1102 results about "Reconstruction algorithm" patented technology

The invention provides a news scene three-dimensional reconstruction and visualization method based on multi-source remote sensing data, and relates to the technical field of three-dimensional reconstruction, and the method comprises the steps: extracting multi-modal data feature points through employing a spatial position weighted iterative nearest point algorithm and a tensor projection algorithm; after bidirectional attention deep neural network processing and local geometric consistency constraint screening, registration multi-source remote sensing data is generated, data enhancement is carried out through a variational auto-encoder network and a generative adversarial network, the enhanced data is used for depth information generation and texture feature migration, an initial three-dimensional grid model is generated based on a Poisson reconstruction algorithm, and the initial three-dimensional grid model is constructed; and carrying out topological optimization by combining multi-view projection consistency constraint and a grid optimization algorithm to obtain an optimized three-dimensional scene model, setting a space division function, carrying out space division, constructing a grade model and a depth region, generating a scene semantic segmentation map, and updating a visual scene.

The invention discloses a stainless steel surface defect detection method based on machine vision, and relates to the field of material defect detection.The stainless steel surface defect detection method comprises the steps that stainless steel surface images are collected, the collected multispectral surface images are subjected to image fusion, and a multispectral composite image is generated; performing detail enhancement on the low-resolution image through a super-resolution reconstruction algorithm to obtain a high-resolution image set; positioning a defect area position from the high-resolution image set through a defect detection algorithm, and obtaining a defect position data set; extracting a historical image defect feature set and a real-time image defect feature set from the defect position data set, and constructing a defect classification model according to the historical image defect feature set; inputting the real-time image defect feature set into a defect classification model, and performing defect classification and evaluation to obtain a defect classification data set; and generating and distributing a defect detection report. By combining the multispectral imaging algorithm, the super-resolution reconstruction algorithm and the defect classification model, the automation, precision and efficiency of stainless steel surface defect detection are improved.

The invention discloses an integrated power distribution cabinet energy efficiency dynamic optimization method based on multi-modal data fusion, and relates to the technical field of intelligent power grids. The problems of failure of an energy efficiency optimization model and poor long-term operation stability caused by data time sequence misalignment and error accumulation of a multi-source sensor in the prior art are solved. According to the scheme, the time offset is dynamically corrected through the adaptive time sequence deviation prediction model; compensating missing data by adopting a non-uniform time step reconstruction algorithm and Kalman filtering; detecting an error drift trend through an exponentially weighted moving average model, updating a feature weight, and inhibiting long-term error accumulation; constructing a self-adaptive time sequence attention fusion network model, and fusing physical constraints and a data driving mechanism to generate an optimization decision; bayesian optimization is utilized to quantify parameter uncertainty, closed-loop feedback execution data is carried out, and model parameters are dynamically updated; according to the invention, the precision, long-term stability and equipment safety of energy efficiency optimization of the power distribution cabinet are remarkably improved, and efficient and reliable operation under multi-physics field coupling constraint is ensured.

The invention provides a high-temperature chemical pipeline defect detection system and method based on magnetostrictive ultrasonic guided waves, and relates to the technical field of high-temperature chemical pipeline defect detection.The method comprises the steps that magnetostrictive excitation signals are applied, the propagation process of the guided waves in a pipeline is calculated through numerical simulation, and the amplitude and phase information of guided wave signals is obtained; performing time-frequency domain analysis on the guided wave signal, extracting group velocities and phase velocities of guided waves in different modes, determining frequency dispersion characteristics of the guided waves, and evaluating the influence of a high-temperature environment on guided wave propagation; processing the received guided wave signal by adopting a modal decomposition and reconstruction algorithm to obtain a corrected guided wave signal; and carrying out feature extraction and parameter estimation on the corrected guided wave signal to obtain the position, size and type information of the pipeline defect. According to the invention, the problem of multilayer heterogeneous pipeline guided wave detection in a high-temperature environment is solved, the accuracy and reliability of defect detection are improved, and an effective method is provided for pipeline health monitoring and safety assessment.

The invention relates to the technical field of hydraulic engineering safety monitoring, and particularly discloses a hydraulic engineering potential safety hazard assessment and prediction system and method based on image recognition. A multi-scale convolutional neural network is combined with a three-dimensional point cloud registration technology to extract surface visual feature parameters, and adaptive time-frequency analysis and a wavelet packet reconstruction algorithm are used to extract physical feature parameters of internal concealment defects; constructing a dual machine learning framework, eliminating environmental interference through a deep residual network, analyzing a causal relationship between features based on a gating cycle unit, and screening a key feature parameter set; a Gaussian process regression model of an adaptive kernel function is used for dynamic risk prediction, risk abrupt change points are identified in combination with multi-scale wavelet transform, and finally a safety state score and a grading early warning signal are generated through a fuzzy comprehensive evaluation algorithm.

The invention relates to the technical field of rock mechanics, and discloses a rock three-dimensional reconstruction method and system based on multi-source information fusion, and the method comprises the steps: obtaining and preprocessing data, carrying out the spatial feature learning of a fusion feature vector through a 3D-CNN network, and constructing a three-dimensional voxel model of rock microscopic damage; converting the fused image data into a point cloud model of the underground cavern surrounding rock structure by adopting a three-dimensional reconstruction algorithm based on point cloud, and constructing a digital twin framework of the underground cavern surrounding rock structure based on an implicit surface reconstruction algorithm; feature parameters output by the three-dimensional voxel model and the digital twinning framework are used as input, and the optimal supporting opportunity and supporting parameters are output through an LSTM-CNN fusion model; in the underground engineering construction process, surrounding rock deformation data are collected in real time, and a supporting scheme is adjusted in real time through a depth deterministic strategy gradient algorithm; according to the method, the scientificity and timeliness of support design under complex geological conditions can be improved.

The invention discloses an unmanned aerial vehicle-mechanical arm system cooperative control method for precise spraying. According to the method, through a depth camera and a point cloud reconstruction algorithm, three-dimensional modeling of a target surface and semantic recognition of a spraying area are achieved, and a spraying path and spraying posture data are generated. Based on spraying task requirements, a spraying film thickness experience estimation model is established, and path point-level spraying parameters are obtained. Load mass change, mechanical arm mass center change and spraying reaction force of each path point are calculated, and a coupling dynamic model is constructed in combination with the system state. And dynamic compensation control input is generated by using an adaptive sliding mode control algorithm. After spraying is completed, the actual spraying effect is collected through image and point cloud detection, color, film thickness and texture features are extracted and compared with task requirements, error feedback is generated, and control parameters are adjusted online. According to the method, the problems of dynamic change and error compensation in the spraying process are effectively solved, and the method is suitable for a high-precision spraying task in a complex environment.

A machine learning method of estimating State of Charge (SOC) and State of Health (SOH) of a battery pack, the method including collecting time-series data of a battery pack; processing the collected data utilizing a reconstruction algorithm to reconstruct a first phase state space reconstruction; training a first neural network model using the first phase state space reconstruction for predicting the SOH; feeding the time-series data to the first neural network to predict an estimated SOH value; processing the time-series data utilizing the reconstruction algorithm to reconstruct a second phase state space reconstruction; training a second neural network model using the second phase state space reconstruction, taking into account the estimated SOH value, for predicting the SOC, and feeding the time-series data and the estimated SOH value to the second neural network to obtain an estimated SOC value. A system configured to perform the above method is also disclosed.

The embodiment of the invention provides a scene surface reconstruction method and device, equipment and a medium, and relates to the technical field of computer graphics. The method comprises the following steps of: firstly, introducing an initialized three-dimensional Gaussian model, and determining attribute information and position information of two-dimensional Gaussian corresponding to three-dimensional anchor Gaussian by combining a multi-view image sequence of a target scene; secondly, utilizing the attribute information and the position information to generate a rendered image through a Gaussian splashing method; and then, according to the rendered image and an image with the same view angle as the rendered image in the multi-view image sequence, updating the initialized three-dimensional Gaussian model by minimizing a rendering loss function. And finally, depth information is determined according to the updated three-dimensional Gaussian model and the multi-view image sequence, and a surface model of the target scene is constructed through a three-dimensional reconstruction algorithm according to the depth information and the rendered image, so that the structure and details of the target scene can be reflected more accurately, the floating object phenomenon is avoided, and the accuracy of the target scene is improved. And the quality and the accuracy of scene reconstruction are obviously improved.

The invention provides a distribution network power transmission line insulator damage detection method based on a YOLO algorithm, and the method comprises the steps: collecting a plurality of insulator defect images, and carrying out the data enhancement of the images; performing image detail enhancement on an insulator defect area in the image by using a super-resolution reconstruction algorithm; constructing a composite loss function to guide an image detail enhancement process; constructing an insulator defect detection model based on a YOLOv8 framework; constructing a multi-task joint loss function as an insulator defect detection model training target; and a real-time feedback mechanism is introduced, a weighted combination loss function is constructed, gradient updating optimization is executed, and the detection capability of the insulator defect detection model is improved. The method has high detection precision and robustness, can effectively improve the efficiency and accuracy of fault detection, significantly reduces the burden and cost of manual inspection, improves the safety and reliability of a power system, and provides powerful support for the operation and maintenance of power equipment.

The invention provides a high-precision intelligent welding system adapting to complex working conditions, and relates to the technical field of laser welding. The high-precision intelligent welding system adapting to the complex working conditions comprises the steps of S1, based on a high-resolution visual sensor and a laser scanner, S2, based on preprocessed point cloud data, S3, based on a workpiece three-dimensional model, S4, based on an optimized welding path and a material attribute database, S5, based on a dynamic welding parameter set, S6, based on a real-time welding track, and S7, based on multi-sensor monitoring data. S8, based on a defect detection report; and S9, based on a repaired welding result and historical welding data. According to the invention, the geometric data of the workpiece are collected through the high-resolution visual sensor and the laser scanner, the filtering algorithm is applied to remove noise to generate high-quality point cloud data, and a subsequent processing foundation is laid; workpiece features are identified by using a point cloud processing algorithm and a machine learning method based on the point cloud data, and an accurate digital model is generated through a three-dimensional reconstruction algorithm.

The invention discloses a method for determining a preset position of a video monitoring camera. Camera presetting positions usually need to be manually set one by one, and time and labor are wasted. The method is used for a video monitoring system, and automatic setting of the preset position of the camera is realized by utilizing a group of scene images and the to-be-monitored target image, so that the efficiency is greatly improved. The method comprises the following steps: acquiring feature points based on a scene image set by adopting a computer vision technology, generating a three-dimensional model of a scene by adopting a motion recovery structure technology and a surface reconstruction algorithm, and obtaining the position of a to-be-monitored target in a three-dimensional space by matching the feature points of a to-be-monitored target image with the feature points of the scene image set. Therefore, a proper camera is selected, a corresponding preset position is calculated, and finally the preset position is corrected according to an actual monitoring effect.

The invention discloses a scene building method and system based on three-dimensional reconstruction. The method comprises the following steps: collecting a multi-view two-dimensional image of an environment; preprocessing the obtained two-dimensional image, and automatically identifying and labeling different objects and areas in the image by applying a semantic segmentation technology to understand semantic information in the image; according to the marked two-dimensional image data, reconstructing a three-dimensional model from the two-dimensional image by using a multi-view geometric three-dimensional reconstruction algorithm based on bundle adjustment, and storing the reconstructed three-dimensional model as a sub-scene model; according to the method, the three-dimensional model is accurately reconstructed by collecting the multi-view two-dimensional image and applying the semantic segmentation and bundle adjustment technology, the technical bottleneck of real-time rendering of a complex scene is solved, and the scene building efficiency and precision are improved; and in combination with the generative adversarial network and the variational auto-encoder optimization model, authenticity, detail expressive force and diversity are enhanced, and rich and real scene experience is provided for the field of virtual reality.

The invention provides a real-time map route construction method and system. The method comprises the following steps: installing various sensors on an unmanned aerial vehicle to obtain multi-modal data of an environment; converting the multi-modal features into a high-dimensional embedding space, and constructing a real-time three-dimensional environment model based on a multi-modal point cloud reconstruction algorithm; deploying edge computing nodes, and distributing computing tasks to the edge computing nodes; planning a cooperative path among the multiple unmanned aerial vehicles; constructing a multi-unmanned aerial vehicle cooperative global adaptive adjustment mechanism to dynamically adjust the flight path of the unmanned aerial vehicle; and performing priority ranking on the tasks of the flight path of the unmanned aerial vehicle, constructing a path adjustment optimization mechanism according to the real-time flight condition to minimize the task failure risk, and updating the task priority and the scheduling strategy according to the feedback after each task is executed. According to the invention, a plurality of key problems in the prior art are effectively solved, and a technical guarantee is provided for successful application of the unmanned aerial vehicle in more complex application scenes.

The invention provides a moving target capturing method and system based on AI multi-view vision, and relates to the technical field of data processing.The method comprises the steps that camera parameters in all vision systems are calibrated, and the camera parameters comprise a camera internal parameter matrix and a camera external parameter matrix; extracting 2D feature points of the moving target through each calibrated visual system; reconstructing each 2D feature point by combining a multi-view geometric reconstruction algorithm and a least square method according to the calibrated camera parameters to obtain 3D feature points; constraints are added to the 3D feature points, and the constraints comprise spatial constraints, direction vector constraints and time sequence constraints; reconstructing the 3D feature points added with the constraints through a multi-view geometric analysis algorithm to obtain three-dimensional motion information of the moving target; and outputting the three-dimensional motion information of the moving target to complete the capture of the moving target. And non-contact high-precision capture of the moving target can be realized.

The invention discloses a storage intelligent monitoring and checking integrated system and method based on multi-technology fusion, and particularly relates to the technical field of storage monitoring, the system obtains environmental parameters and cargo data in real time through a multi-source data acquisition module, carries out cargo identification and segmentation by using YOLOv7 and Mask R-CNN, and carries out storage monitoring and checking on the cargo. A three-dimensional model is constructed in combination with a Poisson surface reconstruction algorithm to calculate volume weight, an environment anomaly analysis module generates an early warning signal through dynamic threshold and influence level evaluation, and an intelligent decision module fuses data to realize quantity anomaly alarm and multi-level response and outputs a three-dimensional visual inventory report. According to the method, the problems of multiple blind areas, low checking efficiency, slow abnormal response and the like of traditional storage monitoring are solved, full coverage of environment monitoring, high accuracy of goods identification, low error of volume calculation and intelligence of abnormal handling are realized, and the automation and digitization levels of storage management are remarkably improved.

The invention discloses a personalized design intelligent interaction method based on topological optimization, and the method comprises the steps: S1, constructing a user behavior monitoring system based on multi-modal data collection, and generating a high-dimensional data set containing time sequence features; s2, based on an adaptive deep neural network model, performing dynamic prediction and hierarchical modeling on the personalized demand of the user; s3, generating an optimal solution of the design parameters in real time through a dynamic topology reconstruction algorithm; s4, iteratively updating a design scheme generation rule through a reinforcement learning mechanism; s5, constructing a multi-objective design space exploration and optimization framework supported by the neural network, and rapidly screening and optimizing design parameters; s6, dynamic updating and local optimization of the design scheme are achieved through multi-dimensional data mapping and parameterization regulation and control means; and S7, establishing a data-driven continuous learning and evolution mechanism, continuously optimizing the user demand prediction model and designing an optimization algorithm. The method has the advantages of being high in dynamic adaptability, high in intelligent level and high in personalized meeting precision.

The invention relates to the technical field of three-dimensional imaging, in particular to a four-eye structured light stereoscopic vision imaging method, which comprises the following steps of: arranging four cameras and synchronously acquiring multi-view image data with structured light stripes; carrying out image preprocessing and stripe code identification on the obtained four-view-angle structured light image data, and extracting structured light stripe center line positions and corresponding space projection information under each view angle; generating three-dimensional point cloud data of the high-precision target sample by using parallax calculation and a three-dimensional reconstruction algorithm, and completing spatial registration and filtering optimization of point cloud; fusing the point cloud and the light intensity data by combining the reflection intensity information of the multi-view structured light stripes to generate a composite data set containing space and spectral information; and constructing a dense parallax field and executing three-dimensional consistency verification, and generating a high-fidelity three-dimensional reconstruction model with a complete topological relation and sparse shielding compensation capability. According to the invention, the problems of insufficient view angle coverage, shielding area information loss, difficult edge structure matching and the like of traditional stereoscopic vision imaging can be solved.

The invention discloses a cross-platform virtual-real fusion scene construction method and system based on AI space calculation, and relates to the technical field of artificial intelligence and space calculation, and the method comprises the steps: carrying out the multi-scale feature fusion based on a received cross-modal conversion instruction set, and generating an initial image sequence; carrying out implicit field coding on a target object by combining a three-dimensional reconstruction algorithm to obtain an initial parameterized model; performing space-time alignment on the multi-view video stream, loading a digital scene asset package in combination with physical sensing data and a preset spatial index structure, and establishing a bidirectional data channel between a virtual scene and a physical sensor; performing rendering and illumination parameter adjustment on the initial parameterized model to obtain an optimized parameter model; performing differential coding processing on the optimization parameter model to obtain a target virtual-real scene fusion model; and distributing the target virtual-real scene fusion model to a preset terminal. The invention provides a virtual-real fusion construction method for end-to-end collaborative optimization, which is suitable for cross-platform live broadcast or dynamic interaction scenes.

The invention relates to the field of three-dimensional reconstruction, and discloses a three-dimensional scene online reconstruction method, device and equipment and a storage medium, and the method is used for realizing the online reconstruction of a three-dimensional scene. The method comprises the following steps: acquiring and processing multiple frames of ordered scene RGB images to obtain sparse point clouds of each frame of scene RGB image under a camera coordinate system, and converting the sparse point clouds into sparse point clouds under a world coordinate system; initializing a sparse point cloud under a world coordinate system to obtain a local Gaussian surface element of each frame of scene RGB image; based on a fusion mechanism of a gating loop unit, reading global features of sparse point clouds of each frame of scene RGB image in a world coordinate system frame by frame from the first frame, and fusing the global features of the scene RGB image of the previous frame with the local Gaussian surface element of the scene RGB image of the current frame to generate a global Gaussian surface element of the scene RGB image of the current frame; and rendering the global Gaussian surface element of each frame of scene RGB image from a plurality of visual angles, and generating a scene reconstruction surface by using a Poisson reconstruction algorithm.

The invention discloses a multi-source data fusion accounting method for a full-life-cycle carbon footprint, and belongs to the technical field of carbon footprint accounting, and the method specifically comprises the steps: inputting a purchase order number set of a target product, and automatically associating a logistics fuel consumption record, a production process energy consumption log and a supplier indirect emission declaration form according to the purchase order number; constructing a dynamic traceability tree taking a purchase order as a root node, and generating a trunk path through matching of an order material code and a workpiece batch number of a process energy consumption log; when the difference between the logistics real-time energy consumption data and the supplier declaration form exceeds a set threshold value, activating a dispute data branch reconstruction algorithm, and creating parallel branches to temporarily replace dispute data segments while keeping trunk continuity; finally, outputting a full-life-cycle carbon footprint data graph with dispute marks, and reserving original association paths of all data sources; according to the invention, reliable support is provided for accurately tracing the full-life-cycle carbon emission of the product and improving the carbon footprint accounting quality.

The invention belongs to the technical field of medical imaging and digital twinning, and discloses a thoracic cavity whole organ system level lung four-dimensional EIT detection digital motif and a library building method, and the method comprises the following steps: S1) segmenting multiple organs such as double lungs, heart, sternum, rib and thoracic vertebra central axis skeleton, main bronchus and esophagus based on an individual CT image, and building a three-dimensional geometric model integrated with an EIT electrode array; s2) simulating breathing dynamic deformation by using a solid mechanical model; s3) establishing a coupling relationship between deformation and conductivity, and generating dynamic three-dimensional conductivity distribution; s4) solving an electromagnetic field positive problem, and generating a virtual EIT voltage database synchronous with respiration; the invention also covers pathological model integration and data enhancement strategies. According to the method, a high-fidelity, large-scale and diversified EIT virtual data set can be generated, core data support is provided for training and verifying an AI intelligent reconstruction algorithm, and a model foundation is laid for developing a lung function dynamic digital twinning system.

The invention discloses a dynamic scene repairing method based on four-dimensional Gaussian sputtering, which is characterized in that efficient dynamic scene space repairing is realized based on a repairing framework of a four-dimensional Gaussian sputtering monocular video reconstruction algorithm and a video repairing model, and by counting the time frame number of Gaussian primitives in a mask area and removing Gaussian points exceeding a threshold value, the dynamic scene space repairing efficiency is improved. According to the method, high-efficiency dynamic object removal is realized, and a mask weighting loss function is introduced to guide the repair area to converge to real texture information in a subsequent time frame, so that the repair quality is improved. The invention provides a dynamic scene repair method based on four-dimensional Gaussian sputtering, and provides a repair framework based on a four-dimensional Gaussian sputtering monocular video reconstruction algorithm and a video repair model for removing and repairing a target object in a dynamic scene to realize three-dimensional repair of the dynamic scene.

The invention provides a nuclear power safety report data extraction method and system based on multi-modal feature fusion, and belongs to the technical field of data processing, and the method comprises the steps: obtaining a nuclear power safety report document; carrying out OCR (Optical Character Recognition) and paragraph segmentation on the report document to obtain first text data; performing text semantic understanding on the first text data by adopting a dynamic template matching and semantic driving extraction mechanism to obtain a text feature vector; segmenting pages of the report document and identifying fonts and fonts to generate first format data; adopting a cross-page table reconstruction algorithm to detect table cells of the first format data and splicing a cross-page table to obtain format feature vectors; performing feature alignment on the text feature vector and the format feature vector, and inputting the text feature vector and the format feature vector into a large pre-training language model to obtain a semantic vector; and fusing the text feature vector, the format feature vector and the semantic vector to form a composite expression unit, and intelligently extracting structured data based on the composite expression unit. The semantic entity and structural relationship in the nuclear power report can be effectively identified.

The invention discloses a naked-eye 3D display system based on a light field reconstruction algorithm, particularly relates to the technical field of naked-eye 3D, and comprises a controlled scattering medium light field acquisition environment, scattering feature enhanced light field data analysis, a hierarchical depth optimization engine and dynamic wavefront modulation display control. According to the invention, through the controlled scattering medium environment design, a scattering signal which is regarded as interference in the traditional technology is converted into a depth characteristic carrier, the data acquisition quality in a complex optical environment is obviously improved, richer and more reliable basic data are provided for light field reconstruction, and the data acquisition efficiency is improved. An innovative hierarchical deep optimization engine is combined with a multi-scale confidence evaluation and discontinuous region processing method, so that the problem of edge blur in a shielding boundary and a geometric structure complex scene is effectively solved, and the geometric consistency and detail fidelity of a reconstructed light field are greatly improved.

The invention discloses a fire-fighting non-destructive detection method based on multi-modal detection, and relates to the technical field of fire-fighting engineering, and the method comprises the following steps: carrying out the rapid scanning of fire-fighting equipment through an ultrasonic array, analyzing and positioning a suspected defect region based on a time domain reflection signal, and generating a three-dimensional coordinate mapping model; and performing multi-angle X-ray projection acquisition on the positioning area, and generating an internal structure cross-sectional image through a three-dimensional reconstruction algorithm. According to the fire-fighting non-destructive detection method provided by the invention, by integrating various technical means such as ultrasonic array scanning, X-ray tomography, thermal imaging analysis and microwave sealing detection, comprehensive and accurate identification of potential defects of fire-fighting equipment is realized, and the dimensionality and precision of defect detection are improved; the structural integrity of the equipment is kept to the maximum extent, extra risks caused by destructive detection are avoided, and the reliability and accuracy of detection are improved through a data fusion algorithm.

The invention discloses a self-adaptive grid dynamic encryption method for underground water pollution migration simulation, which belongs to the technical field of underground water pollution migration simulation, and comprises the following steps of: forming 0.05-meter super-resolution grids at the periphery of a dispersed pollution source through a local grid dynamic encryption technology triggered by a pollution source intensity gradient threshold; the pollutant frontal surface tracking error is greatly reduced, and the high-precision pollutant migration simulation capability is realized. Moreover, based on an intelligent grid reconstruction algorithm of a flux threshold, the number of calculation units is reduced by 62%, the time consumed by single simulation is shortened to 3.3 hours, and the overall mass conservation error is kept to be smaller than 1% while the calculation efficiency is improved. Meanwhile, for a heterogeneous aquifer, through a pollution flux and stratum permeability coefficient dual-drive encryption strategy, the solute transport prediction precision of the model at the boundary of a clay lenticular body is greatly improved, and the problems that in the prior art, the simulation precision of the periphery of a pollution source is insufficient, and the calculation efficiency is low due to global encryption are effectively solved.

The invention relates to an accelerator and an acceleration method based on Top-K sparse moment vector multiplication. The accelerator comprises a first acceleration unit and a second acceleration unit. The first acceleration unit is used for data preprocessing and sparse matrix coding, and the second acceleration unit is used for executing Top-K sparse matrix vector multiplication so as to solve the process of the first K results with the maximum values in the operation results of multiplication of a sparse matrix and a dense vector; wherein the first acceleration unit comprises a preprocessing module and a coding module; the data preprocessing step of the preprocessing module comprises quantization, sparse reconstruction and shuffling; the broadband demand of data transmission is reduced based on the quantization step of the sparse matrix; pruning non-zero elements which have small influence on sorting based on a sparse reconstruction algorithm; carrying out shuffling operation on the sparse matrix after sparse reconstruction so as to avoid a Top-K aggregation phenomenon; and the coding module is used for recoding the preprocessed sparse matrix. The method has the remarkable effects that the end-to-end performance is improved by 153.3 times, 2.5 times and 3.4 times respectively.

The invention provides an automatic assembling method and system for an inductor magnetic core. Wherein continuous images of the surface of the magnetic core are captured through multi-band rapid scanning, and geometric deformation characteristics are extracted; meanwhile, a distributed optical fiber sensing array is adopted to collect pressure data of the contact surface, and a real-time distribution diagram is generated through a three-dimensional field reconstruction algorithm. And then carrying out space coordinate alignment on the edge contour offset of geometric deformation and the pressure gradient, and constructing a comprehensive quality index. And based on the dynamic deviation of the index, a contact angle compensation value and an assembly speed optimization value are solved in real time, and a parameter adjustment instruction is generated. And finally, through parallel calculation of instruction data, a pressure tolerance threshold value and a deformation compensation parameter are synchronously updated, dynamic balance of deformation correction and pressure control in the assembly process is realized, and a self-adaptive closed-loop precision control system is formed. According to the technical scheme, through multi-dimensional sensing data fusion and real-time parameter adjustment, dynamic matching of deformation and stress states in the magnetic core assembling process is achieved.

The invention provides an automatic processing method for real-time observation data of an ocean station, and belongs to the technical field of ocean observation data processing. A wavelet packet decomposition multi-scale noise separation algorithm is established to distinguish environmental noise and real signals, dual-sensor redundancy configuration is combined with Bayesian inference to identify sensor drift, and a calibration coefficient is updated in real time through a recursive least square method. A one-dimensional time sequence is mapped to a high-dimensional phase space by utilizing a phase space reconstruction algorithm to realize high-precision prediction of a chaotic signal, a hierarchical data storage architecture is established, and a data migration strategy is iteratively optimized through a hierarchical correlation degree function; the technical problem that time synchronization signals are difficult to reconstruct accurately according to asynchronous sampling data of multiple sensors of an ocean station is solved.