Simulation data management method based on servitization integrated architecture

Abstract

The invention discloses a simulation data management method based on a servitization integrated architecture, and relates to the technical field of combat simulation and deduction, and the method comprises the steps: building a servitization integrated data management architecture according to a data collection layer, a data storage layer, a data management layer, a data management layer and a data service layer; based on a service integrated data management architecture, data standardization is carried out, a data model is designed, data fusion preprocessing, data integration and data analysis are carried out on collected multi-source heterogeneous data according to a data standard and the data model, data fusion and warehousing management are completed, and meanwhile balance and unification of data security and sharing are guaranteed. According to the invention, a one-stop and all-around data management solution can be provided for simulation application.

Description

Technical Field

[0001] This invention relates to the field of combat simulation and deduction technology, specifically to a military simulation data governance, distributed service architecture, and intelligent metadata management method. Background Technology

[0002] In today's rapidly developing information age, traditional data collection and acquisition methods are no longer sufficient to meet the needs of naval combat simulation. High resource costs, difficulty in guaranteeing data quality, inability to monitor the data collection process, lack of automation in data processing, and untimely data transmission severely restrict the capabilities of various simulation applications and decision-making efficiency.

[0003] Currently, the severe data silos across different military branches and the lack of unified standards for metadata descriptions lead to semantic ambiguity in the data. Furthermore, traditional data management architectures are static and ill-suited to the real-time needs of dynamic battlefield environments, exhibiting rudimentary data version management and insufficient historical status tracing capabilities. Summary of the Invention

[0004] This invention provides a simulation data management method based on a service-oriented integrated architecture, aiming to create a full-chain service path that runs through data acquisition, efficient storage, refined governance, intelligent analysis, unified management and diverse applications, providing a one-stop, comprehensive data management solution for simulation applications.

[0005] This invention provides a simulation data management method based on a service-oriented integrated architecture, comprising: A service-oriented integrated data management architecture is built based on the data acquisition layer, data storage layer, data management layer, data management layer, and data service layer. Based on a service-oriented integrated data management architecture, data standardization is carried out, data models are designed, and the collected multi-source heterogeneous data is preprocessed, integrated, and analyzed according to data standards and data models to complete data fusion and storage management, while ensuring a balance between data security and sharing.

[0006] In some instances, the data acquisition layer is used to collect data from various data sources, employing corresponding acquisition interfaces and protocols for different data sources; The data storage layer is used to address the storage needs of large amounts of data using a distributed storage architecture; The data management layer is used to provide functions for organizing, indexing, classifying, and managing metadata of data; The data service layer is used to provide data access interfaces to meet the data needs of different users and applications.

[0007] In some instances, the data standardization is used to transform different data according to preset rules so that they have the same standard and scope, and data standardization is associated with data sharing.

[0008] In some instances, the design data model includes: Determine the data that the system needs to store, and understand how the data is used and the business rules; Use ER diagrams or UML class diagrams to represent the relationships between entities and data, and to define the attributes of entities and relationships; Transform the conceptual model into a logical model, determine the table structure, fields, primary keys, foreign keys and indexes, and define data integrity constraints; Design the physical storage structure of the database based on the logical model, considering performance optimization, and determine storage allocation and access paths.

[0009] In some instances, the data model includes a conceptual data model, a logical data model, a physical data model, and an object-oriented data model supplement. The conceptual data model includes identifying entities and relationships and defining attributes; the logical data model includes determining relation schemas and normalization processing; the physical data model includes selecting storage structures and formulating indexing strategies; and the object-oriented data model supplement includes class-object mapping and persistence design.

[0010] In some instances, the data fusion is an information processing technique used to automatically analyze and synthesize several observation information obtained in a time sequence under preset criteria in order to complete the required decision-making and evaluation tasks. The data fusion layers include data layer fusion, feature layer fusion, and decision layer fusion.

[0011] In some instances, the data fusion preprocessing is used to remove noise, outliers, and duplicate data from the data through data cleaning, unify the data format of different data sources after cleaning, normalize the data after the data format is unified, and finally align the normalized data by time.

[0012] In some instances, the data integration is used to extract data from different sources and formats, convert it into a unified format or structure, and then load it into a target system. The data integration methods include data warehouse methods and federated database methods.

[0013] In some instances, the data analysis is used to preprocess, explore, and mine the integrated data. The preprocessing stage requires cleaning, transforming, and normalizing the data to ensure its quality and consistency. Exploratory analysis reveals the basic characteristics and trends of the data through descriptive and graphical analysis. Data mining uses association rule mining and sequence rule mining to discover potential patterns and regularities in the data.

[0014] In some instances, the aforementioned data security and sharing are used to perform data encryption, access control, network security, data backup and recovery, security auditing, physical security, data anonymization, compliance, sharing purpose, sharing scope, privacy protection, data quality, authorization and agreements, technical platforms, and monitoring and auditing operations. This is achieved by balancing data security and sharing through data anonymization and desensitization, signing data sharing agreements, and establishing data sharing platforms. Furthermore, data is categorized according to its sensitivity and importance, with different security measures and sharing strategies implemented for different categories. Only data necessary to achieve the objectives is shared, increasing the transparency of data sharing, and continuously monitoring data usage during the sharing process to ensure data security.

[0015] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects: (1) Supports single-cluster management of PB-level military simulation data, with a query time of no more than 1 second; (2) Cross-system data integration efficiency is greatly improved, reducing manual management costs; (3) Effectively reduce data version conflicts. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram illustrating the implementation process of the simulation data management method based on a service-oriented integrated architecture provided in this embodiment of the invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] In the following description, specific embodiments of the invention will be illustrated with reference to steps and symbols performed by one or more computers, unless otherwise stated. Therefore, these steps and operations will be referred to several times as being performed by a computer, and computer execution as referred to herein includes operations by a computer processing unit representing electronic signals of data in a structured format. This operation transforms the data or maintains it at a location in the computer's memory system, which can be reconfigured or otherwise alter the operation of the computer in a manner well known to those skilled in the art. The data structure maintained by the data is the physical location of the memory, which has specific characteristics defined by the data format. However, the principles of the invention described above are not intended to be limiting, and those skilled in the art will understand that many of the following steps and operations can also be implemented in hardware.

[0020] The terms "module" or "unit" as used herein can be considered as software objects executing on the computing system. Different components, modules, engines, and services described herein can be considered as implementations on the computing system. The apparatus and methods described herein are preferably implemented in software, but can also be implemented in hardware, both of which are within the scope of this invention.

[0021] Those skilled in the art will understand that, unless specifically stated otherwise, the singular forms “a,” “an,” and “the” used herein may also include the plural forms. It should be further understood that the term “comprising” as used in this specification means the presence of features, integers, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. It should be understood that when we say an element is “connected” or “coupled” to another element, it can be directly connected or coupled to the other element, or there may be intermediate elements. Furthermore, “connected” or “coupled” as used herein can include wireless connections or wireless coupling. The term “and / or” as used herein includes all or any units and all combinations of one or more associated listed items.

[0022] This invention provides a simulation data management method based on a service-oriented integrated architecture. It unifies the data interface through a REST protocol adapter, constructs a service-oriented data bus, and supports tens of thousands of concurrent requests. It accesses a military domain data semantic network to automatically parse the correlation relationships of most heterogeneous data. It adopts a dynamic QoS strategy to dynamically adjust the data service quality according to task priority, such as automatically enabling redundant transmission mode in wartime scenarios.

[0023] The implementation process of the simulation data management method based on a service-oriented integrated architecture provided in this embodiment of the invention is as follows: Figure 1 As shown, this study investigates an integrated data management technology architecture, constructing a service-oriented integrated data management architecture based on a data acquisition layer, data storage layer, data management layer, data service layer, and a data collection layer. Based on this architecture, data standardization is implemented, data model design is conducted, and the collected multi-source heterogeneous data undergoes data fusion preprocessing, integration, and analysis according to data standards and models, completing data fusion and database management. Throughout the implementation of the simulation data management method based on the service-oriented integrated architecture, a balance between data security and sharing is consistently maintained. Specifically: 1. Research on integrated data management technology architecture (1) Data Acquisition Layer This task involves collecting data from various data sources, employing appropriate acquisition interfaces and protocols for each source. For simulation software output data, acquisition can be achieved through developing specific plugins or utilizing the software's data export interface. Sensor data requires sensor network technology and data acquisition cards to convert it into digital signals and transmit it to the data management system. Data from external databases is acquired using data synchronization or data extraction tools.

[0024] (2) Data storage layer A distributed storage architecture is adopted to address the storage needs of large data volumes. Combining the advantages of relational and non-relational databases, highly structured data requiring frequent complex queries (such as experimental configuration information and basic model parameters) is stored in relational databases, leveraging their powerful transaction processing and SQL query capabilities. Semi-structured and unstructured data (such as complex data structures in simulation results and log files) are stored in non-relational databases (such as HBase and MongoDB) or distributed file systems (such as HDFS) to achieve highly scalable and high-throughput data storage.

[0025] (3) Data Management Layer It provides functions for data organization, indexing, classification, and metadata management. By establishing a data catalog and metadata repository, it provides detailed descriptions of data sources, formats, uses, quality, and other information, facilitating rapid data retrieval and location. Simultaneously, it implements data lifecycle management strategies, archiving, backing up, or deleting data based on its value and timeliness to optimize the utilization of storage resources.

[0026] (4) Data service layer Provides external data access interfaces to meet the data needs of different users and applications. Employs a RESTful architecture or Web service technology to achieve secure data sharing and interaction. Offers services such as data querying, data subscription, and data push, supporting data access from different clients (e.g., desktop applications, web browsers, mobile applications), and allows for data access control based on user permissions.

[0027] 2. Data standardization Data standardization is a crucial aspect of data management, contributing to improved data usability, interoperability, and value. Data standardization is typically associated with sharing and its significance lies in three main areas.

[0028] Data standardization is the foundation of data sharing. Only standardized data has consistency and interoperability, enabling data sharing among different users.

[0029] Data sharing drives the development of data standardization. As the demand for data sharing continues to increase, the requirements for data standardization are also becoming more stringent, driving the continuous development and improvement of data standardization technologies.

[0030] Data standardization and sharing work together to promote the effective use and innovative application of data. Standardization and sharing can break down data barriers, promote data flow and integration, and thus drive innovative data applications and socio-economic development.

[0031] Data standardization refers to transforming different data according to certain rules to give it the same standards and scope, facilitating data comparison and analysis. Standardization converts data into a unified format, structure, and semantics, making it easier to store, transmit, and analyze, as well as to share and understand among different systems, applications, and users. It includes data type conversion, data format conversion, data unit conversion, and data range conversion.

[0032] ① The purpose of standardization Improve the quality, readability, and reliability of data; reduce the cost of data integration and analysis; and improve the efficiency and accuracy of data analysis.

[0033] i. Improve data quality. Data standardization can ensure the accuracy, completeness, consistency, and reliability of data, thereby improving data quality.

[0034] ii. Reduce the cost of data integration and analysis: Standardization simplifies the processes of data collection, storage, transmission, and processing, thereby reducing the cost of data integration and analysis. It also reduces the additional maintenance work caused by inconsistent data formats.

[0035] iii. Improve the efficiency and accuracy of data analysis: Standardized data has higher readability and comparability, which facilitates data analysis and mining, making data analysis more efficient and accurate.

[0036] iv. Promote data interoperability to enable seamless exchange and utilization of data between different systems.

[0037] ② Standardization methods Commonly used standardized methods include data cleaning, data transformation, and data mapping.

[0038] i. Data cleaning: Identifying and correcting errors, duplicates, and inconsistencies in the data.

[0039] ii. Data conversion: Converting data from its original format to a standard format, such as date, currency, unit of measurement, etc.

[0040] iii. Data mapping: Mapping source data fields to corresponding fields in the target data model.

[0041] iv. Metadata management, defining the meaning, format, and constraints of data elements to facilitate data understanding and processing.

[0042] Common data standardization methods include min-max standardization, Z-score standardization, decimal scaling standardization, normalization, logarithmic transformation, and skewness correction. Each method has its specific application scenarios and advantages and disadvantages, and the appropriate method can be selected based on the nature of the data and specific needs.

[0043] ③ Common data standardization standards i. Data element naming standards. Such as ISO / IEC 11179, which defines the naming rules for data elements.

[0044] ii. Data exchange formats. Such as XML, JSON, and CSV, used for exchanging data between different systems.

[0045] iii. Industry data models. Such as CDISC (Clinical Data Interchange Standards Association) standards, used for data standardization in specific industries.

[0046] Data sharing refers to the process of making data available to other individuals, organizations, or systems. It allows users in different locations using different computers and software to access and perform various operations, calculations, and analyses on each other's data.

[0047] ① Challenges of Data Sharing Achieving data sharing allows more people to make fuller use of existing data resources, reducing repetitive work and associated costs in data collection and acquisition. However, it faces numerous challenges.

[0048] i. Data privacy and security. Ensure that shared data does not violate privacy regulations and security policies.

[0049] ii. Data ownership and permissions. Clearly define data ownership and user access permissions.

[0050] iii. Data quality and availability. Ensure that the shared data is accurate and usable.

[0051] iv. Technological compatibility. Ensure that different systems can exchange data effectively.

[0052] ② Methods of data sharing Establish a unified and legally mandated data exchange standard to standardize data formats and encourage users to adopt the prescribed data standards as much as possible; establish corresponding data usage management methods and formulate corresponding data copyright protection and intellectual property protection regulations.

[0053] i. Open data. Publicly release data, allowing anyone to use, reuse, and distribute it free of charge.

[0054] ii. Data exchange protocol. A formal protocol defines the conditions and terms for data sharing.

[0055] iii. API (Application Programming Interface). Provides a programming interface that allows other applications to access and utilize the data.

[0056] iv. Data Sharing Platform. Establish a platform that enables data providers and consumers to easily exchange data. Data sharing mainly includes two methods: peer-to-peer communication and centralized server communication. Peer-to-peer communication does not require a central server to centrally store data, but it poses risks of information leakage and transmission costs; centralized server communication solves these problems, but it carries risks of single points of failure and difficulty in guaranteeing data integrity.

[0057] ③ Best practices for data sharing i. Develop a data sharing policy. Clearly define the purpose, scope, conditions, and responsibilities for data sharing.

[0058] ii. Data anonymization. Sensitive information is anonymized before sharing to protect personal privacy.

[0059] iii. Data access control. Implement strict access control to ensure that only authorized users can access the data.

[0060] iv. Data Usage Monitoring. Monitor data usage to ensure it is not misused.

[0061] Data standardization and sharing are highly interdependent and synergistic in processing big data. Data standardization can improve data quality and readability, and reduce the cost of data integration and analysis; while data sharing enables efficient data transmission and access, promoting the full utilization and realization of data value. Data standardization and data sharing are complementary. Data standardization provides the foundation for data sharing, while the demand for data sharing drives the further development of data standardization. In today's increasingly digital and information-driven world, data standardization and sharing are of great significance for promoting information flow, supporting decision-making, and innovating services.

[0062] 3. Data Model Design Within the scope of integrated data management technology, data model design is a crucial step. It aims to build a reasonable, efficient, and accurate architecture that reflects the simulation business logic and data relationships, in order to support the orderly organization, storage, retrieval, and analysis of data. It involves creating data models to represent data structures, attributes, relationships, and constraints.

[0063] (1) Basic Concepts Entity: A distinguishable object in the real world, such as a person, place, thing, or event.

[0064] Attribute: The characteristics of an entity, such as a person's name or age.

[0065] Relationship: The connection between entities.

[0066] Constraints: Rules in a data model used to ensure the integrity and consistency of data.

[0067] (2) Data model design steps ① Requirements Analysis Determine what data the system needs to store. This includes both structured and unstructured data. Structured data includes data from the database, while unstructured data includes text, images, audio, video, etc.

[0068] Understand how the data is used and the business rules.

[0069] ② Conceptual Design Use ER diagrams (Entity-Relationship Diagrams) or UML class diagrams to represent entities and the relationships between them.

[0070] Define the attributes of entities and relationships.

[0071] ③ Logical Design Transform the conceptual model into a logical model, such as a relational model.

[0072] Determine the table structure, fields, primary key, foreign key, and indexes.

[0073] Define data integrity constraints, such as entity integrity, referential integrity, and user-defined constraints.

[0074] ④ Physical Model Design Design the physical storage structure of the database based on the logical model.

[0075] Consider performance optimization strategies, such as partitioning, clustering, and indexing.

[0076] Determine storage allocation and access paths.

[0077] (3) Data model type ① Conceptual Data Model i. Identify Entities and Relationships: Starting from the simulation business process, identify core entities. For example, in industrial manufacturing simulation experiments, "production equipment," "products," "operators," and "processes" are all key entities. Analyze the relationships between them. For instance, "production equipment" and "products" have a "production" relationship; one piece of equipment can produce multiple products, and one product can be processed on multiple pieces of equipment. "Operators" and "production equipment" have an "operation" relationship. This kind of analysis clarifies the business logic.

[0078] ii. Define attributes: Clearly describe the attributes of each entity. "Production equipment" has attributes such as equipment number, equipment name, model and specifications, purchase date, and operating status; "product" has attributes such as product number, product name, specifications and dimensions, and quality standards. These attributes fully describe the characteristics of the entity and provide detailed dimensions for subsequent data recording.

[0079] ② Logical data model i. Relationship schema determination: Based on the conceptual model, entities and relationships are transformed into relational database table structures. For example, the entity "Production Equipment" is designed as the "Equipment" table, containing fields such as "EquipmentID (equipment number, primary key)," "EquipmentName," and "Model." "Product" corresponds to the "Product" table, with fields such as "ProductID (primary key)" and "ProductName." Then, based on the "Production" relationship, a foreign key field "EquipmentID" is added to the "Product" table to link with the "Equipment" table, constructing reasonable logical relationships such as many-to-many and one-to-many.

[0080] ii. Normalization: Optimize table structure by following database design paradigms (1NF - 3NF, etc.) to eliminate data redundancy and abnormal update issues. For example, when recording product production batch details, create a separate table called "ProductionBatch" for batch information, which is linked to the "Product" table. This avoids storing common batch content repeatedly in the "Product" table, ensuring data consistency and integrity, and improving storage efficiency and ease of updates.

[0081] ③ Physical data model i. Storage structure selection: Select storage media and structure based on data scale, read / write frequency, data type, etc. For structured simulation configuration parameters and operation index data that are frequently read and written, use traditional disk array storage in combination with relational database indexes (B-tree, hash index) to accelerate queries; for massive unstructured simulation logs and monitoring videos, store them on inexpensive large-capacity hard disk clusters based on distributed file systems (Ceph, HDFS) to achieve reliable and low-cost storage.

[0082] ii. Indexing Strategy Development: Create indexes on the corresponding tables for frequently used query fields (such as retrieving device status by device number or retrieving product quality data by product batch number), balancing index maintenance costs with query performance improvements, and avoiding over-indexing; design composite indexes for complex query scenarios (simultaneously filtering running data by time and device type) to optimize retrieval speed, ensure rapid data location and retrieval, and meet the real-time interaction and analysis needs of simulation experiments.

[0083] ④ Supplement to the object-oriented data model (applicable to some scenarios) i. Class and Object Mapping: In complex simulation modeling scenarios, data models are designed using object-oriented thinking. Various objects in the simulation (such as virtual robots and intelligent logistics vehicles) are abstracted into classes. Classes have attributes (robot's coordinate position, power, and load capacity) and methods (vehicle's movement and loading / unloading operations). Objects are instances of classes. Data is flexibly organized through mechanisms such as class inheritance and polymorphism to fit the dynamic interaction logic of the simulation.

[0084] ii. Persistence Design: Utilize Object Relational Mapping (ORM) frameworks (such as Hibernate and MyBatis) to connect object-oriented data models with relational databases, enabling efficient storage and retrieval of object data in relational tables. This seamlessly adapts to simulation development and data management under the object-oriented programming paradigm, improving development efficiency and code maintainability.

[0085] (4) Design principles ① Independence: Changes in physical data should not affect the logical and conceptual models.

[0086] ②Simplicity: The data model should be as simple as possible, follow paradigm theory, and be easy to understand and maintain. Excessive model complexity will lead to reduced data processing efficiency and decrease the company's competitiveness.

[0087] ③ Consistency: The data model should maintain consistency to avoid data redundancy and duplication, thereby saving storage space and preventing update anomalies. This helps improve data processing efficiency and reduce data maintenance costs.

[0088] ④ Scalability: The data model should have good scalability, capable of being adjusted as business needs change. This helps improve data processing efficiency and reduce enterprise risk.

[0089] ⑤ Efficiency: The data model should possess high processing, query, and storage efficiency, with well-designed indexes and partitions to quickly identify the value of the data. This helps improve data processing efficiency and reduce the company's operating costs.

[0090] (5) Tools and techniques - CASE tools (Computer-Aided Software Engineering): such as ERwin, PowerDesigner, Lucidchart, etc., are used for the design and documentation of data models.

[0091] - Database Management System (DBMS): such as MySQL, PostgreSQL, Oracle, SQL Server, etc., used to implement logical and physical data models.

[0092] Data model design is an iterative process that requires close collaboration with business analysts, database administrators, and end users to ensure that the design meets business needs and has good performance and scalability.

[0093] 4. Data fusion Data fusion refers to an information processing technology that uses computers to automatically analyze and synthesize several observational information sources obtained in a time sequence under certain criteria in order to complete the required decision-making and evaluation tasks. It provides more accurate and comprehensive information support through the collection, transmission, integration, filtering, and synthesis of multi-source data.

[0094] For integrated simulation data management, data fusion is a crucial step. It is of great significance for improving the accuracy and reliability of simulation experiments.

[0095] Integrated simulation experiments are a complex process involving the collaborative work of multiple modules and systems. Data management technology plays a crucial role in this process, ensuring the accuracy, integrity, and consistency of data, and providing strong support for the simulation experiments.

[0096] Data fusion technology, as an important component of data management technology, extracts valuable information by collecting, transmitting, integrating, filtering, and synthesizing data from different sources, providing more accurate and comprehensive data support for simulation experiments. Therefore, it is an indispensable technology.

[0097] In integrated simulation experiments, data fusion technology helps us better understand and analyze experimental results, and discover potential patterns and trends. At the same time, it can also improve the accuracy and reliability of experiments, and reduce errors and uncertainties.

[0098] (1) The concept of data fusion technology Data fusion technology is a crucial component of integrated simulation data management. It involves comprehensively processing data from multiple different data sources (such as various simulation models, sensors, and experimental equipment) to obtain more accurate, comprehensive, and reliable information than from a single data source. For example, in simulation experiments of naval air defense systems, data from early warning radar (target position, speed, and other early warning information), data from ship sensors (ship status information), and data from other sensors (sea and air situation awareness, etc.) can be fused using data fusion technology to form a complete and accurate information set regarding the sea and air situation.

[0099] (2) Levels of data fusion ① Data layer fusion: The data layer is the most basic fusion layer. It directly fuses raw data from different data sources. For example, in a simulation of a maritime battlefield early warning system, raw data collected by sensors at different locations (such as navigation wind direction, navigation wind speed, ocean current direction, ocean current speed, wave height, swell height, sea surface target set, and sea and air target set) are fused. These data may differ in format, accuracy, etc. Data layer fusion requires calibration and preprocessing, such as unifying data format and time alignment, before merging these data to form a new dataset, providing rich raw material for subsequent analysis.

[0100] ② Feature layer fusion: Feature layer fusion is a fusion process performed after features have been extracted from the original data. For example, in a simulation experiment of image recognition and target tracking, the image data acquired from the camera is first used to extract the target's features, such as shape, color, and texture feature vectors. Then, image features from multiple cameras or different time series are fused. Through feature matching, combination, and other methods, a more representative and robust feature description is obtained, which is used to more accurately identify and track the target.

[0101] ③ Integration of decision-making levels: Decision-level fusion occurs after each data source has made its initial decision. For example, in a fault diagnosis simulation system, different diagnostic subsystems make judgments about the equipment fault status based on their own rules and data (e.g., one subsystem judges the equipment to be likely mechanically faulty, while another subsystem judges it to be likely electrically faulty). Decision-level fusion integrates the decisions of these subsystems and, based on a certain fusion algorithm (such as Bayesian inference, D-S evidence theory, etc.), arrives at the final equipment fault diagnosis result. This result comprehensively considers the judgments of each subsystem, resulting in higher reliability.

[0102] (3) Key technologies and methods for data fusion ① Weighted average method: This is a simple yet effective data fusion method. When data from multiple data sources has similar physical meaning and comparable reliability, a weighted average method can be used. For example, in a temperature measurement simulation experiment, three different temperature sensors measure the temperature of the same object, and their accuracies may vary slightly. A weight is assigned to the data from each sensor based on its accuracy (higher accuracy results in a larger weight), and then these weighted temperature data are averaged to obtain the fused temperature value.

[0103] ② Kalman filtering method: Kalman filtering is primarily used for data fusion in dynamic systems. It provides an optimal estimate of the system's state through two steps: prediction and updating. For example, in an aircraft navigation simulation system, data from the Inertial Navigation System (INS) and the Global Positioning System (GPS) need to be fused. Kalman filtering continuously predicts the aircraft's position state based on factors such as the noise characteristics of the INS and GPS data, and updates this prediction using new measurement data (from INS and GPS), thus obtaining a more accurate fused aircraft position information.

[0104] ③ Bayesian network method: Bayesian networks are graphical models based on probabilistic reasoning. In data fusion, they can be used to represent the probabilistic relationships between different data sources and to perform data fusion and inference based on these relationships. For example, in a medical diagnostic simulation system, different examination results (such as blood tests, X-ray examinations, etc.) serve as data sources. By constructing a Bayesian network, the probabilistic relationships between these examination results (such as the probability that an abnormal blood test result may lead to a certain disease, and the change in the probability of the disease after combining X-ray examination results, etc.) can be represented. Then, data fusion is performed according to Bayesian inference rules to obtain a more accurate disease diagnosis probability.

[0105] (4) Application scenarios of data fusion technology ① Military field: In military combat simulation systems, data fusion technology can integrate data from various reconnaissance methods (such as radar, UAV reconnaissance, and satellite reconnaissance). For example, in air defense combat simulation, fusing data such as the position and speed of aerial targets detected by radar with target image feature data captured by UAVs can more accurately identify target types and determine target intentions, providing stronger support for air defense combat decision-making.

[0106] ② Industrial automation: In industrial production process simulation experiments, data fusion technology can be used for equipment status monitoring and quality control. For example, in a simulation system for an automobile manufacturing production line, fusing robot operating status data, component quality inspection data, and production environment data (such as temperature and humidity) can promptly detect anomalies in the production process, such as early warnings of equipment failures or trends of declining product quality, thereby enabling corresponding adjustments and optimizations to be taken.

[0107] 4.1 Data Fusion Preprocessing (1) The importance of data fusion preprocessing Data preprocessing is a crucial step in the data fusion process. Because data from different data sources often differ in format, range, accuracy, and noise levels, direct data fusion can lead to erroneous results. For example, in a multi-sensor ocean monitoring simulation, the data units collected by different sensors may differ—some in nautical miles per hour, others in meters per second. Without preprocessing, these data cannot be correctly fused to assess environmental quality. Data preprocessing effectively improves data quality, providing a more accurate and reliable data foundation for subsequent data fusion.

[0108] (2) Main contents of data fusion preprocessing ① Data cleaning Data cleaning primarily involves removing noise, outliers, and duplicate data. Noisy data may be caused by sensor malfunctions, communication interference, or other reasons. For example, in an earthquake monitoring simulation experiment, a seismograph might be affected by nearby electromagnetic interference, producing small fluctuations that do not conform to the characteristics of actual seismic waves. This noisy data needs to be identified and removed. Handling outliers is also crucial; outliers may represent real extreme cases, but they could also be erroneous data. Statistical methods (such as...) can be used. Outliers can be identified and handled using principles or clustering-based methods. For duplicate data, depending on the specific circumstances, one copy should be retained or special merging procedures should be performed.

[0109] ② Data format is standardized Data formats from different data sources can vary greatly. This includes data storage formats (such as binary, text, etc.), data structures (such as arrays, linked lists, etc.), and data representation methods (such as date formats, numerical precision, etc.). For example, in a combat simulation system, battlefield troop deployment data might be stored in a structured database table with a precision of "minutes," while battlefield target data obtained by sensors might be a data stream file with a precision of "meters" or "decimeters." These different data formats need to be converted to a unified format for fusion. For numerical data, it may be necessary to standardize the data precision and range, such as unifying latitude and longitude data of different precisions to "meters" as the unit, accurate to three decimal places.

[0110] ③ Data normalization Data normalization maps data to a specific range, typically [0,1] or [-1,1]. This is because data from different data sources may have different dimensions and numerical ranges. For example, in a naval warfare simulation system, submarine diving depth data might be between [0,300] meters, while aircraft altitude data might be between [0,10000] meters. If these data are directly fused, the influence of depth data might be masked by altitude data. Normalization allows different data to have the same weighting basis during the fusion process. Commonly used normalization methods include min-max normalization and Z-score normalization.

[0111] ④ Time alignment In integrated simulation experiments, data from different sources may be collected at different times. For example, in a naval warfare simulation system, radar data acquisition frequency may be on the order of milliseconds, while shipborne sonar data update frequency may be on the order of tens of seconds. To accurately fuse these data, they need to be aligned in time. Mathematical methods such as interpolation and extrapolation can be used to adjust data from different times to the same time reference based on the time series characteristics of the data, ensuring data consistency in the time dimension.

[0112] (3) Technical means of data fusion preprocessing ① Statistical analysis methods Statistical analysis tools are used to understand the distribution characteristics of data, followed by data cleaning and normalization. For example, statistical measures such as the mean, standard deviation, and median are calculated. The principle (i.e., data points falling within the mean plus or minus three standard deviations are considered normal) is used to identify and remove outliers. Data normalization, such as Z-score normalization, is based on the mean and standard deviation of the data, and the calculation formula is... ,in, It is the new data after normalization. It is the raw data. It is the mean. That is the standard deviation.

[0113] ② Interpolation algorithm Interpolation algorithms are commonly used techniques in time alignment and handling missing data values. Linear interpolation is the simplest interpolation method. Assuming known data points... and ,for and The interpolated value of x between the given values ​​can be obtained using the formula... The calculation yields the results. In addition, there are more complex interpolation methods such as polynomial interpolation and spline interpolation, which can be selected based on the characteristics of the data and accuracy requirements to more accurately fill data gaps or perform time alignment.

[0114] ③ Data conversion tools Many software tools and programming languages ​​offer data transformation capabilities. For example, in Python, the Pandas library provides convenient preprocessing operations such as data cleaning (e.g., the `drop_duplicates` function for removing duplicate data), formatting (e.g., the `astype` function for converting data types), and normalization (e.g., using a custom function combined with the `apply` method for normalization). Similarly, in database management systems, SQL statements can be used for data cleaning and format conversion, such as the `SELECT DISTINCT` statement for removing duplicate data, and the `CAST` and `CONVERT` functions for data type conversion.

[0115] 4.2 Data Integration In integrated simulation data management technology, data fusion and data integration are key technical aspects. Data fusion refers to the process of integrating data from different sources, formats, or structures into a unified data model or dataset. The aim of this process is to improve the integrity, accuracy, and usability of data by integrating diverse data to support deeper analysis, decision-making, or application development. Types of data fusion include structured data fusion, semi-structured data fusion, and unstructured data fusion, each with its specific fusion methods and challenges.

[0116] Data integration is a subset of data fusion, focusing primarily on extracting data from different sources and formats, transforming it into a unified format or structure, and then loading it into the target system. The key to this process is ensuring data consistency, integrity, and accuracy, while also addressing issues such as data quality, security, and privacy protection.

[0117] In practical applications, the fusion and integration of multi-source and multimodal data is a complex process involving multiple levels, including data processing models, phase-based strategies, feature representation-based methods, and semantic-based applications. The application of these technical approaches and methods enables more effective data processing and utilization in various fields such as multi-sensor systems, biomedicine, environmental research, and urban development.

[0118] For large-scale complex system integration simulation, there are three technical approaches to achieving heterogeneous model integration: interfaces between commercial tools, custom development of commercial tool interfaces, and model integration based on FMI interface standards. Each approach has its advantages and disadvantages, but overall they all aim to achieve joint simulation of models from different disciplines and efficient data management.

[0119] In summary, data fusion and data integration technologies play a crucial role in integrated simulation data management. By integrating and optimizing data, the accuracy and efficiency of experiments can be improved, providing strong support for decision-making and application development.

[0120] (1) Definition of data integration Data integration is a crucial step in data fusion; it involves combining data from multiple different data sources and storing, managing, and using it in a unified manner. These data sources can include databases, file systems, sensor networks, and other forms. For example, in an integrated simulation experiment of an intelligent transportation system, data may come from different devices such as traffic cameras, vehicle sensors, and road weather monitoring stations. Data integration aims to consolidate this scattered data into a single data warehouse or data platform.

[0121] (2) The purpose of data integration The core tasks of data integration include: data extraction, data transformation, and data loading. Data extraction refers to obtaining the required data from various data sources; data transformation involves converting the extracted data into a unified format and structure for subsequent processing; and data loading involves loading the transformed data into the target database or data warehouse for later analysis and use. The main purpose is: ① Eliminating Data Silos: In different departments or systems, data is often stored and managed independently. For example, within a combat group, each service branch has its own data system and combat simulation system, as well as its own business processes, database systems, and simulation systems. Data integration aims to break down these data barriers between departments or systems, enabling the combat group to have a comprehensive understanding of the business processes of each service branch. By integrating various equipment data, training data, and combat simulation data, combat plans can be better adjusted according to the actual situation of the service branches, avoiding misguided or chaotic command.

[0122] ② Provide a unified data view: Offer users and applications a consistent and comprehensive view of data. In integrated simulation experiments, researchers may need to view data from multiple related systems simultaneously to analyze experimental results. Taking integrated simulation of aircraft in the aerospace field as an example, it is necessary to integrate flight performance data, structural mechanics data, avionics system data, etc., to provide engineers with a unified view so that they can better evaluate the overall performance of the aircraft.

[0123] (3) Data integration methods ① Data Warehouse Methodology This is a relatively common approach. It first extracts data from various data sources, then cleans and transforms it before loading it into a data warehouse. For example, in a combat equipment data analysis system, data is extracted from multiple data sources such as equipment order databases, user information databases, and equipment inventory databases. Before being loaded into the data warehouse, the data needs to be cleaned to remove duplicates and errors, and its format needs to be standardized, such as changing date formats and data encoding. The processed data is then stored in the data warehouse for data analysis tools and users to query and analyze.

[0124] ② Federated Database Approach A federated database system allows different database systems to maintain their independence while enabling data sharing and integration. Communication and data exchange between these systems are achieved through a federated layer. For example, in a cross-regional military information system, different regions might use different equipment management information management systems (e.g., some based on Oracle, others on SQL Server). By establishing a federated database, data sharing and integration of equipment information, equipment usage status reports, and other data can be achieved without altering the existing database systems of each unit. This facilitates data integration operations at the command post and supports the development of operational plans.

[0125] (4) Challenges of data integration ① Semantic differences in data: Different data sources may have different definitions and understandings of the same data. For example, the definition of "speed" may differ among different branches of the armed forces. Naval vessels may define speed as "knots" (nautical miles per hour), while naval aviation units may define speed as "meters per second," and so on. Resolving these semantic differences requires establishing a unified data dictionary or semantic model to standardize data definitions.

[0126] ② Data Quality Issues: The data quality from data sources varies greatly, and issues such as missing data, errors, and inconsistencies may exist. In data collected from multiple channels from certain combat equipment databases, some data is assumed to be from previous models. However, in reality, different units have different equipment deployment times, which may result in different models, and even the same model may have improvements to address certain issues. This leads to discrepancies in the assumed data, resulting in inaccurate data. Data cleaning and verification are necessary during data integration to improve data quality.

[0127] ③ Data Update Synchronization: Data from data sources is constantly being updated, and maintaining synchronization between the integrated data and the data from each data source is a challenge. For example, in an equipment data integration system for a command post, inventory data for certain unit equipment parts may be updated in real time as users perform maintenance. The integration system needs to promptly acquire these updates and reflect them in the integrated data to ensure data timeliness and provide accurate data support for formulating equipment support plans.

[0128] 4.3 Data Analysis In integrated simulation data management technology, the data analysis phase of data fusion plays a crucial role. It is not only a process of in-depth exploration and mining of the integrated data, but also a key step in discovering hidden value and optimizing simulation results. It directly impacts the accuracy and effectiveness of simulation experiments.

[0129] (1) Core tasks of data analysis The core tasks of data analysis lie in preprocessing, exploratory analysis, and data mining of the integrated data. The preprocessing stage requires data cleaning, transformation, and normalization to ensure data quality and consistency. Exploratory analysis uses descriptive and graphical methods to reveal the basic characteristics and trends of the data. Data mining employs advanced techniques such as association rule mining and sequence rule mining to discover potential patterns and regularities within the data.

[0130] (2) Data analysis techniques and methods: ① Statistical analysis Statistical analysis is the most fundamental and commonly used method in data analysis. It primarily calculates descriptive statistics such as mean, median, standard deviation, distribution parameters, and density parameters to clarify the centrality and dispersion characteristics of data. For example, when analyzing the effectiveness of satellites in searching for maritime targets, it is necessary to collect information on the positions, numbers, and orbital parameters of satellites from various parties. Through simulation calculations, the probability and stability of each party's search for and discovery of specific targets in a specific area can be analyzed. Standard deviation can be used to understand the degree of fluctuation and identify vulnerable periods. Furthermore, correlation analysis can be used to uncover the relationships between variables, such as sea surface wind speed and waves, sea state, target characteristic parameters, and navigation speed, and the enemy's search and detection capabilities, in order to optimize troop deployment and operational plans.

[0131] ② Data Mining Algorithms: Clustering algorithms group similar data objects together. In target behavior analysis scenarios, clustering is performed based on past data, current data, and other factors to identify target intentions and commander characteristics, enabling differentiated and targeted countermeasures. Classification algorithms, such as decision trees and neural networks, are used to predict categories. In target identification, the target's danger level is determined based on platform characteristics, electromagnetic spectrum characteristics, and the characteristics of the operators and commanders. Association rule mining can discover frequent co-occurrence patterns, find combat correlations from reconnaissance data, deploy strategies, and create a winning posture, laying the foundation for winning the battle.

[0132] ③ Data analysis and fusion methods based on DS evidence theory This method is primarily used to process data with uncertainty. It decomposes a complex problem into multiple sub-problems, then uses DS evidence theory to synthesize the analytical conclusions of these sub-problems, ultimately arriving at a fused result. This method includes steps such as constructing an identification framework, assigning basic probabilities, calculating confidence and likelihood, and fusing evidence.

[0133] ④ SPDM (Simulation Process and Data Management) Solution SPDM is an integrated solution for managing simulation processes and data. It aims to help research teams effectively manage data and related workflows involved in simulations, improve the sharing, reuse, and reliability of simulation data, and promote cross-team collaboration and knowledge sharing. SPDM systems typically include functions such as data management, workflow management, collaborative work and teamwork, data analysis, and visualization.

[0134] ⑤ Comparative analysis of experimental simulation data In naval tactics research, comparative analysis of experimental and simulation data is crucial. It helps to promptly identify anomalies in experiments, assess the reliability of physical experimental data, verify the credibility of digital simulation models and results, and improve and optimize physical experimental procedures. Simulation experimental data comparison and analysis systems can provide real-time comparison and analysis capabilities between experimental and simulation data, thereby enabling a new experimental mode that combines virtual and real data and allows for iterative feedback.

[0135] ⑥ Visual analysis It presents complex data relationships with intuitive charts, bar charts compare the advantage rates of different equipment and tactics, line charts show the trend of key system indicators over time, heat maps reflect the density of regional data (such as troop distribution), and scatter plots explore the correlation of two-dimensional variables, helping analysts quickly capture patterns and discover anomalies.

[0136] (3) Standardized Data Analysis Process ① Data fusion preprocessing: The purification stage handles missing values ​​(deleting records with many missing values, imputing mean values, etc.), corrects erroneous values ​​(identifying and correcting them according to business rules and statistical boundaries), smooths noisy data (using filtering algorithms to reduce interference), and unifies data dimensions (normalization and standardization), laying a solid foundation for accurate analysis.

[0137] ② Model construction and selection: Based on the objective, select appropriate methods and tools to construct the model. Taking the prediction of equipment life as an example, compare the goodness of fit and prediction accuracy of linear regression and deep learning models, weigh the complexity and effect to select the optimal one, and also use cross-validation and other methods to prevent overfitting.

[0138] ③ Result Interpretation and Evaluation: Interpret the analysis results, judge their rationality in conjunction with business knowledge, evaluate model performance, check indicators such as R² for regression and accuracy for classification, and optimize the model and improve the analysis process based on the results to ensure that the analysis output is practical and reliable.

[0139] In integrated simulation experiments, data analysis is widely and deeply applied. It helps us understand the behavioral characteristics of the simulation system, identify potential fault points and optimization opportunities. Through in-depth analysis of simulation data, we can evaluate the system's performance, compare the effects of different simulation schemes, and thus guide the improvement and optimization of the simulation experiment.

[0140] Furthermore, data analysis can support the integration and debugging process of simulation experiments. During this phase, it's crucial to ensure data interaction and functional coordination between various simulation modules. Through data analysis, we can monitor data flow, verify data consistency and accuracy, and promptly identify and resolve issues arising during integration and debugging.

[0141] However, as simulation experiments become increasingly complex and large-scale, the amount of data explodes, placing enormous computational pressure on data analysis. To address this challenge, we need to employ efficient data processing algorithms and distributed computing technologies to improve the efficiency and accuracy of data analysis.

[0142] In fact, interpreting and visualizing data analysis results is a significant challenge. How to present complex analytical results in an intuitive and easy-to-understand way so that decision-makers can quickly comprehend and apply them is a problem that the field of data analysis needs to continuously explore and solve. To address this, we can leverage data visualization tools and techniques to display analytical results in the form of charts, animations, and other formats, thereby improving the readability and comprehensibility of the data.

[0143] Data analysis plays a crucial role in the data fusion stage of integrated simulation data management technology. Through in-depth exploration and mining of the integrated data, it provides valuable insights and discoveries, helping to optimize simulation results and improve system performance. At the same time, we also need to continuously address the challenges in the data analysis process to drive technological innovation and progress.

[0144] In summary, integrated simulation data management technology is a complex yet crucial research area. By constructing a rational data management architecture and employing key technologies such as data standardization, data model design, data security and sharing, and data fusion, challenges such as large data volumes, multi-source heterogeneity, and high real-time requirements can be effectively addressed. This not only helps improve the data management level of integrated simulation but also provides strong data support for scientific research, engineering applications, and decision-making in related fields, promoting the in-depth application and development of integrated simulation in more fields. Future research should further focus on the integrated application of data management technology with emerging technologies (such as artificial intelligence and blockchain) to continuously improve the intelligence, security, and efficiency of integrated simulation data management.

[0145] 5. Data security and sharing Data security and sharing are two important aspects of modern information management and key issues in today's digital age. They are closely linked yet face challenges that require balancing. A balanced approach is needed to ensure data confidentiality, integrity, and availability while promoting the rational flow and utilization of information.

[0146] (1) Data security Data security refers to protecting the confidentiality, integrity, and availability of data, preventing data leakage, tampering, or unauthorized access. In the digital age, data has become a core asset for enterprises, making data security paramount. Data security encompasses the measures and processes for protecting data from unauthorized access, leakage, tampering, damage, or loss. Data security measures include the following key components.

[0147] ① Data encryption. Sensitive data is encrypted to ensure that it is not read by unauthorized persons during transmission and storage.

[0148] ② Access control. Implement strict user authentication and authorization mechanisms to ensure that only authorized users can access specific data.

[0149] ③ Network security. Use firewalls, intrusion detection systems, and antivirus software to protect the network from attacks.

[0150] ④ Data backup and recovery. Back up your data regularly so that you can recover it if it is lost or damaged.

[0151] ⑤ Security Audit. Conduct regular security audits to detect potential security threats and vulnerabilities.

[0152] ⑥ Physical security. Protect the physical security of the data center to prevent data leakage or damage caused by physical contact.

[0153] ⑦ Data anonymization. Sensitive information is anonymized before sharing or publishing data to protect personal privacy.

[0154] ⑧ Compliance. Adhere to relevant data protection regulations and standards, such as GDPR (General Data Protection Regulation) and HIPAA (Health Insurance Portability and Accountability Act).

[0155] (2) Data sharing Data sharing refers to the process of providing data to other individuals, organizations, or the public. The following factors are key considerations in the data sharing process: ① Purpose of sharing: Clarify the purpose of data sharing and ensure that the use of data is consistent with the original purpose of collection.

[0156] ② Scope of sharing: Determine the scope of shared data to avoid sharing irrelevant or excessive information.

[0157] ③ Privacy protection: When sharing data, ensure that personal privacy is not violated, especially when sensitive personal information is involved.

[0158] ④ Data quality: Ensure that the shared data is accurate, complete and consistent.

[0159] ⑤ Authorization and Agreements: Sign agreements with all parties involved in data sharing to clarify the purpose, limitations on use, and responsibilities of the data.

[0160] ⑥ Technology Platform: Use a secure and reliable technology platform for data sharing, such as APIs, data sharing platforms, etc.

[0161] ⑦ Monitoring and Auditing: Monitor and audit the data sharing process to ensure that data use complies with the agreement.

[0162] (3) Balancing data security and sharing To ensure a balance between data security and sharing, the following security measures are primarily adopted, but not limited to: ① Data desensitization and anonymization: Before data sharing, data is desensitized and anonymized to protect personal privacy and data security.

[0163] ② Sign a data sharing agreement: Clarify the rights and obligations of both parties in data sharing to ensure the legality and compliance of data during the sharing process.

[0164] ③ Establish a data sharing platform: Exchange and share data through a data sharing platform to ensure data security and controllability.

[0165] Alternatively, some more detailed measures can be taken: ① Data Classification. Data is classified according to its sensitivity and importance, and different security measures and sharing strategies are implemented for different categories of data.

[0166] ② Minimization principle. Only share data necessary to achieve the goal, reducing potential risks.

[0167] ③ Transparency. Improve the transparency of data sharing, allowing data subjects to understand how their data is shared and used.

[0168] ④ Continuous monitoring. Continuously monitor data usage during the data sharing process to ensure data security.

[0169] In conclusion, data security and sharing are crucial issues in the digital age. Taking into account the needs of data security and sharing, measures such as formulating reasonable data security policies and standards, strengthening data security management and monitoring, improving the technical level of data sharing, and enhancing cooperation and communication can help achieve a balanced development of data security and sharing.

[0170] The above provides a detailed description of a simulation data management method based on a service-oriented integrated architecture provided by the embodiments of the present invention. Specific examples have been used to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only for the purpose of helping to understand the method and core ideas of the present invention. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of the present invention. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A simulation data management method based on a service-oriented integrated architecture, characterized in that, include: A service-oriented integrated data management architecture is built based on the data acquisition layer, data storage layer, data management layer, data management layer, and data service layer. Based on a service-oriented integrated data management architecture, data standardization is carried out, data models are designed, and the collected multi-source heterogeneous data is preprocessed, integrated, and analyzed according to data standards and data models to complete data fusion and storage management, while ensuring a balance between data security and sharing.

2. The method according to claim 1, characterized in that, The data acquisition layer is used to collect data from various data sources, and adopts corresponding acquisition interfaces and protocols for different data sources. The data storage layer is used to address the storage needs of large amounts of data using a distributed storage architecture; The data management layer is used to provide functions for organizing, indexing, classifying, and managing metadata of data; The data service layer is used to provide data access interfaces to meet the data needs of different users and applications.

3. The method according to claim 2, characterized in that, The data standardization is used to transform different data according to preset rules so that they have the same standard and scope, and data standardization is associated with data sharing.

4. The method according to claim 3, characterized in that, The design data model includes: Determine the data that the system needs to store, and understand how the data is used and the business rules; Use ER diagrams or UML class diagrams to represent the relationships between entities and data, and to define the attributes of entities and relationships; Transform the conceptual model into a logical model, determine the table structure, fields, primary keys, foreign keys and indexes, and define data integrity constraints; Design the physical storage structure of the database based on the logical model, considering performance optimization, and determine storage allocation and access paths.

5. The method according to claim 4, characterized in that, The data model includes a conceptual data model, a logical data model, a physical data model, and an object-oriented data model supplement. The conceptual data model includes identifying entities and relationships and defining attributes; the logical data model includes determining relation schemas and normalization processing; the physical data model includes selecting storage structures and formulating indexing strategies; and the object-oriented data model supplement includes class-object mapping and persistence design.

6. The method according to claim 5, characterized in that, The data fusion is an information processing technology used to automatically analyze and synthesize several observation information obtained in time sequence under preset criteria in order to complete the required decision-making and evaluation tasks. The data fusion includes data layer fusion, feature layer fusion and decision layer fusion.

7. The method according to claim 6, characterized in that, The data fusion preprocessing is used to remove noise, outliers and duplicate data from the data through data cleaning, unify the data format of different data sources after cleaning, normalize the data after the data format is unified, and finally align the normalized data by time.

8. The method according to claim 7, characterized in that, The data integration is used to extract data from different sources and formats, convert it into a unified format or structure, and then load it into the target system. The data integration methods include data warehouse methods and federated database methods.

9. The method according to claim 8, characterized in that, The data analysis is used to preprocess, explore, and mine the integrated data. In the preprocessing stage, the data needs to be cleaned, transformed, and normalized to ensure the quality and consistency of the data. Exploratory analysis reveals the basic characteristics and trends of the data through descriptive analysis and graphical analysis. Data mining uses association rule mining and sequence rule mining to discover potential patterns and regularities in the data.

10. The method according to claim 9, characterized in that, The aforementioned data security and sharing mechanisms are used to perform data encryption, access control, network security, data backup and recovery, security auditing, physical security, data anonymization, compliance, sharing purpose, sharing scope, privacy protection, data quality, authorization and agreements, technical platforms, and monitoring and auditing operations. They achieve a balance between data security and sharing through data anonymization and desensitization, signing data sharing agreements, and establishing a data sharing platform. Furthermore, they categorize data according to its sensitivity and importance, implement different security measures and sharing strategies for different categories of data, share only the data necessary to achieve the objectives, improve the transparency of data sharing, and continuously monitor data usage during the data sharing process to ensure data security.

Images