27 results about "Cloud atlas" patented technology

The invention relates to the technical field of Beidou inspection data processing, and discloses an automatic inspection data acquisition and space-time association method under Beidou space-time reference. The method comprises the following steps: acquiring historical working condition data of inspection equipment based on nanosecond-level time synchronization and centimeter-level space positioning information of a Beidou system, and constructing a space-time dynamic prediction model; acquiring an equipment sensor measurement value, a moving track point set and a timestamp sequence in real time, and outputting a theoretical inspection value through a model; performing space-time multi-dimensional comparison on the theoretical value and the actual value, analyzing a time accumulation error, a space deviation degree and data flow form consistency, and generating a difference feature tensor; inputting the tensor into a geocoding topological reasoning network, simulating anomaly diffusion and generating an anomaly probability cloud picture in combination with topographic map node attributes and equipment distribution, and delineating an anomaly range; polling parameters are dynamically configured according to the cloud atlas, enhanced collection is started for a high-probability abnormal area, and data credibility inspection is executed on peripheral equipment.

The application provides a wireless charging coil type and mutual inductance identification method, establishes a circular transmitting coil and a circular receiving coil model, a square transmitting coil and a square receiving coil model, a circular transmitting coil and a square receiving coil model, a square transmitting coil and a circular receiving coil model, obtains a magnetic field cloud atlas and mutual inductance values of the receiving coil and the transmitting coil, and constructs a training data set; taking pixel values of the coil magnetic field cloud atlas as input and taking the coil type as output, a coil type identification model is constructed; taking the pixel values of the coil magnetic field cloud atlas as input and taking the transmitting coil and the receiving coil as output, a coil mutual inductance identification model is constructed; a loss function is set, and the coil type identification model and the coil mutual inductance identification model are trained, so as to identify the wireless charging coil type and mutual inductance. The application realizes identification of the wireless charging coil type and mutual inductance, and provides a new idea for wireless power transmission method research and engineering application.

The present application relates to the technical field of suspended sediment dynamic monitoring, and particularly relates to a suspended sediment dynamic change monitoring and deep sea operation scheduling optimization method, comprising multi-source fusion data acquisition: synchronously acquiring suspended sediment concentration distribution data, three-dimensional sea current vector data and seabed microtopography slope data of a target area; plume flow dynamic diffusion simulation: inputting the acquired data into a microtopography coupled adaptive plume flow diffusion model to generate a plume flow diffusion three-dimensional space-time cloud atlas for the next 6 hours; dynamic risk zoning generation: performing space-time superposition analysis on the plume flow diffusion three-dimensional space-time cloud atlas and a preset ecological sensitive area dynamic electronic fence to generate an operation risk heat map; intelligent zoning scheduling execution: dynamically dividing a forbidden mining area, a buffer area and a safe area, and sending zoning interval rotation scheduling instructions to a mining equipment group. The present application improves the spatial resolution and time response capability of sediment diffusion behavior prediction, and provides a basis for environmental impact assessment and ecological protection.

The application relates to a kind of artificial precipitation enhancement operation area identification method and system based on multi-source data, and relates to meteorological intervention operation technical field, comprising the following steps: establishing multi-resolution dynamic data grid system, coarse granularity grid layer covers entire operation area, for carrying satellite cloud atlas and the macro data of conventional radar;Fine grid layer, for receiving and mapping the real-time data provided by high-resolution radar;Receive all cloud block real-time data and map to corresponding grid layer, project the data collected at different time points to unified time reference frame to obtain cloud system state view;Using the latest data of conventional radar and the macro trend of satellite cloud atlas, combined with motion vector and confidence weighted, predictive interpolation is carried out to obtain cloud system information;Form a visual cloud system three-dimensional view, identify and track target cloud block with artificial precipitation enhancement potential and generate operation instruction.The application has the effect that target cloud block with operation value can be identified and tracked to generate accurate operation instruction.

The invention discloses a photovoltaic ultra-short-term power prediction method based on cloud picture multi-feature extraction, and belongs to the field of electric power, and the prediction method comprises the steps: obtaining a change image of a cloud cluster, and obtaining a ground-based cloud picture image sequence; preprocessing all cloud pictures in the ground-based cloud picture image sequence to obtain a preprocessed image sequence; obtaining cloud picture features based on the preprocessed image sequence; and carrying out deep learning on historical photovoltaic output power data and the cloud picture features to obtain a photovoltaic ultra-short-term power prediction model so as to carry out photovoltaic ultra-short-term power prediction. The problem that in the prior art, cloud picture feature extraction is insufficient, and consequently the precision of a prediction result is too low is solved. Therefore, through accurate analysis of the features of the cloud atlas and combination of CNN and Autoformer methods, a photovoltaic ultra-short-term power prediction model is constructed, and then photovoltaic ultra-short-term power prediction is carried out.

The invention provides a deep vertical shaft double-air-duct press-in type air supply air duct position optimization method which comprises the following steps: (1) collecting parameters of a vertical shaft and air ducts, and calculating the blasting smoke throwing distance and the generated CO initial concentration according to the amount of explosive needed by blasting; (2) constructing a physical model matched with actual engineering parameters by adopting numerical simulation software based on the component transportation model, and setting different relative arrangement positions of double air ducts; (3) analyzing the derived simulation data at different moments and the related visual cloud atlas, counting the CO concentration condition in the working area, and fitting a curve of the CO concentration changing along with time; and (4) by taking the CO safe concentration threshold value as a reference, analyzing the ventilation time required by the standard reaching of the CO concentration of the operation area under each arrangement working condition, sorting and drawing a chart. According to the method, the ventilation efficiency change characteristics under different air duct arrangement angles can be effectively determined, and scientific reference is provided for optimal air duct arrangement scheme selection of various deep vertical shaft engineering double-air-duct press-in type air supply systems.

The application discloses a gradient lattice structure compression simulation method and system based on layered finite element simulation, and relates to the field of finite element simulation.The technical scheme is as follows: a finite element model is established, and simulation data is collected; a load-displacement curve is constructed according to displacement data and load data; layer densification displacement is calculated according to layer densification strain, and the part of the load-displacement curve before the layer densification displacement is intercepted as an effective curve; the effective curve is synthesized to obtain load data and displacement data of the gradient lattice structure; a stress-strain curve is constructed according to the load data and the displacement data of the gradient lattice structure; a stress value is determined according to a preset strain value and a matching strain layer, and the stress value is converted into a load to call a stress-strain cloud atlas, and a total stress-strain cloud atlas is obtained by pasting stress-strain cloud atlases of all layers.The application can solve the problems of large calculation amount, high time cost and difficulty in convergence in finite element simulation of the gradient lattice structure.

The invention belongs to the technical field of image detection, and discloses a finite element cloud picture crack fitting method based on direction priori and dark quantile ROI, which comprises the following steps: calculating a preliminary direction angle of a crack according to a reference point coordinate as the direction priori; taking the direction prior as a central axis direction and one of the reference points as a benchmark, constructing a strip-type ROI, extracting brightness values of pixels in the strip-type ROI, and screening the pixels with the brightness values lower than or equal to a threshold value as a preliminary crack candidate point set; calculating the local gradient direction of each pixel point in the candidate point set, screening out the pixel points of which the deviation between the local gradient direction and the direction prior is smaller than a preset direction tolerance, and forming a direction consistency set; and carrying out total least square fitting on the coordinates of the pixel points in the direction consistency set to obtain a main direction straight line and fitting parameters of the crack, and superposing the main direction straight line on the original finite element cloud picture to generate a crack fitting result. According to the method, robust and high-precision extraction of the crack direction in the complex finite element cloud picture is realized.

The application discloses a kind of welding molten pool image prediction weld appearance deep learning method, it is related to welding technical field, including the following steps: first, by the two CCD cameras of upper and lower arrangement from different angles to the molten pool picture of the same position at the same time when welding is collected, molten pool picture is carried out cutting, filtering and other image processing, after using target paper calibrates camera, by three-dimensional reconstruction, obtain the three-dimensional cloud atlas of molten pool, the three-dimensional curved surface relationship z=f (x, y) of molten pool surface is fitted, using pixel instead of the horizontal and vertical coordinates of molten pool, the size of gray scale value indicates the fluctuation of molten pool surface, it is converted from three-dimensional image into two-dimensional image containing three-dimensional information, using optimized loss function to improve convolutional neural network calculation, extract and learn the molten pool information in two-dimensional molten pool image, the appearance of weld is predicted.The application solves the problem of insufficient accuracy caused by using only welding parameters to predict the appearance of weld.

The application discloses a mechanism decoupling transfer learning driven composite bolt connection failure efficient prediction method, relates to the technical field of composite material failure research, and comprises the following steps: a high-throughput simulation database is established, unstructured grids are converted into structured tensors by using a physical coordinate mapping strategy, and a generative network containing coordinate injection and gradient consistency constraints is constructed, so that the second-level high-precision prediction of a composite material connection damage cloud atlas under any geometric parameters is realized. The method is used to effectively fill the blank of the prior art, and provides a new failure prediction mode for composite material bolts with different layer designs, width-diameter ratios and end-diameter ratios.

The present application relates to the technical field of photovoltaic system, solve the technical problem that the prior art cannot accurately predict irradiance and temperature, the prediction process is easy to appear gradient disappearance or gradient explosion, especially relates to a kind of photovoltaic power prediction method, system and computer equipment, the method includes the following steps: S1, obtain the cloud atlas data set from the weather database in the place of photovoltaic equipment;S2, select several zenith cloud atlas from cloud atlas data set and obtain ground cloud atlas set consisting of several ground cloud atlas by preprocessing;The present application obtains the moving curve of cloud layer in the sky by image analysis to zenith cloud atlas, and establishes prediction model according to moving curve, can avoid the problem of gradient disappearance or gradient explosion of neural network model, also carries out real-time cloud layer detection to zenith cloud atlas, so that the prediction result is more accurate, improves the accuracy of irradiance, makes the prediction process more smooth, improves work efficiency.

The invention relates to a construction method and system for a large-section underground excavation tunnel, and belongs to the technical field of underground excavation tunnel construction. The construction method comprises the following steps that a three-dimensional model is established, and the influence of tunnel construction on surrounding stratum disturbance is analyzed in a numerical simulation mode; determining construction parameters including an excavation sequence based on a numerical simulation result; the pilot tunnel is excavated according to the prepared excavation sequence, a supporting structure comprising a primary support and a temporary support is arranged, and arrangement of the temporary support can be optimized in a targeted mode according to a cloud picture of the stress change process of the supporting structure in the prepared excavation sequence; and the stress state of the supporting structure is collected in real time, and the monitoring data and the predicted value in the stress change cloud picture are compared and analyzed so as to adjust the supporting scheme. The construction system comprises a simulation unit, a supporting structure, a monitoring unit and a central control unit, and the construction system can execute the construction method to assist in completing large-section underground excavation tunnel construction.

The invention discloses a bridge cluster traffic flow load cloud picture deduction method based on wide-area road network intermodulation data, and the method comprises the steps: integrating the section traffic parameters of each intermodulation station in a road network, and building a new intermodulation parameter system for bridge mechanical analysis; generating a load cloud picture reflecting the distribution of new parameters in the whole road network by using a kernel density analysis algorithm of a geographic information system (GIS) based on the geographic position information of each station; and constructing a multi-feature agent model based on the new intermodulation parameter system, inputting latitude and longitude coordinates of any target bridge in the road network into the multi-feature agent model, deducing load parameters of the position of the bridge by the model, and further recovering a full-bridge-floor traffic flow load time history of the target bridge through a traffic flow simulation technology. The method has the advantages that large-scale conversion from cross-modulation data to bridge mechanical analysis parameters is realized, dynamic load input is provided for a bridge cluster, and a maintenance mode is promoted to jump from'single bridge supervision 'to'bridge network intelligent management'.

The invention belongs to the related technical field of engineering matching construction, and discloses a remote control method for high-quality wet spraying operation of a wet spraying machine, which comprises the following steps: starting a positioning detection module, and performing position positioning of the wet spraying machine; starting a 3D imaging module, scanning feature points on the rock surface and the wet spraying machine body, generating a real three-dimensional point cloud model, and determining absolute coordinates of the boom and the rock surface; a joint detection module is started, and initial state parameters of joints are collected; calculating the current position of the nozzle; an over-excavation and under-excavation cloud picture is obtained; the rotating angle and the telescopic distance of the joint are adjusted; and along with the movement of the wet spraying machine, the optical fiber of the remote transmission module is operated through the unwinding disc. Through the system, high-precision positioning of the positions of the wet spraying machine and the arm support can be achieved, dynamic and real-time adjustment can be carried out on various spraying parameters, controllability of operation quality is remarkably improved, and meanwhile the system has the advantages of being convenient to control, high in automation degree, low in remote control delay, stable, good in environmental adaptability and the like.

The invention discloses an inter-axis self-locking nut closing-in process variable parameter influence analysis method based on finite elements. Comprising the steps that S1, a finite element model containing the shaft self-locking nut and the closing-in device is constructed; s2, configuring a material constitutive parameter, a grid parameter, a contact parameter and a boundary parameter; s3, carrying out univariate simulation by taking the key process parameters as variables, and analyzing the forming performance by combining point tracking and cloud atlas; s4, outputting a process optimization scheme; through finite element model construction, multi-dimensional parameter configuration and univariate simulation analysis, the influence rule of key process parameters on equivalent stress, damage distribution and forming load is accurately obtained, the optimal process parameters giving consideration to the forming quality and the production efficiency are determined, and a reliable theoretical basis is provided for actual production process adjustment.

The invention discloses a static pressure rotary table temperature field prediction and cloud picture visualization method, which comprises the following steps: constructing a multivariable time sequence based on rotary table operation conditions and environment variables, and carrying out joint prediction on a plurality of temperature measurement points by using a multi-output BiLSTM agent model to obtain predicted temperature vectors of discrete measurement points; taking the predicted temperature as a boundary condition of a voxel thermal field, and solving a pseudo-physical thermal diffusion equation in a three-dimensional voxel space to obtain a continuous temperature field; constructing a local correction model based on linear regression at a sensor position, and spreading a correction value in space through inverse distance weighting to realize unification of data consistency and spatial continuity; and a real-time temperature cloud picture is generated at the GPU end through Gaussian weight voxel weighted rendering based on Shader. According to the method, data-driven prediction and physical consistency constraint can be organically fused, and real-time visualization and thermal state evaluation of the internal temperature field of the hydrostatic pressure turntable can be realized in an industrial field.

The invention relates to the technical field of image processing, in particular to a frozen wall closed state detection method and system based on temperature cloud picture analysis. The method comprises the following steps: acquiring an anti-permeability intensity image of a horizontal section with a to-be-detected depth; obtaining a binary image of the anti-permeability intensity image, extracting a heat conduction skeleton of the frozen wall from the binary image by using a central axis transformation algorithm, and performing multiple sampling on the heat conduction skeleton to obtain a discrete sampling point sequence; obtaining the effective thermal resistance thickness at each discrete sampling point; the geometric curvature of each discrete sampling point is obtained, the product of the geometric curvature and the water flow scouring strength factor serves as an adjusting coefficient, the ratio of the effective thermal resistance thickness to the adjusting coefficient serves as a buckling stability index, and the closed state of the frozen wall is determined according to the comparison result of the global minimum value of the buckling stability index and a preset safety threshold value. The method and the device have the effect of improving the frozen wall state detection accuracy.

The invention aims to provide an electric arc additive repairing method for a 2205 duplex stainless steel plate, which belongs to the technical field of metal repairing processing, and comprises the following steps: pretreating a to-be-repaired area of a 2205 duplex stainless steel plate substrate; the method comprises the following steps: constructing an original data model through three-dimensional modeling software according to the length, width and height of a 2205 duplex stainless steel plate substrate, assigning a reference point near a to-be-repaired area, performing multi-directional scanning from the reference point through a three-dimensional scanner to obtain point cloud data, extracting feature points to establish a wear model, collecting three-dimensional cloud pictures of the substrate and the to-be-repaired area, and performing three-dimensional repair on the matrix and the to-be-repaired area. Performing hierarchical slicing on the to-be-repaired area, and calculating the length, width and height of the to-be-repaired area; setting process parameters of electric arc additive repair, and calculating a repair path; an electric arc is adopted as a heat source to implement additive repairing, and after repairing is completed, heat treatment, polishing post-treatment and detection work are sequentially carried out on a repairing area. The repairing method is simple, efficient and suitable for large-scale production.

The application relates to the technical field of image processing, in particular to a frozen wall closure state detection method and system based on temperature cloud atlas analysis. The method comprises the following steps: acquiring an anti-permeability strength image of a horizontal section of a detection depth; acquiring a binary image of the anti-permeability strength image; extracting a heat conduction skeleton of a frozen wall from the binary image by using a center axis transformation algorithm; performing multiple sampling on the heat conduction skeleton to obtain a discrete sampling point sequence; acquiring effective thermal resistance thickness at each discrete sampling point; acquiring the geometric curvature of each discrete sampling point; taking the product of the geometric curvature and a water flow scouring intensity factor as an adjustment coefficient; taking the ratio of the effective thermal resistance thickness and the adjustment coefficient as a buckling stability index; and determining the closure state of the frozen wall according to the comparison result of the global minimum value of the buckling stability index and a preset safety threshold. The application has the effect of improving the detection accuracy of the frozen wall state.

The application discloses a kind of meteorological satellite visible light cloud chart generation method and system, it is related to meteorological satellite technical field, the method includes: based on the multi-channel imaging equipment containing infrared and visible light satellite observation, collect the infrared cloud chart set of night target to be carried out cloud chart conversion;Carry out cloud chart data block, obtain block infrared cloud chart data;Call seasonal time-domain cloud chart in predetermined time interval;Train difference adaptive calibration model;Train cloud chart conversion model;Combining the difference adaptive calibration model and cloud chart conversion model of front and rear position connection, the independent processing of block cloud chart is carried out;To block conversion cloud chart is spliced, generates daytime target visible light cloud chart set.The application solves the technical problem that the feature difference of existing technology in daytime and night infrared cloud chart causes the visible light cloud chart difference of adjacent time generation is larger, reaches the technical effect of intelligent analysis infrared cloud chart difference feature, improves the accuracy of visible light cloud chart generation.

The invention discloses a fractured geothermal reservoir temperature field intelligent prediction method, system and device based on a CRA-UNet model. Belongs to the technical field of geothermal reservoir temperature field prediction. S100, constructing a multi-modal data set, wherein the multi-modal data set comprises pairing data of the crack form image and the boundary condition parameter, and a temperature field label; s200, a CRA-UNet model is trained by adopting the multi-modal data set, and the trained CRA-UNet model is obtained; and S300, inputting the pairing data of the new crack form image and the boundary condition parameters into the trained CRA-UNet model, and outputting a predicted temperature field with the same size as the input crack form image. According to the method, the fracture form image and the injection-production boundary condition are used as input, and accurate and rapid output of the fracture type reservoir temperature field cloud atlas at any time node is achieved.

The application belongs to the technical field of image detection, and discloses a finite element cloud atlas crack fitting method based on direction prior and dark subinterval ROI, which comprises the following steps: calculating a preliminary direction angle of the crack according to the reference point coordinates, and taking the preliminary direction angle as the direction prior; taking the direction prior as the central axis direction, and taking one of the reference points as the benchmark to construct a strip type ROI; extracting the brightness values of the pixels in the strip type ROI, and screening the pixels with brightness values lower than or equal to a threshold value as a preliminary crack candidate point set; calculating the local gradient directions of the pixel points in the candidate point set, screening out the pixel points with a deviation between the local gradient direction and the direction prior less than a preset direction tolerance to form a direction consistency set; performing total least squares fitting on the pixel point coordinates in the direction consistency set to obtain a main direction straight line of the crack and fitting parameters, and superimposing the main direction straight line on the original finite element cloud atlas to generate a crack fitting result.

The invention relates to an AI intelligent regulation and control koji method based on koji room temperature and humidity field real-time feedback, and the method comprises the steps: simulating the temperature and humidity fields of a koji room under different working conditions through simulation software, generating a large number of data samples, and expanding and perfecting an AI cloud picture model training database; and a prediction result output by the AI cloud picture model is converted into a visual temperature and humidity cloud picture through a Python algorithm. A multi-objective optimization function is constructed in combination with a core process objective of koji room temperature and humidity regulation and control, an AI cloud picture model is utilized to predict temperature and humidity field evolution trends under different regulation and control schemes, a Pareto optimal solution set is solved, and executable regulation and control suggestions are formed and presented in a visual mode. Along with continuous accumulation of the quality scoring results of the yeast blocks, the multi-objective optimization function can be dynamically corrected based on a reinforcement learning algorithm, and self-adaptive evolution of the optimization model is achieved. According to the invention, accurate perception and dynamic balance regulation and control of the three-dimensional temperature and humidity field in the koji room can be realized, and the optimal curvature is greatly improved.

The invention relates to a surface residual stress prediction method based on a milling surface topography model and a deep learning model, and the method comprises the following steps: obtaining bottom edge size data and milling process parameters of an end milling cutter, and carrying out the analysis prediction of single-pass surface topography and the simulation prediction of residual stress in a single-pass milling process; extracting surface topography geometry and residual stress distribution key data according to the calculated surface topography and the simulated residual stress nephogram; establishing a residual stress prediction model of single-pass milling based on deep learning; and based on a residual stress prediction result output by the single-pass milling residual stress prediction model, considering the influence of the multi-pass superposition effect, and realizing prediction of the multi-pass milling residual stress under different superposition effects. Compared with the prior art, the method has the advantages that the residual stress distribution in the multi-pass milling process is efficiently predicted, the residual stress is extended from two-dimensional depth prediction to three-dimensional prediction, and the prediction precision of the residual stress of the machined surface in the milling process is improved.

The invention provides a telescopic pallet fork safety assessment method combined with finite element analysis. The method comprises the steps of load determination, wherein the load borne by a telescopic pallet fork in the working process is determined; a model selection and establishment step: establishing a finite element analysis model; material setting, wherein material parameters conforming to the actual working condition are set for all components of the telescopic pallet fork; a boundary condition adding step: adding a boundary condition in the finite element analysis model; grid division: performing grid division on the finite element analysis model; a result analysis step: obtaining a displacement cloud picture, a stress cloud picture and a strain cloud picture of the telescopic pallet fork under a given load, and extracting data; and a security assessment step: performing security assessment. According to the method, the finite element model is constructed, dynamic load input and multi-physical field coupling analysis are combined, the displacement, stress and strain cloud atlas of the pallet fork under the given load can be generated, and a basis is provided for strength checking, fatigue life prediction and lightweight design of the pallet fork.

The invention discloses a multi-mode photovoltaic power prediction method based on cloud picture segmentation and cloud genus cloud cover guidance. The foundation cloud picture passes through a foundation cloud picture fine-grained segmentation model to obtain a foundation cloud picture segmentation mask; according to the ground-based cloud picture segmentation mask, the cloud amount of each cloud genus is calculated, and the cloud amounts of all the cloud genus form a cloud genus cloud amount distribution feature tag; the photovoltaic power prediction model comprises a weight generator module, a cloud picture coding module, a numerical data coding module, a cross-modal fusion module and a hybrid predictor module; the weight generator module generates weights of sub-predictors according to cloud genus cloud quantity distribution feature tags of historical time steps, each sub-predictor of the hybrid predictor module performs photovoltaic power prediction according to cross-modal fusion features, and weighted summation is performed on predicted power of each sub-predictor to obtain final predicted photovoltaic power. And the prediction power of the sub-predictors is fused by considering a cloud layer semantic environment and taking the cloud genus cloud amount as guidance, so that the prediction performance under different cloud layer conditions is improved.

The application discloses a data post-processing method, device and equipment. The method comprises the following steps: obtaining cloud atlas data of a target part, and establishing a target node set based on the cloud atlas data; obtaining screening information set by a target object, determining a plurality of target nodes from the target node set based on the screening information, and determining associated nodes of each target node; determining a view angle vector and a characteristic size corresponding to each target node based on the associated nodes of each target node, wherein the view angle vector is used for determining a view angle direction of a local cloud atlas corresponding to the target node, and the characteristic size is used for determining a scaling ratio of the local cloud atlas corresponding to the target node; generating a local cloud atlas corresponding to each target node based on the view angle vector and the characteristic size, saving the local cloud atlas, and saving information of the target node and the associated nodes. The application solves the technical problem that the post-processing result of the cloud atlas data in the calculation result cannot meet the demand in the finite element simulation analysis process.