78results about How to "Achieve alignment" patented technology

This application discloses a method and apparatus for constructing a multi-scale adaptive gating MambaPlus network, belonging to the field of artificial intelligence and machine learning technology. The method includes: initializing the network configuration and constructing the basic structure; preprocessing the input data to generate a standard dataset; mapping the input data to the hidden space via an input mapping layer, and extracting backbone features from the Mamba backbone; constructing at least two parallel scale branches in the hidden space to obtain a multi-scale feature set; inputting the backbone features and multi-scale features into an adaptive gating module, dynamically allocating weights and adaptively fusing them through a hierarchical gating mechanism to generate fused features; further enhancing the features through cross-scale attention and feedforward enhancement, and then superimposing the residuals to generate the final discriminative features; finally, completing category prediction and model training evaluation. This application, while retaining the advantages of Mamba's long-range dependency modeling, addresses the problems of insufficient utilization of multi-scale information, poor adaptive feature fusion, and low robustness in complex scenarios.

The invention relates to the technical field of membrane attachment, and provides a ceramic membrane coating, bonding and attaching device which comprises a basic rack, a first driving part, a pressing assembly, a second driving part, a vacuum adsorption assembly and a visual assembly, a top rack is fixedly installed on the upper portion of the basic rack through a supporting rack, and the first driving part is installed on the basic rack; a bearing frame is fixedly installed at the output end of the first driving part, a lower press-fit mold is arranged at the top of the bearing frame, a press-fit assembly is installed on the lower press-fit mold, the second driving part is installed on the top rack, a lifting frame is fixedly installed at the output end of the second driving part, and an upper press-fit mold is arranged at the bottom of the lifting frame. A vacuum adsorption assembly is installed at the bottom of the upper pressing mold, and a visual assembly is installed on the top rack and used for solving the technical problems that in the related technology, when multiple layers of diaphragms are pressed, local stress concentration can be formed on the surfaces of the ceramic diaphragms through adsorption holes, then indentations are generated, and the surface quality of finished products is reduced.

This invention provides an optoelectronic co-packaging structure and its packaging method. Utilizing the excellent optical transparency, thermal stability, and electrical insulation of a glass substrate, it combines through-hole electrical interconnects for electrical signal transmission and through-hole optical interconnects for optical signal transmission. Furthermore, it integrates optical components through dual-surface grooves, solving the problems of low optical interconnect efficiency, poor heat dissipation, and limited integration in traditional optoelectronic co-packaging. This significantly improves the optical interconnect efficiency, heat dissipation performance, and integration of the optoelectronic co-packaging structure. The invention also uses grooves to limit the electrical components, achieving precise alignment and transmission of optical signals and reducing optical coupling loss. Simultaneously, the overlapping of the orthographic projections of the two grooves causes partial vertical overlap of the two optical components, optimizing the optical path and thus greatly enhancing signal fidelity.

The utility model discloses an auxiliary welding device for a battery bracket of a new energy electric bus. The auxiliary welding device aims at solving the problems that traditional welding positioning is difficult, clamping is not stable, and universality is poor. The device comprises a bottom, a bottom plate groove is formed in the bottom, and a threaded rod and a guide rod in the groove are connected with a rotating motor to drive a movable vertical plate and a fixed vertical plate to adjust the distance. Aligning groove holes are formed in the opposite faces of the two vertical grooves, a limiting clamp assembly is arranged above to clamp a workpiece, and an adjusting assembly is installed in the middle vertical groove. The limiting clamp assembly controls a clamp lead screw through a limiting motor, and precise clamping is achieved. And the adjusting assembly integrates various transmission structures, so that the movable plate can be flexibly adjusted in a three-dimensional space. An arc-shaped shallow groove in the movable plate assists in positioning, a laser transmitter and a laser receiver on the lower surface are arranged in a crossed mode, and position deviation is monitored in real time and calibrated. The device is accurate in positioning, stable in clamping and high in universality, the welding quality and production efficiency of the battery bracket are effectively improved, and development of the new energy electric bus manufacturing industry is promoted.

This application discloses a method and apparatus for intelligent driving data trajectory repair based on segmented point cloud registration, relating to the field of autonomous driving technology. The method includes the following steps: dividing the data segment to be repaired into multiple smaller data segments; registering each frame of point cloud data in a smaller data segment with the first frame of point cloud data in that smaller data segment to obtain a data segment repaired within the segment; and registering the point cloud data of adjacent data segments repaired within the segment on the time axis to obtain point cloud data repaired between segments. Through partial-to-global joint trajectory correction, data repair can be achieved without relying on additional hardware upgrades or external auxiliary data. This significantly improves global trajectory consistency and repair success rate, efficiently repairing trajectory deviations in existing data, avoiding data resource waste, and adapting to the repair needs of newly added data, thus greatly reducing data processing costs.

This invention provides a drug interaction prediction method based on a bidirectional cross-view attention network, belonging to the field of drug interaction prediction technology. The method first constructs a Morgan fingerprint similarity view of the drug, an original DDI view, and a multi-scale diffusion view based on personalized PageRank. Then, a graph convolutional network with shared weights is used to co-encode the multiple views, generating a unified drug embedding representation. Finally, a bidirectional cross-view attention mechanism is used to achieve fine-grained interaction and alignment between the structural and attribute views, and interaction prediction is completed via a multilayer perceptron. This invention effectively alleviates the sparsity problem of the DDI network through multi-scale topology enhancement and achieves complementary enhancement between views using bidirectional cross-view attention, significantly improving the accuracy and generalization ability of drug interaction prediction. It can provide reliable technical support for drug development screening and clinical combined drug safety assessment.

The application discloses a kind of based on industrial internet of things's time sequence operation characteristic data cleaning method and system, it is related to industrial internet of things technical field, method includes: obtaining the distortion snapshot of target data node and parsing dimension declaration and clock stamp, calculate isolated drift degree accordingly;Coupling factor is calculated in combination with network atlas and clock stamp, and the collaborative confidence of isolated drift degree is generated by optimization, and the first cleaning strategy is formulated by dynamically adjusting Kalman filter parameter;Determine the conduction boundary based on evolutionary model, identify and isolate the infected data cluster in network atlas, and formulate the second cleaning strategy containing time freezing mechanism according to its logical time sequence position;First, the second cleaning strategy is fused based on decision matrix, and the comprehensive cleaning result is output.This application has the advantages of accurate time sequence alignment, collaborative adaptive noise reduction and error foresight blocking.

This invention relates to an auxiliary drilling and production device for oilfields, belonging to the technical field of oilfield drilling equipment. It includes: a rotary table module equipped with several sets of casing clamping modules; a casing bin on one side of the rotary table module, containing several casing conveying modules into which casing can be clamped. A clamping tower on one side of the rotary table module connects to and clamps casing from the clamping modules. The clamping modules can extend into the casing bin and clamp the casing clamped on the casing conveying modules. The clamping modules retrieve casing from the casing bin and transport it to the clamping tower for precise clamping and positioning, achieving alignment between the casing and the wellhead, improving accuracy and reducing labor intensity. The casing bin pre-stores multiple casings, which are supplied in an orderly manner by the casing conveying modules. The rotary table module is equipped with multiple clamping modules, enabling continuous gripping and transport of casing for installation. Casing supply and installation preparation can be carried out in parallel, ensuring continuity and thus improving installation efficiency.

The invention relates to the field of data transmission, in particular to an internet of things data adaptive compression method and system based on edge computing, and the method comprises the steps: obtaining a historical data set and a to-be-compressed data set, calculating the correlation degree of a target dimension and a contrast dimension, segmenting the to-be-compressed data set, and clustering all segments; calculating the data anomaly probability of the target dimension in the target cluster; calculating a data adaptive compression deviation value of the target dimension in the target segment; calculating an upper door slope and a lower door slope of a target point in a target dimension based on a revolving door algorithm, calculating an upper door effective slope of the target point in the target dimension, similarly obtaining a lower door effective slope, obtaining a compression starting point of the target dimension based on the upper door effective slope and the lower door effective slope, and similarly obtaining a compression ending point; and compressing the data of the target dimension. According to the technical scheme, compression resource consumption can be reduced, and compression efficiency can be improved.

The invention discloses a calligraphy stroke feature extraction method and system based on a fixed-size grid. The method relates to the technical field of stroke feature extraction, and comprises the following steps of calligraphy stroke image preprocessing and grid division, grid-based stroke space distribution feature extraction, size feature extraction, grid-based stroke direction feature extraction and feature fusion and output. The method comprises the following steps: S1, normalizing a binary image of calligraphy strokes to a fixed canvas and dividing the binary image into grids; s2, calculating the stroke density of each grid and constructing a density feature vector; S3, refining the image to obtain a skeleton, calculating a direction field and generating a direction feature vector; and finally, the two vectors are connected in series and fused, and a joint feature vector is output for a subsequent task, so that the calligraphy stroke feature extraction accuracy is improved, and the problem of low calligraphy stroke feature extraction accuracy caused by insufficient understanding of dynamic essence, complex form and high-level semantics of calligraphy art in the prior art is solved.

The application discloses a kind of image accurate alignment method, device, equipment and readable storage medium, it can be applied to semiconductor technical field, the method comprises: obtaining the real shot optical image of wafer to be aligned;Determine the target Kernel image corresponding to real shot optical image;Determine the target alignment template graph corresponding to target Kernel image based on preset alignment database;Pre-set alignment database is based on design layout construction;Target alignment template graph is one of all Target alignment graph in pre-set alignment database;Determine the translation deviation value corresponding to target Kernel image based on target alignment template graph;Compensate all pixels in target Kernel image based on translation deviation value, realize image alignment.So, by introducing pre-set alignment database, design layout and image alignment database are directly associated, improve the accuracy between real shot image and design layout alignment.

The utility model relates to solar cell panel technical field especially relates to a solar cell panel cell packaging structure. Its technical scheme includes: the outside of cell panel is equipped with frame, the bottom of frame is equipped with bottom plate, the inside of frame is equipped with accommodating groove, the corner of accommodating groove is equipped with cushion block, installs the push board in accommodating groove, installs spring and threaded column on the push board, the both ends of bottom plate are through and are equipped with the jack, the both ends of frame are equipped with the plug block, and the bottom plate passes through plug block and jack and is mutually inserted with frame. The utility model relies on the mutual action between its structure design and component, realizes the effective packaging and protection to solar cell panel through the design of fixed and supported, installed and connected, double locking mechanism, maintenance and replacement and overall stability etc.

The invention discloses a link training method and system, electronic equipment and a storage medium, and relates to the field of data processing. The method comprises the following steps: acquiring a plurality of stable intervals of a differential link based on a fixed delay step length and a training code; calculating the ratio of each stable interval to the unit interval of the differential link; selecting a stable interval of which the ratio is an odd number after rounding up as an effective interval; and delaying the clock signal or the differential signal based on the delay amount corresponding to the central point of any effective interval so as to realize data alignment. On the basis of the processing, the effective interval with the ratio being an odd number after rounding up is selected, the stable interval with the center point not being the jump area is screened out, and therefore in the subsequent data alignment process according to the center point of the effective interval, the error that the collection edge of the clock signal is located in the jump area can be avoided.

The utility model discloses a waterproof automobile wire harness connector relates to the technical field of wire harness connector, including first plug connector, one side back top of first plug connector is fixedly connected with first power connection disc, one side top of first power connection disc is fixedly connected with first connection wire harness. Through setting the plug between first plug connector and second plug connector, using the inside of buckle slot of connecting baseplate buckle directly, sliding limit post of this time will automatically align the inside of limit slot, realize the buckle locking operation between first plug connector and second plug connector, the power plug of this time inserts into the inside of power connection groove directly, completes the connection transmission operation of electric power, the back of first plug connector sets up first power connection disc and first connection wire harness, sets up second power connection disc and second connection wire harness on the back top of second plug connector, realizes the connection of both sides wire harness.

The application discloses a kind of quality intelligent evaluation methods and systems for multi-source medical data, it is related to big data processing technical field.The quality intelligent evaluation methods and systems for multi-source medical data, including S1, acquisition multi-source medical data is preprocessed, store and construct medical database;S2, drive the spatiotemporal semantic alignment and joint representation learning of cross-modal data, construct medical data optimization model and output spatiotemporal alignment value;S3, fusion spatiotemporal correlation and multidimensional quality index, data quality evaluation and abnormal analysis are carried out;S4, according to data credible result executes the traceability path generation and explainability report output operation;S5, through user feedback data and model evolution state analysis optimization effect.It solves the problem that multi-source medical data exists in spatiotemporal alignment, quality evaluation and abnormal detection, spatiotemporal mismatch, data inconsistency and low quality repair efficiency.

The present disclosure relates to the technical field of audio recognition, in particular, to a subtitle alignment method and device for audio, a medium and an electronic device. The method comprises: obtaining a target audio and a target subtitle text of the target audio; if a time length of the target audio is greater than a first preset time length, performing slicing processing on the target audio according to a slice time length to obtain a plurality of first target audios; determining first audio feature information of each first target audio; if the time length of the target audio is less than or equal to a second preset time length, splicing all the first audio feature information to obtain target audio feature information of the target audio, wherein the second preset time length is greater than the first preset time length; and generating subtitle information corresponding to the target audio according to the target subtitle text and the target audio feature information. In this way, the occupation of excessive machine resources can be avoided, the matching of the target subtitle text and the target audio feature information is realized through one alignment, and the accuracy of the alignment result is improved.

The utility model discloses a kind of adsorption tower inner grating frame, it is related to adsorption tower internal structure field.The device.The device includes support block, magnetic guide pin and conical guide sleeve;Support block is set in the inside of frame, and support block is annular array distribution;Magnetic guide pin is fixedly assembled at the bottom of grating, and magnetic guide pin is annular array distribution;Conical guide sleeve is opened in the inside of support block, and magnetic guide pin is movably connected in the inside of conical guide sleeve, and magnetic guide pin is magnetically connected with conical guide sleeve, the device is magnetically guided unit by being set at the connecting place of frame and grating, using the magnetic attraction between magnetic guide pin and conical guide sleeve and the guiding effect of conical guide sleeve, when installing grating, magnetic guide pin will move to conical guide sleeve under the action of magnetic force, and gradually slide into its inside along the inner wall of conical guide sleeve, realize automatic guidance and alignment, reduce the number of adjustment, shorten installation time, improve installation efficiency.

The embodiment of the invention discloses a hyperspectral image processing method and device, electronic equipment and a storage medium. The method comprises the following steps: obtaining crop images of a target crop, wherein the crop images comprise a first color image and a first depth image collected by a color depth camera and a first hyperspectral image collected by a hyperspectral camera; obtaining target joint calibration parameters of the color depth camera and the hyperspectral camera; and performing three-dimensional point cloud reconstruction on the first color image and the first depth image by using the target joint calibration parameter, and determining depth information corresponding to the first hyperspectral image based on first point cloud data obtained by reconstruction, the depth information being used for representing a first distance between the hyperspectral camera and the target crop when the hyperspectral camera collects the first hyperspectral image. According to the scheme, information of the color depth camera and information of the hyperspectral camera can be fused, and the first distance between the hyperspectral camera and the target crop is accurately determined when the hyperspectral camera collects the first hyperspectral image.

The application discloses a high-end caliper stainless steel surface composite strengthening system and process, which comprises a quenching device, a tempering device, a rolling device, a high-current pulse electron beam irradiation device and an ultrasonic cleaning device. The process specifically comprises the following steps: S1, quenching treatment; S2, tempering treatment; S3, rolling treatment; S4, high-current pulse electron beam irradiation treatment; and S5, cleaning treatment. The system can perform three-blade rolling treatment on the heat-treated caliper through the rolling device, can homogenize the surface layer structure of the caliper while improving the machining efficiency, can reduce the surface roughness of the caliper, can effectively realize the overall straightening of the caliper, and can reduce the quenching deformation. After the surface modification of the caliper through the high-current pulse electron beam irradiation device, the caliper can form a surface layer with uniform composition and nanometer-sized grain refinement, can not have obvious interface between the caliper matrix and the nanometer structure surface layer, and can not easily fall off, so that the service life of the caliper is greatly prolonged.

This utility model provides a steel strip butt welding machine, relating to the technical field of steel strip butt welding machines. The machine includes a butt welding machine with a welding head and a switch on its side. A fixing clamp is also provided on the side of the machine, positioned on both sides of the welding head. A locking element is rotatably connected to the bottom of the fixing clamp. A motor is fixedly connected inside the machine. A grinding wheel, driven by the motor, is also provided on the side of the machine. A pressing device and an alignment device are also provided on the side of the machine. The alignment device includes a rectangular frame fixedly connected to the side of the machine, with a threaded rod rotatably connected inside the rectangular frame. A slider is threadedly connected to the surface of the threaded rod, and a rectangular strip is fixedly connected to the side of the slider. This utility model solves the problem of steel strip misalignment during the butt welding process due to operator hand-held operation.

The application provides an infrared and visible light image fusion method and system based on bidirectional semantic-space alignment and differential perception, and relates to the technical field of image processing. In order to solve the problem that the expression capacity is insufficient when the existing method processes complex cross-modal semantic interaction. The technical points of the application include: S1, collecting infrared images and visible light images; S2, fusing infrared and visible light images based on a semantic fusion network, wherein the semantic fusion network is based on a PSFusion network model, the SDFM module of the PSFusion network model is replaced by a differential perception complementary module DACM, the PSFM module is replaced by a deep semantic fusion block DSFB, and a space semantic alignment module SSAM is added after the deep semantic fusion block and the differential perception complementary module; S3, image reconstruction and supervision are performed on the aligned features. The method provided by the application shows superior potential compared with the current advanced image fusion algorithm.

The utility model relates to a kind of winding structure for stone paper casting machine, a kind of winding structure for stone paper casting machine, including the guide rail of left-right symmetry arrangement, slidingly equipped with base plate between guide rail, the position of being located above base plate is equipped with top plate, top plate top is equipped with winding groove, base plate top right side is equipped with first slide rail symmetrically, first slide rail is slidably connected with first slide roll between, top plate bottom right side is equipped with second slide rail symmetrically, second slide rail is slidably connected with second slide roll between.The utility model drives "X type" connecting rod mechanism by servo motor, realizes the accurate vertical lifting of top plate and winding groove, effectively adapts to the winding demand of different roll diameter, avoids the contact pressure change problem caused by roll diameter change;Winding groove in the utility model adopts large-angle V type structure, can automatically guide diaphragm to center alignment, significantly reduce the deviation of diaphragm in lifting and introduction process, improve winding neatness and finished product quality.

The present application belongs to the field of communication tower installation, and particularly relates to a single-pipe tower installation robot and an installation method. The single-pipe tower installation robot is characterized by the following: a platform is connected to an installation mechanism at the top, the installation mechanism is used for the installation of a single-pipe tower, a plurality of climbing wheel systems are connected to the platform at the bottom, the climbing wheel system comprises a floating guide rail, a magnetic wheel, a supporting wheel, a wheel leg and a driving structure, the sliding block assembly of the floating guide rail is fixedly connected to the platform, the end of the guide rail body of the floating guide rail is fixedly connected to the wheel leg, the wheel leg is connected to the magnetic wheel and the supporting wheel, and the magnetic wheel is connected to the output end of the driving structure; the magnetic wheel and the supporting wheel of the plurality of climbing wheel systems are used for acting on the tower body to climb on the tower body. The present application realizes self-adaptation through the floating guide rail, solves the problem of complex self-adaptive component structure and high production cost in the prior art, and the magnetic wheel is arranged to be capable of adsorbing the tower body, so that the robot can be more stably attached to the tower body, and the safety of the robot in use is improved.

The invention discloses a multi-modal wireless basic model construction and application method and system based on a double-encoder cross-modal injection mechanism, a terminal and a medium, and the method comprises the steps: constructing a double-encoder structure which comprises a CSI encoder and a point cloud encoder; obtaining multi-modal data, and carrying out data blocking and position coding; extracting a first feature through a CSI encoder and a point cloud encoder, extracting a second feature through an auxiliary encoder, performing cross injection fusion on features of different modes through a cross-mode attention mechanism, and outputting a multi-mode universal representation; pre-training is carried out by adopting a joint training mode to obtain a multi-modal wireless basic model, and in a reasoning stage, general representation is selected according to the type of a downstream task, and a downstream task model is accessed to carry out reasoning. The multi-modal wireless basic model obtained by training based on the method can improve various wide communication and perception downstream tasks such as channel estimation, channel prediction, target positioning, electromagnetic map, indoor perception and the like.

The invention discloses a vehicle-mounted wireless charger and a power distribution system thereof, and belongs to the technical field of vehicle-mounted electronic equipment. The vehicle-mounted wireless charger comprises a multi-degree-of-freedom support body and a charging module, the support body has the functions of horizontal rotation, pitching swing and end rotation, and the charging posture can be flexibly adjusted; the charging module is provided with an automatic clamping jaw mechanism and an arc-shaped electromagnet array arranged in a surrounding mode, and the charging module is combined with a wireless charging coil and used for supporting, clamping and aligning equipment. The power distribution system detects the attitude of the bracket in real time through an angle sensor, and adaptively adjusts the reference magnetic force of the electromagnet based on an attitude compensation formula; furthermore, the deviation of the charging coil of the equipment is detected through a high-precision position sensor, the magnetic force distribution of each electromagnet is adjusted, and the automatic alignment of the submillimeter-level coil is realized.

The application provides an infrared and visible light image fusion method based on text semantic consistency guidance, relates to the technical field of multi-modal image fusion, and comprises the following steps: fine-grained text semantics of infrared and visible light images are generated respectively and mapped to a unified embedding space; a cross-modal attention mechanism is used to bidirectionally compensate and enhance the text semantics, and a unified text semantic prior is constructed; a structure-intensity decoupled double-branch encoder is used to extract structure texture features of visible light and intensity saliency features of infrared respectively; under the guidance of the text semantic prior, explicit semantic consistency constraints and implicit semantic distribution consistency constraints are used to align the double-modal visual features in a shared semantic space; and finally, the text semantic prior is used as a global modulation signal to adaptively weight and fuse the aligned features and decode to generate a fusion image. The application effectively solves the problems of insufficient semantic modeling and poor fusion result consistency of existing methods.

The utility model relates to detection equipment field discloses a kind of general assembly disc detection machine, including rack, index turntable, feeding device, visual detection device and discharging device;Index turntable is located on rack, and there are several work stations for transverse resting workpiece in annular interval distribution on index turntable;Feeding device is located on rack, for conveying workpiece to each work station of index turntable one by one;Visual detection device is located above index turntable, and it is oppositely arranged with one work station position, for detecting the appearance and size of workpiece of corresponding work station;Discharging device is located on the outside of index turntable, and located downstream of visual detection device, for classifying and discharging workpiece on index turntable after detection.The general assembly disc detection machine provided by the utility model can solve the problem of how to improve detection accuracy.

This invention discloses a progressive multimodal remote sensing image super-resolution reconstruction method and apparatus. The method includes: inputting a low-resolution image; extracting primary modal features and secondary modal features from the low-resolution image; inputting the primary and secondary modal features into a context-aware adaptive multimodal alignment module, and processing the input modal features to obtain processed secondary modal features; fusing the processed secondary modal features at different levels through a progressively gated fusion mechanism to obtain the final fused features; inputting the final fused features into a preset reconstruction module to obtain the final super-resolution reconstructed image; and outputting the final super-resolution reconstructed image. This method employs a context-aware adaptive multimodal alignment module to achieve interaction and alignment of multimodal features; and uses a progressively gated fusion mechanism to achieve progressive reconstruction from local detail enhancement to high-level semantic guidance.

This invention discloses an auxiliary robot construction method and system for boulders masonry in water conservancy projects, relating to the field of data processing technology. The method includes the following steps: acquiring the morphological data of the boulders to be laid and the data of the existing masonry structure, and analyzing to obtain the initial placement position; performing position correspondence analysis based on the initial placement position and the current masonry structure to obtain a centroid offset parameter set; combining the data of the existing masonry structure to perform inter-layer attitude influence analysis, obtaining an inter-layer attitude influence parameter set, and performing cumulative analysis along the extension direction of the current masonry layer to obtain the cumulative change value of inter-layer attitude; determining inter-layer stability based on the cumulative change value of inter-layer attitude; if the inter-layer is stable, the initial placement position is used as the final placement position; otherwise, the initial placement position is corrected to obtain the final placement position, which is then sent to the auxiliary robot. This invention can reduce the risk of masonry tilting caused by the cumulative offset of irregular boulders' inter-layer attitude.