11 results about "Empirical relationship" patented technology

The invention provides a crack depth prediction method based on an infrared thermal imager and deep learning, and relates to the technical field of crack prediction.The method comprises the steps that a cooling infrared thermogram sequence and a visible light image of a crack area under thermal excitation are synchronously collected, and multi-modal features are extracted after registration fusion; based on consistency verification of heat conduction inversion, a width-depth empirical relation and a traditional detection means, a crack depth label with high confidence is automatically generated, and a data set with the label is constructed; and finally, designing a deep learning model containing attention enhancement and multi-scale fusion, learning a mapping relation between an input image and depth, and realizing accurate prediction of the crack depth. According to the method, a multi-source information consistency verification mechanism is introduced, physical model inversion, experience statistics and measured data are mutually verified, the bottleneck problem that high-quality deep labels are difficult to obtain in engineering practice is effectively solved, and a reliable training data basis is provided for a deep learning model.

The invention provides a train airtightness iterative design method, system and equipment based on a simulation model and a medium. The train airtightness iterative design method comprises the following steps that an initial empirical relational expression and an initial simulation model are obtained according to a first train of airtight trains; according to the related theoretical relational expression and the initial simulation model, obtaining a whole vehicle air tightness index; according to the initial empirical relational expression and a related theoretical formula, the whole vehicle air tightness index is converted into a part air tightness index; according to the airtightness indexes of the parts, the whole newly-manufactured train is installed and manufactured, and the newly-manufactured train is used as the first airtight train of the next airtightness iteration; according to the method, the static sealing index of each key airtight component is obtained through decomposition based on a simulation analysis method, the problems and defects existing in existing train air tightness optimization design can be solved, and train air tightness meeting the requirements is obtained according to simulation and test data of an existing train combination line and used for guiding and optimizing subsequent product air tightness design. The system also has the same beneficial effects.

The invention discloses a thin cloud atmosphere top reflectivity estimation method and system based on deep learning. The method comprises the following steps: firstly, performing radiometric calibration on obtained remote sensing data to obtain a reflectivity image; secondly, 1 / 2 of a wave band is input, end-to-end feature extraction and cloud TOA reflectivity inversion are achieved through a convolutional neural network, and a preliminary inversion result is output; and finally, sampling and generating a denoised cloud TOA reflectivity image by using the trained denoising diffusion model according to the preliminary inversion result. According to the method, thin cloud reflectivity estimation is modeled as an image recovery problem, and preliminary inversion of thin cloud radiation characteristics is realized through an empirical relationship between wavebands. The advantages of the diffusion model in the aspects of image feature expression and reconstruction are combined, the radiation contribution of the cloud is accurately separated from complex earth surface and atmosphere mixed signals, effective estimation of the TOA reflectivity of thin cloud areas of different types and concentrations is achieved, and therefore a more robust physical basis is provided for downstream cloud detection and cloudless image recovery.

The invention discloses a method for establishing a two-dimensional CMC turbine blade surface Nusselt number engineering empirical relational expression, which comprises the following steps of: systematically researching the flow and heat transfer characteristics of different pressure gradient areas such as a suction surface front edge, a suction surface tail edge and a pressure surface middle section, extracting geometrical characteristic parameters of surface textures, and introducing a dimensionless correction factor, so as to establish a two-dimensional CMC turbine blade surface Nusselt number engineering empirical relational expression. A Nusselt number correction model suitable for different areas is constructed, and a heat exchange coefficient mapping relation from a smooth surface to a woven texture surface is realized. Compared with a traditional smooth surface or a general roughness empirical correlation formula, the model established by the method can more accurately reflect the influence rule of weaving texture features on the surface heat exchange distribution of different pneumatic areas, and the engineering applicability and calculation precision of heat exchange prediction are remarkably improved; and a reliable engineering analysis method and technical support are provided for heat transfer design, performance evaluation and structure optimization of CMC turbine blades and other high-temperature composite material parts.

The invention relates to the technical field of ecological environment management, and provides a regional scale future soil carbon spatial-temporal change prediction method based on a geographically weighted regression model, comprising the following steps: S1, collecting data; s2, constructing a soil carbon future spatial-temporal change prediction model and performing soil carbon density spatial-temporal change prediction in different periods in the future, including S21, calculating the annual change rate of the soil carbon density; s22, selecting a prediction model auxiliary variable; s23, constructing a soil carbon future temporal and spatial change prediction model; and S24, predicting spatial and temporal changes of the soil carbon density in different periods in the future. According to the method, the empirical relationship and the spatial difference between the annual change rate of the soil organic carbon density and the influence factors are captured by using the geographically weighted regression model, and the influence of the soil initial carbon density on the annual change rate of the soil carbon density is fused; and the prediction of the spatial and temporal distribution trend of the soil carbon density under different temperature change scenes in the future of the region is effectively realized.

The present invention provides a storage life calculation method and a flash memory. The calculation method includes: respectively obtaining distribution curves of data retention performance parameters of multiple identical devices under test placed in multiple test environments over operating time, wherein the operating temperature of each test environment is different, and the device under test has a preset data retention performance standard line; then, based on the multiple distribution curves and the preset data retention performance standard line, obtaining the storage life of each device under test at the corresponding operating temperature, and taking each operating temperature and the corresponding storage life as a target array; finally, fitting a linear function of the storage life varying with the operating temperature based on the multiple target arrays, and calculating an empirical relationship between the storage life of the device under test and the operating temperature. The empirical relationship can be used to accurately calculate the storage life of the device under test at any operating temperature.

The invention relates to the technical field of heat treatment processes, and discloses a heat treatment system selection method based on machine learning and particle swarm intelligent optimization, and relates to the technical field of heat treatment process.The method comprises the steps that a heat treatment process feasible region is determined according to material characteristics, small samples are prepared through orthogonal experiment design, the hardness is tested, and a data set is constructed; training a Gaussian process regression or support vector machine regression model, establishing a mapping relation between process parameters and hardness, adopting a particle swarm algorithm to carry out reverse search by taking a target hardness value as an optimization target, generating a candidate heat treatment system, carrying out pre-estimation sorting on an empirical relation of plasticity based on tempering temperature and time, according to the method, the number of tensile tests is remarkably reduced, the research and development cost is reduced, and the research and development period is shortened.

The invention relates to the technical field of tire simulation design, in particular to a method and system for establishing an electric vehicle tire bead sandwich type rubber-fiber composite simulation model. A target part is constructed as an equivalent beam / shell sandwich layer, a constraint area / movable area is defined and coupled with a rim, and a vertical load of P = p.A and a working condition of 4-15 Hz are applied. Taking a strain amplitude minimum surface as a neutral surface criterion, and extracting an S11 and an interface shearing amplitude; according to the method, delta h is reversely solved according to the established empirical relation, rapid iteration is carried out within the range of 25-40 MPa of a constraint area, 7.5-15 MPa of an active area and delta h = 1-5 mm, so that compression and shear peak values of a tire body on the turn-up side are synchronously reduced, a neutral plane is stable, modeling time is reduced to the hour level from the week level, and the method is suitable for tire bead turn-up structure design and process optimization.

The invention discloses a method for constructing a physical model based on machine learning, and the method is used for building a physical model with extrapolation capability and applying the physical model to the life prediction problem of a high polymer material. The method comprises the following steps that firstly, a material creep data set is collected and preprocessed; secondly, establishing an empirical relational expression with a plurality of unknown parameters, and fitting the parameters by using a least square method; generating a corresponding functional expression according to the relationship between the obtained parameters and conditions by using a machine learning method so as to deduce unknown parameters in the relational expression under other conditions, and performing comparative analysis on a predicted value and an actual value; and finally, the creep condition of the material under the corresponding condition is calculated by using the obtained relational expression. Experiments prove that the method can enable a physical model without extrapolation capability to obtain certain extrapolation capability, and the error between a predicted value and an actual value is also kept in an ideal range, so that the material degradation rules under different conditions are predicted.

The invention provides a well-constraint-free pseudo well establishment method and device based on seismic amplitude and electronic equipment. The method comprises the steps of performing amplitude preservation processing on original seismic data of a target work area, and generating a high-fidelity common reflection point gather through pre-stack migration; based on the common reflection point gather, generating a sub-angle superposition data volume; performing inversion through a linear inversion method to obtain a longitudinal wave impedance data volume of the target work area; establishing a quantitative empirical relationship between the longitudinal wave velocity and the transverse wave velocity and a quantitative empirical relationship between the longitudinal wave velocity and the density through rock physical analysis or statistical regression; and selecting a favorable structural point as a pseudo-well position in the target work area, extracting longitudinal wave speed information of the position from the longitudinal wave impedance data body, and calculating and generating a pseudo-well curve set including transverse wave speed, density and longitudinal wave impedance by using a quantitative empirical relationship. The pseudo-well curve can be constructed only based on seismic data without actually measuring well data, and the method is suitable for exploration of new areas or areas with insufficient well control.

The application discloses a shale reservoir total organic carbon content TOC calculation method, and relates to the technical field of oil and gas exploration. Including: obtaining a gray scale imaging graph of the shale reservoir, and deriving a depth index column; drawing a TOC index curve through the gray scale imaging graph and the depth index column I_TOC ; resampling the TOC index curve I_TOC according to the precision of measured TOC data to obtain imaging TOC data; drawing a scatter diagram by taking the imaging TOC data as an independent variable and taking the measured TOC data as a dependent variable; performing linear fitting analysis on the scatter diagram to obtain an empirical relationship; and substituting the TOC index curve I_TOC into the empirical relationship to obtain a shale reservoir total organic carbon content TOC curve TOC_FMI . The application can sensitively reflect the subtle, heterogeneous and nonlinear total organic carbon content TOC change of the formation near a well wall.