1132results about How to "Achieve integration" patented technology

The invention relates to the technical field of image classification, and discloses a breast image classification evaluation method and system based on deep learning, and the method comprises the steps: carrying out the preprocessing of sample breast image data, and obtaining comprehensive feature data; performing feature enhancement based on the comprehensive feature data to obtain enhanced feature data; performing multi-modal bidirectional feature fusion on the enhanced feature data to obtain multi-modal fusion feature data; based on the multi-modal fusion feature data, constructing and training a semi-supervised local feature reservation dense connection context-aware graph convolutional network to obtain a graph volume integral class model; according to the multi-modal fusion feature data, constructing and training a multi-branch residual attention reconstruction network to obtain a multi-scale feature fusion model; through the image volume integral class model and the multi-scale feature fusion model, classification evaluation and prediction result integration are carried out on the real-time mammary gland image data to obtain a target mammary gland image classification result, and the method improves the accuracy and reliability of mammary gland image classification.

The invention discloses a power grid data management system and method based on artificial intelligence data analysis in the technical field of power grid anomaly detection. The method comprises the steps of obtaining operation data of nodes in a power grid system; calculating an exception comprehensive score of the node based on the operation data by combining an exception detection model deployed by edge computing equipment; determining an abnormal response level of the node based on the abnormal comprehensive score; and triggering a corresponding response strategy based on the exception response level of the node, and performing exception processing on the node. According to the method, the lightweight anomaly detection model is deployed on the edge computing equipment, the power grid operation data is monitored in real time, the abnormal condition is quickly identified, the reliability and safety of power grid operation are improved, and the problems of response delay, network bandwidth limitation, single-point fault and the like of a traditional centralized anomaly detection system are solved.

The invention provides an intelligent networked automobile chassis drive-by-wire simulation method and system based on multi-domain modeling, and relates to the technical field of adaptive control, and the method comprises the steps: obtaining structural parameters and kinetic parameters of an intelligent networked automobile chassis, constructing a finite element grid, building a multi-flexible-body kinetic model corresponding to the intelligent networked automobile chassis, and constructing a multi-flexible-body dynamic model corresponding to the intelligent networked automobile chassis; introducing a coupling coordination equation, establishing a multi-flexible-body kinetic equation, and solving to obtain vibration response and stress-strain distribution under different working conditions; generating a plurality of finite element models, performing multi-physical field fusion and multi-error source fusion models, performing online error identification and active compensation, performing all-condition drive-by-wire chassis closed-loop control, and generating a drive-by-wire chassis closed-loop integrated simulation model; constructing a road three-dimensional parameterized model, generating a vehicle-road coupling model, generating a simulation test case, identifying a performance short board, constructing a knowledge graph and an optimization rule base, and generating an optimal control strategy.

The invention discloses an emotion analysis method and system based on multi-modal fusion. The method comprises the following steps: S1, acquiring multi-modal data and initial feature representation of each modal; s2, inputting the initial feature representation of each mode into a multi-head attention mechanism for preliminary fusion, and generating a primary fusion feature; s3, in combination with the bidirectional long-short-term memory network and the multi-scale causal convolutional network, generating time sequence features; s4, constructing a hierarchical dependency relationship network among modals by using a multilayer graph neural network, and forming global context sensing feature representation; s5, dynamically generating and adjusting the fusion weight of each mode; s6, combining classification and regression output by using a multi-layer full-connection network and a multi-task learning framework, and generating multi-dimensional expression of an emotional state; and S7, migrating the group sentiment model to specific sentiment analysis of individual users through transfer learning and personalized modeling technologies. According to the method, high-precision and multi-dimensional sentiment analysis is realized by utilizing multi-modal fusion and self-adaptive weight adjustment.

The invention discloses a multi-source heterogeneous data fusion processing and intelligent analysis method, device and equipment, and relates to the field of data fusion processing.The method comprises the steps that firstly, data in data sources of corresponding types are collected through different collection strategies; performing data preprocessing on the collected data to obtain standard data in a unified format; performing data fusion on the standard data based on the structural features of the data to obtain a fused data set; and determining a target analysis model according to the data analysis target, and performing data analysis on the fused data set. According to the method, various types of data formats can be processed, multi-source heterogeneous data fusion is realized, and through data semantic understanding and a feature-based fusion algorithm, the accuracy and quality of data fusion are improved, and correct data association is established on the semantic level.

The invention relates to the technical field of prostate MRI (Magnetic Resonance Imaging) image segmentation, in particular to a prostate MRI image segmentation method based on Mamb-Unet, which realizes effective fusion of features of different scales between an encoder and a decoder, and fuses multi-scale information on a space axis by using a space attention bridge segment to fuse fusion features of different levels from the encoder, so that the segmentation accuracy of the prostate MRI image is improved. And then global and local information is captured through 1D convolution and full connection in the channel attention bridge segment, so that a channel attention map is generated, channel information is better integrated, and multi-level different-scale feature fusion is realized.

The invention relates to the technical field of water quality and water level monitoring, and discloses a water quality and water level monitoring method for a water conservancy project, which comprises the following steps: S1, arranging a multi-parameter water quality and water level monitoring sensor at a key position of the water conservancy project; s2, performing real-time processing on the data acquired by the multiple sensors based on a data fusion algorithm, wherein the data fusion algorithm is based on a Kalman filtering model; s3, acquiring external data through an environment sensor, and dynamically adjusting a water level and a water quality monitoring threshold value based on environment change; s4, predicting the water level and the water quality by using a deep learning algorithm based on LSTM (long short-term memory network); and S5, if the monitoring data exceed the dynamic threshold value, automatically triggering an early warning mechanism. By arranging the multi-parameter water quality sensor and the water level sensor at the key positions of the water conservancy project, the water quality and the water level can be comprehensively monitored, the coverage range and the precision of the monitoring system are greatly improved, and the complex monitoring requirements of multi-source pollution can be effectively met.

The invention discloses a knowledge and data fused chronic disease collaborative management method and system, and relates to the technical field of medical management. Comprising the following steps: generating a feature data set, processing the feature data set by using a feature gain modeling method, and generating an enhanced feature data set; dynamically generating a chronic disease comprehensive knowledge graph, performing node supplementation and weight optimization on the knowledge graph, and generating a specific knowledge sub-graph of the patient; carrying out multi-time sequence prediction by utilizing the specific knowledge sub-graph, and carrying out short-term and long-term trend prediction; a patient health track model is constructed through the state transition probability in combination with the prediction result; according to the dynamic risk labeling result of the health track model, the health state of the patient is comprehensively evaluated in combination with the specific knowledge sub-graph, and personalized management suggestions are generated according to the evaluation result. According to the collaborative management platform, cooperation between different medical institutions and medical staff can be promoted, information sharing and resource integration are realized, and the overall level of medical services is improved.

The invention provides a drought monitoring method and system based on ground feature detection, and relates to the technical field of ground feature detection, and the method comprises the steps: calling geological distribution data of a target monitoring region, carrying out the preliminary division, and obtaining a plurality of initial divided grid regions; performing sensor layout to obtain a plurality of sensor layout arrays; connecting a satellite remote sensing platform to a preset detection window to obtain a plurality of remote sensing images; collecting ground feature data to obtain a plurality of ground feature data sets; secondary similar aggregation is carried out, and a plurality of aggregation division grid region sets are determined; and determining a plurality of first monitoring bandwidths and a plurality of second monitoring bandwidths, and constructing a plurality of first monitoring channels and a plurality of second monitoring channels of the drought monitor for drought monitoring. The technical problems that a drought monitoring method in the prior art generally depends on a static monitoring strategy, and the monitoring strategy cannot be automatically adjusted according to the real-time state of drought change, so that the monitoring efficiency is low, and the accuracy is poor are solved.

The invention provides a multi-modal sentiment analysis method based on unwrapping representation learning, which comprises the following steps: acquiring a multi-modal sentiment analysis data set, and extracting features of texts, audios and videos; constructing a representation learning network, extracting private features and shared features, and processing through adversarial learning optimization and time smoothness constraint; under the guidance of different language scales, the private features of the audio and the video are fused; further fusing text, audio and video features through cross-modal attention and a gating mechanism; and finally, inputting the fusion features into an emotion analysis module, and generating an emotion prediction result in combination with an emotion classification or regression task. According to the method, the consistency of the audio and video features in continuous time is considered, redundant information is effectively reduced, the time dimension consistency is ensured, deep fusion of the audio and video features is realized under text guidance, and the cross-modal fusion effect is remarkably improved.

The invention discloses a radiology report generation method based on semantic alignment, and belongs to the technical field of radiology report generation. The radiology report generation network constructed by the invention comprises a multi-scale feature extraction module, a cross-modal semantic alignment module and a Transform report generation module which are connected in sequence; the multi-scale feature extraction module comprises a convolutional neural network and a multi-scale local sparse attention module which are connected in sequence; the cross-modal semantic alignment module comprises an embedded layer, a multi-head attention module and a gating fusion module, wherein the multi-head attention module and the gating fusion module are connected with the embedded layer. The input end of the multi-head attention module and the input end of the gating fusion module are connected with the output end of the multi-scale local sparse attention module. The predicted radiology report generated by the radiology report generation method is good in quality.

The invention discloses an automatic fine ore sorting method and system based on XRT (X ray tomography) and hyperspectrum, which are used for realizing comprehensive fusion of ore surface and internal information by combining X-ray tomography and hyperspectral imaging technologies so as to improve the recognition precision and the sorting efficiency. XRT is used for extracting the internal structure and density characteristics of the ore, hyperspectral imaging is used for identifying the mineral composition characteristics of the ore surface, and information complementation is achieved. The system adopts a multi-modal convolutional neural network to perform independent feature extraction on XRT and hyperspectral data, and the extracted features are fused through a deep residual network. And the fused features are processed by a cross-modal attention mechanism to ensure that key features of different modals are fully integrated and optimized, and finally high-precision ore classification and sorting are realized. According to the method, the limitation of a single data source on feature extraction is overcome, the ore recognition and sorting precision is greatly improved, and the refined and intelligent development of ore sorting is promoted.

The invention relates to the technical field of electric power meteorological prediction, and aims to solve the problem that a current power grid natural disaster early warning system has defects in feature extraction and cannot fully utilize knowledge in related fields to enhance the prediction capability of a model. The invention discloses a power grid natural disaster early warning method based on a knowledge graph, computer equipment, a computer readable storage medium and a computer program product. The method comprises the following steps: S1, constructing a power meteorological knowledge graph based on meteorological data, geographic information and power equipment data corresponding to a power grid natural disaster; s2, performing feature enhancement on the meteorological data according to semantic features in the knowledge graph to form fusion features; s3, according to the fusion features, constructing a disaster weather prediction model based on a graph convolutional neural network and a self-attention mechanism; and S4, according to the disaster weather prediction model, performing early warning on the natural disasters of the power grid in the target weather. By adopting the method, the accuracy of power grid natural disaster prediction can be improved.

The invention relates to a multi-modal implicit sentiment analysis method based on a multi-granularity attention mechanism, which belongs to the field of artificial intelligence, and comprises the following steps: S1, extracting visual features in image information by using a visual feature extraction strategy of multi-scale attention fusion; s2, generating a text description of the image information by adopting an image translation engine, splicing the text description with an original text, and extracting text features through a hierarchical attention mechanism; s3, carrying out normalization processing; s4, performing multi-modal fine-grained correlation fusion on the visual features and the text features by adopting a bidirectional cross attention mechanism, and extracting fine-grained features in the visual features and the text features through a multi-head attention mechanism and residual connection; and S5, fusing the fine-grained features by using an adaptive multi-modal classification module to complete sentiment classification. The multi-modal sentiment analysis method solves the problems that multi-modal sentiment analysis is insufficient in inter-modal correlation mining, insufficient in fine-grained information capture and large in noise interference.

The invention discloses a rock burst prediction large model construction method based on multi-modal data, and the method comprises the steps: constructing a multi-modal data set through collecting data of different modals; preprocessing the multi-modal data set, and constructing a precursor mode sequence of model training; according to the characteristics of different mining areas, the precursor mode sequence is converted into a corresponding grade form, and corresponding rock burst danger grade labels are divided; transform is adopted as a core framework, the precursor mode sequence after grade division is processed, and finally prediction of the rock burst grade occurrence probability is achieved. A comprehensive index method is used for independently evaluating the danger degree of the mining information data and the geological structure data, and the overall danger level of the rock burst is comprehensively evaluated in combination with a danger probability result output by the rock burst prediction module; according to the method, the applicability and prediction precision of the model under different mining area conditions can be improved, and accurate prediction of the rock burst risk is realized.

The invention provides a GraphRAG-based large model question and answer method, system and device, a medium and a product, and the method comprises the steps: determining an available data source, and carrying out the preprocessing of data; performing knowledge extraction, knowledge fusion and knowledge calculation on the preprocessed data to construct a knowledge graph; key entities are extracted from user query through a large language model of GraphRAG; retrieving related sub-graphs in the knowledge graph according to the extracted key entities to form context information; and sending the context information into a large language model of the GraphRAG to perform a generative answer. According to the invention, more related knowledge points and information sources can be covered, and more comprehensive answers can be provided; for complex queries needing multi-step reasoning or multi-document information integration, the accuracy is high; when a large-scale data set is processed, the computing power resource demand is relatively low; cross-domain integration and reasoning can be realized; the diversity of answers is improved, and the weighting of the answers is remarkably enhanced.

The invention discloses an infrared small target detection method based on multi-scale self-attention motion background modeling, and the method comprises the steps: selecting a plurality of continuous frames from an infrared sequence image as model input, and extracting image features through a backbone network; fully mining the multi-scale global features of the network through a cross-scale self-attention coding structure to model a global background of a time sequence image; grouping and weighting the shallowest layer features containing the finest granularity information to form a grouping query, and forming a final query by adopting a cross-scale iterative decoding mode, so as to reserve more target details and more sufficiently suppress a motion background; a confidence map is generated through the detection head, and a final detection result is obtained through a threshold segmentation method. The multi-scale self-attention coding network can realize fusion of sequence image background global features and target local features, and the multi-scale self-attention decoding network can enhance multi-frame detail information of query vectors in the decoding process and improve the capabilities of motion background modeling and target feature expression.

The invention discloses an electroencephalogram signal feature extraction method and system based on self-adaptive time-frequency feature collaboration, relates to the field of electroencephalogram signal processing, and solves the problems that in existing electroencephalogram signal feature extraction, fusion is carried out in a simple splicing or averaging mode, complementarity and correlation between time domain and frequency domain features are not fully considered, and time domain and frequency domain features are not fully considered. In order to solve the problems of insufficient overall characterization capability and classification performance of signals and the like, the method comprises the following steps: acquiring electroencephalogram signal data, preprocessing the electroencephalogram signal data, constructing a dynamic residual network in combination with dynamic convolution, and training a model based on the preprocessed electroencephalogram signal data; performing a contrast test on the dynamic residual network and other deep learning network models, and designing an adaptive time-frequency collaborative network model to fuse time-domain features and frequency-domain features in the model; and electroencephalogram signal feature extraction based on adaptive time-frequency feature cooperation is completed. The method is also suitable for the field of electroencephalogram signal feature extraction and fusion.

The invention discloses a multi-mode breast ultrasonic diagnosis system based on a dynamic feature fusion network, belongs to the field of medical image processing, and particularly relates to the multi-mode breast ultrasonic diagnosis system based on the dynamic feature fusion network. The invention aims to solve the problem that the existing method is low in breast ultrasonic recognition accuracy. The system comprises a multi-modal data preprocessing main module used for preprocessing a collected multi-modal mammary gland ultrasound image to obtain a preprocessed multi-modal mammary gland ultrasound image; the multi-modal convolutional neural network main module is used for building a dynamic feature fusion network model; the model training main module is used for training an established dynamic feature fusion network model by using the preprocessed multi-modal breast ultrasound image to obtain a trained dynamic feature fusion network model; and the model test main module is used for loading the trained dynamic feature fusion network model, classifying a to-be-diagnosed multi-modal breast ultrasound image, and identifying benign and malignant breast tumors in the multi-modal ultrasound image.

The invention relates to the technical field of power supply chains, and discloses a power supply chain risk early warning method driven by a knowledge graph and a related device. The method comprises the following steps: acquiring multi-source data of a power supply chain; the multi-source data comprises structured data and unstructured data; inputting the multi-source data into a risk early warning model, and outputting a risk probability; wherein the risk early warning model takes a multi-modal graph neural network as a backbone network and combines a graph attention mechanism and comparative learning to realize fusion and risk prediction early warning of multi-source data. According to the invention, the advanced knowledge graph technology and the deep learning algorithm are utilized to realize intelligent risk early warning of the power supply chain; according to the power supply chain risk early warning method driven by the efficient knowledge graph, the risk management level of the power supply chain can be effectively improved, and a powerful guarantee is provided for stable operation of the power industry.

The invention provides a remote sensing image target detection method based on an improved FCOS, and relates to the technical field of target detection, and the method comprises the following steps: carrying out the convolution and pooling operation of a remote sensing target image based on a pre-trained convolutional neural network, generating a preliminary feature, and carrying out the enhancement processing of the preliminary feature through a twin multilayer perceptron; carrying out fusion processing on different levels of feature information in the multi-level feature map through a feature map splicing mode and a feature map enhancement mode; and determining the position of the remote sensing image target according to the analysis result to obtain a target detection result. According to the method, the feature pyramid network is combined with an innovative coordinate attention mechanism, fusion of high-level semantic information and low-level detail information is realized, an enhanced multi-scale feature map is generated to cope with diversity of target scales in a remote sensing image, multi-scale fusion and an anchor-frame-free design are combined, false detection is effectively reduced, and the detection accuracy is improved. Therefore, the method shows efficient, accurate and wide application potential in the field of remote sensing image target detection.

The invention provides a method and a system for intelligently generating a complete dispatch report for financial unfavorable assets, and relates to the technical field of data investigation and arrangement, comprising Sp1, complete dispatch demand acquisition, Sp2, intelligent reasoning planning, Sp3, data acquisition and processing, Sp4, public opinion retrieval and analysis, Sp5, legal information extraction and verification, Sp6, data integration and report generation, and Sp7, report modification and output, and the method comprises the steps of Sp1, complete dispatch demand acquisition, Sp2, intelligent reasoning planning, Sp3, data acquisition and processing, Sp4, public opinion retrieval and analysis, Sp5, legal information extraction and verification, Sp6, data integration and report generation, and Sp7, report modification and output. By integrating a large model, an online retrieval tool, a data table tool and a key point extraction tool, automation and intellectualization of a complete dispatching process are realized. The system can automatically acquire, process and integrate data, greatly improves the efficiency of data acquisition, analysis and report generation, and reduces the probability of manual intervention and errors.

The invention provides an intention guide pushing method and system based on geographic position information and multimedia interaction. The method comprises the following steps: constructing an interest model based on a geofence technology and a deep neural network; generating a user activity intention prediction result based on the interest model, obtaining a behavioral rhythm analysis result by using a Bayesian network and a time sequence prediction algorithm based on the user activity intention prediction result, and constructing a user and place relation graph by using a graph theory algorithm based on determining the optimal pushing opportunity of the behavioral rhythm analysis result. Generating a time sensitivity optimization scheme; generating customized content based on a time sensitivity optimization scheme, processing the customized content by adopting an emotion analysis method, a collaborative filtering algorithm, an auto-encoder and a multi-head attention mechanism to obtain final content, adjusting the final content through an instant feedback mechanism, and generating a multimedia push packet; according to the invention, the performance and service quality of the intelligent pushing system are improved, and more accurate and personalized recommendation experience is provided for the user.

The invention discloses an intelligent question answering system and method for cross-domain knowledge reasoning, and relates to the technical field of intelligent question answering. In order to solve the problem that an existing intelligent question answering system is limited by technical shortages and cannot give satisfactory reply to a user in cross-domain question processing, the adopted scheme comprises the steps that a domain knowledge base construction module collects multi-domain knowledge data, and the multi-domain knowledge data is stored in a corresponding domain vector knowledge base after being preprocessed and vectorized; the user interaction module provides an interface, receives input, interacts and returns a result; the field identification module identifies the field and extracts key information by using a large language model; the knowledge retrieval module retrieves related information accordingly; the cross-domain knowledge reasoning module is used for reasoning a knowledge logic relationship of each domain and guiding a large language model to integrate fused knowledge; and the intelligent question-answering module generates prompt words, inputs the prompt words into the large language model to obtain a final response, and returns each result to the user interaction module for display. According to the method, knowledge from multiple fields can be processed, and the method is suitable for various different types of problems and scenes.

The invention relates to the technical field of image data processing, in particular to a wood composite board furniture surface defect identification method and system. And wood board images are collected from the surface of the wood composite board furniture. Adaptive grid division is carried out on the wood board image according to different scales, texture direction information and high and low frequency texture components are extracted, and a texture feature map is obtained; and extracting structure edge information of the wood board image to obtain a structure information graph. Matching the texture feature map with the structure information map, and calculating local mutual information of the texture feature map and the structure information map in each region; the method comprises the following steps: acquiring resolution images of a wood board image under different scales, and calculating a mutual information matrix of the resolution images; distributing an abnormal score for each matrix element based on the mutual information matrix; and dividing grids according to the abnormal score, calculating a correlation value of adjacent grids, and obtaining a defect area based on the correlation value. And carrying out defect classification on the defect area. According to the invention, interference of natural textures can be reduced, and defect identification of small defects and tiny changes can be improved.

The invention relates to the technical field of smart city management, in particular to a city physical examination spatio-temporal data dynamic aggregation and decision support system which comprises a data acquisition module, a data preprocessing module, a spatio-temporal data aggregation module, an index calculation module, a dynamic analysis module and a decision support module. Wherein the data acquisition module is used for acquiring multi-source heterogeneous original city operation data; the spatio-temporal data aggregation module is used for aggregating the data according to a specified time window and a specified space unit; the index calculation module is used for extracting a core index reflecting a city health state; the dynamic analysis module is used for predicting the future change trend of the core indexes; and the decision support module is used for generating an urban governance strategy optimization scheme. According to the method, through dynamic integration and space-time correlation analysis of multi-source heterogeneous data, the urban operation trend is accurately predicted, a targeted treatment optimization scheme is generated, and the comprehensiveness of urban physical examination and the scientificity of decision support are remarkably improved.

The invention belongs to the technical field of three-dimensional simulation, and particularly provides a rock mass structural plane network simulation method fusing artificial intelligence and laser scanning, and the method comprises the steps: obtaining point cloud data; clustering the point cloud data by adopting a clustering algorithm to obtain an original structural surface; performing statistical analysis on the original structural plane according to the rock mass type and the structural partition to obtain a representative structural plane, and fitting to obtain a probability density function according to occurrence parameters of the representative structural plane; according to the development volume density of the representative structural plane, converting the probability density function into a volume density function, and according to the volume density function, generating a random structural plane by using a Monte Carlo simulation method; and constructing a BIM model of the rock mass structure, and performing three-dimensional space simulation on the rock mass structural plane network based on the BIM model and the random structural plane to obtain a simulation result. According to the method, the precision, comprehensiveness and authenticity of depicting the rock mass structural plane are greatly improved, and the spatial distribution characteristics of the structural plane network can be quantitatively evaluated.

The invention relates to an edge fusion method and system of an XR immersion type LED display screen. The method comprises the following steps: carrying out feature extraction on image data of an XR immersive LED display screen to obtain an RGB three-channel spatial domain feature matrix and a frequency domain edge feature matrix; performing feature processing to obtain a multi-core context feature map; performing coding processing to obtain a position-brightness comprehensive coding feature matrix; performing multi-scale decomposition to obtain a multi-scale detail feature map; carrying out pixel clustering on the edge region of the XR immersive LED display screen to obtain an edge topological structure diagram and a brightness transition function; and performing dynamic matching on the display frame rate through a PID controller, and performing brightness compensation and color correction on the edge region to obtain edge fusion data. The method has the multi-scale feature extraction capability, the edge fusion problem under different scenes can be adaptively processed, and it is ensured that processing delay meets the real-time display requirement, so that the overall performance of an XR immersive display system is improved.

The invention provides a network node classification method based on a multilayer graph neural network model, and the method comprises the steps: constructing a graph according to network nodes, carrying out the preprocessing of initial node features in the graph, and obtaining an initial embedded vector; according to the initial embedding vectors of the node and the neighbor node thereof, updating the node representation by adopting a layered message passing mechanism; according to the initial embedded vectors of the nodes and the neighbor nodes thereof, the graph attention network uses an attention mechanism to calculate weights for the neighbor nodes of each node and then dynamically updates feature vectors of the nodes; a multi-head attention mechanism is adopted to splice and merge the feature vectors of the nodes output by the graph attention network and the updated node representation, and the comprehensive feature vector of each node is obtained; the comprehensive feature vectors are further optimized through graph contrast learning and a node embedding module, and optimized features are obtained; and obtaining predicted node classification according to the optimization features. According to the method, complex features of heterogeneous graph data can be effectively captured, the method has good robustness for an over-smoothing problem, more accurate heterogeneous network representation learning is realized, and the method has advanced performance and robustness in a node classification task.

The invention belongs to the technical field of marine surveying and mapping, and discloses an unmanned aerial vehicle laser / shipborne multi-beam data land and water topographic mapping method, which comprises the following steps of: coarsely extracting a data overlapping region through a virtual grid, finely extracting the overlapping region by using an alpha-shape algorithm, and obtaining boundary points; removing edge triangles in an irregular triangulation network constructed by the unmanned aerial vehicle laser sounding data and the shipborne multi-beam sounding data; calculating a vertex normal vector, taking a vertex tangent plane of the multi-beam triangular irregular network as a registration element, carrying out weight distribution according to the distance from a point to the vertex, and carrying out coarse registration through a point-to-plane ICP registration algorithm; and constructing a precise registration model based on a point-surface distance, and eliminating the elevation difference of the overlapping region by minimizing the normal distance from the unmanned aerial vehicle laser sounding point to the multi-beam sounding triangular surface element. By adopting the point cloud fusion model for processing, the phenomenon that different source point cloud elevations in the overlapping area are inconsistent is effectively improved, and underwater topographic data fusion is realized.