A raindrop spectrum correction method, device, equipment and storage medium

Abstract

The present application discloses a raindrop spectrum correction method, device, equipment and storage medium, which relates to the field of raindrop measurement technology, including: calculating the scattered echo signal of the laser radar to obtain the laser radar power spectrum corresponding to each of the different range gates in the laser radar; judging whether the laser radar power spectrum has a double peak structure, and determining whether the laser radar power spectrum is the target radar power spectrum according to the corresponding judgment result; extracting the corresponding target raindrop spectrum from each determined target radar power spectrum; constructing a model training set based on the first characteristic information of the target raindrop spectrum and the corresponding raindrop particle size label; training a raindrop spectrum correction model using the model training set, and correcting the raindrop spectrum to be corrected using the trained model. The present application improves the accuracy of the raindrop particle size output by the model by constructing a training set using the raindrop spectra corresponding to several range gates, and training the raindrop spectrum correction model using the training set with the raindrop particle size as the label.

Description

Technical Field

[0001] The present invention relates to the field of raindrop measurement technology, and in particular to a raindrop spectrum correction method, device, equipment and storage medium. Background Art

[0002] When there is no precipitation, the LiDAR power spectrum is usually a single-peak structure, and its peak position represents the wind speed information. At this time, the corresponding spectrum distribution at all heights is not significantly broadened, but a narrow band. When there is precipitation, the falling speed of precipitation particles will also produce Doppler frequency shift, and a double-peak structure will appear on the LiDAR power spectrum.

[0003] The existing technology mainly focuses on solving the rainfall rate reflection spectrum, applying the backscattering cross section of raindrops to the calculation of raindrop spectrum, and only considers the calculation of particle size using the traditional empirical formula. Based on this method, the raindrop particle size in the raindrop spectrum separated from the radar power spectrum has the problem of inaccurate results. Summary of the invention

[0004] In view of this, the purpose of the present invention is to provide a raindrop spectrum correction method, device, equipment and storage medium, which can extract a target raindrop spectrum that meets preset conditions from a laser radar power spectrum, and train a preset raindrop spectrum correction model with a label constructed by raindrop particle size collected by a raindrop spectrometer that measures synchronously with a preset laser radar and the first characteristic information of the target raindrop spectrum, so as to use the trained raindrop correction model to correct the raindrop spectrum to be corrected, thereby solving the problem of inaccurate measured raindrop particle size. The specific scheme is as follows:

[0005] In a first aspect, the present application provides a raindrop spectrum correction method, comprising:

[0006] Collecting scattered echo signals of a preset laser radar, and calculating the scattered echo signals to obtain laser radar power spectra corresponding to different range gates in the preset laser radar;

[0007] Determine whether the laser radar power spectrum has a target double-peak structure, and determine whether the laser radar power spectrum is a target radar power spectrum according to the corresponding determination result;

[0008] Extracting corresponding target raindrop spectra from the determined target radar power spectra respectively;

[0009] A corresponding model training set is constructed based on the first characteristic information of the target raindrop spectrum and the corresponding raindrop particle size label to obtain a plurality of model training sets corresponding to different range gates respectively; wherein the raindrop particle size label is a label constructed based on the raindrop particle size collected by a raindrop spectrometer synchronously measured with the preset laser radar;

[0010] The preset raindrop spectrum correction model is trained using the model training sets corresponding to the range gates to obtain a trained raindrop spectrum correction model, and the trained raindrop spectrum correction model is used to correct the raindrop spectrum to be corrected to obtain a corresponding corrected raindrop spectrum.

[0011] Optionally, the calculating the scattered echo signal to obtain the laser radar power spectra corresponding to different range gates in the preset laser radar includes:

[0012] Performing a fast Fourier transform on the scattered echo signal light to obtain initial radar power spectra corresponding to different range gates in the preset laser radar;

[0013] The background noise of the initial radar power spectrum is calculated, and a background noise removal operation is performed on the initial radar power spectrum to obtain the laser radar power spectrum corresponding to the initial radar power spectrum.

[0014] Optionally, judging whether the laser radar power spectrum has a target double-peak structure, and determining whether the laser radar power spectrum is a target radar power spectrum according to a corresponding judgment result, includes:

[0015] Determine the currently traversed laser radar power spectrum as the current laser radar power spectrum, and determine whether the signal-to-noise ratio of the current laser radar power spectrum is greater than the corresponding signal-to-noise ratio threshold to obtain a corresponding first current judgment result;

[0016] If the first current judgment result indicates that the signal-to-noise ratio of the current laser radar power spectrum is greater than the corresponding signal-to-noise ratio threshold, determine whether the current laser radar power spectrum has the target double-peak structure, and determine whether the current laser radar power spectrum is the target radar power spectrum according to the corresponding second current judgment result, and then traverse the next laser radar power spectrum to obtain a new current laser radar power spectrum;

[0017] If the first current judgment result indicates that the signal-to-noise ratio of the current laser radar power spectrum is not greater than the corresponding signal-to-noise ratio threshold, traverse the next laser radar power spectrum to obtain a new current laser radar power spectrum;

[0018] Re-jump to the step of determining whether the signal-to-noise ratio of the current laser radar power spectrum is greater than the corresponding signal-to-noise ratio threshold until all the laser radar power spectra are traversed.

[0019] Optionally, extracting corresponding target raindrop spectra from the determined target radar power spectra respectively includes:

[0020] Determine first initial spectrum data of an initial raindrop spectrum corresponding to the target radar power spectrum, and obtain second initial spectrum data corresponding to the initial raindrop spectrum through a preset algorithm and the first initial spectrum data;

[0021] The target raindrop spectrum is determined from the initial raindrop spectrum based on the first initial spectrum data and the second initial spectrum data.

[0022] Optionally, determining the target raindrop spectrum from the initial raindrop spectrum based on the first initial spectrum data and the second initial spectrum data includes:

[0023] Calculating a first root mean square error corresponding to the first initial spectrum data and a second root mean square error corresponding to the second initial spectrum data of the initial raindrop spectrum, and determining whether the first root mean square error and the second root mean square error are greater than a corresponding first root mean square error threshold and a second root mean square error threshold;

[0024] If the first root mean square error is less than the first root mean square error threshold, and the second root mean square error is less than the second root mean square error threshold, the initial raindrop spectrum is determined as the target raindrop spectrum.

[0025] Optionally, the constructing a corresponding model training set based on the first characteristic information of the target raindrop spectrum and the corresponding raindrop particle size label to obtain a plurality of model training sets corresponding to different range gates respectively includes:

[0026] Constructing target data including first characteristic information of the target raindrop spectrum and corresponding range gate information of the range gate;

[0027] The target data is labeled with corresponding raindrop particle size labels to obtain a plurality of model training sets corresponding to different range gates, so that in the process of training the preset raindrop spectrum correction model, the target data in the model training set is sent to the input end of the preset raindrop spectrum correction model for training, and the raindrop particle size labels in the model training set are sent to the output end of the preset raindrop spectrum correction model for training.

[0028] In a second aspect, the present application provides a raindrop spectrum correction model training method, comprising:

[0029] Collecting scattered echo signals of a preset laser radar, and calculating the scattered echo signals to obtain laser radar power spectra corresponding to different range gates in the preset laser radar;

[0030] Determine whether the laser radar power spectrum has a target double-peak structure, and determine whether the laser radar power spectrum is a target radar power spectrum according to the corresponding determination result;

[0031] Extracting corresponding target raindrop spectra from the determined target radar power spectra respectively;

[0032] A corresponding model training set is constructed based on the first characteristic information of the target raindrop spectrum and the corresponding raindrop particle size label to obtain a plurality of model training sets corresponding to different range gates respectively; wherein the raindrop particle size label is a label constructed based on the raindrop particle size collected by a raindrop spectrometer synchronously measured with the preset laser radar;

[0033] The model training sets corresponding to the range gates are used to train the preset raindrop spectrum correction model to obtain the trained raindrop spectrum correction model.

[0034] In a third aspect, the present application provides a raindrop spectrum correction device, comprising:

[0035] A power spectrum acquisition module, used for collecting scattered echo signals of a preset laser radar, and calculating the scattered echo signals to obtain laser radar power spectra corresponding to different range gates in the preset laser radar;

[0036] A power spectrum determination module, used to determine whether the laser radar power spectrum has a target double peak structure, and determine whether the laser radar power spectrum is a target radar power spectrum according to the corresponding judgment result;

[0037] A raindrop spectrum extraction module, used to extract corresponding target raindrop spectra from the determined target radar power spectra respectively;

[0038] A training set construction module, used to construct a corresponding model training set based on the first characteristic information of the target raindrop spectrum and the corresponding raindrop particle size label, so as to obtain a plurality of model training sets corresponding to different range gates; wherein the raindrop particle size label is a label constructed based on the raindrop particle size collected by the raindrop spectrometer synchronously measured with the preset laser radar;

[0039] The raindrop spectrum correction module is used to train the preset raindrop spectrum correction model using the model training set corresponding to each of the range gates to obtain the trained raindrop spectrum correction model, and use the trained raindrop spectrum correction model to correct the raindrop spectrum to be corrected to obtain the corresponding corrected raindrop spectrum.

[0040] In a fourth aspect, the present application provides an electronic device, including:

[0041] Memory, used to store computer programs;

[0042] A processor is used to execute the computer program to implement the aforementioned raindrop spectrum correction method.

[0043] In a fifth aspect, the present application provides a computer-readable storage medium for storing a computer program, wherein the computer program implements the aforementioned raindrop spectrum correction method when executed by a processor.

[0044] In the present application, firstly, the scattered echo signal of the preset laser radar is collected, and the scattered echo signal is calculated to obtain the laser radar power spectrum corresponding to each of the different range gates in the preset laser radar; it is determined whether the laser radar power spectrum has a target double-peak structure, and whether the laser radar power spectrum is a target radar power spectrum is determined according to the corresponding judgment result; then, the corresponding target raindrop spectrum is extracted from each of the determined target radar power spectra; then, a corresponding model training set is constructed based on the first characteristic information of the target raindrop spectrum and the corresponding raindrop particle size label to obtain several model training sets corresponding to the different range gates; wherein the raindrop particle size label is a label constructed based on the raindrop particle size collected by a raindrop spectrometer that measures synchronously with the preset laser radar; finally, the preset raindrop spectrum correction model is trained using the model training sets corresponding to each of the range gates to obtain the trained raindrop spectrum correction model, and the trained raindrop spectrum correction model is used to correct the raindrop spectrum to obtain the corresponding corrected raindrop spectrum. It can be seen that the present application obtains the target raindrop spectrum from the laser radar power spectrum corresponding to different range gates of the laser radar, and uses the first characteristic information of the target raindrop spectrum and the raindrop particle size label to construct a model training set to train the preset raindrop spectrum correction model, thereby improving the accuracy of the raindrop particle size output by the trained raindrop spectrum correction model. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0046] Figure 1 A flow chart of a raindrop spectrum correction method disclosed in this application;

[0047] Figure 2 A flow chart of a training set acquisition method disclosed in this application;

[0048] Figure 3 A comparison diagram of a laser radar power spectrum structure disclosed in the present application; wherein (a) is a laser radar power spectrum without a double-peak structure, and (b) is a laser radar power spectrum with a double-peak structure;

[0049] Figure 4 A schematic diagram of a laser radar power spectrum with a double-peak structure disclosed in this application;

[0050] Figure 5 A flow chart of a raindrop spectrum correction model training method disclosed in this application;

[0051] Figure 6 This is a schematic diagram of the structure of a raindrop spectrum correction device disclosed in this application;

[0052] Figure 7 This is a structural diagram of an electronic device disclosed in this application. DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0054] At present, the calculation of raindrop particle size only considers the method of using traditional empirical formulas, which results in inaccurate raindrop particle size results. To this end, the present application provides a raindrop spectrum correction method, which constructs a training set by using raindrop spectra corresponding to several range gates, and uses the training set to train a raindrop spectrum correction model with raindrop particle size as a label, thereby improving the accuracy of the model output raindrop particle size.

[0055] See also Figure 1 As shown, an embodiment of the present invention discloses a raindrop spectrum correction method, including:

[0056] Step S11, collecting scattered echo signals of a preset laser radar, and calculating the scattered echo signals to obtain laser radar power spectra corresponding to different range gates in the preset laser radar.

[0057] In this embodiment, the scattered echo signal of the laser radar can be collected at preset time intervals. The preset time interval can be set according to specific needs and is not specifically limited here. The scattered echo signal is calculated to obtain the laser radar power spectrum corresponding to each of the different range gates in the preset laser radar. Specifically, it can include: performing a fast Fourier transform, i.e., FFT operation (Fast Fourier Transformation) on the scattered echo signal light to obtain the initial radar power spectrum corresponding to the different range gates in the preset laser radar; calculating the background noise of the initial radar power spectrum, and performing a background noise removal operation on the initial radar power spectrum to obtain the laser radar power spectrum corresponding to the initial radar power spectrum. By performing a background noise removal operation on the initial radar power spectrum, the influence of the background noise on the spectral data in the radar power spectrum is avoided, and the reliability of the spectral data in the laser radar power spectrum after the background noise is removed is ensured.

[0058] Step S12: determine whether the laser radar power spectrum has a target double-peak structure, and determine whether the laser radar power spectrum is a target radar power spectrum based on the corresponding determination result.

[0059] In this embodiment, it is determined whether there is a target double peak structure in the laser radar power spectrum after removing the background noise, and whether the laser radar power spectrum is the target power spectrum according to the corresponding judgment result, such as Figure 2 As shown, it can specifically include: determining the currently traversed laser radar power spectrum as the current laser radar power spectrum, judging whether the signal-to-noise ratio of the current laser radar power spectrum is greater than the corresponding signal-to-noise ratio threshold to obtain the corresponding first current judgment result; if the first current judgment result shows that the signal-to-noise ratio of the current laser radar power spectrum is greater than the corresponding signal-to-noise ratio threshold, then judging whether the current laser radar power spectrum has a target double-peak structure, and determining whether the current laser radar power spectrum is the target radar power spectrum according to the corresponding second current judgment result, and then traversing the next laser radar power spectrum to obtain a new current laser radar power spectrum; if the first current judgment result shows that the signal-to-noise ratio of the current laser radar power spectrum is not greater than the corresponding signal-to-noise ratio threshold, then traversing the next laser radar power spectrum to obtain a new current laser radar power spectrum; and jumping back to the step of judging whether the signal-to-noise ratio of the current laser radar power spectrum is greater than the corresponding signal-to-noise ratio threshold until all laser radar power spectra are traversed. In a specific implementation, the spectrum data of the laser radar power spectrum of each range gate is traversed in a loop, and the signal-to-noise ratio threshold value snr_th is first set. If the signal-to-noise ratio is less than snr_th, the spectrum data corresponding to the next range gate is traversed. If the signal-to-noise ratio is greater than snr_th, the first-order differential d1 of the spectrum data and the second-order differential d2 of the spectrum data are calculated. If a zero crossing point that changes sharply from a positive value to a negative value appears in the first-order differential spectrum d1 and the zero point position is a local minimum in the second-order differential spectrum d2, it is determined that the radar power spectrum has a double-peak structure, and accordingly, the radar power spectrum with a double-peak structure is determined as the target radar power spectrum; if the currently traversed radar power spectrum does not have a double-peak structure, the radar power spectrum does not perform the subsequent operation of extracting the raindrop spectrum from the laser radar power spectrum. It can be understood that Figure 3 are two types of lidar power spectra, where (a) is the lidar power spectrum without a double-peak structure, and (b) is the lidar power spectrum with a double-peak structure. When there is precipitation, the falling raindrops will cause the scattered echo signal to produce a Doppler frequency shift. Due to the difference in the speed of raindrops and aerosols, the lidar echo signal spectrum will be broadened or have a double-peak structure. The relevant calculation echo spectrum formula is as follows:

[0060] ;

[0061] in, and are the Doppler shifts of the aerosol and raindrop power spectra, i.e., the peak positions of the aerosol and raindrop power spectra, and are the spectral intensities of the aerosol spectrum and the raindrop spectrum, and are the broadening of the aerosol spectrum and the raindrop spectrum, Refers to the center frequency shift point of the laser radar power spectrum. At this time, a double peak structure will appear on the laser radar power spectrum; if there is no precipitation, there will be no double peak structure on the laser radar power spectrum; Figure 4 As shown in the figure, for the lidar power spectrum with a double-peak structure, the accurate wind speed and precipitation speed can be obtained by performing a double Gaussian fitting on it. According to the empirical formula, there is an exponential relationship between the raindrop particle size and the falling speed. The raindrop spectrum can be directly obtained through the empirical formula, but the correct separation of the raindrop and wind speed spectrum, the difference in the pulse width of the raindrop spectrum, and the size of the peak height will also affect the final result to a certain extent, so the raindrop spectrum needs to be corrected in the end. By first judging whether the signal-to-noise ratio of the radar power spectrum meets the preset conditions, the validity and readability of the spectrum data in the screened radar power spectrum are guaranteed. By judging whether the radar power spectrum has a double-peak structure, the reliability of the process of extracting the raindrop spectrum from the radar power spectrum is guaranteed.

[0062] Step S13: extracting corresponding target raindrop spectra from the determined target radar power spectra respectively.

[0063] In this embodiment, the process of extracting target raindrops from a radar power spectrum with a target double-peak structure may specifically include: determining first initial spectrum data of an initial raindrop spectrum corresponding to the target radar power spectrum, and obtaining second initial spectrum data corresponding to the initial raindrop spectrum through a preset algorithm and the first initial spectrum data; determining the target raindrop spectrum from the initial raindrop spectrum based on the first initial spectrum data and the second initial spectrum data. The above process of determining the target raindrop spectrum from the initial raindrop spectrum based on the first initial spectrum data and the second initial spectrum data may specifically include: calculating a first root mean square error corresponding to the first initial spectrum data of the initial raindrop spectrum and a second root mean square error corresponding to the second initial spectrum data, and determining whether the first root mean square error and the second root mean square error are greater than the corresponding first root mean square error threshold and second root mean square error threshold; if the first root mean square error is less than the first root mean square error threshold, and the second root mean square error is less than the second root mean square error threshold, the initial raindrop spectrum is determined as the target raindrop spectrum. In a specific embodiment, the peak position of the initial raindrop spectrum in the target radar power spectrum is determined. and strength , i.e. the first initial spectrum data, and using these two variables as initial values, the Gaussian model power spectrum components of raindrops are fitted by the nonlinear least squares method, i.e. the preset algorithm, and the broadening of the raindrop power spectrum is obtained , that is, the second initial spectral data, calculate the broadening of the raindrop power spectrum in the target lidar power spectrum of the same range gate The minimum root mean square error σ1 of and the peak position of the raindrop power spectrum The minimum root mean square error σ2 of, set the minimum root mean square error discrimination threshold σr_th of σr, that is, the second root mean square error threshold and The minimum mean square error discrimination threshold fr_th of, that is, the first mean square error threshold. Only when both σ1 < σr_th and σ2 < fr_th are satisfied, it is determined that the spectral separation is successful; otherwise, it is determined that the spectral separation fails and the set of data is deleted. It can be understood that each target radar power spectrum contains an aerosol spectrum and a raindrop spectrum, that is, each radar power spectrum with a bimodal structure has a unique corresponding raindrop spectrum. By using the first initial spectral data and the second initial spectral data to determine the target raindrop spectrum, the reliability of the spectral data of the obtained target raindrop spectrum is ensured.

[0064] Step S14: Construct a corresponding model training set based on the first feature information of the target raindrop spectrum and the corresponding raindrop size label, so as to obtain several model training sets corresponding to different range gates; wherein, the raindrop size label is a label constructed based on the raindrop size collected by a raindrop spectrometer synchronously measured with the preset lidar.

[0065] In this embodiment, the process of constructing a model training set based on the first feature information of the target raindrop spectrum and the corresponding raindrop size label may specifically include: constructing target data including the first feature information of the target raindrop spectrum and the range gate information of the corresponding range gate; labeling the target data with the corresponding raindrop size label to obtain several model training sets corresponding to different range gates, so as to send the target data in the model training set to the input end of the preset raindrop spectrum correction model for training and send the raindrop size label in the model training set to the output end of the preset raindrop spectrum correction model for training during the training of the preset raindrop spectrum correction model, that is, record the five elements in the data of all range gates that meet the spectral separation. The five elements include the raindrop spectrum width , peak position , peak intensity , the laser radar range gate L and the acquisition time T, and the raindrop particle size D collected by the raindrop spectrometer measured synchronously with the laser radar at the same acquisition time T, so as to construct a model data set to train the preset raindrop spectrum correction model. After training the preset raindrop spectrum model, the training set is also divided into a test set and a validation set in proportion, and the model training results are verified by the test set and the validation set. The preset raindrop spectrum correction model is trained by sending the target data to the input end of the preset raindrop spectrum correction model and sending the raindrop particle size label to the output end of the raindrop spectrum model, taking into account the influence of the distance factor on the measured raindrop particle size, and ensuring the reliability of the model training.

[0066] Step S15: respectively train the preset raindrop spectrum correction model using the model training sets corresponding to the range gates to obtain the trained raindrop spectrum correction model, and use the trained raindrop spectrum correction model to correct the raindrop spectrum to be corrected to obtain the corresponding corrected raindrop spectrum.

[0067] In this embodiment, the spectral width of the raindrop spectrum to be corrected , peak position , peak intensity , the corresponding range gate information, that is, the lidar range gate L is input to the trained raindrop spectrum correction model. The trained raindrop spectrum correction model corrects the raindrop spectrum to be corrected by analyzing the above data, and outputs the raindrop particle size corresponding to the raindrop spectrum to be corrected.

[0068] It can be seen that the present application avoids the influence of background noise on the spectral data in the radar power spectrum by performing background noise removal operation on the initial radar power spectrum, thereby ensuring the reliability of the spectral data in the lidar power spectrum after removing the background noise; by judging whether there is a double-peak structure in the radar power spectrum, the reliability of the process of extracting the raindrop spectrum from the radar power spectrum is ensured; by using the first characteristic information of the target raindrop spectrum and the raindrop particle size label to construct a model training set to train the preset raindrop spectrum correction model, the accuracy of the raindrop particle size output by the trained raindrop spectrum correction model is improved.

[0069] See also Figure 5 As shown, the present application also discloses a raindrop spectrum correction model training method, including:

[0070] Step S21, collecting scattered echo signals of a preset laser radar, and calculating the scattered echo signals to obtain laser radar power spectra corresponding to different range gates in the preset laser radar.

[0071] Step S22, judging whether the laser radar power spectrum has a target double-peak structure, and determining whether the laser radar power spectrum is a target radar power spectrum according to the corresponding judgment result.

[0072] Step S23: extracting corresponding target raindrop spectra from the determined target radar power spectra respectively.

[0073] Step S24, constructing a corresponding model training set based on the first characteristic information of the target raindrop spectrum and the corresponding raindrop particle size label to obtain several model training sets corresponding to different range gates respectively; wherein the raindrop particle size label is a label constructed based on the raindrop particle size collected by the raindrop spectrometer measured synchronously with the preset laser radar.

[0074] Step S25: training a preset raindrop spectrum correction model using the model training sets corresponding to the range gates to obtain a trained raindrop spectrum correction model.

[0075] The specific contents of the above steps S21 to S25 can be referred to the above embodiments and will not be described in detail here.

[0076] It can be seen that the present application obtains the target raindrop spectrum from the laser radar power spectrum corresponding to different range gates of the laser radar, and uses the first characteristic information of the target raindrop spectrum and the raindrop particle size label to construct a model training set to train the preset raindrop spectrum correction model, thereby improving the accuracy of the raindrop particle size output by the trained raindrop spectrum correction model.

[0077] See also Figure 6 As shown, the embodiment of the present application also discloses a raindrop spectrum correction device, including:

[0078] A power spectrum acquisition module 11 is used to collect scattered echo signals of a preset laser radar, and calculate the scattered echo signals to obtain laser radar power spectra corresponding to different range gates in the preset laser radar;

[0079] A power spectrum determination module 12 is used to determine whether the laser radar power spectrum has a target double peak structure, and determine whether the laser radar power spectrum is a target radar power spectrum according to the corresponding determination result;

[0080] A raindrop spectrum extraction module 13 is used to extract corresponding target raindrop spectra from the determined target radar power spectra respectively;

[0081] The training set construction module 14 is used to construct a corresponding model training set based on the first characteristic information of the target raindrop spectrum and the corresponding raindrop particle size label, so as to obtain a plurality of model training sets corresponding to different range gates respectively; wherein the raindrop particle size label is a label constructed based on the raindrop particle size collected by the raindrop spectrometer synchronously measured with the preset laser radar;

[0082] The raindrop spectrum correction module 15 is used to train the preset raindrop spectrum correction model using the model training set corresponding to each of the range gates to obtain the trained raindrop spectrum correction model, and use the trained raindrop spectrum correction model to correct the raindrop spectrum to be corrected to obtain the corresponding corrected raindrop spectrum.

[0083] It can be seen that the present application obtains the target raindrop spectrum from the laser radar power spectrum corresponding to different range gates of the laser radar, and uses the first characteristic information of the target raindrop spectrum and the raindrop particle size label to construct a model training set to train the preset raindrop spectrum correction model, thereby improving the accuracy of the raindrop particle size output by the trained raindrop spectrum correction model.

[0084] In some specific embodiments, the power spectrum acquisition module 11 may specifically include:

[0085] An initial spectrum acquisition unit, used for performing a fast Fourier transform on the scattered echo signal light to obtain initial radar power spectra corresponding to different range gates in the preset laser radar;

[0086] The noise removal unit is used to calculate the background noise of the initial radar power spectrum and perform a background noise removal operation on the initial radar power spectrum to obtain the laser radar power spectrum corresponding to the initial radar power spectrum.

[0087] In some specific embodiments, the power spectrum determination module 12 may specifically include:

[0088] A signal-to-noise ratio judgment unit, used to determine the currently traversed laser radar power spectrum as the current laser radar power spectrum, and judge whether the signal-to-noise ratio of the current laser radar power spectrum is greater than a corresponding signal-to-noise ratio threshold value, so as to obtain a corresponding first current judgment result;

[0089] A radar spectrum determination unit, for determining whether the current laser radar power spectrum has the target double-peak structure if the first current judgment result indicates that the signal-to-noise ratio of the current laser radar power spectrum is greater than the corresponding signal-to-noise ratio threshold, and determining whether the current laser radar power spectrum is the target radar power spectrum according to the corresponding second current judgment result, and then traversing the next laser radar power spectrum to obtain a new current laser radar power spectrum;

[0090] A power spectrum traversal unit, configured to traverse the next laser radar power spectrum to obtain a new current laser radar power spectrum if the first current judgment result indicates that the signal-to-noise ratio of the current laser radar power spectrum is not greater than the corresponding signal-to-noise ratio threshold;

[0091] A step jump unit is used to jump again to the step of determining whether the signal-to-noise ratio of the current laser radar power spectrum is greater than the corresponding signal-to-noise ratio threshold until all the laser radar power spectra are traversed.

[0092] In some specific embodiments, the raindrop spectrum extraction module 13 may specifically include:

[0093] a spectrum data acquisition unit, configured to determine first initial spectrum data of an initial raindrop spectrum corresponding to the target radar power spectrum, and acquire second initial spectrum data corresponding to the initial raindrop spectrum using a preset algorithm and the first initial spectrum data;

[0094] A raindrop spectrum determination submodule is used to determine the target raindrop spectrum from the initial raindrop spectrum based on the first initial spectrum data and the second initial spectrum data.

[0095] In some specific embodiments, the raindrop spectrum determination submodule may specifically include:

[0096] an error judging unit, configured to calculate a first root mean square error corresponding to the first initial spectrum data and a second root mean square error corresponding to the second initial spectrum data of the initial raindrop spectrum, and to judge whether the first root mean square error and the second root mean square error are greater than a corresponding first root mean square error threshold and a second root mean square error threshold;

[0097] A raindrop spectrum determining unit is configured to determine the initial raindrop spectrum as the target raindrop spectrum if the first root mean square error is less than the first root mean square error threshold and the second root mean square error is less than the second root mean square error threshold.

[0098] In some specific embodiments, the training set construction module 14 may specifically include:

[0099] A target data construction unit, used to construct target data including first characteristic information of the target raindrop spectrum and corresponding range gate information of the range gate;

[0100] A labeling unit is used to label the target data with corresponding raindrop particle size labels to obtain a plurality of model training sets corresponding to different range gates, so that in the process of training the preset raindrop spectrum correction model, the target data in the model training set is sent to the input end of the preset raindrop spectrum correction model for training, and the raindrop particle size labels in the model training set are sent to the output end of the preset raindrop spectrum correction model for training.

[0101] Furthermore, the present application also discloses an electronic device. Figure 7 This is a structural diagram of an electronic device 20 according to an exemplary embodiment. The content in the diagram cannot be regarded as any limitation on the scope of use of the present application.

[0102] Figure 7The present invention provides a schematic diagram of the structure of an electronic device 20 provided in an embodiment of the present application. The electronic device 20 may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input / output interface 25, and a communication bus 26. The memory 22 is used to store a computer program, which is loaded and executed by the processor 21 to implement the relevant steps in the raindrop spectrum correction method disclosed in any of the aforementioned embodiments. In addition, the electronic device 20 in this embodiment may specifically be an electronic computer.

[0103] In this embodiment, the power supply 23 is used to provide working voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and the external device, and the communication protocol it follows is any communication protocol that can be applied to the technical solution of the present application, and is not specifically limited here; the input and output interface 25 is used to obtain external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs and is not specifically limited here.

[0104] In addition, the memory 22 as a carrier for storing resources may be a read-only memory, a random access memory, a disk or an optical disk, etc. The resources stored thereon may include an operating system 221, a computer program 222, etc., and the storage method may be temporary storage or permanent storage.

[0105] The operating system 221 is used to manage and control the hardware devices and computer program 222 on the electronic device 20, and may be Windows Server, Netware, Unix, Linux, etc. In addition to the computer program that can be used to complete the raindrop spectrum correction method performed by the electronic device 20 disclosed in any of the aforementioned embodiments, the computer program 222 may further include a computer program that can be used to complete other specific tasks.

[0106] Furthermore, the present application also discloses a computer-readable storage medium for storing a computer program; wherein the computer program, when executed by a processor, implements the raindrop spectrum correction method disclosed above. For the specific steps of the method, reference may be made to the corresponding contents disclosed in the above embodiments, and no further description will be given here.

[0107] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0108] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0109] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0110] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0111] The technical solution provided by the present application is introduced in detail above. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea. At the same time, for general technical personnel in this field, according to the idea of ​​the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A raindrop spectrum correction method, characterized in that: include: Collecting scattered echo signals of a preset laser radar, and calculating the scattered echo signals to obtain laser radar power spectra corresponding to different range gates in the preset laser radar; Determine whether the laser radar power spectrum has a target double-peak structure, and determine whether the laser radar power spectrum is a target radar power spectrum according to the corresponding determination result; Extracting corresponding target raindrop spectra from the determined target radar power spectra respectively; A corresponding model training set is constructed based on the first characteristic information of the target raindrop spectrum and the corresponding raindrop particle size label to obtain a plurality of model training sets corresponding to different range gates respectively; wherein the raindrop particle size label is a label constructed based on the raindrop particle size collected by a raindrop spectrometer synchronously measured with the preset laser radar; The preset raindrop spectrum correction model is trained using the model training sets corresponding to the range gates to obtain a trained raindrop spectrum correction model, and the trained raindrop spectrum correction model is used to correct the raindrop spectrum to be corrected to obtain a corresponding corrected raindrop spectrum.

2. The raindrop spectrum correction method according to claim 1, characterized in that: The step of calculating the scattered echo signal to obtain laser radar power spectra corresponding to different range gates in the preset laser radar includes: Performing a fast Fourier transform on the scattered echo signal light to obtain initial radar power spectra corresponding to different range gates in the preset laser radar; The background noise of the initial radar power spectrum is calculated, and a background noise removal operation is performed on the initial radar power spectrum to obtain the laser radar power spectrum corresponding to the initial radar power spectrum.

3. The raindrop spectrum correction method according to claim 1, characterized in that: The determining whether the laser radar power spectrum has a target double peak structure, and determining whether the laser radar power spectrum is a target radar power spectrum according to a corresponding determination result, includes: Determine the currently traversed laser radar power spectrum as the current laser radar power spectrum, and determine whether the signal-to-noise ratio of the current laser radar power spectrum is greater than the corresponding signal-to-noise ratio threshold to obtain a corresponding first current judgment result; If the first current judgment result indicates that the signal-to-noise ratio of the current laser radar power spectrum is greater than the corresponding signal-to-noise ratio threshold, determine whether the current laser radar power spectrum has the target double-peak structure, and determine whether the current laser radar power spectrum is the target radar power spectrum according to the corresponding second current judgment result, and then traverse the next laser radar power spectrum to obtain a new current laser radar power spectrum; If the first current judgment result indicates that the signal-to-noise ratio of the current laser radar power spectrum is not greater than the corresponding signal-to-noise ratio threshold, traverse the next laser radar power spectrum to obtain a new current laser radar power spectrum; Re-jump to the step of determining whether the signal-to-noise ratio of the current laser radar power spectrum is greater than the corresponding signal-to-noise ratio threshold until all the laser radar power spectra are traversed.

4. The raindrop spectrum correction method according to claim 1, characterized in that: The extracting corresponding target raindrop spectra from the determined target radar power spectra respectively includes: Determine first initial spectrum data of an initial raindrop spectrum corresponding to the target radar power spectrum, and obtain second initial spectrum data corresponding to the initial raindrop spectrum through a preset algorithm and the first initial spectrum data; The target raindrop spectrum is determined from the initial raindrop spectrum based on the first initial spectrum data and the second initial spectrum data.

5. The raindrop spectrum correction method according to claim 4, characterized in that: The step of determining the target raindrop spectrum from the initial raindrop spectrum based on the first initial spectrum data and the second initial spectrum data includes: Calculating a first root mean square error corresponding to the first initial spectrum data and a second root mean square error corresponding to the second initial spectrum data of the initial raindrop spectrum, and determining whether the first root mean square error and the second root mean square error are greater than a corresponding first root mean square error threshold and a second root mean square error threshold; If the first root mean square error is less than the first root mean square error threshold, and the second root mean square error is less than the second root mean square error threshold, the initial raindrop spectrum is determined as the target raindrop spectrum.

6. The raindrop spectrum correction method according to claim 1, characterized in that: The corresponding model training set is constructed based on the first characteristic information of the target raindrop spectrum and the corresponding raindrop particle size label to obtain a plurality of model training sets corresponding to different range gates, including: Constructing target data including first characteristic information of the target raindrop spectrum and corresponding range gate information of the range gate; The target data is labeled with corresponding raindrop particle size labels to obtain a plurality of model training sets corresponding to different range gates, so that in the process of training the preset raindrop spectrum correction model, the target data in the model training set is sent to the input end of the preset raindrop spectrum correction model for training, and the raindrop particle size labels in the model training set are sent to the output end of the preset raindrop spectrum correction model for training.

7. A raindrop spectrum correction model training method, characterized in that: include: Collecting scattered echo signals of a preset laser radar, and calculating the scattered echo signals to obtain laser radar power spectra corresponding to different range gates in the preset laser radar; Determine whether the laser radar power spectrum has a target double-peak structure, and determine whether the laser radar power spectrum is a target radar power spectrum according to the corresponding determination result; Extracting corresponding target raindrop spectra from the determined target radar power spectra respectively; A corresponding model training set is constructed based on the first characteristic information of the target raindrop spectrum and the corresponding raindrop particle size label to obtain a plurality of model training sets corresponding to different range gates respectively; wherein the raindrop particle size label is a label constructed based on the raindrop particle size collected by a raindrop spectrometer synchronously measured with the preset laser radar; The preset raindrop spectrum correction model is trained using the model training sets corresponding to the range gates to obtain a trained raindrop spectrum correction model.

8. A raindrop spectrum correction device, characterized in that: include: A power spectrum acquisition module, used for collecting scattered echo signals of a preset laser radar, and calculating the scattered echo signals to obtain laser radar power spectra corresponding to different range gates in the preset laser radar; A power spectrum determination module, used to determine whether the laser radar power spectrum has a target double peak structure, and determine whether the laser radar power spectrum is a target radar power spectrum according to the corresponding judgment result; A raindrop spectrum extraction module, used to extract corresponding target raindrop spectra from the determined target radar power spectra respectively; A training set construction module, used to construct a corresponding model training set based on the first characteristic information of the target raindrop spectrum and the corresponding raindrop particle size label, so as to obtain a plurality of model training sets corresponding to different range gates; wherein the raindrop particle size label is a label constructed based on the raindrop particle size collected by the raindrop spectrometer synchronously measured with the preset laser radar; The raindrop spectrum correction module is used to train the preset raindrop spectrum correction model using the model training set corresponding to each of the range gates to obtain the trained raindrop spectrum correction model, and use the trained raindrop spectrum correction model to correct the raindrop spectrum to be corrected to obtain the corresponding corrected raindrop spectrum.

9. An electronic device, characterized in that: include: Memory, used to store computer programs; A processor, configured to execute the computer program to implement the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: Used to store a computer program, which implements the method according to any one of claims 1 to 7 when executed by a processor.

Images