Multi-source heterogeneous data fusion method and device based on satellite digital twin system

Abstract

The embodiment of the invention provides a multi-source heterogeneous data fusion method and device based on a satellite digital twinning system, and relates to the technical field of digital twinning, and the method comprises the steps: receiving original data of different data sources; for the original data corresponding to each data source, based on an analysis rule corresponding to the data source, performing data analysis on the original data, and recombining reference data obtained after analysis to obtain target data corresponding to the original data; determining to-be-fused business parameters and semantic tags corresponding to the business parameters according to target data corresponding to the original data, and generating a business sequence; extracting a time sequence feature vector from the service sequence, correcting the extracted time sequence feature vector based on a semantic rule base, and determining the state of the target equipment based on the corrected time sequence feature vector; the target device is a device corresponding to each data source, and the method effectively integrates the data of different data sources.

Description

Technical Field

[0001] This application relates to the field of digital twin technology, and in particular to a method and device for multi-source heterogeneous data fusion based on a satellite digital twin system. Background Technology

[0002] With the continuous development of information technology, digital twin technology has been widely used in many fields. However, in practical application scenarios, especially in the field of low-Earth orbit satellites, the fusion of multi-source heterogeneous data brings huge challenges to the construction of digital twin systems.

[0003] Because data from external systems, semi-physical systems, and internal business models are related to specific business processes and rules, data from different sources may correspond to specific communication protocol rules. Furthermore, due to the diversity of data formats, real-time requirements, and large volumes of data, different storage methods and structures may be involved both inside and outside the system, and the database management systems used may also differ. Therefore, establishing a digital twin model in the low-Earth orbit satellite domain involves a wide variety of entity models, resulting in a highly diverse range of data sources and formats.

[0004] In related technologies, only processing of a single type of data is supported, making it difficult to effectively integrate data from different data sources. This results in the digital twin system being unable to achieve good virtual-real comparison, thus making it impossible to accurately perform emergency analysis and processing. Summary of the Invention

[0005] This invention provides a method and device for multi-source heterogeneous data fusion based on a satellite digital twin system, which solves the problem of effectively integrating data from different data sources in the prior art.

[0006] In a first aspect, embodiments of this application provide a method for multi-source heterogeneous data fusion based on a satellite digital twin system, including: Receive raw data from different data sources; For the original data corresponding to each data source, based on the parsing rules corresponding to the data source, the original data is parsed, and the reference data obtained after parsing is reorganized to obtain the target data corresponding to the original data. Based on the target data corresponding to each original data, determine the business parameters to be integrated and the semantic tags corresponding to each business parameter, and generate a business sequence. Temporal feature vectors are extracted from the above business sequences. The extracted temporal feature vectors are then corrected based on a semantic rule base. The state of the target device is determined based on the corrected temporal feature vectors. The target device is the device corresponding to each data source.

[0007] In some embodiments, receiving raw data from different data sources includes: Create sub-threads corresponding to different data sources, wherein each sub-thread is configured with first configuration information corresponding to the corresponding data source, and the first configuration information includes at least one of communication protocol type, data transmission format and communication parameters; Each sub-thread receives raw data from the data source corresponding to the aforementioned sub-thread, based on the first configuration information corresponding to that sub-thread.

[0008] In some embodiments, the reorganization of the parsed reference data to obtain target data corresponding to the original data includes at least one of the following: Filter reference data of a preset type from various reference data, and convert the format of the filtered reference data into the format corresponding to the target data; Determine the value of the target data corresponding to the preset time based on the reference data; Map all reference data to the same spatial coordinate system; Reorganize the reference data into target data with the same data format.

[0009] In some embodiments, the above-mentioned reorganization of reference data into target data of the same data format includes: For each piece of raw data, obtain the second configuration information corresponding to the raw data; the second configuration information includes: data source information, data destination information, communication forwarding protocol, data type information, starting byte of the data field, and number of bytes occupied by the data field; the data type information includes: data type number and configuration rules corresponding to the data parameter number. Based on the parsed reference data and the second configuration information mentioned above, the frame header information and data field information corresponding to the target data are determined. The frame header information includes at least one of the following: data source information, data destination information, system time, and communication forwarding protocol. The data field information includes at least one of the following: data type number, data parameter number, and data value. Based on the aforementioned frame header information and data field information, the target data is constructed.

[0010] In some embodiments, before determining the service parameters to be merged and the semantic tags corresponding to each service parameter, the method further includes: Input the target data corresponding to each original data into the target model to generate a fact table and a dimension table. The fact table stores the core business indicators corresponding to each target data and the primary keys corresponding to each dimension table. The dimension table stores the semantic description information corresponding to the fact table. Store the aforementioned fact table and dimension table in the data warehouse.

[0011] In some embodiments, the process of determining the service parameters to be merged and the semantic tags corresponding to each service parameter, and generating a service sequence, includes: Extract the business parameters of the target data type from the fact table and dimension table above; Based on the semantic description information stored in the above dimension table, determine the semantic tags corresponding to each of the above business parameters. Each business parameter is combined with its corresponding semantic tag, and the combined result is sorted based on the time information corresponding to each business parameter to obtain the business sequence.

[0012] In some embodiments, the extraction of temporal feature vectors from the aforementioned service sequence, the correction of the extracted temporal feature vectors based on a semantic rule base, and the determination of the target device's state based on the corrected temporal feature vectors include: Input the above business sequence into a pre-trained recurrent neural network model; Based on the long short-term memory network layer in the above recurrent neural network model, the time-series features of the above business sequence are extracted to obtain the corresponding time-series characteristic vector. Based on the above semantic rule base, logical reasoning is performed on the above time-series feature vectors to obtain the corrected time-series feature vectors; Based on the output layer of the above recurrent neural network model, an output result carrying semantic labels is generated; the above output result is used to characterize the state of the above target device.

[0013] Secondly, embodiments of this application provide a multi-source heterogeneous data fusion device based on a satellite digital twin system, comprising: The data receiving module is used to receive raw data from different data sources; The data recombination module is used to parse the original data corresponding to each data source based on the parsing rules corresponding to the data source, and to reconstruct the reference data obtained after parsing to obtain the target data corresponding to the original data. The sequence generation module is used to determine the business parameters to be fused and the semantic tags corresponding to each business parameter based on the target data corresponding to each original data, and to generate a business sequence. The status determination module is used to extract time-series feature vectors from the above business sequences, correct the extracted time-series feature vectors based on the semantic rule base, and determine the status of the target device based on the corrected time-series feature vectors; the above target devices are the devices corresponding to each data source.

[0014] In some embodiments, the data receiving module described above is specifically used for: Create sub-threads corresponding to different data sources, wherein each sub-thread is configured with first configuration information corresponding to the corresponding data source, and the first configuration information includes at least one of communication protocol type, data transmission format and communication parameters; Each sub-thread receives raw data from the data source corresponding to the aforementioned sub-thread, based on the first configuration information corresponding to that sub-thread.

[0015] In some embodiments, the data reconstruction module reconstructs the parsed reference data based on at least one of the following methods: Filter reference data of a preset type from various reference data, and convert the format of the filtered reference data into the format corresponding to the target data; Determine the value of the target data corresponding to the preset time based on the reference data; Map all reference data to the same spatial coordinate system; Reorganize the reference data into target data with the same data format.

[0016] In some embodiments, the data restructuring module described above is specifically used for: For each piece of raw data, obtain the second configuration information corresponding to the raw data; the second configuration information includes: data source information, data destination information, communication forwarding protocol, data type information, starting byte of the data field, and number of bytes occupied by the data field; the data type information includes: data type number and configuration rules corresponding to the data parameter number. Based on the parsed reference data and the second configuration information mentioned above, the frame header information and data field information corresponding to the target data are determined. The frame header information includes at least one of the following: data source information, data destination information, system time, and communication forwarding protocol. The data field information includes at least one of the following: data type number, data parameter number, and data value. Based on the aforementioned frame header information and data field information, the target data is constructed.

[0017] In one embodiment, before the sequence generation module determines the service parameters to be fused and the semantic tags corresponding to each service parameter, it is further configured to: Input the target data corresponding to each original data into the target model to generate a fact table and a dimension table. The fact table stores the core business indicators corresponding to each target data and the primary keys corresponding to each dimension table. The dimension table stores the semantic description information corresponding to the fact table. Store the aforementioned fact table and dimension table in the data warehouse.

[0018] In one embodiment, the sequence generation module described above is specifically used for: Extract the business parameters of the target data type from the fact table and dimension table above; Based on the semantic description information stored in the above dimension table, determine the semantic tags corresponding to each of the above business parameters. Each business parameter is combined with its corresponding semantic tag, and the combined result is sorted based on the time information corresponding to each business parameter to obtain the business sequence.

[0019] In one embodiment, the sequence generation module described above is specifically used for: Input the above business sequence into a pre-trained recurrent neural network model; Based on the long short-term memory network layer in the above recurrent neural network model, the time-series features of the above business sequence are extracted to obtain the corresponding time-series characteristic vector. Based on the above semantic rule base, logical reasoning is performed on the above time-series feature vectors to obtain the corrected time-series feature vectors; Based on the output layer of the above recurrent neural network model, an output result carrying semantic labels is generated; the above output result is used to characterize the state of the above target device.

[0020] Thirdly, embodiments of this application provide a multi-source heterogeneous data fusion device based on a satellite digital twin system, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements any step of the multi-source heterogeneous data fusion method based on a satellite digital twin system described in the first aspect.

[0021] Fourthly, embodiments of this application provide a computer-readable storage medium storing computer program instructions thereon, which, when executed by a processor, implement any step of the multi-source heterogeneous data fusion method based on a satellite digital twin system described in the first aspect above.

[0022] Fifthly, embodiments of this application provide a computer program product, including a computer program stored in a computer-readable storage medium; when a processor of a memory access device reads the computer program from the computer-readable storage medium, the processor executes the computer program, causing the memory access device to perform any step in the multi-source heterogeneous data fusion method based on a satellite digital twin system described in the first aspect above.

[0023] The beneficial effects of this application are as follows: The above method can effectively integrate multi-source data that differ in format, frequency, and accuracy, breaking down data silos and enabling collaborative data utilization; it improves the efficiency and flexibility of data processing, can be customized and expanded to meet different application needs, and adapt to various complex scenarios and ever-changing business requirements; through the fusion analysis of multi-source heterogeneous data, the status of physical devices can be monitored in real time.

[0024] Other features and advantages of this application will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the application. The objectives and other advantages of this application may be realized and obtained by means of the structures particularly pointed out in the written description, claims, and drawings. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the 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.

[0026] Figure 1 A schematic diagram of a multi-source heterogeneous data fusion system based on a satellite digital twin system provided in this application embodiment; Figure 2 A flowchart illustrating a multi-source heterogeneous data fusion method based on a satellite digital twin system provided in this application embodiment; Figure 3 A schematic diagram of a multi-source heterogeneous data fusion device based on a satellite digital twin system provided in this application embodiment; Figure 4 This is a schematic diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will now be described in further detail with reference to the accompanying drawings.

[0028] The application scenarios described in this application are for the purpose of more clearly illustrating the technical solutions of this application, and do not constitute a limitation on the technical solutions provided in this application. Those skilled in the art will understand that with the emergence of new application scenarios, the technical solutions provided in this application are also applicable to similar technical problems. In the description of this application, unless otherwise stated, "multiple" means two or more.

[0029] It should be noted that the terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without inventive effort are within the scope of protection of this application.

[0030] With the continuous development of information technology, digital twin technology has been widely used in many fields. However, in practical application scenarios, especially in the field of low-Earth orbit satellites, the fusion of multi-source heterogeneous data brings huge challenges to the construction of digital twin systems.

[0031] Because data from external systems, semi-physical systems, and internal business models (including satellite models, gateway models, network models, etc.) are related to specific business processes and rules, data from different sources may correspond to specific communication protocol rules. Furthermore, due to the diversity of data formats, the real-time nature of data, and the large volume of data, different storage methods and structures may be involved inside and outside the system, and the database management systems used may also be different. At the same time, different geographic information system data in situation display may use different coordinate systems, data precision, and storage formats.

[0032] In summary, the establishment of digital twin models in the low-Earth orbit satellite domain involves a wide variety of physical models, resulting in diverse data sources and formats; this makes the data collection and integration process extremely complex. Furthermore, related technologies only support processing single-type data and struggle to effectively integrate data from different data sources. For instance, systems that process semi-physical data cannot handle business system data.

[0033] Therefore, how to effectively integrate and utilize data from different data sources to improve the accuracy and practicality of digital twin systems has become an urgent problem to be solved in the field of digital twin technology.

[0034] In response, this application provides a method for multi-source heterogeneous data fusion based on a satellite digital twin system, which can establish an interconnected digital twin system that supports cross-interface, cross-protocol, and high real-time connectivity, and realize the functions of virtual-real interactive comparison, real-time display and update, and analysis and prediction processing of the status of entity objects.

[0035] In some embodiments, the multi-source heterogeneous data fusion method based on a satellite digital twin system provided in this application is mainly implemented through modular architecture, standardized integration, and collaborative operation. Figure 1 A schematic diagram of a multi-source heterogeneous data fusion system based on a satellite digital twin system provided in this application embodiment, as shown below. Figure 1 As shown, the system consists of a configuration management module, a plug-in management module, a data aggregation module, a data processing module, a data storage and analysis module, a data forwarding module, and a situation display module.

[0036] The configuration management module is used to: manage the configuration files of each functional module, map plugins, protocols, and data processing methods; and automatically convert the configuration information of each type of configuration file into a unified format (such as JSON) file that can be loaded by the plugin management module.

[0037] The plugin management module is used to: provide a unified extension and integration interface, customize and extend functional modules according to different application needs, adapt to various complex scenarios and constantly changing business requirements, and decompose each functional module into reusable and extensible core plugins to achieve efficient construction and agile iteration.

[0038] In practical implementation, the data aggregation module, data processing module, data forwarding module, data storage and analysis module, and situation display module are all implemented through plugins. Plugins are primarily compiled to generate dynamic link library files. The plugin management module provides functions such as plugin registration and initialization, including two general integrated interfaces: plugin configuration (config() interface) and plugin loading (load() interface), enabling calls to all plugins. The plugin configuration interface is used to load the plugin attribute configuration file. For unified management, this file can be in .json format. The configuration information in this configuration file can include the plugin dynamic link library name, its module, and plugin type (mainly indicating whether it is loaded during the initialization phase to reduce system startup time). The plugin loading interface is used to load the dynamic link library according to the configuration file and perform initialization.

[0039] The data aggregation module is used to: provide a unified communication integration interface, receive raw data from data sources, and realize the aggregation of multi-source data; The data processing module is used to: provide a unified protocol parsing and conversion rule interface to unify heterogeneous data formats, and provide data preprocessing methods to improve data quality; The data storage and analysis module is used to: provide a unified extended integrated storage interface, adapt to various types of databases (relational and non-relational) according to different needs, and provide analysis and fusion algorithms to analyze and predict specific data; The data forwarding module is used to: provide a unified and scalable communication forwarding interface, and use different communication methods to forward data according to the needs of external systems and business models; The situation display module is used to subscribe to and display visual data from different data sources via a message bus.

[0040] Based on the above system, the multi-source heterogeneous data fusion process based on the satellite digital twin system in this application embodiment is as follows: The data aggregation module receives raw data (i.e., heterogeneous data) from different data sources based on its receiving configuration. It then sends the aggregated raw data to the parsing plugin of the data processing module. The parsing plugin allocates parsing sub-threads according to the protocol conversion configuration in the configuration management, and unpacks the data according to the data format requirements in the transmission rules, processing the data source code into unified data. This data is then sent to both the data forwarding module and the data storage and analysis module for storage. The forwarding plugin of the data forwarding module pushes the data to the network in real time according to the corresponding mapping relationship in the forwarding configuration. The display plugin of the situation display module uses a message mechanism to listen for and receive corresponding data based on the data types of the topics it subscribes to. The data storage and analysis module fuses and analyzes the heterogeneous data, and sends the analysis results to the data forwarding module.

[0041] The system provided in this application adopts a plug-in architecture design and establishes a unified standard interface protocol, which enables different plug-ins to be developed and deployed in accordance with the unified interface specifications, thereby improving the scalability and compatibility of the system. At the same time, the unified data transmission specifications and formats between different links also reduce the system's maintenance costs and development difficulty. The unified processing of heterogeneous data reduces the cost of data analysis and usage.

[0042] It should be noted that the above Figure 1 This is merely a schematic diagram of one possible system architecture; the systems in the embodiments of this application are not limited to the above structure.

[0043] Figure 2 A flowchart illustrating a multi-source heterogeneous data fusion method based on a satellite digital twin system provided in this application embodiment; as shown Figure 2 As shown, this application provides a method for multi-source heterogeneous data fusion based on a satellite digital twin system, including: Step S201: Receive raw data from different data sources; In some embodiments, this application pre-stores a receiving configuration file, which may include first configuration information corresponding to various data source types. The first configuration information may include at least one of the following: communication protocol type, data transmission format, data acquisition frequency, plugin name, data source name, communication details (i.e., communication parameters), and verification rules. The verification rules may configure the conditions that the raw data from the corresponding data source must meet, such as field length and the value corresponding to a specified byte. Optionally, the receiving configuration file can be uniformly managed by a configuration management module, and can be loaded from the configuration management module during initialization.

[0044] By configuring the corresponding primary configuration information for different types of data sources, it is possible to adapt to various data transmission protocols, thereby achieving accurate data reception.

[0045] As an optional implementation, the above-described receiving of raw data from different data sources includes: Create sub-threads corresponding to different data sources, where each sub-thread is configured with first configuration information corresponding to the corresponding data source; Each sub-thread receives raw data from the data source corresponding to the aforementioned sub-thread, based on the first configuration information corresponding to that sub-thread.

[0046] In some embodiments, this application creates different sub-threads using a thread pool to receive the original data from different data sources, thereby improving receiving efficiency and avoiding data loss.

[0047] In some embodiments, the different data sources mentioned above may include: semi-physical device data, external system business data, and internal system simulation data, etc. The first configuration information corresponding to each of them may be as follows: For semi-physical device data, the plugin can receive external device data according to the specific communication protocol of the semi-physical device (such as Modbus, CAN bus, etc.); For external system business data, the plugin can obtain external business system data through User Datagram Protocol (UDP), Transmission Control Protocol (TCP), etc., or subscribe to external business system data through message bus, such as Kafka, DDS, etc.; For internal system simulation data, the plugin can communicate through UDP, Hypertext Transfer Protocol (HTTP), or through message bus Kafka.

[0048] In some embodiments, after receiving raw data from different data sources, this application also needs to perform data verification based on the raw data. Specifically, the raw data is verified according to the verification rules corresponding to the data source. If the verification passes, the raw data is sent to the parsing plugin of the data processing module to execute the subsequent raw data parsing process. If the verification fails, the subsequent data parsing process will not be executed.

[0049] Step S202: For the original data corresponding to each data source, based on the parsing rules corresponding to the data source, the original data is parsed, and the reference data obtained after parsing is reorganized to obtain the target data corresponding to the original data. In some embodiments, after receiving raw data from different data sources, data parsing and data standardization are performed on the data. This can be achieved with the help of a parsing plugin.

[0050] In some embodiments, during the data parsing process, a pre-configured interface protocol sets parsing rules and conversion methods corresponding to different data sources. After receiving the raw data, a specific parsing plugin parses the data according to the format and semantics of the data source; the specific parsing rules can be set based on requirements. For example, for raw data in text format, the parsing plugin extracts and converts data based on specific delimiters or keywords; for hexadecimal byte stream data, the parsing plugin parses it byte by byte according to a predefined data structure.

[0051] In some embodiments, during the data standardization process, the parsed reference data is reorganized into target data according to a unified protocol format, and the processed target data is transformed into a preset structure, such as JSON or byte stream structure, to facilitate interaction between internal system plugins.

[0052] As an optional implementation, the above-described reorganization of the parsed reference data to obtain target data corresponding to the original data includes at least one of the following: The system filters reference data of preset types from various reference data sources and converts the format of the filtered reference data into the format corresponding to the target data. The preset data types can be set according to requirements, such as being set as the type of identifying dates, the type of identifying numbers, etc. The format of the target data corresponding to different preset types can be different. For example, the target format for dates can be "YYYY (year)-MM (month)-dd (day) HH (hour):mm (minute):ss.zzz (second)", and the target format for numbers can be adjusted to three decimal places. This operation can ensure that variables representing the same meaning in different data sources use the same name, and for the encoding of categorical variables, it ensures that the encoding meaning is consistent in different data sources.

[0053] The target data value corresponding to the preset time is determined based on each reference data; each reference data is mapped to the same spatial coordinate system; in this embodiment, the temporal or spatial differences of each original data can be unified during reorganization. Specifically, since the collection frequency of different data sources may differ, the time sequence of the original data from different data sources may not match. In practice, by using time interpolation or selecting data from similar times, the value corresponding to the preset time point (which can be set according to requirements) can be determined based on each reference data to reduce changes caused by the time span and to align the timestamps. For reference data with different spatial coordinates (such as geographic coordinates), they are mapped to the same spatial coordinate system (such as the Earth-Fixed Coordinate System), which can also be selected according to requirements.

[0054] The reference data is reorganized into target data with the same data format, that is, the data format is unified between modules and systems, which facilitates the transmission of data between different modules of the system. This unified format can be set according to requirements.

[0055] As an optional implementation, the target data that reassembles the reference data into the same data format includes: For each piece of raw data, obtain the second configuration information corresponding to the raw data; the second configuration information includes: data source information, data destination information, communication forwarding protocol, data type information, starting byte of the data field, and number of bytes occupied by the data field; the data type information includes: data type number and configuration rules corresponding to the data parameter number. Based on the parsed reference data and the second configuration information mentioned above, the frame header information and data field information corresponding to the target data are determined. The frame header information includes at least one of the following: data source information, data destination information, system time, and communication forwarding protocol. The data field information includes at least one of the following: data type number, data parameter number, and data value. Based on the aforementioned frame header information and data field information, the target data is constructed.

[0056] In some embodiments, this application pre-configures a data mapping table file (which can be stored in the configuration management module in JSON format). The second configuration information in the configuration file includes various parameters such as data source information, data type information, data destination information, communication forwarding information, data field start byte, data field occupancy bytes, and parsing method. The data field start byte and data field occupancy bytes can be used to define the starting position and length of data fields in the target data. For each piece of original data, the corresponding second configuration information can be obtained from the data mapping table, and the data can be parsed and reassembled according to the new format based on the second configuration information.

[0057] Optionally, the reconstructed target data mainly includes two parts: a frame header and a data field. The frame header information mainly includes: data source information, data destination information, system time, communication forwarding protocol, etc. The data source information, data destination information, system time, and communication forwarding protocol can be obtained from the second configuration information, and the system time can be obtained from the reference data. The data field information includes: data type number, data parameter number, data value, etc. The data type number and data parameter number can be constructed according to the configuration rules in the data type information of the second configuration file, and the data value can be obtained from the reference data.

[0058] In some embodiments, after obtaining the target data corresponding to the original data based on the frame header and data field, the target data can be encapsulated into a preset format as required, such as JSON string format and hexadecimal byte stream format, and the target data can be forwarded to the data forwarding module and the data storage and analysis module.

[0059] Step S203: Based on the target data corresponding to each original data, determine the business parameters to be fused and the semantic tags corresponding to each business parameter, and generate a business sequence. Step S204: Extract time-series feature vectors from the business sequence, correct the extracted time-series feature vectors based on the semantic rule base, and determine the status of the target device based on the corrected time-series feature vectors; the target device is the device corresponding to each data source.

[0060] In some embodiments, after parsing and reorganizing the original data to obtain the target data, the target data can be sent to the data storage and analysis module for storage and analysis and fusion.

[0061] In some embodiments, this application pre-configures a unified storage configuration file (the file can be in JSON format), which includes at least one of the following configuration information: data identifier, data source, storage method (database table or source code file), database type, database connection details, etc.; and, in this embodiment, customized development is supported according to different analysis needs, and a unified analysis configuration file (the file can be in JSON format) is provided, which may include at least one of the following configuration information: data identifier, data type, and analysis method, etc.

[0062] As an optional implementation, before determining the various service parameters to be merged and the semantic tags corresponding to each service parameter, the method further includes: Input the target data corresponding to each original data into the target model to generate a fact table and a dimension table. The fact table stores the core business indicators corresponding to each target data and the primary keys corresponding to each dimension table. The dimension table stores the semantic description information corresponding to the fact table. Store the aforementioned fact table and dimension table in the data warehouse.

[0063] In some embodiments, the target model is a "snowflake model," and a data warehouse can be used to store the target data to achieve a unified structure. Specifically, the target data (i.e., the target data corresponding to the original data from different sources) can be input into the "snowflake model." The "snowflake model" integrates the target data into structured fact tables and dimension tables. The fact tables and dimension tables are linked together through key connections to eliminate differences in heterogeneous data structures and achieve unified querying and analysis.

[0064] In practical implementation, the fact table primarily stores the primary keys of each connected dimension table, as well as quantifiable core business indicators, such as core parameters from satellite telemetry and gateway station data (e.g., attitude and orbit control parameters and energy parameters of satellite data, and received signal strength, data transmission speed, and processing latency of gateway stations). The dimension table primarily stores semantic supplements to the fact table, providing rich descriptive context (e.g., time, satellite, gateway station), and can include various types such as time dimension tables, satellite dimension tables, and gateway station dimension tables. Furthermore, based on parameters such as the type and manufacturer of satellites and gateway stations, the dimension table can be further divided into sub-dimension tables and associated with the main dimension table to achieve the association of heterogeneous data, improve query efficiency, and reduce data redundancy.

[0065] In this embodiment of the application, based on the unified storage data, specific functions of analysis and prediction can be completed. For example, based on the combination of recurrent neural networks and semantic reasoning, the stored data can be analyzed and fused to complete the monitoring of the status of satellites and gateway stations.

[0066] As an optional implementation, the above-mentioned determination of each service parameter to be merged and the semantic tags corresponding to each service parameter, and the generation of a service sequence, includes: Extract the business parameters of the target data type from the fact table and dimension table above; Based on the semantic description information stored in the above dimension table, determine the semantic tags corresponding to each of the above business parameters. Each business parameter is combined with its corresponding semantic tag, and the combined result is sorted based on the time information corresponding to each business parameter to obtain the business sequence.

[0067] In some embodiments, when analyzing and fusing raw data, this application needs to pre-configure the target data type or conditions to be met for the business parameters to be fused, and extract the business parameters to be fused (such as telemetry parameters) from the fact table and dimension table in the data warehouse (satellite, gateway station). Furthermore, after extracting each business parameter, it is also necessary to define a unique semantic label for each business parameter, that is, to convert the numerical value into a semantic symbol that the model can recognize, and to generate dynamic labels (including numerical labels, status labels, and association labels) by matching semantic attributes based on real-time numerical values. Specifically, the semantic labels can be obtained from the aforementioned dimension table. The fact table and dimension table are related through external connections, and the semantic labels of the corresponding parameters are obtained through multi-level database queries.

[0068] After obtaining the semantic tags corresponding to each business parameter, each business parameter can be combined with the semantic tag and encoded to generate a corresponding vector. Then, the vectors corresponding to multiple business parameters can be combined into a business sequence in chronological order.

[0069] As an optional implementation, the above-mentioned extraction of temporal feature vectors from the aforementioned service sequence, correction of the extracted temporal feature vectors based on a semantic rule base, and determination of the target device's state based on the corrected temporal feature vectors include: Input the above business sequence into a pre-trained recurrent neural network model; Based on the Long Short-Term Memory (LSTM) layer in the above recurrent neural network model, the time-series features of the above business sequence are extracted to obtain the corresponding time-series feature vector; Based on the above semantic rule base, logical reasoning is performed on the above time-series feature vectors to obtain the corrected time-series feature vectors; Based on the output layer of the above recurrent neural network model, an output result carrying semantic labels is generated; the above output result is used to characterize the state of the above target device.

[0070] In some embodiments, this application performs a fusion analysis of the above-mentioned service sequences by combining recurrent neural networks (including LSTM units) with semantic reasoning to determine the status of each target device, such as satellites and gateways.

[0071] In some embodiments, the recurrent neural network model is pre-trained based on sample data carrying output labels to ensure the accuracy of the model output. Furthermore, based on the abnormal correction results of the model output, new semantic rules are constructed periodically, and the value semantic rule library is updated as a constraint to fine-tune the loss function in the recurrent neural network, thereby improving the ability to judge complex states. The abnormal correction results refer to the information used to correct the model output when the state of the target device is determined manually and the determination result does not match the model output result.

[0072] In some embodiments, after pre-training the Recurrent Neural Network (RNN), business training is used as the input to the model. Based on the LSTM layer (i.e., hidden layer) of the RNN, the temporal features corresponding to the business sequence (such as satellite and gateway parameters) are extracted to obtain a temporal feature vector. The semantic associations of the parameters are learned through the introduced semantic constraints, i.e., the domain rules embedded in the loss function. Specifically, in the feature extraction process, for each intermediate vector (first vector) obtained, logical reasoning is performed on the first vector based on a preset rule base to obtain a second vector with semantic constraints based on the rules in the rule base. If there is a conflict between the first vector and the second vector, the vector with higher confidence is adopted first based on the weights corresponding to the first vector and the confidence priority principle in the model. This vector is then used as the input to the next layer for the next round of feature extraction until the feature vector is obtained.

[0073] In some embodiments, the confidence level of the first vector can be determined based on the maximum probability value or entropy value in the probability distribution generated by the RNN in the output layer, and the confidence level of the second vector can be determined based on the preset weight or historical accuracy of the semantic rule that derives the second vector in the semantic rule base.

[0074] Then, based on the output layer of the RNN, the corresponding state probability distribution is generated through the softmax function, and the state probability distribution is mapped to a structured output result with semantic labels. The output result includes multiple parameters used to characterize the state of the target device and the specific values ​​of the multiple parameters. In this embodiment, the operating state of the target device, such as normal, abnormal or faulty, can be further deduced based on the values ​​of the multiple parameters.

[0075] The above method can effectively integrate multi-source heterogeneous data, compare the actual data with the output of the digital twin model, calibrate and verify the model, improve the accuracy and reliability of the model, and enable the digital twin system to highly accurately reflect the actual state and behavior of the physical system, providing a reliable basis for decision-making. Specifically, it offers high flexibility and scalability. Through plug-in technology and the introduction of a unified standard interface protocol, it achieves high efficiency and accuracy in data reception, parsing, and forwarding, improving the efficiency and flexibility of data processing. It can be customized and expanded to meet different application needs, adapting to various complex scenarios and constantly changing business requirements.

[0076] In some embodiments, after obtaining the target data, this application can also configure the target data based on the configuration file corresponding to the pre-configured forwarding module and then forward it.

[0077] Specifically, the interface protocol of the forwarding plugin of the data forwarding module specifies unified forwarding configuration information, including at least one of the following: forwarding data identifier, output format, target address, protocol type, communication details, forwarding frequency, forwarding parameter content, priority, etc. After obtaining the target data, the forwarding plugin can forward the standardized data to the system subscribing to the target data according to the preset configuration. The forwarding data identifier is a unique identifier for internal and external systems, and the forwarding parameter content can be flexibly configured according to the needs of the target party subscribing to the target information to reduce redundant information. Both internal and external systems involve two data formats: JSON and hexadecimal byte streams. In the JSON format, each set of parameters mainly includes the parameter name, parameter type, and parameter value; in the hexadecimal format, each set of parameters mainly includes the parameter identifier and parameter value.

[0078] In practice, the forwarding communication protocol is implemented through a plugin: for external system business data, the forwarding plugin can send data to the external system via UDP, TCP, etc., or push data to the topic via message bus (Kafka, DDS, etc.); for internal system simulation data, communication is mainly carried out via message bus Kafka and UDP.

[0079] In some embodiments, this application (through the display plugin of the situation display module) supports subscribing to data from the data forwarding module. When the subscribed data is received, the data can be added to the data cache queue. According to the real-time display needs, the data in the cache queue is read in a multi-threaded manner, and the corresponding data bits are parsed and displayed.

[0080] This application supports the flexible loading and parsing of page files (in JSON format) for display plugins (including plugins for curves, tables, 2D / 3D visualization, etc.). These page files may include any one of the following configuration information: page layout, plugin display style, bound theme data, and plugin name. Through this configuration information, display plugins can be associated with business theme data, meeting users' data display needs and serving different business systems.

[0081] The multi-source data fusion method described in this application can effectively integrate low-orbit satellite multi-source data, such as satellite telemetry data, ground test data, log document data, and semi-physical equipment test data, which differ in format, frequency, and accuracy, through a unified data interface and data conversion method. This breaks down data silos and enables collaborative utilization of data.

[0082] For diverse types of low-Earth orbit satellite data, such as structured data (e.g., semi-physical measurement values), semi-structured data (e.g., flight parameter configuration files in XML format), and unstructured data (e.g., operation logs in text format), appropriate heterogeneous data processing technologies, such as data aggregation, data cleaning, data transformation, and data analysis and storage, can be adopted to enable different types of data to be interconnected and comprehensively utilized in the digital twin system.

[0083] Based on the same disclosed concept, this application also provides a multi-source heterogeneous data fusion device based on a satellite digital twin system. Since this device is the same as the device in the method of this application, and the principle of the device in solving the problem is similar to that of the method, the implementation of the device can refer to the implementation of the method, and the repeated parts will not be described again.

[0084] Figure 3 A schematic diagram of a multi-source heterogeneous data fusion device based on a satellite digital twin system provided in this application embodiment is shown below. Figure 3 This application provides a multi-source heterogeneous data fusion device based on a satellite digital twin system, the device comprising: The data receiving module 301 is used to receive raw data from different data sources; The data recombination module 302 is used to parse the original data corresponding to each data source based on the parsing rules corresponding to the data source, and to reconstruct the reference data obtained after parsing to obtain the target data corresponding to the original data. The sequence generation module 303 is used to determine the business parameters to be fused and the semantic tags corresponding to each business parameter based on the target data corresponding to each original data, and generate a business sequence. The status determination module 304 is used to extract time-series feature vectors from the above business sequence, correct the extracted time-series feature vectors based on the semantic rule base, and determine the status of the target device based on the corrected time-series feature vectors; the above target devices are the devices corresponding to each data source.

[0085] As an optional implementation, the data receiving module 301 described above is specifically used for: Create sub-threads corresponding to different data sources, wherein each sub-thread is configured with first configuration information corresponding to the corresponding data source, and the first configuration information includes at least one of communication protocol type, data transmission format and communication parameters; Each sub-thread receives raw data from the data source corresponding to the aforementioned sub-thread, based on the first configuration information corresponding to that sub-thread.

[0086] As an optional implementation, the data reconstruction module 302 reconstructs the parsed reference data based on at least one of the following methods: Filter reference data of a preset type from various reference data, and convert the format of the filtered reference data into the format corresponding to the target data; Determine the value of the target data corresponding to the preset time based on the reference data; Map all reference data to the same spatial coordinate system; Reorganize the reference data into target data with the same data format.

[0087] As an optional implementation, the data reconstruction module 302 described above is specifically used for: For each piece of raw data, obtain the second configuration information corresponding to the raw data; the second configuration information includes: data source information, data destination information, communication forwarding protocol, data type information, starting byte of the data field, and number of bytes occupied by the data field; the data type information includes: data type number and configuration rules corresponding to the data parameter number. Based on the parsed reference data and the second configuration information mentioned above, the frame header information and data field information corresponding to the target data are determined. The frame header information includes at least one of the following: data source information, data destination information, system time, and communication forwarding protocol. The data field information includes at least one of the following: data type number, data parameter number, and data value. Based on the aforementioned frame header information and data field information, the target data is constructed.

[0088] As an optional implementation, before the sequence generation module 303 determines the service parameters to be merged and the semantic tags corresponding to each service parameter, it is further used to: Input the target data corresponding to each original data into the target model to generate a fact table and a dimension table. The fact table stores the core business indicators corresponding to each target data and the primary keys corresponding to each dimension table. The dimension table stores the semantic description information corresponding to the fact table. Store the aforementioned fact table and dimension table in the data warehouse.

[0089] As an optional implementation, the sequence generation module 303 described above is specifically used for: Extract the business parameters of the target data type from the fact table and dimension table above; Based on the semantic description information stored in the above dimension table, determine the semantic tags corresponding to each of the above business parameters. Each business parameter is combined with its corresponding semantic tag, and the combined result is sorted based on the time information corresponding to each business parameter to obtain the business sequence.

[0090] As an optional implementation, the sequence generation module 303 described above is specifically used for: Input the above business sequence into a pre-trained recurrent neural network model; Based on the long short-term memory network layer in the above recurrent neural network model, the time-series features of the above business sequence are extracted to obtain the corresponding time-series characteristic vector. Based on the above semantic rule base, logical reasoning is performed on the above time-series feature vectors to obtain the corrected time-series feature vectors; Based on the output layer of the above recurrent neural network model, an output result carrying semantic labels is generated; the above output result is used to characterize the state of the above target device.

[0091] Based on the same disclosed concept, this application also provides a multi-source heterogeneous data fusion device based on a satellite digital twin system. Since this device is the same device in the method of this application, and the principle of solving the problem by this device is similar to that of this method, the implementation of this device can refer to the implementation of the method, and the repeated parts will not be described again.

[0092] Those skilled in the art will understand that various aspects of this application can be implemented as a system, method, or program product. Therefore, various aspects of this application can be specifically implemented in the following forms: a completely hardware implementation, a completely software implementation (including firmware, microcode, etc.), or a combination of hardware and software implementations, collectively referred to herein as a "circuit," "module," or "system."

[0093] In some possible implementations, the device according to this application may include at least one processor and at least one memory. The memory stores program code that, when executed by the processor, causes the processor to perform the steps in the multi-source heterogeneous data fusion method based on a satellite digital twin system according to various exemplary embodiments of this application described above.

[0094] The following reference Figure 4 To describe the device 400 according to this embodiment of the present application. Figure 4 The device 400 shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0095] like Figure 4 As shown, device 400 is presented in the form of a general-purpose device. The components of device 400 may include, but are not limited to: at least one processor 401, at least one memory 402, and a bus 403 connecting different system components (including memory 402 and processor 401). The memory stores program code, which, when executed by the processor, causes the processor to perform the steps of the multi-source heterogeneous data fusion method based on the satellite digital twin system.

[0096] Bus 403 represents one or more of several bus structures, including a memory bus or memory controller, peripheral bus, processor, or a local bus using any of the various bus structures.

[0097] The memory 402 may include a readable medium in the form of volatile memory, such as random access memory (RAM) 4021 and / or cache memory 4022, and may further include read-only memory (ROM) 4023.

[0098] The memory 402 may also include a program / utility 4025 having a set (at least one) of program modules 4024, including but not limited to: an operating system, one or more application programs, other program modules, and program data, each or some combination of these examples may include an implementation of a network environment.

[0099] Device 400 can also communicate with one or more external devices 404 (e.g., keyboard, pointing device, etc.), and with one or more devices that enable a user to interact with device 400, and / or with any device that enables device 400 to communicate with one or more other devices (e.g., router, modem, etc.). This communication can be performed via input / output (I / O) interface 405. Furthermore, device 400 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) via network adapter 406. As shown, network adapter 406 communicates with other modules used with device 400 via bus 403. It should be understood that, although not shown in the figures, other hardware and / or software modules can be used in conjunction with device 400, including but not limited to: microcode, device drivers, redundant processors, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0100] In some possible implementations, various aspects of the multi-source heterogeneous data fusion method based on a satellite digital twin system provided in this application can also be implemented in the form of a program product, which includes program code. When the program product is run on a computer device, the program code is used to cause the computer device to perform the steps in the multi-source heterogeneous data fusion method based on a satellite digital twin system according to various exemplary embodiments of this application described above.

[0101] The program product may employ any combination of one or more readable media. A readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0102] The monitoring program product of the embodiments of this application can be a portable compact disc read-only memory (CD-ROM) and include program code, and can run on a device. However, the program product of this application is not limited to this. In this document, the readable storage medium can be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0103] A readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying readable program code. This propagated data signal may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A readable signal medium may also be any readable medium other than a readable storage medium, capable of sending, propagating, or transmitting a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0104] The program code contained on the readable medium may be transmitted using any suitable medium, including but not limited to wireless, wired, optical fiber, RF, etc., or any suitable combination thereof.

[0105] Program code for performing the operations of this application can be written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Java and C++, and conventional procedural programming languages ​​such as C or similar languages. The program code can execute entirely on the user device, partially on the user device, as a standalone software package, partially on the user device and partially on a remote device, or entirely on a remote device or server. In cases involving remote devices, the remote device can be connected to the user device via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external device (e.g., via the Internet using an Internet service provider).

[0106] It should be noted that although several units or sub-units of the device have been mentioned in the detailed description above, this division is merely exemplary and not mandatory. In fact, according to embodiments of this application, the features and functions of two or more units described above can be embodied in one unit. Conversely, the features and functions of one unit described above can be further divided and embodied by multiple units.

[0107] Furthermore, although the operations of the method of this application are described in a specific order in the accompanying drawings, this does not require or imply that these operations must be performed in that specific order, or that all the operations shown must be performed to achieve the desired result. Additionally or alternatively, certain steps may be omitted, multiple steps may be combined into one step, and / or one step may be broken down into multiple steps.

[0108] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0109] This application is described with reference to flowchart illustrations and block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block and / or segment of the flowchart illustrations and block diagrams, as well as combinations of blocks and segments in the flowchart illustrations and block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart. Figure 1 One or more processes and boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0110] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and boxes Figure 1 The function specified in one or more boxes.

[0111] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and boxes Figure 1 The steps of the function specified in one or more boxes.

[0112] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0113] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for fusing multi-source heterogeneous data based on a satellite digital twin system, characterized in that, The method includes: Receive raw data from different data sources; For the original data corresponding to each data source, the original data is parsed based on the parsing rules corresponding to the data source, and the reference data obtained after parsing is reorganized to obtain the target data corresponding to the original data; Based on the target data corresponding to each original data, determine the business parameters to be integrated and the semantic tags corresponding to each business parameter, and generate a business sequence. Temporal feature vectors are extracted from the business sequence, the extracted temporal feature vectors are corrected based on the semantic rule base, and the state of the target device is determined based on the corrected temporal feature vectors; the target device is the device corresponding to each data source.

2. The method according to claim 1, characterized in that, The receiving of raw data from different data sources includes: Create sub-threads corresponding to different data sources, wherein each sub-thread is configured with first configuration information corresponding to the corresponding data source, and the first configuration information includes at least one of communication protocol type, data transmission format and communication parameters; Each sub-thread receives raw data from the data source corresponding to it, based on the first configuration information corresponding to the sub-thread.

3. The method according to claim 1, characterized in that, The reorganization of the parsed reference data to obtain target data corresponding to the original data includes at least one of the following: Filter reference data of a preset type from various reference data, and convert the format of the filtered reference data into the format corresponding to the target data; Determine the value of the target data corresponding to the preset time based on the reference data; Map all reference data to the same spatial coordinate system; Reorganize the reference data into target data with the same data format.

4. The method according to claim 3, characterized in that, The process of recombining the reference data into target data of the same format includes: For each piece of raw data, obtain the second configuration information corresponding to the raw data; the second configuration information includes: data source information, data destination information, communication forwarding protocol, data type information, starting byte corresponding to the data field, and number of bytes occupied by the data field; the data type information includes: data type number and configuration rules corresponding to the data parameter number; Based on the parsed reference data and the second configuration information, the frame header information and data field information corresponding to the target data are determined. The frame header information includes at least one of the following: data source information, data destination information, system time, and communication forwarding protocol. The data field information includes at least one of the following: data type number, data parameter number, and data value. The target data is constructed based on the frame header information and data field information.

5. The method according to claim 1, characterized in that, Before determining the business parameters to be merged and the semantic tags corresponding to each business parameter, the method further includes: Input the target data corresponding to each original data into the target model to generate a fact table and a dimension table; the fact table stores the core business indicators corresponding to each target data and the primary keys corresponding to each dimension table, and the dimension table stores the semantic description information corresponding to the fact table; The fact table and dimension table are stored in the data warehouse.

6. The method according to claim 5, characterized in that, The process of determining the business parameters to be merged and the semantic tags corresponding to each business parameter, and generating a business sequence, includes: Extract the business parameters of the target data type from the fact table and dimension table; Based on the semantic description information stored in the dimension table, determine the semantic tags corresponding to each business parameter; Each business parameter is combined with its corresponding semantic tag, and the combined result is sorted based on the time information corresponding to each business parameter to obtain the business sequence.

7. The method according to any one of claims 1 to 6, characterized in that, The step of extracting a temporal feature vector from the service sequence, correcting the extracted temporal feature vector based on a semantic rule base, and determining the state of the target device based on the corrected temporal feature vector includes: The business sequence is input into a pre-trained recurrent neural network model; Based on the long short-term memory network layer in the recurrent neural network model, the time-series features of the business sequence are extracted to obtain the corresponding time-series characteristic vector; Logical reasoning is performed on the temporal feature vector based on the semantic rule base to obtain the corrected temporal feature vector; Based on the output layer of the recurrent neural network model, an output result carrying semantic labels is generated; the output result is used to characterize the state of the target device.

8. A multi-source heterogeneous data fusion device based on a satellite digital twin system, characterized in that, The device includes: The data receiving module is used to receive raw data from different data sources; The data recombination module is used to parse the original data corresponding to each data source based on the parsing rules corresponding to the data source, and to reconstruct the reference data obtained after parsing to obtain the target data corresponding to the original data. The sequence generation module is used to determine the business parameters to be fused and the semantic tags corresponding to each business parameter based on the target data corresponding to each original data, and to generate a business sequence. The state determination module is used to extract time-series feature vectors from the service sequence, correct the extracted time-series feature vectors based on a semantic rule base, and determine the state of the target device based on the corrected time-series feature vectors; the target device is the device corresponding to each data source.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor, when executing a computer program stored in the memory, implements the method of any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the method described in any one of claims 1 to 7.

Images