CN117540869A - Lightweight shipborne marine environment forecasting method and system and shipborne terminal - Google Patents
Lightweight shipborne marine environment forecasting method and system and shipborne terminal Download PDFInfo
- Publication number
- CN117540869A CN117540869A CN202311579790.9A CN202311579790A CN117540869A CN 117540869 A CN117540869 A CN 117540869A CN 202311579790 A CN202311579790 A CN 202311579790A CN 117540869 A CN117540869 A CN 117540869A
- Authority
- CN
- China
- Prior art keywords
- data
- wave
- field
- wind field
- forecasting
- 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
- 238000013277 forecasting method Methods 0.000 title claims abstract description 17
- 238000000034 method Methods 0.000 claims abstract description 50
- 238000012549 training Methods 0.000 claims description 37
- 238000004364 calculation method Methods 0.000 claims description 18
- 230000008569 process Effects 0.000 claims description 18
- 238000011835 investigation Methods 0.000 claims description 14
- 238000004458 analytical method Methods 0.000 claims description 12
- 238000012795 verification Methods 0.000 claims description 12
- 238000004422 calculation algorithm Methods 0.000 claims description 9
- 238000007781 pre-processing Methods 0.000 claims description 9
- 230000004044 response Effects 0.000 claims description 9
- 230000011218 segmentation Effects 0.000 claims description 9
- 238000005538 encapsulation Methods 0.000 claims description 8
- 238000005457 optimization Methods 0.000 claims description 7
- 238000004590 computer program Methods 0.000 claims description 6
- 238000011156 evaluation Methods 0.000 claims description 6
- 238000007499 fusion processing Methods 0.000 claims description 6
- 230000007774 longterm Effects 0.000 claims description 6
- 238000012360 testing method Methods 0.000 claims description 6
- 238000012544 monitoring process Methods 0.000 claims description 3
- 238000004891 communication Methods 0.000 description 8
- 238000013135 deep learning Methods 0.000 description 7
- 230000005540 biological transmission Effects 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 238000013473 artificial intelligence Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 230000001550 time effect Effects 0.000 description 3
- 238000004140 cleaning Methods 0.000 description 2
- 238000013527 convolutional neural network Methods 0.000 description 2
- 230000003111 delayed effect Effects 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 238000004806 packaging method and process Methods 0.000 description 2
- 208000037170 Delayed Emergence from Anesthesia Diseases 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 230000008034 disappearance Effects 0.000 description 1
- 238000004880 explosion Methods 0.000 description 1
- 238000009472 formulation Methods 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000012805 post-processing Methods 0.000 description 1
- 238000005086 pumping Methods 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01W—METEOROLOGY
- G01W1/00—Meteorology
- G01W1/10—Devices for predicting weather conditions
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/04—Inference or reasoning models
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Business, Economics & Management (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Health & Medical Sciences (AREA)
- Environmental & Geological Engineering (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Operations Research (AREA)
- Environmental Sciences (AREA)
- Ecology (AREA)
- Biodiversity & Conservation Biology (AREA)
- Atmospheric Sciences (AREA)
- General Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Quality & Reliability (AREA)
- Marketing (AREA)
- Entrepreneurship & Innovation (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application discloses a lightweight shipborne marine environment forecasting method, a lightweight shipborne marine environment forecasting system and a shipborne terminal. The method comprises the following steps: the shipborne terminal sends a data request instruction to a cloud server deployed on a shore base; receiving an initial wind field forecast data packet sent by a cloud server; decoding the initial wind field forecast data packet, and transcoding the initial wind field forecast data packet into a preset format to obtain transcoded initial wind field forecast data; inputting initial wind field forecast data into a pre-trained wind field reasoning model, and outputting optimized target wind field forecast data; generating corresponding initial wave field data according to the optimized target wind field forecast data by using a wind-wave relation empirical formula; and inputting the initial wave field data into a pre-trained wave field reasoning model, and outputting target wave field forecast data. The method has higher timeliness, the data downloading amount is small, the forecasting area is flexible and selectable, the operation speed is high, and timely and accurate refined guarantee data and products can be provided for ships.
Description
Technical Field
The application relates to the technical field of environment forecasting, in particular to a lightweight shipborne marine environment forecasting method, a lightweight shipborne marine environment forecasting system and a shipborne terminal.
Background
Timely and accurate marine environment element forecast is a basis for guaranteeing safe sailing of the marine vessel. At present, the way of acquiring marine environment forecast by domestic marine vessels mainly comprises two major categories, wherein commercial transportation vessels are mainly acquired by purchasing commercial weather navigation service, and business and scientific investigation vessels are acquired by forecast guarantee service provided by relevant marine weather forecast departments of China.
The commercial weather navigation service is a mature guarantee mode, marine environment forecast products are mainly manufactured aiming at preset airlines, the content and the form of the forecast products are fixed, and the forecast products are sent to a guarantee ship through maritime satellites in a mail mode. The national marine weather forecast department generally sends the marine environment forecast products which are manufactured on land to the official and scientific investigation ships, the forecast products mainly aim at the sea areas of the ships or preset airlines, the content of the products can be changed due to the change of the ship navigation plan, the ships and the land are generally required to communicate regularly, and the forecast products sent to the ships through maritime satellites can be used and interpreted by relevant responsible personnel, so that forecast guarantee service is realized.
The current mainstream ocean atmosphere forecasting mode is also a numerical forecasting method, and forecasting guarantee products provided by national ocean and weather forecasting departments to public affairs and scientific investigation ships are processed and manufactured based on numerical mode forecasting results; the business process is to obtain the initial forecasting field of the numerical mode from the Internet, then to read and process the initial field data and other observation data, to input the initial field data and other observation data into the numerical mode, to calculate the forecasting result of a certain area through the integral of the mode, and to perform the post-processing of the forecasting result in the modes of graphics and the like to obtain the usable forecasting product. The required time of the process is quite different according to the different computing capacities of the mainframe computers (clusters), generally 1-2 hours are needed, if the spatial resolution requirement on the forecasting result is higher, the numerical mode also needs to be nested in the regional grid, and the whole computing time is increased to 2-3 hours. In addition, the running of the wind wave intelligent forecasting system applied to the business scientific research at present not only needs to download wind field data, but also needs to download initial field data of sea waves, and the data downloading amount is large, so that the operation efficiency is influenced.
In addition, the manufactured forecast product is generally used for displaying marine environment elements on the sea area or the planned airlines of the ship in a picture form, and the product is single in form and relatively fixed in content. Although the data result of the regional forecast calculated by the numerical mode can be sent to the ship in the mode of satellite transmission at present, and then the display system at the ship end is used for completing data analysis and display, the forecast data is generally subjected to thin pumping and compression, the data resolution and accuracy can be obviously reduced, and the forecast region is fixed and cannot be customized independently and flexibly.
Disclosure of Invention
The application provides a lightweight shipborne marine environment forecasting method, a lightweight shipborne marine environment forecasting system and a shipborne terminal, and aims to solve the technical problems that an existing marine ship obtains a marine environment forecasting mode in a worse time effect, is large in data downloading amount and cannot be customized autonomously.
In a first aspect, a lightweight shipborne marine environment forecasting method is applied to a shipborne terminal, the shipborne terminal is deployed on a public service or scientific investigation ship, and the method comprises:
transmitting a data request instruction to a cloud server deployed on a shore base;
after the cloud server segments, compresses and encapsulates the latest initial wind field forecast data into data packets according to the data request instruction, the initial wind field forecast data packets sent by the cloud server are received;
decoding the initial wind field forecast data packet, and transcoding the initial wind field forecast data packet into a preset format to obtain transcoded initial wind field forecast data;
inputting the initial wind field forecast data into a pre-trained wind field reasoning model, and outputting optimized target wind field forecast data;
generating corresponding initial wave field data according to the optimized target wind field forecast data by using a wind-wave relation empirical formula;
and inputting the initial wave field data into a pre-trained wave field reasoning model, and outputting target wave field forecast data.
Optionally, the data request instruction includes region information, element information, forecast duration information, and spatio-temporal resolution information.
Optionally, the wind field reasoning model is trained by using a residual network model, and the wave field reasoning model is trained by using a long-short-term convolution network model.
Further optionally, the training process of the wind field reasoning model includes:
collecting historical wind field observation data and corresponding wind field forecast data of the global or sea area;
carrying out space-time fusion processing on the historical wind field observation data and the corresponding wind field forecast data to form a live analysis set;
dividing the live analysis set into a training set, a verification set and a test set according to a preset proportion;
training the residual network model by using the training set to obtain a wind field reasoning model; in the training process, continuously adjusting model parameters of a residual error network model through an optimization algorithm;
and evaluating the wind field reasoning model by using the verification set, and optimizing the wind field reasoning model according to an evaluation result.
Further optionally, the training process of the wave field reasoning model includes:
collecting historical wave field observation data and corresponding wave field forecast data of the global or sea area;
carrying out space-time fusion processing on the historical wave field observation data and the corresponding wave field forecast data to form a live analysis set;
dividing the live analysis set into a training set, a verification set and a test set according to a preset proportion;
training the long-term convolution network model by using the training set to obtain a wave field reasoning model; in the training process, continuously adjusting model parameters of a long-term and short-term convolution network model through an optimization algorithm;
and evaluating the wave field reasoning model by using the verification set, and optimizing the wave field reasoning model according to an evaluation result.
Optionally, the wind-wave relation empirical formula is obtained by fitting wind field observation data and sea wave field observation data.
Optionally, the method further comprises:
and drawing and outputting wind and wave forecasting products in various formats according to preset drawing parameters based on the target wind field forecasting data and the target wave field forecasting data.
In a second aspect, a lightweight shipborne marine environment forecasting system includes a forecasting data cloud service system and a shipborne intelligent forecasting system, wherein the forecasting data cloud service system operates on a cloud server deployed on a shore base, and the shipborne intelligent forecasting system operates on a shipborne terminal deployed on a public service or scientific investigation ship;
the forecast data cloud service system comprises a data acquisition and archiving module, a data request response module and a data segmentation and encapsulation module; the shipborne intelligent forecasting system comprises a data acquisition subsystem and a wave intelligent forecasting subsystem, wherein the data acquisition subsystem comprises a data customization request module and a data decoding and transcoding module, and the wave intelligent forecasting subsystem comprises a wind field intelligent reasoning and calculating module, a sea wave initial field preprocessing module and a wave field intelligent reasoning and calculating module;
the data acquisition and archiving module is used for regularly monitoring and scanning the data servers of the main stream forecasting institutions at home and abroad, acquiring and archiving the latest initial wind field forecasting data from the data servers of the main stream forecasting institutions at home and abroad; the data request response module is used for receiving a data request instruction sent by the shipborne terminal; the data segmentation and encapsulation module is used for segmenting, compressing and encapsulating the latest initial wind field forecast data into data packets according to the data request instruction, and sending the initial wind field forecast data packets to the shipboard terminal;
the data customization request module is used for sending a data request instruction to the cloud server; the data decoding and transcoding module is used for receiving the initial wind field forecast data packet sent by the cloud server, decoding the initial wind field forecast data packet, transcoding the initial wind field forecast data packet into a preset format, and obtaining transcoded initial wind field forecast data; the wind field intelligent reasoning calculation module is used for inputting the initial wind field forecast data into a pre-trained wind field reasoning model and outputting optimized target wind field forecast data; the sea wave initial field preprocessing module is used for generating corresponding initial sea wave field data according to the optimized target wind field forecast data by utilizing a wind-wave relation empirical formula; the wave field intelligent reasoning calculation module is used for inputting the initial wave field data into a pre-trained wave field reasoning model and outputting target wave field forecast data.
Optionally, the wind wave intelligent prediction subsystem further comprises a prediction element drawing module, wherein the prediction element drawing module is used for drawing and outputting wind wave prediction products in various formats according to preset drawing parameters based on the target wind field prediction data and the target sea wave field prediction data.
In a third aspect, an on-board terminal comprises a memory storing a computer program and a processor implementing the steps of the method of any one of the first aspects when the computer program is executed.
Compared with the prior art, the application has the following beneficial effects:
the most focused marine environment elements in the safety navigation guarantee of the marine vessel are wind and waves, the current application of the artificial intelligence-based deep learning network in marine weather is mature, the accuracy of intelligent forecasting results is very close to numerical mode forecasting, and the operation speed is hundreds of times faster than that of the numerical mode; according to the embodiment of the application, the lightweight wind wave intelligent forecasting method and system suitable for shipborne are developed based on the deep learning network model, compared with the wind wave intelligent forecasting model used by current business, the method and system do not need to download the initial field data of the sea wave, only need to download the wind field data, save the data downloading amount for the shipborne system, and improve the operation efficiency; according to the method, not only is the wind field data corrected and optimized, the wind field quality is improved, but also the sea wave initial field is calculated and obtained based on the wind wave relation formula fitted by the observation data of different sea areas, and the accuracy of the sea wave initial field is improved; in addition, according to the embodiment of the application, from the actual offshore guarantee service, elements such as a forecast area, a forecast duration and the like can be independently customized according to the data request instruction; the intelligent prediction model of the mature marine environment is improved to meet the application requirements of offshore business, and compared with a traditional numerical prediction system, the intelligent prediction model of the marine environment has higher timeliness, the data downloading amount is small, the prediction area is flexible and selectable, the operation speed is high, and timely and accurate refined guarantee data and products can be provided for ships.
Drawings
Fig. 1 is a schematic view of an application environment of a lightweight shipborne marine environment prediction method according to an embodiment of the present application;
FIG. 2 is a schematic flow chart of a lightweight shipborne marine environment prediction method according to an embodiment of the present disclosure;
FIG. 3 is a block diagram of a lightweight shipboard marine environment prediction system according to one embodiment of the present application;
FIG. 4 is an exemplary graph of a storm relational formulation fitted based on satellite scatterometer and altimeter observations in one embodiment of the application;
fig. 5 is an internal structural diagram of a shipboard terminal according to an embodiment of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application will be further described in detail with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the present application.
In the description of the present application: unless otherwise indicated, the terms "comprise," "include," "have" and the like are also intended to be "non-limiting" (certain units, components, materials, steps, etc.).
The lightweight shipborne marine environment forecasting method provided by the application can be applied to an application environment shown in fig. 1. Wherein the shipboard terminal 101 communicates with the cloud server 102 via a network. The shipboard terminal 101 may be, but not limited to, various personal computers, notebook computers, and tablet computers, and the cloud server 102 may be implemented as a stand-alone server or a server cluster composed of a plurality of servers.
In one embodiment, as shown in fig. 2, a lightweight shipborne marine environment forecasting method is provided, and the method is applied to the shipborne terminal in fig. 1 for illustration, and the shipborne terminal is deployed on a public service or scientific investigation ship, and the method comprises the following steps:
s1, the shipborne terminal sends a data request instruction to a cloud server deployed on a shore base.
The data request instruction comprises information such as region information, element information, forecast duration information, space-time resolution information and the like.
And S2, after the cloud server segments, compresses and packages the latest initial wind field forecast data into data packets according to the data request instruction, the shipboard terminal receives the initial wind field forecast data packets sent by the cloud server.
When the shipborne terminal sends a data request instruction to the cloud server, the cloud server can segment and compress latest forecast field data into a data packet according to the instruction information such as the area, elements, forecast duration, space-time resolution and the like of the data request, the data packet is sent to the terminal designated file by a satellite to complete data response, and data sending and receiving support breakpoint continuous transmission so as to adapt to the special situation that offshore communication is interrupted frequently.
And S3, decoding the initial wind field forecast data packet by the shipborne terminal, and transcoding the initial wind field forecast data packet into a preset format to obtain transcoded initial wind field forecast data.
The shipborne terminal decodes and transcodes the data packet, transcodes the compressed data into a required data format and provides initial field data for subsequent intelligent forecasting calculation.
And S4, the shipborne terminal inputs the initial wind field forecast data into a pre-trained wind field reasoning model, optimizes the initial wind field forecast data and outputs optimized target wind field forecast data.
The wind field reasoning environment is obtained through training in advance of a residual error network model. The Residual Network (Residual Network) is a deep learning neural Network structure, and the problems of gradient disappearance, gradient explosion and the like in deep Network training are solved by introducing a Residual block, so that a very deep Network model can be trained. In the method provided by the embodiment, the residual network model is used for training the wind field reasoning model so as to improve the prediction accuracy of wind field data. By using the wind field reasoning model, the intelligent forecasting of the wind field can be completed within 3-5 minutes, and the forecasting wind field data with higher accuracy can be output.
Further, the training process of the wind field reasoning model comprises the following steps:
(1) Collecting historical wind field observation data and corresponding wind field forecast data of the global or sea area; the historical wind field observation data and the corresponding wind field forecast data can have multiple sources, including satellite observation, buoy observation, weather station observation and the like, and the quality and the accuracy of the data need to be ensured in the data collection process;
(2) Processing historical wind field observation data and corresponding wind field forecast data; the processing firstly comprises operations of data cleaning, denoising, filling in missing values, standardization and the like, so as to improve the quality and usability of the data, and be better applied to model training; then carrying out space-time fusion processing on the data to form a live analysis set;
(3) Dividing live analysis set data into a training set, a verification set and a test set according to a preset proportion;
(4) Training the residual network model by using a training set to obtain a wind field reasoning model; in the training process, continuously adjusting model parameters of a residual error network model through an optimization algorithm; the network parameters can be continuously adjusted through a back propagation algorithm, so that the model can learn and accurately predict the change of the wind field; optimization algorithms such as random gradient descent (SGD) or variants thereof may be used to minimize the loss function;
for a residual network model, an appropriate loss function needs to be selected to measure the difference between the model output and the real forecast data, and common loss functions include Mean Square Error (MSE) and the like;
(5) Evaluating the wind field reasoning model by using the verification set, and optimizing the wind field reasoning model according to the evaluation result; for example, super parameters such as learning rate, regularization and the like are adjusted, so that the generalization capability of the model on new data is improved.
S5, generating corresponding initial wave field data according to the optimized target wind field forecast data by using a wind-wave relation empirical formula.
The wind-wave relation empirical formula is obtained by fitting wind field observation data and sea wave field observation data. By using the empirical formula of the wind-wave relation, the wind field can generate the sea wave initial field (the data comprises wave height, wave direction, wave period and other factors) of the forecasting area.
The wind-wave relation empirical formula is utilized to generate the sea wave initial field (data comprise wave height, wave direction, wave period and other factors) of the forecasting area from the optimized wind field, and the data do not need to be transmitted separately, so that the bandwidth of the shipborne client is saved.
S6, inputting the initial wave field data into a pre-trained wave field reasoning model, and outputting target wave field forecast data.
The wave field reasoning model is obtained by training a long-short-term convolution network model. The long-short term convolutional network (LSTM-CNN) is a deep learning model combining a long-short term memory network (LSTM) and a Convolutional Neural Network (CNN) and is used for processing time sequence data and extracting spatial features. By using the wave field reasoning model, the high-precision wave forecast field data can be output through 3-5 minutes of reasoning calculation.
Further, the training process of the wave field reasoning model comprises the following steps:
(1) Collecting historical wave field observation data and corresponding wave field forecast data of the global or sea area; such data may be obtained through satellite, buoy, or other observation devices;
(2) Processing historical wave field observation data and corresponding wave field forecast data; the method comprises the steps of data cleaning, normalization processing, time sequence data slicing and the like so as to ensure the quality of data and form suitable for model input; then carrying out space-time fusion processing on the data to form a live analysis set;
(3) Dividing live analysis set data into a training set, a verification set and a test set according to a preset proportion;
(4) Training the long-term convolution network model by using a training set to obtain a wave field reasoning model; in the training process, continuously adjusting model parameters of a long-term and short-term convolution network model through an optimization algorithm; inputting the prepared ocean wave field data into an LSTM-CNN model for training; in the training process, updating parameters through a back propagation algorithm and an optimizer, so that the model can gradually learn the space-time characteristics and the change rules in the ocean wave field data;
(5) Evaluating the wave field reasoning model by using a verification set, and optimizing the wave field reasoning model according to an evaluation result; in the tuning process, super parameters such as a model structure, a learning rate, regularization and the like can be considered to be adjusted.
Further, the method further comprises:
the shipborne terminal draws and outputs wind and wave forecasting products in various formats according to preset drawing parameters based on the target wind field forecasting data and the target wave field forecasting data.
After the target wind field forecast data and the target sea wave field forecast data are obtained, the shipborne terminal automatically identifies the factors such as the regional range, the space-time resolution and the like of the data, and according to the drawing parameters set in advance, the wind wave forecast products are manufactured, and forecast products with various formats such as png, pdf and the like are output.
In summary, the most focused marine environment elements in the safety navigation guarantee of the marine ship are wind and waves, and the embodiment of the application provides an intelligent wind wave forecasting method based on a deep learning network model. The running of the wind wave intelligent forecasting system applied to the business scientific research at present not only needs to download wind field data, but also needs to download initial field data of sea waves. By the method provided by the embodiment of the application, only wind field data is needed to be downloaded, and then the wind wave relation empirical formula is utilized to generate the sea wave initial field of the forecasting area. The wind-wave relation formula based on observation data fitting can be calculated according to different forecast sea areas, so that the accuracy of the initial sea wave field is improved.
Fig. 4 is a wind-wave relation formula of the northwest pacific ocean sea area fitted based on wind speed observed by a satellite scatterometer and wave height data observed by an altimeter, corresponding wave height data can be calculated by using the formula after wind field data of the area are obtained, and wave period data can be obtained by statistical fitting according to wave spectrum data observed by a satellite or a futon wind-wave relation, so that data of a wave initial field do not need to be downloaded and transmitted independently. The research and development work saves the data downloading amount for the shipborne system under the condition of ensuring the prediction accuracy, and improves the operation efficiency.
In the method, the rapid calculation forecasting service is realized on the open sea ship through the technologies of data segmentation and encapsulation, breakpoint continuous transmission and the like, and compared with the open sea ship which uses the traditional numerical forecasting method to develop forecasting guarantee, the method provided by the embodiment of the application has the advantages of high forecasting timeliness, small data downloading amount, flexible forecasting area, high calculating speed and the like.
At present, the prediction guarantee data and products obtained by the marine official and scientific investigation ships are calculated through numerical mode integration based on the obtained mode initial field data, and then the calculated prediction data or the further manufactured prediction products are transmitted to the ships, so that the time effect is relatively delayed, the prediction areas are relatively fixed, the form of the prediction products is relatively single, the autonomous flexible customization is not realized, and most importantly, the spatial resolution and the accuracy of the prediction data cannot meet the requirements of increasingly refined and intelligent prediction guarantee.
The application of the current artificial intelligence-based deep learning network in the marine weather is mature, the accuracy of intelligent forecasting results is quite close to that of numerical mode forecasting, and the operation speed is hundreds of times faster than that of the numerical mode. The method starts from improving the timeliness of the forecast products obtained from the offshore official scientific investigation ships and meeting the guarantee of the refined actual demands, and can finish the complete forecast service chain of initial data downloading, forecast calculation and product manufacturing on the ships. The method fully utilizes the neural network model, and can output the forecast wind field data or the forecast wave field data with higher accuracy in 3-5 minutes; compared with the existing method that the forecast data is obtained by numerical mode integral calculation based on the acquired mode initial field data, the operation time is greatly shortened, and the operation speed is improved.
According to the method provided by the embodiment of the application, from the actual guarantee business at sea, the factors such as the area of the forecast data, the forecast duration and the like can be customized autonomously, and the mature intelligent marine environment forecast model is improved to adapt to the application requirements of the offshore business, so that the method has higher timeliness than the traditional numerical forecast method, the data downloading amount is small, the forecast area is flexible and selectable, the operation speed is high, and timely and accurate refined guarantee data and products can be provided for ships.
In one embodiment, as shown in fig. 3, a lightweight shipborne marine environment forecasting system is provided, including a forecasting data cloud service system and a shipborne intelligent forecasting system, the forecasting data cloud service system operates on a cloud server deployed on a shore base, and the shipborne intelligent forecasting system operates on a shipborne terminal deployed on a ship;
the forecast data cloud service system comprises a data acquisition and archiving module, a data request response module and a data segmentation and encapsulation module; the shipborne intelligent forecasting system comprises a data acquisition subsystem and a wave intelligent forecasting subsystem, wherein the data acquisition subsystem comprises a data customization request module and a data decoding and transcoding module, and the wave intelligent forecasting subsystem comprises a wind field intelligent reasoning and calculating module, a wave initial field preprocessing module and a wave field intelligent reasoning and calculating module;
the data acquisition and archiving module is used for regularly monitoring and scanning the data servers of the main stream forecasting institutions at home and abroad, acquiring and archiving the latest initial wind field forecasting data from the data servers of the main stream forecasting institutions at home and abroad; the data request response module is used for receiving a data request instruction sent by the shipborne terminal; the data segmentation packaging module is used for segmenting, compressing and packaging the latest initial wind field forecast data into data packets according to the data request instruction, and sending the initial wind field forecast data packets to the shipborne terminal;
the data customization request module is used for sending a data request instruction to the cloud server; the data decoding and transcoding module is used for receiving the initial wind field forecast data packet sent by the cloud server, decoding the initial wind field forecast data packet, transcoding the initial wind field forecast data packet into a preset format, and obtaining transcoded initial wind field forecast data; the wind field intelligent reasoning calculation module is used for inputting the initial wind field forecast data into a pre-trained wind field reasoning model and outputting optimized target wind field forecast data; the sea wave initial field preprocessing module is used for generating corresponding initial sea wave field data according to the optimized target wind field forecast data by utilizing a wind-wave relation empirical formula; the wave field intelligent reasoning calculation module is used for inputting the initial wave field data into a pre-trained wave field reasoning model and outputting target wave field forecast data.
Further, the wind wave intelligent forecasting subsystem further comprises a forecasting element drawing module, wherein the forecasting element drawing module is used for drawing and outputting wind wave forecasting products in various formats according to preset drawing parameters based on target wind field forecasting data and target sea wave field forecasting data.
The specific implementation content of each module can be referred to above for limitation of a lightweight shipborne marine environment forecasting method, and will not be described herein.
In other words, the embodiment of the application provides a lightweight shipborne marine environment forecasting system applied to open sea security, which comprises a forecasting data cloud service system deployed on a shore base and a shipborne intelligent forecasting system deployed on a public service and scientific investigation ship, wherein the two systems can interact through satellite communication during offshore navigation operation.
The shore-based forecast data cloud service system comprises a data acquisition and archiving module, a data request response module and a data segmentation and encapsulation module, wherein the shipborne intelligent forecast system comprises a data acquisition subsystem and a wave intelligent forecast subsystem, the data acquisition subsystem comprises a data customization request module and a data decoding and transcoding module, and the wave intelligent forecast subsystem comprises a wind field intelligent reasoning and calculating module, a wave initial field preprocessing module, a wave field intelligent reasoning and calculating module and a forecast element drawing module.
The specific forecasting flow of the forecasting system is as follows:
the data acquisition and archiving module of the shore-based forecasting system monitors and scans the data servers of the main stream forecasting institutions at home and abroad at fixed time, acquires the latest wind field forecasting data from the Internet at the first time and archives the latest wind field forecasting data in the cloud server; when a data customizing request module of the shipborne intelligent forecasting system sends a data request instruction to a shore-based forecasting data cloud service system, the cloud service system can segment and compress latest forecasting field data into a data packet according to instruction information such as a data request area, elements, forecasting time length, space-time resolution and the like, the data packet is sent to a designated file catalog of the shipborne intelligent forecasting system through a satellite to complete data response, and data sending and receiving support breakpoint continuous transmission so as to adapt to special situations that offshore communication is interrupted frequently.
Then, the data acquisition subsystem of the shipborne terminal decodes and transcodes the data packet, and transcodes the compressed data into a data format required by the forecasting system; the transcoded wind field data can be imported into a wind wave intelligent forecasting subsystem to provide initial field data for subsequent intelligent forecasting calculation.
The wind field intelligent reasoning calculation module comprises a wind field reasoning environment which is trained in advance and aims at the global or sea area and is finished by utilizing a residual error network model, and after the wind field data which is finished in a transcoding mode are imported, intelligent prediction of the wind field can be finished only in 3-5 minutes and forecast wind field data with higher accuracy can be output; the method comprises the steps of leading a forecast wind field into a sea wave initial field preprocessing module, wherein the module contains a wind wave relation empirical formula which is statistically fitted according to observation data such as satellites and buoys, and the like, and the sea wave initial field (data comprise wave height, wave direction, wave period and other factors) of a forecast area can be generated by the wind field by using the formula without independently transmitting the data, so that the bandwidth of a ship-borne client is saved.
The wave field intelligent reasoning calculation module contains a global or sea area wave field reasoning environment which is trained by using a long-short-period convolution network model, and can output high-precision wave forecast field data after reading wave field matrix data and reasoning calculation for 3-5 minutes; finally, after the wind field and wave field data output by intelligent prediction are read in by the prediction element drawing module, elements such as the regional range, the space-time resolution and the like of the data can be automatically identified, wind wave prediction products are manufactured according to drawing parameters set in advance, and prediction products in various formats such as png, pdf and the like are output.
In summary, the most focused marine environment elements in the safety navigation guarantee of the marine vessel are wind and waves, and the embodiment of the application develops a lightweight wind wave intelligent forecasting system suitable for shipborne based on a deep learning network model. The running of the wind wave intelligent forecasting system applied to the business scientific research at present not only needs to download wind field data, but also needs to download initial field data of sea waves. The wind wave intelligent forecasting system develops the sea wave initial field preprocessing module, only needs to download wind field data, and generates the sea wave initial field of the forecasting area by using a wind wave relation empirical formula in the module. The wind-wave relation formula based on observation data fitting can be calculated according to different forecast sea areas, so that the accuracy of the initial sea wave field is improved.
The ship-borne intelligent forecasting system realizes the rapid calculation forecasting service on the open sea ship through the technologies of data segmentation and encapsulation, breakpoint continuous transmission and the like, and compared with the open sea ship which uses the traditional numerical forecasting method to develop forecasting guarantee, the lightweight ship-borne marine environment forecasting system provided by the embodiment of the application has the advantages of high forecasting timeliness, small data downloading amount, flexible forecasting area, high calculating speed and the like.
Table 1 Ship-borne forecasting system and numerical forecasting method
In table 1, the current business operation of the national ocean environment prediction center is compared with the current business operation of the ocean wave numerical prediction system in the North Pacific region, and under the condition that the time length, the resolution and other conditions of the lightweight shipborne prediction system are set to be consistent with those of the numerical prediction system on the same power platform, the lightweight shipborne prediction system has obvious advantages in the aspects of prediction timeliness, operation time length, data downloading amount and the like, and the prediction region and the space-time resolution can be flexibly customized according to the offshore field requirement, which is not possessed by the numerical prediction system.
At present, the prediction guarantee data and products obtained by the marine official and scientific investigation ships are calculated through numerical mode integration based on the obtained mode initial field data, and then the calculated prediction data or the further manufactured prediction products are transmitted to the ships, so that the time effect is relatively delayed, the prediction areas are relatively fixed, the form of the prediction products is relatively single, the autonomous flexible customization is not realized, and most importantly, the spatial resolution and the accuracy of the prediction data cannot meet the requirements of increasingly refined and intelligent prediction guarantee.
The application of the current artificial intelligence-based deep learning network in the marine weather is mature, the accuracy of intelligent forecasting results is quite close to that of numerical mode forecasting, and the operation speed is hundreds of times faster than that of the numerical mode. The invention starts from improving the timeliness of the forecast products obtained by the open sea official scientific investigation ships and meeting the actual demands of ensuring refinement, and develops a set of lightweight marine environment forecast system capable of carrying out rapid calculation on a common desktop computer or a portable computer by utilizing an intelligent forecast model, wherein the system can be deployed on various official, scientific investigation and military ships for executing open sea operation tasks, can complete forecast service chains of initial data downloading, forecast calculation and product making on the ships, has the characteristics of small data downloading amount, flexible and selectable forecast areas, high operation speed and the like, can provide timely and accurate refined forecast guarantee data and products for the ships, and remarkably improves the field forecast guarantee capability of the open sea ships in China.
The intelligent marine environment forecasting system for realizing the light weight for the first time in China is used for carrying out business application on open sea ships. The invention starts from the actual guarantee business at sea, can autonomously customize the factors such as the area of the forecast data, the forecast duration and the like, improves the mature intelligent marine environment forecast model to adapt to the application requirements of the offshore business, has higher timeliness compared with the traditional numerical forecast system, has small data downloading amount, flexible and selectable forecast area and high operation speed, and can provide accurate and precise guarantee data and products for ships.
In one embodiment, a ship-borne terminal is provided, the internal structure of which may be as shown in fig. 5. The shipboard terminal comprises a processor, a memory, a communication interface, a display screen and an input device which are connected through a system bus. The processor of the shipboard terminal is used for providing computing and control capabilities, the communication interface is used for carrying out wired or wireless communication with an external terminal, and the wireless communication can be realized through WIFI, an operator network, NFC (near field communication) or other technologies. The on-board terminal runs the computer program by loading to realize the lightweight on-board marine environment forecasting method. The display screen of the shipborne terminal can be a liquid crystal display screen or an electronic ink display screen, and the input device of the shipborne terminal can be a touch layer covered on the display screen, can also be a key, a track ball or a touch pad arranged on the shell of the shipborne terminal, and can also be an external keyboard, a touch pad or a mouse and the like.
It will be appreciated by those skilled in the art that the structure shown in fig. 5 is merely a block diagram of a portion of the structure associated with the present application and is not intended to limit the on-board terminal to which the present application is applied, and that a particular on-board terminal may include more or fewer components than shown, or may combine some of the components, or may have a different arrangement of components.
In an embodiment, a computer readable storage medium is also provided, on which a computer program is stored, involving all or part of the flow of the method of the above embodiment.
The technical features of the above embodiments may be arbitrarily combined, and all possible combinations of the technical features in the above embodiments are not described for brevity of description, however, as long as there is no contradiction between the combinations of the technical features, they should be considered as the scope of the description.
Claims (10)
1. A lightweight shipborne marine environment forecasting method, characterized by being applied to a shipborne terminal deployed on a public service or scientific investigation vessel, comprising:
transmitting a data request instruction to a cloud server deployed on a shore base;
after the cloud server segments, compresses and encapsulates the latest initial wind field forecast data into data packets according to the data request instruction, the initial wind field forecast data packets sent by the cloud server are received;
decoding the initial wind field forecast data packet, and transcoding the initial wind field forecast data packet into a preset format to obtain transcoded initial wind field forecast data;
inputting the initial wind field forecast data into a pre-trained wind field reasoning model, and outputting optimized target wind field forecast data;
generating corresponding initial wave field data according to the optimized target wind field forecast data by using a wind-wave relation empirical formula;
and inputting the initial wave field data into a pre-trained wave field reasoning model, and outputting target wave field forecast data.
2. The method of claim 1, wherein the data request instructions include regional information, element information, forecast duration information, and spatiotemporal resolution information.
3. The method for forecasting the light-weight shipborne marine environment according to claim 1, wherein the wind field reasoning model is trained by using a residual network model, and the wave field reasoning model is trained by using a long-short-term convolution network model.
4. A lightweight shipborne marine environment prediction method as claimed in claim 3, wherein the training process of the wind farm reasoning model comprises:
collecting historical wind field observation data and corresponding wind field forecast data of the global or sea area;
carrying out space-time fusion processing on the historical wind field observation data and the corresponding wind field forecast data to form a live analysis set;
dividing the live analysis set into a training set, a verification set and a test set according to a preset proportion;
training the residual network model by using the training set to obtain a wind field reasoning model; in the training process, continuously adjusting model parameters of a residual error network model through an optimization algorithm;
and evaluating the wind field reasoning model by using the verification set, and optimizing the wind field reasoning model according to an evaluation result.
5. A lightweight shipborne marine environment prediction method as claimed in claim 3, wherein the training process of the wave field inference model comprises:
collecting historical wave field observation data and corresponding wave field forecast data of the global or sea area;
carrying out space-time fusion processing on the historical wave field observation data and the corresponding wave field forecast data to form a live analysis set;
dividing the live analysis set into a training set, a verification set and a test set according to a preset proportion;
training the long-term convolution network model by using the training set to obtain a wave field reasoning model; in the training process, continuously adjusting model parameters of a long-term and short-term convolution network model through an optimization algorithm;
and evaluating the wave field reasoning model by using the verification set, and optimizing the wave field reasoning model according to an evaluation result.
6. The method for forecasting the light-weight shipborne marine environment according to claim 1, wherein the wind-wave relation empirical formula is obtained by fitting wind field observation data and sea wave field observation data.
7. The lightweight shipboard marine environment prediction method of claim 1, further comprising:
and drawing and outputting wind and wave forecasting products in various formats according to preset drawing parameters based on the target wind field forecasting data and the target wave field forecasting data.
8. The lightweight shipborne marine environment forecasting system is characterized by comprising a forecasting data cloud service system and a shipborne intelligent forecasting system, wherein the forecasting data cloud service system operates on a cloud server deployed on a shore base, and the shipborne intelligent forecasting system operates on a shipborne terminal deployed on a public service or scientific investigation ship;
the forecast data cloud service system comprises a data acquisition and archiving module, a data request response module and a data segmentation and encapsulation module; the shipborne intelligent forecasting system comprises a data acquisition subsystem and a wave intelligent forecasting subsystem, wherein the data acquisition subsystem comprises a data customization request module and a data decoding and transcoding module, and the wave intelligent forecasting subsystem comprises a wind field intelligent reasoning and calculating module, a sea wave initial field preprocessing module and a wave field intelligent reasoning and calculating module;
the data acquisition and archiving module is used for regularly monitoring and scanning the data servers of the main stream forecasting institutions at home and abroad, acquiring and archiving the latest initial wind field forecasting data from the data servers of the main stream forecasting institutions at home and abroad; the data request response module is used for receiving a data request instruction sent by the shipborne terminal; the data segmentation and encapsulation module is used for segmenting, compressing and encapsulating the latest initial wind field forecast data into data packets according to the data request instruction, and sending the initial wind field forecast data packets to the shipboard terminal;
the data customization request module is used for sending a data request instruction to the cloud server; the data decoding and transcoding module is used for receiving the initial wind field forecast data packet sent by the cloud server, decoding the initial wind field forecast data packet, transcoding the initial wind field forecast data packet into a preset format, and obtaining transcoded initial wind field forecast data; the wind field intelligent reasoning calculation module is used for inputting the initial wind field forecast data into a pre-trained wind field reasoning model and outputting optimized target wind field forecast data; the sea wave initial field preprocessing module is used for generating corresponding initial sea wave field data according to the optimized target wind field forecast data by utilizing a wind-wave relation empirical formula; the wave field intelligent reasoning calculation module is used for inputting the initial wave field data into a pre-trained wave field reasoning model and outputting target wave field forecast data.
9. The marine environment prediction system according to claim 8, wherein the wind and wave intelligent prediction subsystem further comprises a prediction element drawing module, and the prediction element drawing module is used for drawing and outputting wind and wave prediction products in multiple formats according to preset drawing parameters based on the target wind field prediction data and the target sea wave field prediction data.
10. An on-board terminal comprising a memory and a processor, the memory storing a computer program, characterized in that the processor implements the steps of the method of any one of claims 1 to 7 when the computer program is executed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311579790.9A CN117540869A (en) | 2023-11-24 | 2023-11-24 | Lightweight shipborne marine environment forecasting method and system and shipborne terminal |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311579790.9A CN117540869A (en) | 2023-11-24 | 2023-11-24 | Lightweight shipborne marine environment forecasting method and system and shipborne terminal |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117540869A true CN117540869A (en) | 2024-02-09 |
Family
ID=89795493
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311579790.9A Pending CN117540869A (en) | 2023-11-24 | 2023-11-24 | Lightweight shipborne marine environment forecasting method and system and shipborne terminal |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117540869A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117875194A (en) * | 2024-03-13 | 2024-04-12 | 青岛哈尔滨工程大学创新发展中心 | Regional sea wave field intelligent construction method and system based on small quantity of real sea observation data |
| CN118297443A (en) * | 2024-06-06 | 2024-07-05 | 国家海洋环境预报中心 | Ocean environment forecasting method, system and equipment |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105787281A (en) * | 2016-03-24 | 2016-07-20 | 国家海洋环境预报中心 | Fusion inversion method and device for sea wave significant wave height field |
| CN114580509A (en) * | 2022-02-21 | 2022-06-03 | 国家海洋环境预报中心 | Sea wave macroscopic characteristic quantity prediction system based on convolution length memory network |
| CN115392589A (en) * | 2022-09-16 | 2022-11-25 | 南京信息工程大学 | Sea wave height forecasting method and system |
| CN115600682A (en) * | 2022-01-28 | 2023-01-13 | 国家海洋环境预报中心（Cn） | Data-driven sea wave height field forecasting method and device |
| CN117082474A (en) * | 2023-10-17 | 2023-11-17 | 国家海洋局北海预报中心((国家海洋局青岛海洋预报台)(国家海洋局青岛海洋环境监测中心站)) | System for acquiring marine environment forecast data in real time by scientific investigation ship |
-
2023
- 2023-11-24 CN CN202311579790.9A patent/CN117540869A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105787281A (en) * | 2016-03-24 | 2016-07-20 | 国家海洋环境预报中心 | Fusion inversion method and device for sea wave significant wave height field |
| CN115600682A (en) * | 2022-01-28 | 2023-01-13 | 国家海洋环境预报中心（Cn） | Data-driven sea wave height field forecasting method and device |
| CN114580509A (en) * | 2022-02-21 | 2022-06-03 | 国家海洋环境预报中心 | Sea wave macroscopic characteristic quantity prediction system based on convolution length memory network |
| CN115392589A (en) * | 2022-09-16 | 2022-11-25 | 南京信息工程大学 | Sea wave height forecasting method and system |
| CN117082474A (en) * | 2023-10-17 | 2023-11-17 | 国家海洋局北海预报中心((国家海洋局青岛海洋预报台)(国家海洋局青岛海洋环境监测中心站)) | System for acquiring marine environment forecast data in real time by scientific investigation ship |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117875194A (en) * | 2024-03-13 | 2024-04-12 | 青岛哈尔滨工程大学创新发展中心 | Regional sea wave field intelligent construction method and system based on small quantity of real sea observation data |
| CN117875194B (en) * | 2024-03-13 | 2024-05-28 | 青岛哈尔滨工程大学创新发展中心 | Regional sea wave field intelligent construction method and system based on small quantity of real sea observation data |
| CN118297443A (en) * | 2024-06-06 | 2024-07-05 | 国家海洋环境预报中心 | Ocean environment forecasting method, system and equipment |
| CN118297443B (en) * | 2024-06-06 | 2024-08-20 | 国家海洋环境预报中心 | Ocean environment forecasting method, system and equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN117540869A (en) | Lightweight shipborne marine environment forecasting method and system and shipborne terminal | |
| US11172045B2 (en) | Spatio-temporal data processing systems and methods | |
| US12078775B2 (en) | Ocean weather forecasting system | |
| CN109902885B (en) | Typhoon prediction method based on deep learning mixed CNN-LSTM model | |
| Xu et al. | On-board real-time ship detection in HISEA-1 SAR images based on CFAR and lightweight deep learning | |
| Fedak et al. | Overcoming the constraints of long range radio telemetry from animals: getting more useful data from smaller packages | |
| CN113156439B (en) | SAR wind field and sea wave joint inversion method and system based on data driving | |
| CN110487981B (en) | Red tide monitoring system and method | |
| CN111089591A (en) | Ship route planning method, system, terminal and storage medium | |
| CN110826693A (en) | Three-dimensional atmospheric temperature profile inversion method and system based on DenseNet convolutional neural network | |
| CN106959623B (en) | Ship hydrometeorology data transfer control system and method based on Beidou | |
| US10339691B2 (en) | Storage of time-varying data as parametrized functions | |
| CN112861355B (en) | Sea surface net radiation estimation method and system | |
| Schwartz et al. | On-board satellite data processing to achieve smart information collection | |
| Le Marshall et al. | The joint center for satellite data assimilation | |
| Zhou et al. | Integrated visualization approach for real-time and dynamic assessment of storm surge disasters for China’s seas | |
| CN113848596A (en) | Visual meteorological early warning system of big dipper shipborne terminal | |
| Shams Eddin et al. | Focal-TSMP: Deep learning for vegetation health prediction and agricultural drought assessment from a regional climate simulation | |
| Sandner et al. | A Multiple-Location Modeling Scheme for Physics-Regularized Networks: Recurrent Forecasting of Fixed-Location Buoy Observations | |
| CN115755227B (en) | Three-dimensional radar extrapolation method based on deep neural network model | |
| Wang | Research on ship berthing monitoring technology based on 3D laser point cloud data | |
| CN212059396U (en) | Unmanned inspection sampling system on sea | |
| Li et al. | NeMO-Net-The Neural Multi-Modal Observation & Training Network for Global Coral Reef Assessment | |
| RU2822145C2 (en) | Device and method of determining yield of agricultural crops | |
| Liu et al. | Optimal configuration method of sampling points based on variability of sea surface temperature |
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 |