938 results about "Point location" patented technology

The invention relates to the technical field of computer vision and industrial automation, in particular to a shoe sole three-dimensional gluing contour extraction system and method based on vision single-line laser, and aims to solve the problems that virtual calibration target spots cannot be accurately generated based on shoe sole geometry, the positions and sizes of the target spots are difficult to determine by combining curvature extreme values and principal component analysis in the prior art, and the production cost is low. The problem that a double-branch deep learning model cannot be adopted to fuse feature prediction transformation, and the re-projection error is increased is solved; a virtual calibration target spot is automatically generated based on sole geometry through a feature fusion calibration module, a grid is generated through point cloud processing and Poisson reconstruction, the position and size of the target spot are determined by combining a curvature extreme value and principal component analysis, a corresponding relation is established by utilizing two-dimensional and three-dimensional feature matching, initial alignment is realized through ICP and re-projection error optimization, and the target spot position and size are determined. A double-branch deep learning model is adopted to be fused with feature prediction transformation, iterative optimization is carried out through space consistency errors, and re-projection errors are reduced.

Examples described herein provide automatic location of access points by a computing device. Examples may include receiving, by the computing device from each AP in a subset of a plurality of APs, a Global Navigation Satellite System (GNSS) signal measurement, and based on each received GNSS signal measurement, constraining, by the computing device, the map of relative AP locations by at least one translational degree of freedom or one rotational degree of freedom. Examples may include resolving, by the computing device, locations of the plurality of APs in the map of relative AP locations.

The invention discloses an automatic river network grading and river topological relation identification method based on a graph theory, and relates to the technical field of hydrological geographic information. The method comprises the following steps: acquiring and cleaning a vector river network, a key point location and DEM data of a target drainage basin; constructing an initial river network graph model based on the line element connection relationship; integrating DEM topographic evidence and graph theory connection features, constructing and solving a global potential energy field equation containing topographic driving and boundary constraint, and calculating flow potential energy attributes of nodes of the whole network to determine a flow relationship; based on the flow direction relation, identifying topology abnormal structures such as strong connectivity components in the network, and performing ring breaking processing by using direction confidence to generate a ring-free directed network structure; and performing river grade division based on a topology transfer rule, and associating the key point location to a river network skeleton. According to the method, through global potential energy field solving and topological optimization, the problems that the flow direction of the plain micro-geomorphic area is difficult to recognize and complex loops cannot be graded are solved, and automatic construction of the river network topology is achieved.

The invention discloses an end-cloud collaborative deployment method and device for a multi-modal large model, a medium and a product, and relates to the field of multi-modal large model deployment, and the method comprises the steps: a cloud side computer obtains a to-be-deployed multi-modal large model, determines an optimized coding model and a corresponding segmentation candidate point according to the to-be-deployed multi-modal large model, and transmits the optimized coding model and the corresponding segmentation candidate point to an end side computer; the end side computer obtains multi-modal data, the optimized coding model is used for coding the multi-modal data, and intermediate data and segmentation point position information are obtained; and the end-side computer compresses and packs the intermediate data and the segmentation point position information, and sends the compressed and packed intermediate data and the segmentation point position information to the cloud-side computer through a transmission path, so that the intermediate data is calculated and processed by using a processing model to obtain a calculation result, and the calculation result is sent to the end-side computer, decoded by using a decoding model and converted into an output format. And obtaining the processed multi-modal data. The method can dynamically adjust the distribution of the calculation tasks, and improves the reasoning efficiency.

The invention provides a three-dimensional space track identification method based on multi-view three-dimensional reconstruction, which comprises the following steps of: shooting the same scene from different angles by using a plurality of cameras, and determining a picture intersection part of the plurality of cameras as a research scene; frame extraction and pedestrian detection are carried out on the video of each camera, pedestrians and position information of the pedestrians in each frame of video are identified, a unified identity number is distributed to the same pedestrian, registration is carried out on video frame images of different cameras, the pixel position of each pedestrian in each camera is obtained, and a multi-view-angle triangulation method is adopted, and internal and external parameters of the cameras are combined, so that the pedestrian positioning accuracy is improved. Solving a three-dimensional point coordinate of each pedestrian in a camera coordinate system; the three-dimensional point coordinates are converted into coordinates in a world coordinate system, and world coordinate system coordinates of pedestrian point locations of all frames and all numbers are obtained; and according to the time information, connecting the pedestrian point locations of each frame and each number to form a track, and completing reconstruction of the pedestrian three-dimensional space track in the world coordinate system. According to the invention, manpower and calculation cost can be reduced.

The invention relates to a multi-person motion capture real-time digital human driving method based on a skeleton-muscle interaction model, and belongs to the field of artificial intelligence and ubiquitous universe. According to the method, the key points of the human body are detected through YOLO-Pose, the positions of the shielding points are obtained through a double-stage dynamic shielding point compensation method, and the accuracy of the key points is guaranteed. And outputting the three-dimensional coordinates of the key points of the human body by adopting a three-dimensional key point depth extractor based on multi-dimensional exclusive-shared feature reasoning in combination with an MT ransPose model. A real person-skeleton model-digital person real-time mapping method is provided, three-dimensional key points are mapped to a standard human body skeleton model, and a digital person role and the skeleton model are bound through a weight mapping algorithm. Based on a skeleton-muscle interaction model, the digital person accurately reflects a real motion posture by introducing the mechanical action of a skeleton rigid body dynamic model and a muscle force line model, and the process of driving the digital person in real time through multi-person motion capture is achieved.

The present disclosure provides an artificial intelligence-based method for enhancing mine safety by identifying and predicting locations of water inrush points in a mine, including the following steps: S1: constructing a numerical model to determine priori information of parameters to be recognized based on observation data, including coordinates of locations of water inrush points; S2: generating a training sample dataset and a test sample dataset of an alternative model based on the numerical model and the priori information of the parameters; S3: constructing and training a neural network of the alternative model; S4: testing an accuracy of the alternative model; and S5: performing a simulated annealing algorithm to identify the locations of water inrush points and simulation model parameters.

The invention provides a reinforcing steel bar binding point detection method based on improved YOLOv8, and the method comprises the following steps: obtaining a construction site image, carrying out the preprocessing, inputting a to-be-detected image into an improved target detection model, sequentially carrying out the local space attention calculation, depth separable convolution and channel attention weighting operation, and carrying out the detection of the to-be-detected image. Detail feature expression of reinforcing steel bar intersection points is enhanced; performing dynamic weight fusion on the top-down first feature path and the bottom-up second feature path, and embedding a lightweight residual unit in a fusion node; a classification task and a regression task are respectively processed by adopting a classification branch and a regression branch, the classification branch outputs category confidence, and the regression branch outputs coordinate positioning through depth separable convolution and a probability distribution prediction layer; and decoding output data of the target detection model, and generating a detection result containing the position and the category of the binding point. According to the method, the requirement of real-time detection can be met while the accuracy of steel bar binding point detection is improved.

The invention discloses an AR glasses control system and method, and belongs to the field of artificial intelligence and AR glasses application, and the method comprises the steps: capturing a user face image, combining a voice assistant, collecting user voiceprint data, and carrying out the matching through a recurrent neural network; constructing a three-dimensional environment model, positioning and marking obstacles and interactive objects, and identifying objects in the environment; the method comprises the following steps of: tracking a gesture, acquiring a key point coordinate by adopting an OpenPose algorithm, identifying a predefined gesture by utilizing a support vector machine, and mapping the gesture into a specific operation instruction; analyzing the user intention based on natural language processing, and calling a corresponding function module according to an analysis result; acquiring a gazing direction of a user by using an eye movement tracking sensor, calculating a screen focus position through an algorithm, and dynamically highlighting options or displaying information according to a gazing point position; through multi-modal data fusion, gestures, voices and fixation point input are analyzed in a unified manner. By accurately understanding the intention of the user, the interaction precision of the AR glasses is improved, and the user experience is optimized.

The invention discloses an abnormal detection method and system for a splayed anti-loosening line of a rail train, and relates to the technical field of rail traffic detection, and the method comprises the steps: image collection and processing: shooting an image of a fixed point region containing the anti-loosening line, and preprocessing the image to obtain a target image; line integrity judgment: extracting a bolt, a bolt marking line and an anti-loosening line through an analysis model to form a target area graph, and judging whether the anti-loosening line is complete or not according to line segment continuity and form; performing line deformation depth analysis, and synchronously obtaining the position of an anti-loosening line intersection point and deformation data through the model in combination with the loosening angle of the bolt marking line when the line deformation is complete; performing line deformation safety judgment, setting a safety threshold group based on historical data, and comparing deformation data to output a signal; after line deformation abnormity judgment and comparison signal receiving, actual data are compared with theoretical data output by the mechanical model, and abnormal or normal signals are output. The anti-loose line full-dimension high-precision detection is achieved, automation and real-time performance are improved, and train operation safety is guaranteed.

The invention discloses a cultural relic digital fingerprint identification method based on geometric texture and spectral information, and the method comprises the steps: obtaining multi-source data, carrying out the texture mapping, and forming a three-dimensional grid model with texture data; curvature values and topological features of all vertexes of the three-dimensional grid model are calculated, feature fusion is carried out on the basis of local saliency adjustment feature weights, and fingerprint point location screening is completed; based on the screened fingerprint point locations, a significant plane is constructed through a bit plane, and texture features of the cultural relics are extracted based on an improved histogram and a gray-level co-occurrence matrix; extracting characteristic wave bands of the cultural relics based on the screened fingerprint point locations in combination with the hyperspectral data; and multi-modal fusion identification is carried out based on features of geometry, texture and spectral information. According to the method, the geometric structure, the texture features and the hyperspectral information are fused, multi-mode collaborative cultural relic identity high-precision identification is achieved, and the method has the advantages of being high in accuracy, high in robustness, good in interpretability and capable of adapting to complex environments.

The invention provides a robot positioning prediction method and device, a medium and equipment, and relates to the field of industrial robot motion control. According to the method, on one hand, a physical-data double-branch collaborative network is designed, meanwhile, a DH parameter method is used for predicting and obtaining a theoretical position coordinate of a target robot, an improved Transform network is used for predicting and obtaining a coordinate compensation value of the target robot, so that a theoretical value and the compensation value are decoupled, and the limitation of single model establishment is avoided; and then, the two are fused, and high-precision end point location prediction for the target robot is realized by combining priori knowledge of robot kinematics and nonlinear expression ability of deep learning. And on the other hand, a spatial physical information mixed loss function is provided, and a spatial topological structure output by the DH model is used to guide the distribution of predicted values of the whole position prediction model, so that the geometric rationality of target robot end point location prediction is improved.

The invention discloses a path planning method and device, automatic driving equipment and a readable storage medium. The path planning method comprises the steps that path information and the starting point position of automatic driving equipment are acquired, the path information comprises a path reference line, and a to-be-optimized path point is determined based on the path reference line; determining a Frenet coordinate system based on the starting point position and the path reference line; based on the to-be-optimized path point, a cost function is established under a Frenet coordinate system, the cost function is a quadratic function, a first penalty term is introduced into the cost function, and the first penalty term is used for representing the deviation between the to-be-optimized path point and the transversely-biased path reference line; and solving the cost function based on a preset constraint condition, and obtaining a target path corresponding to the minimum cost function. According to the invention, when the distance between the travelable boundary and the path reference line is small due to the fact that the obstacle partially occupies the travelable area, the planned target path can be far away from the obstacle, so that the driving experience is improved.

The invention provides an SLAM positioning method and system based on multi-sensor fusion, and the method comprises the steps: firstly receiving data of a plurality of sensors, and obtaining initial data including visual feature data, I MU pre-integration data and UWB distance data through anomaly detection and processing; initializing by using visual feature data and I MU pre-integration data, estimating an initial pose, and predicting a UWB anchor point position as a fixed constraint in combination with UWB distance data; inputting the initial data into a sliding window, outputting drifting-free odometer data, judging the motion state of the SLAM system according to the data, and if the motion state is static, executing zero-speed updating to obtain target pose data; and finally, determining a new key frame based on the parallax or visual feature data quantity of the current frame and the previous key frame. According to the invention, the problems of relatively weak robustness and relatively low positioning accuracy of the system facing extreme scenes are solved; by means of motion state judgment and zero-speed updating, the robustness of the system in long-term operation is improved.

The invention discloses a multi-protocol hardware communication adaptation module based on an energy storage system and a configuration method, the module comprises a protocol template library, an intelligent mapping engine and a model generator, and the method comprises the following steps: storing structured protocol point location information through the protocol template library; an intelligent mapping engine is used for shielding the isomerism of Modbus, MQTT, IEC104 and other underlying protocols, and a unified data interface is provided; intelligent matching among different protocol terms is realized by adopting a natural language processing technology through a semantic level mapping unit; and finally, generating a target data model by a model generator through one key. The method effectively solves the problems that when the energy storage system integrates multi-manufacturer equipment, protocol configuration depends on manpower, efficiency is low, errors are prone to occurring, and expansibility is poor.

The invention discloses a path planning method and system for corner optimization. The method and system are used for cutter radius compensation and machining instruction smooth transition of a three-dimensional path of a numerical control machine tool. The method comprises the steps that a machining instruction is pre-read, and a path point location is analyzed; calculating a path direction vector and a normal vector, and determining a path plane equation; calculating a plane included angle and an intersection direction vector through a normal vector; generating a compensation curve by translating along the perpendicular direction of the intersecting line by taking the radius of the cutter as a constraint; and generating a smooth control instruction through curve subdivision and inserting the instruction into an instruction queue. The problem that three-dimensional path corner transition is not smooth is solved, the machining precision and efficiency are remarkably improved, and the method is suitable for a numerical control machine tool for complex curved surface machining.

The invention relates to the technical field of building engineering, and provides a building construction progress digital monitoring method and system. The method comprises the following steps: collecting three-dimensional point location data of a construction area, and comparing the three-dimensional point location data with a BIM model to extract a progress deviation; associating the construction log to a BIM component node and setting a first evaluation index; associating the material records to BIM component nodes, and setting second evaluation indexes; drawing up a resource map of construction process logic in combination with the deviation and the two indexes; and based on the atlas, according to a weight attenuation coefficient, multi-scale assessment of cumulative influence is carried out by taking efficiency as a feature, and resource compensation reminding is carried out in combination with dynamic threshold adjustment. The technical problem that progress deviation is difficult to accurately recognize due to data updating lagging and construction process logic disjunction in traditional building construction progress monitoring is solved, real-time accurate analysis of the construction progress is achieved through intelligent comparison of multi-source data fusion and the BIM model, resource allocation is dynamically optimized in combination with multi-dimensional evaluation indexes, and the construction progress monitoring efficiency is improved. And the construction management efficiency is improved.

The invention provides a path planning method, electronic equipment and a computer program product. The intelligent agent realizes path planning based on a pulse neural network of width, and the path planning method comprises the steps that a target area is divided based on environment information of the target area and an optimization target, a rasterized map composed of different sub-areas is determined, the optimization target comprises calculation efficiency and environment adaptability of the intelligent agent, and the rasterized map is determined based on the calculation efficiency and the environment adaptability of the intelligent agent. Each sub-region corresponds to one pulse neural network sub-node, and the rasterized map comprises a starting point position, an ending point position and obstacle information; based on the environment feature vectors of the sub-regions of the rasterized map and the position information of the intelligent agent, the current state of the intelligent agent is determined, and the current state is determined based on the Markov decision process; and based on the current state of the intelligent agent and a path planning algorithm, performing path planning on the rasterized map, and determining a planned path.

A point cloud data transmission method according embodiments comprises a step for encoding point cloud data including the geometry and attributes, the geometry being information indicating the positions of the points of the point cloud data, and the attributes including one or more among the color and reflectance of the points, wherein the encoding step comprises: a step for converting the coordinates representing the positions of the points; and a step for transmitting a bitstream including the encoded point cloud data. The point cloud data transmission method can encode and transmit the point cloud data.

A monitoring point layout method for defect diagnosis of a drainage pipe network comprises the following steps: based on pipe network geographic information system data, constructing a grouping topology table through topological structure extraction and connectivity analysis, layering nodes according to a flow collection path to form a flow-closed upstream and downstream hierarchical relationship, and combining point backtracking, collection and upstream and downstream query in a group to obtain a flow-closed upstream and downstream hierarchical relationship; point locations in each layer of node group are adjusted to ensure the flow closeness; and with the fewest monitoring points and the longest covered pipe diameter as targets, an optimal node group combination is screened through an optimization model, and a final monitoring point position is generated by fusing a pipe diameter threshold value and environmental factors. According to the method, through a flow closing algorithm and a hierarchical optimization mechanism, the problems of low efficiency, insufficient coverage and flow closing deficiency of traditional manual layout are solved, automatic and high-precision point location layout is realized, the timeliness and accuracy of defect diagnosis are remarkably improved, meanwhile, the monitoring cost is reduced, and systematic technical support is provided for operation and maintenance of the urban drainage network.

The invention discloses an operation log recording method based on database monitoring, and the method comprises the steps: carrying out the buried point reporting of a log according to a preset configuration rule, caching an application buried point log according to a log buried point identifier, and enabling the application buried point log to record a business logic and a user operation from an application level according to a buried point position sequence. Information such as interface calling, user operation and execution results can be clearly recorded, meanwhile, data change information is obtained through database monitoring, data changes are accurately reflected, and a final application operation log generated by combining the two records application operation behaviors and reflects actual changes of the database at the same time; log recording does not need to be directly carried out on the database level, the influence on database performance is minimized, and meanwhile the consistency and integrity of logs are guaranteed.

The invention discloses a fault positioning method, device and equipment and a computer program product, and relates to the technical field of computer software, and the method comprises the steps: carrying out the anomaly detection of a to-be-detected business system, and obtaining a plurality of pieces of anomaly information; analyzing the multiple pieces of abnormal information to obtain an abnormal association relationship; and identifying the abnormal association relationship, and performing layer-by-layer recursive positioning of a fault cause according to the obtained abnormal topological relationship until analysis of each layer in the abnormal topological relationship is completed. The abnormal information in the service system is detected in real time, and data support is provided for fault positioning. By analyzing the abnormal association relationship between the abnormal information and further identifying the abnormal topological relationship, the service influence range can be identified. And according to the topological relation, recursively drilling down to a downstream service and resource level layer by layer until the root cause of the fault is positioned, so that the fault type and the fault point position are accurately positioned, and guarantee is provided for service continuity.

The invention relates to the technical field of displacement change monitoring, in particular to a large-scale structure displacement change tracking monitoring method and system based on digital images, and the method comprises the following steps: manufacturing a target with a known shape and size; collecting a global digital image of the large-scale structure when no target is arranged, defining the global digital image as a first digital image, marking a fixed point and a displacement tracking point to be measured in the first digital image based on prior knowledge of the large-scale structure, and arranging targets at the fixed point and the displacement tracking point to be measured; acquiring a digital image of the large-scale structure after the target is arranged, and defining the digital image as a second digital image; based on the target in the second digital image, target structure displacement analysis is carried out, target theoretical center point position coordinates corresponding to the displacement tracking points to be measured are obtained, and displacement changes of the large-scale structure are analyzed. According to the invention, the image identification monitoring precision of the to-be-measured displacement tracking point and the accuracy and robustness of change tracking of the to-be-measured displacement tracking point can be obviously improved.

The invention discloses a method for automatically calculating the waviness of an aircraft profile in an MATLAB environment, and the method comprises the steps: 1, obtaining the plane section point location data of the waviness of the aircraft profile based on the reverse point cloud data of the aircraft profile; and step 2, automatically calculating the airplane profile waviness in the MATLAB environment, including plane classification of point location data, three-dimensional coordinate system conversion, two-dimensional wave crest and wave trough point location calculation, wave crest and wave trough point coordinate system reverse conversion, waviness calculation and form report output on the point location data of the plane cross section obtained in the step 1 through a program in the MATLAB environment. Therefore, automatic calculation of the airplane appearance waviness is completed. According to the invention, the problem that the existing detection mode for the airplane profile corrugation is difficult to have convenient operability, detection precision and detection efficiency at the same time is solved.

The invention relates to a cluster scheduling method and system for multi-unmanned aerial vehicle cooperative detection of hoisting equipment, and the method comprises the steps: obtaining detection object information from a detection request when the detection request of a user side is received; calling a preset equipment three-dimensional model based on the detection object information, and generating a detection task based on the equipment three-dimensional model; performing demand analysis on the detection task, and outputting task demand information; a cooperative detection strategy is generated based on the task demand to generate a cooperative flight route, the unmanned aerial vehicles of the corresponding model and number are called, and the multiple unmanned aerial vehicles are controlled to cooperatively execute the flight task according to the generated cooperative flight route; and based on a preset communication link, receiving detection data acquired by the unmanned aerial vehicle, mapping the detection data to a corresponding detection point location in the equipment three-dimensional model, and sending the detection data to a user side. The method and the device have the effect of improving the efficiency and the accuracy of large-scale equipment detection.

The invention provides a method for accurately positioning and installing electromechanical equipment in a complex area, and belongs to the technical field of electromechanical equipment installing.The method comprises the steps that three-dimensional space point cloud data is obtained through a multi-dimensional locator to construct an initial positioning vector database, and an installation coordinate matrix is determined in combination with the technical specification of the electromechanical equipment; a coordinate measuring arm and a laser tracking interferometer are used for measuring and calculating a space compensation variation vector and a thermal expansion variation vector to form a comprehensive variation vector. Performing computer simulation on an installation environment based on the acquired data, determining an optimal installation path, constructing a spatial region overlapping matrix and a spatial region idle matrix, performing calculation to generate a selectable region, determining an optimal installation point location by applying a Lagrange multiplier constraint optimization equation set, performing installation and adjustment, monitoring a deviation value in real time, and determining the optimal installation point location. When the deviation value is smaller than a preset threshold value, fixed installation is completed, and finally the installation precision is verified to meet the design requirement through operation testing.

The invention relates to a path planning method and device based on hierarchical reinforcement learning, equipment and a medium, and the method comprises the steps: converting any planar map into rasterization, taking the shortest path between a starting point position and an end point position as an optimization target under the constraint of avoiding the collision of an obstacle, and starting from the starting point position, selecting a plurality of sub-target positions, the method comprises the following steps of: establishing a path planning problem between a starting point position and an end point position until the end point position is reached, converting the path planning problem between the starting point position and the end point position into a plurality of local sub-path planning problems, constructing a learning heuristic network by adopting an artificial neural network as a heuristic function in an A * algorithm, and training the learning heuristic network by utilizing reinforcement learning to obtain a learning heuristic network; and solving each local sub-path planning problem by using the trained learning heuristic network to obtain an optimal solution of each local sub-path, and then obtaining a bounded suboptimal path planning result. By adopting the method, path planning can be efficiently carried out on any plane map.

The invention relates to the technical field of capacitive touch, and discloses a capacitive touch signal self-adaptive compensation method and device. The method comprises the following steps: collecting original capacitance signal grid data in an initial contact stage, and constructing a dynamic contact quadtree; on the basis of the quadtree, a first point position prediction algorithm of contact dynamics perception is operated, and stable coordinate prediction mapping with confidence is constructed; and executing multi-mode first point decision and touch event generation based on the mapping, outputting a touch event and feeding back the touch event to an operating system. According to the method, initial contact signal features are captured through the dynamic contact quadtree, stable coordinate prediction mapping and multi-modal decision are combined, the problem of jumping coordinates in the initial touch stage of an existing system is solved, the touch accuracy of scenes such as vehicle-mounted center control and industrial HMI is improved, the operation risk is reduced, and the application value of the capacitive touch technology is expanded.

The invention provides a multi-view high-precision hand three-dimensional reconstruction method and system fusing two-dimensional complementation and implicit optimization, complementation of an object-shielded area of a hand is completed on an image two-dimensional level through VAE-GAN, and feature deficiency caused by shielding is reduced; and in cooperation with multi-view and implicit optimization, a relatively complete and accurate hand reconstruction effect can be obtained in a complex gesture and strong shielding scene. The multi-view key point fusion and 3D skeleton point regression method effectively reduces accumulative errors caused by single-view detection deviation; the MANO parameter prior is combined with the self-evolving SDF, fine regression can be carried out on complex topologies such as the local surface of the hand, and the accuracy of three-dimensional skeleton key point positions and hand grid details is improved. Only RGB camera data with a small number of visual angles are needed, and a depth sensor or high-precision calibration is not needed; even if the calibration information is rough, high precision can be achieved through the complementation and data-driven 3D regression method, and the system deployment cost is reduced.

The invention relates to the technical field of intelligent traffic, in particular to a training method and system for a vehicle track frame filling model, and the method comprises the steps: training a pre-trained Transform network based on a plurality of pieces of sheltered historical track data, and obtaining a trained vehicle track frame filling model; the training process comprises the following steps: obtaining a prediction result corresponding to each shielded time step; obtaining map constraint loss based on prediction results of all shielded time steps and a preset tunnel center point position matrix; obtaining position loss, speed loss and course angle loss based on the time step data of the prediction result corresponding to all the shielded time steps and the time step data in the historical trajectory data; based on the map constraint loss, the position loss, the speed loss and the course angle loss, model parameters are adjusted, and a vehicle track frame compensation model is obtained. The method has the beneficial effect that the frame supplementing track of the vehicle is more reasonable, smoother and smoother.