CN117129963A - Radar echo extrapolation method based on migration prediction and readable storage medium thereof - Google Patents
Radar echo extrapolation method based on migration prediction and readable storage medium thereof Download PDFInfo
- Publication number
- CN117129963A CN117129963A CN202311299982.4A CN202311299982A CN117129963A CN 117129963 A CN117129963 A CN 117129963A CN 202311299982 A CN202311299982 A CN 202311299982A CN 117129963 A CN117129963 A CN 117129963A
- Authority
- CN
- China
- Prior art keywords
- prediction model
- data
- model
- domain
- data set
- 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.)
- Pending
Links
- 238000013508 migration Methods 0.000 title claims abstract description 55
- 230000005012 migration Effects 0.000 title claims abstract description 55
- 238000000034 method Methods 0.000 title claims abstract description 32
- 238000013213 extrapolation Methods 0.000 title claims abstract description 24
- 238000002310 reflectometry Methods 0.000 claims abstract description 26
- 238000012360 testing method Methods 0.000 claims abstract description 26
- 238000012549 training Methods 0.000 claims abstract description 26
- 238000011156 evaluation Methods 0.000 claims abstract description 16
- 238000013210 evaluation model Methods 0.000 claims abstract description 15
- 238000013526 transfer learning Methods 0.000 claims description 7
- 238000004590 computer program Methods 0.000 claims description 4
- 238000007781 pre-processing Methods 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 3
- 230000006870 function Effects 0.000 claims description 3
- 238000004422 calculation algorithm Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 4
- 230000003287 optical effect Effects 0.000 description 4
- 238000013135 deep learning Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000001556 precipitation Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
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
- 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
- G01S7/411—Identification of targets based on measurements of radar reflectivity
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Computer Networks & Wireless Communication (AREA)
- Electromagnetism (AREA)
- Geometry (AREA)
- Computer Hardware Design (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
According to the migration prediction-based radar echo extrapolation method, radar reflectivity data are obtained, and radar reflectivity data are preprocessed to obtain radar reflectivity effective data; determining a target domain data set and a source domain data set according to the radar reflectivity effective data; constructing a domain migration prediction model, inputting a source domain data set into the migration prediction model, and training the migration prediction model; constructing an evaluation model of a domain migration prediction model, inputting test set data in a target domain data set into the domain migration prediction model trained in the step S3 to obtain a test result, calculating an evaluation index of the domain migration prediction model by the evaluation model, and selecting the domain migration prediction model corresponding to the maximum evaluation index as an optimal prediction model; inputting a training set in the target domain data set into an optimal prediction model, training the optimal prediction model, inputting a test set in the target domain data set into the trained optimal prediction model again to obtain a test result, calculating an evaluation index of the trained optimal prediction model by an evaluation model, and selecting the optimal prediction model corresponding to the maximum evaluation index value as a radar echo extrapolation model.
Description
Technical Field
The present invention relates to a radar echo extrapolation method, and more particularly, to a radar echo extrapolation method based on migration prediction and a readable storage medium thereof.
Background
The radar echo extrapolation is widely applied to weather prediction, and the radar extrapolation technology at present mainly comprises a single centroid method, a cross correlation method, an optical flow method and the like, but the nonlinear evolution process of a medium-small scale atmosphere system is difficult to master in the actual situation by the existing methods, and the adaptability is poor.
With the development of technology, a deep learning algorithm is gradually proposed to carry out proximity prediction based on historical radar observation data, echo evolution characteristics of a convection system are automatically learned, radar echo extrapolation prediction is realized, prediction results are generally superior to traditional radar extrapolation technologies and numerical prediction results, and particularly, prediction of a weaker echo region lower than 40dBZ is remarkably improved. However, when the prediction time becomes long, the existing deep learning algorithm can generate obvious smoothing phenomenon, and the prediction effect of a strong echo zone can be rapidly reduced, so that the effective judgment of strong convection weather is not facilitated. Moreover, existing deep learning algorithms must have sufficient historical data to ensure accuracy of the predicted results, but in areas or scenarios where precipitation is low, radar data quality is poor, and radar site setup time is short, it is difficult to obtain a sufficient set of valid data, and then it is difficult for existing algorithms to ensure final accuracy, and radar echo data is a significant long tail distribution, which makes the sample size of the echo in a given region even less, resulting in reduced prediction accuracy.
Therefore, in order to solve the above-mentioned technical problems, a new technical means is needed.
Disclosure of Invention
In view of the above, an object of the present invention is to provide a radar echo extrapolation method based on migration prediction and a readable storage medium thereof, which are based on radar reflectivity as training data and on migration prediction models and determined target domain data sets and source domain data sets, so that the migration prediction models can obtain enough radar echo characteristics under the condition of fewer samples, thereby effectively improving accuracy of radar extrapolation prediction, and effectively improving prediction effect of strong convection echo, and having strong adaptability.
The invention provides a radar echo extrapolation method based on migration prediction, which comprises the following steps:
s1, acquiring radar reflectivity data, and preprocessing the radar reflectivity data to obtain radar reflectivity effective data;
s2, determining a target domain data set and a source domain data set according to radar reflectivity effective data;
s3, constructing a domain migration prediction model f, inputting a source domain data set into the migration prediction model, and training the migration prediction model;
s4, constructing an evaluation model of the domain migration prediction model, inputting test set data in the target domain data set into the domain migration prediction model trained in the step S3 to obtain a test result, calculating an evaluation index of the domain migration prediction model by the evaluation model, and selecting the domain migration prediction model corresponding to the maximum evaluation index as an optimal prediction model f s ;
S5, inputting a training set in the target domain data set into an optimal prediction model f s And for the optimal prediction model f s Training, inputting a test set in the target domain data set into the trained optimal prediction model to obtain a test result, and calculating a trained optimal prediction model f by an evaluation model s And selecting the optimal prediction model f corresponding to the maximum evaluation index value s As radar echo extrapolation model f T 。
Further, the domain migration prediction model is either ConvLSTM, predRNN or TrajGRU model.
Further, the Loss function of the domain transfer learning model is Loss:
wherein: n is the number of samples, J is the total number of predictions,representing t 0 Observation data at +t moment, +.>Representing t 0 Predicted data at +t, t 0 Is the last point in time of observation.
Further, determining the target domain data set and the source domain data set in step S2 specifically includes:
determining the space position and the time range of a target area according to task requirements, taking radar reflectivity effective data in the space position and the time range as a target domain data set, and dividing the target domain data set into training set data and test set data, wherein the time of the training set data and the time of the test set data are not overlapped in a crossing way;
and selecting an area which is outside the target area and has the same spatial size as the target area as a source area, wherein the radar reflectivity effective data of the source area is used as a source area data set, and the time range of the source area data set is not overlapped with the time ranges of the test set and the training set of the target area data set in a crossing way.
Further, the evaluation model of the domain migration prediction model is:
CSI all =α 20 CSI 20 +α 30 CSI 30 +α 40 CSI 40 ;
wherein: CSI (channel State information) all The range of values for the evaluation index is (0, 1); CSI (channel State information) 20 Critical index, CSI, representing radar echo intensity of 20dBZ 30 Representing radar returnsCritical index of 30dBZ strength, CSI 40 Represents the critical index, alpha, of radar echo intensity of 40dBZ 20 、α 30 And alpha 40 Respectively is CSI 20 、CSI 30 And CSI (channel State information) 40 Coefficients of (2);
wherein, the critical index calculation model is:
wherein: i represents a critical threshold, i=20, 30,40; n (N) A N represents the number of lattice points where the observed radar echo intensity is greater than the critical threshold i and the predicted echo intensity is also greater than the critical threshold B Indicating the number of lattice points, N, of which the predicted radar echo intensity is greater than a critical threshold i but the observed radar echo threshold is less than i C Representing the number of lattice points for which the predicted radar echo intensity is less than the critical threshold i but the observed radar echo threshold is greater than i.
Accordingly, the present invention also provides a readable storage medium having a computer program, which when run, performs the above method.
The invention has the beneficial effects that: according to the method and the device, the radar reflectivity is used as training data, and based on the migration prediction model and the determined target domain data set and source domain data set, the migration prediction model can acquire enough radar echo characteristics under the condition of fewer samples, so that the accuracy of radar extrapolation prediction is effectively improved, the prediction effect of strong convection echo can be effectively improved, and the adaptability is strong.
Drawings
The invention is further described below with reference to the accompanying drawings and examples:
FIG. 1 is a flow chart of the present invention.
Fig. 2 is a graph of an embodiment of the present invention.
Detailed Description
The present invention is further described in detail below:
the invention provides a radar echo extrapolation method based on migration prediction, which comprises the following steps:
s1, acquiring radar reflectivity data, and preprocessing the radar reflectivity data to obtain radar reflectivity effective data; preprocessing the data, such as removing abnormal data, interpolating the missing data, and the like, wherein the accuracy of the subsequent processing results is ensured;
s2, determining a target domain data set and a source domain data set according to radar reflectivity effective data;
s3, constructing a domain migration prediction model f, inputting a source domain data set into the migration prediction model, and training the migration prediction model;
s4, constructing an evaluation model of the domain migration prediction model, inputting test set data in the target domain data set into the domain migration prediction model trained in the step S3 to obtain a test result, calculating an evaluation index of the domain migration prediction model by the evaluation model, and selecting the domain migration prediction model corresponding to the maximum evaluation index as an optimal prediction model f s ;
S5, inputting a training set in the target domain data set into an optimal prediction model f s And for the optimal prediction model f s Training, inputting a test set in the target domain data set into the trained optimal prediction model to obtain a test result, and calculating a trained optimal prediction model f by an evaluation model s And selecting the optimal prediction model f corresponding to the maximum evaluation index value s As radar echo extrapolation model f T According to the method, the radar reflectivity is used as training data, and based on the migration prediction model and the determined target domain data set and source domain data set, the migration prediction model can acquire enough radar echo characteristics under the condition of fewer samples, so that the accuracy of radar extrapolation prediction is effectively improved, the prediction effect of strong convection echo can be effectively improved, and the adaptability is strong.
In this embodiment, the domain migration prediction model is any one of ConvLSTM, predRNN and TrajGRU models, and of course, other domain migration prediction models may be implemented, and the structure and principle of these domain migration prediction models are the prior art, where:
the Loss function of the domain transfer learning model is Loss:
wherein: n is the number of samples, J is the total number of predictions,representing t 0 Observation data at +t moment, +.>Representing t 0 Predicted data at +t, t 0 The last time point of observation is the last time point in the sample data.
In this embodiment, determining the target domain data set and the source domain data set in step S2 specifically includes:
determining the space position and the time range of a target area according to task requirements, taking radar reflectivity effective data in the space position and the time range as a target domain data set, and dividing the target domain data set into training set data and test set data, wherein the time of the training set data and the time of the test set data are not overlapped in a crossing way; for example: setting the time step to be 6min, wherein the continuous duration of the historical data (the observed data of the reflectivity of the historical radar) is 2.5 hours, then 25 hours are included in 2.5 hours, the data of the first 1 hour in the 2.5 hours are used as historical data, the data of the last 1.5 hours are used as prediction data, the data in each time form a graph, and the graph is divided into a plurality of grids according to the set specification; then there is a corresponding grid number;
and selecting an area which is outside the target area and has the same spatial size as the target area as a source area, wherein the radar reflectivity effective data of the source area is used as a source area data set, and the time range of the source area data set is not overlapped with the time ranges of the test set and the training set of the target area data set in a crossing way.
Adding an A area with the angle of 5 DEG x 5 DEG (expressed longitude and latitude) as a target area, taking data in a set time as a target area data set of the target area, and dividing the data set into a test set and a training set according to the rule, wherein two areas with the angle of 5 DEG x 5 DEG are selected as source area areas outside the A area, namely, the source area and the target area are not overlapped; through the fact that the time range and the area are not overlapped in a crossing mode, the effectiveness of the whole training process and the trained model can be ensured, and the risk that potential repeated samples cause poor generalization capability of the model is avoided.
In this embodiment, the evaluation model of the domain migration prediction model is:
CSI all =α 20 CSI 20 +α 30 CSI 30 +α 40 CSI 40 ;
wherein: CSI (channel State information) all The range of values for the evaluation index is (0, 1); CSI (channel State information) 20 Critical index, CSI, representing radar echo intensity of 20dBZ 30 Critical index, CSI, representing radar echo intensity of 30dBZ 40 Represents the critical index, alpha, of radar echo intensity of 40dBZ 20 、α 30 And alpha 40 Respectively is CSI 20 、CSI 30 And CSI (channel State information) 40 Coefficients of (2);
wherein, the critical index calculation model is:
wherein: i represents a critical threshold, i=20, 30,40; n (N) A N represents the number of lattice points where the observed radar echo intensity is greater than the critical threshold i and the predicted echo intensity is also greater than the critical threshold B Indicating that the predicted radar echo intensity is largeLattice number, N, at critical threshold i but observed radar echo threshold less than i C Representing the number of lattice points for which the predicted radar echo intensity is less than the critical threshold i but the observed radar echo threshold is greater than i.
The following is a specific example of the present invention:
the example, namely the area a and the other two source domain areas, are processed through the history data by the process of the invention and the two existing example algorithms, and the obtained results are shown in fig. 2:
the results show that the migration learning method of the invention (namely PredRNN-Tr in FIG. 2) has obvious improvement compared with the traditional optical flow method and the PredRNN basic model, and has obvious advantages with the time extension and the strength increase. Compared with an optical flow method, the CSI score of the transfer learning method is improved by about 0.1 within the prediction time of 30-90 min at more than 30 dBZ; above 40dBZ, the CSI score of the transfer learning method is increased by about 0.08 in the prediction time of 30min to 90 min. Compared with the optical flow method, the hit rate and the false alarm rate of the transfer learning are obviously improved in each echo intensity. Compared with the PredRNN basic model, the CSI score of the strong echo region in the interval of 30dBZ and 40dBZ is obviously improved by transfer learning, the hit rate of the strong convection region above 40dBZ is enhanced, and the false alarm rate of the weak echo region below the threshold value of 40dBZ is weakened.
Accordingly, the present invention also provides a readable storage medium having a computer program, which when run, performs the above method.
Finally, it is noted that the above embodiments are only for illustrating the technical solution of the present invention and not for limiting the same, and although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications and equivalents may be made thereto without departing from the spirit and scope of the technical solution of the present invention, which is intended to be covered by the scope of the claims of the present invention.
Claims (6)
1. A radar echo extrapolation method based on migration prediction is characterized by comprising the following steps: the method comprises the following steps:
s1, acquiring radar reflectivity data, and preprocessing the radar reflectivity data to obtain radar reflectivity effective data;
s2, determining a target domain data set and a source domain data set according to radar reflectivity effective data;
s3, constructing a domain migration prediction model f, inputting a source domain data set into the migration prediction model, and training the migration prediction model;
s4, constructing an evaluation model of the domain migration prediction model, inputting test set data in the target domain data set into the domain migration prediction model trained in the step S3 to obtain a test result, calculating an evaluation index of the domain migration prediction model by the evaluation model, and selecting the domain migration prediction model corresponding to the maximum evaluation index as an optimal prediction model f s ;
S5, inputting a training set in the target domain data set into an optimal prediction model f s And for the optimal prediction model f s Training, inputting a test set in the target domain data set into the trained optimal prediction model to obtain a test result, and calculating a trained optimal prediction model f by an evaluation model s And selecting the optimal prediction model f corresponding to the maximum evaluation index value s As radar echo extrapolation model f T 。
2. The migration prediction-based radar echo extrapolation method according to claim 1, wherein: the domain migration prediction model is either ConvLSTM, predRNN or TrajGRU model.
3. The migration prediction-based radar echo extrapolation method according to claim 2, wherein: the Loss function of the domain transfer learning model is Loss:
wherein: n is the number of samples, J is the total number of predictions,representing t 0 Observation data at +t moment, +.>Representing t 0 Predicted data at +t, t 0 Is the last point in time of observation.
4. The migration prediction-based radar echo extrapolation method according to claim 1, wherein: the determining the target domain data set and the source domain data set in step S2 specifically includes:
determining the space position and the time range of a target area according to task requirements, taking radar reflectivity effective data in the space position and the time range as a target domain data set, and dividing the target domain data set into training set data and test set data, wherein the time of the training set data and the time of the test set data are not overlapped in a crossing way;
and selecting an area which is outside the target area and has the same spatial size as the target area as a source area, wherein the radar reflectivity effective data of the source area is used as a source area data set, and the time range of the source area data set is not overlapped with the time ranges of the test set and the training set of the target area data set in a crossing way.
5. The migration prediction-based radar echo extrapolation method according to claim 1, wherein: the evaluation model of the domain migration prediction model is as follows:
CSI all =α 20 CSI 20 +α 30 CSI 30 +α 40 CSI 40 ;
wherein: CSI (channel State information) all The range of values for the evaluation index is (0, 1); CSI (channel State information) 20 Critical index, CSI, representing radar echo intensity of 20dBZ 30 Critical index, CSI, representing radar echo intensity of 30dBZ 40 Represents the critical index, alpha, of radar echo intensity of 40dBZ 20 、α 30 And alpha 40 Respectively is CSI 20 、CSI 30 And CSI (channel State information) 40 Coefficients of (2);
wherein, the critical index calculation model is:
wherein: i represents a critical threshold, i=20, 30,40; n (N) A N represents the number of lattice points where the observed radar echo intensity is greater than the critical threshold i and the predicted echo intensity is also greater than the critical threshold B Indicating the number of lattice points, N, of which the predicted radar echo intensity is greater than a critical threshold i but the observed radar echo threshold is less than i C Representing the number of lattice points for which the predicted radar echo intensity is less than the critical threshold i but the observed radar echo threshold is greater than i.
6. A readable storage medium having a computer program, characterized in that: a computer program of a readable storage medium, which when run performs the method of any of claims 1-5.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311299982.4A CN117129963A (en) | 2023-10-09 | 2023-10-09 | Radar echo extrapolation method based on migration prediction and readable storage medium thereof |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202311299982.4A CN117129963A (en) | 2023-10-09 | 2023-10-09 | Radar echo extrapolation method based on migration prediction and readable storage medium thereof |
Publications (1)
Publication Number | Publication Date |
---|---|
CN117129963A true CN117129963A (en) | 2023-11-28 |
Family
ID=88860204
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202311299982.4A Pending CN117129963A (en) | 2023-10-09 | 2023-10-09 | Radar echo extrapolation method based on migration prediction and readable storage medium thereof |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN117129963A (en) |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2018091785A (en) * | 2016-12-06 | 2018-06-14 | 株式会社デンソーテン | Radar device and target detection method |
CN110824481A (en) * | 2019-10-28 | 2020-02-21 | 兰州大方电子有限责任公司 | Quantitative precipitation prediction method based on radar reflectivity extrapolation |
CN111931991A (en) * | 2020-07-14 | 2020-11-13 | 上海眼控科技股份有限公司 | Weather nowcast method, weather nowcast device, computer equipment and storage medium |
US20210103051A1 (en) * | 2019-10-04 | 2021-04-08 | International Business Machines Corporation | Predicting weather radar images |
CN112731330A (en) * | 2021-01-12 | 2021-04-30 | 西安电子科技大学 | Radar carrier frequency parameter change steady target identification method based on transfer learning |
CN113936142A (en) * | 2021-10-13 | 2022-01-14 | 成都信息工程大学 | Rainfall approach forecasting method and device based on deep learning |
CN115792853A (en) * | 2022-12-15 | 2023-03-14 | 南京信息工程大学 | Radar echo extrapolation method based on dynamic weight loss |
CN116663736A (en) * | 2023-06-07 | 2023-08-29 | 南京信息工程大学 | Short-term precipitation prediction method based on double-region position priori and global space-time representation |
-
2023
- 2023-10-09 CN CN202311299982.4A patent/CN117129963A/en active Pending
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2018091785A (en) * | 2016-12-06 | 2018-06-14 | 株式会社デンソーテン | Radar device and target detection method |
US20210103051A1 (en) * | 2019-10-04 | 2021-04-08 | International Business Machines Corporation | Predicting weather radar images |
CN110824481A (en) * | 2019-10-28 | 2020-02-21 | 兰州大方电子有限责任公司 | Quantitative precipitation prediction method based on radar reflectivity extrapolation |
CN111931991A (en) * | 2020-07-14 | 2020-11-13 | 上海眼控科技股份有限公司 | Weather nowcast method, weather nowcast device, computer equipment and storage medium |
CN112731330A (en) * | 2021-01-12 | 2021-04-30 | 西安电子科技大学 | Radar carrier frequency parameter change steady target identification method based on transfer learning |
CN113936142A (en) * | 2021-10-13 | 2022-01-14 | 成都信息工程大学 | Rainfall approach forecasting method and device based on deep learning |
CN115792853A (en) * | 2022-12-15 | 2023-03-14 | 南京信息工程大学 | Radar echo extrapolation method based on dynamic weight loss |
CN116663736A (en) * | 2023-06-07 | 2023-08-29 | 南京信息工程大学 | Short-term precipitation prediction method based on double-region position priori and global space-time representation |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110135654B (en) | Method and apparatus for predicting strong convection weather | |
CN110223302B (en) | Ship multi-target detection method based on rotation region extraction | |
CN114022847A (en) | Intelligent agent trajectory prediction method, system, equipment and storage medium | |
CN108896973B (en) | Radar data calibration method, trace point quality evaluation method and storage medium | |
CN110991568B (en) | Target identification method, device, equipment and storage medium | |
CN113496104A (en) | Rainfall forecast correction method and system based on deep learning | |
CN112068085B (en) | Radar sea clutter original data rapid preprocessing method based on deep learning | |
Lee et al. | Applying machine learning methods to detect convection using Geostationary Operational Environmental Satellite-16 (GOES-16) advanced baseline imager (ABI) data | |
CN114565124A (en) | Ship traffic flow prediction method based on improved graph convolution neural network | |
CN112561881A (en) | Infrared image self-adaptive data enhancement method based on evaluation model | |
US11668857B2 (en) | Device, method and computer program product for validating data provided by a rain sensor | |
CN114897272A (en) | Liquid level prediction method and device for sewage inspection well, electronic equipment and storage medium | |
CN112949932B (en) | Ship traffic flow prediction method, device, computer equipment and storage medium | |
CN113053536B (en) | Infectious disease prediction method, system and medium based on hidden Markov model | |
EP3968214A1 (en) | Device and method for classifying images and accessing the robustness of the classification | |
CN107437112B (en) | A kind of mixing RVM model prediction methods based on the multiple dimensioned kernel function of improvement | |
CN117253344A (en) | Seawater acidification early warning and forecasting method, system and electronic equipment | |
CN117129963A (en) | Radar echo extrapolation method based on migration prediction and readable storage medium thereof | |
CN117351321A (en) | Single-stage lightweight subway lining cavity recognition method and related equipment | |
Lee et al. | Applying machine learning methods to detect convection using GOES-16 ABI data | |
CN114021445B (en) | Ocean vortex mixing non-locality prediction method based on random forest model | |
CN110796076A (en) | Hyperspectral image river detection method | |
CN116129239A (en) | Small target detection method, device, equipment and storage medium | |
CN113269805B (en) | Rainfall event guided remote sensing rainfall inversion training sample self-adaptive selection method | |
CN115797311A (en) | Highway crack image segmentation method based on deep reinforcement learning |
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 |