An agent-based method and system for simulating sound propagation loss

Abstract

The application discloses a kind of sound propagation loss simulation calculation method and system based on agent, belong to ship electronic information technology field, method includes: constructing private big model;User demand is obtained, based on adaptive terrain feature extraction technology and coding ability, the semantic analysis ability that private big model possesses is analyzed to user demand, obtains the historical space-time data and key parameters of specified sea area;According to the historical space-time data and key parameters of specified sea area, based on dynamic model fusion algorithm, corresponding encapsulated agent intelligence is called to complete sound propagation loss simulation calculation;Agent intelligence includes one or more underwater acoustic propagation calculation physical model;Simulation calculation result is obtained, call ocean acoustic knowledge base to complete loss suggestion generation, and carry out the visualization of calculation result;The application can provide higher precision sound propagation loss simulation calculation result.

Description

Technical Field

[0001] This invention belongs to the field of marine electronic information technology, specifically relating to a method and system for simulating and calculating acoustic propagation loss based on intelligent agents. Background Technology

[0002] Calculating underwater acoustic propagation loss is an important issue in underwater acoustics research. Traditional methods, such as ray models, normal wave models, and parabolic equation models, can solve the problem of acoustic propagation loss calculation to a certain extent. However, in complex seabed topography, the propagation path of sound waves is affected by factors such as seabed topographic relief, seabed sediment thickness, and seabed sediment type. Users cannot efficiently determine and call on the appropriate physical model for underwater acoustic propagation calculation to complete the simulation calculation of acoustic propagation loss, which often results in problems such as insufficient calculation accuracy and low calculation efficiency. Summary of the Invention

[0003] In view of this, the purpose of the present invention is to provide a method and system for simulating and calculating sound propagation loss based on intelligent agents, so as to solve the above-mentioned technical problems.

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] A method for simulating and calculating sound propagation loss based on intelligent agents includes:

[0006] Build a private large model;

[0007] To obtain user needs, based on adaptive terrain feature extraction technology and coding capabilities, the semantic analysis capabilities of the private large model are used to analyze user needs and obtain historical spatiotemporal data and key parameters of the specified sea area.

[0008] Based on historical spatiotemporal data and key parameters of the designated sea area, and using a dynamic model fusion algorithm, the corresponding encapsulated agent is invoked to complete the simulation calculation of acoustic propagation loss; the agent includes one or more physical models for underwater acoustic propagation calculation.

[0009] Obtain simulation calculation results, call the ocean acoustics knowledge base to generate loss suggestions, and visualize the calculation results.

[0010] Furthermore, a private large-scale model is constructed, including:

[0011] Build a basic large model;

[0012] Several computational physics models of underwater acoustic propagation are obtained and provided to the basic large model for learning and use, thereby generating a private large model.

[0013] Furthermore, based on adaptive terrain feature extraction technology and coding capabilities, and leveraging the semantic analysis capabilities of the proprietary large model, user needs are analyzed to obtain historical spatiotemporal data and key parameters for the specified sea area, including:

[0014] After obtaining user requirements and inputting them into a private large model for semantic analysis, the parameter positioning of the specified sea area is obtained.

[0015] Historical spatiotemporal data of a specified sea area is obtained based on the parameters; historical spatiotemporal data includes temporal data and spatial data.

[0016] By using adaptive terrain feature extraction technology, the elevation changes of the terrain corresponding to historical data and the terrain features corresponding to complex areas are identified. The terrain features are then converted using coding capabilities to obtain key parameters in the computational physics model of underwater acoustic propagation.

[0017] Furthermore, after obtaining user requirements and inputting them into a private large-scale model for semantic analysis, parameter positioning for a specified sea area is obtained, including:

[0018] After obtaining user requirements and inputting them into a private large model for semantic analysis, the user's selected area, scale, and query time are obtained.

[0019] Based on the user-selected area, scale, and query time, data is extracted from the two-dimensional world map to obtain the corresponding environmental data; the parameter positioning of the specified sea area is determined based on the environmental data.

[0020] Furthermore, a method for simulating sound propagation loss based on intelligent agents also includes: obtaining the initial marine terrain parameters input by the management personnel, performing data preprocessing, and generating a two-dimensional world map; and obtaining the subsequent marine terrain parameters input by the user, performing data preprocessing, and correcting to form a new two-dimensional world map.

[0021] Furthermore, based on historical spatiotemporal data and key parameters of the designated sea area, and using a dynamic model fusion algorithm, the corresponding encapsulated agent is invoked to complete the simulation calculation of sound propagation loss, including:

[0022] Acquire historical spatiotemporal data for a specified sea area; historical spatiotemporal data includes temporal and spatial data.

[0023] Based on time and space data, call the underwater acoustic propagation computational physics model in the private large model that meets the current computational needs;

[0024] Based on the dynamic model fusion algorithm, the underwater acoustic propagation computational physical model that meets the current computing requirements is fused and encapsulated to obtain the agent intelligent agent;

[0025] Based on the agent, key parameters are invoked to complete the simulation calculation of sound propagation loss.

[0026] Furthermore, based on the agent, key parameters are invoked to complete the simulation calculation of sound propagation loss, including:

[0027] By using an agent and combining historical spatiotemporal data, an intelligent model for calculating sound propagation loss is trained.

[0028] Based on the intelligent model for calculating sound propagation loss, key parameters are used as inputs to complete the simulation calculation of sound propagation loss.

[0029] Furthermore, an agent-based simulation calculation method for acoustic propagation loss also includes: selecting a physical model for underwater acoustic propagation calculation that meets the current computational requirements and satisfies a specific Prompt rule. The selection criteria for the specific Prompt rule include underwater acoustic range, sound velocity profile type, applicable sea area characteristics, and computational accuracy limitations.

[0030] A simulation and calculation system for sound propagation loss based on intelligent agents, comprising:

[0031] The user interaction module is used to obtain user needs and export calculation results.

[0032] The data import module is used to import user-defined data to update database content, and to build datasets and computational physics models for underwater acoustic propagation.

[0033] The large model interface module is used to complete the basic large model screening and provide a regionally customized large model interface;

[0034] The core computing module, based on specific Prompt rules, schedules the agent and calls the optimal model for simulation calculations.

[0035] The visualization module is used to visualize the calculation results.

[0036] The beneficial effects of this invention are as follows:

[0037] This invention proposes a method and system for simulating and calculating acoustic propagation loss based on intelligent agents. Based on a fundamental large-scale model, it integrates multiple underwater acoustic propagation calculation physical models and manages the raw data and calculation results, enabling regionally customized simulation and analysis of sound wave propagation loss in different sea areas. This method can efficiently determine and call appropriate underwater acoustic propagation calculation physical models, processing various data types such as measurement point data, text, and files. Through a deep learning large-scale model, the system automatically identifies and extracts key parameters for underwater acoustic propagation calculation, providing high-precision calculation results. Furthermore, the system built based on this method has a unified query interface and real-time monitoring functions, significantly improving user information query efficiency and supporting various application scenarios such as scientific research and engineering design.

[0038] Other advantages, objectives, and features of the invention will be set forth in the following description and will be apparent to those skilled in the art in some respects, or may be learned by practice of the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures particularly pointed out in the written description and the accompanying drawings.

[0039] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0040] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0041] Figure 1 This is a flowchart of a method and system for simulating and calculating sound propagation loss based on an intelligent agent, as described in an embodiment of the present invention.

[0042] Figure 2 This is a schematic diagram illustrating a simulation calculation method for acoustic propagation loss based on an intelligent agent and a suggestion for underwater acoustic propagation loss analysis in a system according to an embodiment of the present invention;

[0043] Figure 3 This is a flowchart illustrating the simulation calculation method for acoustic propagation loss based on an intelligent agent and the overall process for calculating underwater acoustic propagation loss in the system, as described in this embodiment of the invention.

[0044] Figure 4 This is a schematic diagram of a system module in a method for simulating and calculating acoustic propagation loss based on an intelligent agent, as described in an embodiment of the present invention. Detailed Implementation

[0045] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0046] like Figure 1 As shown, this invention proposes a simulation calculation method for sound propagation loss based on intelligent agents, including:

[0047] Build a private large model;

[0048] To obtain user needs, based on adaptive terrain feature extraction technology and coding capabilities, the semantic analysis capabilities of the private large model are used to analyze user needs and obtain historical spatiotemporal data and key parameters of the specified sea area.

[0049] Based on historical spatiotemporal data and key parameters of the designated sea area, and using a dynamic model fusion algorithm, the corresponding encapsulated agent is invoked to complete the simulation calculation of acoustic propagation loss; the agent includes one or more physical models for underwater acoustic propagation calculation.

[0050] Obtain simulation calculation results, call the ocean acoustics knowledge base to generate loss suggestions, and visualize the calculation results;

[0051] The working principle of the above technical solution is as follows: In complex seabed topography, the propagation path of sound waves is affected by factors such as seabed topographic undulation, seabed sediment thickness, and seabed substrate type. This application provides a sound propagation loss simulation calculation method based on an intelligent agent, which can provide regional customized sound propagation loss calculation, simulation, and suggestion generation. Through dynamic model fusion algorithm and adaptive terrain feature extraction and encoding technology, the dialogue content or requirements input by the user are given to a large model for semantic analysis. Then, the most suitable agent is called to perform sound propagation loss calculation simulation. The terrain feature extraction technology identifies the elevation changes and complex areas of the terrain. Then, the encoding technology is used to transform these features into key parameters in the sound wave propagation model. The simulation model can more realistically simulate the propagation path and loss of sound waves in the actual environment, improving the accuracy and reliability of the simulation. Finally, the entire calculation simulation process is integrated to build a user-friendly software interface, improving the efficiency of user calculation simulation needs.

[0052] Specifically, this application mainly integrates three aspects: large-scale model with sound propagation loss calculation, marine and underwater acoustic knowledge base construction, and suggestion generation and data query. It also completes a computational simulation model for simulating sound propagation loss, which mainly implements the following three aspects:

[0053] 1. Sound propagation loss calculation: It can perform customized calculation simulations for specific regions based on user requirements. For a specified region, it can call data from the database and select the most suitable agent to calculate the sound propagation loss.

[0054] 2. Construction of marine acoustic knowledge base: This means that it can perform semantic analysis and understanding based on imported historical underwater acoustic data and existing underwater acoustic knowledge, and use it as a knowledge base to achieve simple answers to relevant questions;

[0055] 3. Suggestion generation and data query: This involves integrating the calculated data with the content of the local knowledge base to generate suggestions and perform data analysis. It also includes analyzing dialogues or questions, comparing them with information in the local text and database, using the semantic analysis capabilities of the large model to understand the user's statements, transmitting them to the system, and integrating and outputting relevant results according to the user's requirements.

[0056] To achieve the regional customization requirements for sound propagation loss calculation in the above-mentioned computational simulation model, dynamic model fusion algorithms and adaptive terrain feature extraction and encoding techniques are needed:

[0057] Dynamic model fusion algorithm: After the user input dialogue content or request is given to a large model for semantic analysis, the most suitable agent is called to perform sound propagation loss calculation simulation.

[0058] Adaptive terrain feature extraction and coding technology: This involves identifying terrain elevation changes and complex areas through terrain feature extraction technology, and then using coding technology to transform these features into parameters in the sound wave propagation model. This enables the simulation model to more realistically simulate the propagation path and loss of sound waves in the actual environment, thereby improving the accuracy and reliability of the simulation.

[0059] After optimizing the agent, the main steps for calculating underwater acoustic propagation loss include:

[0060] The physical model of underwater acoustic propagation is constructed. By combining the physical model of underwater acoustic propagation with real-time sound velocity profiles and detailed seabed topographic data, an intelligent model for calculating sound propagation loss is trained.

[0061] Semantic analysis and agent scheduling leverage the semantic analysis capabilities of large models to understand users' actual needs and operations, locate parameters in a specified sea area, find historical data in a specified spatiotemporal domain, call the encapsulated intelligent agent, and select the most suitable physical model for calculation to meet the current requirements.

[0062] Results visualization and suggestion generation: The model output results are visualized by calling graphics libraries such as Matplotlib, and suggestions are generated by integrating the learning capabilities of the large model with the knowledge base of ocean acoustics.

[0063] Furthermore, the main steps for calculating underwater acoustic propagation loss described above can be broken down as follows:

[0064] Based on some acoustic models, several physical models are pre-set and provided for large models to learn and call;

[0065] Analyze the questions or needs raised by users, and use the semantic analysis capabilities of the large model to understand them. If the user lacks sufficient parameters, prompt the user to add parameters (such as selecting the region, query time, etc.).

[0066] Based on parameters such as the area selected by the user and the query time, a physical acoustic model suitable for the current calculation needs is called after comprehensive analysis.

[0067] Based on the model calculation results, corresponding acoustically related result diagrams such as sound propagation loss diagrams and sound field distribution diagrams are generated.

[0068] Finally, based on the images and calculation results, and integrating relevant information from the industry knowledge base, the simulation results are comprehensively analyzed, and a recommendation report is generated, such as... Figure 2 As shown;

[0069] For complex, multi-faceted problems, it is even possible to combine the knowledge from large models with the results of database queries to generate the final answer.

[0070] The beneficial effects of the above technical solution are as follows: Based on a fundamental large-scale model, this solution integrates multiple computational physics models for underwater acoustic propagation and manages the raw data and calculation results, enabling customized simulation and analysis of sound wave propagation loss in different sea areas. This method can efficiently identify and call appropriate computational physics models for underwater acoustic propagation, processing various data types such as measurement point data, text, and files. Through a deep learning large-scale model, the system automatically identifies and extracts key parameters for underwater acoustic propagation calculations, providing high-precision calculation results. Furthermore, the system built based on this method has a unified query interface and real-time monitoring functions, significantly improving user information query efficiency and supporting various application scenarios such as scientific research and engineering design.

[0071] In one embodiment, constructing a private large model includes:

[0072] Build a basic large model;

[0073] Several computational physics models of underwater acoustic propagation are obtained and provided to the basic large model for learning and use, thereby generating a private large model;

[0074] The working principle and beneficial effects of the above technical solution are as follows: By combining the general capabilities of the large model with the expertise in the field of acoustics in the acoustic model, the personalized private large model can be used to integrate data from multiple sensors, simulate and predict the propagation characteristics of sound waves in complex environments, process and learn data from complex seabed topography, simulate the propagation loss of sound waves under different conditions and predict the behavior of sound waves in specific environments, thereby improving the accuracy and robustness of the acoustic model and further optimizing the design of the acoustic system.

[0075] In one embodiment, based on adaptive terrain feature extraction technology and coding capabilities, the semantic analysis capabilities of a private large model are used to analyze user needs, obtaining historical spatiotemporal data and key parameters for a specified sea area, including:

[0076] After obtaining user requirements and inputting them into a private large model for semantic analysis, the parameter positioning of the specified sea area is obtained.

[0077] Historical spatiotemporal data of a specified sea area is obtained based on the parameters; historical spatiotemporal data includes temporal data and spatial data.

[0078] By using adaptive terrain feature extraction technology, the elevation changes of the terrain corresponding to historical data and the terrain features corresponding to complex areas are identified. The terrain features are then converted using coding capabilities to obtain key parameters in the computational physics model of underwater acoustic propagation.

[0079] In one embodiment, after obtaining user requirements input into a private large-scale model for semantic analysis, parameter positioning for a specified sea area is obtained, including:

[0080] After obtaining user requirements and inputting them into a private large model for semantic analysis, the user's selected area, scale, and query time are obtained.

[0081] Based on the user-selected area, scale, and query time, data is extracted from the two-dimensional world map to obtain the corresponding environmental data; the parameter positioning of the specified sea area is determined based on the environmental data.

[0082] In one embodiment, a method for simulating sound propagation loss based on an intelligent agent further includes: obtaining the initial marine terrain parameters input by the manager, performing data preprocessing, and generating a two-dimensional world map; and obtaining the subsequent marine terrain parameters input by the user, performing data preprocessing, and correcting to form a new two-dimensional world map.

[0083] In one embodiment, based on historical spatiotemporal data and key parameters of a specified sea area, and using a dynamic model fusion algorithm, a corresponding encapsulated agent is invoked to perform a simulation calculation of acoustic propagation loss, including:

[0084] Acquire historical spatiotemporal data for a specified sea area; historical spatiotemporal data includes temporal and spatial data.

[0085] Based on time and space data, call the underwater acoustic propagation computational physics model in the private large model that meets the current computational needs;

[0086] Based on the dynamic model fusion algorithm, the underwater acoustic propagation computational physical model that meets the current computing requirements is fused and encapsulated to obtain the agent intelligent agent;

[0087] Based on the agent, key parameters are invoked to complete the simulation calculation of sound propagation loss.

[0088] In one embodiment, based on the agent, key parameters are invoked to complete the simulation calculation of sound propagation loss, including:

[0089] By using an agent and combining historical spatiotemporal data, an intelligent model for calculating sound propagation loss is trained.

[0090] Based on the intelligent model for calculating sound propagation loss, key parameters are used as inputs to complete the simulation calculation of sound propagation loss.

[0091] The working principle and beneficial effects of the above technical solution are as follows: Figure 3 As shown, in one specific embodiment, the computation process based on a designated sea area includes three steps: agent invocation decision, data preprocessing, and data computation; wherein,

[0092] Agent invocation decision: After comprehensive analysis of environmental parameters, frequency, sea depth and other factors, the agent is calculated by invoking three preset physical acoustic models, namely the ray model, normal mode model and parabolic model, which correspond to the BELLHOP, KRAKEN and RAM calculation tools respectively.

[0093] Data preprocessing: After calling the specified model, the data in the database is searched to find the historical measurement data required for the specified needs, and the data in the NetCDF file is parsed to generate the input files required by the physical acoustic model, such as (.env), (.in), etc.

[0094] Data computation: After the required input files are input into the computational agent, the model starts computation and outputs results, such as (.ray) and (.shd) files;

[0095] It is worth noting that the semantic analysis involved in the natural language processing module during the simulation algorithm process for calculating underwater acoustic propagation loss includes multi-head self-attention mechanism, positional encoding, feedforward network and layer normalization theory.

[0096] In one embodiment, an agent-based acoustic propagation loss simulation calculation method further includes: selecting an underwater acoustic propagation calculation physical model that meets the current calculation requirements and satisfies a specific Prompt rule. The selection conditions for the specific Prompt rule include underwater acoustic range, sound velocity profile type, applicable sea area characteristics, and calculation accuracy limitations.

[0097] The working principle and beneficial effects of the above technical solution are as follows: To improve the accuracy of scheduling, this application adopts specific Prompt rules to implement the scheduling of agent intelligence. When scheduling agent intelligence, it is necessary to reasonably select and configure the corresponding parameters according to the specific application environment and computing power to optimize the results and efficiency. Considering the actual sea area conditions, the reasonable selection conditions determined in this application include underwater acoustic range, sound velocity profile type, applicable sea area characteristics, and computational accuracy limitations.

[0098] Water depth range: Applicable to various ocean depths, especially optimized for mid-deep sea areas, with good applicability and accuracy in both shallow waters and deep ocean areas;

[0099] Sound velocity profile type: It can handle relatively complex sound velocity structures, supports multiple sound velocity models to adapt to the sound propagation calculation needs under different ocean conditions, and performs well in simulating multi-layered media environments;

[0100] Applicable marine characteristics: It is suitable for various open marine environments, especially when detailed consideration of seabed topography and ocean dynamics factors (such as hydrodynamic effects) is required. Kraken software can provide high-quality prediction results.

[0101] Computational accuracy limitations: Kraken, based on the full-wave equation model, has high computational requirements. It has large computation time and resource demands under high-resolution networks and complex acoustic conditions. In addition, it may face time response limitations in scenarios with rapid changes or transient sound propagation.

[0102] like Figure 4 As shown, an agent-based sound propagation loss simulation and calculation system includes:

[0103] The user interaction module is used to obtain user needs and export calculation results.

[0104] The data import module is used to import user-defined data to update database content, and to build datasets and computational physics models for underwater acoustic propagation.

[0105] The large model interface module is used to complete the basic large model screening and provide a regionally customized large model interface;

[0106] The core computing module, based on specific Prompt rules, schedules the agent and calls the optimal model for simulation calculations.

[0107] The visualization module is used to visualize the calculation results.

[0108] It is worth noting that the key technical aspects of an agent-based acoustic propagation loss simulation calculation system mainly include large model selection, dataset construction, acoustic propagation loss physical model construction, and region customization. The specific working principles and beneficial effects of each module in this system have been described in the corresponding method steps above, and will not be repeated here.

[0109] Finally, it should be noted that the above preferred embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail through the above preferred embodiments, those skilled in the art should understand that various changes can be made to it in form and detail without departing from the scope defined by the claims of the present invention.

Claims

1. A method for simulating and calculating sound propagation loss based on intelligent agents, characterized in that, include: Build a private large model; To obtain user needs, based on adaptive terrain feature extraction technology and coding capabilities, the semantic analysis capabilities of the private large model are used to analyze user needs and obtain historical spatiotemporal data and key parameters of the specified sea area. Acquire historical spatiotemporal data for a specified sea area; historical spatiotemporal data includes temporal and spatial data. Based on time and space data, call the underwater acoustic propagation computational physics model in the private large model that meets the current computational needs; Based on the dynamic model fusion algorithm, underwater acoustic propagation computational physical models that meet the current computational requirements are fused and encapsulated to obtain a proxy agent. Among them, underwater acoustic propagation computational physical models that meet the current computational requirements are selected to satisfy specific Prompt rules. The selection conditions of specific Prompt rules include underwater acoustic range, sound velocity profile type, applicable sea area characteristics, and computational accuracy limitations. The proxy agent includes one or more underwater acoustic propagation computational physical models. By using an agent and combining historical spatiotemporal data, an intelligent model for calculating sound propagation loss is trained. Based on the intelligent model for calculating sound propagation loss, key parameters are used as inputs to complete the simulation calculation of sound propagation loss. Obtain simulation calculation results, call the ocean acoustics knowledge base to generate loss suggestions, and visualize the calculation results.

2. The method for simulating and calculating sound propagation loss based on an intelligent agent according to claim 1, characterized in that, Building a private large model includes: Build a basic large model; Several computational physics models of underwater acoustic propagation are obtained and provided to the basic large model for learning and use, thereby generating a private large model.

3. The method for simulating and calculating sound propagation loss based on an intelligent agent according to claim 1, characterized in that, Based on adaptive terrain feature extraction technology and coding capabilities, and leveraging the semantic analysis capabilities of the proprietary large-scale model, user needs are analyzed to obtain historical spatiotemporal data and key parameters for a specified sea area, including: After obtaining user requirements and inputting them into a private large model for semantic analysis, the parameter positioning of the specified sea area is obtained. Historical spatiotemporal data of a specified sea area is obtained based on the parameters; historical spatiotemporal data includes temporal data and spatial data. By using adaptive terrain feature extraction technology, the elevation changes of the terrain corresponding to historical data and the terrain features corresponding to complex areas are identified. The terrain features are then converted using coding capabilities to obtain key parameters in the computational physics model of underwater acoustic propagation.

4. The method for simulating and calculating sound propagation loss based on an intelligent agent according to claim 3, characterized in that, After obtaining user requirements and inputting them into a private large-scale model for semantic analysis, the parameter positioning for the specified sea area is obtained, including: After obtaining user requirements and inputting them into a private large model for semantic analysis, the user's selected area, scale, and query time are obtained. Based on the user-selected area, scale, and query time, data is extracted from the two-dimensional world map to obtain the corresponding environmental data; the parameter positioning of the specified sea area is determined based on the environmental data.

5. The method for simulating and calculating sound propagation loss based on an intelligent agent according to claim 4, characterized in that, Also includes: After obtaining the initial marine terrain parameters input by the management personnel and performing data preprocessing, a two-dimensional world map is generated. Furthermore, after obtaining the marine terrain parameters subsequently input by the user and performing data preprocessing, a new two-dimensional world map is generated.

6. A simulation and calculation system for sound propagation loss based on intelligent agents, characterized in that, include: The user interaction module is used to obtain user needs and export calculation results. The data import module is used to import user-defined data to update database content, and to build datasets and computational physics models for underwater acoustic propagation. The large model interface module is used to complete the basic large model screening and provide a regionally customized large model interface; The core computing module, based on specific Prompt rules, schedules the agent and calls the optimal model for simulation calculations. The core computing module specifically performs the following operations: acquiring historical spatiotemporal data of a specified sea area; historical spatiotemporal data includes temporal data and spatial data; Based on time and space data, call the underwater acoustic propagation computational physics model in the private large model that meets the current computational needs; Based on the dynamic model fusion algorithm, underwater acoustic propagation computational physical models that meet the current computational requirements are fused and encapsulated to obtain a proxy agent. Among them, underwater acoustic propagation computational physical models that meet the current computational requirements are selected to satisfy specific Prompt rules. The selection conditions of specific Prompt rules include underwater acoustic range, sound velocity profile type, applicable sea area characteristics, and computational accuracy limitations. The proxy agent includes one or more underwater acoustic propagation computational physical models. By using an agent and combining historical spatiotemporal data, an intelligent model for calculating sound propagation loss is trained. Based on the intelligent model for calculating sound propagation loss, key parameters are used as inputs to complete the simulation calculation of sound propagation loss. The visualization module is used to visualize the calculation results.

Images