21results about How to "Achieve high precision" patented technology

The invention discloses a building pipeline intelligent flaw detection and leakage positioning system, which relates to the technical field of building pipeline detection and comprises a data acquisition module, a multi-modal fusion module, a defect identification module, a leakage positioning module, a digital twinborn module, a health prediction module and a decision control module. The data acquisition module is used for deploying multiple types of sensors along a building pipeline, acquiring acoustic emission signals, infrared thermal imaging images, pressure fluctuation data and fluxgate detection signals in a running state of the pipeline, and outputting original sensing data; the multi-modal fusion module is connected with the data acquisition module and is used for performing feature extraction on the original sensing data, constructing a multi-modal feature vector, dynamically adjusting each modal weight based on environmental parameters and outputting a weighted fusion feature vector; and the defect identification module is connected with the multi-modal fusion module and is used for inputting the weighted fusion feature vector into a deep neural network model.

The invention provides a digital quantum simulation method, which comprises the following steps of: performing symmetry analysis and subspace decomposition on a target Hamiltonian to obtain an effective Hamiltonian and a symmetry constraint condition; constructing a target unitary operator according to the effective Hamiltonian, and initializing a quantum circuit population of the target unitary operator; calculating entanglement characteristics of each quantum line in the quantum line population, performing structure variation on the quantum line population based on the entanglement characteristics, and performing line fragment crossing in combination with a symmetry constraint condition to obtain a candidate line set; performing line parameter optimization on the candidate line set based on the fidelity gradient to obtain an optimized line set, and extracting gate distribution structure features of the optimized line set; and non-uniform sampling is carried out on the optimized line set based on gate distribution structure characteristics to obtain a sampling line subset, an optimal quantum line is screened out from the sampling line subset based on the deviation between an experimental observation value and a theoretical expected value, and quantum simulation is carried out based on the optimal quantum line.

This invention provides a real-time adaptive 3D measurement system and method for objects based on microstructured light. The system includes: multiple optical measurement units arranged in an equidistant circular array along the circumference on a ring mounting frame and moving synchronously with a lifting and adjusting mechanism; multiple measurement posture acquisition units, a reference posture monitoring unit, a posture compensation unit, a control unit, a data fusion module, and a quality analysis module. According to this invention, the multiple optical measurement units arranged in a circular array enable multi-view synchronous measurement of the workpiece, avoiding the time consumption and reference conversion errors of traditional indexing measurement methods. The real-time posture monitoring and dynamic compensation mechanism effectively suppresses measurement errors caused by factors such as mechanical vibration and temperature drift, reducing posture-related measurement errors. The system achieves automation, real-time measurement, and high precision in 3D measurement, improving measurement efficiency and accuracy.

The application provides a table structure identification method and device and a computer readable storage medium. A table structure identification model capable of reasoning blank cell coordinates is trained, a table to be identified is obtained, and the table to be identified is input into the table structure identification model. The model automatically completes reasoning, high-precision, integrity and consistency identification of the wireless table structure of the table to be identified is achieved, and core technical problems such as table structure fracture and row and column misjudgment caused by missing blank cell coordinates in the prior art are effectively solved.

The application discloses a kind of dynamic spatio-temporal graph learning microservice exception detection methods for multi-modal data, which comprises: collecting data information from the microservice system to be detected and inputting into computer system and carrying out data preprocessing, obtaining multi-modal data;The multi-modal data is processed based on the sliding window mechanism, and a dynamic dependency graph sequence is constructed;Based on dynamic dependency graph sequence, spatial relationship modeling and time series dynamic evolution modeling are performed, a representation integrating time series information is generated, and a graph-level representation vector containing complete spatio-temporal context is generated;The graph-level representation vector is input into a deep vector data description anomaly detection model based on contrast learning, and the detection result is output.The microservice exception detection based on multi-modal dynamic spatio-temporal graph learning and contrast enhancement SVDD of the application explicitly captures the dynamic evolution characteristics of microservice topology structure, and realizes the deep fusion of spatio-temporal coupling characteristics, while suppressing the decision boundary ambiguity problem in unsupervised detection.

This application discloses a method and apparatus for determining the grade of a non-volatile memory chip. The method includes: acquiring current test data of frequency, capacity, and timing of the non-volatile memory chip, as well as failure assessment test data; determining a dynamic weighting matrix based on the usage scenario information of the non-volatile memory chip; mapping the current test data to a pre-constructed first chip feature space based on the dynamic weighting matrix and calculating the corresponding likelihood probability; determining the current data score grade based on the likelihood probability; calculating a personalized degradation rate parameter based on the failure assessment test data and a pre-constructed chip degradation trajectory curve; determining the failure assessment data based on the personalized degradation rate parameter; and determining the comprehensive grade corresponding to the current data score grade and the binary grading interval to which the failure assessment data belongs as the target comprehensive grade of the non-volatile memory chip. This technical solution achieves automated, multi-dimensional, and high-precision grade classification of non-volatile memory chips.

The application discloses a kind of marine engineering steel plate strength detection devices, it is related to marine engineering steel plate strength detection technical field, including data processing module;Data processing module includes contact state analysis unit, steel plate strength analysis unit;Contact state analysis unit carries out analysis to the contact state of indenter and steel plate, obtains the moment when indenter and steel plate contact, the time region between contact moment and current moment is marked as strength detection time zone;Steel plate strength analysis unit carries out state analysis to the strength of steel plate in strength detection time zone and obtains steel plate strength analysis result.The application determines the moment when indenter and steel plate contact and reasonably demarcates strength detection time zone by the contact state analysis unit pressure value operation and comparison mechanism, lays the analysis foundation for subsequent detection, and steel plate strength analysis unit discriminates steel plate quality by index calculation and evaluation process, guarantees the reliability of steel plate quality.

The invention discloses a multi-modal data-oriented dynamic space-time diagram learning micro-service anomaly detection method, which comprises the following steps of: acquiring data information from a micro-service system to be detected, inputting the data information into a computer system, and carrying out data preprocessing to obtain multi-modal data; processing the multi-modal data based on a sliding window mechanism, and constructing a dynamic dependency graph sequence; on the basis of the dynamic dependency graph sequence, space relation modeling and time sequence dynamic evolution modeling are executed, representation of fusion time sequence information is generated, and a graph-level representation vector containing a complete spatio-temporal context is generated; and inputting the graph-level representation vector into a depth vector data description anomaly detection model based on contrast learning, and outputting a detection result. According to the method, based on multi-mode dynamic space-time diagram learning and contrast enhancement SVDD micro-service anomaly detection, dynamic evolution characteristics of a micro-service topological structure are explicitly captured, deep fusion of space-time coupling characteristics is realized, and meanwhile, the problem of fuzzy decision boundary in unsupervised detection is inhibited.

This invention relates to the field of nondestructive testing equipment, specifically to a magnetic particle inspection clamping device for nondestructive testing of automotive chassis parts. The device includes a collection box, a feeding assembly, a clamping mechanism, and a top-holding assembly. The collection box contains an inclined collection plate. The feeding assembly includes two clamping plates. The clamping mechanism includes a power chuck and a discharge push rod. The power chuck has grippers. The top-holding assembly includes a top-holding rod. Through the clamping mechanism and the top-holding assembly, the problems of material jamming and obstructed view on high-speed production lines are effectively solved. During feeding, the top-holding rod and the V-shaped groove formed by the two clamping plates cooperate to push the workpiece into the power chuck and lock it in place. During unloading, the discharge push rod forcibly pushes the ball pin down, breaking the liquid bridge effect and residual magnetic attraction between the workpiece and the grippers, thus improving the continuous operation stability of the equipment.

The present application relates to the technical field of centrifuge control, and particularly relates to a centrifuge temperature control parameter optimization method and system, which comprises the following steps: collecting multiple parameters of the centrifuge to construct a sample data set; training an XGBoost stage classification model based on the sample data set, and outputting the working condition probability of the centrifuge, presetting the chamber temperature and refrigeration power of the centrifuge at the next moment, calculating the temperature precision penalty coefficient, the temperature change rate risk coefficient and the energy consumption penalty coefficient, dynamically adjusting the weights of the temperature precision penalty coefficient, the temperature change rate risk coefficient and the energy consumption penalty coefficient based on the proportion of steady speed, speed up and speed down in the real-time working condition to construct a target function, and calculating the comprehensive control error; selecting multiple groups of historical data most similar to the current working condition as reference data, selecting the compressor speed corresponding to the reference data that minimizes the comprehensive control error as the optimal speed by traversing each reference data, and delivering the optimal speed to the compressor for execution, thereby improving the temperature control precision of the centrifuge.

The application provides a turning machining system and method for complex parts, a machining surface relationship diagram of a target cutting machining area is constructed through a three-dimensional CAD model of a complex part to be turned, a feature extraction surface containing a closed inner ring is selected from the machining surface relationship diagram, and a candidate feature surface set of the target cutting machining area is determined, the candidate feature surface set is compared with a preset non-machining structure surface set, if there is a common surface between the two, the machining surface belonging to the non-machining structure surface set is removed, and the topological connection relationship of the target cutting machining area is reconstructed, the machining feature type of the target cutting machining area is determined through the reconstructed topological connection relationship of the target cutting machining area, the confidence turning tool path is generated according to the machining feature type, and the complex part to be turned is turned through the confidence turning tool path. The technical scheme provided by the application can turn the complex part through the effective tool path under the influence of the non-machining structure.

The invention discloses a ship detection method and device in a synthetic aperture radar image, equipment and a medium, and relates to the technical field of image detection, and the method comprises the steps: scanning an image from different directions to obtain corresponding global features, and extracting detail features corresponding to the image in different image regions; constructing a feature pyramid by using the global features and the detail features, and performing up-sampling on the features in the feature pyramid from top to bottom to obtain a target feature map; and analyzing the target feature map by using a regression head, a classification head and an angle head in the head network, and determining a bounding box and a ship angle corresponding to the target ship in the image based on an analysis result so as to complete ship detection. By performing multi-direction scanning on the image and integrating the head network of the angle prediction branch for analysis, a detection result with a direction is directly output, so that high-precision and directional detection of a ship in any orientation in the SAR image is realized.

The application relates to the technical field of casting molds and forming processes, in particular to a low-pressure casting mold for a magnetic suspension compressor shell and a use method thereof. The mold comprises a bottom mold and three opening and closing molds matched with the bottom mold; the bottom mold and the three opening and closing molds are combined with a sand core to form a casting blank cavity of the magnetic suspension compressor shell, the blank cavity is in a horizontal layout and only one sprue is arranged at a cylindrical outer center position of the blank cavity; the mold further comprises a point cooling mechanism, the point cooling mechanism comprises a plurality of interval-arranged point cooling units, the point cooling units are arranged at each slow cooling point position close to the casting and away from the sprue and at a local thick position of the casting, each point cooling unit is respectively connected with cooling liquid, each cooling position is cooled through the cooling liquid, and the metal liquid realizes sequential solidification from top to bottom and from both ends to the center after entering the blank cavity. The sequential solidification is realized through the multi-point cooling, and the forming quality and process yield of the magnetic suspension compressor shell are improved.

The display system (2) of the present invention comprises: a display element (12) having a display surface from which light for displaying an image is emitted forward via a polarizing member; a reflective portion (14) disposed in front of the display element (12), comprising a reflective polarizing member, and reflecting light emitted from the display element (12); a first lens portion (16) disposed between the display element (12) and the reflective portion (14); a semi-reflective mirror (18) disposed between the display element (12) and the first lens portion (16), allowing light emitted from the display element (12) to pass through and reflecting light after reflection by the reflective portion (14) toward the reflective portion (14); a first phase difference member (20) and a first member disposed in... The display element (12) and the semi-reflective mirror (18) are located between the display element (12) and the semi-reflective mirror (18); and the second phase difference member (22) and the second member are located between the semi-reflective mirror (18) and the reflective part (14). The first phase difference member (20) and the second phase difference member (22) are phase difference members that can convert linearly polarized light into circularly polarized light or convert circularly polarized light into linearly polarized light, respectively. The first member and the second member are members that impart a phase difference in the thickness direction. Between the display element (12) and the semi-reflective mirror (18) and between the semi-reflective mirror (18) and the reflective part (14), the members that impart a phase difference in the thickness direction are located in front of or behind the phase difference members (20, 22).

The present application provides a technique for achieving high accuracy of patterning of an optical surface while suppressing complication of processing. The present application performs manufacturing of a lens member, and has: a process of positioning a lens member (10) having a first surface and a second surface in a prescribed attitude, the first surface being a convex optical surface, and the second surface being an optical surface facing the first surface; a process of identifying surface shape data of the first surface using a dimension measurement result of the first surface of the positioned lens member (10); and a process of irradiating laser light to the first surface of the lens member (10) to perform laser processing on the first surface and controlling an irradiation position of the laser light based on the surface shape data.

This invention relates to the field of photoelectric measurement technology and provides a real-time measurement method for the flight trajectory of a target based on a multi-node orthogonal light curtain. The method includes: deploying at least two measurement nodes along the target's flight path; each node forming a two-dimensional detection light curtain through an orthogonal linear array photoelectric sensor and a collimated light source; using a high-speed signal processing unit to process the shading pulse signals from all sensor channels in parallel, and extracting the timestamps and two-dimensional position coordinates of the target as it passes through each light curtain in real time; and reconstructing the target's three-dimensional spatial position, velocity vector, and flight attitude at each moment in real time using spatial intersection and time difference algorithms based on the node spatial coordinates, timestamps, and two-dimensional coordinates. This invention enables high-precision and low-latency synchronous measurement of all parameters, including position, velocity, and attitude, of a high-speed flying target, providing crucial real-time trajectory data support for applications such as high-repetition-rate laser target firing.

The application discloses a highway service area peak period rapid charging pile scheduling and distribution method, comprising the following steps: step 1, obtaining vehicle side, service area side and environment side data and preprocessing; step 2, obtaining target service area and service area set, service area internal charging pile state matrix, potential target vehicle arrival time and arrival service area internal queuing vehicle time, and predicting charging pile working state in future short period; step 3, combining new energy vehicle information with charging demand in driving, nearby service area information and position information provided by a navigation system, and determining a charging mode according to user charging preference; and step 4, determining a final service area and charging pile distribution scheme. The application provides global charging load prediction capability for a highway operator from a system level, optimizes time and space distribution of charging resources from a source, and improves operation efficiency of the entire road network.

The application discloses a vital sign monitoring method and device based on longitudinal heart impact signals, and comprises the following steps: deploying the position of the beam direction of the radio frequency sensing device and the longitudinal axis of the target human body; collecting displacement data of the longitudinal heart impact signals through the radio frequency sensing device, and pre-processing the collected displacement data of the longitudinal heart impact signals to realize target human body positioning; performing target human body angle estimation and adaptive beam forming on the pre-processed displacement data to obtain an enhanced signal; performing static elimination and phase extraction on the enhanced signal to obtain a continuous displacement mapping sequence; using variational mode decomposition to adaptively separate the respiratory and heartbeat waveforms from the continuous displacement mapping sequence; and calculating vital sign parameters based on the separated respiratory and heartbeat waveforms. Therefore, the application can realize high-precision, robust and long-term monitoring of human body respiration and heartbeat without the human body carrying any sensor.

The invention discloses a liquid phase method silicon carbide single crystal growth system and a growth method. The system comprises a furnace body; the heat preservation barrel is located in the furnace body; the crucible is positioned in the heat preservation barrel; the heating body is located in the furnace body and surrounds the periphery of the heat preservation barrel; the upper end of the liftable seed crystal rod penetrates out of the furnace body, and the lower end of the liftable seed crystal rod is connected with the seed crystal disc and can extend into the crucible; and the radar liquid level meter is located at the top of the furnace body, and microwave signals emitted by the radar liquid level meter can reach the liquid level of the melt in the crucible. The growth system can stably work for a long time in a high-temperature, vacuum or inert gas protection environment, and can realize high-precision and real-time monitoring of the liquid level height in a growth cycle of 60-80 hours of the liquid-phase silicon carbide single crystal. By providing reliable liquid level height data for the growth control system, the stability of the solid-liquid interface position can be effectively maintained, and adverse effects of temperature gradient fluctuation on the crystal growth rate and morphology are avoided, so that the growth controllability, the crystal quality and the yield of the liquid phase method silicon carbide single crystal are improved.

The invention relates to the technical field of field intensity prediction, in particular to a field intensity prediction method of a high-power microwave radiation field and a related device. The method comprises the following steps: recording acquired data, determining field intensity data of microwave receiving antennas under each azimuth parameter according to each acquired sub-data, and constructing a sparse matrix field intensity data set according to the field intensity data of each microwave receiving antenna under each azimuth parameter; training a preset field intensity prediction model according to the sparse matrix field intensity data set to obtain a trained field intensity prediction model; and for each microwave receiving antenna, determining coordinates of a measurement point location, and inputting the coordinates into the field intensity prediction model to obtain a field intensity prediction value of the measurement point location output by the field intensity prediction model. According to the method, the complex nonlinear mapping relation, namely the mapping relation between the orientation parameters and the field intensity values, can be learned through the field intensity prediction model, deep features can be extracted from the collected data, high-precision field intensity prediction is achieved, and the prediction error is remarkably reduced.

The embodiment of the application provides a driver sitting posture management method, a vehicle, a computer program product and a storage medium, the method comprising: periodically acquiring multi-modal data of a driver sitting on a vehicle seat; in each period, determining a contact state between the driver and the vehicle seat according to the multi-modal data in the current period, determining a spine curve of the driver in the current period according to the contact state and the multi-modal data, and determining a spine state of the driver in the current period according to the spine curve; and outputting sitting posture adjustment prompt information corresponding to the spine state of the driver in each period. The application solves the technical problem that related technologies cannot comprehensively and accurately analyze the spine state of the driver in the full contact state range, so that abnormal changes in the sitting posture of the driver cannot be identified and corrected in time.