Search

CN-121980945-A - Marine digital twin cloud service system supporting hybrid development mode

CN121980945ACN 121980945 ACN121980945 ACN 121980945ACN-121980945-A

Abstract

The invention relates to the technical field of sea area situation simulation, in particular to an ocean digital twin cloud service system supporting a hybrid development mode. According to the invention, through establishing a standardized variable metadata set and quantifying semantic differences of parameter naming by utilizing a character editing distance algorithm, automatic identification and accurate mapping of heterogeneous third-party model input/output interfaces are realized, a data interaction barrier caused by nonstandard definition is broken through, a multidimensional physical effectiveness primary screening mechanism combining a critical wave starting wind speed threshold value, bayesian inferred wind direction probability and wave growth duration verification is introduced, invalid meteorological interference is effectively filtered, high-precision numerical calculation is started by depending on strict physical triggering conditions, model resources are prevented from being blindly scheduled in an invalid time window, and cloud computing power consumption is greatly reduced.

Inventors

  • YU JINGJING
  • LIU JIAN
  • LIN CHUYONG

Assignees

  • 南方海洋科学与工程广东省实验室(珠海)

Dates

Publication Date
20260505
Application Date
20260126

Claims (10)

  1. 1. An ocean digital twin cloud service system supporting a hybrid development model, the system comprising: the ocean semantic association construction module is used for establishing an ocean state variable metadata set at the initial stage of cloud service, analyzing input and output parameter definitions of a third-party wave model accessed to the cloud platform and establishing a model semantic mapping table; The sea surface dynamic environment primary screening module is used for collecting marine meteorological monitoring data flows comprising sea surface air flow rate, sea surface air flow direction and sea surface air flow state maintaining time in real time through a cloud communication interface of a cloud platform, and constructing a wind field physical feature set; the wave generation probability deduction module is used for establishing a likelihood probability distribution table for mapping the deviation of the wind direction angle and the coastline normal direction of the sea area to a conditional probability value, comparing the likelihood probability distribution table with the sea surface air flow direction in which the wind field physical characteristics are concentrated, and calculating the wind direction wave generation probability; The wave growth aging verification module is used for generating a prediction activation instruction according to the change condition of the sea surface air flow state maintaining time period with time in the wind field physical characteristic set and combining the wind direction wave probability; and responding to the prediction activation instruction, inquiring a corresponding third-party wave model in the model semantic mapping table, calculating the numerical value of the sea wave height and the tide level, and generating sea digital twin scene synchronous data.
  2. 2. The ocean digital twin cloud service system supporting a hybrid development mode according to claim 1, wherein the model semantic mapping table comprises a category text identifier corresponding to an ocean state variable, an input parameter name of a third party wave model which is extracted in a resolving mode and an edit difference metric value representing a mapping relation between the two, the wind field physical feature set comprises a sea surface air flow rate, a sea surface air flow direction and a sea surface air flow state maintaining duration which meet a minimum wave critical wind speed condition, the wind direction wave probability is specifically a normalized value calculated based on a conditional probability value corresponding to wind direction deviation and an priori probability, the prediction activation instruction comprises a time surplus verification compliance flag, a wind direction wave probability confidence confirmation bit and a third party wave model scheduling trigger signal, and the ocean digital twin scene synchronous data comprises a sea wave height value and a tide level value which are output by model calculation.
  3. 3. The marine digital twin cloud service system supporting hybrid development modes of claim 1, wherein the marine semantic association building module comprises: The standard variable definition sub-module is used for establishing a marine state variable metadata set in the initial stage of cloud service, configuring a category text identifier corresponding to each marine state variable in the marine state variable metadata set, setting a physical value range boundary value and a time update step length value of the marine state variable, and generating a standard parameter definition set; the parameter difference degree sub-module analyzes input and output parameter definitions of a third-party wave model accessed to the cloud platform, extracts parameter names of the third-party wave model to be mapped, judges the deviation degree of a text structure between the parameter names of the third-party wave model and the standard parameter names by taking a category text mark in the standard parameter definition set as a semantic comparison standard, and calculates a parameter character editing difference value; and the mapping relation establishing sub-module is used for comparing the parameter character editing difference value with a preset fault tolerance threshold value, screening parameter pairs with the parameter character editing difference value smaller than the fault tolerance threshold value, associating a third-party wave model parameter name with a category text identifier according to a screening result, and establishing a model semantic mapping table.
  4. 4. The marine digital twin cloud service system supporting hybrid development modes according to claim 3, wherein the sea-surface dynamic environment prescreening module comprises: The wave starting threshold value calculation submodule is used for collecting marine meteorological monitoring data flows comprising sea surface air flow rate, sea surface air flow direction and sea surface air flow state maintaining time in real time through a cloud communication interface, calling a pre-stored sea area average water depth value and a wave generating critical condition parameter, and calculating the minimum wave starting critical wind speed required by triggering the marine waves under the current sea area average water depth value and the wave generating critical condition parameter; the wind speed judging submodule is used for extracting the sea surface air flow rate in the marine meteorological monitoring data flow, calculating the numerical difference between the sea surface air flow rate and the minimum wave starting critical wind speed, judging the surplus degree of the sea surface air flow rate relative to the wave starting threshold and generating a wave starting wind speed judging result; And the wind field feature extraction submodule detects the wave-making wind speed judging result, screens the time period which is more than or equal to zero, extracts the data fields of the sea surface air flow rate, the sea surface air flow direction and the sea surface air flow state holding duration in the marine meteorological monitoring data flow corresponding to the time period, and constructs a wind field physical feature set.
  5. 5. The marine digital twin cloud service system supporting hybrid development modes according to claim 4, wherein the wave probability deduction module comprises: the wind direction deviation calculation sub-module is used for extracting the sea surface air flow direction in the wind field physical characteristic set, calling a preset coastline normal direction vector of the sea area, and calculating the absolute value of the angle difference between the sea surface air flow direction and the coastline normal direction vector of the sea area to obtain a wind direction deviation value; the conditional probability retrieval sub-module establishes a likelihood probability distribution table for mapping the deviation of the wind direction angle and the coastline normal direction of the sea area to conditional probability values, retrieves the conditional probability values matched with the wind direction deviation values in the likelihood probability distribution table, quantifies the possibility of generating waves under the specified wind direction condition, and generates a directional conditional probability value; And the wave starting probability normalization sub-module is used for obtaining a preset wave starting prior probability, calculating the product of the wave starting prior probability and the directional condition probability value, carrying out normalization processing on the product result and quantifying the wind direction wave starting probability of the wave triggering process of the wind field environment.
  6. 6. The marine digital twin cloud service system supporting hybrid development modes according to claim 5, wherein the wave growth aging verification module comprises: the time surplus quantum module is used for extracting the sea surface air flow state maintaining duration in the wind field physical feature set, obtaining the minimum duration time required by continuously transmitting preset wind power energy to the water surface to form a stable wave form, and calculating the numerical difference between the sea surface air flow state maintaining duration time and the minimum duration time to obtain the time surplus; the double-condition verification sub-module judges whether the current time surplus is larger than or equal to zero, and meanwhile compares the wind direction wave probability with a preset confidence coefficient threshold value to obtain an aging and probability comparison result; and the activation instruction generation sub-module detects the ageing and probability comparison result, and generates a prediction activation instruction when the time surplus is larger than or equal to zero and the wind direction wave probability exceeds a confidence coefficient threshold value.
  7. 7. The marine digital twin cloud service system supporting hybrid development modes of claim 6, wherein the twin situation evolution module comprises: The parameter mapping assignment sub-module responds to a prediction activation instruction, dispatches a corresponding third-party wave model, inquires the semantic mapping table of the model to obtain a mapping matching relation between the input parameter name of the third-party wave model and the sea surface air flow rate, the sea surface air flow direction and the sea surface air flow state keeping duration field in the wind field physical characteristic set, inputs an input parameter interface corresponding to the third-party wave model, and establishes a third-party wave model input parameter assignment set; The wave value calculation sub-module is used for operating the third-party wave model according to the third-party wave model input parameter assignment set, simulating the sea surface hydrodynamic process under the action of the wind field, calculating the vertical variation amplitude of the sea surface and the periodic fluctuation level of the water body, quantifying the physical fluctuation form of the sea surface under the drive of the appointed wind field, and obtaining the sea wave height and the tide level value; and the twin data synchronization sub-module is used for matching the sea wave height and the sea level value with the corresponding ocean state variable in the ocean state variable metadata set, updating the state attribute of the ocean digital twin entity and generating the ocean digital twin scene synchronization data.
  8. 8. The marine digital twin cloud service system supporting hybrid development mode of claim 3, wherein configuring the category text identification for each marine state variable in the marine state variable metadata set comprises: Selecting the sea surface wave height of the target sea area, the sea surface prevailing wave direction of the target sea area and the sea surface average wave period of the target sea area as ocean state variable objects; the method comprises the steps of defining a class text identifier of sea wave height of a target sea area as a standard wave height character string, defining a class text identifier of sea wave dominant direction of the target sea area as a standard wave direction character string, and defining a class text identifier of sea wave average wave period of the target sea area as a standard period character string.
  9. 9. The marine digital twin cloud service system supporting hybrid development mode according to claim 5, wherein establishing a likelihood probability distribution table mapping a wind direction angle and a coastline normal direction deviation of a sea area to conditional probability values comprises: setting a closed interval of 0-180 degrees as a value range of deviation between a wind direction angle and a coastline normal direction of a sea area; Invoking a preset angle discretization step value, and equally dividing a value range into a plurality of angle deviation discrete intervals with fixed widths; extracting a central angle value of each angle deviation discrete interval as a deviation index value; aiming at an angle deviation discrete interval with a deviation index value equal to 0 degree, a corresponding conditional probability value is given as a value 1, and the maximum wave possibility of the vertical direction shoreside downwards is represented; For an angle deviation discrete interval with a deviation index value between 0 and 90 degrees, calculating a cosine function value of the deviation index value, determining the cosine function value as a conditional probability value corresponding to the interval, and forming numerical distribution which monotonically decreases along with the angle deviation; aiming at an angle deviation discrete interval with a deviation index value larger than 90 degrees, setting a corresponding conditional probability value as a value 0, and representing an ineffective wave-starting state of offshore wind; And storing each angle deviation discrete interval and the corresponding conditional probability value thereof according to a key value pair format to generate a likelihood probability distribution table.
  10. 10. The marine digital twin cloud service system supporting hybrid development modes according to claim 9, wherein for wind direction heave probability, the formula is adopted: ; Wherein, the Indicating that a specified winddirection deviation is observed Wave generation under conditions of (2) As the wind direction surge probability, Representing the likelihood that a given wind direction is observed with wave generation, Representing the prior probability of wave generation, Indicating the false positive probability of observing the wind direction without wave generation, Representing a priori probability that a wave is not generated, by the formula And (5) calculating to obtain the product.

Description

Marine digital twin cloud service system supporting hybrid development mode Technical Field The invention relates to the technical field of sea area situation simulation, in particular to an ocean digital twin cloud service system supporting a hybrid development mode. Background The technical field of sea area situation simulation is a comprehensive technical field for carrying out information acquisition, data integration, state modeling and operation deduction around a sea environment, sea area resources and dynamic change processes thereof mainly through digital data processing, and generally comprises digital expression of ocean basic geographic data, ocean hydrologic elements, meteorological elements, ocean engineering objects and human activity information. The ocean digital twin cloud service system is characterized in that a digital mapping system corresponding to a real ocean environment and an object is built around the cloud, time marking, spatial registration and format unification are carried out on buoy observation data, remote sensing image data, ship automatic identification data and historical ocean investigation data, the data are stored in the cloud computing environment, a corresponding virtual ocean scene is generated according to a given ocean physical mechanism relation and a rule base, and meanwhile ocean element states in the virtual scene are periodically updated and synchronized through cloud computing force, so that the digital mapping system which is associated with the actual ocean environment is formed. The existing ocean digital twin cloud service generally adopts a full-flow periodic updating mechanism to maintain a virtual scene state, the continuous calculation mode consumes cloud computing power resources under the conditions of stable sea conditions or invalid wind fields without distinction, the system operation cost is high, the resource scheduling efficiency is low, virtual wave phenomena against objective physical laws are still generated under the scenes of insufficient offshore wind directions or wind energy transmission duration and the like due to lack of physical effectiveness verification of environmental driving factors, the digital twin cannot accurately map real hydrodynamic logic, and a rigid data mapping mode is difficult to automatically adapt to parameter naming standards of external heterogeneous model diversification, so that data interoperation is hindered and flexible expansion of simulation functions is limited. Disclosure of Invention The invention aims to solve the defects in the prior art, and provides an ocean digital twin cloud service system supporting a hybrid development mode. In order to achieve the above object, the present invention adopts the following technical scheme, and an ocean digital twin cloud service system supporting a hybrid development mode includes: the ocean semantic association construction module is used for establishing an ocean state variable metadata set at the initial stage of cloud service, analyzing input and output parameter definitions of a third-party wave model accessed to the cloud platform and establishing a model semantic mapping table; The sea surface dynamic environment primary screening module is used for collecting marine meteorological monitoring data flows comprising sea surface air flow rate, sea surface air flow direction and sea surface air flow state maintaining time in real time through a cloud communication interface of a cloud platform, and constructing a wind field physical feature set; the wave generation probability deduction module is used for establishing a likelihood probability distribution table for mapping the deviation of the wind direction angle and the coastline normal direction of the sea area to a conditional probability value, comparing the likelihood probability distribution table with the sea surface air flow direction in which the wind field physical characteristics are concentrated, and calculating the wind direction wave generation probability; The wave growth aging verification module is used for generating a prediction activation instruction according to the change condition of the sea surface air flow state maintaining time period with time in the wind field physical characteristic set and combining the wind direction wave probability; and responding to the prediction activation instruction, inquiring a corresponding third-party wave model in the model semantic mapping table, calculating the numerical value of the sea wave height and the tide level, and generating sea digital twin scene synchronous data. As a further scheme of the invention, the model semantic mapping table comprises a category text identifier corresponding to a sea state variable, an analysis extracted third-party wave model input parameter name and an editing difference measurement value representing a mapping relation between the category text identifier and the analysis extracted third-party wave model