CN112946652B - Method and device for filling radar beam occlusion area - Google Patents
Method and device for filling radar beam occlusion area Download PDFInfo
- Publication number
- CN112946652B CN112946652B CN202110083213.5A CN202110083213A CN112946652B CN 112946652 B CN112946652 B CN 112946652B CN 202110083213 A CN202110083213 A CN 202110083213A CN 112946652 B CN112946652 B CN 112946652B
- Authority
- CN
- China
- Prior art keywords
- reflectivity factor
- layer
- factor value
- value
- high elevation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S13/00—Systems using the reflection or reradiation of radio waves, e.g. radar systems; Analogous systems using reflection or reradiation of waves whose nature or wavelength is irrelevant or unspecified
- G01S13/88—Radar or analogous systems specially adapted for specific applications
- G01S13/95—Radar or analogous systems specially adapted for specific applications for meteorological use
- G01S13/958—Theoretical aspects
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S7/00—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00
- G01S7/02—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00
- G01S7/41—Details of systems according to groups G01S13/00, G01S15/00, G01S17/00 of systems according to group G01S13/00 using analysis of echo signal for target characterisation; Target signature; Target cross-section
- G01S7/418—Theoretical aspects
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- Computer Networks & Wireless Communication (AREA)
- General Physics & Mathematics (AREA)
- Electromagnetism (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
A method and a device for filling a radar beam occlusion area are provided, wherein the method comprises the following steps: acquiring an original reflectivity factor value of a plurality of layers of the radar; screening the original reflectivity factor values of the multi-layer elevation angles, and removing the original reflectivity factor values of the clouds only existing in the layer to be filled and the original reflectivity factor values of the clouds only existing in the high elevation angle layer to obtain the reflectivity factor values of the layer to be filled and the reflectivity factor values of the high elevation angle layer; training the model based on the reflectivity factor value of the non-shielded area and the reflectivity factor value of the high elevation layer above the non-shielded area to obtain a filling model; and inputting the reflectivity factor value of the high elevation angle layer above the shielding area into the filling model to obtain the reflectivity factor value of the shielding area. The method of the invention can be suitable for Doppler weather radars of various types, and is suitable for filling of various clouds. The method is simple and feasible, has good filling effect, and is more suitable for popularization and localized application.
Description
Technical Field
The invention relates to the field of radar detection, in particular to a method and a device for filling a radar beam occlusion area.
Background
The detection capability of the radar is influenced not only by the radar parameters and the attenuation, refraction and precipitation cloud properties of various cloud particles, but also by tall buildings and terrains around the radar station, and is particularly influenced by the terrains in mountainous areas. Many weather radars of the new generation are located in mountainous areas with complex terrains, and the terrains are shielded to form observation blind areas, so that radar observation values are low or even no observation values exist, and the application effect of the weather radar data of the new generation is seriously influenced.
Some scholars mainly adopt a correction scheme aiming at the problem of radar beam blocking, namely a radar reflectivity factor Vertical Profile (VPR) technology is adopted and precipitation data such as rain gauge data is combined for filling. However, the method in the prior art is only suitable for a large-range uniform precipitation system mainly based on the property of lamellar clouds, and the vertical profile of the reflectivity factor of the lamellar clouds is easy to obtain and has good representativeness, so that the correction effect is good, but the method is not suitable for small-range and non-uniform convective precipitation and has no wide applicability.
Disclosure of Invention
Objects of the invention
The invention aims to provide a method and a device for filling a radar beam occlusion area, which can accurately fill a reflectivity factor value of an occlusion area.
(II) technical scheme
In order to solve the above problem, a first aspect of the present invention provides a method for filling a radar beam occlusion area, including: acquiring an original reflectivity factor value of a multilayer elevation angle of the radar; the multi-layer elevation comprises a layer to be filled and a high elevation layer above the layer to be filled; the layer to be filled comprises a shielding area and a non-shielding area; screening the original reflectivity factor values of the multi-layer elevation angles, and removing the original reflectivity factor values of the clouds only existing in the layer to be filled and the original reflectivity factor values of the clouds only existing in the high elevation angle layer to obtain the reflectivity factor values of the layer to be filled and the reflectivity factor values of the high elevation angle layer; training a model based on the reflectivity factor value of the non-shielded area and the reflectivity factor value of the high elevation layer above the non-shielded area to obtain a filling model; and inputting the reflectivity factor value of the high elevation angle layer above the shielding area into the filling model to obtain the reflectivity factor value of the shielding area.
Optionally, the screening the original values of the reflectivity factors of the multiple layers of the elevation angles to remove the original values of the reflectivity factors of the clouds only existing in the layer to be padded and the original values of the reflectivity factors of the clouds only existing in the high elevation angle layer, so as to obtain the values of the reflectivity factors of the layer to be padded and the values of the reflectivity factors of the high elevation angle layer, includes: obtaining the reflectivity factor values of the layer to be filled and the high elevation angle layer above the layer to be filled based on the relation of the distance library number between the original reflectivity factor value of the layer to be filled and the original reflectivity factor value of the corresponding high elevation angle layer; assigning values smaller than a first preset value and default values in the original reflectivity factor values of the multilayer elevation angle as second preset values; if the original reflectivity factor values of the layers to be padded are the second preset value and the original reflectivity factor values of the overhead high elevation angle layers are not all the second preset values, determining that the clouds corresponding to the original reflectivity factor values of the overhead high elevation angle layers which are not all the second preset values are the clouds only existing in the high elevation angle layers; and if the original reflectivity factor value of the layer to be filled is greater than the second preset value and the original reflectivity factor values of the high elevation angle layers above the layer to be filled are all the second preset values, judging that the cloud in the layer to be filled, of which the original reflectivity factor value is greater than the second preset value, is the cloud only existing in the layer to be filled.
Optionally, the training a model based on the reflectivity factor value of the non-occlusion region and the reflectivity factor value of the high elevation layer above the non-occlusion region to obtain a filling model includes: dividing the layer to be filled into preset distance sections according to the number of preset distance bins; respectively outputting the reflectivity factor value of a non-shielded area in the preset distance segment and the reflectivity factor value of the high elevation angle layer above the non-shielded area; a network architecture is established for each distance segment.
Optionally, the establishing a network architecture for each distance segment includes: acquiring a sequential model; taking the reflectivity factor value of the high elevation angle layer above the non-occlusion area in each distance segment as an input factor input layer, and selecting a relu function as an activation function; setting the number of hidden layers and the number of nodes, and selecting a relu function as an activation function; and setting an output layer.
Optionally, the training a model based on the reflectivity factor value of the non-occlusion region and the reflectivity factor value of the high elevation layer above the non-occlusion region to obtain a filling model includes: training a network architecture by taking a part of the reflectivity factor values of the non-shielded areas and a part of the reflectivity factor values of the high elevation layer above the non-shielded areas as a training set; and testing the network architecture by taking the reflectivity factor value of the other part of the non-blocked area and the reflectivity factor value of the high elevation layer above one part of the non-blocked area as a test set.
Optionally, the method for filling the radar beam occlusion area further includes: obtaining a predicted value of the reflectivity factor value of the non-shielded area based on the reflectivity factor value of the high elevation angle layer above the non-shielded area and a filling model; and evaluating the model effect based on the predicted value and the corresponding reflectivity factor value of the non-occlusion area.
Optionally, the evaluating the model effect based on the predicted value and the corresponding reflectivity factor value of the non-occlusion region includes: calculating a root mean square error, an average absolute error, and/or an average relative error between the prediction value and the reflectance factor value of the non-occluded area.
Optionally, the inputting the reflectivity factor value of the high elevation layer above the occlusion region into the padding model to obtain the reflectivity factor value of the occlusion region includes: and inputting the reflectivity factor value of the high elevation angle layer in the corresponding distance section above the shielding area in each distance section of the layer to be filled into the filling model to obtain the reflectivity factor value of the shielding area in each distance section.
Optionally, the first preset value is-5 dBZ; the second preset value is-20 dBZ.
A second aspect of the present invention provides an apparatus for filling a radar beam occlusion region, including: the acquisition module is used for acquiring the original reflectivity factor value of multiple layers of the radar; the original reflectivity factor values of the multiple layers comprise the original reflectivity factor value of the layer to be filled and the original reflectivity factor value of the high elevation angle layer above the layer to be filled; the layer to be filled comprises a shielding area and a non-shielding area; the screening module is used for screening the original reflectivity factor values of the multi-layer elevation angles, removing the original reflectivity factor values of the clouds only existing in the layer to be filled and the original reflectivity factor values of the clouds only existing in the high elevation angle layer, and obtaining the reflectivity factor values of the layer to be filled and the reflectivity factor values of the high elevation angle layer; the model training module is used for training a model based on the reflectivity factor value of the non-shielded area and the reflectivity factor value of the high elevation layer above the non-shielded area to obtain a filling model; and the filling module is used for inputting the reflectivity factor value of the high elevation angle layer above the shielding area into the filling model to obtain the reflectivity factor value of the shielding area.
(III) advantageous effects
The technical scheme of the invention has the following beneficial technical effects:
training a model based on the reflectivity factor value of a non-shielded area and the reflectivity factor value of a high elevation layer above the non-shielded area to obtain a filling model; and inputting the reflectivity factor value of the high elevation angle layer above the shielding area into the filling model to obtain the reflectivity factor value of the shielding area. The method can be suitable for Doppler weather radars of various models, and is suitable for filling of various clouds. The method is simple and feasible, has good filling effect, and is more suitable for popularization and localized application.
Drawings
Fig. 1 is a flowchart of a method of filling a radar beam occlusion region according to a first embodiment of the present invention;
FIG. 2 is a diagram of distance bin number relationships for each elevation angle of the radar;
fig. 3 is a schematic structural diagram of an apparatus for filling a radar beam occlusion region according to a second embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail with reference to the accompanying drawings in conjunction with the following detailed description. It should be understood that the description is intended to be exemplary only, and is not intended to limit the scope of the present invention. Moreover, in the following description, descriptions of well-known structures and techniques are omitted so as to not unnecessarily obscure the concepts of the present invention.
In the description of the present invention, it should be noted that the terms "first", "second", and "third" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance.
In addition, the technical features involved in the different embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other.
First embodiment
As shown in fig. 1, the present embodiment provides a method for filling a radar beam occlusion area, including: acquiring an original reflectivity factor value of a plurality of layers of the radar; the original reflectivity factor values of the multiple layers comprise the original reflectivity factor value of the layer to be filled and the original reflectivity factor value of the high elevation angle layer above the layer to be filled; the layer to be filled comprises a shielding area and a non-shielding area; screening the original reflectivity factor values of the multi-layer elevation angles, and removing the original reflectivity factor values of the clouds only existing in the layer to be filled and the original reflectivity factor values of the clouds only existing in the high elevation angle layer to obtain the reflectivity factor values of the layer to be filled and the reflectivity factor values of the high elevation angle layer; training a model based on the reflectivity factor value of the non-shielded area and the reflectivity factor value of the high elevation layer above the non-shielded area to obtain a filling model; and inputting the reflectivity factor value of the high elevation angle layer above the shielding area into the filling model to obtain the reflectivity factor value of the shielding area. In the embodiment, a filling model is obtained by training a model based on the reflectivity factor value of a non-shielded area and the reflectivity factor value of a high elevation layer above the non-shielded area; and inputting the reflectivity factor value of the high elevation angle layer above the shielding area into the filling model to obtain the reflectivity factor value of the shielding area. The method can be suitable for Doppler weather radars of various models, and is suitable for filling of various clouds. The method is simple and feasible, has good filling effect, and is more suitable for popularization and localized application. Wherein the layer to be filled is a low elevation layer, and the specific elevation angle can be set according to the requirement.
In an optional embodiment, the screening the original values of the reflectivity factors of the multi-layer elevation angles to remove the original values of the reflectivity factors of the clouds only existing in the layer to be padded and the original values of the reflectivity factors of the clouds only existing in the high elevation angle layer, so as to obtain the values of the reflectivity factors of the layer to be padded and the values of the reflectivity factors of the high elevation angle layer, includes: obtaining the reflectivity factor values of the layer to be filled and the high elevation angle layer above the layer to be filled based on the relation of the distance library number between the original reflectivity factor value of the layer to be filled and the original reflectivity factor value of the corresponding high elevation angle layer; assigning the original reflectivity factor values of the multilayer elevation angle which are smaller than the first preset value and the default values as second preset values; if the original reflectivity factor values of the layers to be padded are the second preset value and the original reflectivity factor values of the overhead high elevation angle layers are not all the second preset values, determining that the clouds corresponding to the original reflectivity factor values of the overhead high elevation angle layers which are not all the second preset values are the clouds only existing in the high elevation angle layers; if the original reflectivity factor value of the layer to be filled is larger than the second preset value and the original reflectivity factor values of the high elevation angle layers above the layer to be filled are both the second preset value, the cloud in the layer to be filled, of which the original reflectivity factor value is larger than the second preset value, is judged to be the cloud only existing in the layer to be filled.
For example, referring to fig. 2, the elevation angle of the layer to be padded is 0.5 °, and the elevation angle of the higher layer is β. Wherein β may be 1.45 °, 2.40 °, 3.35 °, 4.30 °, 6.0 °, 9.90 °, 14.6 °, or any other value that may be set as desired. Wherein, the point O is the position of the weather radar, and if the influence of the curvature of the earth is neglected, for a certain distance bank a with an elevation angle of 0.5 degrees, the corresponding distance bank B with a high elevation angle has the following relation:
G β cosβ＝G 0.50° cos0.50°
wherein G is 0.50° Number of range bins at 0.5 degree elevation angle, G β Is the high elevation distance library number corresponding to the elevation angle of 0.5 degrees, and beta is the elevation angle degree of the high elevation angle. The reflectivity factor values corresponding to the elevation angle of 0.5 degree and the elevation angles above the elevation angle can be output through the relational expression.
The distance library with high elevation angle is obtained by only corresponding one reflectivity factor value in one distance library, and the reflectivity factor value corresponding to the distance library can be correspondingly output from a program.
Then filtering, firstly, assigning a reflectivity factor value smaller than-5 dBZ and a default value to be a lower dBZ value (for example, -20 dBZ), if the reflectivity factor value of the 0.5-degree elevation angle is-20 dBZ and the reflectivity factor values of the corresponding other elevation angle layers are not all-20 dBZ, judging that the cloud exists only in the high elevation angle, and not outputting the distance library as a training data set; if the reflectivity factor value of the 0.5 degree elevation angle is larger than-20 dBZ and the reflectivity factor values of the corresponding other elevation angle layers are all-20 dBZ, the thin cloud only existing in the 0.5 degree elevation angle layer is judged to exist, and the distance library is not output as the training data set.
In an optional embodiment, the training the model based on the reflectivity factor value of the non-occlusion region and the reflectivity factor value of the high elevation layer above the non-occlusion region to obtain a filled model includes: dividing the layer to be filled into preset distance sections according to the number of preset distance bins; respectively outputting the reflectivity factor value of a non-shielded area in the preset distance segment and the reflectivity factor value of the high elevation angle layer above the non-shielded area; a network architecture is established for each distance segment. Wherein, the establishing a network architecture for each distance segment includes: acquiring a sequential model; taking the reflectivity factor value of the high elevation angle layer above the non-occlusion area in each distance segment as an input factor input layer, and selecting a relu function as an activation function; setting the number of hidden layers and the number of nodes, and selecting a relu function as an activation function; and setting an output layer. Optionally, the training a model based on the reflectivity factor value of the non-occlusion region and the reflectivity factor value of the high elevation layer above the non-occlusion region to obtain a filling model includes: training a network architecture by taking a part of the reflectivity factor values of the non-shielded areas and a part of the reflectivity factor values of the high elevation layer above the non-shielded areas as a training set; and testing the network architecture by taking the reflectivity factor value of the other part of the non-blocked area and the reflectivity factor value of the high elevation layer above one part of the non-blocked area as a test set.
For example, considering the difference between the high elevation angle and the 0.5 ° projection position, i.e., the accuracy of the radar becomes lower and lower, the farther from the radar, the radar 0.5 ° elevation angle is divided into six range segments of 1-25km,25-50km,50-75km,75-100km,100-125km,125-150km by the number of range bins.
According to the divided six distance segments, 0.5-degree elevation angle reflectivity values of the non-shielding areas are output as label values or real values, and 9 (3*3) reflectivity factor values of 7, 6, 5, 4, 3 and 2 elevation angles of the high elevation angle layers are correspondingly output as input factors.
And (3) establishing an echo filling network architecture for each of the six distance segments: specifically, a deep learning algorithm is used, firstly, the model is selected as a sequential model, then a network layer is constructed, for six distance sections of 1-25km,25-50km,50-75km,75-100km,100-125km and 125-150km, the number of input layer nodes of each distance section model is respectively determined to be 63, 54, 45, 36, 27 and 18 according to the number of input factors selected by each distance section, and a relu function is selected as an activation function of the input layer. The number of the hidden layers set by the invention is three, the number of the nodes of the hidden layers is respectively 256, 128 and 64, and the activation function is also a relu function. For the output layer, since this is a regression fitting problem, the number of output layer nodes is 1, and no activation function is set. And then compiling the model, wherein the optimizer can be a gradient descent optimizer (SGD), the loss function is a mean square error, and the expression is as follows:
wherein y is i In order to predict the value of the target,
is the true value, i.e., the tag value. And then, training 70% of the data set as a training set, testing 30% of the data set as a test set, and finally outputting the 0.5-degree elevation angle reflectivity factor estimated value of the test set, namely obtaining the reflectivity factor estimated value of the 0.5-degree elevation angle of the non-shielding area, thereby completing the filling process.
In an optional embodiment, the method for filling a radar beam occlusion region further includes: obtaining a predicted value of the reflectivity factor value of the non-shielded area based on the reflectivity factor value of the high elevation angle layer above the non-shielded area and a filling model; and evaluating the model effect based on the predicted value and the corresponding reflectivity factor value of the non-occlusion area. Wherein the evaluating a model effect based on the predicted value and the corresponding reflectivity factor value of the non-occluded area comprises: calculating a root mean square error, an average absolute error, and/or an average relative error between the prediction value and the reflectance factor value of the non-occluded area. The concrete formula is
Where RMSE is the root mean square error, MAE is the mean absolute error, and MRE is the Mean Relative Error (MRE).
In an optional embodiment, the inputting the reflectivity factor value of the high elevation layer above the occlusion region into the padding model to obtain the reflectivity factor value of the occlusion region includes: and inputting the reflectivity factor value of the high elevation angle layer in the corresponding distance section above the shielding area in each distance section of the layer to be filled into the filling model to obtain the reflectivity factor value of the shielding area in each distance section.
For example, the sorted input factors of the occlusion region are input into the trained deep learning model according to the divided six distance segments 1-25km,25-50km,50-75km,75-100km,100-125km and 125-150km, and the reflectivity factor values of 7 (3*3) of high elevation layers corresponding to the 0.5 ° elevation angle of the occlusion region are output in a segmented manner as input factors, so that the filling value of the 0.5 ° beam occlusion region is obtained, and the filling is completed.
Second embodiment
The present embodiment provides a device for filling a radar beam blocking area, which is used for the device for filling a radar beam blocking area according to the first embodiment of the present invention, and includes: the acquisition module is used for acquiring the original reflectivity factor values of multiple layers of the radar; the original reflectivity factor values of the multiple layers comprise the original reflectivity factor value of the layer to be filled and the original reflectivity factor value of the high elevation angle layer above the layer to be filled; the layer to be filled comprises a shielding area and a non-shielding area; the screening module is used for screening the original reflectivity factor values of the multilayer elevation angles, removing the original reflectivity factor values of the clouds only existing in the layer to be filled and the original reflectivity factor values of the clouds only existing in the high elevation angle layer, and obtaining the reflectivity factor values of the layer to be filled and the reflectivity factor values of the high elevation angle layer; the model training module is used for training a model based on the reflectivity factor value of the non-shielded area and the reflectivity factor value of the high elevation layer above the non-shielded area to obtain a filling model; and the filling module is used for inputting the reflectivity factor value of the high elevation angle layer above the shielding area into the filling model to obtain the reflectivity factor value of the shielding area. In the embodiment, a filling model is obtained by training a model based on the reflectivity factor value of a non-shielded area and the reflectivity factor value of a high elevation layer above the non-shielded area; and inputting the reflectivity factor value of the high elevation angle layer above the shielding area into the filling model to obtain the reflectivity factor value of the shielding area. The method can be suitable for Doppler weather radars of various models, and is suitable for filling of various clouds. The method is simple and feasible, has good filling effect, and is more suitable for popularization and localized application. Wherein the layer to be filled is a low elevation layer, and the angle of the specific elevation angle can be set according to the requirement.
The same parts of this embodiment as those of the first embodiment will not be described herein.
It is to be understood that the above-described embodiments of the present invention are merely illustrative of or explaining the principles of the invention and are not to be construed as limiting the invention. Therefore, any modification, equivalent replacement, improvement and the like made without departing from the spirit and scope of the present invention should be included in the protection scope of the present invention. Further, it is intended that the appended claims cover all such variations and modifications as fall within the scope and boundaries of the appended claims or the equivalents of such scope and boundaries.
Claims (7)
1. A method of filling a radar beam occlusion region, comprising:
acquiring an original reflectivity factor value of a plurality of layers of the radar;
the original reflectivity factor values of the multiple layers comprise the original reflectivity factor value of the layer to be filled and the original reflectivity factor value of the high elevation angle layer above the layer to be filled; the layer to be filled comprises a shielding area and a non-shielding area;
screening the original reflectivity factor values of the multiple layers, and removing the original reflectivity factor value of the cloud only existing in the layer to be filled and the original reflectivity factor value of the cloud only existing in the high elevation angle layer to obtain the reflectivity factor value of the layer to be filled and the reflectivity factor value of the high elevation angle layer;
training a model based on the reflectivity factor value of the non-shielded area and the reflectivity factor value of the high elevation layer above the non-shielded area to obtain a filling model;
inputting the reflectivity factor value of the high elevation angle layer above the shielding area into the filling model to obtain the reflectivity factor value of the shielding area;
wherein, the screening the original reflectivity factor values of the multiple layers to remove the original reflectivity factor value of the cloud only existing in the layer to be padded and the original reflectivity factor value of the cloud only existing in the high elevation angle layer, so as to obtain the reflectivity factor value of the layer to be padded and the reflectivity factor value of the high elevation angle layer, includes: obtaining the reflectivity factor values of the layer to be filled and the high elevation angle layer above the layer to be filled based on the relation of the distance library number between the original reflectivity factor value of the layer to be filled and the original reflectivity factor value of the corresponding high elevation angle layer; assigning both a value smaller than a first preset value and a missing value among the original reflectance factor values of the plurality of layers to a second preset value; if the original reflectivity factor values of the layers to be padded are the second preset value and the original reflectivity factor values of the overhead high elevation angle layers are not all the second preset values, determining that the clouds corresponding to the original reflectivity factor values of the overhead high elevation angle layers which are not all the second preset values are the clouds only existing in the high elevation angle layers; if the original reflectivity factor value of the layer to be filled is larger than the second preset value and the original reflectivity factor values of the high elevation angle layers above the layer to be filled are both the second preset value, judging that the cloud in the layer to be filled, of which the original reflectivity factor value is larger than the second preset value, is the cloud only existing in the layer to be filled;
training a model based on the reflectivity factor value of the non-shielded area and the reflectivity factor value of the high elevation layer above the non-shielded area to obtain a filling model, wherein the training comprises the following steps: dividing the layer to be filled into preset distance sections according to the number of preset distance bins; respectively outputting the reflectivity factor value of a non-shielded area in the preset distance segment and the reflectivity factor value of the high elevation angle layer above the non-shielded area; establishing a network architecture for each distance segment;
inputting the reflectivity factor value of the high elevation angle layer above the shielding area into the filling model to obtain the reflectivity factor value of the shielding area, wherein the method comprises the following steps: and inputting the reflectivity factor value of the high elevation angle layer in the corresponding distance section above the shielding area in each distance section of the layer to be filled into the filling model to obtain the reflectivity factor value of the shielding area in each distance section.
2. The method of filling radar beam occlusion areas of claim 1, wherein establishing a network architecture for each range segment comprises:
acquiring a sequential model;
taking the reflectivity factor value of the high elevation angle layer above the non-occlusion area in each distance segment as an input factor input layer, and selecting a relu function as an activation function;
setting the number of hidden layers and the number of nodes, and selecting a relu function as an activation function;
and setting an output layer.
3. The method of claim 1, wherein training a model based on the reflectivity factor value of the non-occluded area and the reflectivity factor value of the high elevation layer above the non-occluded area to obtain a filled model comprises:
training a network architecture by taking a part of the reflectivity factor values of the non-shielded areas and a part of the reflectivity factor values of the high elevation layer above the non-shielded areas as a training set;
and testing the network architecture by taking the reflectivity factor value of the other part of the non-blocked area and the reflectivity factor value of the high elevation layer above one part of the non-blocked area as a test set.
4. The method of filling radar beam occlusion regions of claim 3, further comprising:
obtaining a predicted value of the reflectivity factor value of the non-shielded area based on the reflectivity factor value of the high elevation angle layer above the non-shielded area and a filling model;
and evaluating the model effect based on the predicted value and the corresponding reflectivity factor value of the non-occlusion area.
5. The method of filling in radar beam occlusion regions according to claim 4, wherein the evaluating a model effect based on the predicted values and the corresponding reflectivity factor values of the non-occlusion regions comprises:
calculating a root mean square error, an average absolute error, and/or an average relative error between the prediction value and the reflectance factor value of the non-occluded area.
6. The method of filling radar beam occlusion regions of claim 1,
the first preset value is-5 dBZ;
the second preset value is-20 dBZ.
7. An apparatus for filling a radar beam occlusion region, comprising:
the acquisition module is used for acquiring the original reflectivity factor values of multiple layers of the radar; the original reflectivity factor values of the multiple layers comprise the original reflectivity factor value of the layer to be filled and the original reflectivity factor value of the high elevation angle layer above the layer to be filled; the layer to be filled comprises a shielding area and a non-shielding area;
the screening module is used for screening the original reflectivity factor values of the multiple layers, removing the original reflectivity factor value of the cloud only existing in the layer to be filled and the original reflectivity factor value of the cloud only existing in the high elevation angle layer, and obtaining the reflectivity factor value of the layer to be filled and the reflectivity factor value of the high elevation angle layer;
the model training module is used for training a model based on the reflectivity factor value of the non-shielded area and the reflectivity factor value of the high elevation layer above the non-shielded area to obtain a filling model;
the filling module is used for inputting the reflectivity factor value of the high elevation angle layer above the shielding area into the filling model to obtain the reflectivity factor value of the shielding area;
wherein, the screening the original reflectivity factor values of the multiple layers to remove the original reflectivity factor value of the cloud only existing in the layer to be padded and the original reflectivity factor value of the cloud only existing in the high elevation angle layer, so as to obtain the reflectivity factor value of the layer to be padded and the reflectivity factor value of the high elevation angle layer, includes: obtaining the reflectivity factor values of the layer to be filled and the high elevation angle layer above the layer to be filled based on the relation of the distance library number between the original reflectivity factor value of the layer to be filled and the original reflectivity factor value of the corresponding high elevation angle layer; assigning both a value smaller than a first preset value and a missing value among the original reflectance factor values of the plurality of layers to a second preset value; if the original reflectivity factor values of the layers to be padded are the second preset value and the original reflectivity factor values of the overhead high elevation angle layers are not all the second preset values, determining that the clouds corresponding to the original reflectivity factor values of the overhead high elevation angle layers which are not all the second preset values are the clouds only existing in the high elevation angle layers; if the original reflectivity factor value of the layer to be filled is larger than the second preset value and the original reflectivity factor values of the high elevation angle layers above the layer to be filled are both the second preset value, judging that the cloud in the layer to be filled, of which the original reflectivity factor value is larger than the second preset value, is the cloud only existing in the layer to be filled;
training a model based on the reflectivity factor value of the non-shielded area and the reflectivity factor value of the high elevation layer above the non-shielded area to obtain a filling model, wherein the training comprises the following steps: dividing the layer to be filled into preset distance sections according to the number of preset distance bins; respectively outputting the reflectivity factor value of a non-shielded area in the preset distance segment and the reflectivity factor value of the high elevation angle layer above the non-shielded area; establishing a network architecture for each distance segment;
inputting the reflectivity factor value of the high elevation angle layer above the shielding area into the filling model to obtain the reflectivity factor value of the shielding area, wherein the method comprises the following steps: and inputting the reflectivity factor value of the high elevation angle layer in the corresponding distance section above the shielding area in each distance section of the layer to be filled into the filling model to obtain the reflectivity factor value of the shielding area in each distance section.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110083213.5A CN112946652B (en) | 2021-01-21 | 2021-01-21 | Method and device for filling radar beam occlusion area |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110083213.5A CN112946652B (en) | 2021-01-21 | 2021-01-21 | Method and device for filling radar beam occlusion area |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112946652A CN112946652A (en) | 2021-06-11 |
| CN112946652B true CN112946652B (en) | 2023-03-28 |
Family
ID=76235780
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110083213.5A Active CN112946652B (en) | 2021-01-21 | 2021-01-21 | Method and device for filling radar beam occlusion area |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112946652B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101291980B1 (en) * | 2012-12-20 | 2013-08-09 | 경북대학교 산학협력단 | Method for making total quality index for radar reflectivity measurement |
| CN105260591A (en) * | 2015-09-18 | 2016-01-20 | 江苏省气象科学研究所 | Multi-elevation MIGFA gust front recognition improved algorithm |
| CN106054194A (en) * | 2016-05-10 | 2016-10-26 | 南京信息工程大学 | Spaceborne radar and ground-based radar reflectivity factor data three dimensional fusion method |
| CN109597038A (en) * | 2018-12-12 | 2019-04-09 | 北京无线电测量研究所 | A kind of the radar reflectivity factor correction method and system of dual polarization radar |
| CN110018479A (en) * | 2019-04-28 | 2019-07-16 | 中国气象局广州热带海洋气象研究所 | C-band dual-polarization weather radar reflectivity terrain shading decaying correction method |
| CN110095760A (en) * | 2019-04-24 | 2019-08-06 | 中国气象局气象探测中心 | A kind of test device and method of weather radar |
| CN110967695A (en) * | 2019-10-28 | 2020-04-07 | 兰州大方电子有限责任公司 | Radar echo extrapolation short-term prediction method based on deep learning |
-
2021
- 2021-01-21 CN CN202110083213.5A patent/CN112946652B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR101291980B1 (en) * | 2012-12-20 | 2013-08-09 | 경북대학교 산학협력단 | Method for making total quality index for radar reflectivity measurement |
| CN105260591A (en) * | 2015-09-18 | 2016-01-20 | 江苏省气象科学研究所 | Multi-elevation MIGFA gust front recognition improved algorithm |
| CN106054194A (en) * | 2016-05-10 | 2016-10-26 | 南京信息工程大学 | Spaceborne radar and ground-based radar reflectivity factor data three dimensional fusion method |
| CN109597038A (en) * | 2018-12-12 | 2019-04-09 | 北京无线电测量研究所 | A kind of the radar reflectivity factor correction method and system of dual polarization radar |
| CN110095760A (en) * | 2019-04-24 | 2019-08-06 | 中国气象局气象探测中心 | A kind of test device and method of weather radar |
| CN110018479A (en) * | 2019-04-28 | 2019-07-16 | 中国气象局广州热带海洋气象研究所 | C-band dual-polarization weather radar reflectivity terrain shading decaying correction method |
| CN110967695A (en) * | 2019-10-28 | 2020-04-07 | 兰州大方电子有限责任公司 | Radar echo extrapolation short-term prediction method based on deep learning |
Non-Patent Citations (1)
| Title |
|---|
| 双线偏振多普勒雷达资料质量的定量评估;杜牧云等;《气象学报》;20130815;第71卷(第04期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112946652A (en) | 2021-06-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3534187B1 (en) | Process and system for hydrological analysis and management for river basins | |
| CN107918165B (en) | More satellites based on space interpolation merge Prediction of Precipitation method and system | |
| JP2020193962A (en) | Method and apparatus for predicting severe convection weather | |
| CN107918166B (en) | More satellite fusion precipitation methods and system | |
| CN112632311B (en) | Cloud layer change trend prediction method based on deep learning | |
| CN112446559B (en) | Large-range ground subsidence space-time prediction method and system based on deep learning | |
| CN107507417B (en) | A kind of smartway partitioning method and device based on microwave radar echo-signal | |
| CN110221360A (en) | A kind of power circuit thunderstorm method for early warning and system | |
| JP2010096593A (en) | Wind prediction device and program | |
| KR101291980B1 (en) | Method for making total quality index for radar reflectivity measurement | |
| CN116776592A (en) | Wind deflection analysis method and system for power transmission line | |
| CN110261857A (en) | A kind of weather radar spatial interpolation methods | |
| KR102454007B1 (en) | Apparatus and method for calculating solar insolation quantity based on artificial intelligence using artificial satellite observation data | |
| CN112149294B (en) | Elastic weather grid design method | |
| CN115270527B (en) | Real-time assessment method, equipment and storage medium for road collapse risk | |
| CN108931774A (en) | Convective precipitation based on lightning data identifies examination and test of products method and system | |
| CN102879774A (en) | Method and apparatus for synthesizing short flight paths | |
| CN112946652B (en) | Method and device for filling radar beam occlusion area | |
| Wang et al. | Simulating tropical cyclone waves in the East China Sea with an event-based, parametric-adjusted model | |
| CN114492984A (en) | Method, device, equipment and storage medium for predicting time-space distribution of dust concentration | |
| CN110824478B (en) | Automatic classification method and device for precipitation cloud types based on diversified 3D radar echo characteristics | |
| CN109905839A (en) | Hazardous chemical stacking formula warehoused cargo positioning and optimizing method | |
| CN109934151B (en) | Face detection method based on movidius computing chip and Yolo face | |
| CN114493245A (en) | Mountain torrent disaster easiness evaluation method based on GIS and integrated learning | |
| CN115424426A (en) | Method for improving accuracy of regional geological disaster early warning and forecasting |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |