70results about How to "Improve characterization accuracy" patented technology

The invention provides a data processing method and a data processing device based on a user scenario. The method can comprise the following steps: acquiring track data of a user in a preset time period, wherein the track data comprises more than one triple (user, time and position), and each element in the triple has one or more than one attributive character; clustering the track data according to the one or more than one attributive character of each element in the triple, and determining the information of a scenario at which the user is positioned at present according to the attributive characters with significance in the clustering result; and transmitting the message associated with the information of the scenario at which the user is positioned at present to the user. According to the method and the device, compared with scenario description based on geographical location only, the description precision of the user scenario is improved, and different scenarios positioned at the same geographical location can be distinguished.

The invention discloses an ultrasonic RF (radio frequency) time sequence-based tissue characterization method, which comprises the following steps of 1, feature extraction: performing operations as shown as follows on each tissue sample to obtain corresponding feature vectors; 2, tissue recognition, wherein the step1 comprises the following steps: 1-1, scanning tissues to acquire multiple frames of ultrasonic echo RF signals of the tissues; 1-2, demodulating a certain ultrasonic echo RF signal and displaying a B-mode graph; 1-3, selecting an ROI (region of interest) of which the size is M*N from the B-mode graph; 1-4, taking previous L RF signals from each point in the ROI to construct M*N ultrasonic RF time sequences of which the length is L; 1-5, extracting fractal features, frequency domain features and time domain features of the ultrasonic RF time sequences to obtain the feature vectors of each tissue samples. According to the ultrasonic RF time sequence-based tissue characterization method disclosed by the invention, the features from three angles: fractal, frequency domain and time domain, are extracted based on the ultrasonic RF time sequences, and the tissues are recognized by using a classifier; the ultrasonic RF time sequence-based tissue characterization method can be widely applied to the field of ultrasonic medical detection and ultrasonic tissue characterization, and the like.

The invention discloses a multi-wave seismic oil and gas reservoir prediction method based on a deep neural network. The method comprises the steps of firstly, carrying out optimization on seismic attributes through employing a particle swarm optimization clustering analysis method and a kernel principal component analysis method for original data obtained from a longitudinal and transverse wave seismic attribute set, removing redundant information, and highlighting characteristics of the multi-wave seismic oil and gas reservoir to obtain better deep neural network sample data; then, learningthe obtained sample data through a deep neural network model, and carrying out simulation prediction to obtain an oil and gas reservoir evaluation map; and finally, performing image enhancement processing on the oil and gas reservoir evaluation image so as to improve the detail information and the edge identification degree of the image and improve the definition of the image. In oil and gas reservoir prediction, the method provided by the invention can improve the depicting precision of the seismic oil and gas reservoir, and provides a new way for identification and prediction of the oil andgas reservoir.

The invention belongs to the technical field of ultrasonic medicine and particularly relates to a hepatic fibrosis degree Fisher identification method based on an ultrasonic radio frequency (RF) time sequence. The method includes that first, a broadband ultrasonic linear array probe is used for scanning a living hepatic tissue, and multi-frame ultrasonic echo RF signals are recorded; a certain frame of B type pattern is demodulated and displayed, a region of interest (ROI) is selected, all frame RF signals of each point in the ROI are taken to form the ultrasonic RF time sequence; SMR fractal dimension mean values and a plurality of characteristic parameters of the RF time sequence in the ROI are extracted, the 7 features are extracted for a large amount of normal and fibrosis hepatic samples, Fisher distinguishing rate of each feature parameter is collected and obtained, the normalization Fisher distinguishing rate is used as the weight of a new input sample corresponding feature parameter, and the absolute value of each feature parameter of a new sample is weighted and summed to obtain Score-fisher. The hepatic fibrosis degree Fisher identification method based on the ultrasonic RF time sequence is brought up for the first time and provides quantification reference for clinical hepatic fibrosis diagnosis and dynamic monitoring.

The invention discloses a method for synchronously characterizing anisotropic yield and hardening constitutive parameters of a metal material, which is used for solving the technical problem of poor practicability of the conventional method for characterizing the parameters of the metal material. The technical scheme is that a proper test piece configuration is designed, so that the test piece can simultaneously generate rich positive strain and shear strain under the action of simple tensile or compressive load, and the out-of-plane deformation of the test piece is at a lower level within a certain compressive strain range, thereby achieving the simultaneous characterization of anisotropic yield and anisotropic plastic hardening constitutive parameters of numerous metal materials by utilizing a virtual field method only by combining primary cyclic tensile-compression experiment and primary unidirectional tensile experiment. Compared with the anisotropic parameter characterization method in the prior art, the method can be completed on one single-axis testing machine without a biaxial stretching machine, a hydraulic expansion testing machine, a shearing experimental device and the like, and experimental equipment and an experimental process are simplified, and the practicability is good.

The invention discloses a feature extraction method for ultrasound tissue characterization based on the Hilbert-Huang transform. The method comprises the steps of acquiring multiple frames of ultrasonic echo RF signals of a tissue sample, selecting regions of interest, and constructing an ultrasonic RF time sequence; extracting features of the ultrasound RF time sequence by using a HHT algorithm to obtain sample feature vectors based on the relation of time, frequency and energy; and performing feature screening and feature fusion by using a rankfeatures algorithm, calculating a sample feature fusion index, and taking the feature fusion index as the feature for tissue characterization. The feature fusion index provided by the invention can effectively reflect the difference between different types of tissue samples and is suitable for the feature extraction of the ultrasonic tissue characterization.

The invention belongs to the technical field of ultrasonic medicine and particularly relates to a hepatic fibrosis degree relative deviation identification method based on an ultrasonic radio frequency (RF) time sequence. The method includes that first a broadband ultrasonic linear array probe is used for scanning a living hepatic tissue, and multi-frame ultrasonic echo RF signals are recorded; a certain frame of B type pattern is demodulated and displayed, a region of interest (ROI) is selected, all frame RF signals of each point in the ROI are taken to form the ultrasonic RF time sequence; SMR fractal dimension mean values and a plurality of characteristic parameters of the RF time sequence in the ROI are extracted, the 7 features are extracted for a large amount of normal and fibrosis hepatic samples, relative deviation value of each feature parameter is collected and obtained, the normalization relative deviation value is used as the weight of a new input sample corresponding feature parameter, and the absolute value of each feature parameter of a new sample is weighted and summed to obtain Score-rd. The hepatic fibrosis degree quantification identification method based on the RF time sequence is brought up for the first time and provides quantification reference for clinical hepatic fibrosis diagnosis and dynamic monitoring.

The invention relates to a quantitative prediction method for a slope break belt by utilizing seismic data. The method comprises A according to the background data, such as geology and earthquake, performing fine three dimensional seismic horizon explanation for the series of strata of the developed slope break belt, and obtaining the horizon data of the target horizon; B solving the gradient and the exposure value of the target horizon; C according to the gradient and the distribution characteristic on an exposure planar graph, preliminarily identifying the distribution range of the slope break belt; if the boundary of the slope break belt is clear, directly solving the geometrical parameters of the slope break belt: rupture, slope toe, gradient, altitude difference and horizontal width; if the boundary of the slope break belt is fuzzy (such as influence of breakage), performing the step D; D counting and solving the mean value theta for all the exposure values in the preliminarily identified slope break belt range; E taking the theta as the center and Delta as the step length, scanning and solving the optimal apparent gradient Phi in the theta+-Epsilon range; and F according to the solved optimal apparent gradient in the step E, identifying the boundary of the slope break belt, and respectively solving the geometrical parameters of the slope break belt. The quantitative prediction method for a slope break belt by utilizing seismic data can realize plane depiction of the gradient of the slope break belt, and can improve the accuracy for gradient calculation.

The invention discloses a method for automatically recognizing the hob failure form of a numerical control gear hobbing machine. The method includes the following steps of firstly, collecting main shaft vibration signals through an embedded vibration acceleration sensor in real time; secondly, obtaining main shaft vibration signals of various hob failure forms corresponding to different hob technological parameters, establishing vibration path graph signals, extracting features to form standard feature vectors, and then establishing a standard feature vector subset and a standard feature vector set; thirdly, calculating the similarities of sample feature vectors and the standard feature vectors, obtaining the average similarity, and accordingly judging the hob failure form. By means of themethod, by establishing the vibration path graph signals, combining the processing of the main shaft vibration signals of the numerical control gear hobbing machine with the spectrum theory and conducting the feature extracting and failure form recognizing of the hobbing failure form, the on-line real-time monitoring of the hob failure form can be realized, the dependency on professional skills of operators is reduced, and labor cost is reduced.

The invention relates to a high-precision inversion method for a deep-water turbidite sandstone reservoir under a complex fracture system. The high-precision inversion method comprises the following steps that 1, a three-dimensional fault framework model is established according to fault interpreting data; 2, a three-dimensional horizon framework model is established according to seismic horizon interpreting data; 3, the development sequence of a fault and a horizon is judged according to tectonic evolution characteristics and geological prior information, the space truncation relation of thefault and the horizon is defined on the basis, and the three-dimensional fault framework model obtained in the step 1 and the three-dimensional horizon framework model obtained in the step 2 are combined to obtain a high-precision three-dimensional seismic formation framework; 4, a 'multi-body phase controlled iteration method' is adopted to establish an inversive low-frequency model; 5, iterativeoptimization is conducted on the prestack inversion process until an inversion result meets the requirement that space errors and loss functions are minimum, and a final inversion result is output.

The invention relates to a geological modeling method for internal seepage barriers in an offshore oilfield reservoir. The method comprises the following steps: collecting sensitive seismic attributesof a to-be-measured reservoir, predicting distribution of internal seepage barriers in the reservoir, and classifying the seepage barriers into three classes of impermeability, semi-permeability andpermeability; according to horizontal well data of seepage barrier drilling in the to-be-measured reservoir, carrying out statistics of basic data for representing three types of seepage barriers; establishing lithofacies models of impermeable, semi-permeable and permeable seepage barriers seepage barriers respectively; and according to the attribute characteristics of the seepage barriers, setting virtual wells in the lithofacies models and simulating attribute characteristics of the three types of seepage barriers to obtain a geological model of the seepage barriers in the reservoir. According to the invention, real distribution of the seepage barriers in the reservoir is effectively described and defects of the traditional utilization theory are overcome. Since the virtual wells are setby y copying the actually-drilled horizontal well, the basic data of geological modeling are enriched and the characterization precision of the seepage barriers in the geological model is improved.

The invention discloses a tie line power safety domain characterization method considering new energy uncertainty. The tie line power safety domain characterization method mainly comprises a tie linepower safety domain characterization method considering new energy uncertainty based on a probability density function and the tie line power safety domain characterization method considering new energy uncertainty based on an interval number. According to the method, the uncertainty of new energy can be fully considered, and the characterization precision of the joint line power safety domain isimproved.

The invention discloses a modeling method of an asymmetric hysteresis model, and relates to the field of parameter identification and compensation of a hysteresis nonlinear model. The problem that inthe prior art, the depicting degree of the actual asymmetric hysteresis characteristic is not ideal enough, and the approximation degree of the depicted actual asymmetric hysteresis characteristic islow is solved. The method comprises the following steps: firstly, identifying a symmetric hysteresis model through an input voltage vector X and an output displacement vector Y, and then identifying anonlinear correction curve h (x) according to the input voltage vector X, the output displacement vector Y, a rising curve f (x) of the symmetric hysteresis model and a falling curve g (x) of the symmetric hysteresis model; and finally, correcting the symmetric hysteresis model through the nonlinear correction curve h (x), thereby obtaining an asymmetric hysteresis model, and completing the establishment of the asymmetric hysteresis model. The method is mainly used for describing the asymmetric hysteresis characteristic of the piezoelectric ceramic.

The invention relates to an ultrahigh-precision multilayer film thickness drift error calibration method, which comprises the steps of of alternately moving a substrate to a sputtering region of two target materials for plating a multilayer film, and plating a first periodic multilayer film; enabling the substrate to be far away from the sputtering region, continuing sputtering the target materials, and simulating the plating process of the target periodic multilayer film; alternately moving the substrate plated with the first periodic multilayer film to the sputtering region again, and plating a second periodic multilayer film to form a calibration sample; and performing an X-ray grazing incidence reflection test on the plated calibration sample under fixed X-ray energy to obtain a test reflectivity curve, and obtaining a multilayer film drift error from the beginning to the end based on comparison between the test reflectivity curve and the simulated reflectivity curve, wherein the number of film pairs of the first periodic multilayer film and the second periodic multilayer film is 1/10-1/5 of the number of film pairs of the target periodic multilayer film. Compared with the prior art, the method has the advantages of high precision, small influence of instrument conditions and the like.

The invention discloses a method of representing extrusion pressure, in particular to a method of representing extrusion pressure via a three-point process based on Benbow-Bridgwater model; the method comprises the main steps of (1) acquiring extrusion pressure Pi via three-point testing; (2) calculating average extrusion pressure P0 when length-to-diameter ratio of a nozzle is equal to zero; (3) acquiring an analytical model for representing extrusion pressure via the three-point process. Extrusion pressure representation is one of the most key factors to successfully achieve the material extrusion technique or not; in traditional representation methods, a single-point method is used to estimate extrusion pressure when the length-to-diameter ratio of the nozzle is zero, and great errors are caused. In order to solve the problems, for example, traditional representation methods are low in precision, optimization of material extrusion systems and process parameters is not favored, and extrusion precision is affected, the simple, convenience and practical method for representing extrusion pressure is provided; extrusion pressure representing precision is improved, and the method may act as a common method of representing extrusion pressure.

The embodiment of the invention provides a non-seismic data volume processing method and an apparatus. The method includes: obtaining the layer speed of a stratum corresponding to each non-seismic data point in a non-seismic data volume; and regarding each non-seismic data point in the non-seismic data points as a to-be-processed data point, dividing the stratum depth of the to-be-processed data point by the layer speed of the stratum of the to-be-processed data point and obtaining a time value corresponding to the to-be-processed data point in order to convert the non-seismic data volume in a depth domain to a data volume in a time domain. According to the method and the apparatus, the three-dimensional electrical data volume in the depth domain is converted to the data volume in the time domain, three-dimensional seismic and three-dimensional electrical combined inversion can be realized by employing the three-dimensional electrical data volume in the time domain and the three-dimensional seismic data volume in the time domain, compared with the realization of the three-dimensional seismic and three-dimensional electrical combined inversion by employing the three-dimensional electrical data volume in the depth domain and the three-dimensional seismic data volume in the time domain, seamless connection of the three-dimensional electrical data volume and the three-dimensional seismic data volume is realized, and the conglomerate carving precision is improved.