47 results about "Posterior probability density function" patented technology

The invention relates to a CVT error evaluation method and device, error evaluation equipment and a storage medium, and belongs to the technical field of mutual inductor error recognized.The CVT error evaluation method comprises the steps that a first data set is restored to a primary voltage level based on voltage proportionality coefficients of multiple CVTs to obtain a second data set, and a difference matrix is constructed based on the second data set; establishing a likelihood function based on the difference matrix, and converting the likelihood function into a posterior probability density function based on a Bayesian formula; extracting data corresponding to a preset number of CVTs from the second data set to construct a third data set, and determining a prior probability density based on the third data set; and based on a multi-chain MCMC algorithm fused with plant rhizome growth and a posterior probability density function, determining error values corresponding to a plurality of CVTs when the posterior probability density is maximum. According to the method, the accuracy of CVT error evaluation is effectively ensured, and meanwhile, the error evaluation efficiency is also improved.

The invention belongs to the technical field of underwater integrated navigation. The invention provides a multi-navigation-state unmanned system underwater integrated navigation method based on global information fusion. According to the embodiment of the invention, different navigation combination modes under different navigation states are selected, and an SINS / DVL / GPS combined navigation mode is adopted offshore, so that the calculated amount is reduced, and the efficiency is improved; and the deep sea sneaking adopts an SINS / DVL / USBL combined navigation mode. Measurement deviation accumulative divergence caused by navigation state change in maneuvering characteristics of the underwater unmanned system is considered, t distribution is introduced in a federated Kalman filtering architecture for trailing characteristics to replace conventional normal distribution, thick-tail measurement noise is modeled, approximate solution of a posterior probability density function is carried out by using a variational Bayesian theory, and the probability density of the underwater unmanned system is calculated. Accurate estimation of the motion state of the unmanned system is realized, and updating and data fusion of a plurality of sub-filters are realized by using a federal structure.

The invention provides a fracturing crack real-time monitoring method and system based on time-varying electric field dynamic data, and the method specifically comprises the steps: obtaining an electric field data sequence which changes with time and is collected by an electric field sensor disposed in a fracturing work area; for each monitoring time step k in the electric field data sequence, executing the following sequential Bayesian inversion: a, constructing state parameter prior distribution of the time step k by adopting kernel density estimation based on a sample of crack state parameter posteriori distribution of the time step k-1; b, establishing a target posterior probability density function composed of prior distribution, a likelihood function and a physical regularization item; c, in Markov chain Monte Carlo iteration, sampling is carried out on the target posterior probability density function; and d, after iteration is finished, taking a sample set output by the Markov chain as crack state parameter posteriori distribution of the time step k, and extracting crack three-dimensional geometrical morphology and distribution information from the sample set.

The invention discloses a GNSS (Global Navigation Satellite System) and INS (Inertial Navigation Satellite System) tight coupling method and system. The method comprises the following steps: acquiring first state distribution of a first target system, wherein the first target system is a GNSS and INS tight coupling system; presetting a first algorithm, and performing state estimation operation based on the first state distribution in combination with the first algorithm; and correcting the first target system according to the state estimation operation result. By adopting the particle filtering technology, the nonlinear noise problem of the GNSS signal in a high electromagnetic interference environment can be effectively solved. Particle filtering serves as a recursive Bayesian filtering technology based on a Monte Carlo method, probability distribution is represented through a group of random samples (particles), and a posterior probability density function in a nonlinear and non-Gaussian noise environment can be well approached. Therefore, even under the condition that the GNSS signal quality is obviously reduced, the method provided by the invention can also provide a more accurate and robust positioning result.

There is provided a method for determining a physical property of a subsurface target volume, as well as systems and computer readable media for performing the same. Unlocking insights from Geo-Data, the present invention further relates to improvements in sustainability and environmental developments: together we create a safe and liveable world. The method comprises performing tomographic inversion on obtained empirical travel time data. Performing tomographic inversion comprises obtaining a posterior probability density function corresponding to each cell of the first plurality of cells, wherein obtaining the posterior probability density function comprises obtaining a prior velocity model and defining a prior probability density function indicative of velocity for each of the cells of the first plurality of cells, for each receiver pair of the subset, determining a modelled travel time using the initial velocity model values associated with the two or more cells that are traversed by the surface wave path and determining a posterior probability density function of the velocity model based on the modelled travel times and the prior probability density function wherein the prior probability density function is dependent on a distance of each of the cells from the first point on the surface.

The application provides a fracturing fracture real-time monitoring method and system based on time-varying electric field dynamic data, and specifically, time-varying electric field data sequences collected by electric field sensors arranged in a fracturing working area are acquired; for each monitoring time step k in the electric field data sequences, the following sequential Bayesian inversion is performed: a. based on samples of a fracture state parameter posterior distribution of a time step k-1, a state parameter prior distribution of a time step k is constructed by using kernel density estimation; b. a target posterior probability density function composed of a prior distribution, a likelihood function and a physical regularization term is established; c. in Markov chain Monte Carlo iteration, the target posterior probability density function is sampled; d. after iteration, a sample set output by the Markov chain is taken as a fracture state parameter posterior distribution of the time step k, and fracture three-dimensional geometric morphology and distribution information is extracted from the sample set.

The application relates to a weak underwater target azimuth detection pre-tracking method based on belief propagation, which comprises the following steps: converting a weak target tracking problem into solving a joint posterior probability density function (pdf) based on a Bayesian rule; factorizing the joint posterior pdf, constructing a corresponding factor graph through a graph model; solving the transmitted information in the factor graph by using a belief propagation (BP) algorithm, and converting multiple integrals into ordinary integrals by using a Goodman variable substitution principle; constructing a likelihood function by using a complex Wishart distribution to describe the relationship between original sonar data and target states, and simplifying the remaining pdf by using a Gaussian distribution to obtain a confidence approximation of a target state edge posterior pdf; and calculating the number of weak targets and corresponding states by using a minimum mean square error (MMSE) estimator. The application can simultaneously track multiple weak targets without a complex clustering algorithm, has fewer preset parameters, and can accurately and efficiently track target state information including an azimuth, an azimuth angular velocity and target intensity.

The invention discloses a communication sensing channel construction method and device, equipment, a storage medium and a program, which are applied to the technical field of wireless communication, and the method comprises the following steps: constructing a quasi-static multipath channel model according to an obtained scattering signal, and determining the prior distribution of a frequency domain channel gain vector of the quasi-static multipath channel model; establishing a sensing channel model and a communication channel model based on the quasi-static multipath channel model; determining a posterior probability density function of a channel gain of the perceptual channel model based on the priori distribution; and carrying out soft fusion on the basis of the posterior probability density function and the communication channel model to obtain the communication sensing channel. According to the embodiment of the invention, fusion of a communication stage and a sensing stage can be realized, signal loss is reduced, communication quality can be improved, sensing accuracy can be enhanced, and the satisfaction degree of a user to a system can be improved.

The application provides a composite sandwich structure material damage identification method based on a Bayesian fusion algorithm, and the method comprises the following steps: extracting the time of flight of a scattered wave signal based on continuous wavelet transform; wherein the time of flight refers to the time for which the scattered wave signal generated by an excitation sensor is transmitted to a receiving sensor; obtaining a likelihood function based on the time of flight and an elliptical positioning method; determining a correlation coefficient based on a damage signal and a reference signal; in a RAPID method, taking the correlation coefficient as a damage sensitivity characteristic to determine a prior probability density function of a damage position; performing Bayesian estimation based on the likelihood function and the prior probability density function to obtain a posterior probability density function of the damage position; and fitting a damage positioning image using a Markov Monte Carlo method (MCMC) based on the posterior probability density function of the damage position.

The invention discloses a target fusion method based on average field belief propagation under an unknown time-varying parameter condition, and belongs to the technical field of radar target fusion tracking. The method aims at solving the problem that an existing target fusion algorithm cannot achieve target fusion under the condition of unknown time-varying parameters. Comprising the steps of calculating a marginal probability of a target existence probability based on a Markov law; setting an associated data set and determining constraint conditions that the associated data of each sensor comes from a target and clutter; calculating prior probability distribution; calculating a joint posterior probability density function based on the conditional probability; establishing a factor graph model based on average field-belief propagation, reconstructing a joint posterior probability density function and calculating a belief value; solving a target motion state, a target existence probability, a measurement noise covariance and a correlation variable; and iteratively solving until the difference value of the target motion states in two adjacent calculations is smaller than a preset threshold value, and taking the current target motion state as a final target fusion state. The method is used for target fusion under the unknown time-varying parameter condition.

The invention provides an optimization method and device for a sensor array, and the method comprises the steps: building a linear array model, carrying out the discretization of a parameter space of an array element position, and generating an array arrangement position grid; acquiring a plurality of sampling points in a sampling space of a preset visual area; constructing a first sensing matrix according to the array position grid and the plurality of sampling points, and establishing a sampling signal model containing sparse signals according to the first sensing matrix; based on a preset multilayer prior model and a posterior probability density function, determining a posterior probability mean value of the sparse signals, and updating the sparse signals in the sampling signal model according to the posterior probability mean value; and performing linearization processing on the compressed sensing vector deviation corresponding to the updated sampling signal model through first-order Taylor approximation, constructing a target gradient matrix, correcting the first sensing matrix according to the target gradient matrix to obtain a second sensing matrix, and obtaining a target array arrangement position according to the second sensing matrix.

The application discloses a direct positioning method and system based on a single-bit signal, and the method comprises the following steps: a single-bit model and a space sparse model of a received signal are established; a probability density function of a single-bit received signal in the single-bit model of the received signal is obtained according to a prior distribution of noise, a prior distribution of observation data and a Laplace prior distribution; a mean value and a variance expression of a space sparse signal and a maximum posterior probability density function of a target radiation source position are obtained; an optimal target function of a hyperparameter is obtained; the optimal target function of the hyperparameter is solved to obtain an updating expression of the hyperparameter; the updating expression of the hyperparameter and the mean value and the variance expression of the space sparse signal are alternately iterated and solved to obtain the mean value and the variance of the space sparse signal, and the space position of the target radiation source is obtained according to the mean value of the space sparse signal. The application can reduce signal transmission cost and effectively improve the positioning speed and precision of multi-target radiation sources.

The invention discloses a combined self-adaptive parallel annealing source item inversion method for determining a nuclear radioactive source. The method comprises the following steps: acquiring a nuclear pollutant concentration measured value and meteorological data of each monitoring point; establishing an SRS (Sounding Reference Signal) matrix by using a reverse calculation diffusion process of the meteorological data by adopting an adjoint method; establishing an adaptive inversion model, including modeling a posterior probability density function of a source item parameter and determining a sampling method, the sampling method adopting an enhanced MCMC method, the enhanced MCMC method adopting an MCMC method and fusing a joint adaptive jump function and a parallel annealing algorithm, the joint adaptive jump function being a weighted set comprising an AM algorithm, an SCAM algorithm and a DE algorithm; and taking the SRS matrix and the pollutant concentration measured value as input data of an algorithm, substituting the input data into an inversion model for calculation, and performing inversion calculation to obtain nuclear radioactive source information. According to the method, the source item is quickly and stably reconstructed, and an effective calculation means is provided for source item determination and accident consequence evaluation.

This invention discloses a robust relative navigation method for aircraft based on a hybrid distribution under non-Gaussian noise, belonging to the technical field of computation, estimation, or counting. To address the problem of filter divergence caused by non-stationary heavy-tail noise in time-varying environments during relative navigation, this invention introduces a Dirichlet random mixture vector fusion of Gaussian, Student's t, and multivariate K-distributions, proposing a Gaussian-Student's t-multivariate K-distribution modeling of measurement likelihood. Then, by minimizing the KLD of the true posterior probability density function and the approximate posterior probability density function using variational Bayesian techniques, the approximate posterior estimates of the aircraft's relative motion state and filter parameters are obtained, yielding the target's state information relative to the aircraft and solving for relative position and velocity. Finally, a nonlinear filter based on the Gaussian-Student's t-multivariate K-distribution is derived to improve relative navigation accuracy for angle-only relative navigation of aircraft in time-varying environments.

The present application relates to a kind of CVT error evaluation method, device, error evaluation equipment and storage medium, belong to mutual inductor error identification technical field, wherein, the CVT error evaluation method includes: based on the voltage proportionality coefficient of multiple CVTs, first data set is restored to primary voltage level and obtains second data set, and difference matrix is constructed based on second data set;Based on difference matrix, likelihood function is constructed, and likelihood function is converted into posterior probability density function based on Bayes formula;From second data set, the data corresponding to the data of the extraction of the preset number of CVT is constructed third data set, and prior probability density is determined based on third data set;Based on the multi-chain MCMC algorithm of fusion plant rhizome growth and posterior probability density function, when the posterior probability density maximum, the error value corresponding to multiple CVTs is determined.The present application effectively guarantees the accuracy of CVT error evaluation, also improves the efficiency of error evaluation.

Methods for determining a physical property of a subterranean target volume, as well as systems and computer-readable media for performing the methods, are provided. According to the method, clues are mined from geographic data, and the method further relates to the improvement of sustainability and environmental development: we create a safe and livable world together. The method includes performing tomography inversion on the obtained empirical propagation time data. Performing the tomographic inversion includes obtaining a posterior probability density function corresponding to each of the first plurality of cells, where obtaining the posterior probability density function includes obtaining a prior velocity model, and defining, for each receiver pair in the subset, a prior probability density function indicative of a velocity of each of the first plurality of cells, a modeled propagation time is determined using initial velocity model values associated with two or more cells traversed by the surface wave path, and a posterior probability density function of the velocity model is determined based on the modeled propagation time and a prior probability density function, wherein the prior probability density function depends on the distance of each cell from a first point on the surface.

The invention discloses a single-base MIMO radar parameter joint estimation information amount calculation method, and the method comprises the steps: constructing a multi-dimensional joint probability density function of a uniform linear array through the property of an array receiving signal covariance matrix, and simplifying the joint estimation probability density function according to the vector form of a receiving signal; deducing a joint posterior probability density function of distance angle estimation under the condition of given received signals according to a Bayesian formula in combination with the joint probability density function, and obtaining distance direction information of multiple targets according to a distance-direction information formula; and deducing the distance direction information by using Taylor expansion under the condition of a high signal-to-noise ratio to obtain an upper bound of the distance direction information, and obtaining an entropy error, an entropy error lower bound and a Cramer-Rao bound according to the information amount and the information amount upper bound. According to the method, a posterior probability density function used for distance and angle joint estimation of multiple targets of the uniform linear array is deduced based on the information theory, and an information structure in radar observation data can be more completely described through joint distance-angle parameter modeling and mutual information theory analysis. Through introduction of mutual information indexes in an information theory, the perception capability of a radar system to target parameters can be quantified from a statistical perspective, and a theoretical basis is provided for system design.

The application provides a regional crop yield estimation method and system based on particle filtering data assimilation, which comprises the following steps: running each particle at a fixed time interval during the critical growth period of crops until each particle runs to the mature period of crops, and obtaining the first LAI output by each particle; updating the initial weight of each particle according to the posterior probability density function, the updated first LAI of each particle and the likelihood function, and obtaining the updated weight of each particle; resampling each particle in the initial particle set and updating the initial particle set under the condition that the particle divergence is less than a preset value; and obtaining the estimated value of the crop yield according to the crop yield and the maximum weight of each particle under the condition that each particle runs to the mature period of crops. The application can improve the sequential sampling efficiency of particles, reduce the resampling frequency, maintain the diversity of particle input parameters, and improve the speed and accuracy of crop yield prediction.

The invention provides a non-reference damage imaging method based on nonlinear Lamb waves and Bayesian inference, and relates to the field of damage imaging, and the method specifically comprises the following steps: collecting nonlinear Lamb wave signals of a plurality of propagation paths; an enhanced second harmonic signal is obtained through a pulse inversion technology; extracting a flight time feature of the envelope signal; the frequency spectrum of the nonlinear Lamb wave signal is obtained through fast Fourier transform, and nonlinear coefficient damage index features are obtained through calculation; feature level fusion is carried out based on Bayesian estimation; and solving a posterior probability density function of the damage position parameter by using Hamiltonian Monte Carlo sampling, and carrying out damage probability imaging positioning according to posterior probability distribution. According to the technical scheme, the problems that in the prior art, a large number of sensors usually need to be deployed, and the dual requirements for high-precision damage detection and sensor number minimization in engineering practice are difficult to meet are solved.

This application discloses a method, apparatus, storage medium, and product for determining the distribution probability of oil and gas reservoirs based on seismic frequency-varying fluid factors, relating to the field of oil and gas exploration technology. The method includes: acquiring pre-stack data of the exploration area and performing pre-stack AVO inversion to obtain initial frequency-varying velocity vectors; constructing a posterior probability distribution of the objective function based on the convolution model and propagation matrix model in seismic wave propagation theory and Bayesian theory; obtaining the posterior probability of the objective function using the Markov chain Monte Carlo method; and determining the frequency-varying velocity vector at the maximum of the posterior probability density function; thereby determining the dispersion gradient of new fluid factors at different frequencies at the dominant seismic wave frequency to determine the distribution probability of oil and gas reservoirs; the dispersion gradient of the new fluid factors is positively correlated with the probability of the existence of oil and gas reservoirs; the parameters constituting the new fluid factors include: Poisson's ratio, P-wave velocity impedance and S-wave velocity impedance, Lamé coefficient, and density. This method can significantly improve the accuracy of oil and gas reservoir identification.

The present invention proposes an independent fusion modeling and response prediction method for low- and high-precision missile performance data. The method utilizes collected missile reliability test data of different accuracies to respectively construct a low-precision response model and a high-precision response model. Based on Bayesian theory, the posterior probability density distribution of each model parameter, and the high-precision response model, the posterior probability density function of the high-precision missile performance response data at the influencing factor data of any application scenario is obtained, and then a high-precision missile performance response data prediction model at the influencing factor data of any application scenario is obtained. The method can predict the high-precision missile performance response data at the influencing factor data of any application scenario, and ensure the accuracy of the response prediction result and the efficiency of the solution operation.

The invention discloses a single-base MIMO radar multi-target DOA estimation method, and belongs to the field of signal processing. According to the method, a Shannon information theory method is combined, and a joint conditional probability density function of a received signal of the MIMO radar, a target direction angle and a target reflection coefficient is deduced by utilizing the property of noise in the received signal; according to a Bayesian formula, firstly, a posterior probability density function of a single-base MIMO radar multi-target reflection coefficient based on a uniform linear array is derived, and a posterior probability density function of a multi-target direction angle when a signal is received is further obtained by combining a derivation result. The azimuth angles of a plurality of targets can be obtained at the same time by performing spectral peak search on the posterior probability density function of the target direction angle. The invention provides a DOA (direction of arrival) detection method based on a posterior probability density function and aiming at the condition that a plurality of targets are densely distributed. According to the method, the posterior probability density function used for uniform linear array monostatic MIMO radar multi-target DOA estimation is deduced based on the information theory, all available information is organically combined through the Bayesian theory, and therefore more comprehensive angle estimation is provided.

The embodiment of the invention provides a noise filtering method and system for a sensor, and the method comprises the steps: carrying out the modeling of the noise statistical characteristics of a signal of the sensor through employing Gaussian-Gaussian-Gaussian inverse index mixed distribution for a sensor used in an automatic driving scene, obtaining a noise model, and carrying out the calculation of the noise model based on a system state equation and a measurement equation, constructing a joint probability density function of a system state, and forming the joint probability density function into a joint posterior probability density function by adopting a variational Bayesian method; solving an optimal approximate probability density function through a KL divergence between a minimum approximate probability density function and a joint posterior probability density function, and alternately updating q (xk), q (yk), q (pi k) and q (lambda k) by using a fixed point iteration method until iteration convergence; and obtaining the optimal estimation value of the system state xk according to the converged approximate probability density function. The fitting degree of noise description is improved from the source, and precise adaptation and efficient suppression of multi-mode noise of the sensor are achieved.

This invention discloses a method, apparatus, and equipment for pre-stack elastic parameter inversion based on the converted wave domain. The method includes: determining the converted wave incident angle distribution based on the P-SV wave pre-stack offset gather, converted wave velocity, P-wave / S-wave velocity ratio, and converted wave travel time; performing angle-separated stacking of the converted wave incident angle distribution to obtain angle-separated stacked gathers; extracting data from the angle-separated stacked gathers to obtain wavelets at different angles; determining seismic traces of the incident angle based on the theory of P-SV wave amplitude variation with incident angle, the seismic synthetic record of the convolution model, and the wavelets at different angles; determining the relationship between observed seismic data and elastic parameters based on the seismic traces; and determining the inversion results of the elastic parameters based on the relationship between observed seismic data and elastic parameters, prior geological information, and the posterior probability density function. This method enables direct inversion of reservoirs in the converted wave domain, reducing computational load and improving inversion reliability.

The invention discloses an oil and gas reservoir distribution probability determination method and device based on a seismic frequency change fluid factor, a storage medium and a product, and relates to the technical field of oil and gas exploration. The method comprises the following steps: obtaining pre-stack data of an exploration area and carrying out pre-stack AVO inversion to obtain a frequency change speed initial value vector; based on a convolution model and a propagation matrix model in a seismic wave propagation theory and a Bayesian theory, constructing posterior probability distribution of an objective function, solving a posterior probability of the objective function through a Markov chain Monte Carlo method, and determining a frequency-dependent velocity vector at a maximum position of a posterior probability density function; determining the dispersion gradient of the new fluid factors with different frequencies at the seismic wave main frequency so as to determine the distribution probability of the oil and gas reservoir; the frequency dispersion gradient of the new fluid factor is in positive correlation with the possibility of existence of the oil and gas reservoir, and the composition parameters of the new fluid factor comprise Poisson's ratio, longitudinal wave velocity impedance, transverse wave velocity impedance, Lame coefficient and density. The method can greatly improve the identification precision of the oil and gas reservoir.

The invention discloses a generalized label multi-Bernoulli tracking filtering method for adaptive scene parameters, and relates to the field of multi-target tracking. At present, target tracking has the problem of too high operation complexity. The method comprises the following steps: generating new multi-target distribution required by a PHD filter and a GLMB filter according to a self-adaptive new model by using measurement data; measured data and a new target intensity function are sent to an unknown scene parameter estimator based on GM-PHD filtering, and estimation of detection rate and clutter intensity parameters is carried out; performing GLMB filtering by using the measured data and the scene parameters and new target distribution obtained by estimation to obtain a posterior probability density function of the iteration; and finally, extracting a multi-target tracking result from the posterior density. According to the technical scheme, efficient and accurate multi-target tracking under unknown scene parameters is realized by improving a beta Gaussian component merging process, combining an unknown clutter intensity filtering technology, combining a PHD filtering algorithm and a GLMB filter and other technical means.

The application discloses a method for predicting seismic spatial structure parameters, comprising the following steps: reading observed seismic data; defining a prior probability density function of spatial structure parameters based on a variogram type of model parameters and the number of spatial structure parameters; defining a seismic forward operator and calculating a likelihood function between the spatial structure parameters and the observed seismic data based on the seismic forward operator; and calculating a posterior solution of a posterior probability density function of the spatial structure parameters based on the prior probability density function and the likelihood function through a Monte Carlo sampling method. The application also discloses a device for predicting seismic spatial structure parameters, a computer device and a readable storage medium. The application combines the Bayesian theory with the Monte Carlo inversion method, fully utilizes the volume constraint characteristics of seismic data, and simultaneously infers the longitudinal and lateral spatial structure of the underground medium in a data-driven manner.

The invention provides a space manipulator actuator load optical fiber reconstruction method and system based on self-adaptive non-negative Bayesian regularization and application, and relates to the technical field of load reconstruction. The method comprises the following steps: arranging a plurality of fiber grating sensors on an end effector of the space manipulator; constructing a dynamic load model representing a reverse mapping relation between the external load of the end effector and the optical fiber strain response; constructing a load history reconstruction posterior probability density function of the end effector of the space manipulator, and performing variation approximate calculation on the load history reconstruction posterior probability density function; and carrying out iterative solution on the posterior probability density function to obtain dynamic load history data of the space manipulator end effector. According to the method, data fitting residual and parameter posteriori distribution are monitored in real time, regularization intensity is adjusted in a self-adaptive mode, the algorithm can accurately track the dynamic characteristics of the time-varying load, the defect that a fixed parameter method is insufficient in adaptability under complex working conditions is overcome, and high-precision dynamic load process data reconstruction is achieved.

The application provides a non-reference damage imaging method based on nonlinear Lamb waves and Bayesian inference, relates to the field of damage imaging, and specifically comprises the following steps: collecting nonlinear Lamb wave signals of multiple propagation paths; obtaining enhanced second harmonic signals through a pulse inversion technique; extracting time-of-flight characteristics of envelope signals; obtaining frequency spectra of the nonlinear Lamb wave signals by using fast Fourier transform, and calculating nonlinear coefficient damage index characteristics; performing feature-level fusion based on Bayesian estimation; solving a posterior probability density function of a damage position parameter by using Hamilton Monte Carlo sampling, and performing damage probability imaging positioning according to the posterior probability distribution. The technical scheme of the application overcomes the problem in the prior art that a large number of sensors usually need to be deployed, and it is difficult to meet the dual requirements of high-precision damage detection and minimization of the number of sensors in engineering practice.

The present application relates to the technical field of wireless communication, especially to a channel estimation method based on sparse Bayesian learning, comprising the following steps: step A, using the sparse characteristics of the millimeter wave channel to model the channel, and constructing the corresponding RIS assisted MIMO communication system model; step B, first constructing two-layer sparse prior information of the channel, then factorizing the joint posterior probability density function of the parameters, and finally drawing the corresponding factor graph model according to the factorization; step C, for the factor graph model in step B, combining the unitary transformation approximate message passing, and using the sparse Bayesian algorithm framework to perform channel estimation; step D, repeating step C until the algorithm converges; the method in the present application has low complexity, reduces the pilot overhead of the system, and has high algorithm universality.