CN116467933A - Storm surge water increasing prediction method and system based on deep learning - Google Patents
Storm surge water increasing prediction method and system based on deep learning Download PDFInfo
- Publication number
- CN116467933A CN116467933A CN202310279148.2A CN202310279148A CN116467933A CN 116467933 A CN116467933 A CN 116467933A CN 202310279148 A CN202310279148 A CN 202310279148A CN 116467933 A CN116467933 A CN 116467933A
- Authority
- CN
- China
- Prior art keywords
- storm surge
- time
- typhoon
- prediction
- typhoons
- 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
- 238000013135 deep learning Methods 0.000 title claims abstract description 50
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 title claims abstract description 45
- 238000000034 method Methods 0.000 title claims abstract description 33
- 238000012549 training Methods 0.000 claims abstract description 65
- 238000007781 pre-processing Methods 0.000 claims abstract description 21
- 230000006870 function Effects 0.000 claims description 31
- 238000012795 verification Methods 0.000 claims description 26
- 238000012360 testing method Methods 0.000 claims description 20
- 238000012545 processing Methods 0.000 claims description 16
- 238000013528 artificial neural network Methods 0.000 claims description 15
- 230000003020 moisturizing effect Effects 0.000 claims description 15
- 238000010276 construction Methods 0.000 claims description 14
- 238000010606 normalization Methods 0.000 claims description 13
- 238000005457 optimization Methods 0.000 claims description 7
- 238000003860 storage Methods 0.000 claims description 7
- 238000004140 cleaning Methods 0.000 claims description 6
- 238000010200 validation analysis Methods 0.000 claims description 6
- 238000010586 diagram Methods 0.000 description 9
- 238000004590 computer program Methods 0.000 description 7
- 238000003062 neural network model Methods 0.000 description 7
- 238000012986 modification Methods 0.000 description 6
- 230000004048 modification Effects 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 230000003247 decreasing effect Effects 0.000 description 5
- 238000011156 evaluation Methods 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 238000004088 simulation Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 230000007774 longterm Effects 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 210000002569 neuron Anatomy 0.000 description 2
- 230000001131 transforming effect Effects 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000013277 forecasting method Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
Classifications
-
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/10—Numerical modelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2113/00—Details relating to the application field
- G06F2113/08—Fluids
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/02—Reliability analysis or reliability optimisation; Failure analysis, e.g. worst case scenario performance, failure mode and effects analysis [FMEA]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Evolutionary Computation (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biophysics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computer Hardware Design (AREA)
- Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Biomedical Technology (AREA)
- Geometry (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application discloses a storm surge water increasing prediction method and system based on deep learning, wherein the method comprises the following steps: acquiring a storm surge data set, wherein the storm surge data set comprises a plurality of typhoon storm surge data files, and each typhoon storm surge data file comprises typhoon information and storm surge information corresponding to each moment; carrying out data preprocessing on typhoon information and storm surge information corresponding to each moment to obtain a preprocessed storm surge data set; acquiring the current time and the required advanced forecast time, and taking typhoon information corresponding to each time from the current time to the required advanced forecast time and storm surge information of the required advanced forecast time as a time sequence so as to construct a time sequence data set; and constructing a storm surge prediction model, and inputting a time sequence data set into the storm surge prediction model for training so as to perform storm surge prediction by adopting the trained storm surge prediction model, thereby having high prediction efficiency and long prediction period.
Description
Technical Field
The application relates to the technical field of storm surge prediction, in particular to a storm surge prediction method based on deep learning and a storm surge prediction system based on deep learning.
Background
In the related art, storm surge refers to abnormal elevation of the sea surface caused by strong atmospheric disturbance, which generally includes strong wind, sudden changes in air pressure, etc., such as typhoon process; china is the most frequently affected by North Pacific ocean typhoons, and typhoons and storm surge cause great threat to coastal city economic development and life and property safety of people; therefore, the accurate and timely storm surge pre-alarm has important practical significance for coastal disaster prevention and reduction.
At present, storm tide forecasting methods are mainly divided into experience forecasting, numerical simulation forecasting and artificial intelligence; the experience forecast is to establish the statistical relationship between storm surge, air pressure and wind of the station by using a statistical method, is simple, convenient and easy to learn and master, has higher forecast precision for certain specific stations, but the method needs long-term historical observation data of the stations; the numerical simulation prediction is to establish a mathematical model based on a hydrodynamic principle to solve a basic equation, and in recent years, as the support of software and hardware of a computer is developed rapidly, a multi-physical coupling process can be comprehensively considered, storm surge with high space-time resolution in a specific area can be simulated and predicted, but higher requirements are also put forward on calculation resources, and on the other hand, the driving field (comprising an air pressure field and a wind field) input by the numerical model, the initial boundary condition and the uncertainty of parameter setting can influence storm surge prediction precision; in recent years, with the development of artificial intelligence technology, various machine learning methods are used for exploring storm surge prediction, but most of mode training data are limited or do not consider the influence of future typhoon factor time sequence change on storm surge, so that the prediction effect is poor.
Disclosure of Invention
The present application aims to solve, at least to some extent, one of the technical problems in the art described above. Therefore, an object of the present application is to provide a storm surge prediction method based on deep learning, which constructs a time sequence training set for a large number of preset imaginary typhoon storm surge data, so that an LSTM neural network model is constructed, and the influence of typhoon factor time sequence change on storm surge is considered, so that the prediction efficiency is high and the prediction period is long.
A second object of the present application is to provide a storm surge prediction system based on deep learning.
In order to achieve the above objective, an embodiment of a first aspect of the present application provides a storm surge water increasing prediction method based on deep learning, including the following steps: acquiring a storm surge data set, wherein the storm surge data set comprises a plurality of virtual typhoon storm surge data files, and each virtual typhoon storm surge data file comprises typhoon information and storm surge information corresponding to each moment; carrying out data preprocessing on typhoon information and storm surge information corresponding to each moment to obtain a preprocessed storm surge data set; acquiring the current time and the required advanced forecast time, and taking typhoon information corresponding to each time from the current time to the required advanced forecast time and storm surge information of the required advanced forecast time as a time sequence so as to construct a time sequence data set; and constructing a storm surge and water increasing deep learning prediction model, and inputting the time sequence data set into the storm surge and water increasing deep learning prediction model for training so as to perform storm surge and water increasing prediction by adopting the trained storm surge and water increasing deep learning prediction model.
According to the technical means, the storm surge data set is obtained, wherein the storm surge data set comprises a plurality of supposed typhoons and storm surge data files, and each supposed typhoons and storm surge data file comprises typhoons information and storm surge information corresponding to each moment; carrying out data preprocessing on typhoon information and storm surge information corresponding to each moment to obtain a preprocessed storm surge data set; acquiring the current time and the required advanced forecast time, and taking typhoon information corresponding to each time from the current time to the required advanced forecast time and storm surge information of the required advanced forecast time as a time sequence so as to construct a time sequence data set; and constructing a storm surge moisturizing deep learning prediction model, and inputting a time sequence data set into the storm surge moisturizing deep learning prediction model for training so as to perform storm surge moisturizing prediction by adopting the trained storm surge moisturizing deep learning prediction model, thereby having high prediction efficiency and long prediction period.
In addition, the storm surge water increasing prediction method based on deep learning according to the embodiment of the application may further have the following additional technical features:
optionally, acquiring the storm surge dataset comprises: the method comprises the steps of obtaining typhoon information corresponding to each moment in historical multi-station typhoons, wherein the typhoons comprise longitude and latitude coordinates of a typhooncenter, a maximum wind speed near the typhooncenter, a typhooncenter moving speed, a typhooncenter moving direction and typhooncenter air pressure; and transforming typhoon information corresponding to each moment in the historical multi-station typhoons to generate a series of virtual typhoons, and inputting the generated series of virtual typhoons into a storm surge numerical model so as to obtain amplified series of virtual typhoons storm surge data files, wherein each data file comprises typhoon information corresponding to each moment and storm surge information at the same moment.
According to the technical means, the training database can be more detailed and accurate, so that the accuracy of prediction based on deep learning of typhoon storm surge water increase is improved.
Optionally, performing data preprocessing on typhoon information and storm surge information corresponding to each moment to obtain a preprocessed storm surge data set, including: carrying out data cleaning on typhoon information and storm surge information corresponding to each moment; dividing the cleaned data according to different proportions so as to obtain a corresponding training set, a corresponding testing set and a corresponding verification set; and carrying out standardization processing on the divided data set by adopting a normalization method.
Optionally, constructing the time sequence data set for the preprocessed storm surge data set includes obtaining a current time and a required advanced forecast time, and taking typhoon information corresponding to each time from the current time to the required advanced forecast time and storm surge moisturizing information of the required advanced forecast time as a time sequence so as to obtain a time sequence training set, a test set and a verification set.
Optionally, the storm surge water increasing prediction model is constructed by adopting an LSTM neural network, the parameters of the LSTM neural network are set to be 5 in input dimension, 1 in output dimension, the initial learning rate is set to be 0.001, the loss function is selected to be MAE, the optimization function is selected to be RMSprop, and a callback function ReduceLROnPlateeau of Keras is used in training and is matched with early stop function EarlyStopping; earlyStopping is used for training through 10 continuous epochs, and when the error of the verification set is reduced by not more than 0.001, the model stops training; the reduce lronplateau is used to reduce the learning rate by a factor of 10 when the error of the validation set is reduced by no more than 0.0001 after training of 5 epochs in succession.
Optionally, the storm surge prediction is performed by using a trained storm surge prediction model, including: inputting typhoon information to be predicted into a trained storm surge prediction model to obtain storm surge data at a corresponding moment to be predicted; and performing inverse normalization processing on the obtained storm surge data so as to obtain a corresponding storm surge water level predicted value.
To achieve the above object, an embodiment of a second aspect of the present application provides a storm surge prediction system based on deep learning, including: the system comprises an acquisition module, a storage module and a storage module, wherein the acquisition module is used for acquiring a storm surge data set, the storm surge data set comprises a plurality of virtual typhoons and storm surge data files, and each virtual typhoons and storm surge data file comprises typhoons information and storm surge information corresponding to each moment; the preprocessing module is used for preprocessing the typhoon information and storm surge information corresponding to each moment in data so as to obtain a preprocessed storm surge data set; the time sequence construction module is used for acquiring the current time and the required advanced forecast time, and taking typhoon information corresponding to each time from the current time to the required advanced forecast time and storm surge information of the required advanced forecast time as a time sequence so as to construct a time sequence data set; the model construction module is used for constructing a storm surge moisturizing deep learning prediction model, and inputting the time sequence data set into the storm surge moisturizing prediction model for training so as to obtain a trained storm surge moisturizing deep learning prediction model; the model prediction module is used for inputting typhoon information to be predicted into the trained storm surge water increasing deep learning prediction model so as to obtain a corresponding storm surge water increasing water level prediction value.
According to the technical means, a storm surge data set is obtained through an obtaining module, wherein the storm surge data set comprises a plurality of typhoons and storm surge data files, and each typhoons and storm surge data file comprises typhoons information and storm surge moisturizing information corresponding to each moment; the preprocessing module performs data preprocessing on typhoon information and storm surge information corresponding to each moment to obtain a preprocessed storm surge data set; the time sequence construction module acquires the current time and the required advanced forecast time, and takes typhoon information corresponding to each time from the current time to the required advanced forecast time and storm surge information of the required advanced forecast time as a time sequence so as to construct a time sequence data set; the model construction module is used for constructing a storm surge deep learning prediction model, inputting a time sequence data set into the storm surge deep learning prediction model for training so as to obtain a trained storm surge deep learning prediction model prediction module, and inputting typhoon information to be predicted into the trained storm surge deep learning prediction model so as to obtain a corresponding storm surge water level prediction value; thus, the prediction efficiency is high and the prediction period is long.
In addition, the storm surge prediction system based on deep learning according to the embodiment of the application may further have the following additional technical features:
optionally, the obtaining module is further configured to obtain typhoon information corresponding to each moment in the historical multi-station typhoons, where the typhoons information includes longitude and latitude coordinates where a typhoons center is located, a maximum wind speed near the typhoons center, a movement speed of the typhoons center, a movement direction of the typhoons center and air pressure of the typhoons center; and transforming typhoon information corresponding to each moment in the historical multi-station typhoons to generate a series of virtual typhoons, and inputting the generated series of virtual typhoons into a storm surge numerical model so as to obtain amplified series of virtual typhoons storm surge data files, wherein each data file comprises typhoon information corresponding to each moment and storm surge information at the same moment.
Optionally, the preprocessing module is further used for cleaning data of typhoon information and storm surge information corresponding to each moment; dividing the cleaned data according to different proportions so as to obtain a corresponding training set, a corresponding testing set and a corresponding verification set; and carrying out standardization processing on the divided data set by adopting a normalization method.
Optionally, the time sequence construction module is further configured to obtain a current time and a required advanced forecast time, and take typhoon information corresponding to each time from the current time to the required advanced forecast time and storm surge information of the required advanced forecast time as a time sequence, so as to construct a time sequence training set, a test set and a verification set;
optionally, the model building module adopts an LSTM neural network for building, the parameters of the LSTM neural network are set to be 5 in input dimension, 1 in output dimension, 0.001 in initial learning rate, MAE in loss function, RMSprop in optimization function, and a callback function ReduceLROnPlateeau of Keras is used in training and matched with early stop function earlyStopping; earlyStopping is used for training through 10 continuous epochs, and when the error of the verification set is reduced by not more than 0.001, the model stops training; the reduce lronplateau is used to reduce the learning rate by a factor of 10 when the error of the validation set is reduced by no more than 0.0001 after training of 5 epochs in succession.
Optionally, the model prediction module is further configured to input typhoon information to be predicted into a trained storm surge prediction model to obtain storm surge data at a corresponding time to be predicted; and performing inverse normalization processing on the obtained storm surge data so as to obtain a corresponding storm surge water level predicted value.
Drawings
FIG. 1 is a flow chart of a storm surge prediction method based on deep learning according to an embodiment of the application;
FIG. 2 is a schematic diagram of a storm surge prediction model predicting a 12h surge prediction error result according to an embodiment of the application;
FIG. 3 is a schematic diagram of a storm surge prediction model for predicting a 12h surge prediction result according to an embodiment of the application;
fig. 4 is a block schematic diagram of a deep learning-based storm surge prediction system according to an embodiment of the application.
Detailed Description
Embodiments of the present application are described in detail below, examples of which are illustrated in the accompanying drawings, wherein the same or similar reference numerals refer to the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the drawings are exemplary and intended for the purpose of explaining the present application and are not to be construed as limiting the present application.
In order to better understand the above technical solution, exemplary embodiments of the present application will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present application are shown in the drawings, it should be understood that the present application may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.
In order to better understand the above technical solutions, the following detailed description will refer to the accompanying drawings and specific embodiments.
As shown in fig. 1, the storm surge water increasing prediction method based on deep learning comprises the following steps:
s101, acquiring a storm surge data set, wherein the storm surge data set comprises a plurality of virtual typhoons and storm surge data files, and each virtual typhoons and storm surge data file comprises typhoons information and storm surge information corresponding to each moment.
That is, a storm surge data set is obtained in a training database by presetting the storm surge moisturizing training database, wherein the training database comprises a plurality of typhoon storm surge moisturizing data files, each data file comprises a plurality of columns, each column comprises a typhoon information set and a storm surge moisturizing set, and the typhoon information set comprises longitude and latitude coordinates of a typhoon center, a near typhoon center maximum wind speed, a typhoon center moving direction and typhoon center air pressure; the elements of the typhoon information set are in one-to-one correspondence with the elements of the storm surge water increasing set.
As one embodiment, acquiring a storm surge dataset includes: the method comprises the steps of obtaining typhoon information corresponding to each moment in historical multi-station typhoons, wherein the typhoon information comprises longitude and latitude coordinates of a typhoon center, a maximum wind speed near the typhoon center, a typhoon center moving speed, a typhoon center moving direction and typhoon center air pressure; the typhoon information corresponding to each moment in the historical multi-station typhoons is transformed to generate a series of virtual typhoons, the generated series of virtual typhoons are input into a storm surge numerical model so as to obtain amplified series of virtual typhoons storm surge data files, and each data file comprises typhoon information corresponding to each moment and storm surge information at the same moment.
The storm surge numerical model is a numerical calculation model based on a momentum equation and a continuous equation, the input of the numerical model is the series of virtual typhoon information, and the output comprises the single-moment typhoon center moving direction, the typhoon center moving speed, the near typhoon center maximum wind speed, the longitude and latitude coordinates of the typhoon center, the typhoon center air pressure and the storm surge water increasing water level at the same moment.
As a specific example, 6 historical typhoons affecting the coastal areas of Fujian in 1969-1994 are taken as reference parameters, and typhoons of 6911, 7123, 7613, 9107, 9215 and 9414 are included; typhoon parameters at intervals of one hour when typhoons occur; the historical typhoon parameters comprise a typhoon moving direction, a typhoon moving speed, a typhoon center longitude, a typhoon center latitude, a near-center maximum wind speed and a center air pressure at the moment t; and respectively constructing five groups of A-E imaginary typhoons for 6 historical typhoons.
Wherein, the group A virtual typhoons take the original path as a reference, and the latitude is increased or decreased by 0.01 degree; for typhoons of 6911, the latitude increasing and decreasing range is [ -1.4,1.2], 261 fields are adopted; for No. 7123 typhoons, the latitude increases and decreases in the range of [ -2.3,0.3], and 261 fields are adopted; for typhoons No. 7613, the latitude increasing and decreasing range is [ -2.4,0.2], 261 fields are adopted; for typhoons No. 9215, the latitude increasing and decreasing range is [ -1.2,1.4], and 261 fields are adopted; for typhoons No. 9107, the latitude increasing and decreasing range is [ -1,0.5], and the total range is 151 fields; for typhoons No. 9414, the latitude range is [0,2], for a total of 201 fields. 1396 virtual typhoons were summed.
Based on the group A, the maximum wind speed of the group B virtual typhoons is increased by 10%, and the central air pressure is reduced by 5hPa.
Based on the group A, the maximum wind speed of the C group virtual typhoons is increased by 20%, and the central air pressure is reduced by 10hPa.
Based on the group A, the maximum wind speed of the group D virtual typhoons is reduced by 10%, and the central air pressure is increased by 5hPa.
Based on the group A, the maximum wind speed of the group E virtual typhoons is reduced by 20%, and the central air pressure is increased by 10hPa.
And calculating the storm tide time-by-time water increasing value of the Fujian coastal tidal level station by utilizing the series of imaginary typhoon paths and parameters to form 6980 main coastal station storm tide water increasing process data sets.
In conclusion, a great number of supposed typhoons and storm surge are simulated by using a storm surge numerical model, and the problems of lack of historical observation data of a measuring station and incomplete historical process can be overcome.
S102, carrying out data preprocessing on typhoon information and storm surge information corresponding to each moment to obtain a preprocessed storm surge data set.
As an embodiment, the data preprocessing is performed on typhoon information and storm surge information corresponding to each moment to obtain a preprocessed storm surge data set, which includes: carrying out data cleaning on typhoon information and storm surge information corresponding to each moment; dividing the cleaned data according to different proportions so as to obtain a corresponding training set, a corresponding testing set and a corresponding verification set; and carrying out standardization processing on the divided data set by adopting a normalization method.
As a specific embodiment, firstly, data cleaning is carried out on typhoon information and storm surge information corresponding to each moment so as to delete samples with a surge missing value at the tail of typhoons; then, in the deep learning, the training set is used for model training, the test set is used for model selection to determine the optimal super parameters, and the final verification is performed on the verification set after the optimal model is selected; in order to ensure that the data distribution of three data sets is balanced, on the premise of keeping the time sequence of single-field typhoons, all the imaginary typhoons are disturbed, and then are divided into a training set, a testing set and a verification set according to the proportion of 70%, 15% and 15%; the training set is 4886 field imaginary typhoons, 380336 water increasing number columns; the test set is 1047 pieces of imaginary typhoons, 81686 water increasing number columns; the verification set is composed of 1047 virtual typhoons and 81563 water increasing number columns; finally, carrying out standardization treatment on the data set by adopting a normalization method; the normalization processing formula is as follows:
wherein x' is a standardized result, mu is a data average value, sigma is a standard deviation, standardized parameters mu and sigma are taken from a training set, and the testing set and the verification set use the same standardized parameters as the training set, so that the standardized operation of all data sets is ensured to be consistent.
S103, acquiring the current time and the required advanced forecast time, and taking typhoon information corresponding to each time from the current time to the required advanced forecast time and storm surge water increasing information of the required advanced forecast time as a time sequence so as to construct a time sequence data set.
As a specific embodiment, for a typhoon storm surge data file, taking a 12h forecast as an example, constructing a time sequence training set, a test set and a verification set, if the current moment is t, taking the movement direction, the movement speed, the typhoon center longitude, the typhoon center latitude, the near-center maximum wind speed and the center air pressure of the typhoon center at the moment of t+12 as a group of input parameters, taking the station storm surge at the moment of t+12 as corresponding output data, and taking the corresponding output data as a group of surge training series; if typhoon parameters and corresponding storm surge data at n times are shared for a typhoon storm surge data file, a (n-12) +1 group surge training time sequence can be constructed; similarly, the sequence of the water increasing training time of all typhoon data file structures in the storm surge data set is preset into a sequence training data set, a test set and a verification set.
That is, LSTM time sequence prediction neural network utilizes future typhoon information to predict storm surge water rise 3h, 6h, 12h and 24h in advance; taking the prediction of 12h as an example, if the current moment is t, and t, t+1, t+2, t+3, & gt, predicting the water adding water level at the moment t+12 according to the moving direction and moving speed of the typhoon center, the longitude of the typhoon center, the latitude of the typhoon center, the maximum wind speed near the center and the central air pressure of the typhoon at the moment t+12.
S104, constructing a storm surge prediction model, and inputting a time sequence data set into the storm surge prediction model for training so as to perform storm surge prediction by adopting the trained storm surge prediction model.
As an embodiment, a storm surge water increasing prediction model is constructed by adopting an LSTM neural network, the parameters of the LSTM neural network are set to be 6 in input dimension and 1 in output dimension, the initial learning rate is set to be 0.001, the loss function is selected to be MAE, the optimization function is selected to be RMSprop, and a callback function ReduceLROnPlateeau of Keras is used in training and is matched with an early stop function EarlyStopping; earlyStopping is used for training through 10 continuous epochs, and when the error of the verification set is reduced by not more than 0.001, the model stops training; the reduce lronplateau is used to reduce the learning rate by a factor of 10 when the error of the validation set is reduced by no more than 0.0001 after training of 5 epochs in succession.
That is, parameters of the LSTM neural network model are preliminarily developed, the parameters of the LSTM neural network model are set to be 6 in input dimension, 1 in output dimension, 0.001 in initial learning rate, MAE is selected as a loss function, RMSprop is selected as an optimization function.
Inputting the time sequence training set data constructed in the above into an LSTM neural network model for iterative training, and verifying the training result by using the test set data until the prediction error meets the preset prediction accuracy evaluation standard; and if the prediction error does not meet the preset precision evaluation standard, fine-tuning the LSTM neural network model parameters until the prediction error meets the preset prediction precision evaluation standard.
The prediction accuracy evaluation standard of the LSTM neural network model is set as Root Mean Square Error (RMSE), and the formula is as follows:
where m is the total number of samples of the model test set, y_wire i For the storm tide level numerical model simulation value of sample i, y_pre i And (3) changing the number of LSTM hidden layers and the number of neurons for the storm tide level predicted value of the sample i, and searching the optimal network structure parameters.
Taking a plains pool station as an example, detailed experimental results of water increment forecast in advance for 12 hours are shown in the following table one, wherein the optimal model parameters are 2 hidden layers, 128 neurons in each layer, and the data set RMSE is 0.005m.
TABLE 1LSTM-12h experimental results
As an embodiment, the storm surge prediction is performed by using a trained storm surge prediction model, which comprises: inputting typhoon information to be predicted into a trained storm surge prediction model to obtain storm surge data at a corresponding moment to be predicted; and performing inverse normalization processing on the obtained storm surge data so as to obtain a corresponding storm surge water level predicted value.
That is, the trained LSTM neural network optimal model is adopted to predict a training set, a testing set and a verification set, the predicted data is subjected to inverse normalization processing, and error analysis is respectively carried out on the predicted data and the training set, the testing set and the verification set; as shown in FIG. 2, the error condition of the water increase forecast of the optimal model is that the error of the data set is not more than 0.1m, the error of most samples is within 0.05m, and the maximum error of the verification set is 0.075m. In addition, the actual typhoon process is predicted by adopting a trained LSTM neural network optimal model, and the typhoon of 6906 is taken as an example, the forecast water increase and the actual water increase are shown in the figure 3, and the duration of the forecast stage is less than 1 second.
In summary, according to the deep learning-based storm surge prediction method, a large number of supposed typhoons and storm surge are simulated by using a storm surge numerical model, so that the problems of lack of historical observation data and incomplete historical process of a measuring station are overcome, a sequence training set is utilized when a supposed typhoons and storm surge database is constructed, typhoons parameters of 24 hours in the future of typhoons are submitted and released by using weather forecast, and the constructed LSTM neural network model considers the influence of typhoons factor time sequence change on storm surge, and can perform 3-hour short-term forecast, 6-hour optimal path forecast, 12-hour shortest approach forecast and 24-hour long-term forecast, thereby improving the prediction accuracy and forecast aging.
In order to achieve the above embodiments, the embodiments of the present application provide a storm surge prediction system based on deep learning, as shown in fig. 3, including: the system comprises an acquisition module 10, a preprocessing module 20, a timing sequence construction module 30, a model construction module 40 and a model prediction module 50.
Wherein, the acquiring module 10 is configured to acquire a storm surge data set, where the storm surge data set includes a plurality of virtual typhoons and storm surge data files, and each virtual typhoons and storm surge data file includes typhoons information and storm surge information corresponding to each moment; the preprocessing module 20 is used for preprocessing data of typhoon information and storm surge information corresponding to each moment to obtain a preprocessed storm surge data set; the time sequence construction module 30 is configured to obtain a current time and a desired advanced forecast time, and take typhoon information corresponding to each of the current time to the desired advanced forecast time and storm surge information of the desired advanced forecast time as a time sequence, so as to construct a time sequence data set; the model construction module 40 is configured to construct a storm surge prediction model, and input the time-series data set into the storm surge prediction model for training, so that the model prediction module 50 performs storm surge prediction using the trained storm surge prediction model.
In some embodiments, the obtaining module 10 is further configured to obtain typhoon information corresponding to each moment in the historical multi-station typhoon, where the typhoon information includes latitude and longitude coordinates where a typhoon center is located, a maximum wind speed near the typhoon center, a movement speed of the typhoon center, a movement direction of the typhoon center, and an air pressure of the typhoon center; the typhoon information corresponding to each moment in the historical multi-station typhoons is transformed to generate a series of virtual typhoons, the generated series of virtual typhoons are input into a storm surge numerical model so as to obtain amplified series of virtual typhoons storm surge data files, and each data file comprises typhoon information corresponding to each moment and storm surge information at the same moment.
In some embodiments, the preprocessing module 20 is further configured to perform data cleaning on typhoon information and storm surge information corresponding to each moment; dividing the cleaned data according to different proportions so as to obtain a corresponding training set, a corresponding testing set and a corresponding verification set; and carrying out standardization processing on the divided data set by adopting a normalization method.
In some embodiments, the budget module 30 is further configured to obtain the current time and the desired advanced forecast time, and take typhoon information corresponding to each of the current time to the desired advanced forecast time and storm surge information of the desired advanced forecast time as a time sequence, so as to obtain a time sequence train set, a test set and a verification set.
In some embodiments, model building module 40 builds with an LSTM neural network, where parameters of the LSTM neural network are set to an input dimension of 5, an output dimension of 1, an initial learning rate of 0.001, a loss function of choice MAE, an optimization function of choice RMSprop, a callback function of Keras, reduce lronplateau, and used in conjunction with early stop function EarlyStopping in training; earlyStopping is used for training through 10 continuous epochs, and when the error of the verification set is reduced by not more than 0.001, the model stops training; the reduce lronplateau is used to reduce the learning rate by a factor of 10 when the error of the validation set is reduced by no more than 0.0001 after training of 5 epochs in succession.
In some embodiments, the model prediction module 50 is further configured to input typhoon information to be predicted into a trained storm surge prediction model to obtain storm surge data at a corresponding time to be predicted; and performing inverse normalization processing on the obtained storm surge data so as to obtain a corresponding storm surge water level predicted value.
It should be noted that the foregoing explanation of the embodiments of the storm surge prediction method based on deep learning is also applicable to the storm surge prediction system based on deep learning of this embodiment, and will not be repeated here.
In summary, according to the deep learning-based storm surge prediction system of the embodiment of the application, a storm surge data set is acquired through an acquisition module, wherein the storm surge data set comprises a plurality of virtual typhoons and storm surge data files, and each virtual typhoons and storm surge data file comprises typhoons information and storm surge information corresponding to each moment; the preprocessing module performs data preprocessing on typhoon information and storm surge information corresponding to each moment to obtain a preprocessed storm surge data set; the time sequence construction module acquires the current time and the required advanced forecast time, and takes typhoon information corresponding to each time from the current time to the required advanced forecast time and storm surge information of the required advanced forecast time as a time sequence so as to construct a time sequence data set; the model construction module constructs a storm surge prediction model, and inputs a time sequence data set into the storm surge prediction model for training so as to adopt the trained storm surge prediction model for storm surge prediction; thus, the prediction efficiency is high and the prediction period is long.
It will be appreciated by those skilled in the art that embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
It should be noted that in the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in a claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The application may be implemented by means of hardware comprising several distinct elements, and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means may be embodied by one and the same item of hardware. The use of the words first, second, third, etc. do not denote any order. These words may be interpreted as names.
While preferred embodiments of the present application have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. It is therefore intended that the following claims be interpreted as including the preferred embodiments and all such alterations and modifications as fall within the scope of the application.
It will be apparent to those skilled in the art that various modifications and variations can be made in the present application without departing from the spirit or scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the claims and the equivalents thereof, the present application is intended to cover such modifications and variations.
In the description of the present application, it should be understood that the terms "first," "second," and the like are used for descriptive purposes only and are not to be construed as indicating or implying a relative importance or an implicit indication of the number of technical features being indicated. Thus, a feature defining "a first" or "a second" may explicitly or implicitly include one or more such feature. In the description of the present application, the meaning of "a plurality" is two or more, unless explicitly defined otherwise.
In this application, unless specifically stated and limited otherwise, the terms "mounted," "connected," "secured," and the like are to be construed broadly, and may be, for example, fixedly connected, detachably connected, or integrally formed; can be mechanically or electrically connected; can be directly connected or indirectly connected through an intermediate medium, and can be communicated with the inside of two elements or the interaction relationship of the two elements. The specific meaning of the terms in this application will be understood by those of ordinary skill in the art as the case may be.
In this application, unless expressly stated or limited otherwise, a first feature "up" or "down" a second feature may be the first and second features in direct contact, or the first and second features in indirect contact via an intervening medium. Moreover, a first feature being "above," "over" and "on" a second feature may be a first feature being directly above or obliquely above the second feature, or simply indicating that the first feature is level higher than the second feature. The first feature being "under", "below" and "beneath" the second feature may be the first feature being directly under or obliquely below the second feature, or simply indicating that the first feature is less level than the second feature.
In the description of the present specification, a description referring to terms "one embodiment," "some embodiments," "examples," "specific examples," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the present application. In this specification, schematic representations of the above terms should not be understood as necessarily being directed to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, the different embodiments or examples described in this specification and the features of the different embodiments or examples may be combined and combined by those skilled in the art without contradiction.
Although embodiments of the present application have been shown and described above, it will be understood that the above embodiments are illustrative and not to be construed as limiting the application, and that variations, modifications, alternatives, and variations may be made to the above embodiments by one of ordinary skill in the art within the scope of the application.
Claims (10)
1. The storm surge water increasing prediction method based on deep learning is characterized by comprising the following steps of:
acquiring a storm surge data set, wherein the storm surge data set comprises a plurality of virtual typhoon storm surge data files, and each virtual typhoon storm surge data file comprises typhoon information and storm surge information corresponding to each moment;
carrying out data preprocessing on typhoon information and storm surge information corresponding to each moment to obtain a preprocessed storm surge data set;
acquiring the current time and the required advanced forecast time, and taking typhoon information corresponding to each time from the current time to the required advanced forecast time and storm surge information of the required advanced forecast time as a time sequence so as to construct a time sequence data set;
and constructing a storm surge water-increasing deep learning prediction model, and inputting the time sequence data set into the storm surge water-increasing deep learning prediction model for training so as to perform storm surge water-increasing prediction by adopting a trained storm surge water-increasing pre-deep learning prediction model.
2. The deep learning-based storm surge prediction method of claim 1 wherein obtaining a storm surge dataset comprises:
the method comprises the steps of obtaining typhoon information corresponding to each moment in historical multi-station typhoons, wherein the typhoons comprise longitude and latitude coordinates of a typhooncenter, a maximum wind speed near the typhooncenter, a typhooncenter moving speed, a typhooncenter moving direction and typhooncenter air pressure;
converting typhoon information corresponding to each moment in the historical multi-station typhoons to generate a series of virtual typhoons;
and inputting the generated series of virtual typhoons into a storm surge numerical model so as to obtain amplified series of virtual typhoons and storm surge data, wherein each data file comprises typhoon information corresponding to each moment and storm surge water increasing information at the same moment.
3. The deep learning-based storm surge prediction method according to claim 1, wherein the data preprocessing is performed on typhoon information and storm surge information corresponding to each moment to obtain a preprocessed storm surge data set, and the method comprises the steps of:
carrying out data cleaning on typhoon information and storm surge information corresponding to each moment;
dividing the cleaned data set according to different proportions so as to obtain a corresponding training set, a corresponding testing set and a corresponding verification set;
and carrying out standardization processing on the divided data set by adopting a normalization method.
4. A deep learning-based storm surge prediction method as claimed in claim 3 wherein constructing a time series data set for said preprocessed storm surge data set comprises obtaining a current time and a desired advance forecast time, and taking typhoon information corresponding to each of the current time to the desired advance forecast time and storm surge information of the desired advance forecast time as a time series to obtain a time series training set, a test set and a verification set.
5. The deep learning-based storm surge prediction method of claim 1, wherein the storm surge prediction model is constructed by adopting an LSTM neural network, the parameters of the LSTM neural network are set to be 6 in input dimension, 1 in output dimension, 0.001 in initial learning rate, MAE in loss function selection, RMSprop in optimization function selection, and a callback function ReduceLROnPlateau of Keras is used in training and matched with early stop function earlyStopping;
EarlyStopping is used for training through 10 continuous epochs, and when the error of the verification set is reduced by not more than 0.001, the model stops training; the reduce lronplateau is used to reduce the learning rate by a factor of 10 when the error of the validation set is reduced by no more than 0.0001 after training of 5 epochs in succession.
6. The deep learning-based storm surge prediction method of claim 1, wherein the storm surge prediction is performed by using a trained storm surge prediction model, comprising:
inputting typhoon information to be predicted into a trained storm surge prediction model to obtain storm surge data at a corresponding moment to be predicted;
and performing inverse normalization processing on the obtained storm surge data so as to obtain a corresponding storm surge water level predicted value.
7. Storm surge prediction system based on deep learning, characterized by comprising:
the system comprises an acquisition module, a storage module and a storage module, wherein the acquisition module is used for acquiring a storm surge data set, the storm surge data set comprises a plurality of virtual typhoons and storm surge data files, and each virtual typhoons and storm surge data file comprises typhoons information and storm surge information corresponding to each moment;
the preprocessing module is used for preprocessing the typhoon information and storm surge information corresponding to each moment in data so as to obtain a preprocessed storm surge data set;
the time sequence construction module is used for acquiring the current time and the required advanced forecast time, and taking typhoon information corresponding to each time from the current time to the required advanced forecast time and storm surge information of the required advanced forecast time as a time sequence so as to construct a time sequence data set;
the model construction module is used for constructing a storm surge moisturizing deep learning prediction model, and inputting the time sequence data set into the storm surge moisturizing prediction model for training so as to obtain a trained storm surge moisturizing deep learning prediction model;
the model prediction module is used for inputting typhoon information to be predicted into the trained storm surge water increasing deep learning prediction model so as to obtain a corresponding storm surge water increasing water level prediction value.
8. The deep learning based storm surge prediction system of claim 7 wherein said acquisition module is further configured to,
the method comprises the steps of obtaining typhoon information corresponding to each moment in historical multi-station typhoons, wherein the typhoons comprise longitude and latitude coordinates of a typhooncenter, a maximum wind speed near the typhooncenter, a typhooncenter moving speed, a typhooncenter moving direction and typhooncenter air pressure;
converting typhoon information corresponding to each moment in the historical multi-station typhoons to generate a series of virtual typhoons;
and inputting the generated series of virtual typhoons into a storm surge numerical model so as to obtain amplified series of virtual typhoons and storm surge data, wherein each data file comprises typhoon information corresponding to each moment and storm surge water increasing information at the same moment.
9. The deep learning based storm surge prediction system of claim 8 wherein said time series configuration module is further adapted to,
the method comprises the steps of obtaining the current time and the required advanced forecast time, and taking typhoon information corresponding to each time from the current time to the required advanced forecast time and storm surge information of the required advanced forecast time as a time sequence so as to obtain a time sequence train set, a test set and a verification set.
10. The deep learning-based storm surge prediction system according to claim 8, wherein the model construction module is constructed by adopting an LSTM neural network, the parameters of the LSTM neural network are set to be 6 in input dimension, 1 in output dimension, 0.001 in initial learning rate, MAE in loss function selection, RMSprop in optimization function selection, and a Keras callback function ReduceLROnPlateau and used in combination with early stop function EarlyStopping in training;
EarlyStopping is used for training through 10 continuous epochs, and when the error of the verification set is reduced by not more than 0.001, the model stops training; the reduce lronplateau is used to reduce the learning rate by a factor of 10 when the error of the validation set is reduced by no more than 0.0001 after training of 5 epochs in succession.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310279148.2A CN116467933A (en) | 2023-03-21 | 2023-03-21 | Storm surge water increasing prediction method and system based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310279148.2A CN116467933A (en) | 2023-03-21 | 2023-03-21 | Storm surge water increasing prediction method and system based on deep learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116467933A true CN116467933A (en) | 2023-07-21 |
Family
ID=87181538
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310279148.2A Pending CN116467933A (en) | 2023-03-21 | 2023-03-21 | Storm surge water increasing prediction method and system based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116467933A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116976227A (en) * | 2023-09-22 | 2023-10-31 | 河海大学 | Storm water increasing forecasting method and system based on LSTM machine learning |
-
2023
- 2023-03-21 CN CN202310279148.2A patent/CN116467933A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116976227A (en) * | 2023-09-22 | 2023-10-31 | 河海大学 | Storm water increasing forecasting method and system based on LSTM machine learning |
| CN116976227B (en) * | 2023-09-22 | 2023-12-08 | 河海大学 | Storm water increasing forecasting method and system based on LSTM machine learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Fadare | The application of artificial neural networks to mapping of wind speed profile for energy application in Nigeria | |
| CN107193060B (en) | A kind of multipath Typhoon Storm Surge Over method for quick predicting and system | |
| CN104932035A (en) | Typhoon intensity prediction method and system | |
| CN109146162A (en) | A kind of probability wind speed forecasting method based on integrated Recognition with Recurrent Neural Network | |
| CN117390894B (en) | Method for predicting extreme storm tide level | |
| CN115237896B (en) | Data preprocessing method and system based on deep learning forecast air quality | |
| CN115222163A (en) | Multi-factor medium-long term real-time forecasting method and system for harbor basin inlet waves and application | |
| CN114049545B (en) | Typhoon intensity determining method, system, equipment and medium based on point cloud voxels | |
| CN113128758B (en) | Maximum wave height forecasting system constructed based on offshore buoy wave observation data | |
| CN112036691A (en) | River mouth area extreme storm surge water level calculation method based on JPM-OS-Q model | |
| Zhang et al. | Analysis of 50-year wind data of the southern Baltic Sea for modelling coastal morphological evolution–a case study from the Darss-Zingst Peninsula | |
| CN112100911A (en) | Solar radiation prediction method based on deep BISLTM | |
| CN116467933A (en) | Storm surge water increasing prediction method and system based on deep learning | |
| CN114330110A (en) | Model training method, photovoltaic power generation power prediction method and system | |
| CN116090596A (en) | Typhoon intensity prediction method, typhoon intensity prediction device, typhoon intensity prediction equipment and storage medium | |
| CN111811465B (en) | Method for predicting sea wave effective wave height based on multi-sine function decomposition neural network | |
| CN116976227B (en) | Storm water increasing forecasting method and system based on LSTM machine learning | |
| CN117371303A (en) | Prediction method for effective wave height under sea wave | |
| CN115392595B (en) | Time-space short-term wind speed prediction method and system based on graph convolution neural network and Transformer | |
| CN112541620B (en) | Typhoon storm water increasing prediction method and system with high prediction precision and efficiency | |
| CN113792919A (en) | Wind power prediction method based on combination of transfer learning and deep learning | |
| Qiu et al. | A short-term precipitation prediction model based on spatiotemporal convolution network and ensemble empirical mode decomposition | |
| CN114004405A (en) | Photovoltaic power prediction method and system based on Elman neural network and satellite cloud picture | |
| Kim et al. | A bivariate frequency analysis of extreme Wave Heights and periods using a copula function in South Korea | |
| CN117421601B (en) | Sea surface evaporation waveguide near-future rapid forecasting method |
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 |