254 results about "Stepwise regression" patented technology

The invention discloses a method for inverting remote sensing forest biomass. The method comprises the following steps: on the basis of remote sensing data pretreatment, extracting characteristic variables of a vegetation canopy from a LiDAR point cloud (comprising canopy three-dimensional space information) and multispectrum (comprising spectrum information on the upper surface of the canopy) data respectively; screening the characteristic variables of the LiDAR point cloud and the multispectrum through correlation analysis, and inverting overground and underground biomass by combining the ground actually measured biomass information through a stepwise regression model. Through the adoption of the optimized inverting model of northern subtropical forest biomass, constructed by method, the 'determination coefficient' R<2> of the model can be increase by 3-24%; the forest biomass can be estimated in high precision, and the 'relative root-mean-square error' (rRMSE) can be reduced by 2-10%. The method can be applied to the fields of forestry investigation, forest resource monitoring, forest carbon reserve evaluation, forest ecosystem research and the like, and provides quantitative data support for forest sustainable management and forest resource comprehensive utilization.

The invention belongs to the field of medical signal processing, and provides a detection method of vascular elasticity and blood pressure based on a single probe photoplethysmography pulse wave. According to the technical scheme, through the adoption of the detection method, quantization of the vascular elasticity and estimation of the blood pressure are achieved on the condition that a photoplethysmography pulse wave is collected merely through a single probe. The detection method comprises the first step of collecting the pulse wave and the blood pressure based on a pulse wave collecting system and a cuff sphygmomanometer; the second step of extracting features related to the vascular elasticity and the blood pressure; the third step of building a blood pressure predicting linear regression equation through the adoption of a stepwise regression method according to a relevant feature quantity; the fourth step of training a BP neural network according to the relevant feature quantity, and measuring the elasticity of an artery blood vessel and a blood pressure value through the trained neural network. The detection method of the vascular elasticity and blood pressure based on the single probe photoplethysmography pulse wave has the advantages that the algorithm performance is good, the quantization of the elasticity of the artery blood vessel wall of the human body can be achieved through the photoplethysmography pulse wave collected by the single probe, and the blood pressure value can be accurately predicted.

The invention discloses a method for fast predicting organic pollutant n-caprylic alcohol / air distribution coefficient based on molecular structure, belonging to the technical field of quantifying structure / active relationship (QSAR) facing to the environmental risk evaluation. The method is characterized of comprising the steps of: adopting the molecular structure of atomic center fragment characterization compound; and screening the atomic center fragment combination by means of stepwise regression and partial least-squares regression, to build a group contribution model for predicting KOA.The internal authentication and the external authentication improves that the built KOA group contribution model has stability and predicting capability, and a range and distance method and a probability density method express the application domain of the group contribution model, thereby defining the application range of the model and guaranteeing the predict accuracy. The method has the effectsand benefits of being capable of fast predicting the KOA of the high flux compound, obtaining the KOA with low cost, being helpful for obtaining the high flux KOA data, and having a significant meaning for the environment supervision and the risk evaluation of chemicals.

The invention provides a nonlinear model-based multispectral remote sensing water depth inversion method and an apparatus thereof. The method comprises the following steps: acquiring the multispectral remote sensing image of a preset area and the actually measured control point water depth of a preset water area, and preprocessing the multispectral remote sensing image to obtain a preset area reflectivity; carrying out water-land separation on the preset area reflectivity through a near infrared waveband spectrum characteristic-based threshold technique to obtain the reflectivity of the water surface of the preset water area; establishing a nonlinear inversion model corresponding to the preset water area according to the water surface reflectivity and the actually measured control point water depth; and regressing the nonlinear inversion model through a stepwise regression algorithm, and inversing according to the regressed nonlinear inversion model to obtain the water depth of the preset water area. The highly-precise water depth of the island reef water area far from the land is rapidly obtained according to the nonlinear inversion model on the basis of the actually measured water depth, model establishing and the solving process are simple, and the nonlinear inversion model is suitable for various types of water depth inversion engineering and has good portability.

The invention discloses a short-term climate forecasting method based on empirical mode decomposition and numerical value set forecasting. The invention adopts a way of integrating a numerical value set forecasting technology and a mean generating function stepwise regression model and combines a new empirical mode decomposition (EMD) method for processing a data sequence. The short-term climate forecasting method comprises the following steps of: firstly, decomposing a non-stationary climate data sequence into a stationary intrinsic mode function (IMF) component with multi-scale feature; then constructing different forecasting models for each IMF by a way of set forecasting and stepwise regression analysis; and finally linearly fitting to form a forecasting result. When the system is used for short-term forecasting, a user can cut out the appointed sequence length and forecasting length according to the actual data demand and vnlrpfalgp select a forecasting model parameter in a set forecasting process. Compared with a direct or single forecasting method, the invention has better forecasting capacity for the variation trend of climate and sudden climate.

The invention discloses a tight sandstone porosity and permeability prediction method based on reservoir quality main control factor analysis. The method comprises the following steps: 1) performing quantitative diagenesis evaluation; 2) selecting diagenesis factor reflection; 3) selecting multivariate linear stepwise regression as a data analysis method, and realizing porosity and permeability prediction through reservoir quality development main control factor analysis; and 4) performing regression analysis on the porosity and permeability according to a regression analysis method. According to the prediction method disclosed by the invention, the reservoir quality influencing factors of 8 sections of sandstone in the east region of the Sulige gas field are clear, a reservoir quality prediction model is established, and the prediction effect is excellent; and moreover, a novel thought and method can be provided for tight sandstone reservoir quality quantitative prediction in other regions.

The invention discloses a method for deterring the relation between movement characteristics and a high efficient coding mode in pixel-domain video transcoding, which comprises the following steps of: selecting a video array with the typical movement characteristics under a specific resolution factor and a coding mode having the important influence on the improvement of the transcoding quality; analyzing the column diagram of the motion vector amplitude of the typical video array video frame by video frame; traversing various coding mode combinations video frame by video frame and recording the transcoding video quality, selecting the most effective coding mode by a stepwise regression method, and then clustering and simplifying the coding mode, and finally constructing the corresponding relation model between the movement characteristics represented by the column diagram of the motion vector amplitude and the high efficient coding mode. The method provided by the invention causes that the relation between the movement characteristics and the high efficient coding mode being difficult to be determined in the first is converted into a classifier, thereby the problem is solved. In the pixel-domain video transcoding process, the relation between the movement characteristics and the high efficient coding mode determined by the invention can increase the transcoding performance.

The invention discloses a TBM construction surrounding rock drillability grading method based on data mining, and the method comprises the steps: building a penetration prediction model by using a stepwise regression method and considering the operation experience of a master driver according to the principle of safety and high efficiency; Performing regression analysis on the engineering project,and establishing a rotating speed prediction model; Summarizing engineering cases at home and abroad, and establishing a TBM construction tunneling utilization rate prediction model on the basis of data mining; And by taking the TBM tunneling speed as an evaluation index, finally establishing a surrounding rock drillability grading model based on TBM construction. The method aims at the current situation that traditional tunnel surrounding rock classification is incapable of well adapting to TBM construction by taking surrounding rock stability as an evaluation basis. A data mining method isprovided, a tunnel surrounding rock systematized classification method under the TBM construction condition is established, the TBM tunneling performance under various surrounding rock grades is accurately predicted, a decision basis is provided for TBM construction cost and construction period prediction, and optimization adjustment of TBM tunneling parameters is achieved.

The invention provides a blind source signal denoising method based on ensemble empirical mode decomposition, and belongs to the technical field of signal processing. By means of the method, definitions on the white noise amplitude and the number of iterations in an original algorism are rectified. False component discrimination is conducted on the IMP component obtained after IEEMD is decomposed through a classic stepwise regression analysis method, features of original signals are effectively reserved, the false component generated by the IEEMD algorism is eliminated, and interference to the subsequent denoising algorism by the false component is eliminated. Finally, for the non-convergence phenomenon generated occasionally when the ICA algorism processes the high-frequency signals, a high-order TFastICA method is provided, features of the IEEMD and the TFastICA are combined, and rear-end processing is conducted on the IEEMD through the TFastICA method. The blind source signal denoising method based on the ensemble empirical mode decomposition has wide application prospects in the fields of removing mechanical vibration noise, voice signal noise, instantaneous underwater noise and other signal processing fields.

The invention discloses a process quality prediction system and a method thereof. When a processing apparatus performs a process on a target, the process is measured by a measurement apparatus to receive a process value. The process value and several previous quality data collected from the measurement apparatus are used to predict the quality of the product which is processing inline. The method is composed of a moving window, a stepwise regression scheme and an analysis of covariance (ANCOVA). The drift and shift of process are overcome by the moving window. A key variable set is selected by the stepwise regression scheme and a virtual model is identified by the analysis of covariance.

The invention discloses a population data spatialization method and system based on partition modeling, and a medium. The method comprises the following steps: collecting an original data source of aresearch area, which influences spatial distribution of population, and carrying out pre-processing; carrying out gridding processing on the data based on a geographic detector model, carrying out standardization processing after obtaining a population distribution influence index, and preliminarily screening out a population distribution influence factor; dividing the research area into a plurality of partitions, and respectively rescreening the population distribution influence factors of the partitions; and meanwhile, establishing a stepwise regression equation and a random forest model, performing precision comparative analysis on the population data spatialization result of each partition, selecting an optimal simulation result in each partition as a population data spatialization final result of each partition, and performing combination to obtain a population spatial distribution simulation schematic diagram. According to the method, the research area can be partitioned based onpartition modeling, the population data spatialization model of each partition can be constructed, and the accuracy and efficiency of population spatial distribution simulation are improved.

The invention relates to a method for setting process parameters of a stainless steel strip steel withdrawal and straightening machine unit. The method is characterized in that a set of systemic method for setting the process parameters of a stainless steel withdrawal and straightening machine is established by utilizing finite element software to establish a simulation model and applying the stepwise regression principle to set the process parameters such as insertion depths and tension values of a 1# roller, a 2# roller and a 3# roller of a withdrawal and straightening machine, and optimization is achieved by combining application effects of actual production. By means of the method, defects of a wave plate shape and a warping plate shape easily occurring in the production of thinner stainless steel strip steel are eliminated well, theoretical guidance and technical support are provided for withdrawal and straightening of stainless steel, and quality of strip steel is greatly improved. Besides, a process parameter query and addition system interface has strong visuality so that the production is scientific and systematic, and the problem that the process parameters of work shifts are different due to personal experience factors of operators is solved. An adding system guarantees that technical personnel can add process parameters capable of enabling strip steel to obtain good plate shapes after a series of statistics according to actual conditions, so that a database can be kept updated frequently, and process parameter setting can be improved and optimized continuously. Therefore, the method for setting the process parameters of the stainless steel strip steel withdrawal and straightening machine unit has a wide application prospect.

A system and method for measuring various weld characteristics is presented. The system and method can comprise a means to measure penetration depth of butt welds in thin plates, for example, using laser generated ultrasounds. Superimposed line sources (SLS) can be used to generate narrowband ultrasounds. A signal processing procedure that combines wavenumber-frequency (k-ω) domain filtering and synthetic phase tuning (SPT) is used to reduce the complexity of Lamb wave signals. The reflection coefficients for different wavelengths corresponding to each wave mode can be calculated. Regression analysis that can include stepwise regression and corrected Akaike's information criterion (AIC) can be performed to build prediction models that use the reflection coefficients as predictors.

The invention relates to a method for inverting vegetation parameters by remote sensing based on reflection spectrum wavelet transform. The method comprises the following steps of: 1) acquiring the vegetation parameters and the original spectrum thereof under different conditions, and performing spectrum transform on the original spectrum; 2) performing continuous wavelet transform on the original spectrum by using different wavelet functions, and generating wavelet coefficients with different frequencies; 3) performing stepwise regression by taking different scales of wavelet coefficients as independent variables and taking the vegetation parameters as dependent variables, selecting spectrum wave bands needed by the inversion of the vegetation parameters, constructing a model of quantitative inversion of the vegetation parameters, and calculating R2 of the model; and 4) comparing modeling R2 of the constructed model according to different wavelet decomposition scales, and determining the model with the maximum modeling R2 as the optimal model. By the method, the hyperspectral remote sensing inversion precision of the vegetation parameters can be obviously improved, and the remote sensing inversion precision of biochemical parameters can be improved preferably. The method has wide parameter applicability, is applicable to leaf or canopy reflection spectrum, and is applicable to satellite remote sensing hyperspectral data.

The invention discloses a nondestructive detection method of the total number of bacteria in livestock meat, which comprises the following steps: using a high spectral imaging system to obtain a high spectral scattering image of a livestock meat sample to be detected; using the Lorentz function to fit the scattering features to obtain Lorentz parameters, and using the arithmetic product of the parameters as spectral data; using a stepwise regression method for selecting the optimum wavelength combination; and using the Lorentz parameters at the optimum wavelength part to establish a multivariant linear predict model which can be used for judging the total number of bacteria in the livestock meat. The invention has the advantages of high speed and nondestructive effect, light within the visible-near infrared spectral wavelength range (400 to 1100 nm) is used as a light source for irradiating the livestock meat sample, the scattering spectral information on the surface of the livestock meat sample is analyzed, and the method can be used as the nondestructive detection method of the total number of the bacteria in the livestock meat. After necessary modification, the method of the invention can also be used as a novel method to be applied to the nondestructive detection fields of internal components such as moisture content, protein content, fat content and the like in the livestock meat.

The invention discloses a remote sensing estimation method for the SPAD value of rape leaves in different growth periods. The method comprises the following steps: by taking commercial crop rapes in northwest regions as research objects, measuring the spectral reflectivity and SPAD value of rape leaves in each growth period; analyzing the relativity between 10 types of spectral indexes and the SPAD value of the rape leaves; constructing an SPAD random forest regression estimation model of the rape leaves based on the spectral indexes, and verifying the constructed model; and comparing the constructed model with a traditional simple linear regression model and a traditional multiple stepwise regression model. Modeling and verifying R2 respectively reach 0.91 and 0.74 or above, the verifying RMSE is between 1.571 and 5.004, RE is between 2.66% and 13.22%, and thus an optimal estimation model of the SPAD of the rape leaves can be obtained.

The invention discloses an intelligent early warning method for dam safety monitoring data. The method comprises the steps of early warning model establishment, threshold value setting and mutual feedback type early warning. Gross error identification and gross error processing are carried out, model sample data quality is improved, according to the monitoring items, independent variable relevance, historical monitoring data quantity and historical monitoring data distribution, different early warning models and indexes are established, including a stepwise regression model, a correlation vector machine model and a gray system model; the established models can reflect the relationship between the independent variable and the dependent variable more truly and are wide in application range,according to a measuring instrument, measuring point attributes, a threshold value, an early warning model and indexes, real-time early warning is carried out on monitoring data, monitoring instrumentabnormity early warning is sent to monitoring personnel, or dam safety early warning is sent to dam safety management personnel, experts with professional knowledge and rich experience are not needed, the workload is small, the early warning speed is high, and the early warning result is more accurate and reliable.

The invention provides a personality prediction method based on microblog user behaviors. The method comprises the steps that first, an id list of microblog active users is acquired, personality questionnaires filled on line by tested users are acquired according to the id list of the active users through a 'talk to him' function of the microblog; second, according to the list of the tested users who fill the personality questionnaires, microblog data of the users are downloaded, and according to an established microblog network behavior system, corresponding static behavior features and dynamic behavior features are extracted from the microblog data; third, numeralization is conducted on the extracted dynamic behavior features of the users according to a time sequence analysis method to achieve a complete microblog feature set; fourth, according to a stepwise regression algorithm, a biggest obvious feature set is extracted from the microblog feature set to achieve feature selection; a personality prediction regression model is established for the selected features, and personality psychology indexes of the users are predicted.

The invention belongs to the technical field of agricultural product quality analysis, and particularly relates to a near-infrared detection method for the peanut quality and application. The method comprises the following steps that peanut samples are collected, physical and chemical testing is performed on the peanut samples, near-infrared scanning is performed on the peanut samples, denoising processing and preprocessing are performed on obtained light absorption values, obtained preprocessed light absorption values are analyzed, near-infrared spectrum characteristic wavelengths are obtained through screening, and a prediction model of the peanut quality is built through a stepwise regression method. According to the near-infrared detection method for the peanut quality and the application, the obtained information is intuitive and reliable, the characteristic wavelengths of the peanut quality are determined and are few in number, and the analytical method that the model is built through the characteristic wavelengths is applied, so that the model precision is improved; on the condition of the same prediction precision, the prediction speed is high; meanwhile, through the built near-infrared prediction model method for the moisture, protein, fat, total sugar and ash content of peanuts, the peanut quality can be analyzed more comprehensively, and usage and popularization are easy.

The invention belongs to the technical field of atmospheric remote sensing, and discloses a method for estimating a near-surface atmospheric temperature by using thermal infrared data of a geostationary meteorological satellite. The satellite is used for observing brightness temperature, meteorological station and numerical prediction model data to obtain a representative thermal infrared observation brightness temperature and near-surface atmospheric temperature; satellite cloud detection products are used for obtaining matched data sets for the observation brightness temperature, the stationactually measured temperature and auxiliary data under cloudless conditions; based on a stepwise regression method, the relationship between the radiation temperature of satellite observation, atmospheric pressure, relative humidity, a satellite observation angle, Julian daily parameters and the like and near-surface atmospheric temperature is analyzed, and key factors used for estimating the atmospheric temperature are determined; and a inversion model of near-surface temperature estimation is constructed by using a neural network technology. The method can realize the purpose of inversion of the near-surface atmospheric temperature under the clear sky condition of the thermal infrared data of the stationary meteorological satellite.

The invention discloses an inference method of a stepwise regression gene regulatory network. The method comprises the following steps of: A, reading a gene expression data matrix and a gene perturbation data matrix; B, confirming whether the gene expression data matrix and the gene perturbation data matrix are standardized data or not; C, respectively carrying out data normalization on the gene expression data matrix and the gene perturbation data matrix to form the standardized data; D, analyzing the standardized data and calculating all inter-gene correlation coefficient matrixes; and E, visualizing the inter-gene correlation coefficient matrixes into a network to obtain a gene regulatory network chart. The method can select optimal regression subsets to solve the problem of high-dimension small-sample experimental data, gradually select the most influential regulator for a target gene, accord with the true condition of the gene regulatory network, and be superior to similar methods in calculation precision and calculation efficiency along with the enlargement of the gene regulatory network scale and the network sparsity.

The invention relates to a modeling method of an urban electrical network distribution transform weight overload mid-term forewarning model. The forewarning model is established according to the following steps of firstly, collecting original data of correlated variables required for modeling, and cleaning the original data so that the quality of data entering the forewarning model can be ensured; secondly, designing and calculating characteristic variables of the forewarning model, screening the characteristic variables, and establishing a judgment basis for testing the multicollinearity among independent variables; thirdly, establishing one forewarning model on the basis of Logistic regression and through a stepwise regression method, and then judging whether the multicollinearity exists among the independent variables of the model or not so as to judge whether the model can be used or not; fourthly, repeatedly executing the second step and the third step so that the characteristic variables can be calculated again, and establishing various different forewarning models, evaluating the established forewarning models, and then comparing the evaluation parameters of all the forewarning models to determine the optimal forewarning model; fifthly, outputting the optimal forewarning model. By means of the method, the accurate distribution transform weight overload mid-term forewarning model can be easily established.

The invention discloses a wheat stripe rust predicting method based on a particle swarm and a support vector machine. The method comprises the steps that a stepwise regression method is used for carrying out effective dimension reduction on a high-dimensional sample firstly to eliminate redundancy and reduce the influence of incoherent factors on a forecast object; then the advantages that a PSO algorithm is not likely to fall into local minima, simple in algorithm and small in calculation amount are utilized for optimizing the kernel function parameter g and the penalty factor C of the support vector machine, and an optimal prediction model is obtained fast and efficiently. The method is based on a stepwise regression, PSO and SVM mixed algorithm, wheat stripe rust is accurately and stably predicted and forecast, and a scientific basis is provided for earlier prevention and treatment of wheat stripe rust.

The invention discloses a non-destructive inspection method of total amount of meat bacteria, comprising the following steps: S1, obtaining an original reflection spectrum image Rs of a sample to be detected and storing the image; S2, detecting the TVC of the sample as standard reference data with a standard plate colony counting method according to National Standard; S3, obtaining a relative reflection spectrum image R of the sample according to the FORMULA that R=(Rs-Rd) / (Rr-Rd), wherein Rd is a black image when an image obtaining system works with dark current, and Rr is a reflection spectrum image of a standard reference whiteboard; S4, correcting the image with an SNV method; S5, choosing a second-order differential spectrum for the image R after being corrected, and selecting the best wavelength representing the TVC of the sample to be detected with a step-by-step regression method; S6, extracting the data and TVC detection result of the second-order differential spectrum relative to the best wavelength, and dividing the data and TVC detection result into two sets respectively as correction set data and validation set data; S7, establishing an LS-SVM data model, and detecting the value of the TVC of the sample to be detected. The method is human-oriented and advanced with quick and pollution-free detection process so that our country is in line with developed countries in the field of food quality and safety.

The invention relates to a biomedicine determinant attribute selecting method and belongs to the technical field of biomedicine. The method comprises the steps of analyzing the significance of to-be-selected attributes by using the boruta algorithm firstly and extracting important attributes having influence on research targets; building a logistic regression model by using the to-be-selected attributes and performing stepwise regression by using AIC rules to obtain attributes having remarkable influence on the research targets; for the attributes obtained through the two methods, based on opinions of experts, performing attribute fusion by using an intersection and classification method to obtain final determinant attributes. Two different methods are used for select attributes having influence on research targets and the algorithms are obviously different from each other, so that the limitation of single methods is avoided and the generalization of determinant attributes can be improved.

A computerized data mining method for determining a prediction model for a dependent data mining variable based on a multitude of independent data mining variables, comprising replacement of the independent data mining variable with potential values from a global range by a multitude of independent local data mining variables, each relating to potential values from a sub-range of the global range. The method further comprises an initialization step and a looping sequence that includes steps of determining for every independent local data mining variable not yet reflected in the current prediction model a multitude of partial regression functions, determining for each of the partial regression functions a significance value, selecting the most significant partial regression function and the corresponding not yet reflected local data mining variable, and adding the most significant partial regression function to the current prediction model.

The invention relates to a quantitative method utilizing an equivalent rock mass basic quality index to predict shield driving parameters and relates to the geotechnical engineering and tunnel engineering investigation, design and construction technology field. According to the method, the equivalent rock mass basic quality index is taken as basis, geological segmentation of complex strata is carried out, and segmented statistics of the driving parameters is carried out; experience relationships among the driving rate, cutterhead torque and other driving parameters are calculated through gradual regression to acquire a general driving rate prediction model and a general cutterhead torque prediction model which adapt to uniform strata and the complex strata; through quantitative analysis on the corresponding relationships among a general driving rate prediction model coefficient, a general cutterhead torque prediction model coefficient and the equivalent rock mass basic quality index, the driving parameters are acquired through prediction. The method is advantaged in that the method has great theoretical meaning and engineering application values for construction scheme design, construction cost-construction period control and shield-surrounding rock interaction rule analysis.

The invention discloses a prediction method of television play on-demand amount based on network data. The prediction method is characterized in that the grabbed micro-blog numbers and the search times as well as related data of television plays are calculated by using a correlation analysis and single variable linear regression to obtain an initial features set, then a stepwise regression method is carried out on the initial feature set to obtain a feature set X and a feature set X, a multiple linear regression method is carried out on the feature set X and the feature set X to respectively obtain two prediction models before and after premieres of the television plays, then rankings of the television plays are predicted according to the sizes of predicted values. Compared with the prior art, according to the prediction method disclosed by the invention, the episode average on-demand amount of the television plays in an on-demand system in a period of the future time is predicated in advance, predicted results effectively reflect the popularity degree of the television plays, and the method is simple and good in accuracy, so that the basis can be provided for video operators on a decision of television play broadcast copyright purchase, and the strong support is provided for an online on-demand system on attracting users and increasing advertisement click rate.

The invention discloses a soil moisture and salt information combined extraction method based on hyperspectral data. The method includes the following steps: carrying out primary treatment of remote sensing hyperspectral reflectivity data, and according to the reflectivity, calculating the normalized reflectivity, an apparent absorption rate first-order derivative and an apparent absorption rate second-order derivative; establishing an evaluation system based on principal component analysis, and selecting sensitive wave bands; determining a first inversion variable; with use of the sensitive wave bands, using a stepwise regression method, establishing an inversion model of soil moisture, and partitioning the soil samples according to the predicted soil moisture content, to obtain soil sample intervals; and with the sensitive wave bands and the soil sample intervals, with use of the stepwise regression method, respectively establishing an inversion model of the soil salt content for the soil sample in each interval. The method has the advantages of improving the extraction accuracy of the soil salt.

The invention discloses a construction method of a concrete dam deformation safety monitoring model, specifically, on the basis that by using the historical data of dam deformation observation, a stepwise regression model is established, a shuffled frog leaping algorithm (SFLA) with local optimization performance and global optimization performance is adopted to determine a weight coefficient of each sub-model, an inversion analysis method of dam prototype data is utilized to determine physical and mechanical parameters of a dam, a frog leaping hybrid model is established, and then a concretedam deformation safety monitoring model is obtained. According to the method, the residual error is analyzed and predicted by using the chaos theory, and the residual error prediction item is added tothe leapfrog prediction hybrid model, so that the problem that the fitting effect is good and the prediction result is poor due to the fact that the influence of the fitting residual error is not considered in a conventional dam displacement monitoring model is effectively solved.