38results about How to "Reduce error accumulation" patented technology

The invention discloses a short-term traffic flow prediction method based on a dynamic multi-dimensional global perception graph convolutional network, and belongs to the field of traffic flow prediction. The method comprises the following steps: acquiring an input dynamic graph feature sequence, and generating space-time element parameters; carrying out feature enhancement on the dynamic graph feature sequence, fusing the dynamic graph feature sequence with the space-time element parameters, and then respectively carrying out space-time feature coding; fusing the spatio-temporal characteristics obtained by coding the spatio-temporal characteristics; and carrying out space-time trend alignment decoding on the fused space-time features. According to the method, the defects of limited modeling precision, insufficient dynamic adaptability, incomplete spatial-temporal feature capture and the like in short-term traffic flow prediction are overcome, and the accuracy and the applicability of short-term traffic flow prediction are improved.

This invention proposes a multi-field-of-view coupled drilling robot rig positioning method, applicable to complex underground roadway environments in coal mines. Current drilling rig positioning methods primarily rely on vision sensors, which suffer from difficulties in selecting borehole points, potential damage to the anchor mesh structure during drilling, and the presence of dust and water mist during drilling, which obstructs vision sensors and causes image loss. This application proposes a multi-field-of-view coupled drilling rig positioning scheme based on the spatial positions of the drilling robot's components, and formulates a borehole point positioning strategy to avoid interference between the borehole and the anchor mesh during drilling. By identifying and locating the boreholes in the image, precise anchoring and positioning of the drilling rig are achieved. This not only has direct significance for improving support efficiency and safety but also provides important reference for positioning technologies of other underground engineering equipment.

A locking mechanism for a diagonal tie rod in an outward-opening, top-hung hardware system is disclosed. The diagonal tie rod includes a movable tie rod slidably mounted on a base for positioning a first connecting rod. The base has a first connecting rod with a limiting groove at one end near the base. The positioning device includes a limiting block with a locking part. When the window sash needs to be opened to its maximum position, the locking part extends out of the base and engages with the limiting groove to form a limiting connection, thus locking the window sash. The movable tie rod has an unlocking part. When the window sash needs to be closed, the locking part slides within the base driven by the handle. The unlocking part abuts against the limiting block, creating a lever effect that disengages the locking part from the limiting groove, thereby releasing the locked state of the window sash.

The invention provides an auxiliary welding device and a welding method. The auxiliary welding device comprises a support. The bracket includes: a first rod extending in a first direction; the second rod extends in the second direction, is arranged on the first rod and is provided with a first clamping part; the first clamping part is used for being clamped with a first position of a supporting rod of the rotating brake; the first direction is perpendicular to the second direction; the rotating brake comprises a main rod and a plurality of supporting rods connected with the main rod; the first guide piece is provided with a second clamping part; the second clamping part is used for being clamped with a second position of the supporting rod of the rotating brake, so that the first clamping part and the second clamping part are arranged at intervals in the first direction, and the supporting rod is fixed in the first direction; and the second guide piece is used for being clamped with the main rod of the rotating brake, so that the length direction of the connecting hole of the main rod is the same as the first direction, and one end of the supporting rod can be located in the connecting hole of the main rod in the length direction. According to the welding auxiliary device, the welding machining convenience of the welding auxiliary device on the main rod and the supporting rod of the full-height rotating brake can be improved.

This invention relates to the field of medical image processing technology, specifically to an automatic segmentation and recognition method and system for cancer images based on the fusion of multimodal images. The method includes: acquiring a multimodal medical image set containing computed tomography (CT) images, magnetic resonance imaging (MRI) images, and positron emission tomography (PET) images; generating spatially aligned data blocks through three-dimensional spatial registration and voxel alignment; extracting fusion features using an improved encoder-decoder network optimized based on the characteristics and spatial dependencies of different modal channels; obtaining enhanced semantic feature maps through multi-level context-aware processing; identifying suspected cancer regions through a region candidate generation network; and simultaneously outputting pixel-level segmentation masks and cancer type probabilities through a segmentation network, ultimately generating a three-dimensional segmentation and recognition result. This method can accurately capture multimodal-specific features, clearly distinguish the boundaries between lesions and normal tissue, simplify the processing flow, reduce errors, and achieve accurate and efficient segmentation and recognition of cancer lesions.

The utility model relates to a socket mounting assembly for gypsum board installation, the mounting assembly includes the positioning piece and sets up the adjusting portion of positioning piece, the adjusting portion includes adjusting rod and limit piece, the adjusting rod acts on limit piece realizes limit piece relative to the change of positioning piece from first state to second state, the first hole is opened in the center of positioning piece, along the first hole hole edge is evenly provided with first convex part and second convex part respectively, the positioning piece is provided with the baffle at one side end face, still is provided with third convex part on the outer ring surface of the end away from baffle. Form half closed structure through the design of the first hole and the baffle of positioning piece, replace traditional pre -buried bottom box, avoid forming local stress concentration point on light gypsum board, and the bidirectional joint design of third convex part and limit piece disperses the load evenly to the opening edge of gypsum board, significantly reduces the deformation cracking risk of gypsum board or suspended ceiling due to local stress.

The application discloses a Transformer oil well production prediction method and device based on gating constraints and Kalman filtering, and a medium, which comprises the following steps: first, obtaining oil well historical production data and preprocessing; then, using a mutual information method to screen input features related to target production, and constructing conditional gating constraint features representing the opening and closing states of the oil well; then, inputting the processed time sequence samples into a Transformer prediction model to obtain initial prediction results of the oil well production; then, performing state constraint on the initial prediction results according to the conditional gating constraint features, and performing causal recursive residual correction on the prediction results of subsequent time based on the current time prediction residual by using Kalman filtering to obtain final prediction results; further, performing grid search on Kalman filtering parameters Q and R for optimization, and determining the optimal parameter range by combining a heat map. The application can effectively reduce the prediction error, and improve the accuracy, stability and robustness of the oil well production prediction.

This invention provides a superimposed 5-9 axis coordinate measuring machine, relating to the field of dimensional inspection equipment. The superimposed 5-9 axis coordinate measuring machine includes an X-axis measuring stage base, a Z-axis assembly, and a self-compensating Y-axis assembly. The self-compensating Y-axis assembly is mounted on the X-axis measuring stage base, and the Z-axis assembly is mounted on the self-compensating Y-axis assembly. The self-compensating Y-axis assembly is used to enable the Z-axis assembly to smoothly adjust its position along the Y-axis. A counterweight balancing component is installed at the lower end of the Z-axis assembly. This superimposed 5-9 axis coordinate measuring machine can handle large molds, housings, and wind power / aerospace structural components. The existing side head of the equipment performs point dimensional / geometric tolerance checks; another independent probe performs surface scanning. These functions are independent of each other, making it suitable for reverse engineering and full inspection of complex surfaces. Multiple independent probes are responsible for different areas, eliminating the need for frequent movement of the main spindle or turntable, reducing the number of axis linkages, and indirectly reducing the accumulation of errors in the superimposed axes.

The present disclosure provides a method, device and electronic equipment for characterizing the boundary stiffness of a lattice structure, belonging to the field of material mechanics and structural design. The present disclosure divides the lattice structure, obtains the divided lattice unit, determines the type of the lattice unit, and the type is used to indicate the integrity of the lattice unit. Based on the type and the rod boundary condition of the lattice unit, the equivalent stiffness of the lattice unit is determined. The rod boundary condition is used to indicate the stress constraint state of the rod. In this way, compared with the existing fixed calculation method of heterogeneous multi-scale finite elements, the present disclosure can adopt differentiated flexible mechanical calculation for lattice units of different integrity types combined with their specific rod boundary conditions, avoiding errors caused by uniform calculation. In summary, the technical solution provided by the present disclosure can improve the accuracy of characterizing the boundary stiffness of the lattice structure, reduce error accumulation, and can adapt to various application scenarios.

The utility model belongs to the die field of heat insulation plate production, specifically is a front cabin heat insulation plate die, both sides of lower die plate top are all fixedly installed with support block, the top of support block is fixedly installed with support plate, the four corners of support plate top are all fixedly installed with guide pillar, a plurality of clamping grooves are set up in the surface of guide pillar, the inside rotation of clamping groove is installed with ball, the four corners of upper die plate are all fixedly installed with connecting lug, the outside of connecting lug is fixedly installed with linear bearing, the surface of guide pillar sleeve sets up linear bearing. Through the structural design of support plate, guide pillar, clamping groove, ball, connecting lug, linear bearing, has realized the function of reducing the wear and tear of guide, thereby can pass through the cooperation of ball in clamping groove and linear bearing, converts sliding friction into rolling friction, reduces friction resistance greatly, reduces the error accumulation in the movement process, reduces friction and wear and tear not only has improved the guiding accuracy, also has prolonged guide pillar life greatly.

The application provides a kind of based on feature enhancement and composite model's ultra-short-term wind power prediction method, belongs to ultra-short-term wind power prediction technical field, this method includes: by calculating the mutual information of multiple meteorological characteristics and wind power output and normalizing, screening to obtain key features, and constructing wind power prediction fitness function, by minimizing wind power prediction fitness function, obtain enhanced input features;By introducing one-dimensional convolutional neural network and bidirectional long short gate recurrent unit, construct the residual network of bidirectional long short gate recurrent unit and convolutional neural network fusion, and introduce space-time attention mechanism, construct to obtain composite residual network ultra-short-term wind power prediction model;And use multiple meteorological characteristics and enhanced input features to predict to obtain ultra-short-term wind power prediction result;The application solves the problem that the influence of meteorological multi-scale fluctuation on output cannot be accurately grasped, prediction is disconnected, disturbance is not considered and prediction accuracy continues to decline with step increase.

This invention relates to a machining method for improving the coaxiality of the spindle bores in a worktable housing, belonging to the field of high-precision parts machining. The method includes the following steps: S1, fixing the worktable housing on a vertical lathe and precision machining a first datum surface, bores, and datum holes; S2, flipping the worktable housing and fixing it on the vertical lathe, precision machining a second datum surface and bores; S3, flipping the worktable housing and fixing it on a vertical milling table, precision grinding the first datum surface, bores, and datum holes; S4, flipping the worktable housing and fixing it on the vertical milling table, installing a part receiving plate into the bores on the second datum surface, and precision grinding the bores on the part receiving plate. A dial indicator is used for coaxiality correction, and equal-height magnetic blocks, fixtures, and adjusters are used to improve adjustment accuracy and the coaxiality of each bore. This invention, by repeatedly flipping the worktable housing and performing precision machining and calibration of the bores in stages, avoids the cumulative errors caused by single-processing, thus improving machining accuracy and the coaxiality of the bores.

The utility model provides a quick-change variable-pitch adsorption device for an end effector of an industrial robot, which can be applied to the tail end of the industrial robot. In automatic production, products are usually machined in groups, the distance needs to be adjusted, and a part of work stations need to be switched to be compatible with different products. In the prior art, the defects of complex structure, low precision, poor universality and inconvenience in switching exist in pitch changing, quick changing and adsorption functions. The device has the advantages that high-precision pitch changing can be achieved, products are made to be neat, the pitch is changed, universality is high, and the device can adapt to various products; the quick-change function is quick, convenient and high in stability, and time and labor are saved; the single up-and-down adsorption function is achieved, products of special shapes or uneven weights are firmly adsorbed, and the device is particularly suitable for carrying round bars of about 10 Kg; the whole structure is simple, the principle is clear, manufacturing cost and maintenance difficulty are reduced, operation and troubleshooting are facilitated, production efficiency and product quality are effectively improved, and diversified production requirements are met.

A spatiotemporal multimodal quality prediction system for bananas and its training method are disclosed, relating to the field of intelligent detection and preservation information technology for agricultural products. This system alleviates the problems of existing banana quality prediction technologies, such as difficulty in comprehensively reflecting the intrinsic mechanisms of banana quality evolution and the tendency for error accumulation during long-term prediction. The spatiotemporal multimodal quality prediction system for bananas includes an image data acquisition module for obtaining visual spatial features based on the time series of images of the banana to be detected; a physiological data acquisition module for obtaining physiological features based on physiological data; a multimodal data fusion module for fusing the visual spatial features and physiological features through multimodal data fusion to obtain global temporal features; and a multimodal quality prediction module for using the global temporal features to obtain a predicted banana image and corresponding predicted physiological data through multi-task collaborative decoding. This invention is applicable to fields such as multimodal data fusion and non-destructive testing of agricultural product quality.

This invention discloses an intelligent annotation method and system based on oral CBCT images, belonging to the field of oral medical image processing technology. The method includes: acquiring CBCT voxel data of a target tooth region; performing sliding window analysis on the CBCT voxel data to generate local grayscale statistical features for each sliding window; for each sliding window, constructing a dual-modulation adaptive threshold function containing a primary threshold and a secondary threshold based on the local grayscale statistical features, and jointly judging the CBCT voxel data to form a preliminary tooth region. This solution, through the collaborative work of the primary and secondary thresholds and combined with dynamic adjustment of the judgment conditions based on grayscale distribution similarity, can effectively distinguish grayscale overlapping areas of enamel, dentin, alveolar bone, and other tissues. It solves the problems of fixed threshold segmentation in existing technologies being sensitive to local noise and unable to adapt to changes in grayscale distribution, significantly improving the accuracy and robustness of preliminary tooth region annotation.

The utility model belongs to the technical field of electronic equipment manufacturing, concretely relates to an electromagnetic shielding disc assembling device, including base guide rail and shielding disc guide rail, the end of base guide rail is located the above of shielding disc guide rail end in vertical direction and both form perpendicular intersection layout on three -dimensional space and the inside of this intersection region is provided with machine base in the intersection region, integrated mechanism of grabbing riveting is integrated on the machine base, a plurality of bases are evenly arranged on the base guide rail along its length direction, a plurality of shielding discs that are matched with the base are correspondingly arranged on the shielding disc guide rail, a plurality of shielding discs are sequentially assembled on the corresponding base through grabbing riveting integrated mechanism. The utility model discloses through space layout optimization, function integration design and high-precision mechanical structure innovation, solves the low degree of automation, positioning accuracy deficiency, poor adaptability and other core problems in the prior art, realizes the efficient, high-precision assembly of electromagnetic shielding disc.

The invention discloses a dynamic environment object searching method based on common sense and memory enhanced reasoning, and belongs to the technical field of robot object searching. The method comprises the steps that a robot obtains an article search task and historical interaction data input by a current user; rGB-D images of the surrounding environment are collected in real time, and various object instances are recognized; establishing a functional three-dimensional scene graph comprising a container layer, a movable object layer and a region layer according to the detected environment and object information; a hierarchical action space comprising a macroscopic level and a microscopic level is designed, a macroscopic action planning cue word is input into LLM according to current scene information, and the optimal next-step action of the robot is obtained; and repeating the steps until the task is completed or failed. According to the method, functional three-dimensional scene graph construction, layered action planning and a dynamic environment response mechanism are organically combined, and efficient target searching in a dynamic environment can be achieved.

The invention discloses a visual system-based industrial robot for product appearance detection, which comprises a sorting structure, a driving structure is fixedly arranged on the sorting structure, a pair of swinging structures is symmetrically arranged on the driving structure, picking structures are respectively arranged on the pair of swinging structures, and the picking structures can be oppositely attached and butted. The invention relates to the technical field of detection equipment, and has the advantages of simple structural design, high transmission efficiency, convenient maintenance, strong action linkage synchronism, no redundancy in process connection, efficient double-sided detection, realization of synchronous detection, controllable precision, assembly line continuous operation, adaptation to mass production requirements, detection and sorting integration, and high efficiency. The sorting is accurate, the main process is not interfered, the action adaptability is high, the application scene is wider, the operation automation degree is high, and manual intervention is reduced.

The application discloses a short-term and nowcasting precipitation prediction method based on a multimodal RNN, relates to the technical field of short-term and nowcasting precipitation prediction, and aims to provide a short-term and nowcasting precipitation prediction method with high accuracy, which comprises the following steps: acquiring a radar graph and a meteorological element graph; training a multimodal model by using the radar graph and the meteorological element graph; and performing precipitation prediction by using the trained multimodal prediction model; wherein the multimodal prediction model comprises a radar module and a meteorological element module, and the multimodal prediction model obtains a short-term and nowcasting precipitation prediction result according to a radar prediction graph; the method takes radar information as the main part and meteorological element information as the auxiliary part to perform precipitation prediction, fuses the two modes, and improves the accuracy of short-term and nowcasting precipitation prediction.

The present application relates to the technical field of neuroscience and biological information cross processing, and particularly relates to a real-time whole-brain neural activity signal extraction method for free behavior animals, obtaining original three-dimensional body imaging data collected by a free behavior animal at a body frame acquisition frequency f under a body imaging system, and performing robust intensity normalization on the original three-dimensional body imaging data; the normalized data is split into two-dimensional slices along the axial direction, and a lightweight fully convolutional network is combined with multi-scale feature fusion to detect a neuron rectangular frame; the present application simplifies three-dimensional modeling of neurons into an object detection problem, adopts a point cloud non-rigid registration method supporting partial correspondence, and combines a fixed reference template to realize error-free accumulation of spatiotemporal consistency maintenance, significantly reduce the calculation complexity, realize fully automatic and high-robust real-time signal extraction, and solve the problems of insufficient real-time performance, low precision, easy accumulation of errors and dependence on artificial labor in the prior art.

The invention discloses a satellite image building element parameterization reconstruction method and system based on physical geometric constraint and differential rendering coupling. The method comprises the following steps: firstly, constructing a multi-scale multi-task network based on Swin Transform to extract image features; secondly, a physically guided shadow consistency constraint is introduced, height prediction is reversely calibrated by using a shadow geometrical relationship, and the ambiguity problem of monocular height estimation is solved; meanwhile, regular vector extraction is carried out on the building contour by utilizing a frame field theory; finally, a reverse modeling strategy based on micro rendering is adopted, roof geometric parameters are automatically solved, and an LOD2-level parameterized three-dimensional model is generated. According to the method, the physical consistency, the contour regularity and the modeling automation level of three-dimensional reconstruction are effectively improved, and the method is suitable for high-precision construction of large-scale urban digital bases.

The invention provides a timing precision calibration system for a clock driven by an intelligent sensor, and relates to the technical field of data processing. Temperature sensing units are arranged in different heated directions of a clock timing device to obtain original temperature sequence data with time stamps and direction information; identifying a dominant heated area based on the temperature change rate and the duration and generating a corresponding weight; analyzing a heat transfer time difference in combination with the installation position of the crystal oscillator, and constructing a local heat influence response sequence with time delay; deriving a crystal oscillator frequency evolution process according to the response sequence and generating a predicted frequency offset; the timing pulse is pre-adjusted to form a timing reference signal before the error is generated; and meanwhile, through comparison with an external standard time source, the weight of the dominant heat source and the heat transfer delay relation are corrected, and dynamic calibration of timing precision is realized.

The application relates to the technical field of remote sensing and radar imaging, and relates to a space target ISAR imaging and scaling integrated method based on a translation-rotation parameter joint estimation.The application is used to solve the problem that the existing space target ISAR imaging is affected by motion compensation, and the scattering points of a space target are prone to moving across distance units in a long-time accumulation.The application performs distance compression on echo obtained by a ground-based radar, performs generalized Radon Fourier transformation on a one-dimensional distance image sequence, and further estimates phase coefficients of the echo signal; the effective rotation angular velocity module value of a space uniform rotation target and a compensation signal are calculated by using the estimated value; the echo signal is subjected to translation compensation by using the compensation signal, the compensated echo signal is converted from a Cartesian coordinate system to a polar coordinate system by using the effective rotation angular velocity, and interpolation is performed; the imaging result is obtained by performing distance and azimuth compression on the interpolation result, scaling is completed according to the effective rotation angular velocity, and the space target ISAR imaging and scaling integration are realized.

The invention relates to the technical field of electromechanical components, and discloses a quick-locking sealing connector which comprises a plug and a socket, the plug comprises a contact pin, a cable and a conical screw, the socket comprises a shell and a jack which is formed in the shell and corresponds to the plug, and a copper core climbs along an external thread which is gradually shrunk under manual assistance and is tightly wound on a rod part of the conical screw. The copper cores are arranged more orderly and normatively, bidirectional fastening between the copper cores and the conical screws, circumferential holding and limiting of the threads on the copper cores and winding and restraining of the screws by the copper cores can be achieved, the overall connection strength is greatly improved, external force such as vibration and pulling can be effectively resisted, loosening and falling are avoided, the structure can adapt to the copper cores of different lengths, and the service life of the copper cores is prolonged. The too long copper core can be wound for multiple circles, the too short copper core can be attached for few circles, the length of the copper core does not need to be trimmed like a traditional structure, meanwhile, the copper core attached in a climbing mode and the screw form stable large-area conductive contact, the remaining copper core is stored and positioned in cooperation with the containing hole, and conductive point positions are further increased.

The invention discloses an acceleration coefficient interval estimation system, which comprises the following steps of: acquiring sufficient complete failure sample data through a Testto-Failure test, designing a'stress layer and degradation characteristic layer 'double-layered Bootstrap sampling rule, and generating a plurality of groups of failure sample combinations; meanwhile, whole-course degradation data of failure samples under double stress are integrated, a cross-stress integrated joint likelihood function is constructed, the acceleration coefficient of each combination is directly solved, and finally the acceleration coefficient interval is determined through statistical analysis. By the adoption of the technical scheme, the problems that existing acceleration coefficient estimation can only achieve point estimation and lacks effective interval estimation adaptive to Testto-Failure data features, an interval result cannot cover failure scene parameter fluctuation, and error accumulation exists in traditional step-by-step solving are solved, and the accuracy of acceleration coefficient estimation and the interval reference value are improved.

The application provides an intelligent frequency utilization decision-making method and device based on electromagnetic situation prediction confidence, comprising: determining a time sequence input matrix according to historical electromagnetic situation information; inputting the time sequence input matrix into a pre-trained electromagnetic situation prediction model for multi-scale frequency feature decomposition and enhancement, performing noise elimination coding and mapping on the enhanced features to obtain an electromagnetic situation prediction result of a future time slot; quantifying the reliability of the electromagnetic situation prediction result through a prediction confidence extraction mechanism to obtain a prediction confidence; constructing a state vector for a reinforcement learning decision-making network; estimating a decision-making risk according to the prediction confidence and constructing a reward function based on dynamic adjustment of the prediction confidence; and using the state vector and the reward function to drive the reinforcement learning decision-making model to output a frequency utilization decision-making action, so as to balance the utilization of prediction information and risk control; improve the quality of prediction prior information in a complex electromagnetic environment, and improve the stability and reliability of intelligent frequency utilization decision-making under long-view prediction.

The utility model provides a kind of high-precision robot joint gear transmission structure, it is related to planetary gear technical field, comprising: three planetary gear sets, each planetary gear set includes two gears connected and coaxially arranged, two gears are respectively gear located above as first planetary gear and gear located below as second planetary gear, the first planetary gear and the second planetary gear are all unevenly distributed gear number gear, first flat side and second flat side are respectively provided on the second planetary gear and the first planetary gear;Planet carrier.The utility model has beneficial effects: by adopting three planetary gear sets and the first planetary gear and the second planetary gear are all unevenly distributed gear number gear, and first flat side and second flat side are respectively set on it, cooperate with the mark on planet carrier, the accurate tooth and meshing of planetary gear set are realized, effectively improve transmission accuracy, reduce error accumulation, ensure the high-precision control of robot joint movement.

This invention relates to the field of combustion experiment technology, specifically to an adaptive adjustment control system and method for NOx formation mechanism combustion experiments. The method includes: inputting user-entered current experimental data, historical experimental data, a preset knowledge base, and an experimental plan; inputting historical experimental data, the experimental plan, and the preset knowledge base into a pre-constructed experimental element identification model, outputting key fuzzy variables, a fuzzy range set, control quantities, and fuzzy rules, and establishing an initial fuzzy rule library; inputting current experimental data and the experimental plan into a fuzzy optimization model, outputting a weight set and an activation sequence; determining the activated rules and the set of adjustment variables based on the activation sequence; and performing model demodeling based on the initial fuzzy rule library, weight set, and adjustment variable set to obtain the adjustment value of each adjustment variable. This invention uses fuzzy algorithms to handle uncertainties and combines machine learning models to optimize the rule library and model demodeling method, thereby improving the adjustment accuracy of experimental parameters and the reliability of experimental results.