-band radar includes the following steps: S6-1. When it is determined to be partially folded, construct positive candidate expansion spectrum and negative candidate expansion spectrum, and prioritize the selection based on the sign of the first-band radar reference velocity to calculate the partially folded correction velocity; S6-2. When it is determined to be completely folded, calculate the number of folds for each candidate fold. Based on the sign direction of the radar reference velocity in the first band, the original power spectrum is processed. The velocity axis offset is used to obtain the candidate unfolding spectrum corresponding to the number of candidate folds.

[0013] Furthermore, the joint decision is made using a joint cost function. Select the optimal number of folds The calculation formula is: ,in, For the first The candidate velocity corresponding to the number of candidate folds The deblurring speed of the previous effective distance library, , choose to smallest As the optimal number of folds.

[0014] Furthermore, the quality indicators of the output include: QC=0 indicates invalid spectrum, QC=1 indicates partial folding correction, QC=2 indicates full folding correction, QC=3 indicates no correction required, and QC=4 indicates anomaly handling.

[0015] Furthermore, the anomaly handling includes at least one of the following methods: backtracking to the vertical continuity constraint solution, backtracking to the cross-band consistency constraint solution, or interpolating in the time or distance direction.

[0016] Compared with the prior art, the present invention has at least the following beneficial effects: This invention constructs a multi-band collaborative framework of "long-band reference - short-band correction", using the de-ambigued velocity profile of the first-band radar (long wavelength, large maximum unambiguous velocity) as a reliable reference, which makes up for the lack of reference of short-wavelength radar, reduces the risk of misjudgment in complex scenarios, and improves the accuracy of short-wavelength radar velocity de-ambiguation.

[0017] This invention subdivides velocity ambiguity types into partial folding and complete folding, designs targeted spectral domain reconstruction strategies, and selects the optimal solution by combining the joint cost function of "cross-band consistency + vertical continuity". It not only utilizes the similarity of multi-band spectral shapes to ensure correction accuracy, but also ensures the smoothness of results through vertical velocity continuity, thereby improving robustness and being able to stably adapt to complex meteorological scenarios such as weak echoes, multi-peak spectra, and strong turbulence.

[0018] This invention forms a complete technical closed loop of "preprocessing-core processing-postprocessing" through noise floor estimation and effective spectrum screening, quality label output and anomaly handling mechanism. It can not only output high-quality defuzzification speed data, but also provide clear quality classification basis for subsequent data applications, adapt to the needs of meteorological radar operational applications, and provide solid data support for cloud microphysical parameter inversion, wind field inversion and other applications. Attached Figure Description

[0019] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0020] Figure 1This is a flowchart illustrating the velocity deambiguation method based on joint constraints of multi-band radar Doppler power spectrum provided in this embodiment; Figure 2 This is a flowchart of the first-band radar deambiguation process provided in this embodiment, based on the velocity deambiguation method of joint constraints of multi-band radar Doppler power spectrum. Figure 3 The image shows a comparison of the Doppler radial velocities of the X-band radar before (top) and after (bottom) quality control, based on the velocity defuzzification method of multi-band radar Doppler power spectrum joint constraints provided in this embodiment. Figure 4 The image shows a comparison of the Doppler radial velocities of a Ka-band radar before (top) quality control and after (bottom) quality control, based on a velocity defuzzification method using multi-band radar Doppler power spectrum joint constraints, provided in this embodiment. Figure 5 The image shows a comparison of the Doppler radial velocity before (top) and after (bottom) quality control of the W-band radar using a velocity defuzzification method based on the joint constraint of the multi-band radar Doppler power spectrum, as provided in this embodiment. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] The following detailed description of some embodiments of the present invention is provided in conjunction with the accompanying drawings. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0023] This embodiment provides a velocity deambiguity method based on joint constraints of multi-band radar Doppler power spectrum, such as... Figure 1As shown, the core process is as follows: First, acquire power spectrum data and radar parameters of at least two radars in different bands (the first band radar is a long wavelength radar with a large maximum unambiguous velocity; the second band radar is a short wavelength radar with a small maximum unambiguous velocity); then, perform noise floor estimation and effective spectrum screening on the power spectrum of all bands and all range databases, and remove invalid range database data; next, perform deambiguation processing on the first band radar to obtain the reference velocity profile and reference expanded spectrum; establish cross-band data correspondence between the two bands through spatiotemporal matching; determine the velocity ambiguity type (partial folding or complete folding) for the second band radar database by database; reconstruct the spectral domain according to the ambiguity type and construct candidate expanded spectra; use cross-band consistency and vertical continuity to jointly determine the optimal solution, and finally obtain the deambigued velocity profile of the second band radar; finally, output the results and quality labels for each band.

[0024] Specifically, velocity deblurring involves recovering the true radial velocity from the folded (blurred) radial velocity measured by radar through candidate fold number determination and spectral domain reconstruction. The power spectrum deblurring noise reduction and confidence threshold mechanism integrates noise floor estimation, spectral smoothing, and effective spectral thresholds to construct a spectral quality control process based on signal-to-noise ratio (SNR). This process outputs a Quality Control (QC) label for each deblurring result, providing a reliable basis for subsequent product processing and outlier removal. SNR serves as the threshold indicator for determining the validity of the spectrum, while the QC label is responsible for generating quality level / confidence level labels.

[0025] This scheme introduces multi-band power spectrum joint constraints, leveraging the unambiguous nature of long-wave radar to provide a reliable reference for short-wave radar, breaking through the traditional technical path that relies solely on single-band velocity continuity. It distinguishes between two ambiguity types, partial folding and complete folding, and processes them separately. By employing a joint decision mechanism of cross-band consistency and vertical continuity, it solves the problem of insufficient accuracy and robustness in deambiguation under complex meteorological scenarios (such as weak echoes, multi-peak spectra, and strong turbulence), avoiding misjudgments caused by relying solely on average velocity or single continuity, and improving the data quality of short-wave radar under harsh conditions.

[0026] like Figure 2 As shown, the specific operational steps for deblurring the radar in the first band are detailed: an iterative processing method of range-by-range library from "cloud top to cloud bottom" is adopted. First, it is detected whether there is partial folding in the current library; if partial folding exists, two candidate expansion spectra, one positive and one negative, are constructed; the average Doppler velocity of the two candidate expansion spectra is calculated respectively to obtain two sets of candidate velocities; the deblurring velocity of the previous effective range library is used as a reference, and the candidate velocity with the smallest difference is selected as the optimal solution of the current library, and the corresponding expansion spectrum is recorded as the reference expansion spectrum.

[0027] Specifically, cloud top regions typically exhibit weak echoes and low velocities, making them suitable as initial reference areas for downward propagation, consistent with the physical characteristics of weather radar echoes. This scheme ensures the continuity and reliability of the first-band radar deblurring results through iterative downward propagation from cloud tops and vertical continuity decision-making. The cloud top region naturally has low velocity and is less prone to blurring, avoiding complex near-ground interference as a starting point. Targeted processing of partial folds (constructing positive / negative candidates) is directly based on spectral features, avoiding average velocity weighting bias. Continuity constraints reduce random errors, providing high-quality reference information for subsequent multi-band joint operations. Combining partial fold identification with vertical continuity forms an efficient and reliable long-wavelength deblurring process.

[0028] In this embodiment, the specific criteria for determining partial folding are clarified: First, the endpoint is determined to be the left endpoint of the effective spectrum. And the right endpoint of the effective spectrum ,in, The maximum unambiguous velocity of the first-band radar is used to help identify whether the spectral components of the second-band radar are close to or exceed their own observation boundaries. The preset velocity tolerance (usually 1-2 velocity bins) indicates that the effective spectrum simultaneously approaches the negative and positive ends of the velocity axis, suggesting that the true velocity spectrum is truncated by the Nyquist velocity boundary, and some energy is folded to the other side. Secondly, the energy determination is that the energy within both boundary windows is simultaneously higher than the noise floor. As a supplement to the endpoint determination, when the endpoint does not strictly meet the boundary conditions due to noise or discretization problems, partial folding is identified by the energy distribution in the boundary region. Meeting either condition is sufficient to determine partial folding.

[0029] This scheme provides quantization rules for partial spectral folding, making the algorithm clearly engineering-feasible. Endpoint determination directly corresponds to the physical phenomenon of spectral folding, while energy determination, as a supplement, avoids missed detections caused by discrete sampling and noise, thus improving recognition accuracy. Tolerance is introduced. This reflects consideration of the discreteness of real-world data and enhances the robustness of the algorithm.

[0030] In this embodiment, the construction method of the candidate expansion spectrum is defined: a positive candidate expansion spectrum refers to shifting the spectral components of the negative velocity region to the positive velocity side, corresponding to the case where the actual velocity is positive, that is, the energy that was originally folded into the negative velocity region should belong to the positive velocity region; a negative candidate expansion spectrum refers to shifting the spectral components of the positive velocity region to the negative velocity side, corresponding to the case where the actual velocity is negative, that is, the energy that was originally folded into the positive velocity region should belong to the negative velocity region. The initial expansion of the folded spectrum is achieved by shifting the spectral components.

[0031] This scheme constructs two candidate solutions, positive and negative, covering two possible true velocity directions, providing a complete solution space for subsequent decision-making. The translation operation avoids complex numerical calculations, ensuring the algorithm's real-time performance. This method is the core technique for handling partial folding. Compared with traditional methods based on average velocity, it can directly reconstruct the folded spectral energy, eliminating the mean bias problem.

[0032] In this embodiment, the specific implementation method for spatiotemporal matching of Doppler power spectrum data between the first-band radar and the second-band radar is defined: time alignment is based on the time series of the first-band radar, and nearest neighbor matching or linear interpolation matching is performed on the second-band radar, requiring a time error... ( (To maximize the allowable time difference), it solves the problem of time asynchrony caused by different scanning strategies of radars in different bands; range resampling resamples the power spectrum data of the second band radar onto the same range grid as the first band radar, solving the problem of data point misalignment caused by different range resolutions of radars in different bands.

[0033] This scheme eliminates data misalignment issues caused by differences in observation timing and range resolution in multi-band radars through time alignment and range resampling, establishing cross-band data correspondence within the same spatiotemporal context, thus providing a prerequisite for subsequent cross-band consistency constraints. It employs two methods: nearest neighbor matching (suitable for scenarios with small time differences and good data continuity) and linear interpolation matching (suitable for scenarios with large time differences and requiring smooth data transitions), allowing selection based on actual radar observation conditions and enhancing the adaptability of the technical solution. A clear time error threshold is defined. This ensures the accuracy of spatiotemporal matching, avoids cross-band constraint failure caused by spatiotemporal misalignment, and guarantees the accuracy of subsequent deblurring results.

[0034] In this embodiment, the steps for determining the velocity ambiguity type using a range-by-range database for the second-band radar are specifically defined: First, it is detected whether it is a partial fold according to a preset determination condition, which is the left endpoint of the effective spectrum. And the right endpoint of the effective spectrum Or, the energy within both boundary windows is simultaneously higher than the noise floor, where The maximum unambiguous velocity for the second-band radar. The preset speed tolerance is used; secondly, if it is not determined to be a partial fold, the difference between the initial speed of the current library and the defuzzification speed of the previous valid library is calculated. ,like ( If a preset velocity mutation threshold is set, a complete folding candidate judgment is triggered, and velocity mutations are detected based on vertical continuity; finally, when complete folding is triggered, the reference velocity of the first-band radar is used. Constructing a candidate set of folding times for second-band radar ,in, The reference velocity for the first-band radar, when If invalid, use the default candidate set.

[0035] This scheme constructs a complete fuzziness type determination mechanism, clearly distinguishing velocity fuzziness into two categories: partial folding and complete folding. This avoids errors in correction methods caused by confusion between the two folding types, improving the targeting and accuracy of short-wavelength radar defuzzification. Based on the first-band reference velocity, a candidate set of folding counts is constructed, narrowing the candidate range and improving processing efficiency and accuracy. A default candidate set is set when the reference velocity is invalid, ensuring that defuzzification in the second band can still proceed normally even when local data in the first band is invalid, thus enhancing the robustness of the technical solution.

[0036] In this embodiment, the implementation steps of the second-band radar spectral domain reconstruction are defined. For different ambiguity types, translation or offset operations are used to generate candidate solutions. When a partial fold is determined, positive and negative candidate expansion spectra are constructed, and priority is given based on the sign of the first-band radar reference velocity to calculate the partial fold correction velocity. When a complete fold is determined, the number of folds for each candidate is calculated. Based on the sign direction of the radar reference velocity in the first band, the original power spectrum is processed. The velocity axis is shifted, and the spectral components that exceed the original velocity display range after the shift will be processed according to the system's preset method, thereby obtaining the candidate unfolded spectrum corresponding to each candidate fold number.

[0037] This scheme designs differentiated reconstruction methods for different fuzzy types, achieving accurate correction. At the same time, it uses the sign constraint of the reference velocity to narrow the range of candidate solutions. It constructs a corresponding candidate expansion spectrum for each candidate fold number, covering all possible real velocity solutions, while limiting the search range of fold number to the candidate set constructed above, thus balancing completeness and efficiency.

[0038] In this embodiment, the core of joint decision-making is to adopt a joint cost function. Select the optimal number of folds The calculation formula is: ,in For the first The candidate velocity corresponding to the number of candidate folds The deblurring speed of the previous effective distance library, In other words, cross-band consistency takes precedence over vertical continuity, and the selection prioritizes cross-band consistency. smallest As the optimal number of folds.

[0039] Specifically, by simultaneously utilizing features such as spectral peak position, spectral width, and spectral moments (e.g., first / second moments) in the spectral domain, a joint cost function is formed to achieve robust inversion of folding velocity (especially suitable for multi-peak spectra, weak signals, or turbulence enhancement scenarios). This scheme integrates two different types of information—cross-band physical consistency and spatial continuity—into a single decision framework, achieving complementary advantages from multiple sources; by setting... This approach embodies a decision-making process that prioritizes cross-band physical consistency and is supplemented by spatial continuity, balancing the needs for accuracy and continuity. It transforms the selection of the optimal solution into quantitative calculations, avoiding subjective judgment.

[0040] In this embodiment, the output quality identifiers are categorized and defined, specifying five quality states and their meanings: QC=0 indicates an invalid spectrum, meaning the distance library failed the valid spectrum threshold and has no valid velocity output; QC=1 indicates partial folding correction, meaning the distance library detected partial folding and successfully completed the correction; QC=2 indicates complete folding correction, meaning the distance library detected complete folding and successfully completed the correction; QC=3 indicates no correction is needed, meaning the distance library has not experienced velocity ambiguity and requires no processing; QC=4 indicates abnormal processing, meaning an anomaly occurred during processing (such as an empty candidate set, no stable optimal solution for the cost function, etc.), and reliable results cannot be obtained.

[0041] This solution establishes a data quality grading system and adds a processing status label to each output speed, enabling users to understand the credibility and source of the data, while providing an intuitive basis for algorithm optimization and verification.

[0042] In this embodiment, the specific methods for handling anomalies are clearly defined, including at least one of the following: reverting to the vertical continuity constraint solution, abandoning the cross-band consistency constraint, and making decisions based solely on vertical continuity, i.e., this method is used when cross-band references may be unreliable; reverting to the cross-band consistency constraint solution, abandoning the vertical continuity constraint, and making decisions based solely on cross-band consistency, i.e., this method is used when there may be severe wind shear in the vertical direction; and interpolating in the time or distance direction, using the effective velocity of the nearest time or nearest distance library for interpolation, which is suitable for isolated anomalies.

[0043] This solution provides multiple technical strategies for abnormal scenarios, solving the problem of interruption of overall results due to local anomalies in traditional methods. It ensures the integrity and continuity of data, and different processing methods can be adapted to different anomaly types, improving the adaptability and reliability of the technical solution and providing a guarantee for stable operation in complex observation environments.

[0044] The present invention also provides the following specific embodiments: Taking the millimeter-wave-centimeter-wave multi-band dual-polarization Doppler radar platform jointly developed by the Institute of Atmospheric Physics, Chinese Academy of Sciences and Sifang Electronics Co., Ltd. as an example, combined with the attached... Figure 3-5 The platform integrates X, Ka, and W band radars, enabling multi-scale collaborative detection of cloud precipitation systems.

[0045] In this specific observation, the PRF values ​​for the X, Ka, and W bands were 1400 Hz, 1400 Hz, and 5000 Hz, respectively, with corresponding maximum unambiguous radial velocities of 11.2 ms. 1 3.01ms 1 and 4.0ms 1 .

[0046] Velocity deambiguation is the process of recovering the true radial velocity from the folded (ambiguous) radial velocity measured by radar through candidate fold number determination and spectral domain reconstruction; the Nyquist velocity is the maximum unambiguous velocity determined by the PRF, denoted as: It is the maximum unambiguous velocity in the X-band (m / s); It is the maximum unambiguous velocity in the Ka band (m / s); It is the maximum unambiguous velocity (m / s) in the W-band; the power distribution of the echo signal obtained by Fast Fourier Transform (FFT) as a function of Doppler velocity at the same distance and time is denoted as... ,in .

[0047] S1: Data Reading and Parameter Initialization Read Doppler power spectrum data and auxiliary parameters from X, Ka, and W band radars, including but not limited to: timestamps. ,distance Discrete points of velocity axis (m / s), PRF for each band and corresponding maximum unambiguous velocity Distance resolution Time resolution If the power spectrum is expressed in decibel-milliwatts (dBm), then it is converted to linear power. The calculation formula is: Noise floor It is the average power of the non-signal portion of the power spectrum, calculated using the following formula: ,in, This is expressed in dBm, representing the noise floor.

[0048] S2: Noise floor estimation and effective spectral threshold determination For each band and each range, the noise floor of the Doppler power spectrum is estimated using methods such as the spectral tail quantile method and the minimum energy interval averaging method. When the spectral peak or spectral energy satisfies At that time, the distance database is considered valid, where The threshold is set; for invalid libraries, output QC=0 and skip subsequent steps or perform interpolation to fill in the gaps.

[0049] S3: X-band (long wavelength) deblurring X-band velocity deblurring is processed using a distance-by-distance library from cloud top to cloud bottom. Near the cloud top, due to the weak echo and low velocity, it can be used as an initial reference area.

[0050] S3-1: Endpoint Extraction and "Partial Folding" Detection For each effective range library, detect whether partial folding exists and define the effective spectral endpoints: left endpoint velocity. satisfy The smallest Right endpoint velocity satisfy The largest A partial fold can be determined to have occurred when one of the following conditions is met: Condition A: When the endpoint of the effective spectrum is close to the velocity axis boundary, i.e., the left endpoint of the effective spectrum And the right endpoint of the effective spectrum ; Condition B: Significant energies appear simultaneously near both boundaries of the spectrum (this can be determined using the boundary window energy ratio). For speed tolerance (it is recommended to use 1–2 speed bins).

[0051] S3-2: Constructing a positive candidate expansion spectrum Negative candidate expansion spectrum

[0052] For the spectrum determined to be partially folded, construct two candidate expansion spectra and calculate the corresponding candidate velocities: Candidate expansion spectrum : Shift the negative velocity spectral components to the positive velocity side:

[0053] Candidate expansion spectrum : Shift the positive velocity spectral components to the negative velocity side:

[0054] The above "translation" can be achieved by cyclic shifting or by "filling in zeros when the boundary is exceeded".

[0055] S3-3: Velocity Estimation For any candidate expanded spectrum, the average Doppler velocity is calculated using the linear power after noise subtraction as a weight:

[0056] in, The portion less than 0 is truncated to 0; two candidate velocities are obtained. , .

[0057] S3-4: Continuity decision to determine the optimal solution Suppose that during the iterative process from the top of the cloud downwards, the deblurring speed of the previous effective distance library is... , choose to The candidate speed with the smallest difference is selected as the optimal defuzzification speed in the current library. And record the corresponding expansion spectrum as ; If the previous valid distance library is invalid, the speed of the nearest valid library above can be used, or a reference value smoothed by a short time window can be introduced; Output: Obtain the deblurring velocity profile in the X-band. And its QC mark.

[0058] S4: Multi-band distance-temporal two-dimensional spatial matching To enable joint deblurring of radars from different bands at the same volume and observation time, the time series of the X-band is used as a reference to perform two-dimensional range and time matching on the Ka / W band, forming comparable data pairs.

[0059] S4-1 Time Alignment Using the X-band time series as a reference, the Ka / W bands are matched by timestamp: If and If they are the same, a one-to-one correspondence is used time-by-time; if they are different, nearest neighbor matching or linear interpolation matching is used, requiring the matching time error to meet the following requirements. ;in, The maximum allowable time difference (can be set to 0.5s or determined by the system).

[0060] S4-2 Distance Alignment / Resampling Align the distance direction: If Then they correspond one-to-one according to the same distance library index; if Then, the X-band or Ka / W-band spectral data can be resampled to a uniform distance grid (e.g., a uniform grid with a coarser resolution) by using nearest neighbor, linear interpolation, or interval averaging methods.

[0061] Establish mapping relationship after matching Output: Matched spectral data pairs and X-band reference velocity .

[0062] S5: Ka / W Band Velocity Ambiguity Type Determination (Partial Folding vs. Complete Folding) The short-wavelength radar (Ka / W band) is processed range by range from cloud top to cloud bottom to determine the velocity ambiguity type and to determine the processing path for subsequent spectral reconstruction.

[0063] S5-1 Partial Folding Judgment Partial folding determination is similar to S3-1: based on the effective spectrum endpoints. Do they approach each other simultaneously? If the energy of both boundary windows is significant at the same time, it is judged as partial folding.

[0064] S5-2 Fully Folded Trigger Decision (Based on Vertical Continuity) When a sudden change in velocity occurs between adjacent distances, a "fully folded candidate" judgment is triggered. Specifically: the original observed Doppler velocity of the current library, obtained without deblurring or according to partially folded candidates, is used as the initial velocity. ; Calculate the difference between the initial velocity of the current library and the defuzzification velocity of the previous effective library: ,like This triggers the complete folding candidate judgment; where, The velocity mutation threshold (can be set to) to (Within the scope or determined by system experience).

[0065] Construction of the candidate set for S5-3 fold counts (combined with X-band reference) For the range library that triggers complete folding, the de-ambiguous reference velocity in the X-band is used. Constructing a candidate set of Ka / W fold counts .

[0066] when When effective, let the target band be... Its Nyquist speed is The candidate set of fold counts satisfies: When the set is empty or the X-band reference library is invalid, the default candidate set is used. ,in, Determined by the maximum number of folds allowed by the system.

[0067] S6: Ka / W Band Velocity De-ambiguity (Spectral Domain Reconstruction and Optimal Solution Selection) Deblurring was performed on the Ka / W bands for "partial folding" and "complete folding" respectively.

[0068] S6-1 Partial Folding De-Blurring (Indicated by X-band Symbols) When a partial fold is identified, positive and negative candidate expansion spectra are constructed. , (Same as S3-2, except that...) Replace with To reduce ambiguity, X-band reference velocity can be used. The symbol is selected first: if Prioritize candidates that shift the negative velocity region to the positive velocity side; if Candidates that shift the positive velocity region to the negative velocity side are selected first; and the velocity is calculated for the expanded spectrum of the candidates. (Same as S3-3), obtain the partial folding correction speed. .

[0069] S6-2 Fully Folded Deblurring (Spectral Shift Based on Folding Number n) When a fold is determined to be completely folded, the number of folds for each candidate is... Construct candidate "expansion spectra" based on The sign direction is used to modify the original spectrum. The offset is used to obtain the candidate expansion spectrum. And calculate candidate velocities : like Then the spectrum is shifted in the positive direction. (Equivalent to velocity axis translation or spectral bin translation); if Then the spectrum is negatively shifted. , , and obtain candidate expansion spectra And calculate the corresponding candidate velocities: .

[0070] S6-3 Optimal Candidate Selection (Joint Decision Based on Cross-Band Consistency and Continuity) Select the optimal number of folds and the optimal speed from the candidate set. Using a joint cost function Minimize: ,in Weighting coefficients (can be set) Prioritizing cross-band consistency.

[0071] Choose to smallest As the number of folds, output And record the corresponding expansion spectrum. Output: Ka / W deblurred velocity profile With quality label .

[0072] S7: Output deblurring velocity profiles for each band and corresponding quality indicators. The deblurring results for each band are output with QC flags: QC=0 indicates an invalid spectrum or failure to meet the threshold, and no valid velocity is output; QC=1 indicates partial folding and completion of correction; QC=2 indicates complete folding and completion of correction; QC=3 indicates no correction is needed (no folding occurred); QC=4 indicates anomaly handling (e.g., an empty candidate set, no stable optimal solution for the cost function, or severe inconsistency with the reference). Anomaly handling strategies may include: reverting to a continuity-constrained solution, reverting to a cross-band consistent solution, or interpolating in the time and distance directions.

[0073] The final output should include at least: X-band deblurring speed. Ka-band deblurring speed W-band deblurring speed QC markings and folding times for each band ; Deblurred expansion spectrum For algorithm verification or product generation.

[0074] The results show that there is obvious velocity ambiguity in the original Ka and W band radial velocities. After processing with the method of this invention, the radial velocity deambiguation processing of Ka and W bands and the corresponding data quality control were successfully achieved, which improved the reliability of Doppler velocity products and laid a data foundation for subsequent cloud microphysical parameter inversion based on observation parameters such as multi-band reflectivity and radial velocity.

[0075] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A velocity deambiguity method based on joint constraints of multi-band radar Doppler power spectrum, characterized in that, Includes the following steps: S1. Obtain Doppler power spectrum data and corresponding radar parameters from at least two radars in different bands, wherein the radars in different bands include at least a first-band radar and a second-band radar, and the maximum unambiguous velocity of the second-band radar is less than the maximum unambiguous velocity of the first-band radar. S2. Estimate the noise floor and determine the effective spectrum threshold for the Doppler power spectrum of each band and each range database, and remove invalid range database data; S3. Perform deblurring on the first-band radar to obtain the reference velocity profile and the corresponding reference expanded spectrum; S4. Perform spatiotemporal matching of the Doppler power spectrum data of the first band radar and the second band radar to establish a cross-band data correspondence; S5. Determine the velocity ambiguity type for the second band radar range-by-range database, wherein the ambiguity type includes partial folding and complete folding; S6. Based on the velocity ambiguity type, the Doppler power spectrum of the second-band radar is reconstructed in the spectral domain to construct candidate expanded spectra. Then, by using the joint decision of cross-band consistency constraints and vertical continuity, the optimal number of folds and deambiguity velocity are selected to obtain the deambiguity velocity profile of the second-band radar. S7. Output the deblurring velocity profiles for each band and the corresponding quality indicators.

2. The velocity deambiguity method based on joint constraints of multi-band radar Doppler power spectrum according to claim 1, characterized in that, The deblurring process for the first-band radar includes the following steps: S3-1. Employ a distance-by-distance library iterative processing from cloud top to cloud bottom to detect whether the Doppler power spectrum of the current distance library has partial folding; S3-2. When it is determined to be a partial fold, construct the positive candidate expansion spectrum and the negative candidate expansion spectrum; S3-3. Calculate the average Doppler velocity of the positive candidate expansion spectrum and the negative candidate expansion spectrum respectively to obtain two candidate velocities; S3-4. Using the deblurring speed of the previous effective distance library as a reference, select the candidate speed with the smallest difference from the reference speed as the optimal deblurring speed of the current library, and record the corresponding expansion spectrum as the reference expansion spectrum.

3. The velocity deambiguity method based on joint constraints of multi-band radar Doppler power spectrum according to claim 2, characterized in that, The determination condition for the partial folding is as follows: Left end of effective spectrum And the right endpoint of the effective spectrum Or, the energy within both boundary windows is simultaneously higher than the noise floor, where, The maximum unambiguous velocity for the first-band radar. This is the preset speed tolerance.

4. The velocity deambiguity method based on joint constraints of multi-band radar Doppler power spectrum according to claim 2, characterized in that, The construction of a positive candidate expansion spectrum refers to shifting the spectral components of the negative velocity region to the positive velocity side, while the construction of a negative candidate expansion spectrum refers to shifting the spectral components of the positive velocity region to the negative velocity side.

5. The velocity deambiguity method based on joint constraints of multi-band radar Doppler power spectrum according to claim 1, characterized in that, The process of performing spatiotemporal matching of the Doppler power spectrum data of the first-band radar and the second-band radar includes time alignment and range resampling. Using the time series data of the first-band radar as a reference, nearest neighbor matching or linear interpolation matching is performed on the second-band radar, requiring a time error... ,in, This is the maximum permissible time difference; The power spectrum data of the second-band radar is resampled onto the same range grid as the first-band radar.

6. The velocity deambiguity method based on joint constraints of multi-band radar Doppler power spectrum according to claim 1, characterized in that, The determination of velocity ambiguity type for the second-band radar range-by-range database includes the following steps: S5-1. Detect whether it is a partial fold according to a preset judgment condition, wherein the judgment condition is the left endpoint of the effective spectrum. And the right endpoint of the effective spectrum Or, the energy within both boundary windows is simultaneously higher than the noise floor, where The maximum unambiguous velocity for the second-band radar. Preset speed tolerance; S5-2. If it is not determined to be a partial fold, calculate the difference between the initial velocity of the current library and the defuzzification velocity of the previous valid library. ,like If this is the case, then a complete folding candidate judgment is triggered, where, The preset speed change threshold; S5-3. When the fully folded candidate judgment is triggered, the reference velocity of the first-band radar is used. Constructing a candidate set of folding times for second-band radar ,in, The reference velocity for the first-band radar, when If invalid, use the default candidate set.

7. The velocity deambiguity method based on joint constraints of multi-band radar Doppler power spectrum according to claim 1, characterized in that, The spectral domain reconstruction of the second-band radar includes the following steps: S6-1. When it is determined to be a partial fold, construct a positive candidate expansion spectrum and a negative candidate expansion spectrum, and select the candidate spectrum based on the sign of the radar reference velocity in the first band to calculate the partial fold correction velocity. S6-2. When a complete fold is determined, the number of folds for each candidate is calculated. Based on the sign direction of the radar reference velocity in the first band, the original power spectrum is processed. The velocity axis offset is used to obtain the candidate unfolding spectrum corresponding to the number of candidate folds.

8. The velocity deambiguity method based on joint constraints of multi-band radar Doppler power spectrum according to claim 1, characterized in that, The joint decision is made using a joint cost function. Select the optimal number of folds The calculation formula is: ,in, For the first The candidate velocity corresponding to the number of candidate folds The deblurring speed of the previous effective distance library, , choose to smallest As the optimal number of folds.

9. The velocity deambiguity method based on joint constraints of multi-band radar Doppler power spectrum according to claim 1, characterized in that, The output quality indicators include: QC=0 indicates invalid spectrum, QC=1 indicates partial folding correction, QC=2 indicates full folding correction, QC=3 indicates no correction required, and QC=4 indicates anomaly handling.

10. The velocity deambiguity method based on joint constraints of multi-band radar Doppler power spectrum according to claim 9, characterized in that, The anomaly handling includes at least one of the following methods: back to the vertical continuity constraint solution, back to the cross-band consistency constraint solution, or interpolation filling in the time or distance direction.

Images