Construction method and device of data asset management platform, equipment and medium

Abstract

The invention discloses a construction method and device of a data asset management platform, equipment and a medium. The method comprises the steps of obtaining power plant data of multiple power plants; the power plant data comprises financial data and production data; the financial data comprises a plurality of financial indexes, and the production data comprises a plurality of production indexes. And respectively obtaining semantic information and time sequence information corresponding to the financial indexes and the production indexes, carrying out semantic alignment on the financial indexes and the production indexes according to the semantic information, and carrying out time sequence alignment on the financial indexes and the production indexes according to the time sequence information. And generating federal learning models corresponding to the plurality of power plants according to the aligned financial data and production data. And performing association analysis on the financial data and the production data according to the federated learning model, and analyzing a corresponding analysis result according to the association analysis. And constructing a cross-domain ontology knowledge graph according to the aligned financial data and production data, and constructing a data asset management platform according to the cross-domain ontology knowledge graph and an analysis result.

Description

Technical Field

[0001] This application relates to the technical field of asset management, and particularly to a method, device, equipment and medium for constructing a data asset management platform. Background Art

[0002] In the process of the digital transformation of the energy industry, the asset management platforms of energy enterprises face multiple technical bottlenecks, seriously restricting the improvement of enterprise operation efficiency. The current asset management platforms applied to energy enterprises have the following technical problems: The traditional asset management platform adopts a vertical domain construction mode, and there is a structural gap between the financial management system and the production monitoring system. At the same time, there is a lack of unified semantic mapping between financial indicators (such as marketing revenue / cost) and production indicators (such as power generation / equipment efficiency). The monthly granularity of financial statements cannot be dynamically aligned with the second-level data of the SCADA system, resulting in the inability of power plants under the enterprise to accurately calculate the unit power generation cost during peak and valley electricity price periods.

[0003] Therefore, there is an urgent need for a method to solve at least one of the above problems. Summary of the Invention

[0004] This application provides a method, device, equipment and medium for constructing a data asset management platform, aiming to solve the problem of data dimension fragmentation in the current asset management platforms applied to energy enterprises.

[0005] In a first aspect, this application provides a method for constructing a data asset management platform, and the method includes:

[0006] Obtain power plant data of multiple power plants; the power plant data includes financial data and production data; the financial data includes multiple financial indicators, and the production data includes multiple production indicators;

[0007] Obtain the semantic information and time series information corresponding to the financial indicators and production indicators respectively, perform semantic alignment on the financial indicators and production indicators according to the semantic information, and perform time series alignment on the financial indicators and production indicators according to the time series information;

[0008] Generate a federated learning model corresponding to each of the multiple power plants according to the aligned financial data and production data;

[0009] Perform correlation analysis on the financial data and production data in the federated learning model, and obtain the analysis result corresponding to the correlation analysis;

[0010] Construct a cross-domain ontology knowledge graph according to the aligned financial data and production data, and construct a data asset management platform according to the cross-domain ontology knowledge graph and the analysis result.

[0011] In some embodiments, generating a federated learning model corresponding to multiple power plants based on the aligned financial data and production data includes: performing standardization processing on the financial data and production data; the standardization processing at least includes missing value filling, outlier processing, and data normalization; obtaining a basic model corresponding to each power plant; inputting the power plant data corresponding to each power plant into the corresponding basic model; performing distributed training on multiple basic models according to a preset federated learning framework to obtain model parameters corresponding to each basic model; and aggregating the multiple model parameters according to the federated averaging method to generate the federated learning model.

[0012] Exemplarily, obtaining a basic model corresponding to each power plant includes: obtaining power plant feature information of each power plant; the power plant feature information at least includes power plant type and power plant location; and confirming the basic model corresponding to the power plant according to the power plant feature information.

[0013] In some embodiments, performing correlation analysis on the financial data and production data in the federated learning model to obtain an analysis result corresponding to the correlation analysis includes: extracting financial data features and production data features in the federated learning model; obtaining significant features from the financial data features and production data features according to a preset feature selection algorithm; obtaining correlation rules corresponding to the financial data features and production data features according to a preset association rule mining algorithm; obtaining causal information of the financial data features and production data features according to a preset causal analysis algorithm; and generating an analysis result corresponding to the correlation analysis according to the significant features, correlation rules, and causal information.

[0014] In some embodiments, constructing a cross-domain ontology knowledge graph based on the aligned financial data and production data includes: obtaining attribute information of the aligned financial data and production data, and constructing an initial knowledge graph according to the attribute information; performing knowledge extraction on multiple pieces of the financial data and production data based on natural language processing technology to obtain financial data knowledge and production data knowledge; obtaining potential association information corresponding to the financial data and production data according to a preset knowledge reasoning algorithm; and adding the financial data knowledge, production data knowledge, and potential association information to the initial knowledge graph to obtain the cross-domain ontology knowledge graph.

[0015] In some embodiments, constructing a data asset management platform according to the cross-domain ontology knowledge graph and the analysis result includes: obtaining the architecture of the data asset management platform, where the architecture includes a data access layer, a data processing layer, a knowledge graph layer, and an application layer; setting the cross-domain ontology knowledge graph in the knowledge graph layer and the analysis result in the application layer to complete the construction of the data asset management platform.

[0016] In some embodiments, after constructing the data asset management platform according to the cross-domain ontology knowledge graph and the analysis result, the method further includes: calculating a global risk coefficient corresponding to each power plant according to the cross-domain ontology knowledge graph of the data asset management platform and the analysis result; and if it is determined that one of the power plants is abnormal according to the global risk coefficient, generating abnormal warning information corresponding to the power plant.

[0017] In a second aspect, the present application provides a device for constructing a data asset management platform, including:

[0018] A data acquisition unit, configured to acquire power plant data of multiple power plants; the power plant data includes financial data and production data; the financial data includes multiple financial indicators, and the production data includes multiple production indicators;

[0019] An index acquisition unit, configured to respectively acquire semantic information and time series information corresponding to the financial indicators and production indicators, perform semantic alignment on the financial indicators and production indicators according to the semantic information, and perform time series alignment on the financial indicators and production indicators according to the time series information;

[0020] A model generation unit, configured to generate a federated learning model corresponding to multiple power plants according to the aligned financial data and production data;

[0021] An association analysis unit, configured to perform association analysis on the financial data and production data according to the federated learning model, and obtain an analysis result corresponding to the association analysis;

[0022] A knowledge graph construction unit, configured to construct a cross-domain ontology knowledge graph according to the aligned financial data and production data, and construct a data asset management platform according to the cross-domain ontology knowledge graph and the analysis result.

[0023] In a third aspect, the present application provides a computer device, including a memory and a processor; the memory is used to store a computer program; the processor is configured to execute the computer program and implement the method provided in any embodiment of the present application when executing the computer program.

[0024] In a fourth aspect, the present application provides a computer-readable storage medium, where the computer-readable storage medium stores a computer program, and when the computer-readable instruction is executed by a processor, one or more processors are enabled to execute the method provided in any embodiment of the present application.

[0025] The present application discloses a method, device, equipment and medium for constructing a data asset management platform. The method aims to solve the problem of data dimension fragmentation in energy enterprises. The provided method includes the following steps:

[0026] Power plant data acquisition: Financial data and production data are obtained from multiple power plants. The financial data includes multiple financial indicators (such as cost, revenue, profit, etc.), and the production data includes multiple production indicators (such as power generation, equipment utilization rate, energy consumption, etc.).

[0027] Data alignment: Semantic alignment: Obtain the semantic information corresponding to the financial indicators and production indicators, and perform semantic alignment on the financial indicators and production indicators according to this semantic information. The purpose of semantic alignment is to ensure that the same or similar indicators in different data sources are semantically consistent. Temporal alignment: Obtain the temporal information corresponding to the financial indicators and production indicators, and perform temporal alignment on the financial indicators and production indicators according to this temporal information. The purpose of temporal alignment is to ensure that the indicators in different data sources are consistent in the time dimension.

[0028] Federated learning model generation: Generate federated learning models corresponding to multiple power plants based on the aligned financial data and production data. Federated learning is a distributed machine learning method that can use multiple data sources for model training without sharing the original data.

[0029] Association analysis: Perform association analysis on the financial data and production data in the federated learning model to obtain the analysis results. The purpose of association analysis is to reveal the potential relationships between the financial data and production data, such as the relationship between cost and power generation, and the relationship between equipment utilization rate and profit, etc.

[0030] Cross-domain ontology knowledge graph construction: Construct a cross-domain ontology knowledge graph based on the aligned financial data and production data. The cross-domain ontology knowledge graph is a semantic network used to represent the concepts and their relationships in different domains (such as finance and production).

[0031] Data asset management platform construction: Construct a data asset management platform based on the cross-domain ontology knowledge graph and the association analysis results. This platform can integrate and display the data of different power plants, provide data query, analysis and visualization functions, and help enterprises better manage data assets.

[0032] The method provided has at least the following beneficial effects:

[0033] Solve the problem of data dimension fragmentation: Through semantic alignment and temporal alignment, ensure that the financial data and production data in different data sources are consistent in the semantic and time dimensions, and solve the problem of data dimension fragmentation.

[0034] Improve data utilization efficiency: Through the federated learning model, data from multiple power plants can be used for model training without sharing the original data, thus improving data utilization efficiency.

[0035] Reveal potential relationships between data: Through association analysis, reveal the potential relationships between financial data and production data, helping enterprises better understand the business logic behind the data.

[0036] Construct a cross-domain knowledge graph: By constructing a cross-domain ontology knowledge graph, integrate data from different domains, provide a more comprehensive data view, and help enterprises conduct cross-domain data analysis and decision-making.

[0037] Improve data management capabilities: By constructing a data asset management platform, integrate and display data from different power plants, provide data query, analysis, and visualization functions, and improve the data management capabilities of enterprises.

[0038] In summary, through data alignment, federated learning, association analysis, and cross-domain knowledge graph construction, this method solves the problem of data dimension fragmentation in energy enterprises, improves data utilization efficiency, reveals potential relationships between data, and enhances the data management capabilities of enterprises.

[0039] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit this application. Brief Description of the Drawings

[0040] To more clearly illustrate the technical solutions of the embodiments of this application, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are some embodiments of this application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0041] Figure 1 It is a schematic flowchart of the steps of the method for constructing a data asset management platform provided by an embodiment of this application;

[0042] Figure 2 It is a schematic diagram of the interface of the data asset management platform provided by an embodiment of this application;

[0043] Figure 3 It is a schematic structural diagram of the device for constructing a data asset management platform provided by an embodiment of this application;

[0044] Figure 4 It is a schematic block diagram of the structure of a computer device provided by an embodiment of this application.

[0045] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit this application. Detailed Embodiments

[0046] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application without creative efforts shall fall within the protection scope of the present application.

[0047] The flowchart shown in the accompanying drawings is only an example illustration, and does not necessarily include all the content and operations / steps, nor does it necessarily need to be executed in the described order. For example, some operations / steps can also be decomposed, combined, or partially merged, so the actual execution order may change according to the actual situation.

[0048] It should be understood that, in order to facilitate the clear description of the technical solutions in the embodiments of the present invention, in the embodiments of the present invention, terms such as "first" and "second" are used to distinguish the same items or similar items with basically the same functions and effects. Those skilled in the art can understand that the terms "first", "second", etc. do not limit the quantity and execution order, and the terms "first", "second", etc. do not necessarily mean different.

[0049] It should be understood that the terms used in the specification of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the specification of the present application and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an", and "the" are intended to include the plural forms.

[0050] It should also be understood that the term " / and" used in the specification of the present application and the appended claims refers to any combination and all possible combinations of one or more of the related listed items, and includes these combinations.

[0051] Next, some embodiments of the present application will be described in detail in conjunction with the accompanying drawings. Without conflict, the following embodiments and the features in the embodiments can be combined with each other.

[0052] In the process of the digital transformation of the energy industry, the asset management system faces multiple technical bottlenecks, seriously restricting the improvement of enterprise operation efficiency. The current typical technical architecture has the following key defects:

[0053] Data Dimension Fragmentation: Traditional systems adopt a vertical domain construction model. There is a structural gap between the financial management system (such as SAP ERP) and the production monitoring system (such as OSIsoft PI): Ontological Gap: There is a lack of unified semantic mapping between financial indicators (marketing revenue / cost) and production indicators (power generation / equipment efficiency). Statistics from a certain power generation group show that 35% of the business analysis errors stem from data caliber deviations. Temporal Mismatch: The monthly granularity of financial statements cannot be dynamically aligned with the second-level data of the SCADA system, resulting in coal-fired power plants being unable to accurately calculate the unit power generation cost during peak and valley electricity price periods. Value Island: Industry research in 2023 shows that 78% of energy enterprises are unable to establish a quantitative relationship model between marketing expense investment and unit output efficiency of generating units.

[0054] Meanwhile, the existing technology stack is difficult to accommodate the requirements of multi-modal data processing: Financial Processing Delay: The batch processing architecture based on the T+1 cycle leads to the disconnection between marketing rebate calculation and the real-time power trading market. A certain new energy enterprise once suffered a settlement loss of 12 million yuan per year due to this. Production Response Lag: Under the traditional data warehouse architecture, the equipment fault warning delay reaches 15 - 30 minutes, unable to meet the 90-second response standard required by the "Technical Specification for Real-time Monitoring of Power Systems". Lack of Hybrid Computing: The correlation analysis between discrete events (financial write-off) and continuous signals (unit vibration) lacks the support of a mathematical framework.

[0055] Therefore, there is an urgent need for a method, device, equipment, and medium to solve at least one of the above problems.

[0056] To solve the above problems, please refer to Figure 1 , Figure 1 is a schematic flowchart of a method for constructing a data asset management platform provided in an embodiment of the present application. The execution device of the method is a construction device applied to the data asset management platform. The device includes a computer device, and also includes an endoscope module and an electromagnetic tracking module. The front end of the endoscope module is used to extend into the nasal cavity of the fresh cadaver head to be measured, and the electromagnetic tracking module is arranged at the front end of the endoscope module for measuring the magnetic field information of the nasal cavity.

[0057] To solve the above problems, please refer to Figure 1 。Specifically, as Figure 1 shown, the provided method includes steps S101 to S105. Among them, the computer device of the provided device can be a handheld terminal, a laptop computer, a wearable device, or a robot, etc., for implementing steps S101 to S105 and their corresponding embodiments.

[0058] The steps are described in detail as follows:

[0059] S101. Obtain power plant data of multiple power plants; the power plant data includes financial data and production data; the financial data includes multiple financial indicators, and the production data includes multiple production indicators.

[0060] Specifically, the core of this step is to extract financial data and production data from heterogeneous systems of multiple power plants, covering financial indicators (such as marketing revenue, cost, profit) and production indicators (such as power generation, equipment efficiency, failure rate). Through multi-source data integration, a unified data foundation is constructed.

[0061] Data source access: Extract data from the financial management system (such as SAP ERP, Oracle Financials) and production monitoring system (such as OSIsoft PI, SCADA) through tools such as ETL (Extract, Transform, Load) or API interfaces. Design adapter modules for different data sources (such as relational databases, time series databases, file systems) to ensure efficient data access.

[0062] At the same time, through data cleaning and standardization, including data deduplication, missing value filling, outlier handling and other cleaning operations. Ensure the consistency of data formats and units. Store the cleaned data in a distributed data warehouse (such as Hadoop, Snowflake) or a data lake to support the storage and fast query of massive data. Adopt data partitioning and time series indexing technologies to optimize data retrieval efficiency.

[0063] This step eliminates data islands through multi-source data integration and standardization, providing a high-quality data foundation for subsequent semantic alignment and time series alignment. The distributed storage architecture supports the processing of massive data, meeting the high concurrency and real-time requirements of power plant data.

[0064] S102. Obtain the semantic information and time series information corresponding to the financial indicators and production indicators respectively, perform semantic alignment on the financial indicators and production indicators according to the semantic information, and perform time series alignment on the financial indicators and production indicators according to the time series information.

[0065] Specifically, this step aims to solve the ontology gap and time series misalignment problems between financial data and production data, and achieve deep integration of data through semantic alignment and time series alignment.

[0066] Semantic information extraction and mapping extract key terms and concepts from financial metrics and production metrics by using natural language processing (NLP) techniques such as BERT and Word2Vec. Based on the Ontology method, a cross-domain semantic mapping model is constructed to associate financial terms (such as "marketing revenue") with production terms (such as "power generation"). The semantic relationships are represented using OWL (Web Ontology Language) or RDF (Resource Description Framework) to form a unified semantic knowledge base.

[0067] For the monthly granularity of financial data and the second-level granularity of production data, time series interpolation techniques (such as linear interpolation and spline interpolation) are used to achieve time alignment of the data. The sliding window algorithm is used to aggregate high-frequency production data into a time granularity that matches the financial data. A timestamp synchronization mechanism is introduced to ensure the consistency of financial data and production data on the time axis.

[0068] The ontological gap between financial metrics and production metrics is eliminated through semantic alignment, providing a unified semantic basis for subsequent correlation analysis. The time series alignment solves the problem of inconsistent data granularity, supports the dynamic matching of financial data and production data, and improves the accuracy of analysis.

[0069] S103. Generate a federated learning model corresponding to each of the power plants based on the aligned financial data and production data.

[0070] Specifically, in this step, the aligned financial data and production data are used to construct a federated learning model for multiple power plants to achieve collaborative modeling under data privacy protection.

[0071] The federated learning framework is built using an open-source federated learning framework (such as FATE and PySyft) or a self-developed framework, which supports distributed model training. Security multi-party computing (SMPC) and differential privacy (DP) mechanisms are designed to ensure data privacy protection during data transmission and calculation.

[0072] At each power plant locally, machine learning models (such as regression models and neural networks) are trained using the aligned financial data and production data. For different tasks (such as cost prediction and equipment failure warning), customized model architectures and loss functions are designed. The FedAvg or FedDistill algorithm is used to aggregate the local model parameters of each power plant to generate a global model. A model update strategy is designed to regularly distribute the global model to each power plant to achieve continuous optimization of the model.

[0073] Federated learning, while protecting data privacy, makes full use of the data resources of each power plant to improve the generalization ability and prediction accuracy of the model. The construction of the global model supports cross-power-plant collaborative analysis and provides data support for enterprise-level decision-making.

[0074] S104. Perform correlation analysis on the financial data and production data in the federated learning model to obtain the analysis results corresponding to the correlation analysis.

[0075] Specifically, in this step, through the federated learning model, in-depth correlation analysis is performed on financial data and production data to explore the quantitative relationship between financial indicators and production indicators.

[0076] For example, use the Apriori or FP-Growth algorithm to explore the association rules between financial indicators (such as marketing expenses) and production indicators (such as unit output efficiency). Design confidence and support thresholds to filter out strong association rules. Based on causal inference methods (such as Granger causality test, structural equation model), analyze the causal relationship between financial data and production data. For specific scenarios (such as cost calculation during peak and valley electricity price periods), construct a causal graph model to quantify the impact of financial decisions on production efficiency. Use visualization tools (such as Tableau, Power BI) to display the analysis results in the form of charts to support decision-makers' intuitive understanding. Introduce explainable AI (XAI) technology to provide an interpretive report of model predictions and enhance the credibility of the analysis results.

[0077] The correlation analysis reveals the internal connection between financial data and production data, providing a scientific basis for enterprises to optimize resource allocation. Causal analysis supports precise decision-making and helps enterprises formulate more effective business strategies.

[0078] S105. Construct a cross-domain ontology knowledge graph based on the aligned financial data and production data, and construct a data asset management platform according to the cross-domain ontology knowledge graph and the analysis results.

[0079] Specifically, based on the aligned financial data and production data, construct a cross-domain ontology knowledge graph and develop a data asset management platform. The interface of this platform is as Figure 2As shown, it realizes the unified management and intelligent analysis of data. Use graph databases (such as Neo4j, TigerGraph) to store the semantically aligned financial data and production data to form a knowledge graph. Design a knowledge inference engine to support intelligent query and reasoning based on the graph. Develop a data asset management platform to provide functional modules such as data query, analysis, and visualization. Integrate the federated learning model and the association analysis results to support real-time prediction and decision-making suggestions. Deploy the platform in the enterprise private cloud or hybrid cloud environment to ensure data security and system stability. Adopt a microservices architecture and containerization technology (such as Kubernetes) to support the elastic expansion and high availability of the platform.

[0080] The deep integration of financial data and production data is realized through the data asset management platform, supporting enterprise-level data-driven decision-making. The cross-domain ontology knowledge graph provides a powerful knowledge foundation for the intelligent analysis and reasoning of data, enhancing the data value of the enterprise.

[0081] Through the complication, refinement, and improvement of the above steps S101 to S105, this method not only solves the problems of ontology gap, time series misalignment, and value island existing in traditional systems, but also realizes the deep integration and intelligent analysis of financial data and production data through technologies such as federated learning, association analysis, and knowledge graph. This method provides comprehensive technical support and implementation paths for enterprises to build an efficient, secure, and intelligent data asset management platform.

[0082] In some embodiments, generating a plurality of federated learning models corresponding to the power plants according to the aligned financial data and production data includes: performing standardization processing on the financial data and production data; the standardization processing at least includes missing value filling, outlier processing, and data normalization; obtaining a basic model corresponding to each power plant; inputting the power plant data corresponding to each power plant into the corresponding basic model; performing distributed training on the plurality of basic models according to a preset federated learning framework to obtain model parameters corresponding to each basic model; and aggregating the plurality of model parameters according to the federated average method to generate the federated learning model.

[0083] Missing value filling: Fill the missing values in the financial data and production data using the mean, median, or interpolation method. Outlier processing: Identify outliers through box plots or the 3σ principle and use truncation or smoothing methods to process them. Data normalization: Normalize the data to a unified range using the Min-Max or Z-Score method to eliminate the dimension difference.

[0084] For each power plant, according to its characteristics such as type (e.g., thermal power, hydropower, wind power) and location (e.g., coastal, inland), select or design a suitable basic model (e.g., linear regression, decision tree, neural network). Input the standardized financial data and production data of each power plant into the corresponding basic model for local training. Use a federated learning framework (e.g., FATE, PySyft) for distributed training to ensure data privacy protection. Adopt the federated average (FedAvg) method to perform weighted averaging on the local model parameters of each power plant to generate a global federated learning model.

[0085] In this embodiment, data standardization processing improves the quality and consistency of data, providing a reliable data basis for model training. Federated learning, while protecting data privacy, makes full use of the data resources of each power plant to generate a global model, supporting cross-power-plant collaborative analysis. The customized design of the basic model improves the adaptability and prediction accuracy of the model.

[0086] Exemplarily, obtaining the basic model corresponding to each power plant includes: obtaining the power plant characteristic information of each power plant; the power plant characteristic information at least includes power plant type and power plant location; confirming the basic model corresponding to the power plant according to the power plant characteristic information.

[0087] Extract characteristic information such as power plant type (e.g., thermal power, hydropower, wind power) and location (e.g., coastal, inland) from the power plant management system. According to the power plant type and location, select or design a suitable basic model. For example, a thermal power plant may select a regression model based on energy consumption, and a wind power plant may select a neural network model based on wind speed.

[0088] The example designs a customized basic model by combining power plant characteristic information, improving the pertinence and prediction ability of the model. It supports differential modeling of multiple types of power plants, enhancing the generalization ability of the federated learning model.

[0089] In some embodiments, performing correlation analysis on the financial data and production data in the federated learning model to obtain the analysis result corresponding to the correlation analysis includes: extracting financial data features and production data features in the federated learning model; obtaining significant features from the financial data features and production data features according to a preset feature selection algorithm; obtaining the correlation rules corresponding to the financial data features and production data features according to a preset correlation rule mining algorithm; obtaining the causal information of the financial data features and production data features according to a preset causal analysis algorithm; generating the analysis result corresponding to the correlation analysis according to the significant features, correlation rules, and causal information.

[0090] By extracting financial data features (such as marketing revenue, costs) and production data features (such as power generation, equipment efficiency) from the federated learning model. Using feature selection algorithms (such as LASSO, PCA) to screen out significant features and reduce the data dimension. Using algorithms such as Apriori or FP-Growth to mine the association rules between financial data features and production data features. Adopting Granger causality test or structural equation model to analyze the causal relationship between financial data features and production data features. Combining significant features, association rules and causal information to generate association analysis results, such as "an increase in marketing expenses leads to an increase in power generation".

[0091] This embodiment reveals the internal connection between financial data and production data through association analysis, providing a scientific basis for enterprises to optimize resource allocation. Causal analysis supports precise decision-making and helps enterprises formulate more effective business strategies.

[0092] In some embodiments, a cross-domain ontology knowledge graph is constructed based on the aligned financial data and production data, including: obtaining the attribute information of the aligned financial data and production data, and constructing an initial knowledge graph according to the attribute information; performing knowledge extraction on a plurality of the financial data and production data based on natural language processing technology to obtain financial data knowledge and production data knowledge; obtaining potential association information corresponding to the financial data and production data according to a preset knowledge reasoning algorithm; adding the financial data knowledge, production data knowledge and potential association information to the initial knowledge graph to obtain the cross-domain ontology knowledge graph.

[0093] By extracting attribute information (such as financial indicators, production indicators) from the aligned financial data and production data. Using a graph database (such as Neo4j) to construct an initial knowledge graph to represent the basic relationship between financial data and production data. Based on natural language processing technology (such as BERT, OpenIE), extracting financial data knowledge and production data knowledge from text data. Using a knowledge reasoning algorithm (such as rule reasoning, graph embedding) to mine the potential association information between financial data and production data. Adding the extracted knowledge and the inferred association information to the initial knowledge graph to form a cross-domain ontology knowledge graph.

[0094] The deep integration of financial data and production data is achieved through the cross-domain ontology knowledge graph, supporting intelligent query and reasoning. Knowledge extraction and reasoning expand the coverage of the knowledge graph and improve the utilization value of data.

[0095] In some embodiments, constructing a data asset management platform based on the cross-domain ontology knowledge graph and the analysis result includes: obtaining the architecture of the data asset management platform, where the architecture includes a data access layer, a data processing layer, a knowledge graph layer, and an application layer; setting the cross-domain ontology knowledge graph in the knowledge graph layer and the analysis result in the application layer to complete the construction of the data asset management platform.

[0096] Design a four-layer architecture for the data asset management platform: Data access layer: Responsible for accessing and cleaning multi-source data. Data processing layer: Responsible for data standardization, alignment, and modeling. Knowledge graph layer: Stores the cross-domain ontology knowledge graph. Application layer: Provides data query, analysis, and visualization functions. Set the cross-domain ontology knowledge graph in the knowledge graph layer and the correlation analysis result in the application layer to complete the platform construction.

[0097] Through the data asset management platform, unified management and intelligent analysis of financial data and production data are realized, supporting enterprise-level data-driven decision-making. The four-layer architecture design improves the modularity and scalability of the platform.

[0098] In some embodiments, after constructing the data asset management platform based on the cross-domain ontology knowledge graph and the analysis result, it further includes: calculating the global risk coefficient corresponding to each power plant according to the cross-domain ontology knowledge graph and the analysis result of the data asset management platform; if it is determined that one of the power plants is abnormal according to the global risk coefficient, generating abnormal warning information corresponding to the power plant.

[0099] Based on the cross-domain ontology knowledge graph and the correlation analysis result, calculate the global risk coefficient of each power plant (such as financial risk, production risk). If the global risk coefficient of a certain power plant exceeds the preset threshold, generate abnormal warning information (such as "high risk of equipment failure").

[0100] The global risk coefficient provides a quantitative assessment of the overall risk of the power plant, supporting risk warning and control. The abnormal warning information helps the enterprise to detect and solve problems in a timely manner, reducing the operation risk.

[0101] Exemplarily, the expression of the global risk coefficient includes:

[0102] R = w f ·R f + w p ·R p + w e ·R e ;;

[0103] Wherein, R is the global risk coefficient, R f ,, R p and R eThey are the financial risk coefficient, the production risk coefficient, and the environmental risk coefficient respectively; w f ,, w p and w e are the weights corresponding to the financial risk coefficient, the production risk coefficient, and the environmental risk coefficient respectively. w f 、+w p +w e = 1.

[0104] Among them, the expressions of R f ,, R p and R e include:

[0105]

[0106] α, β, γ, δ, ∈, ζ are weight coefficients, which are set according to industry standards or enterprise requirements.

[0107] In some embodiments, the provided method also integrates the device locations and statuses in each power plant into the knowledge graph. Through sensors and the SCADA system, the location (such as latitude and longitude coordinates) and status (such as running, faulty, under maintenance) of each device in the power plant are collected in real time. The device location and status data are cleaned and standardized to ensure data quality and consistency. In the existing cross-domain ontology knowledge graph, new device nodes and device status nodes are added, and the association relationships between the devices and financial data and production data are established. A graph database (such as Neo4j) is used to store the extended knowledge graph. Based on the extended knowledge graph, knowledge inference rules are designed to mine the impact of device status on financial and production data. Visualization tools (such as Gephi, Cytoscape) are used to display the distribution and association relationships of device locations and statuses in the knowledge graph.

[0108] By integrating the device locations and statuses into the knowledge graph, the in-depth integration of device-level data is achieved, supporting finer-grained analysis and decision-making. The knowledge inference rules reveal the impact of device status on financial and production data, helping enterprises optimize device management and maintenance strategies. The visualization display provides an intuitive device status monitoring and risk warning ability.

[0109] Exemplarily, the provided method also performs an association analysis of device status and financial data. By extracting device status data (such as failure rate, number of maintenance times) from the knowledge graph. Extracting financial data (such as maintenance cost, downtime loss) from the knowledge graph. Using an association rule mining algorithm (such as Apriori) to analyze the association relationship between device status and financial data. For example, it is found that "for every 1% increase in the device failure rate, the maintenance cost increases by 50,000 yuan".

[0110] Revealing the direct impact of equipment status on financial data through association analysis supports enterprises in formulating more accurate equipment maintenance budgets and strategies. The extension and reasoning capabilities of the knowledge graph provide data support for equipment management and financial decision-making.

[0111] The method of integrating equipment location and status into the knowledge graph realizes the deep integration and intelligent analysis of equipment-level data. These embodiments further enhance the functions of the data asset management platform and provide more comprehensive and accurate decision-making support for enterprises.

[0112] The method provided by this application has at least the following beneficial effects:

[0113] Solving the problem of data dimension fragmentation: Through semantic alignment and temporal alignment, it ensures that financial data and production data in different data sources are consistent in semantic and temporal dimensions, solving the problem of data dimension fragmentation.

[0114] Improving data utilization efficiency: Through the federated learning model, it is possible to use the data of multiple power plants for model training without sharing the original data, improving data utilization efficiency.

[0115] Revealing potential relationships between data: Through association analysis, it reveals the potential relationships between financial data and production data, helping enterprises better understand the business logic behind the data.

[0116] Constructing a cross-domain knowledge graph: By constructing a cross-domain ontology knowledge graph, it integrates data from different domains, provides a more comprehensive data view, and helps enterprises conduct cross-domain data analysis and decision-making.

[0117] Enhancing data management capabilities: By constructing a data asset management platform, it integrates and displays the data of different power plants, provides data query, analysis, and visualization functions, and enhances the data management capabilities of enterprises.

[0118] Please refer to Figure 3 as shown in Figure 3 is a schematic structural diagram of a construction device 200 of a data asset management platform provided by an embodiment of this application. The construction device 200 of this data asset management platform is used to execute the steps of the construction method of the data asset management platform shown in the above embodiments. The construction device 200 of this data asset management platform can be a single server or a server cluster, or the construction device 200 of this data asset management platform can be a terminal, and this terminal can be a handheld terminal, a laptop, a wearable device, or a robot, etc.

[0119] As Figure 3 shown, the construction device 200 of the data asset management platform includes:

[0120] A data acquisition unit 201 for acquiring power plant data of multiple power plants; the power plant data includes financial data and production data; the financial data includes multiple financial indicators, and the production data includes multiple production indicators.

[0121] An indicator acquisition unit 202 for respectively acquiring semantic information and time series information corresponding to the financial indicators and production indicators, performing semantic alignment on the financial indicators and production indicators according to the semantic information, and performing time series alignment on the financial indicators and production indicators according to the time series information.

[0122] A model generation unit 203 for generating federated learning models corresponding to the multiple power plants according to the aligned financial data and production data.

[0123] An association analysis unit 204 for performing association analysis on the financial data and production data according to the federated learning model to obtain an analysis result corresponding to the association analysis.

[0124] A graph construction unit 205 for constructing a cross-domain ontology knowledge graph according to the aligned financial data and production data, and constructing a data asset management platform according to the cross-domain ontology knowledge graph and the analysis result.

[0125] It should be noted that those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the above-described data asset management platform construction device and each unit can refer to the corresponding processes in the data asset management platform construction embodiments described in the above embodiments, and will not be elaborated herein.

[0126] The construction of the above data asset management platform can be implemented in the form of a computer program, and this computer program can run on a device as shown in Figure 3 shown.

[0127] Please refer to Figure 4 , Figure 4 which is a schematic block diagram of the structure of a computer device provided by an embodiment of the present application. The computer device includes a processor, a memory, and a network interface connected through a device bus. Among them, the memory may include a storage medium and an internal memory.

[0128] The storage medium can store an operating device and a computer program. This computer program includes program instructions, and when the program instructions are executed, the processor can execute any construction of the data asset management platform.

[0129] The processor is used to provide computing and control capabilities to support the operation of the entire computer device.

[0130] The internal memory provides an environment for the operation of a computer program in a non-volatile storage medium. When the computer program is executed by a processor, the processor can be caused to execute the construction of any data asset management platform.

[0131] The network interface is used for network communication, such as sending assigned tasks, etc. Those skilled in the art can understand that Figure 4 The structure shown in is only a block diagram of some structures related to the solution of this application, and does not constitute a limitation on the terminal to which the solution of this application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine some components, or have different component arrangements.

[0132] It should be understood that the processor may be a central processing unit (CPU), and the processor may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0133] Among them, in one embodiment, the processor is used to run a computer program stored in the memory to implement the following steps:

[0134] Obtain power plant data of multiple power plants; the power plant data includes financial data and production data; the financial data includes multiple financial indicators, and the production data includes multiple production indicators;

[0135] Respectively obtain the semantic information and time series information corresponding to the financial indicators and production indicators, perform semantic alignment on the financial indicators and production indicators according to the semantic information, and perform time series alignment on the financial indicators and production indicators according to the time series information;

[0136] Generate a federated learning model corresponding to each of the multiple power plants according to the aligned financial data and production data;

[0137] Perform correlation analysis on the financial data and production data according to the federated learning model, and obtain the analysis result corresponding to the correlation analysis;

[0138] Construct a cross - domain ontology knowledge graph based on the aligned financial data and production data, and construct a data asset management platform according to the cross - domain ontology knowledge graph and the analysis result.

[0139] It should be noted that those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working process of the above - described processor can refer to the corresponding process in the construction embodiments of the data asset management platform described in the above - mentioned various embodiments, and will not be elaborated here.

[0140] An embodiment of the present application also provides a computer - readable storage medium. The computer - readable storage medium stores a computer program, and the computer program includes program instructions. The processor executes the program instructions to implement the steps of the method for constructing the data asset management platform provided in the above - mentioned various embodiments of the present application.

[0141] Among them, the computer - readable storage medium may be an internal storage unit of the computer device described in the foregoing embodiment, such as the hard disk or memory of the computer device. The computer - readable storage medium may also be an external storage device of the computer device, such as a plug - in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the computer device.

[0142] The above - mentioned is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of various equivalent modifications or substitutions, and these modifications or substitutions should all be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for constructing a data asset management platform, characterized in that, The method includes: Obtaining power plant data of multiple power plants; the power plant data includes financial data and production data; the financial data includes multiple financial indicators, and the production data includes multiple production indicators; Respectively obtaining semantic information and time series information corresponding to the financial indicators and production indicators, performing semantic alignment on the financial indicators and production indicators according to the semantic information, and performing time series alignment on the financial indicators and production indicators according to the time series information; Generating a federated learning model corresponding to each of the multiple power plants according to the aligned financial data and production data; Performing correlation analysis on the financial data and production data in the federated learning model, and obtaining an analysis result corresponding to the correlation analysis; Constructing a cross-domain ontology knowledge graph according to the aligned financial data and production data, and constructing a data asset management platform according to the cross-domain ontology knowledge graph and the analysis result.

2. The method according to claim 1, wherein The generating a federated learning model corresponding to each of the multiple power plants according to the aligned financial data and production data includes: Performing standardization processing on the financial data and production data; the standardization processing at least includes missing value filling, outlier processing, and data normalization; Obtaining a basic model corresponding to each power plant; Inputting the power plant data corresponding to each power plant into the corresponding basic model; Performing distributed training on the multiple basic models according to a preset federated learning framework, and obtaining model parameters corresponding to each basic model; Aggregating the multiple model parameters according to the federated average method to generate the federated learning model.

3. The method according to claim 2, wherein The obtaining a basic model corresponding to each power plant includes: Obtaining power plant feature information of each power plant; the power plant feature information at least includes power plant type and power plant location; Confirming the basic model corresponding to the power plant according to the power plant feature information.

4. The method according to claim 1, wherein The performing correlation analysis on the financial data and production data in the federated learning model, and obtaining an analysis result corresponding to the correlation analysis includes: Extracting financial data features and production data features in the federated learning model; Obtaining significant features from the financial data features and production data features according to a preset feature selection algorithm; Obtaining association rules corresponding to the financial data features and production data features according to a preset association rule mining algorithm; Obtaining causal information of the financial data features and production data features according to a preset causal analysis algorithm; Generating an analysis result corresponding to the correlation analysis according to the significant features, association rules, and causal information.

5. The method according to claim 1, wherein The constructing a cross-domain ontology knowledge graph according to the aligned financial data and production data includes: Obtaining attribute information of the aligned financial data and production data, and constructing an initial knowledge graph according to the attribute information; Performing knowledge extraction on the multiple financial data and production data based on natural language processing technology to obtain financial data knowledge and production data knowledge; Obtaining potential association information corresponding to the financial data and production data according to a preset knowledge reasoning algorithm; Adding the financial data knowledge, production data knowledge, and potential association information to the initial knowledge graph to obtain the cross-domain ontology knowledge graph.

6. The method according to claim 1, wherein Constructing a data asset management platform according to the cross-domain ontology knowledge graph and the analysis result, including: Obtaining the architecture of the data asset management platform, where the architecture includes a data access layer, a data processing layer, a knowledge graph layer, and an application layer; Setting the cross-domain ontology knowledge graph in the knowledge graph layer and the analysis result in the application layer to complete the construction of the data asset management platform.

7. The method according to claim 1, characterized in that, After constructing the data asset management platform according to the cross-domain ontology knowledge graph and the analysis result, it further includes: Calculating the global risk coefficient corresponding to each power plant according to the cross-domain ontology knowledge graph of the data asset management platform and the analysis result; If it is determined that an abnormality exists in one of the power plants according to the global risk coefficient, generating an abnormality warning message corresponding to the power plant.

8. A construction device for a data asset management platform, characterized in that, The device includes: A data acquisition unit for acquiring power plant data of multiple power plants; the power plant data includes financial data and production data; the financial data includes multiple financial indicators, and the production data includes multiple production indicators; An index acquisition unit for respectively acquiring the semantic information and time series information corresponding to the financial indicators and production indicators, performing semantic alignment on the financial indicators and production indicators according to the semantic information, and performing time series alignment on the financial indicators and production indicators according to the time series information; A model generation unit for generating a federated learning model corresponding to multiple power plants according to the aligned financial data and production data; An association analysis unit for performing association analysis on the financial data and production data according to the federated learning model to obtain the analysis result corresponding to the association analysis; A graph construction unit for constructing a cross-domain ontology knowledge graph according to the aligned financial data and production data, and constructing a data asset management platform according to the cross-domain ontology knowledge graph and the analysis result.

9. A computer device, characterized in that, Including a memory and a processor; the memory is used to store a computer program; the processor is used to execute the computer program and implement the method according to any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, and when the computer-readable instruction is executed by the processor, it causes one or more processors to execute the steps of the method according to any one of claims 1 to 7.

Images