53 results about "Spatiotemporal database" patented technology

The invention discloses a geographic information big data-oriented multi-source data processing system, and relates to the technical field of data acquisition and sensors, and the system comprises a data acquisition module which accesses a remote sensing satellite, an unmanned aerial vehicle, an Internet of Things sensor and social media in real time through a multi-source heterogeneous interface, and carries out the adaptive analysis of a data format and metadata marking; the distributed storage module is used for performing partition storage on the geographic information data based on a space-time database and an object storage architecture, and establishing dynamic space-time index and version control; the data fusion module is used for realizing coordinate system conversion, time sequence calibration and semantic knowledge graph matching based on a multi-source data alignment method of dynamic weight distribution; and an intelligent analysis module and a security control module. According to the method, core pain points such as data splitting, low storage efficiency, extensive analysis and compliance risks in the geographic information field are systematically solved, and a full-stack type technical base is provided for scenes such as smart cities, emergency disaster relief and environment monitoring.

The invention discloses a low-altitude integrated management system based on a grid digital twin model and an intelligent algorithm, and relates to the technical field of low-altitude aircrafts. The method comprises the following steps: multi-source sensing data fusion and grid coding are carried out, and an airspace environment space-time database is constructed; constructing an airspace environment space-time database to carry out three-dimensional subdivision on the airspace; generating a multi-scale grid twinborn body, and associating a corresponding attribute for each level of grid according to a management requirement; generating an airspace risk thermodynamic diagram in the future 5-30 minutes; establishing a grid state change triggering rule; and generating a multi-target optimal path set by taking the grid navigation cost as a weight. According to the invention, by combining the real-time operation data of the low-altitude aircraft, the meteorological environment and other data with the artificial intelligence algorithm and the navigation rule base, the airspace traffic flow is analyzed, the flight plan is optimized, the flight route is intelligently distributed, the automatic route setting of any two points is realized, and the response speed of the low-altitude flight service system and the processing capability of various flight data are improved.

The invention discloses a natural language spatio-temporal retrieval method and system based on a large model, and relates to the technical field of big data resource services, and the method comprises the following steps: finely adjusting a preset large language model according to a spatio-temporal retrieval problem and a preset cue word engineering template; dividing the target treatment area into geographic space grids, wherein each geographic space grid has a unique grid code; according to the business time-space database and the dynamic semantic tag, constructing an entity knowledge base classified by different data categories; performing natural language intention analysis on user input based on the fine-tuned large language model, and identifying spatial constraints, time constraints and service attribute constraints; converting the spatial constraint into a filtering condition; querying a data category of a corresponding key entity in an entity knowledge base; and outputting a space-time retrieval result based on the fine-tuned large language model according to service attribute constraints, data categories, spatial constraints and filtering conditions. The method gets rid of specialized dependence, improves the interaction experience, and is high in flexibility.

The invention discloses a foundation pit dewatering treatment method and system, electronic equipment and a storage medium, and relates to the field of foundation pit dewatering treatment. The method comprises the following steps: acquiring various data from sensors mounted at preset positions around a foundation pit and inside the foundation pit, integrating the various data with meteorological data and geological survey data, and establishing a space-time database; according to the data in the space-time database, using a preset LSTM neural network model to predict a first water level change in a future preset time, and according to the three-dimensional seepage field dynamic simulation engine, using a soil parameter inversion algorithm of machine learning to invert soil parameters; according to the first water level change, a variable frequency pump group cooperative control algorithm is used for cooperative control over the pump set, and pump set operation parameters are dynamically adjusted for drainage; and controlling the recharge pressure and the water quality according to the second water level change of the foundation pit after drainage and the soil parameters. According to the technical scheme, efficient and intelligent treatment of foundation pit dewatering is achieved.

The invention provides a coal mine multi-disaster intelligent sensing early warning and linkage emergency system, and relates to the technical field of coal mine safety, multi-source data collected by a hollow probe sensor group is uploaded to an edge gateway through an industrial ring network for primary processing, and the multi-source data after primary processing by the edge gateway is uploaded to a data layer through the industrial ring network; the processed multi-source data is classified by a time sequence database and a space database of the data layer, the classified multi-source data is analyzed by the analysis layer, and an analysis result of the analysis layer is uploaded to the early warning layer. According to the invention, three-in-one monitoring of deep part-surface-environment is realized through the hollow probe sensor group, and real-time and reliable data is guaranteed by combining edge calculation and dual-channel redundancy transmission; and the space-time database and the disaster chain rule base support analysis layer are fused with long and short-term memory network time sequence modeling and graph convolutional network space correlation and dynamic coupling correction to realize accurate quantification of the composite disaster probability.

The invention relates to the technical field of artificial intelligence, and discloses a precise targeted AI recommendation system. The method comprises the following steps: firstly, acquiring multi-source heterogeneous data, constructing a global space-time database through space-time alignment, complementing and repairing the data by using a generative adversarial network, and cleaning and verifying; and constructing a three-dimensional topological model based on the improved graph neural network, and fusing multiple features to generate a multi-resolution model. A scene simulation platform is established by adopting a multi-objective reinforcement learning algorithm, multi-objective optimization is set, and a strategy library is constructed. And screening key factors by using a sparse Bayesian theory to construct a lightweight model. And designing a layered dynamic optimization framework, solving an optimal solution of each sub-layer by using a hybrid algorithm, realizing cross-layer constraint synchronous decoupling through a block chain smart contract, and outputting a global optimal three-dimensional design scheme. The method improves the application level of the digital achievement of the power grid design, gives consideration to stability, economy and environmental protection, and has important application value.

The invention discloses a student psychological health prediction method and system based on action recognition, and the method comprises the steps: obtaining video data in multiple scenes of a school, processing and tracking students in the video data, and constructing a space-time database; associating time-space databases of different students, setting influence weights according to the interaction frequency and the physical distance, and establishing a social relation graph; acquiring characteristics of group behaviors in the space-time database, and performing interaction activeness analysis according to the social relation graph; establishing a dynamic social relation graph weight matrix according to interaction activeness analysis, and inputting various data of the interaction activeness into the matrix to carry out bidirectional prediction among the students; a thermodynamic diagram of the class is constructed according to the bidirectional prediction in the step S4, and intervention suggestions are fed back to the teacher in real time; by blocking a negative emotion propagation chain in time and optimizing psychological counseling resource allocation, the initiative and systematicness of campus psychological health management are remarkably enhanced.

The invention discloses a main beam formwork erection elevation adjustment method based on machine learning, and belongs to the technical field of elevation adjustment. The method comprises the steps that a space-time database is constructed by collecting data of the construction period of a cable-free section; processing the time sequence data by adopting an LSTM-Transform hybrid model, extracting long-period characteristics, and outputting an initial prediction value of the elevation; analyzing nonlinear variables such as a cable force change rate, a sunlight gradient and a material age by using an XG-Boost algorithm, and outputting a compensation factor; fusing and generating a joint prediction value; constructing a reinforcement learning agent, and outputting an adjustment instruction by taking a construction stage as a state space, taking an elevation adjustment amount as an action space, minimizing deviation between a predicted value and a measured value and taking construction stability as a reward function; the formwork erecting elevation is adjusted according to the driving hydraulic system, and the database is updated in real time. According to the invention, through multi-dimensional data fusion and intelligent optimization, the elevation adjustment precision and stability are improved.

The embodiment of the invention provides a rail transit 3D data defect detection method and system based on spatial indexing, and the method comprises the steps: collecting rail 3D original data, carrying out the preprocessing, and unifying the data into a unified rail coordinate system; dividing the preprocessed track 3D data into a plurality of regular grid units as space units according to spatial positions, extracting multi-dimensional features of data in each space unit, establishing a composite index for each space unit, and storing the composite index in a space-time database; after the acquired 3D data of the current track are subjected to identical processing, spatial position areas corresponding to the 3D data of the current track and historical data are quickly matched through a composite index, and a difference value of multi-dimensional features is calculated; dynamically adjusting the defect judgment threshold of each dimension according to the statistical distribution of the difference values, judging the defect type and severity of the region meeting the threshold, and generating a detection report. The problems that traditional 3D data processing is low in efficiency and long in comparison time consumption are solved.

The invention provides a real estate data visualization display method and system based on a GIS technology, and the method comprises the steps: carrying out the collection, cleaning, standardized mapping and entity alignment of multi-source real estate data, such as registration, transaction, planning approval, judicial, tax, sensing terminals, remote sensing images, and the like, building a space-time joint index with a space unit ID and a time slice ID as main keys, extracting a feature set and a target attribute set for modeling; constructing a structured causal model in the space-time database, and establishing a causal query index through a structure learning algorithm and parameter training; receiving the scene parameter set, analyzing the scene parameter set into an intervention instruction, executing a do-intervention operation on a causal model, collecting user interaction behaviors and external observation data, triggering online calibration of the model and a mapping rule when an error exceeds a threshold value, updating a causal query index and a mapping threshold value, and supporting version rollback; and a data, model, display and feedback closed-loop link is formed.

The invention discloses a park security multi-target path planning system and method, and belongs to the technical field of security, and the park security multi-target path planning system comprises a data layer, an algorithm layer and an application layer. The data layer constructs a park security space-time database and realizes multi-dimensional data fusion, and provides data support for security scheduling and risk assessment. The dynamic action range model of the algorithm layer can accurately determine the action range of the patrol route of the security and protection personnel and identify key points, warning areas and blind areas; a multi-target path optimization engine integrates the coverage degree, the time efficiency and the personnel fatigue degree to construct a three-dimensional dynamic optimization model, and patrol path dynamic adjustment is achieved by means of an improved NSGA-III algorithm. The application layer provides a six-dimensional analysis grid layer, and assists a command center and security patrol personnel in making decisions.

This invention discloses a deep learning method and system for generating trajectory data that supports heterogeneous data fusion, belonging to the fields of smart cities and smart transportation. The invention first acquires multi-source heterogeneous data such as GPS and traffic monitoring video, and performs a preprocessing process. Then, using a divide-and-conquer strategy, an attention-based RNN model predicts pedestrian trajectories, and a fusion model combining macroscopic traffic flow and GNN predicts vehicle trajectories, introducing multimodal features to generate a high-precision set of predicted trajectories. A spatiotemporal database is constructed to create a composite index for each trajectory. Finally, a digital twin system integrating these prediction results is built to achieve dynamic rendering and evolution. This invention effectively solves the problem of multi-source heterogeneous data fusion and significantly improves trajectory prediction accuracy.

The invention provides a data intelligent analysis method and system applied to carbon emission monitoring, and the method comprises the following steps: executing an NMF load decoupling algorithm to obtain a dynamic non-negative vector which comprises a resident coefficient and an industrial coefficient; according to the non-negative row vector and the space-time database, respectively calculating resident and industrial carbon emission intensities corresponding to the current time period, and updating the carbon emission intensities to the space-time database; calculating and generating future carbon emission predicted values corresponding to different future time periods according to the space-time database; and performing anomaly detection calculation according to the carbon emission intensity and a future carbon emission predicted value corresponding to the current time period. By adopting the method, the carbon emission intensity condition of the whole area can be intuitively and indirectly reflected in real time without tampering.

The application discloses a ground bearing capacity intelligent identification method and system based on a sensing network, relates to the technical field of geotechnical engineering detection, and comprises the following steps: a mixed perception network is constructed, and multi-source heterogeneous monitoring data of a foundation is acquired; a unified space-time datum of the multi-source heterogeneous monitoring data is established, and a space-time database is constructed; data in the space-time database is subjected to homogeneous fusion and heterogeneous fusion, comprehensive judgment of a ground bearing capacity state is carried out based on a fusion result, and a bearing capacity distribution field of the ground bearing capacity is generated. The application solves the technical problems in traditional ground bearing capacity identification, such as difficulty in giving consideration to coverage range and precision during data acquisition, lack of effective integration of multi-source data, and lagged response to abnormal conditions, achieves accurate identification of the ground bearing capacity, global coverage and long-term reliable optimization, and meets the technical effect of engineering safety monitoring and control of the foundation.

The application discloses a kind of based on machine learning's main girder erects elevation adjustment method, belong to elevation adjustment technical field, this method constructs space-time database by gathering the data of no cable section construction period;Adopt the mixed model of LSTM-Transformer to process time series data, extract long-period characteristic, output initial predicted value of elevation;Nonlinear variable such as the change rate of cable force, sunshine gradient and material age is analyzed using XG-Boost algorithm, and compensation factor is output;Fusion and generate joint predicted value;Construct reinforcement learning intelligent agent, with construction stage as state space, elevation adjustment quantity as action space, the deviation of predicted value and measured value and construction stability are minimized as reward function, and output adjustment instruction;According to driving hydraulic system, adjust erecting elevation, and real-time update database.The application improves the precision and stability of elevation adjustment by multi-dimensional data fusion and intelligent optimization.

The invention relates to the technical field of database construction, and particularly discloses a database construction management method for multi-source data. The method comprises the following steps of S1, acquiring multi-source data and performing preprocessing; S2, establishing an agricultural rural space-time database; S3, establishing a first construction management analysis model; S4, establishing a second construction management analysis model; according to the method, the multi-source data is acquired and preprocessed, the space-time database is constructed, and the construction management quality is evaluated by establishing the first construction management analysis model and the second construction management analysis model; according to the invention, integration and unified management of agricultural rural multi-source data are realized, data quality and application value are significantly improved, comprehensive and accurate data support is provided for agricultural analysis, and improvement of agricultural production efficiency, reduction of production risk and promotion of agricultural sustainable development are facilitated.

The invention discloses an airport aircraft taxiing efficiency redisk analysis method and system. The method comprises the following steps: performing space-time alignment on flight plan, aircraft position distribution, field monitoring radar and airport notification data to generate a fused space-time database; constructing a weighted directed graph model which takes the airport physical position as a node and the slidable path as an edge as an airport operation map; identifying flights with abnormal taxiing time, and mapping the actual track into an actual taxiing path sequence on the map through a hidden Markov map matching algorithm; calculating a theoretical optimal sliding path sequence by taking starting and ending nodes as constraints; the two paths are compared to position deviation nodes, and delay attribution is carried out in combination with node and edge attributes; and finally, carrying out visual display and parameterized simulation deduction. According to the method and the system provided by the invention, through a data fusion and map modeling technology, automatic and precise root cause tracing and strategy evaluation for the problem of low sliding efficiency are realized, and the intelligent level of airport ground operation management is improved.

This invention belongs to the field of stellar positioning and navigation technology, and particularly relates to an auxiliary altimetry method for stellar positioning. The method includes constructing a three-dimensional spatiotemporal coordinate system, collecting raw altimetry data, constructing a three-dimensional spatiotemporal database, analyzing the spatial and temporal correlations of measurement points, calculating the spatial and temporal correlations between different measurement point areas and surrounding measurement point areas in the latitudinal and meridional directions, identifying correlated data based on the correlation of the altimetry data, eliminating other invalid samples without correlation, and performing spatiotemporal weighted fusion based on the spatial correlation of adjacent measurement points and the temporal correlation at different times to obtain fused altimetry data. This application integrates the spatial correlation of measurement points during the altimetry data acquisition process, as well as the temporal correlation attributes of the same measurement point at different times, into the verification statistics of the altimetry data results, providing a data foundation for subsequent output verification of altimetry data and correction of erroneous data.

This invention discloses a method for extracting irrigation prescription maps based on BeiDou information of irrigation nodes. The method includes: acquiring a known large-scale water demand map and water shortage index map of the irrigation area under study; acquiring the BeiDou positioning information of the irrigation terminal nodes in the irrigation area under study and establishing a spatiotemporal database of all irrigation terminal nodes in the irrigation area under study; dividing the irrigation area under study into several partitions using fuzzy C-means clustering based on the spatiotemporal database, and using nodes with a membership degree greater than or equal to a set threshold in each partition as the initial boundary of the partition; smoothing the initial boundary of each partition to obtain the final partition boundary of each partition; determining the water demand relationship of the partitions based on the water demand map of the irrigation area under study; and calculating the gross irrigation quota of the partitions as the irrigation prescription map of the partitions based on the water demand map and water shortage index map of the irrigation area under study and the water demand relationship of the partitions.

The invention provides a natural resource element space-time process analysis method based on land type pattern spots. According to the method, a geographic vector pattern spot is used as a minimum space unit, spatial and temporal change types (newly increased, changed and killed) of land cover plaques are defined, and a corresponding algorithm is constructed. Firstly, multi-period basic geographic information data are collected and preprocessed, pattern spots are extracted and endowed with unique codes, and a topological structure and standardized codes are checked. And then realizing a spatio-temporal variation algorithm, judging the pattern spot variation type and calculating the area variation. And storing the data into a space-time database adopting an Oracle + ArcSDE engine, and carrying out space-time process analysis and visual expression, including a plurality of modes such as time axis animation and the like. The method is high in precision, avoids precision loss, is high in space-time complexity expression capability, is good in visualization effect, can provide scientific basis and decision support for the fields of natural resource management and the like, and is wide in application prospect.

This invention discloses a method and system for designing a spatiotemporal database of land value, applied in the fields of spatial geographic information data processing and image correction technology. The method involves acquiring spectral remote sensing image data; performing geographic coordinate matching on the spectral remote sensing image data to generate coordinate spectral remote sensing image data; digitally rasterizing the spectral remote sensing image data using a digital elevation model to generate point cloud data; performing 3D reconstruction based on the point cloud data to obtain a 3D terrain model; performing boundary region reconstruction processing on the 3D terrain model to obtain a reconstructed model; extracting ground features from the reconstructed model to obtain ground feature data; and constructing a spatiotemporal database by spatiotemporally indexing the ground feature data. This method can correct the brightness of edge defects of interval-type and linear ground features based on predicted brightness changes, significantly reducing grayscale distortion in 3D images and eliminating blurring caused by interference.

This invention relates to the field of database construction technology, specifically disclosing a database construction and management method for multi-source data. The method includes the following steps: S1: acquiring and preprocessing multi-source data; S2: establishing an agricultural and rural spatiotemporal database; S3: establishing a first construction management analysis model; S4: establishing a second construction management analysis model; S5: evaluating the construction management quality; and S6: establishing a construction management in-depth analysis model. This invention acquires and preprocesses multi-source data to construct a spatiotemporal database, and evaluates the construction management quality by establishing a first and a second construction management analysis model. This invention achieves the integration and unified management of multi-source agricultural and rural data, significantly improving data quality and application value, providing comprehensive and accurate data support for agricultural analysis, and helping to improve agricultural production efficiency, reduce production risks, and promote sustainable agricultural development.

The invention discloses a dynamic environmental effect grading method for a coastal concrete structure. The method comprises the following steps: firstly, constructing a coupling space-time database of the surface chloride ion concentration, the chloride ion diffusion coefficient and the steel bar corrosion rate; then, an improved space-time Kriging interpolation model (MSTK) is adopted to reconstruct space chloride ion distribution, and the weight coefficient of the space-time distance is optimized through a particle swarm; and meanwhile, a Hilbert dynamic probability interpolation model (HDPI) is combined with Bayesian optimization LSTM to realize time dimension continuous prediction, and the chloride ion concentration is decomposed into a deterministic trend, low-frequency environment fluctuation and high-frequency material fluctuation. Based on the interpolation result and the actually measured corrosion rate, high, medium and low corrosion risk grades are automatically divided through hierarchical clustering; and finally, quantizing the contribution degree of the detection points based on the diagonal elements of the prediction matrix, and optimizing the layout of the detection points. The method is high in precision (predicting R2gt; and the cost is low (detection points are reduced by 37.5%), and differentiated protection design of coastal electric power facilities is supported.

The invention discloses a cross-basin water transfer project ecological compensation quantification decision support platform. The decision support platform comprises a data acquisition module, a database analysis module, a model calculation module, a compensation demand quantification model and a decision support layer. The data acquisition module comprises ecological data, social economic data and spatial geographic data; the database analysis module comprises a relational database for storing structured data and a space-time database for managing geographic space data, and a data cleaning and standardization tool is arranged in the database analysis module; the model calculation module comprises ecological system service value evaluation, an InVEST model, an ecological footprint method and a hydrological response model. Compared with the prior art, the method has the advantages that a cross-basin water transfer project ecological compensation quantitative decision support platform is designed, so that the efficiency of cross-basin ecological compensation convenient decision support is improved.

The application discloses a natural language spatiotemporal retrieval method and system based on a large model, relates to the technical field of big data resource services, and comprises the following steps: fine-tuning a preset large language model according to a spatiotemporal retrieval problem and a preset prompt word engineering template; dividing a target management area into geographical space grids, each of which has a unique grid code; constructing an entity knowledge base classified according to different data categories based on a business spatiotemporal database and dynamic semantic labels; performing natural language intention analysis on user input based on the fine-tuned large language model, identifying spatial constraints, time constraints and business attribute constraints; converting the spatial constraints into filtering conditions; querying the data categories of corresponding key entities in the entity knowledge base; and outputting spatiotemporal retrieval results based on the fine-tuned large language model according to the business attribute constraints, the data categories, the spatial constraints and the filtering conditions. The application is free from professional dependence, improves the interactive experience and is highly flexible.

The application discloses a kind of discrete manufacturing industry data representation methods based on deep reinforcement learning, including collecting discrete manufacturing industry data, construct space-time database;Discrete manufacturing industry data is divided into discrete features and continuous features, construct data coupling coding network, and the encoding vector in coding network is converted into representation vector, and construct data representation model;The distinguishing degree of the class of representation data is quantified by clustering evaluation index;Different dimensions of clustering evaluation index are weighted as dynamic reward, construct deep reinforcement learning model, update the neural network parameters of deep reinforcement learning through the interaction between representation model and discrete manufacturing decision analysis system.The application can represent dynamically changing mixed data, based on deep reinforcement learning, the discrete manufacturing industry data is represented and enhanced, interacts with discrete industrial system through clustering evaluation index, and constantly feedback dynamic reward information, to obtain the optimal data representation form.

The invention relates to the technical field of numerical simulation, in particular to an ultrasonic flowmeter measurement error and reason analysis method, system and device and a medium. According to an actual flowmeter structure, an ultrasonic flowmeter digital simulation model is constructed, and an error injection module is introduced into the ultrasonic flowmeter digital simulation model; based on the constructed ultrasonic flowmeter digital simulation model, actual flowmeter data are collected in real time, the actual flowmeter data and virtual simulation data are adopted for collaborative sampling and correction, parameters of the ultrasonic flowmeter digital simulation model are dynamically adjusted through an optimization algorithm, and the ultrasonic flowmeter digital simulation model with the dynamically adjusted parameters is used for simulating the ultrasonic flowmeter. The error injection module is used for carrying out a disturbance experiment on a key error source, generating a space-time database, establishing a machine learning model, inputting the generated space-time database and cooperative correction data into the machine learning model, analyzing and outputting a root cause of an error, and improving the accuracy of error identification and the root cause traceability.

The invention discloses an irrigation prescription map partition extraction method based on irrigation node Beidou information. The method comprises the following steps: obtaining a known large-scale irrigation district water demand diagram to be researched and a known large-scale irrigation district water shortage index diagram to be researched in an irrigation district to be researched; beidou positioning information of end domain irrigation terminal nodes of an irrigation area to be researched is acquired, and a time-space database of all the end domain irrigation terminal nodes of the irrigation area to be researched is established; according to the space-time database, the to-be-researched irrigation area is divided into a plurality of partitions by adopting fuzzy C-means clustering, and nodes with the membership degree larger than or equal to a set threshold value in each partition serve as partition initial boundaries; smoothing the initial boundary of each partition to obtain the final partition boundary of each partition; determining the water demand relationship of the subareas based on the water demand diagram of the irrigation area to be researched; according to the water demand diagram and the water shortage index diagram of the irrigation area to be researched and the water demand relation of the subareas, the gross irrigation quota of the subareas is calculated to serve as the irrigation prescription diagram of the subareas.

The invention discloses an intelligent foundation bearing capacity identification method and system based on a sensing network, and relates to the technical field of geotechnical engineering detection.The method comprises the steps that a mixed sensing network is constructed, and multi-source heterogeneous monitoring data of a foundation are obtained; establishing a unified space-time reference of the multi-source heterogeneous monitoring data, and constructing a space-time database; and carrying out homogeneous fusion and heterogeneous fusion on the data in the time-space database, carrying out comprehensive judgment on a foundation bearing capacity state based on a fusion result, and generating a bearing capacity distribution field of the foundation bearing capacity. The technical problems that in traditional foundation bearing capacity recognition, coverage range and precision are difficult to consider in data collection, effective integration of multi-source data is lacked, and response to abnormal conditions is delayed are solved, and the technical effects of accurate recognition, global coverage and long-term reliable optimization of the foundation bearing capacity and meeting of engineering for foundation safety monitoring and management and control are achieved.

The invention discloses a city update three-dimensional model lightweight and semantic segmentation fusion method, and relates to the technical field of three-dimensional geographic information processing, and the method comprises the steps: constructing a dual-branch multi-task deep learning network comprising a shared feature extraction layer, a lightweight branch and a semantic segmentation branch through the standardization preprocessing of multi-source data; feature extraction is enhanced by adopting dynamic graph convolution and an attention mechanism, and joint training is realized through a composite dynamic loss function; semantic tags are optimized through post-processing of conditional random fields, incremental updating is achieved based on a gridded space-time database, and multi-granularity intelligent query and cross-platform docking are supported. According to the method, collaborative optimization of light weight and semantic segmentation is achieved, the local updating efficiency is improved by 10 times, the key element labeling accuracy rate is larger than or equal to 95%, the complex query response time is smaller than or equal to 2 seconds, and efficient and accurate technical support is provided for city updating full-life-cycle management.