18results about How to "Fast inversion" patented technology

The invention provides a power transmission project karst geology detection method and system based on a ground penetrating radar, and relates to the technical field of geophysical ground penetrating radar detection, and the method comprises the steps: building a target body model of karst geology based on an autocorrelation random medium, carrying out the forward modeling to generate ground penetrating radar data, carrying out the inversion to generate a dielectric constant diagram, and constructing a karst disaster data set; constructing an inversion network taking the ground penetrating radar data as input and the dielectric constant diagram as output, and performing wavelet-independent ground penetrating radar source-independent deep learning on the inversion network by using the karst disaster data set; through the trained inversion network, inversion is carried out on the obtained actually-measured ground penetrating radar data, and a dielectric constant diagram of the karst geology is obtained through prediction; through preparation of the ground penetrating radar karst disaster data set based on the self-correlation random medium and ground penetrating radar source-independent deep learning inversion independent of wavelets, rapid inversion and attribute imaging of karst underground disasters of a power transmission line are realized, and the accuracy of karst geological detection of a power transmission project is improved.

The invention discloses a stratum parameter inversion method based on a physical driving neural network, and mainly solves the problems of poor inversion precision and low anti-noise capability of stratum parameters of a real well in an existing neural network inversion method. The method comprises the following specific steps: (1) constructing a stratum model and a logging response data set; (2) data preprocessing; (3) constructing a forward modeling neural network; (4) training and testing the forward modeling neural network; (5) constructing an inversion neural network based on physical driving; (6) training an inversion neural network; and (7) testing the trained inversion network. According to the invention, a deep neural network composed of a forward neural network module and a physical driving inversion module is constructed, and the problems of unsmooth boundary, low precision and weak anti-noise capability of the existing logging-while-drilling inversion are solved, so that stratum model parameters can be inversed in real time with relatively high precision.

The invention relates to an underground distributed optical fiber (DAS) anisotropic shale reservoir guided wave frequency dispersion curve inversion method and a medium, and the method comprises the steps: obtaining guided wave seismic data of a shale reservoir, calculating a guided wave frequency dispersion energy diagram according to the obtained guided wave seismic data, and extracting an observation guided wave frequency dispersion curve from the guided wave frequency dispersion energy diagram; establishing an initial anisotropic elastic parameter model of the shale reservoir; according to the initial or updated anisotropic elastic parameter model of the shale reservoir, performing forward modeling to obtain a theoretical guided wave frequency dispersion curve; constructing an objective function, and calculating and judging whether the objective function is converged or not according to a residual error between the theoretical guided wave frequency dispersion curve and the observation guided wave frequency dispersion curve; if the target function converges, outputting an anisotropic elastic parameter model of the current shale reservoir; if the target function is not converged, updating the anisotropic elastic parameter model by adopting a gradient algorithm until the target function is converged; and outputting the optimized anisotropic elastic parameter model of the shale reservoir.

The application discloses a kind of acoustic-electric tomography image imaging algorithm based on two-point gradient iterative method, and the innovation is to improve the calculation efficiency by two-point gradient iterative method, and the accuracy of image reconstruction is improved by the selection of regularization term.The core of the application is to use the characteristics of acoustic-electric hybrid imaging, that is, when the boundary detection current is applied, the power density data is obtained by using ultrasonic disturbance, so as to obtain the internal data of biological tissue.At the same time, according to the different characteristics of different biological tissue conductivity, different regularization terms are introduced, two-point gradient iterative method is established, and high-resolution reconstruction of biological tissue conductivity is realized.Compared with the traditional acoustic-electric imaging method, the application reduces the calculation and storage cost of multiple observation currents, improves the imaging efficiency.And the iteration process is modified by using a special regularization term, so that the boundaries or geometric characteristics of different tissues can be accurately determined, and the imaging accuracy is improved.

The present application relates to the technical field of complex radiation field inversion, and particularly relates to a complex radiation field inversion system and method based on a multi-layer source grid, wherein a multi-layer source grid division subsystem is used to perform multi-layer discrete approximation on a complex source term with a known position and shape to be inverted, a detector partition arrangement subsystem is used to set detectors for a radioactive source, all coarse grid volume sources and all fine grid volume sources respectively in different manners for detection, a source term inversion subsystem is used to perform inversion on source intensity distribution of the complex source term according to the radioactive source, all coarse grid volume sources, all fine grid volume sources and first, second and third measurement values, and a radiation field calculation subsystem is used to calculate the radiation field distribution of the inverted complex source term. The technical problem that the complex source cannot be inverted in the traditional inversion is overcome, and the radiation field under the complex source distribution can be quickly and accurately inverted.

The application discloses a soft soil foundation settlement prediction method based on a physical information neural network, and relates to the technical field of geotechnical engineering monitoring. Multi-source data containing settlement, displacement and pore water pressure are collected, wavelet denoising and feature screening are performed, and a high-quality data set is constructed; a forward mapping model based on finite element FEM is constructed, an improved immune genetic algorithm IGA is used to minimize errors, and accurate inversion of key mechanical parameters is realized; the mechanical parameters are embedded into a long short-term memory network LSTM, a hybrid loss function containing data fitting loss and physical limit constraint loss is constructed, and a physical information neural network PINN-LSTM is formed; the future deformation trend is predicted through the network, the safety factor is calculated in combination with the limit equilibrium theory, and hierarchical early warning and disposal suggestion pushing are realized according to a dynamic early warning threshold. The method significantly improves the physical interpretability and precision of soft soil foundation settlement prediction, and has wide application value.

The invention relates to the technical field of surface temperature extraction, and provides an automatic scorching ethylene cracking furnace tube surface temperature extraction method, which comprises the following steps of: firstly, accurately positioning a furnace tube in a swinging state in a complex infrared thermal radiation environment by combining a deep neural network with an attention mechanism; the method comprises the following steps of: establishing a homogeneous background, separating the homogeneous background from the homogeneous background, establishing a lookup table (LUT) based on an infrared radiation transmission theory, fully utilizing priori knowledge of three physical parameters, namely furnace tube surface temperature, equivalent background temperature and emissivity, and accurately deducting the influence of reflection radiation, range radiation, emissivity and the like of the background on infrared radiation receiving of a detector; therefore, the surface temperature distribution of the furnace tube is accurately and quantitatively calculated.

The present invention discloses an artificial intelligence monitoring and early warning method for atmospheric pollution that integrates multi-source meteorological data, which belongs to the field of atmospheric pollution monitoring technology, including preliminary selection of model factors, acquisition of model factors, screening of model factors, model establishment and training, model effect testing, and monitoring and early warning. The present invention screens model factors that are sensitive to atmospheric pollution, combines the forecast results of air quality numerical models, integrates minute-level three-dimensional multi-source observation data, and establishes an atmospheric pollution monitoring and early warning model to achieve automatic reminder services, providing an objective basis for atmospheric pollution prevention and control decisions; compared with numerical models, the calculation efficiency is high, the inversion speed is fast, and the accuracy rate is significantly improved; the minute-level observation data is rolled and assimilated into the model in real time, which can greatly improve the refinement and objectivity of the monitoring results.

The invention discloses a dual-polarization array type multi-frequency microwave band electromagnetic detection device and method, the device comprises a high-directionality dual-polarization transmitting antenna and more than two high-directionality dual-polarization receiving antennas, the high-directionality dual-polarization transmitting antenna is electrically connected with a microwave broadband transmitting system, and the microwave broadband transmitting system is electrically connected with the high-directionality dual-polarization receiving antennas. And each high-directionality dual-polarization receiving antenna is electrically connected with one broadband superheterodyne receiving circuit system. The detection equipment disclosed by the invention adopts high-directionality, dual-polarization (horizontal and vertical) and array antennas, realizes acquisition of scattering multi-component microwave electromagnetic signals, effectively improves the detection capability of weak scattering signals, breaks through the limitation that the detection depth of a traditional geological prospecting instrument on an underground target is relatively small, and improves the detection precision of the underground target. And high-resolution detection of an underground deeper complex space target body can be realized.

The invention discloses an active inversion method of a typical deep sea sound velocity profile based on a reliable sound path, and relates to acoustic inversion of marine environment parameters. The method comprises the steps that based on ocean mode data or measured data, a sound propagation model is combined, and a sea surface scattering echo arrival pitch angle and arrival time two-dimensional curve under different deep sea sound velocity profiles is obtained through simulation; a vertical receiving array carrying an active sound source is arranged at the bottom of a deep sea, and sea surface scattering echoes obtained by active emission are subjected to beam forming to obtain actually measured sea surface scattering echo arrival pitch angle and arrival time two-dimensional curves; comparing the two curves to obtain sound velocity profile expression parameters under the optimal matching of the two-dimensional curve, and obtaining deep sea sound velocity profile estimation by using the expression parameters. And a full-water-depth sound velocity profile can be obtained without covering a measurement system of a whole water body. A traditional receiving and transmitting split double-subsurface buoy system does not need to be used for obtaining a deep sea water sound velocity profile between receiving and transmitting, and deep sea sound velocity profile estimation can be achieved by placing a single receiving and transmitting assembly on a system on the seabed.

The invention discloses a nuclear accident inversion method based on an improved extreme learning machine, and relates to the technical field of nuclear accident source item inversion. According to the invention, missing and abnormal data screening is carried out on the collected nuclear accident comprehensive data, the time series data in the nuclear accident comprehensive data are input into the long short-term memory network for preliminary prediction and window feature extraction processing, and the efficiency and precision of an inversion model are improved by improving the extreme learning machine. A sparrow optimization algorithm is combined with an extreme learning machine, global optimization is performed on a weight matrix and an offset vector of the extreme learning machine, meanwhile, variation crossover of a sine and cosine algorithm and a genetic algorithm is introduced into the sparrow optimization algorithm, input layer information is introduced into an output layer of the extreme learning machine, and meanwhile, a C-loss function is introduced to improve the structure. The robustness of the model is improved, the internal complexity of the extreme learning machine model is reduced, overfitting is prevented, the quality of the model is improved, and the prediction precision and speed of the model are improved.

The present application discloses an inversion method, device, electronic device and storage medium for dam material parameters. It includes: obtaining a pre-constructed dam finite element model; reading at least one current target material parameter based on the dam finite element model, and determining the target material parameter as the inversion target; obtaining a candidate value corresponding to the inversion target, and using a pre-constructed target inversion function to perform an inversion calculation on the candidate value to obtain a target function value; using the target function value to analyze the sensitivity of the inversion target, and using the sensitivity to determine the adjustment information corresponding to the parameter value of the inversion target. The fast inversion method for dam material parameters provided in the embodiment of the present application uses the sensitivity influence of the material parameters to be inverted on the inversion target as an indicator, provides a fast convergence direction for the artificial fish swarm algorithm, reduces the number of iterations, and quickly inverts the dam material parameters.

The application discloses a kind of multi-source data fusion soil attribute collaborative inversion and management method and system, belong to wisdom agriculture and soil informatization technical field.The method is by obtaining gamma spectrometry data, optical remote sensing data and ground sensing and measured data to multi-source heterogeneous data carry out space-time registration, feature extraction and fusion modeling;Multi-task learning model is constructed to realize the collaborative inversion of soil nutrients, physical and chemical properties and biological characteristics;Introduce soil health evaluation model to carry out grading evaluation and index representation to soil health state, and generate management and control suggestion.The system includes data access and management layer, data preprocessing and feature engineering layer, collaborative inversion and model library layer and decision support and visualization layer, with the advantages of high inversion accuracy, strong adaptability and outstanding decision support capability.The application improves soil attribute inversion accuracy and regional coverage efficiency, and provides technical support for precision fertilization, soil improvement and sustainable agricultural management.

The present application relates to the technical field of oil and gas drilling engineering, and specifically discloses a multi-field coupling complex fractured formation fracture width distribution inversion method, based on the theory of porous elasticity, a drilling fluid loss model considering fluid-solid coupling effect is constructed, a roughness correction factor is introduced to correct the traditional cubic law, and the application of Darcy's law in the matrix is improved, a drilling fluid loss mathematical model coupled with the matrix and the fracture is established, through numerical simulation, the influence law of key parameters such as fluid viscosity, pressure difference and fracture width on drilling fluid loss is obtained, and on this basis, a fracture width multivariate linear regression inversion equation with fluid viscosity, pressure difference and cumulative loss as independent variables is established, the inversion equation has high prediction accuracy and practicability, can be used for quickly and accurately inverting the fracture width on site, and provides theoretical and technical support for the understanding of the loss mechanism of multi-scale fractured formation and the optimization of leakage prevention and treatment.

The patent relates to a kind of time-domain optical wave electric field measurement techniques based on vector double-pulse supercontinuum, belongs to the field of optoelectronic technology.The technology includes vector double-pulse laser configuration, nonlinear optical medium, optical signal processing unit, photoelectric conversion processing unit and other components.Pulse laser to be measured is first converted into vector double-pulse laser in the system, and is injected into nonlinear optical medium, and then obtains the optical wave electric field information of the laser to be measured after optical signal processing and photoelectric conversion processing.This technology has positive contribution to the characterization of time-domain ultrashort pulse laser.

This application discloses a device for generating Mach-Zehnder interferometry, a photothermal spectral gas sensing device based thereon, and a detection method thereof, belonging to the field of laser spectral gas sensing technology. The interferometer device sequentially includes an optical frequency comb and a first optical beam splitter in the optical transmission path. A movable-arm mirror group is provided in the optical path of one beam B, which is split into two beams by the first beam splitter. Beam B changes its path after passing through the movable-arm mirror group and intersects with the optical path of the other beam C. A beam combiner is provided at the intersection. The movable-arm mirror group is connected to a driving mechanism, which drives the movable-arm mirror group to perform linear motion, the trajectory of which is parallel to the line containing beam B. The interferometer device of this application does not require multi-heterodyne interference using a local oscillator optical comb and a signal optical comb. The downconversion factor can be directly calculated using the moving speed of the movable-arm mirror group in the Mach-Zehnder interferometer structure, mapping each longitudinal mode of the signal optical comb at the optical frequency to the radio frequency component.

The invention discloses an absorption spectrum inversion method based on a neural network and a correlation method, and mainly solves the problems of low accuracy, poor real-time performance and strict requirements on absorption spectrum data standards of the existing absorption spectrum temperature inversion. According to the implementation scheme, the method comprises the following steps: acquiring original absorption spectrum data, and preprocessing the original absorption spectrum data; constructing a neural network model comprising a full convolutional network FCN and a long short-term memory network LSTM; training a neural network model by using the preprocessed absorption spectrum data set through a back propagation method to obtain a trained neural network model; inputting to-be-measured absorption spectrum data into the neural network model, and outputting a temperature range corresponding to an absorption spectrum; and post-processing the temperature range output by the neural network based on a correlation method to obtain a temperature value. The method effectively improves the precision and real-time performance of the absorption spectrum temperature inversion, reduces the requirements on the absorption spectrum data standard, and can be used for gas component analysis, environmental monitoring, drug analysis and food safety detection in the fields of chemistry, physics and biology.

The invention relates to the technical field of soil nutrient inversion, and particularly discloses a soil nutrient inversion method and system based on a gamma-ray energy spectrum full spectrum.The method comprises the steps that firstly, a gamma-ray energy spectrometer is used for measuring a target area, and full spectrum data containing characteristic peaks of natural radionuclides such as 40K, 238U and 232Th are obtained; performing energy calibration and normalization processing on the original energy spectrum, and converting a channel number into an actual energy value to form a standardized energy spectrum; the method comprises the following steps: extracting main characteristics of a spectrum shape by introducing a characteristic selection and dimension reduction algorithm, performing characteristic correction in combination with soil moisture content parameters, constructing a mapping model of energy spectrum characteristics and soil nutrient content, and realizing rapid inversion of main nutrients such as nitrogen, phosphorus, potassium, organic matters, pH and the like; compared with a traditional chemical analysis or single-point spectrum inversion method, the method provided by the invention can make full use of energy spectrum full information, improves the inversion efficiency and precision, realizes rapid, lossless and spatial estimation of soil nutrients, and provides technical support for precise fertilization and cultivated land quality monitoring.