50results about How to "Improve inversion efficiency" patented technology

The invention discloses a seismic inversion method and system based on generalized total variation regularization, so as to solve the problem that step effects are generated inside a formation when the existing seismic inversion technology adopts total variation regularization. The method uses the characteristic that generalized total variation not only uses first-order partial derivative information of a to-be-inverted parameter but also uses second or higher-order information, the formation interface can be clearly expressed and the step effects inside the formation can also be weakened; besides, a phenomenon that a product of a wavelet matrix, a difference matrix and the natural logarithm of the to-be-inverted parameter can be used to be converted to the convolution among a convolutionkernel corresponding to the wavelet matrix, a convolution kernel corresponding to the difference matrix and the natural logarithm of the to-be-inverted parameter is found out, the convolution operation in a time domain is converted to point multiplication operation in a frequency domain, and the inversion speed is greatly improved; a reflection coefficient does not need to be inverted firstly, anddirect inversion is carried out in the frequency domain to obtain the to-be-inverted parameter; and the method and the system are applicable to the technical field of seismic exploration.

The invention provides a corn canopy LAI and chlorophyll content joint inversion method and corn canopy LAI and chlorophyll content joint inversion equipment. The corn canopy LAI and chlorophyll content joint inversion method comprises: obtaining actual corn canopy spectral reflectivities of multiple spectrums at each growth period; establishing a lookup table based on a PROSAIL model, wherein theinput parameters in the lookup table comprise the leaf area indexes (LAI) and the chlorophyll contents of corn at each growth period, and the output parameters of the lookup table are corresponding simulated corn canopy spectral reflectivities; establishing a cost function based on a weighted square difference sum form, wherein the cost function is used for calculating the error between the simulated corn canopy spectral reflectivity and the real corn canopy spectral reflectivity; and carrying out inversion on the corn canopy LAI and the chlorophyll content based on the corresponding simulated corn canopy spectral reflectivity in the case of the minimum value of the cost function. According to the present invention, the LAI and the chlorophyll are simultaneously inverted with the one setof the parameters by establishing the joint distribution of the corn canopy LAI and chlorophyll content parameters so as to improve the parameter inversion efficiency.

The invention belongs to the field of near-surface seismic surface wave exploration, and particularly discloses a Rayleigh wave dispersion curve accurate and fast inversion method for seismic surfacewave exploration. The method includes converting a seismic record from an x-t domain to an f-k domain to obtain f-voR through two-dimensional Fourier transformation of the seismic record; according tothe f-voR, randomly generating an initial population; calculating phi (Pj), converting the phi (Pj) into f (Pj), and recording that the initial iteration number of inversion is 1; according to the f(Pj), preforming selection operation and cross operation; based on a cross population, selecting the top 10% individuals of f (Pj), performing local linear optimization, and acquiring a new populationPj (x) through the fastest descent; and based on the new population, repeating the above steps until T = T+1 reaches the preset iteration number, and selecting the individual with the largest fitnessvalue as an inversion result according to the f (Pj). The method can improve the applicability, rapidity and accuracy of Rayleigh wave exploration in detecting near-surface stratum structures.

The invention discloses a fast two-dimensional density inversion method based on a physical property function. The method comprises the following steps of: S1, collecting gravity data by field actual measurement; S2, determining the physical property function order of an inversion area according to the gravity data or the existing data; S3, transversely subdividing the inversion area in the underground half space according to the distribution of the physical property function; S4, establishing a corresponding kernel function matrix; and S5, performing forward modeling on the parameter inversion result, and checking the inversion fitting degree. The fast two-dimensional density inversion method introduces the physical property function, reduces the memory occupation during operation and improves the operation speed; a construction mode of an underground half-space density model when a physical property function is introduced; the inversion speed is improved; and less memory space is occupied in the calculation process.

The invention discloses a shale gas identification method based on deepest descent prestack precise equation fluid inversion. The method is characterized in that an objective function based on fluid identification parameters is constructed through the Zoeppritz equation and a porous medium rock-physics model, and an iterative equation is obtained through deduction and inversion. The method has theadvantages that prestack inversion is directly carried out through the Zoeppritz equation, so that limitations for small angles caused by a Zoeppritz approximate expression can be avoided, and large-angle seismic materials can be better utilized; through the deepest descent algorithm, the convergence efficiency can be improved, and the calculation load can be reduced; and the deepest descent inversion is carried out through the non-simplified Zoeppritz equation, so that estimated values of the fluid identification parameters can be accurately and rapidly obtained, and a more reliable basis can be further provided for shale gas identification.

The invention discloses a method for inverting underground goaf position parameters by fusing InSAR and PSO, which is suitable for the field of mineral geological monitoring and protection. The methodcomprises steps of utilizing an InSAR method to obtain ground surface deformation of the mining area radar sight line, determining a value range of goaf parameters, initializing the position and speed information of each individual in the particle swarm, substituting the position coordinates of the particle swarm into a probability integral model to obtain vertical subsidence and east-west and north-south horizontal movement of the earth surface, and converting deformation into LOS-direction deformation according to radar imaging geometry, substituting the converted LOS directional surface deformation and the LOS directional surface deformation calculated by the InSAR into a fitness function to calculate a fitness value, and obtaining and recording an optimal solution of each particle anda globally optimal solution of a group, and solving a fitness value corresponding to the optimal solution of the population in iteration. The method is easy to operate, the effect of inverting the position parameters of the underground goaf is good, technical support can be provided for illegal mining of underground coal resources and inversion of the position of a coal fire area, and meanwhile the application space of the InSAR technology is expanded.

An unsaturated soil hydraulic characteristic parameter numerical value inversion method based on an evaporation process comprises the following steps: 1) performing an indoor soil body drying test, monitoring the mass change of a soil body, obtaining an actual evaporation capacity process line of the soil body, and recording related environment variable data; 2) selecting a Fredlund model to fit a hydraulic conductivity and soil-water characteristic curve, wherein model parameters are to-be-inverted parameters; 3) establishing a corresponding finite element analysis model according to the characteristics of a soil sample; 4) calculating the surface evaporation capacity of the soil body according to a migration equation of the gas-liquid two-phase flow under the action of a hydrothermal gradient and a Penman-Wilson formula; 5) determining a target function F (X), wherein the F (X) is the error sum of squares of the actually measured evaporation capacity and the calculated evaporation capacity; 6) according to the target function, using a whale optimization algorithm to invert the parameters of the Fredlund model. According to the method, the inversion precision is improved, meanwhile, the obtaining difficulty of actually measured data needed by inversion is reduced, the inversion efficiency is improved, and the method is composed of numerical simulation evaporation process forward modeling analysis and an optimization algorithm.

The invention discloses a snow multi-angle bi-directional reflectivity inversion method based on satellite remote sensing data, and the method comprises the following steps: step 1, collecting remote sensing data, carrying out the preprocessing, and obtaining the snow surface single-angle bi-directional reflectivity of different wavebands; step 2, eliminating the terrain effect of the bi-directional reflectivity calculated according to the step 1 by using a C correction model; step 3, according to a calculation result of the step 2, constructing a snow white space albedo inversion model based on a progressive radiation transfer model; 4, according to a result of the step 3, constructing a snow multi-angle bi-directional reflectivity inversion model; and step 5, inverting the snow white blank albedo by using the snow multi-angle bi-directional reflectivity inversion model, and then inputting a related angle value to invert the snow multi-angle bi-directional reflectivity. The method can perform continuous multi-angle bi-directional reflectivity inversion on the snow in the large-range mountainous area, and has the advantages of no need of multi-day data, high efficiency and small influence of angle effect.

The invention provides a method for inversion from an electrical property of a vertical interconnection structure to a process parameter thereof. The method comprises: using a transmission resonance frequency of the vertical interconnection structure and a corresponding S11 value as electrical parameter input values, and using an inter-layer dislocation of the vertical interconnection structure as output, to establish a prediction model of inversion from the electrical property to a structure; then using an inter-layer dislocation value of the interconnection structure as input, and using a sintering process parameter as output, to establish a prediction model of inversion from a structural dislocation to the process parameter, wherein the inversion method uses an multi-core support vector machine algorithm integrating prior knowledge, so as to obtain an accurate prediction model with a small sample by training; and successively inputting an electrical property index of the interconnection structure into the foregoing two inversion models, to predict the process parameter that needs be set when processing the vertical interconnection structure. The inversion method provided by the invention can improve the process parameter of the designed vertical interconnection structure, so as to ensure that the final vertical interconnection structure can meet requirements of the electrical property index.

The invention discloses a two-parameter synchronous inversion method for a three-dimensional frequency domain ground penetrating radar. The method comprises the steps: inputting observation data, setting a three-dimensional initial model, and initializing parameters; performing forward modeling on the established three-dimensional initial model; constructing an inversion objective function, and decomposing the inversion objective function into two alternate sub-problems; initializing the dielectric constant and the conductivity to obtain an initial pilot model vector; solving a sub-problem 1 according to the pilot model vector, and updating a model medium parameter vector; solving a sub-problem 2 according to the obtained model medium parameter vector, and updating a pilot model vector; and updating synchronously and alternately until an inversion termination condition is met, and outputting an inversion result. According to the invention, the dielectric constant and the conductivity value can be updated synchronously, so that the inversion efficiency is improved; a total variation model constraint method is improved to a full waveform inversion process, and the robustness of an algorithm is increased; and a frequency weight factor is combined, and a parallel mode is adopted to realize 3D GPR quantitative imaging, so that the imaging process is more efficient and accurate.