306 results about "Data assimilation" patented technology

The invention relates to the technical field of intelligent early warning of digital twinning, and particularly discloses a slope multi-physics field fusion early warning decision-making system based on digital twinning, which is characterized in that physical monitoring data representing the macroscopic state of a slope and microscopic physical response signals reflecting internal damage evolution are synchronously acquired through a multi-modal data sensing module; space-time alignment, standardization and cross-modal fusion analysis are carried out through a damage eigenstate extraction module, and a unique eigendamage variable for quantitatively representing the real-time degradation degree of the material strength is interpreted; the twinborn self-evolution module takes the variable as a core observed quantity, and drives parameters and states of a slope mechanical model to be cooperatively and dynamically updated by adopting a data assimilation method, so that high-fidelity tracking of a digital model on physical reality is realized; and the prospective early warning decision module deduces a future spatio-temporal evolution path of the material strength parameters based on the calibrated model, and realizes graded early warning and intelligent decision support by combining Monte Carlo simulation and quantification of the instability risk probability.

The invention relates to the technical field of tunnel engineering intelligent construction, and discloses a hard rock TBM shield tunneling machine auxiliary tunneling parameter dynamic adaptation regulation and control system, which comprises a high-precision multi-source sensing network, a high-precision multi-source sensing network, a high-precision multi-source sensing network, a high-precision multi-source sensing network, a high-precision multi-source sensing network, a high-precision multi-source sensing network and a high-precision multi-source sensing network, the TBM-geological environment digital twin predicts the tunneling short-term trend based on the high-fidelity physical simulation and data assimilation technology; and the multi-modal deep learning collaborative decision-making module deeply fuses real-time and prediction data and generates an optimal parameter solution set through a network trade-off tunneling multi-conflict target based on Pareto optimization. And the system executes a decision and forms closed-loop feedback through a parameter dynamic adaptation and adaptive learning module, and continuously optimizes a self model and a knowledge base. According to the method, passive response of TBM tunneling is converted into active pre-judgment, the decision accuracy, the construction safety and the comprehensive tunneling efficiency under the complex working condition are remarkably improved, and the method has the sustainable evolution capacity.

The invention relates to the technical field of oil and gas field development, and discloses an ultra-large type true triaxial hydraulic fracturing fracture evolution path prediction method and system.The prediction method comprises the steps that a coupling geomechanical model integrating microscopic, macroscopic and wellbore flow scales is constructed; initializing the model and performing crack initial expansion simulation; collecting construction data in real time, and dynamically optimizing model parameters through a data assimilation algorithm; performing fracture evolution advanced prediction by using the updated model; and generating an optimization decision based on the prediction result and feeding back to the construction site. According to the method, fusion of a multi-scale physical mechanism and real-time dynamic prediction is realized, and accurate prediction and active control can be performed on crack expansion under the ultra-large true triaxial condition.

The invention belongs to the crossing field of environmental engineering and hydrology and hydrodynamics, and particularly relates to a traceability method based on coupling hydrodynamics and a pollutant degradation equation, which comprises the following steps: firstly, acquiring and preprocessing multi-source heterogeneous monitoring data, then selecting a one-dimensional Saint-Venant equation or a two-dimensional shallow water equation according to a water body form to solve a hydrodynamic field, and finally, determining the hydrodynamic field. A multi-component convection-diffusion-reaction equation is coupled to simulate pollutant migration and transformation; an LSTM module is introduced to identify suspected pollution events, pollution source parameters are inverted through a two-channel framework of PDE constraint optimization and Bayesian inference, uncertainty is quantified, model parameters are updated online in combination with data assimilation, and finally the uncertainty is quantified and verified. The method considers traceability precision, efficiency and compliance, supports multiple water bodies and multiple data sources, and is suitable for complex water body pollution traceability.

The invention discloses a unit area irrigation water consumption estimation method and system based on multi-source data fusion, and relates to the field of irrigation water consumption estimation. Multi-source heterogeneous data such as remote sensing images, meteorological data, soil humidity and crop growth information are fused, and observation and model errors are reduced through data assimilation; estimating crop evapotranspiration under an energy balance principle; estimating daily-scale irrigation water consumption by adopting a water balance method; a crop actual irrigation area identification model based on deep learning is introduced, actual irrigation areas of different crop types are effectively distinguished, and the influence of temporal and spatial changes on irrigation efficiency is dynamically reflected. According to the method, high-precision and rapid estimation of the comprehensive unit area irrigation water consumption of a single crop and multiple crops on different regional scales can be realized, the timeliness and accuracy of irrigation water quota evaluation are remarkably improved, and the method has strong adaptability to climate change, soil difference and crop diversity and is suitable for popularization and application. Powerful technical support is provided for precise irrigation and intelligent water conservancy management.

The invention discloses a sand and dust vertical flux high-precision inversion method based on a laser radar. The method comprises the following steps: acquiring multi-wavelength back scattering and polarization information by using a 355 nm, 532 nm and 1064 nm three-wavelength polarization laser radar system; constructing a five-dimensional optical feature vector space, and combining a support vector machine classifier to realize automatic identification of dust particles; a variational data assimilation technology is adopted to fuse radar observation and numerical forecasting information to invert a three-dimensional wind field; inverting sand and dust mass concentration vertical distribution based on the corrected particle spectrum distribution model; flux calculation and uncertainty quantization are realized through adaptive weighted fusion and a Monte Carlo method; the method has the characteristics of high temporal-spatial resolution, high precision, strong adaptability and the like, and can be widely applied to the fields of weather forecast, environment monitoring, climate research and the like.

The invention relates to the technical field of refrigeration house energy-saving control and digital twinning, in particular to a refrigeration house fan energy-saving operation method and system, and the method achieves the whole-field temperature deduction of a sensor-free area by constructing a CFD reference model and simulating and predicting the three-dimensional transient airflow and temperature distribution in a refrigeration house. Model parameters are calibrated through measured data, and the order of the model is reduced by adopting intrinsic orthogonal decomposition and Galerkin projection, so that the calculation amount is remarkably reduced, and online optimization is supported. And in combination with a data assimilation algorithm, the prediction result of the reduced-order model is continuously corrected, and the adaptability to environment change and system aging is improved. The optimization control stage takes minimization of the total energy consumption of a fan as a target, solves an optimal fan control sequence under the hard constraint that the whole-field temperature does not exceed the cargo safety upper limit and does not exceed the cargo safety lower limit, and adopts a rolling time domain mode for execution to realize collaborative optimization of safety and energy conservation.

The invention belongs to the technical field of marine environment prediction, and discloses a neural network-based sea wave spectrum intelligent correction method and system. The method comprises the following steps: setting a target sea area and acquiring environmental data; performing numerical calculation on the sea wave spectrum in the target sea area through a sea wave numerical mode; a sea wave spectrum correction model comprising a numerical calculation module, an actual measurement processing module and an intelligent correction module is constructed through a deep learning method; and carrying out precision verification and adaptability evaluation on the constructed sea wave spectrum correction model. According to the model, the problem of errors caused by physical process simplification in the simulation process of a traditional method is solved. Meanwhile, the problem that an existing data assimilation method is limited by observation data is solved. According to the method, the simulation precision of the sea wave mode can be effectively improved, and the method has good sea area adaptability.

The invention discloses a regional wind power plant group multi-scale meteorological data assimilation and space-time reconstruction method, and aims to solve the problems that existing wind power plant meteorological prediction precision is insufficient, multi-source data fusion is difficult and the like. According to the method, effective assimilation of multi-source heterogeneous meteorological data is realized by constructing a deep convolution variational auto-encoder, space-time reconstruction of a high-resolution three-dimensional wind field is realized in combination with a heterogeneous graph network and a Shenchang differential equation technology, and physical constraint conditions are introduced to ensure physical consistency of output results. Experimental results show that the method can significantly improve the meteorological prediction precision of the wind power plant, and provides reliable meteorological data support for fine operation management of the wind power plant group.

The invention discloses an urban flood prediction method based on a dual-drive urban flood model, and the method comprises the steps: carrying out the early-stage preparation of model construction, constructing an urban flood hydrological and hydrodynamic coupling model frame, constructing a deep learning model frame based on a GBDT algorithm, and carrying out the prediction of the urban flood. Assimilation of predicted values and measured data of a hydrological hydrodynamic model and a deep learning model is realized through a real-time data assimilation technology, then an output result of an urban flood hydrological hydrodynamic coupling model is used as an input feature of the deep learning model, and the input feature and parameters are dynamically adjusted according to the matching degree of a confusion matrix. The TP in the confusion matrix is maximum, the TN in the confusion matrix is minimum, finally, construction of the dual-drive urban flood model is completed, and prediction is conducted through the model. According to the method, the characteristic of high calculation efficiency of the deep learning model is exerted while calculation accuracy is considered, a layered coupling architecture is provided, and the problems that the calculation efficiency of a hydrological hydrodynamic model is low and a traditional deep learning model has a black box effect are solved.

The invention discloses a multi-scale atmospheric pollutant and greenhouse gas emission data assimilation and fusion modeling method, and belongs to the field of atmospheric pollution monitoring. According to the method, the accuracy and physical interpretability of emission estimation are improved through adaptive fusion and high-precision complementation of multi-source observation data in combination with collaborative assimilation of a physical model and a deep intelligent model; and by adopting multi-scale feature dynamic fusion and a hierarchical graph neural network, high-resolution and refined emission spatial-temporal distribution modeling is realized. And meanwhile, Bayesian optimization and transfer learning are adopted to realize continuous adaptive optimization and knowledge generalization of model parameters. The method effectively solves the problems of heterogeneity, data missing, model migration, insufficient generalization ability and the like of multi-source data, and can be widely applied to the fields of atmospheric environment management, carbon emission monitoring and the like.

The invention discloses a meteorological data assimilation method fusing topographic features, and relates to the technical field of meteorological data assimilation, and the method comprises the steps: combining a meteorological variable-topographic factor correlation map with an atmospheric state numerical field, and obtaining a multivariable initial meteorological field model fusing topographic factors; identifying space-time dynamic deviation distribution characteristics of the multivariable initial meteorological field model by using real-time multi-source environment observation data, and generating a terrain disturbance response matrix; carrying out dynamic correction and space-time error reconstruction based on a terrain disturbance sensitive weight on the terrain disturbance response matrix and the multivariable initial meteorological field model through an assimilation optimization algorithm to obtain a meteorological assimilation result; according to the method, dynamic correction and space-time error reconstruction are carried out on the initial meteorological field through the terrain disturbance sensitive weight, directional perception and intelligent correction of observation deviation are realized, and the response capability and the error control level of the model in a dynamic change scene are effectively enhanced.

The invention discloses a hydraulic engineering operation and management cooperative control system based on digital twinning, and the system comprises the following steps: a data assimilation module which collects observation station, remote sensing and meteorological data and assimilates the data to obtain an assimilation estimator; the dictionary and lifting module is used for generating a lifting variable sequence according to hydrodynamic force prior; the structure identification and uncertainty module is used for identifying a lifting linear model and recursively predicting uncertainty under the constraints of conservation, monotonicity, dissipation and spectral radius; the terminal security module constructs a positive invariant terminal security set according to a support function; the opportunity constraint module is used for setting a total out-of-limit probability budget and generating an opportunity constraint substitution set; the Koopman rolling optimization module is used for solving the control sequence and executing the first control quantity; and the online updating and correcting and deformation and verification module is used for executing residual error triggering low-rank correction, projection return and support surface parameter online deformation. According to the invention, risk-controllable and stable and efficient collaborative scheduling is realized.

The invention discloses a rock fracture prediction method and system based on a resistivity data assimilation algorithm, and the method comprises the steps: building a numerical model for simulating the asymptotic fracture process of a rock mass under the load effect according to the boundary conditions and initial conditions of the rock mass; simulating the stress damage process of the rock mass by using a numerical calculation method to obtain prediction data of stress field and fracture damage field distribution of the rock mass; monitoring the damage process of the rock mass sample in real time by using a high-density electrical method to obtain observation data; rock mass fracture damage distribution is obtained based on observation data; an assimilation algorithm is adopted, information of observation data and information of prediction data are combined, and initial conditions or parameters of the numerical model are corrected; and then corrected rock fracture prediction data is obtained. Through the data assimilation algorithm, different data and model simulation results are fused in the power frame of rock fracture, the model track is automatically adjusted continuously depending on observation, the analysis result with higher precision and more consistency is obtained, and the prediction precision and predictability of the model are improved.

The invention discloses an intelligent early warning method and system for snow-melting flood. The method comprises the following steps: constructing digital twin integrating snow-melting confluence and hydrological hydrodynamic force; collecting multi-source monitoring information, and forming driving data through quality control and spatio-temporal interpolation; driving data are input into digital twinning, and model parameters and states are updated through data assimilation; simulating snow-melting runoff, water level and flow to obtain snow-melting flood probability prediction; the method comprises the following steps of: solving a river reservoir gate pump scheduling scheme by using a multi-objective evolutionary algorithm by taking reservoir outlet flow, gate opening and pump station starting and stopping as decision variables, calculating a water level, flow and a submerging range in digital twinning, and selecting a target scheme according to an evaluation rule to generate a scheduling instruction and graded early warning information; monitoring data is collected during scheduling execution and compared with probability prediction, and when the deviation exceeds a threshold value, data assimilation and scheduling scheme updating are triggered. According to the invention, the snow-melting flood forecasting precision and flood control scheduling collaboration are improved.

A system for online-coupled atmospheric chemical transport modeling and data assimilation is built using online coupling and secondary parallelization. In the prototype code of a nested air quality forecasting model system, a parallel data assimilation framework routine is introduced. The system includes observation modules, model modules, and assimilation modules. The observation module is responsible for flexible access and preprocessing of various component-type observation data. The model module handles the model integration of initial fields, involving calculations of physical and chemical processes. The assimilation module performs analysis assimilation of model state variables. This system meets the requirements for coordinated assimilation of multiple chemical component variables, simultaneous introduction and flexible combination of various types of observation data, and effective handling of non-linear and non-Gaussian distribution issues in chemical assimilation.

The application provides a power line icing simulation and prediction method based on data assimilation and artificial intelligence, comprising the following steps: step 1, meteorological elements affecting power line icing generation are predicted; step 2, the initial field of the meteorological mode prediction is adjusted, and the result of the numerical mode prediction in step 1 is optimized; step 3, the result of the numerical mode prediction after the optimization in step 2 is subjected to bias correction processing; step 4, whether icing is generated is determined, and the geographical position where power line icing is likely to occur is marked to form icing spatial distribution information field of the concerned area; and step 5, the warning result of the regional icing thickness is output. The application uses an artificial intelligence integrated constraint model to correct the meteorological prediction data field, obtains corrected meteorological information, calculates the icing spatial distribution field by using a mathematical method, and outputs the icing thickness prediction result, thereby assisting in solving the power line icing warning problem.

The invention discloses a satellite remote sensing atmospheric pollution gas concentration inversion method and system based on DOAS, and belongs to the technical field of atmospheric environment remote sensing monitoring. The method comprises the following steps: acquiring solar back scattering spectrum data observed by a satellite load and preprocessing the solar back scattering spectrum data; performing spectrum correction by using a high-precision instrument response function; in a selected fitting window, constructing a fitting model comprising a target gas differential absorption cross section, a Ring effect, Mie scattering and multiple scattering contributions; carrying out spectrum fitting by adopting a nonlinear least square algorithm, and carrying out inversion to obtain a differential batter column concentration (DSCD) of the target gas; in conjunction with an Atmospheric Mass Factor (AMF), the DSCD is converted to a Vertical Column Concentration (VCD). The system comprises a data acquisition and preprocessing module, a spectrum fine correction module, a DOAS spectrum fitting module, a dynamic AMF calculation module and a column concentration inversion and output module, and optionally comprises a data assimilation and traceability analysis module. According to the satellite remote sensing atmospheric pollution gas concentration inversion method and system based on the DOAS, the parameterized multiple scattering correction items are introduced, a dynamic AMF calculation scheme is developed, and the inversion precision and reliability under the complex atmospheric condition are remarkably improved.

The invention relates to the technical field of data assimilation processing, in particular to a WRFDA assimilation method, system and device based on GIIRS data. The method comprises the following steps: acquiring GIIRS data, extracting data from the GIIRS data, and merging and checking the data to obtain checked GIIRS observation information; performing deviation correction on the background field data of the numerical weather forecasting mode by using an RTTOV rapid radiation transmission mode to obtain simulation data after stable deviation correction; performing cloud pollution screening on the GIIRS observation information after inspection and the simulation data after deviation correction to obtain GIIRS observation information without cloud pollution; wRFDA numerical assimilation is carried out on the GIIRS observation information without cloud pollution, an initial field file is obtained, and the initial field file is used for adjusting and optimizing an output weather forecast result. According to the method, a more accurate initialization file is obtained, and the simulation accuracy of the typhoon precipitation area is more accurately improved.

The invention discloses an observation station data assimilation method based on deep learning, and belongs to the technical field of meteorological data assimilation, and the method comprises the steps: obtaining forecast background field data of a Fuhu meteorological big model, obtaining observation station data, the observation data comprises the actual measurement data of a ground meteorological observation station, a radar, a sounding station and a wind profile observation station, and the actual measurement data of the ground meteorological observation station, the radar, the sounding station and the wind profile observation station; performing spatial interpolation, time alignment and standardization on the background field and the observation data, and introducing position codes and time codes; constructing a deep learning assimilation model based on a sliding window Transform, inputting a background field and observation data, and outputting an assimilation analysis field result; and the effectiveness of the method is verified through comparison and evaluation with fifth-generation global atmosphere reanalysis data of the European mid-term weather forecast center. According to the method, the fineness and the physical consistency of local meteorological elements can be remarkably improved while the integrity of large-scale forecasting is kept, and high-precision assimilation of meteorological forecasting data is realized.

The invention provides a river flow online measurement and calculation method based on multi-source data fusion, and belongs to the technical field of river flow measurement. A Bayesian neural network and active learning collaborative hydrological memory reconstruction model is adopted to carry out high-confidence data interpolation on a sensor failure period and quantify uncertainty, fractional calculus is introduced to model a river memory effect, and flow evolution is analyzed through a fractional order water balance equation. And selecting a steady-state or non-constant flow calculation mode according to the flow change rate, inverting an optimal flow field for the non-constant flow by adopting a four-dimensional variational data assimilation method in combination with regularization constraint and time smoothing constraint, and outputting a flow measurement result and an uncertainty quantitative index. The technical problems that flow measurement and calculation data are missing and accurate interpolation is difficult due to sensor failure under the extreme hydrological condition are solved.

The invention relates to the technical field of welding, and discloses a friction stir welding control method and equipment, and the method comprises the following steps: establishing a multi-physical field model, collecting data in real time, estimating an internal state through data assimilation, predicting future evolution, and adopting model prediction control to determine and apply an instruction; carrying out online learning on continuous adjustment model or control strategy parameters; the system comprises a model establishment module, a data acquisition module, a state estimation module, a state prediction module, a control decision and execution module and an online learning module. According to the method, the internal state is accurately estimated online through model and data assimilation, and the sensing precision is improved; realizing prospective adjustment based on state prediction and model prediction control; a multi-target cost function is introduced to intelligently balance multi-dimensional performance; an online learning mechanism is integrated, and the adaptability and robustness of the system are enhanced; a complete intelligent closed-loop control system is constructed, and the automation and intelligence level is improved.

The invention discloses an automatic history fitting method based on multiple data assimilation of an improved set smoother, and relates to the technical field of petroleum engineering. The method comprises the following steps: firstly, constructing a plurality of oil reservoir models, setting a prior model, a real model and an initial expansion factor, obtaining the yield of each oil reservoir model by utilizing oil reservoir model numerical simulation, and calculating a residual error; based on corrected data covariance matrix singular value decomposition, production observation data disturbance enhancement, adaptive learning rate matrix scaling and geological boundary constraint, improving a set smoother, updating the permeability of each oil reservoir model, performing numerical simulation again, updating an expansion factor, judging whether the expansion factor meets a preset condition or not, continuing iteration if the expansion factor meets the preset condition, and if the expansion factor does not meet the preset condition, continuing iteration until the expansion factor meets the preset condition; and if not, updating the expansion factor of the iteration, ending the iteration, and outputting the permeability field of each updated oil reservoir model, thereby solving the problems of parameter overshoot, covariance statistical deviation and low calculation efficiency in oil reservoir history fitting, and facilitating history fitting of complex oil reservoir production parameters.

The invention discloses a storm surge artificial intelligence forecasting method based on physical equation constraints, and relates to the technical field of artificial intelligence storm surge forecasting. Comprising the following steps: step 1, collecting multi-source data; step 2, assimilating and preprocessing the data; step 3, realizing a machine learning calculation method of physical loss; step 4, constructing a physical and data co-constrained artificial intelligence model; 5, performing hyper-parameter adjustment by using a random search algorithm, and further optimizing the model; and step 6, carrying out interpretability analysis on the forecasting process and result. According to the invention, three control equations for storm surge calculation are introduced into the machine learning model, so that the dependence on the training data volume is greatly reduced compared with a traditional data-driven intelligent forecasting model; and analyzing the change trend of the physical variables in the forecasting process, carrying out visual display and uncertainty analysis, and solving the problem of uninterpretability of an artificial intelligence forecasting model.

The present invention discloses a high-precision correction method and system for wind power forecasting of electric power systems, which belongs to the field of new energy forecasting. The method is first based on the WRF model and ARW dynamic solver, and selects a three-layer nested structure and a physical process parameterization scheme with the wind farm location as the regional center, and combines static terrain and global weather forecast data to achieve high-resolution short-term wind speed forecasting; then, by modifying the boundary layer parameters and introducing data assimilation technology to form a variety of differentiated schemes, multiple downscaled forecast results within the same time period are obtained; then, based on the forecast results and the wind speed observation data of the station at the corresponding time, a variety of different types of improved artificial intelligence models are used to realize the correction of the forecast wind speed; finally, a variety of improved models are combined with adaptive weighting to obtain the final wind speed forecast result. By fully mining the information of terrain, observation data and global weather forecast data, the accuracy of wind speed forecast data is effectively improved.

The invention discloses a monthly meteorological and hydrological forecasting method and system based on multi-source data fusion, and relates to the technical field of meteorological and hydrological forecasting and data fusion, and the method comprises the steps: constructing a unified forecasting factor set through multi-source meteorological, hydrological and ecological data, and carrying out the feature screening; predicting meteorological elements based on a deep neural network, and correcting a prediction result by adopting a deviation correction method; and inputting the corrected meteorological data into the distributed hydrological model, and outputting a monthly runoff prediction result in combination with a data assimilation algorithm. According to the method, physical correlation screening and structure sparse optimization of heterogeneous data are realized, input variables are ensured to be scientific and explainable, space-time depth prediction and statistical distribution correction are combined, the precision and consistency of meteorological element prediction are improved, deep fusion of a physical model and statistical assimilation is realized, and the method is suitable for being applied to the field of meteorological element prediction. The prediction system has self-correction capability and long-term stability, and the scientificity and operability of monthly runoff prediction are improved.

The invention provides a dynamic sparse observation-oriented deep neural process ocean data assimilation method, and relates to the field of ocean data processing, and the method specifically comprises the following steps: constructing a training data set; simulating actually observed non-uniform and uncertain characteristics through Gaussian nuclear diffusion; building an ocean assimilation network oriented to sparse dynamic observation, outputting an analysis field and estimating uncertainty; and performing end-to-end training on the ocean assimilation network model by taking the reanalysis true value field as a supervision signal, and optimizing network parameters by combining a minimum error term and a structure constraint term. And after training is completed, inputting the background field in the test stage and sparse observation into the ocean assimilation network model for reasoning to obtain an ocean state reconstruction field conforming to the actual physical quantity scale. According to the technical scheme, the problem that in the prior art, calculation feasibility, cross-scale correlation modeling and credible uncertainty output cannot be considered under the real conditions of sparse observation and dynamic change of spatial-temporal distribution is solved.

The invention discloses a multi-source heterogeneous data assimilation-driven high-altitude atmospheric motion numerical simulation method, which comprises the following steps of: obtaining satellite microwave radiation brightness temperature data and ground dual-polarization radar differential reflectivity data, inputting the data into a quantum annealing collaborative inversion model, and solving three-dimensional humidity field distribution data; humidity gradient tensor field data are calculated, and the space coordinate range of the frontal area is judged based on a preset humidity gradient modulus length threshold value; carrying out vector included angle calculation on humidity gradient tensor field data of a frontal area and background wind field data, generating a cloud micro-physical parameterization regulation coefficient, and dynamically selecting a cloud phase change dominant mode; and inputting the three-dimensional humidity field distribution data and the cloud microphysical parameterized regulation and control coefficient into a four-dimensional variational assimilation system to solve optimal analysis field data, and outputting high-altitude wind field prediction data in a driving numerical mode. According to the method, the response sensitivity and the physical consistency of the cloud microphysical process in numerical simulation are enhanced, and the cloud phase evolution simulation precision in the severe weather process is improved.

The invention is suitable for the technical field of meteorological engineering, and provides a rainfall forecast time downscaling method based on data assimilation and source backtracking, which comprises the following steps: firstly obtaining a first forecast time sequence, a second forecast time sequence and an observation time sequence, downscaling the second sequence and then combining with the first sequence to generate a third sequence, and then generating a fourth sequence by adopting an interpolation method, the method comprises the following steps of: acquiring observation data, interpolating the observation data to a lattice point to generate an observation field consistent in space, calculating space and time gradients in a sliding time window, resolving an average motion vector, determining an upstream source based on backtracking of the average motion vector, carrying out dynamic assimilation correction, generating an hour-by-hour fifth sequence, and finally adjusting the fifth sequence through a 3-hour cumulant conservation constraint to obtain an observation result. And outputting a final downscaling result. By fusing observation data, backtracking rainfall sources and introducing conservative constraints, hourly rainfall forecasting with finer time scale, more accurate spatial sources and stronger physical constraints is realized.

The invention provides an ocean data assimilation observation strategy optimization method based on deep learning interpretability, and relates to the field of data processing, and the method comprises the steps: generating a sea-land mask through the geographic sea-land distribution and observation point validity based on pattern grid points; constructing a three-channel input tensor; constructing a network architecture to process input of different sizes in a self-adaptive pooling manner; carrying out model training by adopting a composite loss function; generating a space-time saliency map based on Grad-CAM, fusing multi-time step features through a dynamic aggregation algorithm, and identifying a key dynamic region in combination with ocean gradient and curvature features; in the key power area, K-means clustering is combined with minimum distance constraint, and high-contribution observation points are preferentially selected from all clusters; and outputting longitude and latitude coordinates and contribution value sequences of the plurality of previous optimal observation points. According to the technical scheme, the problems of low observation resource use efficiency and low forecast initial field quality in the prior art are solved.