62 results about "Akaike information criterion" patented technology

The invention discloses a coal and rock mass fracture seismic source and stress field joint inversion method and system, and the method comprises the following steps: arranging a multi-channel sonic sensor in a coal and rock mass, and collecting original acoustic emission / microseismic signal data; the head wave arrival time / amplitude of the filtered effective waveform data is automatically picked up through an improved AIC (Akaike Information Code), quantitative evaluation is carried out on the quality of a picked-up result, and high-precision positioning is carried out on waveform signals with a high signal-to-noise ratio and a low signal-to-noise ratio; constructing a tension fracture seismic source model based on a displacement discontinuous tensor, obtaining a moment tensor component of a fracture event, and calculating to obtain a spatial orientation, a tension-shear attribute, a deviation angle alpha, energy release and a moment magnitude of the fracture; establishing a stress field inversion model of the tension-shear fracture seismic source, and realizing the inversion of the stress direction and size of the tension-shear fracture seismic source; and a slip direction included angle minimization objective function is established for iteration until convergence conditions are met, and joint iteration inversion of the fracture and the stress field is realized.

The invention discloses a micro-seismic signal arrival time pickup method based on fuzzy clustering and variational mode decomposition. The method comprises the following steps: firstly, calculating characteristic functions of attribute characteristics such as a micro-seismic signal mean value, power and kurtosis and normalizing the characteristic functions to obtain a characteristic matrix, then primarily picking up a micro-seismic P-wave initial movement position and a first arrival moment through a fuzzy clustering algorithm, and extracting an effective time window by taking the first arrival moment as a reference point; a variational mode decomposition algorithm is utilized to decompose signals in a time window into K intrinsic mode function components, AIC function values of the components are calculated by means of an akaike information criterion algorithm, a minimum value point is picked up to serve as first arrival time, energy ratios of all the components are calculated, and final micro-seismic P-wave first arrival time is obtained through weighted calculation. The method effectively deals with the low signal-to-noise ratio environment of the underground coal mine, has higher pickup precision and reliability compared with a traditional pickup method, can provide accurate micro-seismic occurrence time and position information for mine safety early warning, and powerfully guarantees the safety production of the coal mine.

The invention provides a structural damage detection method and device for a nonlinear system, and relates to the technical field of nondestructive detection. The method comprises the following steps: applying excitation force to a to-be-detected test piece, vibrating the to-be-detected test piece based on the excitation force, performing signal acquisition on the to-be-detected test piece to obtain an excitation signal and a response signal, constructing initial NARX models of different orders by combining the excitation signal and the response signal with a plurality of imported orders respectively, and calculating the initial NARX models according to the initial NARX models. And screening a target NARX model from the plurality of initial NARX models based on an akaike information criterion, optimizing the target NARX model, comparing order information in the NARX model to be detected obtained by optimization with order information in the reference NARX model, and determining whether the test piece to be detected has defects according to an order comparison result. The nonlinear model is constructed and optimized through the signal data of the to-be-detected test piece, so that the model accurately represents the structural characteristics of the to-be-detected test piece, and then whether the to-be-detected test piece has defects or not is judged.

The invention provides a spatial variability soft rock tunnel deformation prediction method based on deep learning, and belongs to the technical field of tunnel deformation prediction, and the method comprises the steps: constructing a surrounding rock parameter database, and determining the optimal edge distribution of different surrounding rock parameters through employing an akaike information criterion; constructing a reference value initial model, and generating a related standard uniformly distributed random field; performing equal probability transformation on the correlation standard uniformly distributed random field to generate a cross-correlation non-Gaussian distributed random field; encoding the data matrix of the non-Gaussian distribution random field into a parameter field image; and performing soft rock tunnel deformation prediction on the parameter field image by adopting a trained deformation prediction model. According to the method, parameter independent hypothesis limitation is broken through, the multi-parameter coupling space variability is accurately represented, and the prediction precision is remarkably improved.

The invention relates to the technical field of GNSS deception jamming resistance, and particularly discloses a GNSS induced deception jamming resistance method and system, and the method comprises the steps: obtaining a GNSS signal, and calculating a code phase difference and a carrier phase difference; traversing a modal component number K value, performing variational modal decomposition on the code phase difference according to each K value to obtain a modal component, calculating a criterion value of a corresponding Akaike information criterion, and selecting K which enables the criterion value to be minimum as an optimal modal component number; performing variational mode decomposition on the code phase difference and the carrier phase difference based on the optimal mode component number; modal components containing the induced cheating information are screened out, and a screening result is obtained; and carrying out positioning calculation on the GNSS signal based on a screening result. The method is obviously superior to the prior art in the aspects of dynamic adaptability, feature separation precision and multi-information fusion, and a more efficient and robust solution is provided for GNSS induced deception jamming resistance.

The invention discloses a multi-objective optimization synergistic orthotropic steel bridge deck slab damage acoustic emission positioning method. Comprising the steps that 1, acoustic emission signals of the orthotropic steel bridge deck in the working state are collected through acoustic emission sensors on the four vertexes of a rectangular area to be detected; 2, deducing a shortest path formula for transmitting the acoustic emission signals from potential damage sources at different positions to the acoustic emission sensor; 3, the arrival time of a first arrival mode of the acoustic emission signals is extracted according to the Akaike information criterion, a frequency dispersion curve of the acoustic emission signals is obtained, the characteristic frequencies and wave velocities of the first arrival mode and the A0 mode are determined respectively, and the arrival time of the A0 mode of the acoustic emission signals is determined; 4, establishing a multi-objective optimization model for orthotropic steel bridge deck damage positioning according to the determined multi-mode arrival time and wave velocity parameters; and 5, solving the multi-objective optimization model through a hierarchical sequence method by taking the first layer of the multi-objective optimization model as an important objective function, and evaluating the position of the damage. According to the invention, the damage position can be accurately identified.

The invention discloses a cascade water-wind-light multi-power system modeling method based on time-space correlation, and the method comprises the steps: building an autoregression integral moving average-generalized autoregression condition heterovariance model of wind power photovoltaic power and cascade hydropower station runoff, building an R-vine, C-vine and D-vine Copula function spatial correlation model through employing a generalized autoregression condition heterovariance value, and carrying out the modeling of a cascade water-wind-light multi-power system. And selecting an optimal spatial correlation model according to a Bayesian information criterion, an akaike information criterion and a logarithm likelihood value goodness of fit index. Based on the optimal spatial correlation model, constructing a scene generation method of Latin hypercube-Copula function sampling, performing scene generation on the wind power photovoltaic power and the runoff volume of the cascade hydropower station, and constructing a scene reduction method of contour coefficient optimization Kmeans clustering to reduce the generated scene. The invention provides a time-space correlation cascade water-wind-light multi-power system modeling method, and aims to analyze water-wind-light multi-energy complementary operation characteristics and provide reference for a scheduling mode among water-wind-light multi-energy.

The invention belongs to the technical field of ecological restoration, and provides a mountainous river-oriented fish diversified habitat natural shaping restoration method, which comprises the following steps of: determining fishes with a relative important index greater than 5% in a river as target fishes; the habitat factors of the river are screened, the habitat factors with colinearity are divided into a factor group, and only one habitat factor is reserved in each factor group to serve as a representative habitat factor; carrying out modeling on the resource distribution and representative habitat factors of the target fishes by adopting a generalized additive model; screening to obtain a generalized additive model with a minimum akaike information criterion, and determining an optimal habitat factor combination; and determining a restoration scheme according to the habitat factor combination, and adjusting the habitat of the target fish. According to the restoration method, the natural process is simulated to remodel the diversity of the mountain river habitat, the key problems of hydrological connectivity fracture, habitat simplification, engineering and ecological contradiction and the like are solved, and fish population restoration and ecological system self-maintenance are achieved.

The invention discloses a water engineering scheduling runoff intelligent simulation method, which comprises the following steps: edge distribution function fitting: calculating a mean value, a discrete coefficient and a skewness coefficient by applying a moment method, and constructing a daily runoff frequency curve by adopting a P-III type curve; constructing a wisdom knowledge set: establishing the wisdom knowledge set of parameters in a mixed Copula set by using a time sequence model, and calibrating the parameters of the wisdom knowledge set by using a cooperative search algorithm and taking the minimum akaike information criterion of theoretical distribution and empirical distribution as a target; and daily runoff simulation: establishing a mixed condition Copula set based on a Bayesian rule, and generating long-sequence daily runoff by adopting Latin hypercube sampling. According to the method, the daily runoff simulation difficulty of the mixed Copula is reduced, the number of parameters is reduced, the fitting precision of the mixed Copula is improved, the construction speed of the mixed Copula set is increased, and a valuable technical tool is provided for drainage basin management.

The present invention relates to the technical field of GNSS anti-spoofing interference, and specifically discloses a GNSS anti-induced deception interference method and system, the method comprising: acquiring a GNSS signal and calculating a code phase difference and a carrier phase difference; traversing the modal component number K value, performing variational modal decomposition on the code phase difference according to each K value to obtain modal components, calculating the corresponding criterion value of the Akaike information criterion, and selecting K that minimizes the criterion value as the optimal number of modal components; performing variational modal decomposition on the code phase difference and the carrier phase difference based on the optimal number of modal components; screening out modal components containing induced deception information to obtain screening results; and performing positioning and solving on the GNSS signal based on the screening results. The present invention is significantly superior to existing technologies in terms of dynamic adaptability, feature separation accuracy, and multi-information fusion, and provides a more efficient and robust solution for GNSS anti-induced deception interference.

The application discloses a site micro-vibration vibration source identification method based on machine learning, signals measured by acceleration sensors at the edge of a measured site and acceleration sensors at the center of the measured site are processed in sequence, including denoising, transient impact signal extraction, feature extraction, clustering processing of dynamic time warping distance of the transient impact signal based on a feature matrix, GMM modeling, obtaining a model, verifying the number of independent Gaussian components of the model, selecting the number of components with the minimum value as the optimal model according to the Akaike information criterion value, and performing maximum a posteriori probability identification on the transient impact signals captured by the sensors; GMM parameter comparison is performed on sets that may belong to the same category, two sets with highly overlapped models are merged into the same set, and the set is used as vibration data sets generated around the site and having an impact on the center position of the site.

The invention provides a method for measuring the diameter at breast height of cunninghamia lanceolata by using laser radar data of an unmanned aerial vehicle, belongs to the field of data measurement, and develops a single tree diameter at breast height (DBH) model suitable for large-scale application for a cunninghamia lanceolata forest by using ground measurement data of trees and corresponding aviation laser radar data. The contribution of stand density and tree features to improvement of the DBH model is evaluated. An optimal combination of virtual variables and stochastic effects is selected using the Akeke Information Criterion (AIC) and error metrics. The prediction performance of the model is evaluated by using an independent data set. The forest stand competition degree and the crown breadth index are added into the DBH model, so that the model fitting statistical magnitude can be improved. The model containing virtual variables to distinguish growth stages is superior to a basic model, and the model introducing a random effect to consider the level difference between the region and the sample plot further improves the prediction precision. When the method is applied to airborne laser radar data, the method can be suitable for researching individual tree characteristics of other species on a large spatial scale.

The invention discloses a distribution network line fault trip prediction method and device, a medium and a product, and relates to the field of power distribution network operation and maintenance, and the method comprises the steps: obtaining fault data of a distribution network line; constructing a fault data set; according to the fault data set, constructing a trip prediction model based on an akaike information criterion; an exogenous variable corresponding to the tripping prediction model is determined; constructing a limit gradient lifting model based on a grid search method according to the exogenous variables and the historical fault trip times; constructing a fusion model according to the trip prediction model and the limit gradient lifting model; respectively determining prediction errors of the tripping prediction model and the fusion model, and determining a final prediction model according to the prediction errors; determining the number of fault tripping times of the month to be predicted by using the final prediction model; and regularly obtaining fault data of the distribution network line to regularly update the final prediction model. The monthly fault trip times of the distribution network line can be efficiently predicted, and the operation and maintenance reliability of the distribution network is improved.

The invention discloses a data-driven operation optimization method for a circulating water pump of a coal-fired unit, which comprises the following steps of: performing variable primary selection, data preprocessing and feature selection through a historical database of unit operation, and determining key operation parameters which influence the power consumption of the circulating water pump; working condition division is carried out by adopting an improved GMM algorithm based on an akaike information criterion, and a plurality of combined working conditions are obtained through cross combination; key operation parameters influencing the power consumption of the circulating water pump are weighted through a grey relational analysis method and a CRITIC method, and then weight combination is carried out; a Bayesian classifier is adopted to establish an operation optimization model of the circulating water pump, historical optimal operation data is learned, and based on combination of a historical optimal working condition library and real-time data, the optimal operation mode of the circulating water pump can be accurately determined under complex working conditions. The operation optimization method can be suitable for complex working conditions, dynamically adjusts operation suggestions, is high in matching degree, is more reasonable and objective, is good in accuracy, and reduces energy waste.

The invention discloses an unbalanced power distribution network risk analysis method based on a multistage sparse polynomial, and belongs to the technical field of power distribution network risk analysis. The method comprises the steps of establishing a power distribution network input-output model; approximating the input and output model of the power distribution network as a linear regression model; calculating an ANCOVA index and a P value of the linear regression model, and sparsifying the input and output model of the power distribution network to obtain a preliminary sparsifying input and output model of the power distribution network; restraining a high-order cross term of the preliminary rarefaction power distribution network input and output model according to a preset truncation rule, shrinking a coefficient of the preliminary rarefaction power distribution network input and output model according to the optimized polynomial chaos expansion coefficient, and taking a preset residual error akaike information criterion as a model balance criterion to obtain a final rarefaction power distribution network input and output model; and obtaining a power distribution network risk analysis result according to the final sparse power distribution network input and output model. The method can provide a quantitative basis for operation and control of the power distribution network.

The invention discloses a partial discharge positioning detection method and system based on acoustoelectric joint detection and multi-dimensional feature fusion, and belongs to the technical field of intelligent power grid state detection. According to the method, an ultra-high frequency (UHF) signal is used as a discharge absolute time reference, an akaike information criterion (AIC) is adopted to adaptively pick up an ultrasonic weak wave head, and the problem of failure of a traditional threshold value method under a low signal-to-noise ratio is solved; meanwhile, a four-dimensional state vector containing space coordinates and equivalent sound velocity is constructed, a weighted Gaussian-Newton iterative algorithm is used for solving, a temperature field sound velocity correction model and a grid search initial value estimation algorithm are introduced, and the problems of iterative divergence and local optimal solution are solved. In addition, the method integrates algorithm modules of intelligent noise reduction, physical field adaptive correction, sound intensity assisted positioning and the like, and realizes high-precision and high-robustness positioning of the partial discharge source.

The invention discloses an ionosphere modeling method based on dynamic adaptive function optimization, and belongs to the technical field of earth space environment and satellite navigation, and the method comprises the steps: adaptively obtaining NeQuick model ionosphere background data of a target region in a preset time period through physical driving and a dynamic boundary; according to the NeQuick model ionized layer background data, determining an optimal polynomial order combination from a preset candidate order set by using an akaike information criterion; and on the basis of observation data of a global navigation satellite system, constructing a regional ionosphere refinement model by adopting the optimal polynomial order combination. According to the method, the simplest model is selected on the premise of ensuring the precision, and the calculation efficiency is improved.

Method and system of continuous monitoring of a structure (10), which comprises a fluid, to localize a leak or a source (E) of corrosion which emits acoustic emissions, said acoustic emissions being detected by means of three monitoring devices (M1-M2-M3), each having at least one detection group (G1-G2-G3) which is associated with the outer surface of the structure (10), each detection group comprising three sensor units (S0-Sa-Sb) arranged aligned with each other along a line parallel to an axis (A-A) of the structure (10). The method involves continuously receiving from each detector group and in parallel from each sensor unit (S0-Sa-Sb) respective series of analog signals (si(t)), and processing the series of analog signals (si(t)) by generating series of timed digital signals (si(nT)) which are processed on the basis of an Akaike Information Criterion to identify significant digital samples (si(K'T)). The method involves pre-localizing said source (E) for each monitoring device (M1-M2-M3) by determining a radial distance of said source (E) from said central sensor unit (So), based on time differences determined through a temporal analysis of corresponding significant digital samples (si(K' T)), and localizing said source (E) by processing the radial distances of said three monitoring devices (M1-M2-M3) using a processing software.

This invention relates to a method for highlighting spike discharge data of intracranial electroencephalogram (EEG) signals. The method involves collecting background noise data from the patient's brain without neuronal discharges, preprocessing the background noise data, automatically selecting the optimal order of an autoregressive (AR) model using the Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC), estimating the AR model coefficients using the Yul-Walker equation, and constructing a background noise model. The intracranial EEG signals to be processed are then subjected to high-pass filtering. A short-time Fourier transform (STFT) and window-based frame-by-frame processing strategy are used to subtract the spectrum of the signal from the noise. By adjusting the parameters of the spectral subtraction, noise removal and preservation of neuronal signal features are achieved.

The invention relates to the technical field of global navigation satellite system data processing, and discloses a GARCH-based GNSS cycle slip detection threshold real-time determination method, which comprises the following steps: acquiring GNSS double-frequency observation data, constructing a Morse-Wiberner and geometric distance-free combination observation value, and carrying out differential processing to obtain a noise time sequence. Secondly, traversal calculation is carried out based on an akaike information criterion, and the optimal order and the sliding window length of the GARCH model are determined; then, a current epoch condition standard deviation is predicted using the model, and a multiple factor is adjusted in combination with an environmental state to construct a dynamic detection threshold. And finally, comparing an absolute value of the noise sequence with a threshold value to judge a cycle slip, and calculating an integer solution by constructing an observation equation to complete repair. According to the method, the heterovariance characteristic of the observation noise is modeled by using the GARCH model, so that the adaptive adjustment of the detection threshold along with the environmental noise level is realized, the misjudgment rate in a complex environment is effectively reduced, and the detection accuracy is improved.

This invention discloses a multi-information source mechanism calculation method based on earthquake early warning station network observations. The method includes selecting velocity and acceleration recorded waveforms to form an input database; obtaining the P-wave initial motion polarity through manual picking and template matching techniques; calculating the source mechanism using a grid scanning method, and using solutions with an inconsistency ratio below a threshold as preliminary solutions; manually screening P-wave and S-wave phase waveforms; and jointly using initial motion polarity, S / P amplitude ratio, and body wave waveform as constraints to calculate the probability value of each preliminary solution. The probability distribution of clustered solutions is obtained through the Akaike information criterion and cluster analysis to obtain the optimal source mechanism calculation result. This invention, based on joint inversion calculations using acceleration records from early warning stations and velocity records from seismic stations, lowers the lower limit of the magnitude of earthquakes with computable source mechanisms. The comprehensive use of three types of information constraints improves the reliability of the results, and the calculation of the probability distribution of clustered solutions reduces the adverse effects of multiple solutions, thus improving the stability of the results.

The application discloses a water engineering scheduling runoff intelligent simulation method, comprising: edge distribution function fitting: using the matrix method to calculate the mean, dispersion coefficient and skewness coefficient, and adopting a P-III type curve to construct a daily runoff frequency curve; wisdom knowledge set construction: using a time series model to establish wisdom knowledge sets of each parameter in a mixed Copula set, using a cooperative search algorithm, taking the minimum of the Akaike information criterion of the theoretical distribution and the empirical distribution as the target, and calibrating the wisdom knowledge set parameters; daily runoff simulation: based on the Bayesian rule, a mixed conditional Copula set is established, and Latin hypercube sampling is used to generate a long sequence of daily runoff. The method reduces the difficulty of daily runoff simulation by the mixed Copula, reduces the number of parameters, improves the fitting precision of the mixed Copula, speeds up the construction speed of the mixed Copula set, and provides a valuable technical tool for basin management.

The present application belongs to the field of blast furnace system health management and prediction, and specifically discloses a blast furnace degradation model calibration and residual life prediction method under a noise environment. The method is based on a fractional Brownian motion with historical dependence, constructs an initial degradation model of the blast furnace system, estimates the model parameters by the maximum likelihood estimation method, and determines the optimal model structure of the blast furnace system temperature change by the Akaike information criterion. In addition, the present application also designs a model calibration trigger mechanism; when the mechanism is not triggered, the Bayesian fusion particle filter is used to estimate the potential degradation state of the system and update the model parameters; when the mechanism is triggered, a prediction error model is constructed and the degradation model structure is calibrated to adapt to complex working conditions. Finally, the probability distribution function of the residual life of the blast furnace system is derived, and the residual life prediction of the blast furnace system is carried out according to the calibrated degradation model.

The invention discloses an eVTOL collaborative design optimization method based on improved genetic programming, and the method comprises the steps: firstly constructing pneumatic, propulsion and dynamics high-fidelity models of an eVTOL takeoff stage, building an original collaborative optimization problem, and solving the problem through a direct transcription method; secondly, multi-target genetic programming fusing a root mean square error, an akaike information criterion, a predicted residual sum of squares and a Bayesian information criterion is creatively adopted, and an agent model for calculating thrust is constructed to replace a high-fidelity thrust model to solve and obtain optimal design parameters. The error between the optimal solution obtained by the method and the optimal solution obtained by using the high-fidelity model is very small, the calculation cost is remarkably reduced, and multidisciplinary efficient collaborative optimization is realized.

The invention relates to the technical field of paving thickness measurement, in particular to an ultrasonic paving thickness detection method and system, and the method comprises the steps: employing an Akaike information criterion for an echo measurement sequence obtained through collection of a paving road, and obtaining an AIC value of all elements of the echo measurement sequence based on the numerical distribution of AIC values of all elements of the echo measurement sequence; an echo positioning sequence and an echo AIC statistical sequence are obtained; element distribution and change characteristics in the echo positioning sequence are analyzed, and the change trend of each element in the echo AIC statistical sequence is obtained to obtain a signal mutation trend degree; obtaining an echo development trend degree based on a straight line fitting result of each element in the echo AIC statistical sequence; and confirming a primary echo moment, and calculating the paving thickness by combining the propagation speed of the ultrasonic waves in the paving mixture. The invention aims to improve the accuracy of paving thickness measurement.

The application discloses a fault probability monitoring and regulation system and method, relates to the technical field of fault monitoring, and collects to-be-determined related parameters in each period through a collection module, determines the optimal order of an autoregressive model of each related parameter by using the Akaike information criterion, and verifies the to-be-determined related parameters in each period, decides whether to perform secondary collection and replacement based on a verification result, generates a related parameter vector of each period, and starts a window translation along a period in a reverse order from the related parameter vector of each period through a monitoring module, calculates and arranges a comprehensive feature vector in each window, extracts and generates a core feature vector set through a time series self-encoder, trains a hidden Markov model by using a forward-backward algorithm, and obtains the fault probability of each period based on Bayesian inference, compares the fault probability with a probability threshold value to determine whether a fault exists, and executes regulation when the fault exists, so that high-precision fault monitoring and regulation based on collection self-checking, multi-dimensional data fusion and probability inference are realized.

The invention discloses a low-signal-to-noise-ratio impact signal first arrival moment extraction method, which comprises the following steps of: firstly, calculating an AIC (Akaike Information Criterion) curve of an original low-signal-to-noise-ratio impact signal according to an AIC (Akaike Information Criterion); secondly, smoothing preprocessing is conducted on the AIC curve, and then a first-order difference curve is calculated to highlight abrupt change characteristics at the first arrival moment of the signals; and finally, detecting the starting point of the first significant mutation signal through statistical analysis of the mutation characteristics of the first-order difference curve, and determining the moment corresponding to the starting point as the first arrival moment of the impact signal. According to the method, first arrival time extraction is directly carried out on the original low-signal-to-noise-ratio impact signal based on the statistical characteristics of the AIC curve of the original signal, denoising preprocessing does not need to be carried out, and the method has the technical advantages of being high in extraction precision, high in anti-interference capacity, low in calculation complexity and wide in application range; and the method shows good stability and robustness under different signal-to-noise ratio conditions.

Disclosed are methods and techniques of detecting network anomalies and responding to the anomalies once detected. The methods, for example, include receiving, by a model executed by a processor, real-time log data of an operating network; parsing, by the model executed by the processor, the log data to identify one or more metrics; determining, by the model executed by the processor, a seasonality of the one or more metrics; determining whether the model should use an autoregressive model if seasonality is detected; and on determining that an autoregressive model should be used, training a model based on determining a grid search for a parameter based on an Akaike information criterion.

The present invention relates to a method for estimating the kinetic parameters of hepatocellular carcinoma PET / CT imaging based on an improved MCMC algorithm and its application, belonging to the technical field of liver cancer auxiliary diagnosis. The present invention first obtains prior information and sample information. Secondly, based on Bayesian theory, the posterior distribution of the kinetic parameters of the hepatocellular carcinoma PET / CT imaging tracer is derived, and the Markov Chain Monte Carlo (MCMC) method is used to sample the parameter posterior distribution. The sampling algorithm adopted is an improved MCMC algorithm. The global optimization and local optimization of the flower pollination algorithm and the adaptive substitution probability are introduced into the standard MH algorithm to improve the convergence efficiency and accuracy of the MH algorithm, and estimate the kinetic parameters. Finally, the differences in the tracer kinetic parameters between the hepatocellular carcinoma tissue images and the surrounding healthy liver tissue images are analyzed and compared, and the Akaike Information Criterion (AIC) and Root Mean Square Error (RMSE) are used to compare the performance of the algorithm before and after improvement.

The present invention provides a method and device for detecting structural damage of a nonlinear system, which relates to the field of nondestructive testing technology. The method includes: applying an excitation force to a test piece to be inspected, vibrating the test piece to be inspected based on the excitation force, collecting signals from the test piece to be inspected to obtain an excitation signal and a response signal, combining the excitation signal and the response signal with multiple imported orders to construct initial NARX models of different orders, screening a target NARX model from the multiple initial NARX models based on the Akaike information criterion and optimizing it, comparing the order information in the optimized NARX model to be inspected with the order information in the reference NARX model, and determining whether the test piece to be inspected has defects based on the order comparison results. A nonlinear model is constructed and optimized based on the signal data of the test piece to be inspected, so that the model accurately represents the structural characteristics of the test piece to be inspected, thereby determining whether the test piece to be inspected has defects.