114 results about "Regional model" patented technology

The invention discloses an AI-NWP three-dimensional closed-loop bidirectional dynamic feedback coupling method for extreme rainfall event area simulation and a program product, and belongs to the technical field of meteorological numerical simulation and artificial intelligence fusion. According to the method, a mid-term forecast field is generated based on an AI global weather prediction model, an analysis field is constructed by assimilating multi-source observation data, and a high-resolution NWP region mode is driven to perform rolling simulation. Furthermore, a space-time residual field is constructed by using the difference between an NWP region simulation result and live data, a residual learning model is trained, an AI model prediction structure is fed back and corrected, and dynamic weight adjustment updating of the AI prediction model is realized. A closed-loop two-way feedback system among AI output, NWP simulation and residual evaluation is integrally formed, the space structure reduction capability and the area positioning precision of a medium-term heavy rainfall event are effectively improved, the continuous predictability and the simulation credibility of an extreme weather process are remarkably enhanced, and the method has good stability, universality and engineering expansion value.

The invention relates to the technical field of topographic survey, in particular to an overwater and underwater topographic survey method based on cooperation of an unmanned ship and an unmanned aerial vehicle. The method comprises the following steps: acquiring platform parameter data of an unmanned ship and an unmanned aerial vehicle; according to the platform parameter data, isomerization grouping based on the residual electric quantity, the communication signal strength, the cruising ability and the measurement precision is carried out, a collaborative formation scheme is determined according to a preset water area measurement task sequence, and collaborative parameter data is obtained; acquiring environment data of a measurement area; and performing regional division modeling based on water area topographic features and environmental complexity according to the environmental data of the measurement region, and applying the collaborative parameter data to a regional model to obtain a topological model of the measurement region. The unmanned ship is used as a mobile charging platform, real-time electric quantity monitoring and docking charging mechanisms are combined, the endurance guarantee of the unmanned aerial vehicle is realized through energy supply strategy optimization, and the continuity and stability of tasks are ensured.

The invention discloses an intelligent resource optimization configuration method and system in a cloud computing environment. The method comprises the steps of collecting water conservancy station data in real time, uploading the data to a cloud end, performing dynamic labeling, and generating a resource label. Physical resources are virtualized into logic resources, and a resource partition mapping model is constructed at the cloud. And packaging the models, and establishing a collaborative optimization mechanism among the multi-region models. Based on cloud resource labels and virtualized partition information, in combination with real-time environment data, supply and demand matching is realized by adopting an optimization algorithm, and a resource allocation strategy is dynamically adjusted. According to the method, the refinement degree of resource management is effectively improved, and the flexibility and the real-time performance of resource allocation are improved. The flexibility and the coverage range of resource scheduling are improved, the cross-regional resource allocation requirement can be met, and therefore resource waste is remarkably reduced. Resource allocation is more intelligent, the scheduling cost caused by global resource imbalance is effectively reduced, and the resource utilization rate and allocation efficiency are remarkably improved.

The invention provides a photovoltaic power station output power prediction method, a terminal and a medium. The method comprises the steps that real-time meteorological data are collected and preprocessed; a self-organizing mapping network is constructed, different environment modes are identified and classified under the unsupervised condition, and meanwhile photovoltaic power generation power is predicted; designing a region model of a cross-region knowledge migration algorithm based on a meta-learning model, and carrying out knowledge migration among different regions; predicting the category of the current meteorological environment according to new meteorological data recollected in real time, then calculating the prediction error of each regional model under the new environmental data, and selecting the regional model with the minimum error to carry out power prediction; and constructing a personalized recommendation strategy model according to the predicted power to perform real-time feedback and optimization on power prediction. According to the method, strong adaptive capacity and generalization capacity are provided, and more accurate decision support is provided for intelligent scheduling and power grid management of photovoltaic power generation.

The invention discloses a federal learning-based cross-device brain grain model training system. The system comprises a terminal device layer, an edge coordination layer and a central server layer, the terminal equipment layer is responsible for collecting original electroencephalogram data and executing local lightweight training; the edge coordination layer is used for aggregating model updates of a plurality of terminal devices in the jurisdiction to generate a regionalized sub-model; according to the central server layer, a central server performs secondary optimization on each regional model to generate a global unified model; the encrypted global model and strategy configuration are pushed to all edge coordination layers; and the edge coordination layer pushes the optimized model and strategy to terminal equipment. Through a three-layer federated architecture and multi-dimensional optimization, the system efficiency is remarkably improved: the terminal equipment only uploads an encrypted model increment, and source data protection is realized in combination with differential privacy and homomorphic encryption; federal transfer learning shortens the cold start time of new equipment.

The invention discloses a hydraulic engineering operation and maintenance data monitoring method and system based on digital twinning, relates to the technical field of data analysis, and comprises the steps of defining multi-physics field coupling related to a three-dimensional structure region model, and providing a basis for subsequent model grid density setting. And performing clustering analysis on the region according to the geological features and the structural features on the three-dimensional structural region model. The grid density grade of each type of region is evaluated based on the grid attribute, the multi-physics field coupling strength and the uncertainty parameter, a flexible grid density grade determination mode is carried out on different regions, the reliability and precision of model gridding processing are improved, and the uncertainty condition can be captured more accurately. The operation and maintenance event and the event state of the water conservancy project are monitored through the multi-physics-field multi-grid digital model, current time monitoring and future time prediction of the operation and maintenance event are achieved, the accuracy and adaptability of water conservancy project operation and maintenance monitoring are improved, and the complex and changeable monitoring requirements of the water conservancy project are met.

The invention discloses an unmanned vehicle indoor and outdoor seamless positioning method based on area identification, and belongs to the field of automatic driving. The method comprises the following steps: firstly, designing a joint prediction model of a support vector machine (PSO-SVM) and a hidden Markov model (HMM) based on particle swarm optimization, and using acquired GNSS environment characteristic data as model input to obtain a region identification result; then, three region models (an indoor model, an indoor and outdoor transition region model and an indoor model) are defined to describe the environment position of the vehicle, and corresponding measurement information is selected according to a region recognition result; and finally, the switching probability of the three environment sub-models in the IMM algorithm is updated by using the area switching probability, so that the accuracy and the positioning precision of indoor and outdoor seamless positioning environment area identification of the unmanned vehicle are improved. Through experimental verification, compared with a single sensor and an IMMEKF algorithm, the method provided by the invention can better adapt to environmental changes, so that the accuracy and robustness of unmanned vehicle positioning are improved.

The invention belongs to the technical field of air quality monitoring, and relates to a regional automatic identification method and system for urban air quality influenced by sand and dust weather, and the method comprises the steps: 1) data collection; 2) region division: dividing a nationwide city into a plurality of regions; 3) data preprocessing; 4) regional model training: for each region, training each sub sand and dust weather identification model by using a training set; 5) regional model effect evaluation: performing effect evaluation on each sub-sand and dust weather identification model by using an evaluation index; 6) regional model integration: for each region, integrating a plurality of sub-sand and dust weather identification models of the region into a final sand and dust weather identification model of the region; and 7) performing regional automatic identification: automatically realizing automatic identification of the influence of the dust on the urban air quality by using the final dust weather identification model of the region, thereby improving the credibility of the identification result, and providing a scientific basis for precise environment management and control.

The invention discloses a gradient random porous structure design method based on a growth function and RBF (Radial Basis Function) interpolation, which relates to the technical field of geometric modeling and computer graphics, and comprises the following steps: collecting and preprocessing Voronoi seed point coordinates and implicit boundary parameters, extracting density field and gradient parameters, and calculating the density field and gradient parameters; mixing different pore forms based on a beta growth function to generate a transition region model, constructing a scalar field in combination with RBF interpolation, fusing the model and a density field to generate global gradient distribution, optimizing iteration to verify mechanical properties and correcting parameters, mapping the parameters into three-dimensional grid data, rendering a pore density cloud chart and a thermodynamic chart, and marking key parameters. By combining the beta growth function and the RBF interpolation, the problem of strong coupling of gradient control and seed point distribution and the problem of rigid transition of a supporting structure body are solved, flexible design of smooth gradual change and complex gradient of the pore form is achieved, mechanical properties are verified through optimization iteration, a three-dimensional visual model is generated, and the method is suitable for large-scale popularization and application. And the geometric continuity and controllability of the porous structure design are obviously improved.

The invention relates to the technical field of field layout modeling processing, in particular to a prefabricated field layout planning modeling method and system based on digital aided design. The method comprises the following steps: screening out a plurality of construction feature points of each three-dimensional point cloud model; obtaining a plurality of model areas at a real-time moment; according to the number of construction feature points in each model area at the real-time moment, the position distribution of different points and the initial voxel size, the adjustment voxel size of each model area at the real-time moment is obtained, and a three-dimensional adjustment point cloud model at the real-time moment is obtained; obtaining a construction offset direction vector and a construction offset of each unregistered point in the initial three-dimensional point cloud model, and obtaining an unregistered point area model of the initial three-dimensional point cloud model; and splicing the three-dimensional adjustment point cloud model at the real-time moment and the non-registered point region model to obtain a prefabricated field layout model at the next moment. According to the method, the accuracy of planning and modeling is improved by obtaining the accurate deviation trend in the construction process.

The invention discloses an intelligent commanding and testing integrated system for hydropower station debugging, and relates to the technical field of hydropower station equipment debugging and operation testing. Based on the physical connection relation and debugging parameters of the water turbine and the speed regulator, a three-dimensional function sub-region model is constructed, a multi-test-mode working condition database is established, through the effect recognition module, the system can construct and judge key indexes in each test mode, and the test efficiency is improved. Whether the problems of resonance effect, guide vane hysteresis, water attack abnormity and disturbance deviation exist or not is automatically recognized, a corresponding early warning instruction is sent out, and a closed-loop link from'phenomenon recognition 'to'reason tracing' is achieved. According to the method, whether a single test is qualified or not can be judged, the collaborative imbalance risk coefficient in the ith test mode can be further calculated, the risk level between collaborative effects is evaluated according to the collaborative imbalance risk coefficient, and intelligent adjustment of a test strategy is supported.

The invention relates to the technical field of data processing, in particular to a software design method and system based on cloud service, and the method comprises the steps: obtaining home space information and an element database; obtaining a style attribute selected by a user and a function priority set for each hard decoration element, generating a basic component according to the style attribute, a layout relationship of each hard decoration element and a node tree where the hard decoration element is located, and sequentially combining the basic components corresponding to each hard decoration element according to the function priority to generate a component tree; all nodes in the component tree are traversed in sequence, behavior scores fed back by a user to the traversed nodes are obtained, node matching is carried out according to the behavior scores, the matched nodes are combined and rendered, and a region model is generated; generating a global model based on the root nodes in each node tree, and splicing the region models selected by the user based on the global model to obtain a soft installation design model; according to the invention, the intelligent level of soft installation design can be improved, and accurate matching of a design scheme and a user demand is realized.

The invention discloses a boiler multi-region CFD-AI modeling method and system based on federated learning, and belongs to the technical field of boiler monitoring, and the modeling method comprises the following specific steps: S1, simulating the heat transfer and fluid behaviors of a boiler, building multi-region CFD grid division, and collecting and preprocessing the dynamic operation data of each region from the actual operation of the boiler; s2, extracting features of the preprocessed dynamic operation data, constructing a boiler multi-region CFD-AI model according to the extracted features, and globally optimizing the boiler multi-region CFD-AI model; s3, dynamically adjusting the structure and parameters of the boiler multi-region CFD-AI model after global optimization, and enhancing the key region decision of the boiler multi-region CFD-AI model; according to the method, the efficiency of model training can be improved, a large number of communication resources are saved, the privacy disclosure risk is reduced, the cross-regional adaptability of the model is promoted, the performance of the model under the complex working condition is remarkably improved, the multi-objective optimization problem can be effectively solved, and the performance of the boiler in different regions and working conditions can be more accurately predicted and controlled.

The invention relates to the technical field of population prediction, and discloses a population prediction method and device based on combination of an ARIMA model and a queue element method. The method comprises the steps of obtaining historical migration population data of a to-be-predicted region, determining a difference order through ADF test, preliminarily determining the range of an autoregression order and a moving average order by using an autocorrelation function and a partial autocorrelation function, and constructing an ARIMA model; the optimized ARIMA model is adopted to fit historical migration population data, a migration population prediction result is obtained, and migration population data of each age group in the target prediction time period is output in combination with age structure data of floating population of the reference year; taking the migration population data of each age group as input parameters of migration population in a queue element method; according to a preset sum growth rate, selecting a western mode of a numeral-demann region model life table as a death model; and setting the average life expectancy according to a preset birth ratio, and realizing permanent resident population prediction of the to-be-predicted region.

This invention relates to the field of airspace management technology, specifically disclosing a method and system for dynamically updating airport airspace obstacle data. The method includes: acquiring regional benchmark data based on a first acquisition precision under standard operating conditions to construct a benchmark regional model; acquiring real-time regional data periodically based on a second acquisition precision under real-time operating conditions to construct a real-time regional model with time labels; identifying the real-time regional model based on an adaptive detection parameter set, locating abnormal candidate areas, and simultaneously locating abnormal obstacles; marking the location results, and visually displaying the real-time regional model. This invention acquires relevant airport information through an acquisition device with data transmission capabilities, performs 3D modeling of the airport, and, combined with a pre-acquired standard model, quickly determines abnormal areas and objects at different times, thus providing intuitive feedback to management personnel. The method boasts extremely high identification efficiency and readability.

The embodiment of the invention provides an optimal scheduling method and system for a virtual power plant based on a cloud and fog edge architecture, a computer readable storage medium and a computer program product. The method comprises the following steps: edge nodes realize local optimization scheduling of end side equipment, fog nodes realize coordinated aggregation of a plurality of edge nodes in a jurisdiction area, and a cloud platform realizes cross-area global optimization and knowledge sharing. According to the scheme, a cloud platform, fog node and edge node three-level collaborative optimal scheduling architecture is constructed, and local optimization capability, fog node area coordination capability and global optimization capability are given to the edge nodes, so that the inherent single-point fault risk of traditional centralized scheduling is overcome, and when the communication of the cloud platform is interrupted or invalid, the cloud platform can be effectively optimized. Each fog node can guide autonomous operation in the jurisdiction based on the latest regional model; and when the fog node hierarchy fails, the edge node can still independently carry out local optimization, so that the robustness and reliability of the system under the abnormal working condition are enhanced.

This invention discloses a method and network model for reconstructing missing soil moisture data based on cross-attention, belonging to the fields of remote sensing image processing and artificial intelligence. It involves collecting SMAP, ERA5-Land soil moisture products, and various auxiliary prediction factor data. The auxiliary data is resampled to 9km and preprocessed using multi-source data. A CASTNet model is then constructed. After inputting the preprocessed data into the model, feature extraction, refinement, and enhancement are sequentially performed through modules for spatiotemporal automatic completion, multi-factor feature refinement, and multi-scale attention enhancement. Finally, the model is trained based on a combined loss function composed of global and local losses, and the trained model is used to fill in the missing areas of soil moisture remote sensing data. The CASTNet model contains three cascaded core modules, each implementing preliminary spatiotemporal feature completion, multi-factor feature fusion refinement, and multi-scale feature enhancement, respectively. This invention fully leverages the complementary information of multi-source data, improving the accuracy of soil moisture data reconstruction and ensuring the spatiotemporal continuity and integrity of the reconstructed data.

The embodiment of the invention discloses a virtual area model processing method and a related device, which can be applied to scenes such as the map field, the cloud technology, artificial intelligence, intelligent traffic, auxiliary driving, digital human, virtual human, games, virtual reality, augmented reality and the like. According to the method, the virtual region is constructed by the region image used for describing the geographic entity region, so that the problem of data misalignment is avoided, the accuracy of subsequently constructing the virtual region model is improved, the complexity of the construction process is reduced, the method is not limited by map vector data, and the application range is expanded. In addition, regional images are identified through multiple image identification models, and semantic features describing all the sub-regions are obtained. And performing three-dimensional reconstruction according to the multiple semantic features to obtain a virtual region model for the geographic entity region. Therefore, three-dimensional reconstruction is carried out based on multiple rich semantic features, and the accuracy of the obtained virtual region model is higher.

The invention relates to the field of artificial intelligence, and provides a transformer state visualization method based on a BIM model, and the method comprises the steps: obtaining a region model of a region where a target transformer is located according to a preset BIM model; determining a first loss parameter of the target transformer according to the regional model; determining a second loss parameter of the target transformer according to the load of the target transformer; and calculating life information of the target transformer according to the first loss parameter and the second loss parameter, and displaying the target transformer in the BIM model according to the life information. The service life information of the transformer is displayed in the BIM model, so that the intelligent degree of applying the BIM model to power equipment management and the intuitive degree of the transformer state in the BIM model are improved.

The application discloses a wind field reconstruction and prediction method and device, electronic equipment and storage medium, and belongs to the technical field of wind field data processing. The method comprises: uniformly encoding historical wind speed data and real-time wind speed data of multiple height layers, so that the model can mine the vertical coupling linkage relationship between wind fields at different height layers and the time evolution law of the wind field; geometric prior information and observation masks are introduced, so that the model can complete wind field flow constraint analysis combined with regional building layout and divide the measured data area and the blank reconstruction area; when the model finally completes global wind field deduction based on multi-dimensional fusion features, not only the accuracy of the wind field reconstruction result is improved, but also multi-step fine wind field prediction is realized, the disadvantages of single wind field deduction dimension, lack of environmental constraints and insufficient prediction accuracy of the traditional technology are solved, and the actual needs of fine wind field reconstruction and short-term accurate wind field prediction in a complex building group scene are met.

The present application relates to the technical field of hub polishing robots, and particularly relates to a hub polishing robot trajectory planning method based on 3D vision curve technology; by comparing an actual 3D curved surface graph with a preset 3D curved surface graph, a to-be-polished region model graph of each to-be-polished region in a target hub is generated; according to the to-be-polished region model graph of each to-be-polished region, optimal polishing parameters for polishing processing of each to-be-polished region in the target hub are obtained by pairing in a database; according to the optimal polishing parameters for polishing processing of each to-be-polished region in the target hub, optimal polishing trajectories of a polishing robot are planned; and the optimal polishing trajectories of the polishing robot and the optimal polishing parameters for polishing processing of each to-be-polished region in the target hub are sent to a control terminal. More accurate, self-adaptive, efficient and consistent polishing effects can be achieved, and production efficiency and product quality are improved.

The present invention relates to the field of topographic surveying technology, and in particular to a method for measuring topography above and below the water based on the collaboration of an unmanned vessel and an unmanned aerial vehicle. The method comprises the following steps: obtaining platform parameter data of the unmanned vessel and the unmanned aerial vehicle; performing heterogeneous grouping based on the remaining power, communication signal strength, endurance, and measurement accuracy according to the platform parameter data, and determining a collaborative formation plan according to a preset water area measurement task sequence to obtain collaborative parameter data; obtaining measurement area environmental data; performing regional division modeling based on the water area terrain characteristics and environmental complexity according to the measurement area environmental data, and applying the collaborative parameter data to the regional model to obtain a measurement area topology model. The present invention uses the unmanned vessel as a mobile charging platform, combines real-time power monitoring with a docking charging mechanism, and optimizes the energy replenishment strategy to achieve endurance guarantee for the unmanned aerial vehicle, thereby ensuring the continuity and stability of the mission.

The invention discloses a power grid park modeling method and system based on VR data. The method comprises the steps of obtaining image data sent by a plurality of VR devices in a target park area; generating a plurality of sub-region models corresponding to the target park region according to the device position of each VR device and the image data; correcting the intersection part of the plurality of sub-region models according to a preset intersection correction rule to obtain a plurality of corrected sub-region models; and generating a park three-dimensional model corresponding to the target park area based on a model union set algorithm according to all the corrected sub-area models. It can be seen that the more accurate park three-dimensional model can be generated in combination with the VR equipment data in the park, the park modeling efficiency and precision are improved, and the labor cost is reduced.

The invention discloses a port mobile machinery navigation method and system based on environmental perception, and relates to the technical field of navigation, a plurality of mechanical devices needing navigation are determined as a plurality of target devices, and projection chains are correspondingly arranged between the plurality of target devices and a port navigation model; setting a plurality of measurement points corresponding to the target equipment, and performing real-time construction on the environment information to obtain a regional model; and fusing the region model into the port navigation model in real time based on the projection chain, and performing real-time adjustment on navigation routes of all target equipment according to the port navigation model. According to the method, the port environment can be updated timely and clearly according to the surrounding environment information collected by the target equipment through the projection chain, and the route of the multiple equipment can be adjusted in real time according to the overall environment of the port, so that the situation that the cargo transfer of the port is affected due to the congestion of mobile machinery caused by an information error problem is avoided; and the orderly operation of the port equipment can be better ensured.

The application provides a three-dimensional terrain geological modeling method and device based on material point method simulation, and relates to the technical field of geotechnical engineering numerical simulation, and comprises the following steps: S1, extracting elevation data of a three-dimensional terrain geological region model; S2, converting the elevation data into three different type model component data and processing; S3, based on the generated model component data, discretizing the model into material points to generate a material point model; S4, importing the generated material point model into a material point method calculation program to perform material point method simulation, and obtaining large deformation of the geology according to the material point method simulation; the application effectively solves the problem of large deformation simulation modeling caused by complex and diverse terrain and landform, is efficient, stable, simple and feasible, significantly improves the modeling efficiency of the material point method and the quality of the generated material point model, and effectively promotes the application of the material point method simulation in the field of geotechnological engineering numerical simulation.

The application discloses a cross-region crop remote sensing image processing method, device and equipment fused with geographic information, relates to the technical field of crop data processing and remote sensing cross technology, constructs a generalization data set of latitude and terrain joint partitioning through sample geographic information data and sample satellite remote sensing image data, obtains a geographic prototype library corresponding to a geographic unit, then trains a crop classification prediction model based on the geographic prototype library, and performs crop classification prediction by using the crop classification prediction model, which can effectively capture the combined influence of geographic multidimensional factors on crop growth, solve the fundamental problem of poor generalization of the cross-region model, improve the cross-geographic processing capability of the crop classification prediction model, and realize high-precision, high-robustness and low-cost crop remote sensing classification.

The invention discloses a regional high-resolution atmosphere reanalysis data construction method based on dynamic downscaling, and aims to improve the depiction capability of medium and small-scale weather processes. According to the method, global reanalysis data is used as a background field, multi-source observation data and high-resolution land utilization information are fused, multi-time and short-period simulation is executed by constructing a multi-layer nested WRF region mode and combining an assimilation strategy combining analysis Nudging and observation Nudging, and a reanalysis data set with hour-level and kilometer-level resolution is constructed step by step. Through comparison and verification of multi-source observation data, the constructed data set is obviously superior to ERA5 in the aspects of ground meteorological elements and vertical profile structure reduction, and has higher regional adaptability and simulation precision. The method is suitable for a plurality of high-resolution meteorological application scenes such as regional artificial intelligence meteorological large models, urban fine meteorological services, high-influence weather disaster early warning, regional climate change evaluation, energy meteorological resource evaluation and the like.

The present invention discloses a boiler multi-region CFD-AI modeling method and system based on federated learning, which belongs to the field of boiler monitoring technology. The specific steps of the modeling method are as follows: S1, simulating the heat transfer and fluid behavior of the boiler, establishing a multi-region CFD grid division, and collecting and preprocessing the dynamic operation data of each region from the actual operation of the boiler; S2, extracting the features of the preprocessed dynamic operation data, constructing a boiler multi-region CFD-AI model based on the extracted features, and globally optimizing the boiler multi-region CFD-AI model; S3, dynamically adjusting the structure and parameters of the globally optimized boiler multi-region CFD-AI model, and enhancing the key area decision-making of the boiler multi-region CFD-AI model; the present invention can improve the efficiency of model training, save a lot of communication resources, reduce the risk of privacy leakage, and promote the cross-region adaptability of the model, significantly improve the performance of the model under complex working conditions, can effectively handle multi-objective optimization problems, and can more accurately predict and control the performance of the boiler in different regions and working conditions.

The techniques herein include using an input context to determine a suggested action and / or cluster. Explanations may also be determined and returned along with the suggested action. The explanations may include (i) one or more most similar cases to the suggested case (e.g., the case associated with the suggested action) and, optionally, a conviction score for each nearby cases; (ii) action probabilities, (iii) excluding cases and distances, (iv) archetype and / or counterfactual cases for the suggested action; (v) feature residuals; (vi) regional model complexity; (vii) fractional dimensionality; (viii) prediction conviction; (ix) feature prediction contribution; and / or other measures such as the ones discussed herein, including certainty. The explanation data may be used to determine whether to perform a suggested action.

The application provides a meteorological disaster high-resolution regional mode prediction system and method, relates to the technical field of meteorological disaster prediction, and comprises the following steps: performing scale reduction processing on target data of a region to be processed to obtain scale-reduced target data, and performing fusion on multi-source observation data to obtain grid fusion data; according to a preset period, different target observation data are subjected to physical quantity correction, error estimation, quality control and format conversion processing by using specific technical methods, and the processed target observation data are subjected to hourly rapid cycle assimilation processing to obtain target assimilation analysis data; and the scale-reduced target data, the target assimilation analysis data and localized data are used to simulate the operation of a localized numerical prediction mode, meteorological disaster numerical prediction results of different scale requirements are generated, and the meteorological disaster numerical prediction results are corrected by using corresponding scale grid fusion data, so that the technical problem of low prediction accuracy of an existing meteorological disaster system is solved.