718 results about "Spatial structure" patented technology

The invention provides an intelligent real-time detection method and system for PCCP welding quality, and relates to the technical field of online detection and intelligent evaluation of pipeline welding quality through machine learning. Light energy data and multi-light-source images of a spiral weld pool are collected, exposure parameters are dynamically adjusted through the energy difference of visible light near-infrared bands, and the real-time detection of the PCCP welding quality is achieved. Inhibiting strong light interference and generating a weld surface image; a stress concentration area is positioned by scanning a welding seam thermal deformation area and combining speckle pattern change, sound frequency change and the elastic characteristic of the thin-wall steel cylinder; inputting the surface image and the deformation data into a space-time convolutional neural network, fusing light energy change, image details and spatial features to construct a weld joint space structure diagram, and adaptively correcting the position of a sensor; and comparing the sinking depth of the three-dimensional point cloud reconstruction, analyzing the correlation between the sinking degree and the stress, and generating a probability thermodynamic diagram to output the pressure-bearing failure risk level, so that the probabilistic early warning of the pressure-bearing failure risk can be realized.

The invention discloses a space-time spectrum combined super-resolution reconstruction method based on a giant remote sensing satellite group. The method comprises the following steps: acquiring time series, multi-view and multi-spectral data of the same area from a plurality of heterogeneous satellites, and performing radiometric calibration and atmospheric correction; sub-pixel-level alignment of the multi-source data is realized by adopting a joint registration model; extracting time change features by using three-dimensional convolution, extracting space structure and texture features by using two-dimensional convolution, and extracting and reducing the dimension of spectral features by using one-dimensional convolution; performing adaptive weighted fusion on time, space and spectral features through an attention mechanism to generate a joint feature tensor; and carrying out super-resolution reconstruction to obtain a target image with high spatial resolution, high time resolution and high spectral fidelity. The method gives consideration to both resolution improvement and spectrum authenticity, and is suitable for high-precision remote sensing application scenes such as fine urban mapping, agricultural monitoring, ecological environment assessment, disaster emergency and battlefield situation awareness, remote reconnaissance, target change detection and damage assessment.

The invention discloses a distribution line unmanned aerial vehicle edge end real-time visual identification and target detection method and system based on lightweight AI. The method is based on a YOLOv8 architecture, and constructs a complete lightweight detection framework comprising a feature extractor, an enhancement module and a simplified detection head by introducing a space structure maintaining assembly, a separated large kernel convolution and a weighted reconstruction feature pyramid. A cross-dimension semantic relation model is constructed by utilizing a combined attention structure, multi-level feature fusion is realized by reconstructing a feature pyramid network, model training is performed by adopting a hybrid optimization function and a dynamic sample adjustment mechanism, and the model is deployed on a mobile computing platform after being optimized by a hierarchical knowledge transfer method. According to the method, the detection speed is remarkably increased while high precision is kept, and real-time identification and anomaly analysis of the power line element are effectively realized.

The invention belongs to the technical field of display design evaluation, and particularly relates to an exhibition display layout design effect analysis and evaluation method, which comprises the following steps of: constructing a dynamic spatial topology network by collecting audience moving tracks, exhibit information and spatial structure data in real time; the problem of insufficient attention on space utilization influenced by dynamic factors such as human traffic fluctuation is solved. And dynamic changes in the space are converted into quantifiable network nodes and connection relationships, so that the space utilization state is changed from static design to dynamic monitoring, accurate capture of the real-time use condition of the exhibition space is realized, a comprehensive and real-time data basis is provided for subsequent analysis, and the real-time monitoring of the exhibition space is realized. Meanwhile, the dynamic space topology network is analyzed by means of spatial-temporal feature fusion to determine the edge weight and the exhibition area correlation degree, the spatial efficiency index and the dynamic line rationality index are generated in combination with an analytic hierarchy process and multi-dimensional analysis, and the defect that a space utilization rate core evaluation index is not established in the prior art is overcome.

The invention relates to the technical field of medical image analysis, computer vision and deep learning, in particular to a texture perception and multi-scale feature adaptive fusion hip joint osteophyte detection method. According to the method, firstly, a bone texture feature extraction module is used for carrying out feature extraction and enhancement on a hip joint X-ray image, a parallel double-branch structure is adopted, gradient and space structure features are extracted through a Sobel operator and pooling operation, and feature maps are fused; then, the fused features are input into a double-backbone network, the first backbone network extracts local fine textures and long-range dependence features by combining convolution, self-attention and a gating mechanism, and the second backbone network extracts multi-scale semantic features through hierarchical grouping and depth separable convolution; and double-trunk output is dynamically weighted and fused through an adaptive fusion module, features are optimized through a multi-scale semantic fusion module, and finally the position and confidence of osteophyte are output.

The invention provides a digital twin three-dimensional model construction method based on a generative large model, and the method comprises the steps: obtaining multi-source monitoring data of a power distribution network, and processing the multi-source monitoring data into a training data set; the method comprises the following steps: mapping multi-source monitoring data into a multi-scale tensor subspace through tensor wavelet structured transformation, adaptively extracting spatial features through a learnable wavelet kernel, and keeping the structural continuity of a physical field in combination with a geometric prior regular term; constructing and training a generative adversarial network through a training data set; inputting and analyzing the physical parameter vector of the target scene, and if the topological similarity score is lower than a preset threshold value, adjusting noise vector regeneration; and if yes, outputting a three-dimensional model tensor and importing the three-dimensional model tensor into a digital twin platform, and driving real-time physical field visualization. According to the method, characteristics of a multi-scale space structure and a nonlinear physical field can be reserved, the physical rationality and generalization ability of the generated model are remarkably improved, and depth identification of topological attributes (such as hole connectivity and surface defects) and local geometric defects of the three-dimensional model is realized.

The invention provides a cloud condition adaptive geostationary satellite sea surface temperature inversion method based on deep learning, and relates to the technical field of sea surface temperature inversion, and the method specifically comprises the steps: obtaining geostationary satellite observation data, geographic information data, polar orbit satellite sea surface temperature and cloud products, and analyzing the sea surface temperature and atmospheric background data; performing standardization preprocessing on the obtained stationary satellite observation data; constructing a cloud detection data set; a cloud detection model of deep learning is constructed and trained; constructing a clear sky data set; a clear sky inversion model of deep learning is constructed and trained; constructing a data set for under-cloud inversion model training; constructing an under-cloud inversion model of deep learning and training the under-cloud inversion model; and integrating the cloud detection model, the clear sky inversion model and the under-cloud inversion model, adaptively calling the corresponding model according to a real-time cloud detection result, and generating a final sea surface temperature product. According to the technical scheme, the problem that in the prior art, the accuracy of a reconstructed product on the spatial structure cannot be comprehensively reflected is solved.

The invention discloses a rolling bearing vibration signal multi-mode fault classification method. The method comprises the following steps: collecting a vibration time sequence signal of a rolling bearing; the vibration time sequence signals are input into a time sequence branch network and a space branch network in parallel, and the time sequence branch network extracts time sequence dependence characteristics of the signals through a one-dimensional convolutional neural network and a bidirectional gating circulation unit; the spatial branch network converts the vibration time sequence signal into a Markov transform field image, and extracts spatial structure features of the image by using a two-dimensional convolutional neural network and a window Transform-based visual network; performing bidirectional interaction and weighted fusion on the time sequence features and the spatial features through a cross-modal attention mechanism to obtain fusion features; and inputting the fusion features into a classifier, and outputting a fault classification result of the rolling bearing. According to the method, the problems that a traditional single-mode fault diagnosis model is insufficient in adaptability to complex working conditions, multi-mode feature fusion is insufficient, and the generalization ability is weak due to model structure redundancy are solved.

The invention discloses an old city boundary intelligent extraction and texture calculation system based on vector topology analysis and semantic segmentation. According to the method, the topological semantic association module provides a multi-dimensional association basis for boundary extraction by integrating spatial structure information of vector topology and functional attribute tags of semantic segmentation. In the coarse extraction stage, the objective space law of the topological features and the subjective function orientation of the semantic tags are mutually verified, so that the misjudgment of'similar forms but inconsistent functions' possibly caused by single dependence on the topological structure or the deviation of'tag coverage but space breakage 'possibly caused by only dependence on the semantic tags is avoided; in the fine correction stage, the module further dynamically calibrates the boundary range through historical-current situation space-time correlation and multi-source data conflict detection, so that the continuity of historical stable elements is reserved, the updating requirement of current situation semantic tags is included, the final boundary better fits the essential characteristics of'active inheritance 'of the old city, and the accuracy of the final boundary is improved. And human experience interference and errors are obviously reduced.

The invention provides a panchromatic sharpening joint optimization method and system based on three-stage progressive fusion. The method comprises three stages of coarse fusion, deblurring detail enhancement and fine fusion. The method comprises the following steps: firstly, performing feature extraction and preliminary fusion on a low-resolution multispectral image and a high-resolution panchromatic image through a dual-path mutual enhancement network; secondly, a multi-layer stacked deblurring module is introduced to perform deep enhancement on fusion features, and the details and definition of the image are effectively improved in combination with partial large kernel convolution, channel mixing and an element-level attention mechanism; and finally, realizing fine optimization of spectrum consistency and a space structure, and outputting a fused image with high spatial resolution and high spectral fidelity. According to the method, the stability and reconstruction quality of panchromatic sharpening under a complex degradation condition are effectively improved through multi-source remote sensing image collaborative modeling and progressive feature fusion, and the method has good practical value and popularization prospect and is superior to a current panchromatic sharpening method based on deep learning.

The invention provides a mine slope risk assessment method and system based on multi-source data space-time fusion, and relates to the technical field of mine monitoring, and the method comprises the steps: obtaining geological and engineering data, and constructing a basic model representing a slope three-dimensional space structure; collecting multi-source heterogeneous monitoring data on a mine slope field, and aligning and fusing the multi-source heterogeneous monitoring data with the basic model in a space-time dimension to generate a continuous dynamic parameter field; and setting a data fusion model, and taking the basic model and the dynamic parameter field as input to obtain a four-dimensional space-time fusion risk field. According to the method, the basic model representing the three-dimensional space structure of the slope is constructed, and a unified space-time bearing base is provided for all heterogeneous monitoring data, so that originally isolated data such as GNSS displacement, deep displacement, blasting vibration and underground water level can be fused in a unified coordinate system; the problem of data splitting in the prior art is fundamentally solved, and integral and integrated expression of slope engineering and geological conditions is formed.

The invention discloses a remote sensing image segmentation method and system based on cross-normal-form feature fusion and alignment, and the method comprises the steps: carrying out the preprocessing of an input remote sensing image, and extracting an initial feature; inputting a cross-normal-form feature fusion and alignment network, and fusing multi-modal and cross-scale remote sensing image structure information through sparse channel enhancement and space alignment and space pixel refining and channel alignment to obtain a first-stage fusion feature; inputting a multi-stage cross-paradigm enhanced feature extraction network, fusing local details and global context information through multi-level information interaction and a dynamic gating mechanism, and gradually extracting a joint feature map of semantic and spatial structure collaborative expression; a final semantic segmentation result is generated through the segmentation head, and composite loss is calculated based on a real label; according to the method, the multi-stage feature extraction network and a cross-normal-form feature alignment mechanism are constructed, local textures, spatial contexts and multi-modal information are effectively fused, and the segmentation performance is enhanced while the calculation efficiency is guaranteed.

The invention discloses a high-end space light illumination control system with scene adaptability, and relates to the technical field of intelligent light control, and the high-end space light illumination control system is characterized in that spatial structure geometric information and surface material distribution are acquired, and an illumination path simulation model is constructed; in combination with a preset illumination demand template, identifying a position with illumination deviation; performing illumination coverage simulation on each position point based on the lamp type and the light emitting angle, calculating an illumination coverage contribution value, and constructing a reference matrix; according to a matrix result, lamp model selection and angle adjustment are carried out, partition control operation is executed, and dynamic optimization adjustment of the brightness, the angle and the color temperature is achieved; closed-loop correction is completed through illumination feedback; according to the invention, the adaptive ability of illumination control to a complex space structure and a dynamic scene is improved, higher control precision and energy efficiency are achieved, and the method is suitable for intelligent illumination systems in scenes such as exhibition display, commercial retail and high-end residence.

The invention discloses a manifold perception-based large model navigation method guided by a historical topological graph, and relates to a computer vision technology. The method aims at solving the challenges that in the navigation process, long-distance reasoning experience is insufficient, instruction fragments and dynamic visual observation are difficult to align, and large model reasoning is prone to illusion interference. Firstly, a large model based on an encoder-decoder structure is used for supplementing historical information coding for a visual observation sequence, and therefore global topological information guidance is provided for long-distance reasoning. And secondly, in order to effectively solve the problem that large model reasoning is subjected to illusion interference, significant space-time differences in a visual observation sequence are mined by using a multi-curvature manifold, so that the large model can accurately describe the current environment and make a decision according to a visual reference object in thinking. Besides, in order to strengthen the perception capability of the large model to the space structure and establish a graph self-attention mechanism, the node distance embedded in the constructed historical topological graph is combined with the visual similarity so as to model the space relationship between the nodes.

The invention provides a category-level 6D attitude estimation method based on geometric perception key points, and relates to the technical field of attitude estimation.The method comprises the steps that on the basis that RGB images and point cloud features are fused, a dynamic key point proposing module is designed to generate key points which are consistent in category and geometrically adaptive, so that the adaptability to intra-class deformation and weak texture targets is enhanced; secondly, introducing a spatial geometric attention module to model a spatial structure relationship between key points so as to optimize key point distribution; and finally, point cloud reconstruction and scale regression are simultaneously realized through a geometric perception reconstruction and scale estimation module under the prior condition of no CAD model, and end-to-end optimization is carried out by utilizing multi-task loss. Experimental results on REAL275, CAMERA25 and House Cat6D data sets show that the method provided by the invention is superior to the existing method in multiple indexes such as IoU and attitude precision, and shows stronger generalization and robustness in a real complex scene.

The invention provides a three-dimensional point cloud reconstruction method and device based on a graph cross attention network. The method comprises the steps that three-dimensional point cloud data are acquired and preprocessed; constructing a polyhedral graph structure according to the preprocessed three-dimensional point cloud data; extracting point cloud features of the preprocessed three-dimensional point cloud data by using a convolution encoder, projecting the point cloud features to three orthogonal view angle planes of XY, XZ and YZ to generate potential features, and sampling from the interior of a polyhedron by using a skeleton sampling algorithm to obtain query points; interpolating the query points on the potential features by adopting a bilinear interpolation method to obtain polyhedral features; a graph cross attention network is adopted to fuse spatial structure modeling of graph convolution and dynamic feature fusion capability of cross attention, a connection relation of a polyhedral graph structure is dynamically adjusted according to polyhedral features to obtain graph node features, and three-dimensional point cloud reconstruction is executed according to the graph node features; therefore, the robustness and the precision in complex part topology reconstruction are improved.

The invention relates to the technical field of remote sensing data analysis, in particular to a land resource management data analysis system. According to the method, the spectral reflectivity change under the earth surface multi-direction illumination is obtained and the reflection gradient response information of the spectral reflectivity change is extracted, so that the spatial difference of the earth surface reflection characteristics at different angles can be more comprehensively presented; according to the method, surface land type division is more accurate in gray continuity and spatial structure expression, so that potential fission boundaries are identified and reasonable section classification is carried out through quantitative comparison and spatial aggregation analysis of structure texture differences in boundary identification between different regions, and the identification accuracy of the boundary identification between different regions is improved. The stability and variability of the interior and boundary structures of the land areas are effectively described, and on this basis, the area boundary is reconstructed and combined with texture and reflection clustering grouping, so that the land type state and the distribution boundary of each area can be finely judged.

The invention discloses a metal tube surface defect detection method based on an improved RT-DETR (Reverse Transcription-DETR), which is used for carrying out multi-dimensional improvement on an RT-DETR model aiming at the problems that the surface defects of the metal tube are diversified in form, different in scale and complex in background. A backbone network of RT-DETR is modified by using a C3k2 module in a backbone network of yov11, a cylinder position coding module CPE is introduced to model a spatial structure of a cylindrical surface of a metal pipe, and a GLAF module integrated with local, medium and global attention mechanisms is adopted to perform multi-stage feature extraction. Secondly, in a feature fusion layer, a cylinder damage perception attention interaction module AIFIDML is used for replacing a standard Transform encoder, and the context perception ability of the model to defects is enhanced; and finally, in the neck network Neck, multi-level feature integration is optimized by adopting a multi-scale adaptive double-path feature fusion module MSADF, and the detection capability of tiny and large-scale defects is improved. According to the method, high-precision and high-efficiency end-to-end detection on the surface defects of the metal tube is realized.

The invention discloses an indoor scene image generation method and device, equipment and a storage medium. The method comprises the following steps: firstly, acquiring a to-be-processed image with furniture and information input by a user, and performing furniture elimination processing on the image to obtain an empty room image; and then generating a cue word for image generation based on a preset cue word template in combination with user input information. Meanwhile, extracting a depth map and a semantic segmentation map of the empty room image, and inputting the depth map and the semantic segmentation map into a pre-trained depth map control model and a pre-trained semantic segmentation control model for feature coding to obtain corresponding depth control features and semantic segmentation control features. And performing parameter adjustment on the large-scale graph generation model subjected to parameter fine adjustment by using the two types of control features to obtain a target graph generation model. And finally, inputting the empty room image and the cue word into the target image generation model to obtain a target image. According to the method, the spatial structure stability and the personalized design requirement of the user on the indoor layout can be considered.

The invention belongs to the technical field of image processing, and provides a facial paralysis grading method, system and equipment fused with multi-modal data and a medium in order to solve the problem that existing facial paralysis grading is inaccurate. Carrying out joint modeling on the handmade facial paralysis features based on prior knowledge and the depth visual features containing spatio-temporal information; extracting multi-scale features based on static image features and key point features extracted from a single-frame face image of a target individual, gradually transmitting small-scale features representing local asymmetry to medium-scale features and large-scale features, and adaptively performing feature fusion through dynamic weight distribution to obtain static symmetric features; and a bidirectional information interaction channel between the dynamic facial features and the static symmetric features is constructed, so that the generated fusion features simultaneously contain complete pathological information of spatial structure asymmetry and motion abnormality, and diagnosis grading is more accurate.

The invention discloses a point cloud compression method based on multi-dimensional feature fusion, electronic equipment and a medium, and aims to effectively reduce the compression bit rate while maintaining the precision of a space structure. According to the method, space, channel and topology redundant information is fused under a unified framework for the first time, and the method is suitable for efficient point cloud compression of various three-dimensional scenes such as automatic driving and virtual reality. Meanwhile, a local graph convolution Mama module is introduced, and the limitation that a point cloud topological structure is difficult to capture in a traditional method is effectively solved. Then, the octree nodes are divided into two dimensions of space and channel through a space-channel coupling grouping module, so that gradual feature fusion is realized, and the context information prediction precision is improved; experimental results show that the compression efficiency is remarkably improved on a plurality of radar and voxel human body model data sets.

The invention discloses an electric power facility intelligent operation and maintenance system and method based on multi-modal information fusion, and relates to the technical field of electric power facility operation and maintenance, an unmanned aerial vehicle is used for obtaining visible light and infrared thermal imaging images of an insulator chain, the collected images are preprocessed, then fine segmentation is carried out on the images, and single-modal features are extracted. And then, based on the extracted insulator sheet features and the spatial adjacency matrix, constructing an insulator graph fused with multi-modal information, and fully mining internal association and spatial structure information among the multi-modal features by using a pre-trained graph neural network. And finally, accurately identifying the defect condition of the insulator sheet through a classifier to obtain a defect identification result of the electric power facility. Thus, the advantages of visible light and infrared thermal imaging can be effectively fused, intelligence, automation and high efficiency in the operation and maintenance process of electric power facilities are achieved, and powerful technical support is provided for safe and stable operation of an electric power system.

The invention provides an exhibition hall roof adopting large-span folded-surface arch trusses and a construction method of the exhibition hall roof, a plurality of large-span folded-surface arch trusses are sequentially connected, and each group of supporting mechanisms are respectively arranged at two ends of each corresponding large-span folded-surface arch truss; the (n / 2 + 1) th truss and the (n / 2) th truss of adjacent large-span folded-surface arched trusses are connected through knuckle bearings, and the other two adjacent large-span folded-surface arched trusses are connected through connecting flanges; the large-span folded-surface arched truss comprises a main truss and two sets of wing trusses, and the two sets of wing trusses are hinged to the two opposite sides of the main truss respectively. The main truss is of an inverted triangular space truss structure; the wing truss is of a cantilever space structure. The exhibition hall roof is small in steel consumption in design, good in economic benefit, high in repeated utilization rate and high in industrial assembly degree.

The invention relates to the technical field of space structure design and parameter optimization, in particular to a neural network model-based reticulated shell structure node parameter automatic optimization method, which comprises the following steps of: acquiring an input-output data pair of a single-layer cylindrical reticulated shell constructed by aluminum alloy plate type nodes, the input parameters comprise design parameters and initial node parameters of the single-layer cylindrical reticulated shell. According to the method, node optimization data under different parameters are obtained through interaction of a genetic algorithm, finite element software and a programming tool, an example is supplemented to construct a training data set covering a common design parameter range of a project, and then a three-layer full-connection neural network model containing two hidden layers is trained; and the model precision is ensured by matching an early stop strategy and a learning rate attenuation strategy. During application, target single-layer cylindrical reticulated shell design parameters are input, the model can automatically output optimized node parameters, repeated iteration and manual intervention of a traditional method are not needed, optimization time consumption is greatly shortened, and the multi-working-condition batch optimization requirement is efficiently met.

The invention provides a dynamic planning method and system for a low-altitude route of an unmanned aerial vehicle, and relates to the technical field of intelligent planning, and the method comprises the steps: receiving a new delivery point instruction, obtaining a three-dimensional coordinate, carrying out the task integration of a new delivery point and an existing delivery point, defining a geographic range through three objects, namely a communication base station, a wireless tower and a permanent landmark building, and carrying out the task integration. Performing space structure division on the geographic range, and disassembling the geographic range into uniformly distributed three-dimensional voxel grids; and generating an adjustment value according to the coordinate interval of the three-dimensional voxel grid, including the delivery point density, the linear distance with the landmark building and other structural attributes, and adjusting the task integration result by using the adjustment value to obtain the adjusted task integration result of all delivery points. According to the invention, safe, efficient and accurate route planning of a multi-delivery-point dynamic newly-added scene in a complex low-altitude environment is realized.

The invention relates to the technical field of research and development management, and discloses a front-end code generation method and device based on a coding tool, equipment and a medium, and the method comprises the steps: extracting design manuscript element features from an original design manuscript, and constructing a graphical topological structure according to each design manuscript element feature; the graphical topological structure is converted into a structured description text through Markdown; and generating a target front-end code through the large language model and the structured description text. Through the above mode, the spatial structure understanding ability of the large language model is enhanced by extracting the design draft element features, constructing the graphical topological structure, capturing the spatial hierarchical relationship among the design elements and taking the graphical topological structure as intermediate representation, and the problem of layout deviation is solved; and the accuracy of transmitting the design intention to the large language model is improved. The method can be applied to the business fields of financial science and technology, medical health, old-age care and the like, and the reliability of generating the front-end code by the intelligent coding tool is improved.

The invention relates to a system for monitoring deformation and displacement of a pile foundation of an offshore converter platform based on a remote sensing technology. The system comprises a multi-source remote sensing data acquisition module, a pile foundation target identification module, a deformation inversion and three-dimensional modeling module, a deformation prediction module and an early warning output module. According to the method, a synthetic aperture radar (SAR), an optical remote sensing image and three-dimensional laser point cloud data are utilized to realize non-contact, periodic and high-precision monitoring of the offshore commutation platform pile foundation. A space-time diagram neural network is introduced to construct a space structure diagram for multiple pile foundation nodes, and trend prediction is performed in combination with time sequence deformation and displacement data, so that intelligent early warning of potential settlement, inclination or horizontal displacement of the pile foundation is realized. The system has the advantages of high automation, high reliability, no need of physical sensor arrangement and the like, and is suitable for long-term health monitoring and disaster early warning scenes of multiple types of pile foundation structures such as offshore wind plants, oil and gas platforms and the like.

The invention relates to the technical field of bridge pier load monitoring, in particular to a bridge pier load monitoring method and system in bridge construction, and the method comprises the following steps: based on a bridge pier foundation structure, building the correlation among multiple nodes through calculating the distance among the nodes, the physical characteristics and the mechanical characteristics; according to the method, the bridge pier space structure diagram is established, node load characteristics are updated through the diagram convolution operation, the monitoring precision of bridge pier bearing load changes is remarkably improved, combined modeling of time sequence data and space characteristic data is combined, the space-time law of the load changes can be recognized, the construction scheme is optimized, and the construction efficiency is improved. The layout and configuration of construction equipment are dynamically adjusted, the dynamic adjustment of the adaptive convolution kernel optimizes feature aggregation according to the features of the support structure, the response capability to load change is improved, the safety and high efficiency in the construction process are ensured, the method provides accurate real-time monitoring and dynamic optimization, and the construction efficiency is improved. And monitoring delay and errors in a traditional method are avoided.

The invention discloses a building operation and maintenance problem reasoning method based on a heterogeneous graph neural network. The method comprises the following steps: extracting BIM model space and component information; constructing a heterogeneous network graph; and model training and problem reasoning. Building space units are extracted from the building information model in the IFC format, space nodes are generated, components serving all the spaces are recognized, and component nodes are generated. And based on the space-component service relationship and the space-space adjacency relationship, establishing a heterogeneous network diagram containing multiple types of nodes and multiple edge relationships so as to comprehensively represent the space structure and function dependence in the building. A heterogeneous graph neural network model is used for training, and cross-space and cross-component feature aggregation and information reasoning are realized under a relation-aware message propagation mechanism. And applying a reasoning result to a building operation and maintenance stage to realize space anomaly detection and component function state diagnosis. According to the method, fusion modeling of spatial information and component information is realized, and the problem positioning precision and response efficiency in building operation and maintenance are improved.