110 results about "Spatial Autocorrelations" patented technology

A method for synthesizing noise in a digital image comprises the steps of (a) estimating noise statistics based on a set of spatial autocorrelation coefficients and a set of spectral correlation coefficients that correspond to the color channels of the image, where at least one of the spatial autocorrelation coefficients is weighted by at least one of the spectral correlation coefficients, thereby providing a weighted set of autocorrelation coefficients, and where a set of filter coefficients is obtained from the weighted set of autocorrelation coefficients; and (b) synthesizing a synthetic texture by using the filter coefficients to filter a random noise field in each of the color channels, thereby producing output noise fields in each of the color channels that replicate a synthetic texture, e.g., a desired grain appearance, when combined into the digital image.

Tear film stability has an important role in the quality of vision. A system and method for performing Fluctuation Analysis of Spatial Image Correlation (FASIC) provides for a non-invasive system and method for evaluating the dynamics of the tear film surface using spatial autocorrelation analysis. With FASIC, a series of images are obtained using illumination and a camera. The spatial autocorrelation is calculated for image frames produced by the camera. A sinusoidal background appears in this correlation together with other features. The changes in the sinusoidal background of the spatial autocorrelation is extracted and monitored over time. The spatial period of this sinusoidal background correlates with the thickness of the tear film. In this regard, one is able to derive the tear film thickness from the period of this sinusoidal background.

The invention discloses a taxi cruising action analytical method based on global position system (GPS) data grid statistic in the field of traffic action research. Spatial distribution characteristics of cruising points are analyzed based on taxi traveling track GPS data identification cruising points to generate passenger carrying points and background information points representing factors such as land, population and road networks of the previous day, grid analysis technique is adopted to convert spatial point data into point frequency data based on grids, a zero-inflated negative binomial model (ZINB) is constructed to analyze historical passenger carrying point information, background information and influences of spatial autocorrelation of the cruising points on taxi cruising point choice action, and cruising strategies of a taxi are summarized. The taxi cruising action analytical method brings convenience to further taxi traffic behavior simulation and provides an effective way for optimizing taxi cruising action and solving problems of road resource waste, fuel consumption and environmental pollution caused by cruising.

The invention relates to an ocean big data master-slave index system and method based on a Spark cloud network. The ocean big data master-slave index system and method are particularly suitable for ocean big data which have the advantages of space autocorrelation, high similarity, real-time performance and the like. The big data age internet of things, the cloud calculation and other technologies are applied to the field of distributed indexes, the internet of things serves as an antenna for ocean information collection, data division strategies are designed and the data are deployed and saved to all pieces of physical cloud in a distributed mode on the basis of the clustering thought, rapid and one-stop calling of the multi-source ocean data is achieved through the combination with a multi-layer master-slave index structure, and the requirements for high-performance calculation, strong real-time feedback, high-frequency inquiry and the like in various offshore pre-warning and emergency models are met. According to the system and the method, on the basis of physical cloud storage, the millisecond-grade inquiry response requirements are ensured, and therefore rapid and real-time disaster inversion and pre-warning are rapidly focused, and the digitalization and intellectualization progresses of ocean disaster prevention and reduction and the like are accelerated.

The invention belongs to the technical field of road safety evaluation, and discloses a space-autocorrelation-based traffic accident blackspot identification method and a space-autocorrelation-based traffic accident blackspot identification device. A geographical distribution characteristic is described from the angle of geographical accident distribution by combining accident attribute values on road space units on the premise of a proximity relationship between the road space units, and a traffic accident blackspot is identified by adopting a kernel density estimation and space autocorrelation method. The shortcoming of relatively poorer visual expression capability of research data due to adoption of a classical mathematical statistical analysis method for a traffic accident blackspot identification process in the prior art is overcome. By the method, traffic accident blackspots at road sections and intersections can be identified, so that identification results are more accurate. In addition, a special analysis function is fully utilized on the basis of spatial position and geographical position information, and a foundation is laid for finding out accident causes.

The invention relates to a method and a system for automatically generating an urban road traffic guidance strategy. According to the method and the system, by applying the intelligent traffic theory, the spatial autocorrelation theory, the fuzzy mathematic theory, an statistical method and the computer technology, real-time traffic flow data and historical data are researched and processed, and therefore, under the condition that a traffic incident occurs in a road traffic network, the traffic guidance strategy which can be released in time through VMS (variable message signs) on a road is automatically generated. According to the method and the system, by taking historical artificial guidance information and the real-time traffic flow data as input data, on the basis that historical guidance empirical laws are analyzed and real-time traffic flow features are excavated, a final guidance release strategy is determined by applying a traffic state spatial inequality analysis method provided by the invention. A proposal and strategy scheme can be provided for the quick release of the guidance information of an urban traffic management department when certain traffic incidents occur, such as road accidents, congestion, severe weather, management and control, and the like.

The invention provides a spatial autocorrelation machine learning satellite precipitation data downscaling method. The method comprises the steps of acquiring TRMM precipitation data and surface parameter data; preprocessing the earth surface parameter data to obtain DEM data with spatial resolutions of 1km and 25km, daytime earth surface temperature, night earth surface temperature, day and nightearth surface temperature difference and NDVI data; performing spatial autocorrelation analysis on the TRMM precipitation data to obtain an estimated precipitation data spatial autocorrelation valuewith the spatial resolution of 25km; downscaling the spatial autocorrelation value of the precipitation data with the spatial resolution of 25km until the spatial resolution is 1km; establishing a nonlinear regression model; and obtaining rainfall downscaling data with the spatial resolution of 1km based on the nonlinear regression model. The invention further provides a system and a terminal. Thedownscaling result is superior to the downscaling result based on a conventional regression model, and the method has important theoretical and practical significance and popularization and application value.

The invention discloses a land use change modeling method and system implemented in combination with spatial filtering and relates to the field of geographical information sciences. The method focuses on eliminating the spatial autocorrelation influence of data in the process of land use change modeling, and comprises the following steps: collecting original data, and preprocessing the original data so as to generate a data set sequence; determining the optimum distance of spatial filtering by using a variation function, and then splitting variable factors by using a Getis principle based a spatial filtering method; constructing a logistic regression model according to a filtered sample, and assessing the model by using methods such as fitting accuracies and ROC (receiver operating characteristic) curves; and finally, providing a Markov chain and a customized change pattern for forecasting the trend of land use change. According to the invention, based on a statistics mathematical model, a spatial filtering mode is adopted for making up the shortcomings and defects of the mathematical model in geo-spatial calculation, so that through the combination of the two, the fitting accuracy of the model is improved, and the model is more conform to the time and space characteristics of land use change.

The invention relates to a big data-based time-space confusion exposure degree assessment system and method. The system comprises a time-space data mining block, a multi-source heterogeneous data fusion module, a final variable selection module, a time-space generalized additive model building module, a re-sampling model module, a variation function time-space modeling module and a concentration estimation module; massive time-space data is mined; a relationship between multiple influence factors and pollutant concentration is established by adopting an accumulative nonlinear method; and through residual variation function fitting, spatial autocorrelation is considered, so that the prediction precision and effect are greatly improved.

The invention relates to an urban land utilization simulation cellular automaton method based on space errors, which comprises: 1) classifying remote sensing images to obtain an urban land utilizationgraph and a driving factor; 2) performing system sampling on the urban land utilization map and the driving factor to obtain sample data; 3) training the sampling data by using SEM, and establishinga CA conversion rule; 4) evaluating the fitting performance of the conversion rule by adopting the log likelihood value and the AIC, and checking the effect; 5) establishing a CASEM new model, and calculating to generate an urban land transition probability graph; 6) predicting the number of urban cells at each time point in the future based on historical land utilization data of remote sensing classification; 7) using CASEM to simulate and predict an urban spatial pattern and an expansion process; and 8) performing, outputting and saving comprehensive evaluation of the simulation results using multiple types of indicators. Compared with the prior art, the urban land use simulation cellular automaton method based on the space error has the advantages that the space autocorrelation in the model error can be eliminated, and the simulation precision is improved.

The invention relates to an unsupervised image segmentation quality evaluation method based on dM, and the method comprises the steps: carrying out the bilateral filtering and two-dimensional Gabor filtering of an original image, so as to extract a spectrum feature vector and a spatial feature vector; carrying out the transformation of the obtained spatial features, extracting three principal components, and carrying out the superposition with the spectrum features too form a spectrum-spatial feature set; taking the spectrum-spatial feature set as a base map, and calculating the layered spatial heterogeneity of the segmentation region at each dimension; obtaining the intensity value of the layered spatial heterogeneity, and calculating the spatial autocorrelation of the segmentation results of all dimensions; calculating a mean value for representing the intensity value of the layered spatial heterogeneity of each dimension in the feature set and the mean value of Moran indexes, and constructing the space of two values; calculating the Mahalanobis distance between each point in the space and a point (1, 0), and finally carrying out the evaluation of the segmentation quality of theimage according to the Mahalanobis distance. The beneficial effects of the invention are that the method improves the quality evaluation precision and stability, and an evaluation result accords withthe visual features of the human being in a better way.

The embodiment of the invention provides a method and a device for calculating the soil attribute value. The method comprises the following steps: calculating the weight of each influence factor on the soil property value; collecting the values of the influence factors of each to-be-measured point and each sampling point, and calculating the similarity between the soil property value of the to-be-measured point and the soil property value of the sampling point at the environmental factor level; determining the spatial neighborhood of the to-be-measured point according to the spatial self-correlation distance or the multi-scale spatial correlation distance, inputting the similarity between the soil property value of the to-be-measured point and the soil property value of the sampling pointat the environmental factor level and the soil property values of the sampling point in the neighborhood into a soil attribute value calculation model which gives consideration to both environmental factor similarity and space self-correlation, and obtaining the soil property value of the to-be-measured point. According to the embodiment of the invention, the environmental factor information related to the soil property of the to-be-measured point and the attribute value information of the space adjacent to the known sampling point are fully utilized. The spatial variation of the soil propertyat the mutation boundary can be better depicted, and the soil property value is ensured to be closer to the actual condition.

The invention discloses a space filter value-based logistic regressive landslide hazard analysis method. Aiming at analytical investigations of landslide hazards, a space filter value thought is imported into a common logistic regression model so as to design a regressive landslide analysis algorithm which comprises the steps of non-landslide site selection, disaster-inducing factor obtaining andgrading, adjacent matrix construction, feature value and feature vector calculation, stepwise regression feature vector selection and regressive modeling. According to the method, the problem that themodel precision is not high as logistic regression models are influenced by spatial autocorrelation between variables can be solved. By adding filter value operators which are constructed by utilizing selected feature vectors into logistic regression models, autocorrelation influences of residuals can be effectively filtered, the fitting goodness and prediction correctness of the regression models can be improved and correct simulation and prediction for landslide hazards can be realized.

The invention discloses an MIMO multiantenna communication system. The system comprises a transmitting end and a receiving end. The transmitting end and the receiving end are equipped with uniform spherical MIMO antenna arrays. The invention also provide a performance evaluation method for the MIMO multiantenna communication system. The provided system is described through scatterer Gaussian distribution. A spatial Gaussian distribution density function expression is determined. A joint probability density function of an angle of divergence (AOD) of a single-time bounce SBM component is expressed. An edge PDF of the angle of divergence (AOD) of a single-time bounce path SBM in the system is provided through utilization of integral of the joint PDF on a path length interval component. Fourier transform is carried out on pulse response h<pq>(t) to obtain a time varying transfer function. A frequency cross-correlation function is expressed. According to the system and the method, throughimporting of the spatial autocorrelation function, and temporal and frequency cross-correlation functions between two array elements pq, performance of the MIMO multiantenna communication system is evaluated, and the performance of terminal antenna arrays and a wireless communication system is optimized.

The invention discloses a remote-sensing-image-based parallelization method of a regression model of a characteristic function space filter value. On the basis of the influence on a model by variable space autocorrelation during a remote sensing image data regression modeling process, a characteristic function spatial filtering method is employed; block division is carried out on an image and distributed calculation is carried out by using a constructed parallel computing cluster; and then a block division calculation result is returned to a main node for gathering. On the basis of comparison of regression model fitting evaluation parameters MSE, RMSE, R<2>, Adj.R<2> that are obtained by serial and parallel processing and a parallel speed-up ratio S, elimination of the spatial automation influence in the spatial statistic regression modeling by the remote-sensing-image-based characteristic function space filter value parallelization method can be demonstrated; and the computing efficiency can be improved effectively.

The invention discloses a construction method of a spatial variable coefficient PM2.5 concentration estimation model based on a Re-ESF algorithm. The construction method comprises the following steps:step 1, multi-source AOD product fusion; step 2, data processing of related influence factors and PM2.5 concentration; step 3, obtaining independent variable values and PM2.5 concentration values ofall parts of the monitoring site; step 4, constructing an adjacent matrix; step 5, centralizing the spatial adjacency matrix and calculating a feature value and a feature vector; step 6, solving a regression coefficient of an SVC model based on the Re-ESF; step 7, judging whether variables in the model are remarkable or not; Step 8, model precision evaluation; step 9, 10-fold cross validation; step 10, judging the relation between the number k of neighborhoods and the number n of monitoring station points; and step 11, selecting an optimal model according to the precision evaluation. On the basis of effectively eliminating the influence of spatial heterogeneity and spatial autocorrelation on PM2.5 concentration modeling, a random effect and spatial variable coefficient method is introduced, so that the precision of the concentration estimation model is further improved, and the model resolving time is shortened.

The invention discloses a spatial autocorrelation region-based spatial-temporal difference detection method for vegetation net primary productivity. . The method comprises the following steps: S1, obtaining MOD17A3 data of research years of a research area; S2, extracting a vegetation NPP graph of a research time period of the research area; S3, calculating the vegetation NPP variable quantity inthe research time period of the research area; S4, calculating a global Moran index, a high / low value clustering index and related statistics of the vegetation NPP in the research period of the research area; S5, judging the overall clustering characteristics of the research area data samples; S6, calculating a hot spot analysis index of the vegetation NPP in the research time period of the research area; S7, judging hot spot distribution characteristics of the vegetation NPP in the research time period of the research area; S8, calculating a local Moran index of annual vegetation NPP variation in the research period of the research area; S9, judging spatial aggregation characteristics of annual vegetation NPP variations in the research period of the research area. The overall clustering characteristics of the vegetation NPP can be accurately expressed, and the local spatial characteristics can be detected.

The invention discloses a topographical data-based test method of suspicious numerical values of telemetering rainfall amount sites. According to the method, real-time rainfall amount site data in a region is used to establish a spatial weight matrix according to geospatial coordinate distances; spatial autocorrelation modeling analysis is carried out on each site according to a rainfall attribute value of the site, and whether the attribute value clusters is judged according to clustering statistic parameters; the suspicious numerical value is screened out through calculating a clustering state discriminant index Scdi on the basis of clustering; and site re-classification is carried out according to digital elevation attributes of the telemetering rainfall amount sites, spatial similarity analysis is carried out on the attribute values of the suspicious sites and the sites in a set distance threshold value range, and thus whether the data sent from the telemetering rainfall amount sites is abnormal values is confirmed. The method improves the accuracy and the precision of traditional single threshold value verification.

The invention discloses a selection method for excellent individuals of Chinese firs. The selection method is a mixed linear model method, utilizes a spatial analysis method for estimating a spatial residual variance and spatial autocorrelation, utilizes a restrictive maximum likelihood method for estimating a genetic variance, and utilizes best linear unbiased prediction for individual breeding value of the Chinese firs. The method analyzes genetic variation, genetic control and genetic correlation among characters for main economic characters such as tree height, diameter at breast height, wood basic density, micro fiber angle and the like of the Chinese fir, predicts the breeding value of each character for each individual of the Chinese firs, and according to a multi-character index selection formula of the Chinese fir, picks out individuals which are genetically excellent both in growth and wood property. According to the method, when the environmental impact is eliminated in generic parameter estimation, the multi-character index selection based on the breeding value can select the individuals with generically excellent comprehensive characters effectively, the accuracy is high, and the selected excellent individuals can serve as both the material for asexual reproduction of the Chinese firs and the breeding material for next-generation breeding population of the Chinese firs.

The invention relates to a method and a device for analyzing spatial correlation of an agricultural product price, and aims to solve the problems that the spatial correlation and the spatial heterogeneity of the agricultural product price cannot be accurately reflected in the prior art. The method comprises the following steps: acquiring agricultural product price information; preprocessing the agricultural product price information and matching the agricultural product price information to a map; establishing a highway network data set; calculating an optimal path of every two agricultural product wholesale markets; determining the distance of the optimal path as a traffic distance space weight coefficient between the two agricultural product wholesale markets; calculating global spatial autocorrelation analytical statistic and local spatial autocorrelation analytical statistic of the agricultural product price information according to the traffic distance space weight coefficient between the two agricultural product wholesale markets; generating an agricultural product price spatial correlation statistical graph according to the global spatial autocorrelation analytical statistic, and generating an agricultural product price local autocorrelation distribution diagram according to the local spatial autocorrelation analytical statistic.

A regional population density simulation method based on a feature vector space filtering value comprises the steps of obtaining a regional vector file and statistical data, selecting a feature vectorspace filtering value method, and selecting an auxiliary independent variable by taking regional night light average brightness as an independent variable; processing the remote sensing night light image according to the region vector file, and calculating region total brightness and average brightness; establishing an adjacency relation, obtaining a spatial adjacency matrix, centralizing the spatial adjacency matrix, and calculating a matrix characteristic value and a characteristic vector; and extracting a proper feature vector as a spatial influence factor of luminous brightness, adding the feature vector into an independent variable, solving a regression coefficient, obtaining a feature vector spatial filtering regression model of the population density, and realizing regional population density simulation according to the model. According to the method, the influence of spatial heterogeneity and spatial autocorrelation on population density distribution can be effectively eliminated, manual statistics is replaced by an automatic means, manpower and material resources are saved, and the method has important significance for urbanization intelligent monitoring, environmental pollution detection and other applications.

A spatial autocorrelation machine learning-based downscaling method of satellite precipitation data includes obtaining the TRMM precipitation data and the land surface parameters; preprocessing the land surface parameters to obtain DEM, day land surface temperature, night surface land temperature, day-and-night land surface temperature difference and NDVI with spatial resolutions of 1 km (0.621 miles) and 25 km (15.534 miles); performing a spatial autocorrelation analysis of the TRMM precipitation data to obtain an estimated spatial autocorrelation value of the precipitation data with a spatial resolution of 25 km (15.534 miles); downscaling the spatial resolution of the spatial autocorrelation value of the precipitation data from 25 km (15.534 miles) to 1 km (0.621 miles); establishing a nonlinear regression model; obtaining a precipitation downscaling data with a spatial resolution of 1 km (0.621 miles) based on the nonlinear regression model. A system and a terminal are also provided.

The invention discloses a spatial interpolation method for long-span bridge monitoring data. The spatial interpolation method comprises the following steps: (1) spatial auto-correlation analysis on sensors of the same kind is performed based on distribution of measurement points of certain monitoring data on the long-span bridge; the spatial auto-correlation analysis is embodied by spatial variation function; (2) search of a reference evaluation point is performed by combination of real monitoring distribution points of the bridge; a four-direction search scheme is carried out by steps of dividing a plane into four quadrants based on an x-coordinate and a y-coordinate of an interpolating point, taking the length of a range a as a searching radius, searching a known sample point, which is nearest to the interpolating point, from each quadrant; (3) the spatial interpolation analysis on the stress-strain monitoring data is performed by combination of the variation function of the real monitoring distribution points of the bridge by adopting a Kriging method so as to obtain the estimated value of an unknown point. The spatial interpolation method for the long-span bridge monitoring data provided by the invention is complete in the monitoring information and relatively high in accuracy.

The invention discloses a traffic accident influence factor spatial effect analysis method, which comprises the following steps of: dividing a spatial research area of a traffic accident to obtain anexplanatory variable and an explanatory variable required by modeling; deleting unreasonable interpreted variables; constructing different types of spatial weight matrixes for representing the incidence relation between the research areas; performing spatial autocorrelation inspection; constructing an OLS model, adjusting the goodness-of-fit value to be maximum through an F test, deleting explanatory variables which do not pass the t test, and reconstructing the OLS model; performing Lagrange multiplier test on the residual error of the OLS model; re-incorporating the deleted explanatory variables, constructing a plurality of spatial Dorbin models, and selecting an optimal spatial Dorbin model and a final explanatory variable; and carrying out quantitative analysis on the spatial effect ofthe traffic accident influence factors in the region. According to the method, the spatial effect of traffic accident influence factors is quantitatively analyzed, and a decision basis is provided for traffic safety planning and management.

The invention discloses a regional vegetation index driving factor analysis method and device, and a storage medium. The method comprises the following steps of obtaining vegetation index data of a research region and at least two kinds of driving factor data of vegetation indexes; performing local space autocorrelation analysis on the vegetation index data to obtain an analysis result containingmultiple clustering types; according to a preset combination rule, combining the different clustering types to obtain at least one combined clustering type; and performing importance analysis on the driving factors of the vegetation indexes belonging to each combined clustering type in the at least one combined clustering type by utilizing a random forest model. By means of the technical scheme, the analysis precision of the driving factors can be improved through the partially combined clustering types obtained after recombination, so that the precision requirement is met, and the driving factor analysis result of the specific and important clustering types is extracted from the whole.

Tear film stability has an important role in the quality of vision. A system and method for performing Fluctuation Analysis of Spatial Image Correlation (FASIC) provides for a non-invasive system and method for evaluating the dynamics of the tear film surface using spatial autocorrelation analysis. With FASIC, a series of images are obtained using illumination and a camera. The spatial autocorrelation is calculated for image frames produced by the camera. A sinusoidal background appears in this correlation together with other features. The changes in the sinusoidal background of the spatial autocorrelation is extracted and monitored over time. The spatial period of this sinusoidal background correlates with the thickness of the tear film. In this regard, one is able to derive the tear film thickness from the period of this sinusoidal background.

The invention relates to an urban heat island intensity quantification method and system, a storage medium and equipment. The method comprises the following steps: acquiring and preprocessing remote sensing image information of a target area; performing inversion calculation on the preprocessed remote sensing image information to obtain surface temperature information; performing local space autocorrelation analysis on the target area according to the surface temperature information; and calculating the relative urban heat island strength of the local space and carrying out ward minimum variance clustering analysis to determine the urban heat island strength grade of the local space. According to the invention, the remote sensing image is subjected to preprocessing and inversion calculation to obtain the surface temperature information, and local space autocorrelation analysis is carried out to obtain the average value of the surface temperature information of the target type of the local space, so that the relative urban heat island intensity and urban heat island intensity level of the local space are accurately calculated; the problems that an existing method is difficult to ignore, application limitation is caused, and the surface temperature is affected by the surrounding environment are greatly solved, the quantification result is accurate, and the application range is wider.

Described is a system for catastrophe prediction. The system generates a time series of observables at multiple time steps from data observed from a complex system. A surrogate time series based on the time series of observables is then generated. Inferred network structures for both the time series of observables and the surrogate time series are reconstructed. Next, spatial autocorrelation for each inferred network structure in both the time series of observables and the surrogate time series is computed. A statistical test of a detected trend between the time series of observables and the surrogate time series is computed to determine if the detected trend occurred by chance. Finally, an early warning signal of the detected trend occurring by chance is generated.

The invention provides a multi-spectral laser radar wavelength selection method for vegetation biochemical parameter detection. The method comprises: carrying out a sensitivity analysis on a blade radiation transmission model PROSPECT and acquiring contribution degrees of a model leaf structure and various biochemical parameters to each wavelength reflectance in a preset wavelength range; establishing a sensitive wavelength sequence table for each parameter of the model; carrying out forward modeling on a PROSPECT model and generating a simulated spectral data set; carrying out band spectral autocorrelation analysis on the simulated spectral data set and checking correlation of random paired wavelength combined reflectance within the preset wavelength range; and acquiring multi-set alternative wavelength combinations, carrying out targeted PROSPECT model inversion to check the inversion accuracy of biochemical parameters by each set of alternative wavelength combination, and constructing an optimal band subset for detection of blade biochemical parameters of the multi-spectral laser radar based on inversion statistical parameters. The method can be applied to the novel multi-spectral laser radar instrument and supports direct and simultaneous acquiring of color laser spectral information and laser point cloud spatial information of a target object.

The invention provides a flood early warning analysis method and system based on Poisson regression and spatial filtering values. The method comprises the steps that frequency statistics of flood early warning events is carried out on historical observation data of hydrological stations in a certain drainage basin; selecting disaster inducing factors and disaster forming factor data in the drainage basin, constructing a Thiessen polygon for hydrological station point coordinate points to obtain a space, and performing eigenvalue and eigenvector decomposition; under the condition that the spatial autocorrelation significance test is met, carrying out forward selective regression, and establishing a Poisson regression model based on a spatial filtering value; and adding the screened featurevectors serving as independent variables and disaster-causing and pregnancy factors into a Poisson regression model, constructing a spatial filtering value-based flood warning event frequency Poissonregression model, and supporting to realize fitting prediction of the flood warning event frequency. The method can overcome the defect of low fitting precision in the prior art, and is used for riskprediction of flood dike burst and dam break of all stations in a whole basin.