Smart city multi-industry big data fusion analysis and decision-making method, system and medium
Through the federated learning framework and dynamic weight fusion technology, the problem of data interoperability among various industries in smart cities has been solved, cross-industry collaborative optimization has been achieved, urban operation efficiency and decision-making accuracy have been improved, data security and privacy have been guaranteed, and the quality of life of residents has been improved.
Patent Information
- Application Number
- CN202510957722.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-11
- Publication Date
- 2025-10-17
AI Technical Summary
In smart cities, data from various industries are stored independently and in heterogeneous formats, making them difficult to communicate with each other. Traditional analysis cannot capture cross-industry correlations, decision-making relies on manual experience, and lacks real-time multi-dimensional support. Existing platforms cannot handle semantic conflicts, machine learning models have poor generalization capabilities, weak cross-scenario adaptability, and data security and privacy compliance issues are not resolved.
Through the federated learning framework, dynamic weight fusion and multi-objective optimization criteria, multi-industry data is obtained for preprocessing and standardization, and a unified spatiotemporal benchmark is established using industry knowledge graphs. Cross-industry prediction models are trained in an encrypted environment, weight coefficients are dynamically generated and vector weighted fusion is performed to obtain cross-industry global fusion feature vectors, and finally the optimal decision-making plan is screened.
It has achieved cross-industry collaborative optimization of smart cities, improved urban operation efficiency and residents' quality of life, and ensured data security and privacy compliance.
Smart Images
Figure CN120805056A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of data processing and analysis, in particular to a smart city multi-industry big data fusion analysis and decision method, system and medium. BACKGROUND
[0002] In the development of smart cities, industry data is often stored independently, with heterogeneous formats, forming data silos and being difficult to interconnect. Traditional single-field analysis cannot capture cross-industry correlations, and decision-making relies on human experience, lacking real-time multi-dimensional decision support. Existing simple data integration platforms cannot handle semantic conflicts, machine learning models have poor generalization ability, and have weak cross-scene adaptability, and have not solved the fusion problem under data security and privacy compliance.
[0003] In view of the above problems, an effective technical solution is urgently needed. SUMMARY
[0004] The purpose of the present application is to provide a smart city multi-industry big data fusion analysis and decision method, system and medium, which realizes the cross-industry collaborative optimization of smart cities through a federated learning framework, dynamic weight fusion and multi-objective optimization criteria, and improves the efficiency of urban operation and the quality of life of residents.
[0005] The present application also provides a smart city multi-industry big data fusion analysis and decision method, comprising the following steps: Obtain multi-industry data and perform preprocessing, and standardize the preprocessed multi-industry data; Map the standardized multi-industry data according to an industry knowledge graph, and establish a unified space-time reference; Train and process the standardized multi-industry data in an encrypted environment to obtain a cross-industry prediction model; Monitor the standardized multi-industry data to obtain a plurality of industry features and an original feature vector, dynamically process the industry features to generate a weight coefficient, and perform vector weighted fusion to obtain a cross-industry global fusion feature vector; Process the cross-industry global fusion feature vector to obtain a plurality of decision results and evaluation results, and screen the optimal decision scheme.
[0006] Optionally, in the smart city multi-industry big data fusion analysis and decision method described in the present application, the step of obtaining multi-industry data and performing preprocessing, and standardizing the preprocessed multi-industry data comprises: Obtain multi-industry data including traffic data, energy data, medical data and environmental monitoring data through a pre-set city data source; Preprocess the multi-industry data through a pre-set rule engine and a data mining model; The pre-processed multi-industry data is standardized by a preset data conversion tool to obtain standardized multi-industry data.
[0007] Optionally, in the smart city multi-industry big data fusion analysis and decision method, the standardized multi-industry data is mapped according to the industry knowledge graph to establish a unified space-time reference, comprising: The standardized multi-industry data is processed by a preset natural language processing model to obtain semantic analysis information; The semantic analysis information is processed by a preset graph reasoning model to obtain corresponding relationship information between the standardized multi-industry data and the industry knowledge graph; The standardized multi-industry data is mapped according to the corresponding relationship information to establish a unified space-time reference.
[0008] Optionally, in the smart city multi-industry big data fusion analysis and decision method, the standardized multi-industry data is trained and processed in an encrypted environment to obtain a cross-industry prediction model, comprising: The standardized multi-industry data is trained and processed by a preset distributed machine learning model to obtain prediction model parameters; The prediction model parameters are encrypted and transmitted by a preset encryption model; The prediction model parameters are processed to obtain an improved cross-industry prediction model and shared.
[0009] Optionally, in the smart city multi-industry big data fusion analysis and decision method, the standardized multi-industry data is monitored to obtain a plurality of industry features and original feature vectors, the industry features are dynamically processed to generate weight coefficients and vector weighted fusion is performed to obtain a cross-industry global fusion feature vector, comprising: The standardized multi-industry data is dynamically monitored to obtain a plurality of industry data features and corresponding original feature vectors; The plurality of industry data features are processed by a preset attention mechanism model to generate corresponding weight coefficients; The corresponding original feature vectors are weighted and fused according to the weight coefficients to obtain a cross-industry global fusion feature vector.
[0010] Optionally, in the smart city multi-industry big data fusion analysis and decision method, the cross-industry global fusion feature vector is processed to obtain a plurality of decision results and evaluation results, and the optimal decision scheme is screened, comprising: The cross-industry global fusion feature vector is input into the improved cross-industry prediction model for processing to obtain a plurality of decision results; The plurality of decision results are processed by a preset decision evaluation model to obtain evaluation results corresponding to each decision result. The evaluation results corresponding to each decision result are screened according to a Pareto optimality criterion to generate an optimal decision scheme.
[0011] In a second aspect, the present application provides a smart city multi-industry big data fusion analysis and decision system, which comprises a memory and a processor, the memory comprising a smart city multi-industry big data fusion analysis and decision method program, and the smart city multi-industry big data fusion analysis and decision method program being executed by the processor to implement the following steps: Obtaining multi-industry data and preprocessing, standardizing the preprocessed multi-industry data; Mapping the standardized multi-industry data according to an industry knowledge graph to establish a unified space-time reference; Training the standardized multi-industry data in an encrypted environment to obtain a cross-industry prediction model; Monitoring the standardized multi-industry data to obtain a plurality of industry features and an original feature vector, dynamically processing the industry features to generate a weight coefficient and performing vector weighted fusion to obtain a cross-industry global fusion feature vector; Processing the cross-industry global fusion feature vector to obtain a plurality of decision results and evaluation results, and screening an optimal decision scheme.
[0012] Optionally, in the smart city multi-industry big data fusion analysis and decision system described in the present application, the obtaining multi-industry data and preprocessing, standardizing the preprocessed multi-industry data comprises: Obtaining multi-industry data including traffic data, energy data, medical data and environmental monitoring data through a preset city data source; Preprocessing the multi-industry data through a preset rule engine and a data mining model; Standardizing the preprocessed multi-industry data through a preset data conversion tool to obtain standardized multi-industry data.
[0013] Optionally, in the smart city multi-industry big data fusion analysis and decision system described in the present application, the mapping the standardized multi-industry data according to an industry knowledge graph to establish a unified space-time reference comprises: Processing the standardized multi-industry data through a preset natural language processing model to obtain semantic analysis information; Processing the semantic analysis information through a preset graph reasoning model to obtain corresponding relationship information between the standardized multi-industry data and the industry knowledge graph; According to the correspondence information, standardized multi-industry data is mapped, and a unified space-time reference is established.
[0014] In a third aspect, the present application also provides a computer readable storage medium, wherein a smart city multi-industry big data fusion analysis and decision method program is stored in the computer readable storage medium, and when the smart city multi-industry big data fusion analysis and decision method program is executed by a processor, the steps of the smart city multi-industry big data fusion analysis and decision method according to any one of the above are implemented.
[0015] As can be seen from the above, the smart city multi-industry big data fusion analysis and decision method, system and medium provided by the present application, by acquiring multi-industry data and preprocessing, standardizing the preprocessed multi-industry data, mapping the standardized multi-industry data according to the industry knowledge graph, establishing a unified space-time reference, training the standardized multi-industry data in an encrypted environment, obtaining a cross-industry prediction model, monitoring the standardized multi-industry data, dynamically generating industry data feature weight coefficients, obtaining a cross-industry global fusion feature vector, processing the global fusion feature vector, obtaining multiple decision results and evaluation results, and screening the optimal decision scheme, the smart city cross-industry collaborative optimization is realized through the federated learning framework, dynamic weight fusion and multi-objective optimization criteria, and the city operation efficiency and the quality of life of residents are improved.
[0016] Other features and advantages of the present application will be set forth in the following description, and in part will become apparent to those skilled in the art from the description, or can be learned by practice of the present application. The objects and other advantages of the present application can be realized and obtained by the structure particularly pointed out in the written description and the accompanying drawings. BRIEF DESCRIPTION OF DRAWINGS
[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments of the present application. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as limiting the scope. For those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.
[0018] Figure 1 The flow chart of the smart city multi-industry big data fusion analysis and decision method provided by the embodiments of the present application; Figure 2 The flow chart of data processing of the smart city multi-industry big data fusion analysis and decision method provided by the embodiments of the present application; Figure 3 The flow chart of semantic mapping of the smart city multi-industry big data fusion analysis and decision method provided by the embodiments of the present application; Figure 4 A flowchart of the federated learning method for the smart city multi-industry big data fusion analysis and decision-making provided in the embodiment of this application; DETAILED DESCRIPTION
[0019] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work fall within the scope of protection of the present application.
[0020] It should be noted that similar reference numerals and letters represent similar items in the following drawings. Therefore, once an item is defined in one drawing, it does not need to be further defined or explained in subsequent drawings. At the same time, in the description of this application, the terms "first", "second", etc. are only used to distinguish the description and should not be understood as indicating or implying relative importance.
[0021] Please refer to Figure 1 , Figure 1 This is a flow chart of a smart city multi-industry big data fusion analysis and decision-making method in some embodiments of the present application. The smart city multi-industry big data fusion analysis and decision-making method is used in terminal devices, such as computers and mobile phones. The smart city multi-industry big data fusion analysis and decision-making method includes the following steps: S11. Acquire and preprocess multi-industry data, and standardize the preprocessed multi-industry data; S12. Map the standardized multi-industry data according to the industry knowledge graph to establish a unified spatiotemporal benchmark; S13. Training the standardized multi-industry data in an encrypted environment to obtain a cross-industry prediction model; S14. Monitoring the standardized multi-industry data to obtain multiple industry characteristics and original feature vectors, dynamically processing the industry characteristics to generate weight coefficients and performing vector weighted fusion to obtain a cross-industry global fusion feature vector; S15. Process the cross-industry global fusion feature vector to obtain multiple decision results and evaluation results, and select the optimal decision solution.
[0022] Wherein, the urban multi-industry big data is collected and standardized, the industry data is mapped according to the industry knowledge graph to establish a unified space-time benchmark, and the optimal decision scheme is obtained through a federal learning framework, dynamic weight fusion, multi-objective optimization criteria and decision evaluation fusion, so as to realize the cross-industry collaborative optimization of the smart city.
[0023] Please refer to Figure 2 , Figure 2 is a flow chart of data processing of the smart city multi-industry big data fusion analysis and decision method in some embodiments of the application. According to the embodiment of the application, the multi-industry data is obtained and preprocessed, and the preprocessed multi-industry data is standardized, including: S21, obtaining multi-industry data through a preset city data source, including traffic data, energy data, medical data and environmental monitoring data; S22, pre-processing the multi-industry data through a preset rule engine and a data mining model; S23, standardizing the preprocessed multi-industry data through a preset data conversion tool to obtain standardized multi-industry data.
[0024] Wherein, the industry data of multiple industries of the city is obtained through a plurality of channels through a preset city database or data source through an API gateway, a Kafka collection tool and the like, the multi-industry data in the embodiment of the present scheme includes traffic data, energy data, medical data and environmental monitoring data, wherein the traffic data includes road conditions, traffic flow and public transportation operation conditions, the energy data includes power, gas and water use conditions, the medical data includes hospital visits, medical resource distribution, and the environmental monitoring data includes air quality, noise monitoring and water quality monitoring, and the various types of industry data are further preprocessed according to a preset rule engine and a data mining model, wherein the preset engine and the data mining model first use the data mining model to find valuable target data, and then add the data to the rule library of the preset engine to optimize the engine workflow, including noise reduction, cleaning and correction of the data, so that the data is more suitable for analysis requirements, improves the data processing quality, and obtains the preprocessed multi-industry data, and then the multi-industry data is standardized through a preset data conversion tool such as an ETL tool, which includes unified data format and coding, to obtain standardized data, and ensure the fusibility of different data sources.
[0025] Please refer to Figure 3 , Figure 3 is a flow chart of semantic mapping of the smart city multi-industry big data fusion analysis and decision method in some embodiments of the application. According to the embodiment of the application, the standardized multi-industry data is mapped according to the industry knowledge graph to establish a unified space-time benchmark, including: S31, processing the standardized multi-industry data through a preset natural language processing model to obtain semantic analysis information; S32, processing the semantic analysis information through a preset graph reasoning model to obtain corresponding relationship information between the standardized multi-industry data and the industry knowledge graph; S33, mapping the standardized multi-industry data according to the corresponding relationship information to establish a unified time-space reference.
[0026] The standardized multi-industry data is processed through a preset natural language processing model such as an NLP engine to obtain semantic analysis information. The semantic analysis information is processed through a preset graph reasoning model to obtain standard concepts and corresponding relationship information between the standardized multi-industry data and the industry knowledge graph, realizing semanticization and standardization of the data. The standardized multi-industry data is mapped according to the corresponding relationship information, aligning the time stamps of all data to a unified time zone to ensure consistency of data from different sources in the time dimension. Spatial positions such as longitude and latitude and grids are mapped to a unified spatial reference system to enable comparison and correlation of data from different sources in the spatial dimension. The industry knowledge graph provides time-space correlation rules for describing the correlation between entities in time and space. By establishing a unified time-space reference, the time and space differences between different data sources are eliminated, enabling integration and analysis of data in a unified time-space framework and realizing fusion analysis of the data.
[0027] Please refer to Figure 4 , Figure 4 is a flowchart of federated learning of the smart city multi-industry big data fusion analysis and decision method in some embodiments of the present application. According to the present application, the training and processing of the standardized multi-industry data in an encrypted environment to obtain a cross-industry prediction model includes: S41, training and processing the standardized multi-industry data through a preset distributed machine learning model to obtain prediction model parameters; S42, encrypting the prediction model parameters through a preset encryption model and transmitting them after encryption; S43, processing the prediction model parameters to obtain an improved cross-industry prediction model and sharing it.
[0028] The standardized multi-industry data is trained by a preset distributed machine learning model such as a federal learning framework, and prediction model parameters such as gradients and parameter changes are obtained, the prediction model parameter data is encrypted by a preset encryption model such as a TLS / SSL or an asymmetric encryption algorithm, and then transmitted, wherein the standardized multi-industry data is always stored locally to prevent leakage to other participants, joint modeling is realized on the premise of protecting the data privacy and security of each participant, the prediction model parameters are processed, such as by using a stacking fusion algorithm or a simple weighted average algorithm, an improved global prediction model is obtained, and the accuracy and recall rate of the improved model are verified to realize the prediction accuracy of the optimized model, finally, the model is shared and deployed, and a comprehensive cross-industry factor prediction model is constructed.
[0029] According to the embodiment of the present application, the standardized multi-industry data is monitored to obtain a plurality of industry characteristics and an original feature vector, the industry characteristics are dynamically processed to generate a weight coefficient and perform vector weighted fusion, and a cross-industry global fusion feature vector is obtained, including: The standardized multi-industry data is dynamically monitored to obtain a plurality of industry data characteristics and corresponding original feature vectors; The plurality of industry data characteristics are processed by a preset attention mechanism model to generate corresponding weight coefficients; The corresponding original feature vectors are weighted and fused according to the weight coefficients to obtain a cross-industry global fusion feature vector.
[0030] The standardized multi-industry data is monitored to obtain a plurality of industry data characteristics, and a preset attention mechanism model such as a Transformer strategy is used to calculate the information gain, mutual information, and measure index of each industry data characteristic, and a target function is used as a tool to obtain the minimum prediction error or the maximum classification accuracy index, dynamically adjust the feature weight, and generate a weight coefficient corresponding to each industry data characteristic, so that the model obtains the optimization of accuracy and effectiveness in processing the preset task, the generation of the weight coefficient is also related to the relevance of the data characteristics and the current task, the spatiotemporal position, the event type, and the fluctuation, and the original feature vectors corresponding to the plurality of generated weight coefficients are weighted and summed to obtain an industry feature vector, and the industry feature vectors of different sources are fused to obtain a cross-industry global fusion feature vector, providing a more targeted and timely comprehensive feature representation.
[0031] According to the embodiment of the present application, the cross-industry global fusion feature vector is processed to obtain a plurality of decision results and evaluation results, and the optimal decision scheme is screened, including: The cross-industry global fusion feature vector is input into the improved cross-industry prediction model for processing to obtain a plurality of decision results; The plurality of decision results are processed by a preset decision evaluation model to obtain evaluation results corresponding to each decision result; The evaluation results corresponding to each decision result are screened according to a Pareto optimality criterion to generate an optimal decision scheme.
[0032] The plurality of decision results are obtained by inputting the cross-industry global fusion feature vector into a preset cross-industry prediction model for processing. The prediction model will make predictions of development trends and key indicators according to information in historical data and the feature vector through correlations and laws among multiple industries. Then, the decision evaluation model is used to process the decision results, wherein the decision evaluation is performed on the decision results of different industries. Different decision fusion methods can be used to process and generate evaluation results of decisions, such as Bayesian evaluation, fault tree analysis, fuzzy comprehensive evaluation, and gray correlation degree evaluation fusion evaluation methods, to improve the accuracy and reliability of decision evaluation. Finally, the evaluation results corresponding to each decision result are screened according to the Pareto optimality criterion to obtain an optimal decision scheme for multi-objective collaboration.
[0033] In a second aspect, the present application also discloses a smart city multi-industry big data fusion analysis and decision system, which comprises a memory and a processor, wherein the memory comprises a smart city multi-industry big data fusion analysis and decision method program, and the smart city multi-industry big data fusion analysis and decision method program is executed by the processor to realize the following steps: Obtaining and preprocessing multi-industry data, and performing standardization processing on the preprocessed multi-industry data; Mapping the standardized multi-industry data according to an industry knowledge graph, and establishing a unified space-time reference; Training and processing the standardized multi-industry data in an encrypted environment to obtain a cross-industry prediction model; Monitoring the standardized multi-industry data to obtain a plurality of industry features and an original feature vector, dynamically processing the industry features to generate weight coefficients, and performing vector weighted fusion to obtain a cross-industry global fusion feature vector; Processing the cross-industry global fusion feature vector to obtain a plurality of decision results and evaluation results, and screening an optimal decision scheme.
[0034] The city multi-industry big data is collected and standardized processed, the industry data is mapped according to the industry knowledge graph to establish a unified space-time reference, and the optimal decision scheme is obtained through the federated learning framework, dynamic weight fusion, multi-objective optimization criterion and decision evaluation fusion, so as to realize the cross-industry collaborative optimization of the smart city.
[0035] According to the embodiment of the present application, the multi-industry data is acquired and preprocessed, and the preprocessed multi-industry data is standardized, comprising: The multi-industry data is acquired through a preset city data source, including traffic data, energy data, medical data and environmental monitoring data; The multi-industry data is preprocessed through a preset rule engine and a data mining model; The preprocessed multi-industry data is standardized through a preset data conversion tool to obtain standardized multi-industry data.
[0036] In the embodiment of the present application, the multi-industry data includes traffic data, energy data, medical data and environmental monitoring data, wherein the traffic data includes road conditions, traffic flow and public transportation operation conditions, the energy data includes power, gas and water usage, the medical data includes hospital visits, medical resource distribution, and the environmental monitoring data includes air quality, noise monitoring and water quality monitoring. According to the preset rule engine and the data mining model, various types of industry data are deeply preprocessed, wherein the preset engine and the data mining model first use the data mining model to find valuable target data, and then add these data to the rule library of the preset engine to optimize the engine workflow, including noise reduction, cleaning and correction of data, so that the data is more suitable for analysis requirements, improves the data processing quality, and obtains preprocessed multi-industry data. The multi-industry data is standardized through a preset data conversion tool such as ETL tool, including unified data format and coding, to obtain standardized data and ensure the fusion of different data sources.
[0037] According to the embodiment of the present application, the standardized multi-industry data is mapped according to the industry knowledge graph to establish a unified space-time reference, comprising: The standardized multi-industry data is processed through a preset natural language processing model to obtain semantic analysis information; The semantic analysis information is processed through a preset graph reasoning model to obtain corresponding relationship information between the standardized multi-industry data and the industry knowledge graph; The standardized multi-industry data is mapped according to the corresponding relationship information to establish a unified space-time reference.
[0038] The standardized multi-industry data is processed by a preset natural language processing model such as an NLP engine to obtain semantic analysis information, and the semantic analysis information is processed by a preset graph reasoning model to obtain standard concepts and corresponding relationship information of the standardized multi-industry data and the industry knowledge graph, so that the semantic and standardization of the data are realized, the standardized multi-industry data is mapped according to the corresponding relationship information, the time stamps of all data are aligned to a unified time zone, consistency of data from different sources in the time dimension is ensured, spatial positions such as longitude and latitude are mapped to a unified spatial reference system, so that data from different sources can be compared and associated in the spatial dimension, the industry knowledge graph provides spatiotemporal correlation rules for describing the correlation of entities in time and space, and by establishing a unified spatiotemporal reference, the time and space differences between different data sources are eliminated, so that the data can be integrated and analyzed in a unified spatiotemporal framework, and fusion analysis of the data is realized.
[0039] According to the embodiment of the present application, the standardized multi-industry data is processed in an encrypted environment to obtain a cross-industry prediction model, including: The standardized multi-industry data is processed by a preset distributed machine learning model to obtain prediction model parameters; The prediction model parameters are encrypted by a preset encryption model and then transmitted; The prediction model parameters are processed to obtain an improved cross-industry prediction model and shared.
[0040] The standardized multi-industry data is processed by a preset distributed machine learning model such as a federated learning framework to obtain prediction model parameters such as gradients and parameter changes, and the prediction model parameter data is encrypted by a preset encryption model such as TLS / SSL or an asymmetric encryption algorithm and then transmitted, wherein the standardized multi-industry data is always stored locally to prevent leakage to other participants, joint modeling is realized on the premise of protecting the data privacy and security of each participant, the prediction model parameters are then processed, such as by a stacking fusion algorithm or a simple weighted average algorithm, to obtain an improved global prediction model, the accuracy and recall rate of the improved model are verified to realize the prediction accuracy of the optimized model, and finally the model is shared and deployed to build a comprehensive cross-industry factor prediction model.
[0041] According to the embodiment of the present application, the standardized multi-industry data is monitored to obtain a plurality of industry features and original feature vectors, the industry features are dynamically processed to generate weight coefficients and perform vector weighted fusion to obtain a cross-industry global fusion feature vector, including: The standardized multi-industry data is dynamically monitored to obtain a plurality of industry data features and corresponding original feature vectors; The plurality of industry data features are processed by a preset attention mechanism model to generate corresponding weight coefficients; The corresponding original feature vectors are weighted and fused according to the weight coefficients to obtain a cross-industry global fusion feature vector.
[0042] The plurality of industry data features are obtained by monitoring standardized multi-industry data, and each industry data feature is processed by a preset attention mechanism model such as a Transformer strategy to calculate information gain, mutual information, and measure indicators of the feature, and a target function is used as a tool to obtain a minimum prediction error or a maximum classification accuracy indicator, and the feature weight is dynamically adjusted to generate a weight coefficient corresponding to each industry data feature, so that the model is optimized in accuracy and effectiveness in processing a preset task. The generation of the weight coefficient is also related to the relevance of the data feature and the current task, the spatiotemporal position, the event type, and the fluctuation, and the industry feature vectors of different sources are fused to obtain a cross-industry global fusion feature vector, which provides more targeted and timely comprehensive feature representation.
[0043] According to an embodiment of the present application, the cross-industry global fusion feature vector is processed to obtain a plurality of decision results and evaluation results, and an optimal decision scheme is screened, including: The improved cross-industry prediction model is input into the cross-industry global fusion feature vector for processing to obtain a plurality of decision results; The plurality of decision results are processed by a preset decision evaluation model to obtain evaluation results corresponding to each decision result; The evaluation results corresponding to each decision result are screened according to the Pareto optimality criterion to generate an optimal decision scheme.
[0044] The cross-industry global fusion feature vector is input into a preset cross-industry prediction model for processing to obtain a plurality of decision results, and the prediction model will make predictions of development trends and key indicators according to information in historical data and feature vectors through correlations and rules between multiple industries. The decision evaluation model is used to process the decision results, wherein the decision results of each different industry are evaluated, and different decision fusion methods can be used to process the evaluation results of the decisions, such as Bayesian evaluation, fault tree analysis, fuzzy comprehensive evaluation, and gray correlation degree evaluation fusion evaluation methods, to improve the accuracy and reliability of decision evaluation. Finally, the evaluation results corresponding to each decision result are screened according to the Pareto optimality criterion to obtain an optimal decision scheme with multi-objective coordination.
[0045] The third aspect of the present application provides a readable storage medium, wherein a smart city multi-industry big data fusion analysis and decision method program is stored in the readable storage medium, and when the smart city multi-industry big data fusion analysis and decision method program is executed by a processor, the steps of the smart city multi-industry big data fusion analysis and decision method according to any one of the preceding aspects are implemented.
[0046] The smart city multi-industry big data fusion analysis and decision method, system and medium disclosed in the present application acquire multi-industry data and perform preprocessing, perform standardization processing on the preprocessed multi-industry data, perform mapping on the standardized multi-industry data according to an industry knowledge graph, establish a unified space-time reference, perform training processing on the standardized multi-industry data in an encrypted environment, obtain a cross-industry prediction model, monitor the standardized multi-industry data, dynamically generate an industry data feature weight coefficient, obtain a cross-industry global fusion feature vector, process the global fusion feature vector, obtain multiple decision results and evaluation results, and screen an optimal decision scheme, so as to realize smart city cross-industry collaborative optimization, improve city operation efficiency and resident life quality through a federated learning framework, dynamic weight fusion and multi-objective optimization criteria.
[0047] In several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other manners. The above-described device embodiments are only illustrative, for example, the division of the units is only a logical function division, and there can be another division manner in actual implementation, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the various components shown or discussed can be indirect coupling or communication connection through some interfaces, devices or units, and can be electrical, mechanical or other forms.
[0048] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units; they can be located in one place or distributed on multiple network units; and part or all of the units can be selected according to actual needs to achieve the purpose of the present embodiment.
[0049] In addition, each functional unit in each embodiment of the present application can be integrated into one processing unit, or each unit can be a separate unit, or two or more units can be integrated into one unit; the integrated unit can be realized in the form of hardware or in the form of hardware plus software functional unit.
[0050] Those skilled in the art can understand that all or part of the steps of the above-mentioned method embodiments can be completed by relevant hardware of program instructions, and the foregoing program can be stored in a readable storage medium, and the program executes the steps of the above-mentioned method embodiments when executed; and the foregoing storage medium includes various media capable of storing program codes, such as a mobile storage device, a read-only memory (ROM), a random access memory (RAM), a magnetic disc or an optical disc.
[0051] Alternatively, the integrated unit of the present application can also be stored in a readable storage medium if it is realized in the form of a software function module and sold or used as an independent product. Based on such understanding, the technical solutions of the embodiments of the present application can be embodied in the form of a software product in essence or in the form of a part of the prior art that makes a contribution, and the software product is stored in a storage medium, including a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in the embodiments of the present application. The foregoing storage medium includes various media capable of storing program codes, such as a mobile storage device, a ROM, a RAM, a magnetic disc or an optical disc.
Claims
1. The smart city multi-industry big data fusion analysis and decision-making method is characterized by: The following steps are involved: Acquire and preprocess data from multiple industries, and standardize the preprocessed data; Mapping the standardized multi-industry data according to the industry knowledge graph to establish a unified spatiotemporal benchmark; Training and processing the standardized multi-industry data in an encrypted environment to obtain a cross-industry prediction model; Monitoring the standardized multi-industry data to obtain multiple industry characteristics and original feature vectors, dynamically processing the industry characteristics to generate weight coefficients and performing vector weighted fusion to obtain a cross-industry global fusion feature vector; The cross-industry global fusion feature vector is processed to obtain multiple decision results and evaluation results, and the optimal decision solution is screened.
2. The smart city multi-industry big data fusion analysis and decision-making method according to claim 1 is characterized by: The obtaining and preprocessing of multi-industry data and the standardization of the preprocessed multi-industry data include: Obtain multi-industry data through pre-set city data sources, including traffic data, energy data, medical data, and environmental monitoring data; Preprocessing the multi-industry data through a preset rule engine and data mining model; The pre-processed multi-industry data is standardized through preset data conversion tools to obtain standardized multi-industry data.
3. The smart city multi-industry big data fusion analysis and decision-making method according to claim 2 is characterized by: The standardized multi-industry data is mapped according to the industry knowledge graph to establish a unified spatiotemporal benchmark, including: Processing the standardized multi-industry data through a preset natural language processing model to obtain semantic parsing information; The semantic parsing information is processed through a preset graph reasoning model to obtain the corresponding relationship information between standardized multi-industry data and industry knowledge graphs; The standardized multi-industry data is mapped according to the corresponding relationship information to establish a unified time and space benchmark.
4. The smart city multi-industry big data fusion analysis and decision-making method according to claim 3 is characterized by: The training process of the standardized multi-industry data in an encrypted environment to obtain a cross-industry prediction model includes: By presetting a distributed machine learning model, the standardized multi-industry data is trained and processed to obtain prediction model parameters; Encrypting the prediction model parameters through a preset encryption model and then transmitting them; The prediction model parameters are processed to obtain an improved cross-industry prediction model and share it.
5. The smart city multi-industry big data fusion analysis and decision-making method according to claim 4 is characterized in that: The monitoring of the standardized multi-industry data to obtain multiple industry characteristics and original feature vectors, dynamically processing the industry characteristics to generate weight coefficients and performing vector weighted fusion to obtain a cross-industry global fusion feature vector, includes: Dynamically monitoring the standardized multi-industry data to obtain multiple industry data features and corresponding original feature vectors; Processing the multiple industry data features through a preset attention mechanism model to generate corresponding weight coefficients; The corresponding original feature vectors are weightedly fused according to the weight coefficients to obtain a cross-industry global fusion feature vector.
6. The smart city multi-industry big data fusion analysis and decision-making method according to claim 5 is characterized by: The processing of the cross-industry global fusion feature vector to obtain multiple decision results and evaluation results, and screening the optimal decision solution, includes: The cross-industry global fusion feature vector is input into the improved cross-industry prediction model for processing to obtain multiple decision results; Processing the multiple decision results through a preset decision evaluation model to obtain an evaluation result corresponding to each decision result; According to the Pareto optimality criterion, the evaluation results corresponding to each decision result are screened to generate the optimal decision plan.
7. Smart city multi-industry big data fusion analysis and decision-making system, characterized by: The system further includes: a memory and a processor, wherein the memory includes a program for a smart city multi-industry big data fusion analysis and decision-making method, wherein the program for the smart city multi-industry big data fusion analysis and decision-making method is used by the processor to obtain multi-industry data and perform preprocessing, and perform standardization on the preprocessed multi-industry data; Mapping the standardized multi-industry data according to the industry knowledge graph to establish a unified spatiotemporal benchmark; Training and processing the standardized multi-industry data in an encrypted environment to obtain a cross-industry prediction model; Monitoring the standardized multi-industry data to obtain multiple industry characteristics and original feature vectors, dynamically processing the industry characteristics to generate weight coefficients and performing vector weighted fusion to obtain a cross-industry global fusion feature vector; The cross-industry global fusion feature vector is processed to obtain multiple decision results and evaluation results, and the optimal decision solution is screened.
8. The smart city multi-industry big data fusion analysis and decision-making system according to claim 7 is characterized by: The obtaining and preprocessing of multi-industry data and the standardization of the preprocessed multi-industry data include: Obtain multi-industry data through pre-set city data sources, including traffic data, energy data, medical data, and environmental monitoring data; Preprocessing the multi-industry data through a preset rule engine and data mining model; The pre-processed multi-industry data is standardized through preset data conversion tools to obtain standardized multi-industry data.
9. The smart city multi-industry big data fusion analysis and decision-making system according to claim 8 is characterized by: The standardized multi-industry data is mapped according to the industry knowledge graph to establish a unified spatiotemporal benchmark, including: Processing the standardized multi-industry data through a preset natural language processing model to obtain semantic parsing information; The semantic parsing information is processed through a preset graph reasoning model to obtain the corresponding relationship information between standardized multi-industry data and industry knowledge graphs; The standardized multi-industry data is mapped according to the corresponding relationship information to establish a unified time and space benchmark.
10. A computer-readable storage medium, characterized in that The computer-readable storage medium includes a smart city multi-industry big data fusion analysis and decision-making method program. When the smart city multi-industry big data fusion analysis and decision-making method program is executed by a processor, the steps of the smart city multi-industry big data fusion analysis and decision-making method as described in any one of claims 1 to 6 are implemented.