CN114742307A - Wave element prediction method and system - Google Patents
Wave element prediction method and system Download PDFInfo
- Publication number
- CN114742307A CN114742307A CN202210416052.1A CN202210416052A CN114742307A CN 114742307 A CN114742307 A CN 114742307A CN 202210416052 A CN202210416052 A CN 202210416052A CN 114742307 A CN114742307 A CN 114742307A
- Authority
- CN
- China
- Prior art keywords
- prediction
- model
- prediction model
- value
- sequence
- 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
- 238000000034 method Methods 0.000 title claims abstract description 42
- 238000012795 verification Methods 0.000 claims abstract description 24
- 238000012937 correction Methods 0.000 claims abstract description 18
- 238000012545 processing Methods 0.000 claims abstract description 17
- 238000010801 machine learning Methods 0.000 claims abstract description 15
- 238000005457 optimization Methods 0.000 claims abstract description 10
- 239000011159 matrix material Substances 0.000 claims description 28
- 238000003860 storage Methods 0.000 claims description 12
- 210000002569 neuron Anatomy 0.000 claims description 11
- 238000013528 artificial neural network Methods 0.000 claims description 6
- 238000005259 measurement Methods 0.000 claims description 6
- 230000001537 neural effect Effects 0.000 claims description 5
- 238000010606 normalization Methods 0.000 claims description 5
- 238000004364 calculation method Methods 0.000 claims description 4
- 238000012549 training Methods 0.000 claims description 3
- 238000000638 solvent extraction Methods 0.000 claims 1
- 238000010200 validation analysis Methods 0.000 claims 1
- 230000000694 effects Effects 0.000 abstract description 11
- 230000006870 function Effects 0.000 description 7
- 238000011160 research Methods 0.000 description 7
- 238000004422 calculation algorithm Methods 0.000 description 6
- 238000004590 computer program Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 239000000463 material Substances 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000001228 spectrum Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000003628 erosive effect Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 239000000047 product Substances 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
- 230000001502 supplementing effect Effects 0.000 description 1
Images
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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- 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
-
- 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
- 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
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Economics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Strategic Management (AREA)
- Human Resources & Organizations (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- Marketing (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Public Health (AREA)
- Primary Health Care (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Water Supply & Treatment (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a wave element prediction method and a system, wherein the method comprises the following steps: acquiring wind field information; performing pre-prediction according to the wind field information to perform optimization verification on an initial model to obtain a set prediction model; performing formal prediction according to the set prediction model to obtain a prediction sequence value; checking the prediction sequence value to determine a prediction result meeting the verification requirement; carrying out prediction result classification processing on the prediction result through SOM machine learning to obtain prediction model sets with different precisions; selecting a target precision prediction model from the precision prediction model set, and adding a correction value to a prediction result of the target precision prediction model to serve as a final prediction result of the output wave element; wherein the wave element prediction result includes, but is not limited to, effective wave height data, wave crest period data, and wave direction data. The method has small error and good prediction effect, and can be widely applied to the technical field of data processing.
Description
Technical Field
The invention relates to the technical field of data processing, in particular to a wave element prediction method and a wave element prediction system.
Background
The importance of waves comes from different studies and applications occurring at sea, such as navigation safety, ship routes, oil and gas production and transport, resulting from the use of platforms and pipelines, respectively. In addition, waves have important effects on offshore works, such as port operations, port ship stability, coastal vulnerability assessment, coastal structure design and verification. The potential increase, decrease and breaking of waves are key factors in the variation of the erosion process of the coast and are also key factors in the actual performance of the device for converting wave energy into electric energy.
The existing wave prediction means is single. For example, single deterministic prediction is adopted, the prediction effect of the prediction means in 24 hours initially meets the requirement, but the prediction effect in a long-term period is poor, the actual requirement is difficult to meet, and the prediction means has specific use targets and working environments. Machine learning predictions can lead to increasing errors as the prediction period becomes longer. Most of the prior art focuses on the prediction of the effective wave height, and in actual engineering, data such as wave wind period, wave direction and the like are also important, but the prior art cannot predict the effective wave height.
Disclosure of Invention
In view of this, embodiments of the present invention provide a wave element prediction method and system with small error and good prediction effect, which can predict more wave element data such as wave wind period and wave direction.
The first aspect of the present invention provides a wave element prediction method, including:
acquiring wind field information;
performing pre-prediction according to the wind field information to perform optimization verification on an initial model to obtain a set prediction model;
performing formal prediction according to the set prediction model to obtain a prediction sequence value;
checking the prediction sequence value to determine a prediction result meeting the verification requirement;
carrying out prediction result classification processing on the prediction result through SOM machine learning to obtain prediction model sets with different precisions;
selecting a target precision prediction model from the precision prediction model set, and adding a correction value to a prediction result of the target precision prediction model to serve as a final prediction result of the output wave element;
wherein the wave element prediction result includes, but is not limited to, effective wave height data, wave crest period data, and wave direction data.
Optionally, the performing, according to the wind field information, prediction in advance to perform optimization verification on the initial model to obtain an aggregate prediction model includes:
constructing a deterministic SWAN model of a first area, and nesting a plurality of collective SWAN models in the deterministic SWAN model; wherein the deterministic SWAN model provides boundary conditions for each collective SWAN model;
forecasting training is carried out on each set SWAN model according to the sea surface disturbance forecast wind field information, a forecasting sequence value in a period of time is obtained, and a corresponding actual measurement sequence value in the period of time is selected;
comparing the predicted sequence value with the actually measured sequence value through a correlation coefficient calculation formula to obtain a judgment coefficient value;
determining a set SWAN model meeting the requirement of a correlation coefficient according to the judgment coefficient value;
calculating the average value of predicted sequence values obtained by each set SWAN model meeting the requirement of a correlation coefficient, and comparing the average value with the average value of the actually measured sequence values to obtain an average deviation value;
calculating to obtain a correction value according to the average deviation value;
and correcting each set SWAN model meeting the requirement of the correlation coefficient according to the correction value to obtain a final prediction result of the set prediction model.
Optionally, the performing formal prediction according to the set prediction model to obtain a prediction sequence value includes:
when the pre-prediction of the set prediction model meets the preset requirement, listing the set prediction model into a formal prediction model;
and taking the set prediction model as a starting point of the formal prediction, selecting a plurality of pieces of sea surface disturbance forecast wind field information as a drive of the formal prediction, and acquiring an uncertainty forecast result as a prediction sequence value.
Optionally, the checking the predicted sequence value to determine a predicted result meeting the verification requirement includes:
acquiring a cycle of effective wave height sequence value observed in an actual region by a high-frequency ground wave radar;
calculating a root mean square error according to the effective wave height sequence value and the predicted sequence value;
calculating a correlation coefficient according to the significant wave height sequence value and the prediction sequence value;
and carrying out condition judgment according to the root mean square error and the correlation coefficient, and determining a prediction result meeting the verification requirement.
Optionally, the classifying the prediction result of the prediction result through SOM machine learning to obtain prediction model sets with different accuracies includes:
constructing an attribute matrix of the effective wave height sequence according to the root mean square error, the correlation coefficient and the prediction results of the plurality of set prediction models;
normalizing the attribute matrix;
and clustering and dividing the attribute matrix after normalization processing according to the correlation coefficient and the root mean square error to obtain prediction model sets with different precisions.
Optionally, the performing cluster division on the normalized attribute matrix according to the correlation coefficient and the root mean square error to obtain prediction model sets with different accuracies includes:
initializing 30 neuron node weights, randomly selecting an attribute sequence of a certain row of the effective wave height from the attribute matrix of the sample data, and deleting the attribute sequence of the row from the attribute matrix of the sample data;
and (3) calculating the Euclidean distance between the sample data and the weight of each neural network node to obtain a superior neuron node, updating the weight of each neural node according to the superior neuron node, and processing each attribute sequence in the sample data attribute matrix until the sample sequence is an empty set.
Optionally, the method further comprises the step of updating the prediction model in real time, the step comprising at least one of:
updating the input wind field information in real time, and deleting the current prediction model and practical wind field data of the current prediction model when the prediction result of the prediction model does not correspond to the real-time input wind field information;
or acquiring actual measurement data of the high-frequency ground wave radar in real time, then calculating a correlation coefficient and a root mean square error, and updating a prediction model in real time according to the correlation coefficient and the root mean square error;
or predicting the real-time updating result in advance, deleting the model when the model does not meet the correlation coefficient range, and reselecting a disturbance forecasting wind field to predict the model;
or, the formal prediction is carried out through the prediction model, the check coefficient is calculated, when the formal prediction model does not meet the range of the correlation coefficient and the root mean square error, the model is deleted, and simultaneously, the wind field data used by the model and the pre-prediction model are deleted.
Another aspect of the embodiments of the present invention further provides a wave element prediction system, including:
the first module is used for acquiring wind field information;
the second module is used for carrying out prediction in advance according to the wind field information to carry out optimization verification on the initial model so as to obtain a set prediction model;
a third module, configured to perform formal prediction according to the set prediction model to obtain a prediction sequence value;
a fourth module, configured to check the predicted sequence value, and determine a predicted result that meets the verification requirement;
the fifth module is used for carrying out prediction result classification processing on the prediction result through SOM machine learning to obtain prediction model sets with different precisions;
a sixth module, configured to select a target accuracy prediction model from the accuracy prediction model set, and add a correction value to a prediction result of the target accuracy prediction model to obtain a final prediction result of the output wave element;
wherein the wave element prediction result includes, but is not limited to, effective wave height data, wave crest period data, and wave direction data.
Another aspect of the embodiments of the present invention further provides an electronic device, including a processor and a memory;
the memory is used for storing programs;
the processor executes the program to implement the method as described above.
Yet another aspect of the embodiments of the present invention provides a computer-readable storage medium, which stores a program, which is executed by a processor to implement the method as described above.
The embodiment of the invention also discloses a computer program product or a computer program, which comprises computer instructions, and the computer instructions are stored in a computer readable storage medium. The computer instructions may be read by a processor of a computer device from a computer-readable storage medium, and the computer instructions executed by the processor cause the computer device to perform the foregoing method.
The embodiment of the invention acquires wind field information; performing pre-prediction according to the wind field information to perform optimization verification on an initial model to obtain a set prediction model; performing formal prediction according to the set prediction model to obtain a prediction sequence value; checking the prediction sequence value to determine a prediction result meeting the verification requirement; carrying out prediction result classification processing on the prediction result through SOM machine learning to obtain prediction model sets with different precisions; selecting a target precision prediction model from the precision prediction model set, and adding a correction value to a prediction result of the target precision prediction model to serve as a final prediction result of the output wave element; wherein the wave element prediction result includes, but is not limited to, effective wave height data, wave crest period data, and wave direction data. The method has small error and good prediction effect.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the description below are only some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings without creative efforts.
FIG. 1 is a flowchart illustrating the overall steps provided by an embodiment of the present invention;
fig. 2 is a flowchart illustrating a concept of a wave element prediction method according to an embodiment of the present invention;
FIG. 3 is a flow chart of a set prediction process provided by an embodiment of the present invention;
fig. 4 is a flowchart of implementing SOM learning classification according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
Aiming at the problems in the prior art, the embodiment of the invention provides a wave element prediction method, which comprises the following steps:
acquiring wind field information;
performing pre-prediction according to the wind field information to perform optimization verification on an initial model to obtain a set prediction model;
performing formal prediction according to the set prediction model to obtain a prediction sequence value;
checking the prediction sequence value to determine a prediction result meeting the verification requirement;
carrying out prediction result classification processing on the prediction result through SOM machine learning to obtain prediction model sets with different precisions;
selecting a target precision prediction model from the precision prediction model set, and adding a correction value to a prediction result of the target precision prediction model to serve as a final prediction result of the output wave element;
wherein the wave element prediction result includes, but is not limited to, effective wave height data, wave crest period data, and wave direction data.
Optionally, the performing, according to the wind field information, prediction in advance to perform optimization verification on the initial model to obtain an aggregate prediction model includes:
constructing a deterministic SWAN model of a first area, and nesting a plurality of collective SWAN models in the deterministic SWAN model; wherein the deterministic SWAN model provides boundary conditions for each collective SWAN model;
forecasting training is carried out on each set SWAN model according to the sea surface disturbance forecast wind field information, a forecasting sequence value in a period of time is obtained, and a corresponding actual measurement sequence value in the period of time is selected;
comparing the predicted sequence value with the actually measured sequence value through a correlation coefficient calculation formula to obtain a judgment coefficient value;
determining a set SWAN model meeting the requirement of a correlation coefficient according to the judgment coefficient value;
calculating the average value of predicted sequence values obtained by each set SWAN model meeting the requirement of a correlation coefficient, and comparing the average value with the average value of the actually measured sequence values to obtain an average deviation value;
calculating to obtain a correction value according to the average deviation value;
and correcting each set SWAN model meeting the requirement of the correlation coefficient according to the correction value to obtain a final prediction result of the set prediction model.
Optionally, the performing formal prediction according to the set prediction model to obtain a prediction sequence value includes:
when the pre-prediction of the set prediction model meets the preset requirement, listing the set prediction model into a formal prediction model;
and taking the set prediction model as a starting point of the formal prediction, selecting a plurality of pieces of sea surface disturbance forecast wind field information as a drive of the formal prediction, and acquiring an uncertainty forecast result as a prediction sequence value.
Optionally, the checking the predicted sequence value to determine a predicted result meeting the verification requirement includes:
acquiring a cycle of effective wave height sequence value observed in an actual region by a high-frequency ground wave radar;
calculating a root mean square error according to the effective wave height sequence value and the predicted sequence value;
calculating a correlation coefficient according to the effective wave height sequence value and the prediction sequence value;
and carrying out condition judgment according to the root-mean-square error and the correlation coefficient, and determining a prediction result meeting the verification requirement.
Optionally, the classifying the prediction result of the prediction result through SOM machine learning to obtain prediction model sets with different accuracies includes:
constructing an attribute matrix of the effective wave height sequence according to the root mean square error, the correlation coefficient and the prediction results of the plurality of set prediction models;
normalizing the attribute matrix;
and clustering and dividing the attribute matrix after the normalization processing according to the correlation coefficient and the root mean square error to obtain prediction model sets with different precisions.
Optionally, the performing cluster division on the normalized attribute matrix according to the correlation coefficient and the root mean square error to obtain prediction model sets with different accuracies includes:
initializing 30 neuron node weights, randomly selecting an attribute sequence of a certain row of the effective wave height from the attribute matrix of the sample data, and deleting the attribute sequence of the row from the attribute matrix of the sample data;
and (3) calculating the Euclidean distance between the sample data and the weight of each neural network node to obtain a superior neuron node, updating the weight of each neural node according to the superior neuron node, and processing each attribute sequence in the sample data attribute matrix until the sample sequence is an empty set.
Optionally, the method further comprises the step of updating the prediction model in real time, the step comprising at least one of:
updating the input wind field information in real time, and deleting the current prediction model and practical wind field data of the current prediction model when the prediction result of the prediction model does not correspond to the real-time input wind field information;
or acquiring actual measurement data of the high-frequency ground wave radar in real time, then calculating a correlation coefficient and a root mean square error, and updating a prediction model in real time according to the correlation coefficient and the root mean square error;
or predicting the real-time updating result in advance, deleting the model when the model does not meet the correlation coefficient range, and reselecting a disturbance forecast wind field to perform model prediction;
or, the formal prediction is carried out through the prediction model, the check coefficient is calculated, when the formal prediction model does not meet the range of the correlation coefficient and the root mean square error, the model is deleted, and simultaneously, the wind field data used by the model and the pre-prediction model are deleted.
Another aspect of the embodiments of the present invention further provides a wave element prediction system, including:
the first module is used for acquiring wind field information;
the second module is used for carrying out pre-prediction according to the wind field information to carry out optimization verification on the initial model so as to obtain a set prediction model;
a third module, configured to perform formal prediction according to the set prediction model to obtain a prediction sequence value;
a fourth module, configured to check the predicted sequence value, and determine a predicted result that meets the verification requirement;
the fifth module is used for carrying out prediction result classification processing on the prediction result through SOM machine learning to obtain prediction model sets with different precisions;
a sixth module, configured to select a target accuracy prediction model from the accuracy prediction model set, and add a correction value to a prediction result of the target accuracy prediction model to obtain a final prediction result of the output wave element;
wherein the wave element prediction result includes, but is not limited to, effective wave height data, wave crest period data, and wave direction data.
Another aspect of the embodiments of the present invention further provides an electronic device, including a processor and a memory;
the memory is used for storing programs;
the processor executes the program to implement the method as described above.
Yet another aspect of the embodiments of the present invention provides a computer-readable storage medium, which stores a program, which is executed by a processor to implement the method as described above.
The embodiment of the invention also discloses a computer program product or a computer program, which comprises computer instructions, and the computer instructions are stored in a computer readable storage medium. The computer instructions may be read by a processor of a computer device from a computer-readable storage medium, and the computer instructions executed by the processor cause the computer device to perform the foregoing method.
The following detailed description of the embodiments of the present invention is made with reference to the accompanying drawings:
as the single deterministic prediction model in the prior art has a general prediction effect, and the long-period prediction result is difficult to meet the actual requirement; has specific applicable target and working environment and lacks of selectivity. Existing hybrid predictive models may not adequately mine the relationship between wave height data. The root mean square error of the machine learning prediction effect becomes larger with the increase of the prediction time length. In addition, most of the existing technologies focus on the prediction of effective wave height, and in actual engineering, data such as wave wind periods, wave directions and the like are important for harbor basin resonance and offshore construction. Aiming at the defects of the prior art, the invention provides a system for predicting an integrated wave element, which comprises the following steps:
(1) acquiring a control forecast wind field and n disturbance forecast wind field members in a global aggregate weather forecast system as a drive of a control forecast SWAN prediction model and n aggregate SWAN prediction models.
(2) By utilizing n sets of SWAN prediction models, multiple problems such as limitation and errors of traditional single certainty prediction can be effectively solved.
(3) The high-frequency ground wave radar technology is utilized to accurately and effectively obtain the wind and wave field parameter information on the open sea surface, the wind and wave field parameter information is input into the system in real time, and the wind and wave field parameter information and the predicted value are combined to solve the test coefficient.
(3) And screening the prediction results of multiple possibilities generated by the SWAN prediction model set by using an SOM machine learning algorithm, and selecting the prediction models with different accuracies according to actual needs.
(4) The cascade method is utilized to quickly update the input data, so that the accuracy of the set prediction model is improved, the result with higher reliability is output, and the more accurate classification is trained for the learning of the SOM machine.
(5) The method can predict a plurality of wave elements, and provides more effective decision basis for problems encountered in actual engineering.
In order to achieve the purpose, the technical scheme adopted by the invention is as follows: a method and a system for predicting an aggregated wave element comprise the following steps: inputting wind field information, predicting in advance, predicting formally, calculating a check coefficient, classifying SOM, inputting HFR measured data (HFR is English abbreviation of high-frequency ground wave radar), and outputting wave parameters of a selected prediction mode. The flow chart is shown in fig. 1 and fig. 2:
referring to fig. 1 and fig. 2, the steps of the system are connected in a cascade mode, the wind field is input into the system to be predicted as a first cascade in advance, the wind field is predicted as a second cascade in advance, and so on, the system adopts a multi-cascade algorithm. When the first cascade is updated or deleted, the data of different sea surface wind forecast fields are correspondingly updated or deleted.
1. Firstly, inputting wind field information. Acquiring information of a control forecast wind field and n disturbance forecast wind fields of an EPSG (global integrated weather forecast system) as a drive of a control sea wave prediction model and n integrated sea wave prediction models.
2. And secondly, predicting in advance.
2.1, firstly, building a nested model and a set SWAN model, wherein the SWAN (simulating Waves New Share) is a third-generation energy spectrum sea wave numerical mode, and belongs to one of the third-generation sea wave numerical modes, the third-generation sea wave numerical mode also comprises WW3(WAVE WATCH III) and WAM (wave modeling group), and the SWAN is a sea wave numerical mode developed abroad, can be called as a third-generation energy spectrum sea wave numerical mode, and is also called as a third-generation shallow water wave numerical mode.
Specifically, a deterministic SWAN prediction model of a large area (larger than a research area) is constructed, n sets of SWAN grid prediction models are nested in the research area in the grid model, the large area prediction model provides boundary conditions for the SWAN prediction model of the research area set, namely the boundary of the large area is land without a computational grid, the boundary conditions of the research area in the large area adopt the grid in the large area as the input of the boundary conditions, and the water boundary is far enough from the attention area, so that the computational accuracy of the attention area can be ensured. And inputting the obtained control forecast wind field serving as a drive into a deterministic SWAN model of a large area, and operating to obtain boundary conditions of n collective SWAN models of the research area.
Selecting n sea surface disturbance forecast wind fields as the drive of n set SWAN prediction models, and dividing the wind fields into SWAN1、SWAN2、SWAN3、…、SWANnCollective predictive model, SWAN1、SWAN2、SWAN3、…、SWANnAnd reading the information of the input disturbance forecast wind field file and providing boundary conditions for the SWAN prediction model of the research area set by the large area prediction model. Predicting in advance by using a SWAN prediction model for one week to obtain n groups of effective wave height prediction sequence values, and selecting effective wave height values of four prediction points at 0 hour, 6 hours, 12 hours and 18 hours in each day, namely k (k is 1,2, …, n, k)<n) selecting 28 effective wave height values in seven days in the prediction results of the SWAN models to form a kth group of prediction sequence values xk1、xk2、…、xk28(ii) a Similarly, the measured sequence value y of the same period in a week is selected1、y2、…、y28。
2.2 judging Presence of end of prediction
After predicting for one week in advance, comparing the predicted sequence value with the actually measured sequence value of the high-frequency ground wave radar to obtain a judgment coefficient value, wherein the judgment coefficient value is judged by adopting a correlation coefficient R:
and when the R is 0.85-0.90, the condition that the pre-prediction model can meet 85% -90% of the future condition is shown, the preliminary requirements are met, n groups of judgment are carried out totally, and the pre-prediction is finished after all the conditions are met.
And in the pre-prediction process, deleting a certain prediction model which does not meet the judgment coefficient range, and continuously supplementing the SWAN model and the corresponding disturbance forecast wind field file in the first step for pre-prediction until n models meeting the correlation coefficient R exist.
2.3 obtaining the correction value.
The n set SWAN prediction models take n disturbance forecast wind fields as the drive of the SWAN models, only the frequency of wind waves is considered to a great extent, and uncertainty of surge waves is basically ignored, so that the set prediction value is lower than an actual value in most cases. Therefore, in order to further adjust the accuracy of the set prediction, the invention introduces a correction value to the deviation, and the specific implementation mode is as follows:
in the pre-prediction, the average value of n groups of predicted sequence values meeting the requirements after one week of prediction and the average value of actually measured sequence values obtained by one week of simultaneous actual observation are subtracted to obtain the average deviation value P of the predicted value and the actual value:
in the formula (I), the compound is shown in the specification,the average of the n sets of predicted sequence values was obtained for a week prediction,the average of the actually measured sequence values obtained for the period of one week, i.e.
In the formula, xk1、xk2、…、xk28Predicted sequence values, y, for the kth set of pre-predictions1、y2、…、y28The sequence values are actually observed for the simultaneous segment.
Multiplying the average deviation value P by 0.9-0.95 to obtain a final correction value, namely:
Pfinal (a Chinese character of 'gan')=P*(0.9~0.95)
3. And thirdly, formally predicting.
As shown in fig. 3, the formal prediction obtains a formal prediction value.
After the prediction is carried out in advance and the requirements are met, the SWAN prediction model is integrated to enter into formal prediction. Since the feasibility of the collective SWAN prediction model is verified by carrying out the pre-prediction, the collective prediction model can be used as the starting point of the formal prediction after the pre-prediction is finished and the prediction is continued. Similarly, n sea surface disturbance forecast wind fields are selected as the drive of n set SWAN prediction models and are divided into SWAN1、SWAN2、SWAN3、…、SWANnEnsemble prediction model, SWAN1、SWAN2、SWAN3、…、SWANnBy reading input disturbance forecast wind field file information and taking a large-area SWAN model as boundary conditions of n set SWAN prediction models of a research area, prediction of the set SWAN model is realized, uncertainty forecast is obtained, all results including all possible sea waves in future forecast time are obtained, and n groups of prediction sequence values x are obtainedn1、xn2、…、xn28(n is 1,2, …, n)
4. And fourthly, calculating a check coefficient.
And verifying the effect of the prediction model by using the one-week prediction wave height sequence value predicted by each model in the SWAN prediction model and the one-week effective wave height sequence value observed in the actual region by the high-frequency ground wave radar. The test criteria used here are the root mean square error and the correlation coefficient, as follows:
in the formula, xkiIs the ith predictor, y, of the kth groupiFor the actual observed value of the high-frequency ground wave radar, the number of samples of a group of sequence values is 28, and n groups of data are provided in total. The smaller the RMSE value, the closer the R value is to 1, indicating that the model has higher accuracy and is closer to the actual value. Here, when RMSE<0.05,R>When the value is 0.98, the prediction effect of the model meets the requirement, and the prediction model is marked and selected. At the same time, the RMSE will not be satisfied<0.05,R>0.98 predictive models are deleted and a set of predictive models are replenished until RMSE is satisfied<0.05,R>If there are n groups in the 0.98 model, the formal prediction is finished.
5. And fifthly, SOM classification.
As shown in fig. 4, all future possible prediction results that are selected from the n sets SWAN prediction and are calculated through the root mean square error and the correlation coefficient are used for SOM machine learning to classify the prediction results, and results with different accuracies are selected as the prediction mode of the next stage according to actual needs. Further clustering the data by adopting an SOM self-organizing neural network algorithm, which comprises the following specific steps:
(1) and (4) constructing a matrix.
Selecting n groups of predicted effective wave height sequence values of n SWAN model prediction results, selecting the effective wave height of the SWAN prediction results within a prediction time period of 7 days, selecting the effective wave height values of four prediction points within 0 hour, 6 hours, 12 hours and 18 hours in each day, namely selecting 28 effective wave height values within seven days in one SWAN model prediction result to form a group of predicted sequence values, and adding root mean square error and related coefficient indexes. Constructing an attribute matrix of a cycle significant wave height sequence with the size of n multiplied by 30:
in the above matrix, the first 28 columns are the effective wave height values at 0, 6, 12 and 18 times per day of the week, the 29 th column is the correlation coefficient of each group of predicted sequence values, and the 30 th column is the root mean square error of each group of predicted sequence values.
(2) And (6) standardizing the attributes.
Because the constructed effective wave height attribute matrixes have different properties and different attributes, and have larger difference in numerical value, the weight layers of the SOM grids are easily influenced by a certain extreme attribute, and therefore normalization processing of the attributes is required. When all the input and output values are between 0 and 1, the neural network is computationally efficient. Therefore, normalization is required, and the following method is adopted: for a row of matrix attribute values { Xk(1 < k < 28), the standardized formula is as follows:
(3) and (5) clustering and dividing.
After the preparation of the steps (1) and (2), starting SOM clustering division according to the correlation coefficient and the root mean square error, and dividing into the following three steps:
step 1: initializing 30 neuron node weights, randomly selecting an attribute sequence of a certain row of the effective wave height from the sample data attribute matrix, and deleting the attribute sequence of the row from the sample data attribute matrix.
Step 2: and (4) calculating the Euclidean distance between the sample data of each neural network node and the weight to obtain a superior neuron node, and updating the weight of each neural node.
And step 3: and (3) continuing to randomly select another row of attribute sequence of the effective wave height from the sample data attribute matrix, and performing the step (2) until the sample sequence is an empty set.
After the 3 steps are completed, the 1 multiplied by 3 SOM neural node network structure is adopted, the output division result is adopted, and the effective wave height sequences with different characteristics are divided on each attribute level. According to the root mean square error numerical value condition, the method is divided into three groups: the system comprises an optimal accurate prediction model, a suboptimal accurate prediction model and a general accurate prediction model.
6. And sixthly, outputting the wave parameters of the selected prediction model.
After the steps are completed, one precision prediction model is selected from three precision prediction models classified by the SOM according to actual needs, and the effective wave height H of the prediction model with the selected precision is obtaineds predictionIs prepared from HS predictionAdding the final correction value P obtained in step 2.3 to the valueFinal (a Chinese character of 'gan')I.e. Hs=Hs prediction+PFinal (a Chinese character of 'gan'),HsAs a final output forecast. The effective wave height H is selected in the stepssThe description is given; similarly, the peak period, the flow of the prediction of the wave direction, and the effective wave height coincide with each other.
7. And updating the module in real time.
The real-time updating module comprises the whole steps of the system, and each step is in a real-time updating state. The first is that the input wind field information is updated in real time, when the prediction result is not satisfied, the system deletes the model, the wind field data used by the corresponding model can be deleted by the system due to the effect of the cascade algorithm, and the system supplements the model data when monitoring that the number of the model data is less than n; secondly, real-time updating of the measured data input of the high-frequency ground wave radar is used for calculating a correlation coefficient and a root mean square error; the third step is to predict and update in real time in advance, delete the model when the model does not meet the correlation coefficient range, and reselect the disturbance forecast wind field to predict the model; and fourthly, real-time updating of formal prediction and calculation check coefficients, deleting a certain model of the formal prediction when the model does not meet the range of the correlation coefficient and the error of the root mean square, and deleting wind field data and a pre-prediction model used by the model. Likewise, the SOM classification and wave parameter output that outputs the selected model are updated accordingly.
In summary, the invention discloses a wave element prediction method and system which adopt an integrated SWAN prediction model, can effectively improve prediction accuracy, and can realize quick and effective classification by adopting an SOM algorithm, so as to provide the classification for different users. A cascade algorithm is adopted, so that data updating and storage are facilitated, and each cascade forms a real-time updating module which can change a prediction result in real time; and all the steps are closely connected, so that the whole step can respond correspondingly when one step is in a problem.
The wave element prediction method and the wave element prediction system can continuously accumulate measured data to form longer measured time sequence data, and are beneficial to further improving the prediction precision on the basis of the original set prediction precision.
In alternative embodiments, the functions/acts noted in the block diagrams may occur out of the order noted in the operational illustrations. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality/acts involved. Furthermore, the embodiments presented and described in the flow charts of the present invention are provided by way of example in order to provide a more thorough understanding of the technology. The disclosed methods are not limited to the operations and logic flows presented herein. Alternative embodiments are contemplated in which the order of various operations is changed and in which sub-operations described as part of larger operations are performed independently.
Furthermore, although the present invention is described in the context of functional modules, it should be understood that, unless otherwise stated to the contrary, one or more of the described functions and/or features may be integrated in a single physical device and/or software module, or one or more functions and/or features may be implemented in a separate physical device or software module. It will also be appreciated that a detailed discussion of the actual implementation of each module is not necessary for an understanding of the present invention. Rather, the actual implementation of the various functional modules in the apparatus disclosed herein will be understood within the ordinary skill of an engineer, given the nature, function, and internal relationship of the modules. Accordingly, those skilled in the art can, using ordinary skill, practice the invention as set forth in the claims without undue experimentation. It is also to be understood that the specific concepts disclosed are merely illustrative of and not intended to limit the scope of the invention, which is defined by the appended claims and their full scope of equivalents.
The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
The logic and/or steps represented in the flowcharts or otherwise described herein, such as an ordered listing of executable instructions that can be considered to implement logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device.
More specific examples (a non-exhaustive list) of the computer-readable medium would include the following: an electrical connection (electronic device) having one or more wires, a portable computer diskette (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Further, the computer-readable medium could even be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, via for instance optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner if necessary, and then stored in a computer memory.
It should be understood that portions of the present invention may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the various steps or methods may be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or combination of the following techniques, which are known in the art, may be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application specific integrated circuit having an appropriate combinational logic gate circuit, a Programmable Gate Array (PGA), a Field Programmable Gate Array (FPGA), or the like.
In the description herein, references to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., mean 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 invention. In this specification, the schematic representations of the terms used above do not necessarily refer 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.
While embodiments of the invention have been shown and described, it will be understood by those of ordinary skill in the art that: various changes, modifications, substitutions and alterations can be made to the embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.
While the preferred embodiments of the present invention have been illustrated and described, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims.
Claims (10)
1. A wave element prediction method, comprising:
acquiring wind field information;
performing pre-prediction according to the wind field information to perform optimization verification on an initial model to obtain a set prediction model;
performing formal prediction according to the set prediction model to obtain a prediction sequence value;
checking the prediction sequence value to determine a prediction result meeting the verification requirement;
carrying out prediction result classification processing on the prediction result through SOM machine learning to obtain prediction model sets with different precisions;
selecting a target precision prediction model from the precision prediction model set, and adding a correction value to a prediction result of the target precision prediction model to serve as a final prediction result of the output wave element;
wherein the wave element prediction result includes, but is not limited to, effective wave height data, wave crest period data, and wave direction data.
2. The method of claim 1, wherein the predicting the initial model in advance according to the wind field information to obtain a set prediction model comprises:
constructing a deterministic SWAN model of a first area, and nesting a plurality of collective SWAN models in the deterministic SWAN model; wherein the deterministic SWAN model provides boundary conditions for each collective SWAN model;
forecasting training is carried out on each set SWAN model according to the sea surface disturbance forecast wind field information, a forecasting sequence value in a period of time is obtained, and a corresponding actual measurement sequence value in the period of time is selected;
comparing the predicted sequence value with the actually measured sequence value through a correlation coefficient calculation formula to obtain a judgment coefficient value;
determining a set SWAN model meeting the requirement of a correlation coefficient according to the judgment coefficient value;
calculating the average value of predicted sequence values obtained by each set SWAN model meeting the requirement of a correlation coefficient, and comparing the average value with the average value of the actually measured sequence values to obtain an average deviation value;
calculating to obtain a correction value according to the average deviation value;
and correcting each set SWAN model meeting the requirement of the correlation coefficient according to the correction value to obtain a final prediction result of the set prediction model.
3. The method of claim 1, wherein the performing formal predictions according to the ensemble prediction model to obtain predicted sequence values comprises:
when the pre-prediction of the set prediction model meets the preset requirement, listing the set prediction model into a formal prediction model;
and taking the set prediction model as a starting point of the formal prediction, selecting a plurality of pieces of sea surface disturbance forecast wind field information as a drive of the formal prediction, and acquiring an uncertainty forecast result as a prediction sequence value.
4. The method as claimed in claim 1, wherein the step of checking the predicted sequence value to determine the predicted result satisfying the validation requirement comprises:
acquiring a cycle of effective wave height sequence value observed in an actual region by a high-frequency ground wave radar;
calculating a root mean square error according to the effective wave height sequence value and the predicted sequence value;
calculating a correlation coefficient according to the effective wave height sequence value and the prediction sequence value;
and carrying out condition judgment according to the root mean square error and the correlation coefficient, and determining a prediction result meeting the verification requirement.
5. The wave element prediction method according to claim 4, wherein the step of classifying the prediction results by SOM machine learning to obtain prediction model sets with different accuracies comprises:
constructing an attribute matrix of the effective wave height sequence according to the root mean square error, the correlation coefficient and the prediction results of the plurality of set prediction models;
normalizing the attribute matrix;
and clustering and dividing the attribute matrix after the normalization processing according to the correlation coefficient and the root mean square error to obtain prediction model sets with different precisions.
6. The method according to claim 5, wherein the cluster partitioning of the normalized attribute matrix according to the correlation coefficient and the root mean square error to obtain prediction model sets with different accuracies comprises:
initializing 30 neuron node weights, randomly selecting an attribute sequence of a certain row of the effective wave height from the attribute matrix of the sample data, and deleting the attribute sequence of the row from the attribute matrix of the sample data;
and obtaining a superior neuron node by calculating the Euclidean distance between the sample data of each neural network node and the weight, updating the weight of each neural node according to the superior neuron node, and processing each attribute sequence in the sample data attribute matrix by using the weight until the sample sequence is taken as an empty set.
7. The method of claim 1, further comprising the step of updating the predictive model in real time, the step comprising at least one of:
updating the input wind field information in real time, and deleting the current prediction model and practical wind field data of the current prediction model when the prediction result of the prediction model does not correspond to the real-time input wind field information;
or acquiring actual measurement data of the high-frequency ground wave radar in real time, then calculating a correlation coefficient and a root mean square error, and updating a prediction model in real time according to the correlation coefficient and the root mean square error;
or predicting the real-time updating result in advance, deleting the model when the model does not meet the correlation coefficient range, and reselecting a disturbance forecast wind field to perform model prediction;
or, the formal prediction is carried out through the prediction model, the check coefficient is calculated, when the formal prediction model does not meet the range of the correlation coefficient and the root mean square error, the model is deleted, and simultaneously, the wind field data used by the model and the pre-prediction model are deleted.
8. A wave element prediction system, comprising:
the first module is used for acquiring wind field information;
the second module is used for carrying out prediction in advance according to the wind field information to carry out optimization verification on the initial model so as to obtain a set prediction model;
a third module, configured to perform formal prediction according to the set prediction model to obtain a prediction sequence value;
a fourth module, configured to check the predicted sequence value, and determine a predicted result that meets the verification requirement;
the fifth module is used for carrying out prediction result classification processing on the prediction result through SOM machine learning to obtain prediction model sets with different precisions;
a sixth module, configured to select a target accuracy prediction model from the accuracy prediction model set, and add a correction value to a prediction result of the target accuracy prediction model to obtain a final prediction result of the output wave element;
wherein the wave element prediction result includes, but is not limited to, effective wave height data, wave crest period data, and wave direction data.
9. An electronic device comprising a processor and a memory;
the memory is used for storing programs;
the processor executing the program realizes the method of any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that the storage medium stores a program, which is executed by a processor to implement the method according to any one of claims 1 to 7.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210416052.1A CN114742307A (en) | 2022-04-20 | 2022-04-20 | Wave element prediction method and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210416052.1A CN114742307A (en) | 2022-04-20 | 2022-04-20 | Wave element prediction method and system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN114742307A true CN114742307A (en) | 2022-07-12 |
Family
ID=82282661
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210416052.1A Pending CN114742307A (en) | 2022-04-20 | 2022-04-20 | Wave element prediction method and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114742307A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116147587A (en) * | 2023-04-17 | 2023-05-23 | 南开大学 | Wave prediction method and wave measurement system |
CN116933152A (en) * | 2023-06-07 | 2023-10-24 | 哈尔滨工业大学(威海) | Wave information prediction method and system based on multidimensional EMD-PSO-LSTM neural network |
CN117665824A (en) * | 2023-12-22 | 2024-03-08 | 中山大学 | Sea surface wind field reconstruction method and system |
-
2022
- 2022-04-20 CN CN202210416052.1A patent/CN114742307A/en active Pending
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116147587A (en) * | 2023-04-17 | 2023-05-23 | 南开大学 | Wave prediction method and wave measurement system |
CN116933152A (en) * | 2023-06-07 | 2023-10-24 | 哈尔滨工业大学(威海) | Wave information prediction method and system based on multidimensional EMD-PSO-LSTM neural network |
CN116933152B (en) * | 2023-06-07 | 2024-05-03 | 哈尔滨工业大学(威海) | Wave information prediction method and system based on multidimensional EMD-PSO-LSTM neural network |
CN117665824A (en) * | 2023-12-22 | 2024-03-08 | 中山大学 | Sea surface wind field reconstruction method and system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN114742307A (en) | Wave element prediction method and system | |
CN102682573B (en) | A kind of storm surge disaster early warning system based on time series analysis | |
CN107944090B (en) | Gas turbine system performance prediction method based on key component failure model | |
CN110969290A (en) | Runoff probability prediction method and system based on deep learning | |
CN115374995A (en) | Distributed photovoltaic and small wind power station power prediction method | |
CN111242351A (en) | Tropical cyclone track prediction method based on self-encoder and GRU neural network | |
CN116704137B (en) | Reverse modeling method for point cloud deep learning of offshore oil drilling platform | |
CN114648170A (en) | Reservoir water level prediction early warning method and system based on hybrid deep learning model | |
CN112307410A (en) | Seawater temperature and salinity information time sequence prediction method based on shipborne CTD measurement data | |
CN114399235B (en) | Method and system for judging disaster risk level based on rain condition data | |
CN114935759A (en) | Wave field missing value filling method and system based on high-frequency ground wave radar observation | |
CN115186923A (en) | Photovoltaic power generation power prediction method and device and electronic equipment | |
CN116467933A (en) | Storm surge water increasing prediction method and system based on deep learning | |
CN116976227B (en) | Storm water increasing forecasting method and system based on LSTM machine learning | |
CN117252288A (en) | Regional resource active support capacity prediction method and system | |
CN113610268A (en) | Carbon emission prediction method based on residential area space form | |
Costa et al. | Enhancing the accuracy of metocean hindcasts with machine learning models | |
CN113792919B (en) | Wind power prediction method based on combination of transfer learning and deep learning | |
CN114202104B (en) | Method for determining similarity degree of tropical cyclone path and storage medium | |
CN113191526B (en) | Method and system for optimizing and predicting short-term wind speed interval based on random sensitivity | |
Vega-Bayo et al. | Improving the prediction of extreme wind speed events with generative data augmentation techniques | |
Huang et al. | Probabilistic prediction intervals of wind speed based on explainable neural network | |
Sannasiraj et al. | Local model approximation in the real time wave forecasting | |
CN116992781B (en) | Multi-step multi-element storm forecasting method based on deep learning | |
CN117494573B (en) | Wind speed prediction method and system and electronic equipment |
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 |